THE ART OF MAKING SENSE
A Primer on
Reason and Critical Thinking
Table of Contents
Part One: The Word
Chapter 1 - Logic Isn't Everything-But It Helps
Chapter 2- Words and Ambiguity
Chapter 3- Define Your Terms
Chapter 4- What Kind of Language Are You Using?
Part Two: The Argument
Chapter 5- How Not to Argue
Chapter 6 - Putting Up a Logical argument
Chapter 7 - Some Patterns of Reasoning
Part Three: Truth and Falsity
Chapter 8- Truth and Evidence
Chapter 9- Knowing the Causes of Things
Chapter 10- Are All Generalizations False?
Chapter 11- On Matters of Taste and Opinion
Glossary
PART ONE
The Word
CHAPTER 1
Logic Isn't Everything-But It Helps
Let us begin with a few disclaimers. Study of logic is not going to be a
panacea. It won't make you a better lover, it won't improve your digestion, nor
will it be much comfort when you wreck your car or lose your job. But as an
effective weapon in what James Thurber called "the end-less war between
meaning and gobbledegook," or as a guide when you lose your way in the
morass of doubletalk, conflicting claims, and verbal diarrhea that meets us
almost daily in our society, logic is unbeatable.
What is logic and what isn't it?
Logic is correct reasoning. More precisely, logic is the process by which
statements are supported with adequate proof by being tested against the right
amount and kind of evidence, the process by which knowledge is rendered
reliable-in short, the "science of proof." It follows then, that to
be logical is to argue reasonably. What the logical man insists on is simply
this: if you claim that you have proved a point-about anything at all-then your
conclusion must be examined in terms of the adequacy and reliability of your
evidence. Logic shows us what kinds of tests to apply in each case and how to
apply them. In other words, if you claim that Jones is a thief, or that blondes
are inferior in intelligence to redheads, or that the testing of nuclear
weapons leads to destruction of the environment, logic provides ways for
judging the "truth" of your assertion and tile reliability of your
evidence.
Logic encourages taking a somewhat skeptical attitude toward what-ever you
cannot verify directly by your own experience and in this regard it is akin to
the skepticism implied in phrases ranging from the man from Missouri who
challenges "Show me," to the simple "Oh yeah?" Obviously
this approach is not useful in all situations and the logician acknowledges
that many statements cannot be or do not need to be approached logically. In
fact, all human discourse can be divided into two categories the logical and
the nonlogical When someone says "I'm getting fat. I must eat less,"
he is using a well known form of argument (a unit of discourse that purports to
prove something) called a syllogism lie is claiming a relationship between his
weight and the 'amount of food he eats Such a claim can be subjected to certain
tests, proof can be called for and examined, and an objective judgment made
about its truth. Note that we are not using argument in its popular sense, that
is, a dispute or disagreement. Rather, "argument" in the language of
logic is discourse that contains at least two asserted statements and the claim
that one statement ought to be believed because another is true. Logic is
helpful then in handling certain types of information and resolving certain
kinds of questions. It is appropriate when we deal with patterns of proof or
disproof, claims and counterclaims.
On the other hand, statements like "I like to travel" or "I love
you" are ordinarily regarded as nonargumentative. There can be no useful
debate, in other words, about whether you like to travel the proof such as it
is, is solely in your own head. Such statements are regarded by the logician as
"nonlogical" and do not require supporting evidence. (Note that
"nonlogical" does not mean illogical; an illogical statement is one
which violates the rules of sound reasoning-it's like adding one and one
incorrectly.)
Do we really want or need logic?
Man used logic long before he was defined by Aristotle as "the reasoning
animal." He has continued to develop and refine the rules of logic not as
an intellectual exercise, but as a means of getting at the truth. It must be
admitted at this point, however, that logic is often disregarded when it could
be most helpful in conducting man's business. There are a number of possible
causes for this disinclination to make logic part of our lives. In the first
ace, many persons think of logic as a branch of philosophy, which everyone
knows is an academic pursuit far removed from life and living. According to
this point of view, there is no more reason for the common man to study logic
than to study abstract art or nuclear physics. In an age when swamis,
parapsychologists, back-to-nature enthusiasts, sensitivity group followers, and
mystics of various stripes advocate their respective paths to
"knowledge," it is not surprising that logic, which requires
discipline and eschews emotion, should be brushed aside or ignored. Nor does it
help the case for logic to note that logical thinking can lead to unexpected
and uncomfortable ends. By following the rules of logic we are often led to
unpleasant conclusions, and are sometimes confronted with our own prejudices
and opinions masking as "facts."
What are the advantages of using logic?
The point to remember is that logic is not a straitjacket or a religion. It is
a tool. Without logic you would have available to you no knowledge other than
your own experiences. Most of what you have learned or been taught was compiled
through the process of logic. No one person can personally verify everything:
the number of employed women college graduates in the United States, the
correlation between crime and poverty, the major parties' popular and electoral
votes in the last presidential election. Thus, when a sociology textbook says,
"The lessons of psychology induce a change in people's behavior when they
grasp them," the author is basing his conclusion on data, fact, and
conclusions that other people supplied him. And you, in turn, use his
conclusions to arrive at your own.
Since logic gives you a method for assessing the truth of many of the
statements that other people make to you (whether in print, in person, or oil
television or radio), it is perhaps the most relevant study-and the most basic
one-that you will ever engage in. It is eminently practical in that its whole
focus is on solving problems and arriving at conclusions. In helping you
distinguish between facts and inferences, it will enable you to arrive at a
sound basis for judgment and (if necessary) for action.
We are inundated daily by claims of all kinds, many of them contradictory.
Opposing politicians state conflicting "facts"; advertisers make
soaring claims for their products. Some claims are insidious in their subtle
disregard of logic. Consider the following letter, recently published in a
California newspaper:
Editor:
I wonder how many people, especially women, realize that if the Equal Rights
Amendment is put into law the American woman may look forward to absolute
horror While Russia and Red China favor treating their women as animals I
shudder to think of what will happen to our women Frankly I prefer our girls to
stay as sweet as they are.
A careful reader (and thinker) will ask himself how such words as "horror,"
"animals," and "sweet" are used he will determine what
claim or argument is being asserted by the writer of the letter and he will
determine how this claim or argument can be proven or verified. To be able to
distinguish fact from fiction is more than an abstract exercise. It can
determine the products you purchase, the candidates you elect, even the
safeguarding of your rights as a citizen. Learning more about logic will give
you the tool you need.
How can we become more logical?
We are all capable of being reasonable and logical. But this does not mean that
we are always reasonable, or that our reasoning is always correct. All of us
are guilty of some common bad habits of thinking; by identifying these habits
and examining their causes we can tackle our problems more methodically and
learn to test our conclusions by critical standards and rules.
A starting point would be to eliminate the personal equation from our thinking
when our thinking should be concerned with facts. The anthropologist Franz Boas,
in his Mind of Primitive Man, tells us that primitive man has the same kind of
mind that we have, except that he is more likely to be influenced by emotion.
Civilization, Boas says, does not improve the mind, but it decreases emotional
association with ideas and thus helps us to think more clearly. If we wish to
make sense we must try to eliminate emotion when emotion is irrelevant. Emotion
is out of place when we are considering information contained in the
multiplication table, and it is equally irrelevant in assessing a problem of
legal evidence. Was Ezra Pound a traitor? If we say "No, because I like
his writing," or "Yes, because I can't stand his poetry," we are
evading the fundamental duty of a rational person, which is to study the
evidence.
Another aid to clear thinking is the avoidance of too much rationalization. Our
reasons should lead to our conclusions rather than serve later as an excuse for
our conclusions. And we should be very clear, at least in our ova minds, as to
which came first. We should be honest with ourselves in identifying the reasons
for voting for a particular candidate, purchasing a certain model car, or
selecting a mate.
It is hard to exercise our critical powers in matters that involve our emotions
and self-interest. When asked to do so, we have a tendency to become dogmatic
and to make positive or arrogant assertions that cannot be proved. We may
become blind fanatics and stop thinking altogether, automatically regarding
others' ideas as outlandish or perverse or dangerous. But instead of rejecting
evidence that threatens to disprove our ideas, we should welcome it as an
opportunity to increase our knowledge.
All too often we rely on the word of "authorities" for guidance
without ever inquiring into the validity of their evidence. We should heed
Bertrand Russell's observation concerning the role of expert opinion:
"When the experts are agreed, the opposite opinion cannot be held to be
certain; when the experts are not agreed, then no opinion can be held to be
certain." In other words, let us respect the experts, but let us not
follow them blindly.
While it is hard for human nature to sustain the tension of never having
anything settled with finality or certainty, it is the safest and most
realistic course to reserve final judgment on almost everything. The most
reasonable attitude is that of the scientist who says, "Let us test our
beliefs by the evidence, showing a willingness to revise these beliefs as the
evidence changes, never claiming finality for our beliefs, but recognizing that
the probabilities are sometimes so overwhelming that we can occasionally
identify some truths. Thus we should be ready to believe when there is
sufficient evidence, but we should suspend judgment when evidence is
lacking."
We might start our study of logic by practicing the rules that Rene Descartes,
the French mathematician and philosopher, used to guide his thinking:
1. Never accept anything as true which you do not clearly know to such; that
is, avoid hasty judgments and prejudice.
2. Divide each difficulty under examination into as many parts as possible, or
into as many as necessary for the solution of the problem.
3. Begin with the things that are simplest and easiest to understand, and then
ascend to knowledge of the more complex.
4. Make enumerations so complete, and reviews so comprehensive, that you may be
reasonably assured that nothing is omitted.
Now, whatever we may think of the usefulness of these simple precepts for
attacking a problem, the assumption that underlies them is encouraging. We are
all capable of understanding, and we can improve our understanding by using the
right methods of thinking. These methods will help us solve our problems more
efficiently. Our problems may be "scientific" ones, in the narrow
sense of that term, but they may also arise in business and social relations in
politics and in love
How this book will help
A few words now concerning what you are going to find in this book. The art of
making sense involves an understanding of semantics and scientific methods of
thinking in addition to the analysis of reasoning or logic. The good thinker
will make a threefold analysis of a discussion. He will interest himself in the
meanings of the words used, he will look for the "argument" in what
he reads or hears; and he will ask himself whether what he hears and reads is
true or false, and how it can be verified. Each of these steps represents a
question to be asked, a problem to be solved.
To illustrate the manner in which these questions arise, let us examine an "argument,"
that is, a unit of discourse that purports to prove something: "Polygraph
(lie-detector) tests are not twentieth-century witch-craft. After MI, they are
used by many good police departments, and the results are accepted as evidence
in some courts Let us put this argument to the test of our three questions
1. The semantical question: What do the words mean? For example, what do you
understand by twentieth century witchcraft? What is a "good police
department"? How many are "many"? What is meant by some courts?
The principles which will help us clarify questions of this kind are discussed
in chapters 1 through 4.
2. Let us consider the argument as a whole. What is the author trying to prove?
What evidence does he give? Do you think his argument is logical, that is, does
his conclusion follow from his facts and assumptions? Does your agreement or
disagreement with his assumptions have anything to do with the
"logic" of the argument? A discussion of questions of this kind will
be found in chapters 5 through 7.
3. Finally, another type of question: Is it true that many good police
departments use polygraph tests? Is it true that the results of such tests are
used in some courts? what makes a police department "good"? If you
have beliefs on these matters, and your beliefs are challenged, do you know how
to support your position? The concluding chapters of this book deal with such
questions.
One thing at a time, however. Let us begin our examination with a basic factor
in the human communication process. Charles Lamb said that logic is nothing
more than a knowledge of words, and it is with words that we shall begin.
FOR DISCUSSION AND WRITING
1. This chapter points out that logicians divide all discourse into the
categories of "logical" and "nonlogical." Using these
terms, identify the following statements:
a. The senior senator from California is bald.
b. I was born in East St. Louis, Illinois.
c. I have a stomachache.
d. She said she has a stomachache.
e. A rose is a rose is a rose.
f. According to the Bureau of tile Census, 16 percent of all black males over
twenty-five years of age with a college education earned over $15,000 in 1971.
g. Help!
h. A bird in the hand is worth two in the bush.
i. Please open the door.
j. The newspaper story described the fuel shortage facing this country.
2. Using a dictionary when necessary, define each of the following terms:
hypothesis, inference, argument, guess, intuition, hunch, theory.
3. As you consider the history of man and his actions, does it appear that he
is a logical creature? What evidence can you suggest to support your view?
4. This chapter suggests three questions the careful thinker asks when
analyzing an argument. Apply those questions to the following passages:
a. We must not lose faith in man's future. It was faith that spurred on the
pioneers to settle this country It was faith that made this nation great. That
same faith led man to scientific discoveries and inventions And that same faith
will help him overcome the problems of pollution, crime, the depletion of
natural resources and the threat of nuclear war.
b. When men are spoken of as kings and subjects or when government is mentioned
under the distinct and combined heads of monarchy aristocracy, and democracy
what is it that reasoning man is to under stand by the terms? If there really
existed in the world two or more distinct and separate elements of human power,
we should then see the several origins to which those terms would descriptively
apply, but as there is but one species of man there can be but one element of
human power and that element is man himself. Monarchy, aristocracy, and
democracy are but creatures of imagination, and a thousand such may be
contrived as well as three. (From The' Rights of Man, by Thomas Paine)
5. What are some everyday problems that could profit from the application of
Descartes' four rules? Do they apply chiefly to scientific problems, or to
problems in sociology, politics, and psychology as well?
6. Within one week two newspapers published conflicting accounts of workers'
attitudes toward their jobs. One article reported that "a representative
sampling of adult Americans from coast to coast" indicated that most
people like their jobs and do not find them boring. The second article gave
summaries of interviews with a sampling of production-plant workers and
concluded that most of these workers find that their jobs are "like being
in prison." How would you go about finding the truth concerning the
accuracy of these stories? How would you determine which article more nearly
correctly reports the attitude of most workers in this country?
7. Can you think of any decisions you have made recently that were made
primarily on the basis of reason and logic? In general, how often is logic the
basis of your most important decisions?
CHAPTER 2
Words and Ambiguity
Man has always regarded words with reverence and mystery. In primitive
societies a belief persisted that a person or thing had one right or true name,
and knowledge of that name gave power over the person or thing. In the Gospel
of John we are told that, "In the beginning was the Word, and the Word was
with God, and the Word was God." Because of this tendency to invest words
with a magical quality, man often forgets an important truth: words have little
importance on their own account, or for their own sake. Their importance
derives from their meanings, and their function is to act as signs or symbols
of something outside themselves. Perhaps this is what the English philosopher
John Locke meant when he said, "We should have a great many fewer disputes
in the world if words were taken for what they are, the signs of our ideas
only, and not for things themselves."
In this chapter we shall examine some fallacies and myths about the
relationship words have to the things they stand for. We shall also consider
some of the difficulties and ambiguities that arise-whether in writing a theme
for class or speaking informally to friends-when we are care-less in choosing
our words. Such an examination will begin by reviewing the principles of
semantics: the study of words (and symbols generally) in relation to their
meanings.
Semantics: The study of "the meaning of meaning"
In 1933 Alfred Korzybski, a Polish-American scholar and mathematician,
published Science and Sanity, a difficult and provocative study which presented
a method by which people might think and speak more clearly through an analysis
of language habits. By analyzing how we use words and how we react to words,
semantics (as Korzybski's system is called) has come to designate the study of
words as signs or symbols, that is, as things that usually stand for something
other than themselves. An-other way of describing semantics is to say that it
is the study of "the meaning of meaning."
Although the study of semantics has received a great deal of attention in
recent years from philosophers, logicians, scientists, and teachers of English,
its subject is not a new one. A recognition of the importance of a proper
understanding of language for clear thinking is as old as philosophy itself,
going back to Socrates in ancient Greece. Philosophy, for Socrates, was the
"pursuit of meanings," and he sought adequate definitions of words
like "justice," "good," "right," and "wrong."
He realized that unless there is agreement on the meaning of the words we use,
communication is impossible.
Since the time of Socrates a concern with meaning and language has been a
constant preoccupation of philosophers and logicians, from Aristotle to
Bertrand Russell and Noam Chomsky. Today semantics covers a vast and complex
field of investigation, with widely diverse branches of study, and it often
employs the methods of anthropology, linguistics, logic, psychology, and many
other disciplines. But we shall not be concerned with a systematic
investigation of semantics. Nor do we share the hopes of some semanticists that
this subject offers a panacea for all the world's ills. We are interested in
some practical applications of the subject: to show how an understanding of
semantical principles concerning the uses and functions of words may aid us in
thinking and writing more clearly) and thus in making sense. A sound
understanding of language, the instrument of communication, is a means to this
end.
No one really knows how much thinking human beings could do without language,
but it is undeniable that our thinking would be very limited. Human
intelligence is based on our ability to think and talk about things that are
not in our actual surroundings. We use words to "point to" these Things.
Though animals can think, they have no words, and so their intelligence is
limited. We can think and talk about Mary even though we don't see her. But
suppose we mention the name "Mary" to a dog, and Mary is his
mistress. The dog will react to the familiar sound, and will wag his tail in
eager anticipation of seeing Mary. The dog reacts to the name much as he does
to Mary. But human beings can think and talk about Mary, knowing that she is
absent. In fact, knowing that Mary is absent makes it a good deal easier to
talk about her. A girl can pick at a daisy and say, "He loves me, he loves
me not," but no animal can do that.
Language is thus indispensable to human thinking, but this does not mean that
mental ability is the same thing as having a large vocabulary. A limited
vocabulary, it is true, restricts the range of our thinking. And this may lead
us, mistakenly, to think of people as being unintelligent merely because
certain words are unfamiliar to them. Because vocabulary is determined largely
by cultural and socioeconomic factors, there is a tendency today to regard
intelligence tests that rely heavily on verbal skills as unreliable means of
determining native intelligence. The children in an underprivileged
neighborhood, for example, did very poorly in an intelligence test which
contained questions such as this. "A hand is to an arm as a foot is to a
___." Only a few children filled the blank with "leg," and most
of them were scored low in intelligence. But later it was learned that the expression
"is to" was unfamiliar to these children. They would have said
"goes with," that is, a hand "goes with" an arm, and when
"goes with" was substituted for "is to" in the same test,
they did very well, and scored high in "intelligence."
Language also influences the content of our thoughts. The phrasing of a
question will influence our thinking about the subject matter. It is well to
remember this point when we evaluate the results of public opinion polls. It
makes a difference whether one is asked: "Do you favor clemency for those
American men who obeyed their consciences and refused to participate in what
they believed to be an illegal, immoral war?" or "Are you opposed to
letting draft dodgers and deserters go unpunished?" The phrasing of a
question may convey an emotional tone, and most of us are suggestible, so that
we have a tendency to agree with what the questioner seems to expect of us.
Consider how we are apt to respond to the question: "How do you account
for the fact that the great majority of American people disagree with the
motives and actions of the men who refused to serve in the Vietnam war?"
We may fail to stop short at this point to ask the question: "Is it really
true that the great majority of Americans do disagree with the motives and actions
of the men who refused to serve in the Vietnam war?"
Semantics, then, is concerned with language insofar as language is relevant to
problems of thinking and communication. We shall now examine certain
misunderstandings concerning words, and some of the fallacies and myths that
cluster around the relationship that words have to the things they stand for.
Three semantical errors (or, a warning to the unwary student)
Words are not mysterious things. They are events in space and time; that is,
they have a physical dimension, and they have meanings. Those meanings,
however, are arbitrary. Further, words are merely symbolic sounds, and they do
not possess inherent qualities or defects. Let us now examine the consequences
of these facts.
1. The word and the thing
When we say that words have meanings, we say that human beings agree that a
certain word, like "nylon," for example, shall refer to a certain
kind of material. This material could have been called by any other name, but
the Du Pont Company christened it "nylon." So now, when I think of
this material, and say "nylon," the sound comes to your brain, and
your mind is referred to the material I am thinking of, namely, the fabric used
in women's hosiery, etc.
Now, this relationship of the word to the thing is an "arbitrary"
relationship, in the sense that the word could have been any other word. There
may be aesthetic or other reasons for choosing an "appropriate" name,
but these considerations are never compelling. This principle, that the
relationship of words to things is an arbitrary one, may seem absurdly simple
to you, but it is a basic principle of semantics, and, though simple, it is
unknown to some, and forgotten by others. Children, for example, are apt to be
unfamiliar with the principle. When a French semanticist asked a child whether
the moon could have been called "the sun" and the sun
"moon," the child said, "No, because the sun makes it warm and
the moon gives light." Another little boy once asked his mother:
"Mother, when I was born, how did you know that I was really Charlie, and
not some other little boy?" Children often believe that a word is
necessarily connected with a thing, so that it would be impossible to call the
thing by any name other than the one by which it is known. Now, of course, after
words come to be associated with specific things, a connection is established,
and we would create confusion if we did not use these words in their customary
meanings. But the point is that words, when they first come into being, can be
anything at all.
And here is another amusing example of the same type of error. When the planet
Pluto was discovered in 1930, the story goes, a young lady was reported to have
asked an astronomer: "Professor, when you astronomers discovered the new
planet, how did you know that this planet was really Pluto, and not some other
planet?" Now this new planet, of course, could have been called by an'
other name, even Mickey Mouse, though in some ways that name would not have
been so appropriate, for it is customary to name the planets after ancient
Roman gods.
The kind of error we have just noted was the subject of an amusing bit of
spoofing by Mark Twain. Somewhere in his writings he discussed chapter 2,
verses 19 and 20, of the Book of Genesis, which deal with a semantical matter:
"And out of the ground the Lord God formed every beast of the field, and
every fowl of the air. And brought them unto Adam to see what he would call
them, and whatsoever Adam called every living creature, that was the name
thereof. And Adam gave names to all the cattle and to the fowl of the air, and
to every beast of the field." The semantics of this account is
unexceptionable, but Mark Twain, in his little jest, imagines that these
animals pass in review before Adam. He gets along fine, until he is finally
baffled by one animal. He can't think of a name for it. In desperation (like a
man) he turns to his wife for help. "Eve," he asks, "what name
shall I give this animal?" Without a moment 5 hesitation, Eve answers,
"Call it a horse." "But why a horse?" Adam asks her.
"Well," says Eve, "it looks like a horse, doesn't it?"
We have been dealing with rather obvious examples. But the same kind of error
also occurs on subtler levels than those just considered. For example, the
system of government in the United States is commonly referred to as a
democracy. There are some people, however, who argue that it is wholly improper
to call our system of government a democracy. Sonic have gone so far as to
demand that a law be passed forbidding the use of the word in this sense. These
people argue that we are a republic, that we have a republican rather than a
democratic form of government. The dictionary defines a republic as a system in
which sovereignty resides in the people and in which legislative and administrative
functions are carried on by elected representatives of the people.
"Democracy," the argument goes, "means the rule of the people
directly, and not through representatives," as in a small town where ever'
voter has his say in a town meeting.
Now, there is no question that we are a republic, but we may also be a
democracy as that word is not used. If the people of the United States,
England, and France wish to refer to their systems as democracies, and define a
democracy as "a system of representative government based on the
principles of freedom, and legal and political equality," there is no one
who can or should prevent the people from doing so. In other words, the word
"democracy" has been broadened in its meaning by usage, and today it
usually refers to a system of government in which the people elect
representatives in regularly scheduled free elections, and in which there is a
basic devotion to freedom and political equality. This at least is one meaning
that democracy has acquired, and to argue that it is wrong to use the word in
thismanner represents a failure to acknowledge the arbitrary relationship of
words and things. Language is like a living thing in its growth and
development, but its life depends on human decisions.
The uses and alleged misuses of the word "democracy" may appear to
involve only a theoretical problem of semantics. The heat engendered by this
apparently trivial matter, however, indicates that other issues may be
involved. These other issues are not far to seek. The issue is the age-old one
of conservatism versus liberalism. To call the United States a democracy is to
emphasize our ideals of freedom and equality, in addition to our representative
system. Conservatives, who believe that an emphasis on equality will work
against the public good, prefer the word "republic," which does not
suggest equality. Many liberals, who want more equality, prefer
"democracy."
Words, then, can have as many uses as people give them. As language grows and
develops, it becomes permissible to use new names for things, or to use old
names in new ways. Thus, the word "surgeon," from Greek roots meaning
"one who works with his hands," once meant a laborer; today it means
one who operates on a living organism. The word "doctor" today means
a practitioner of medicine and surgery, among other things; but originally it
meant a teacher, especially one of great learning. These are examples of old
words that have acquired new senses. It would sound rather odd if we said that
medical men were not really doctors, on the ground that "doctor"
really means a teacher of great learning.
New names are invented not only for new things, like new drugs and synthetics
(Streptomycin, Orlon) and space-travelers (astronauts), but we also give old
things new names. Military invasions are called "interdictions," and
bombing raids "protective reactions." Inflammable materials are now
called "flammable," and a question concerning the meaning of a word
is now called a "semantical question."
The principle that words are arbitrarily associated with things should of
course not be abused. If we desire successful communication we should not
capriciously assign new meanings to old words. The use of words in their
customary senses also enables readers or hearers to devote more attention to
the thought and less to the vocabulary. That famous character known as Humpty
Dumpty, however, was unconcerned with whether anyone understood him or not, and
accordingly he was free to abuse the principle, as evidenced by the following
colloquy:
Humpty Dumpty said ...:"There's glory for you."
"I don't know what you mean by glory," Alice said.
Humpty Dumpty smiled contemptuously. "Of course you don't-till I tell you.
I meant, 'There's a nice knock-down argument for you.
"But 'glory' doesn't mean a 'nice knock-down argument,"' Alice
objected.
"When I use a word," Humpty Dumpty said in rather a scornful tone,
"it means just what I choose it to mean-neither more nor less."
(Lewis Carroll, Through the Looking Glass.)
2. The word and magic
We have been discussing the failure to take note of the arbitrary relationship
between words and things. We shall now examine a second type of error which
arises from the failure to recognize the fact that words are merely symbolic
sounds. This error is the belief in the magical power of words. It is the
practice of certain primitive tribesmen to change their names after being
cursed, so that they may escape the evil which has become attached to their
names, and thus to themselves. Another example is the case of the benighted
primitive who, when cursed by a fellow tribesman, dropped flat on the ground so
that the words would fly harmlessly over his head. And do you remember the
story in the Arabian Nights, about Ah Baba and the Forty Thieves? Ah said,
"Open, Sesame," and lo! the cave door opened. These examples indicate
the nature of the belief in the magical power of words: words have potencies to
do things all by themselves; a name can become infected with evil; it can harm
a man if it actually strikes him, and it can open cave doors.
Are these superstitious beliefs in the magic of words confined to primitive
man? I am sure that you are not guilty of similar superstitions, but how about
your friends? And, candidly, aren't we all, just a little bit? Don't we all
know people who say, "Speak of the devil and he's sure to appear, and who
really believe that there way be something in that expression? Don't we go to
the racetrack and find a horse whose name has struck our fancy, and bet on that
horse regardless of the form charts? And if this has happened to you, and your
horse won, weren't you just a little bit persuaded that there really is some
kind of magic in a name? And why do we say, "Knock on wood," when we
express a thought concerning our good fortune? Well, obviously, because we
assume that there are forces in the universe which don't like to hear people
talk about their good luck. Again, the belief in the magic of words.
Have you ever watched a group of men engaged in a form of wagering known as
"shooting craps"? It is a highly instructive area of investigation
for the student of language. One of the players is hoping that he will throw a
seven or eleven. He pleads with the dice, and informs them that his very young
infant is in need of protective covering for its little feet. But alas! The
dice roll regardless of his words. For words have no magical powers.
Or consider the radio broadcasting or the telecasting of baseball games. Many
of the listeners and television viewers believe that words have magical powers.
If the home team's pitcher is pitching a no-hit game, these believers in magic
regard it as a terrible crime for the announcer to mention this fact, for if he
does, then the spell will be broken, and the next batter will be sure to make a
hit. Because of this superstition, which many announcers are afraid to
challenge, millions of listeners are cheated, for they are denied knowledge of
the dramatic intensity on the baseball field.
3. Words, truth, and beauty
A third error is the assumption that words give us guarantees concerning
things. This error usually occurs in this way: we assume that a fine-sounding
name proves the fine quality of the thing referred to. This is a frequent
source of deception. An organization may call itself "The People's
Committee for Peace," or some such name, because the words
"people's" and "peace" sound trustworthy and good. In the
same way, a group seeking to foster race prejudice may use words like
"Christian" or "fellowship in its name in order to convey the
impression that it upholds the principles of religion and love for one's fellow
beings. We all know of the many fancy names which used to be common in the fur
business, such as Hudson Seal for muskrat fur. The Federal Trade Commission may
be going too far, however, in forbidding the expression "imitation
fur" for materials that are not made of fur, on the ground that the word
"fur" cannot be used for materials not made of fur! This reminds one
of a famous cartoon showing a policeman beating a hapless-looking individual at
a Communist demonstration. The man: "But, officer, I'm an
"anti-Communist." The policeman: "I don't care what kind of
Communist you are!" and continued to whack.
One must guard against these sources of deception, but there are also many
harmless forms of this sort of thing which involve willing self-deception. Thus
on transatlantic liners, many years ago, third class" was changed to
"tourist class" because third class sounded too inferior. In the
Soviet Union, we may note, the name "third class" was changed to
"third category," for there can be no class distinctions in a
communist society! The accommodations are the same, objectively, but the change
of name may have a pleasing psychological effect, for it seems :0 eliminate a
stigma.
We also often assume that evil-sounding names imply that the thing is evil. The
implication does not follow. But again, we must emphasize the fact that
"brutal" and ugly names may have undesirable psychological
consequences, and this is particularly unfortunate when such names are not
strictly accurate in their descriptive aspects. For ample, the expression
"home for incurables." This name was based n the assumption that
certain diseases are actually incurable, but this is a big assumption, and the
present tendency is to change the name to institution for chronic
diseases." Not only is such a name truer to the facts, but it also gives
the patients more hope, and hope may have beneficial psychosomatic
consequences. Furthermore, who knows but that science may some day find cures
for such diseases? In other words, let us not call a spade a spade even when it
isn't one.
Names, then, guarantee nothing in themselves. As Shakespeare's Juliet remarked:
What's in a name? That which we call a rose
By any other name would smell as sweet.
Words and etymologies
So much for three semantical errors. One of the basic points of this chapter is
that the meanings of words are based on arbitrary human choices. An
understanding of this point will also help to clarify the semantic relevance of
the etymologies of words. The word "etymology" refers to the study of
the history of words, to their derivations from their roots, with all their
changes of form, spelling, and meaning. Etymology describes the manner in which
words came to acquire their present meanings. For example, the word
"philosophy" is based on two Greek roots, philein and sophia.
Philein means "the love of," and sophia means "knowledge"
or wisdom," so that the etymological derivation of the word
"philosophy" indicates that it means the "love of wisdom."
"Philanthropist" is based on philein and anthropos,
meaning "man," so that a philanthropist means a "lover of
mankind." "Sophomore" comes from sophos and moros,
the latter root meaning "fool." A sophomore is a wise fool.
Other examples could be multiplied endlessly. The word "assassin" is
based on the Arabic word hashahin, which is derived from "hashish,"
an intoxicating drug found in the Middle East. The first assassins were
hashish-addictive eleventh century Syrians and Persians who were members of
secret societies which murdered their political enemies. "Planet"
comes from a Greek word meaning "wanderer," for planets change their
positions among the fixed stars.
But the important point for semantics is this: etymologies enlighten us
concerning meanings, and sometimes give precision to words whose meanings are
somewhat vague to us. The etymology of "definition," for example.
This word comes from the Latin roots de (off) and finis (end, limit, boundary).
A definition, then delimits the meaning of a word. The study of etymologies
will help us in using the right word to express a precise shade of meaning. But
etymologies do not control the use of language. No matter how a word may have
originated, it means today what people use it for. Custom is king in matters of
language, and if human beings customarily use words in new senses, it cannot be
said that they are wrong in doing so, for words are noises arbitrarily
associated with things. An assassin today means one who commits murder because
of fanaticism or for a reward, not a person who smokes hashish. And etymologies
may also be misleading. The word "etymology" itself, for example. The
word is based on etymon, meaning "the true sense", plus logos, or
"word," but there are no "true" or "false" senses
of words. There are only customary or uncustomary senses. And so, though
etymologies illuminate the meaning of many words-like
"philanthropist," and "planet"-they do not establish the
"real" meaning of the word. Words mean what people intend that they
shall mean. We are the masters of the words, not the words of us.
Words are wonderful engines of communication, but we must know what they mean,
and how to handle them. And we must guard against being "taken in" by
them. As that wise old English philosopher Thomas Hobbes wrote: 'Words are wise
men's counters-they do but reckon with them, but they are the money of
fools."
Ambiguity: Or, I wonder what he meant by that?
Words, we have seen, can have any meaning we assign to them. This
characteristic, called "ambiguity," is responsible for many
unnecessary disagreements. We shall now examine the ways in which ambiguity is
a major cause of failures of communication, whether in writing or in
conversation.
Let us imagine that we are listening to a conversation between two men, Bill
and Jim, at one of those informal debating societies known S "cocktail
parties." They are arguing a frequently debated topic, the principle
expressed in our Declaration of Independence, that all men are created equal.
Bill has the floor:
"Jim, I tell you that men aren't equal. Don't let anyone tell you hey are.
They don't know what they are talking about. Use your own yes! Do you see the
equality of mankind? Do you find in your own experience that people have equal
abilities, or equal characters? Do you find that they are equal in any respect
whatsoever? Everyone is different from everyone else. In my opinion Thomas
Jefferson uttered preposterous nonsense when he said, and I quote, 'We hold
these truths to be self-evident, that all men are created equal.' This
so-called truth is not self-evident to me, so it can't be self-evident. In my
opinion this so-called truth is actually a falsehood."
Let us now hear from Jim: "Just a minute, Bill. You are the one who
doesn't know what he is talking about. Men are equal, and I agree completely
with Jefferson. The equality of mankind is the foundation our democracy. No man
has the right to think of himself as better
than any other, or as entitled to special privileges which others aren't
entitled to. Every person is entitled to equal opportunities, and no one should
suffer discrimination because of his race, color, or creed. This is he basis of
our legal system, which tells us that all men are equal before the law. A legal
decision should not depend on the color of a man's ,in. Do you deny that? Are
you in favor of racial and religious discrimination?"
And so on. Let us be merciful, and tune out Bill and Jim at this point, though
they are probably still arguing, unless they have already ached the point of
mutual exhaustion. Now, the argument we have just overhead can never have an
ending, because it was really not an argument at all, but just two fellows
talking at cross-purposes. In order to have a genuine argument there must first
of all be an agreement or meeting of minds about the issue in dispute; that is,
there must be a common understanding of the question to be answered. But there
was no such agreement between Bill and Jim. They were talking about different
things, and so their minds never really met. They were engaged in what we shall
call a "verbal dispute."
A verbal dispute is one in which the two speakers engage in what merely looks
like an argument, but really is not, because the speakers o not understand each
other. The reason why they do not understand each other is that they are using
a key word in two different senses. The key word in this ease was
"equal." In other words, there can be no argument concerning whether
or not men are equal, if Bill means one thing by "equal" and Jim
means something else. Let us recall what they were saying.
Bill said that men were not equal. By "equal" he meant having the
same size, shape, mental and physical powers, talents, and so on. Jim said that
men were equal. By "equal" he meant that all men should be given the
same opportunities, and that they should have the same chance of getting
justice in a court of law. Bill's mind and Jim's mind did not meet, for they
were thinking about different things. Though each used the same word,
"equal," they meant quite different things by the word, and so were
engaged in a verbal dispute, rather than in a genuine discussion or argument.
The point is at we should not disagree with anyone until we first find out what
he means by his words. To understand before we disagree is not only a rule of
courtesy hut also good sense.
The basis for the troubles we have just described is the fact that words are
ambiguous. The key words in many disputes have more than one meaning, or more
than one sense, and this leads to misunderstandings.
Words stand for things, but we don't have just one word for each thing, like a
buttonhole for each button. More than one word may stand for the same thing:
such words are synonyms. One word, on the other hand, may stand for several
different things. When there is uncertainty as to the meaning which e speaker
or writer intends, there is ambiguity. For example, the word
"secretary" usually means "a person who attends to
correspondence." But a big-game hunter may tell you that on his last trip
to Africa he captured two secretaries. If you look the word up in the
dictionary, you will find that "secretary" also means "a South
African bird with very long legs." Or, a business acquaintance may tell
you that lie recently moved his blonde secretary into his home. He is referring
to a writing desk made of light-colored wood. The ambiguity of words may create
embarrassing misunderstandings! The word "equal," similarly, may be
understood in different senses, and this opens the door to misunderstandings of
the kind we have just described. Words are not ambiguous by themselves but only
in a context which makes their meaning uncertain. There is no ambiguity, of
course, in "I wish to dictate a letter to my secretary."
When words are spoken rather than read, their phonetic sounds may be ambiguous.
The sound "tears," for example, in "The audience sat in
tears." Tears or tiers? The next example also involves phonetic ambiguity:
"Some people pray on their knees on Sundays and on their neighbors the
rest of the week."
Let us now look at another example of a discussion in which a key word is used
in different senses. Let us suppose that there is a disagreement over the
number of unemployed in the United States at the present. time, and that two
collectors of statistics have reached different results in counting the
unemployed. The difference between the statisticians may be due to biased
figures, or unrepresentative samples. But the be a semantical one-the
statisticians may have defined the word "unemployed" in different ways.
There are some industries which employ seasonal workers, such as the canning
industry. Is a seasonal worker unemployed during the winter months, when he is
regularly laid off? One statistician may consider him unemployed; the other may
say he is employed, for he expects to return to his job in the spring. Or the
statisticians may differ with respect to the classification of workers who are
ill, or on strike. These matters should be settled by definitions, other-wise a
verbal dispute may occur because of the different meaning given to the word
"unemployed." If we are not agreed on what we mean by a word, we
shall talk at cross-purposes.
A verbal dispute frequently engaged in by college students arises over the old
chestnut: en a tree falls in an uninhabited forest, does the crash make a
sound? The argument goes on and on. One side agrees that there is no sound
because there is no one present to hear it; the other that longitudinal air
waves, known as "sound waves," will occur in the air whether or not
anyone is present to hear them, so sound is present. Now, the sciences of
physics and psychology tell us that "sound" occurs when waves in the
air hit our eardrums and cause motions in our nervous systems, finally reaching
the brain. When motion finally reaches the auditory nerve, we experience what
is called "sound." This analysis reveals the presence of two
elements: (a) a certain kind of mental experience and (b) a physical cause of
that experience. In the dispute we just noted, "sound" was used in
both of these senses: (a) for the experience itself and (b) for the physical
cause, the sound waves that cause the mental experience. In sense (a) the crash
does not make a sound in the uninhabited forest; in sense (b) it does.
Verbal disputes indicate the manner in which the ambiguity of a word may result
in our talking at cross-purposes. Precisely the same sort of thing happens in
verbal agreements, as distinguished from disputes. Verbal agreements are
"merely apparent agreements. We may find ourselves in apparent agreement
with another person only because of ambiguity and speaking at cross-purposes.
Just as a verbal dispute conceals a possible real agreement, so a verbal
agreement may mask a real dispute. Consider the agreements reached between this
country and Russia after World War II, or with No Vietnam. In both instances
ostensible agreements were found to be based on words, rather than on a meeting
of minds. Once again, talking at cross-purposes.
So much for verbal disputes and verbal agreements. The problem of ambiguity has
much wider ramifications, some of which we shall now explore. Ambiguity is an
ever-present aspect of language, for most words have many meanings. This leads
to difficulties in communications, as noted, but it also vastly enriches
language. A word like "fast," for example, which refers to abstention
from food, to a quality of colors, to certain kinds of characters, and so on,
is the equivalent of many words. But we are primarily interested in ambiguity
insofar as it is an obstacle to communication. There are four types of
ambiguity that cause trouble of this kind, and we shall briefly survey each
type. The four types: the ambiguity of single words, of sentences, of emphasis,
and of significance.
1. The ambiguity of single words
Verbal disputes usually involve the ambiguity of a single word or expression.
One of the ways in which we can detect the presence of this kind of ambiguity
is to ask a question containing the suspected word, and phrase the question so
that it can be answered by Yes or No. If a Yes or No answer requires a
specification of the sense, then the question is ambiguous. "Are all men
equal?" Whether we answer by Yes or No, we must specify the sense of
"equal." The best answer: "Yes and No, depending on the sense of
'equal.'"
Similarly, the question, "Do you believe in God?" requires
clarification of the sense of the ambiguous term "God." Now, there
are many people who object to this kind of analysis, and who say that they want
no quibbling; they want a simple Yes or No answer. But philosophers have
defined the word "Cod" in different ways. When the American
philosopher Arthur 0. Lovejoy applied for his first teaching position, the
application asked the question "Do you believe in God?" Lovejoy
appended a list of more than thirty philosophical definitions of God and asked,
"In which of these senses is the question to be answered?" (He got
the job.) Or consider the conception of God held by the philosopher Spinoza,
who was a pantheist. The pantheistic conception of God holds that God is the
system of Nature as a whole, in all its existential and dynamic aspects. For
Spinoza, everything in the world is part of God, and God is everything. Spinoza
regarded himself as a profoundly religious man; for him, God is the only Being
who can be loved by man without fear of man's ever being disappointed, for God
is eternal, infinite, and perfect. But most people think of God as a personal
Being, as the Creator of the Universe, and pantheism denies that God is a
Person. It is thus apparent that one may believe in God, in one sense of that
term, and yet be considered not to believe in God, in a different sense. To the
question, "Do you believe in God?" Spinoza would have answered,
"Yes, in one sense; no, in another."
Unless we are alert to the possibility of ambiguity we may find no sense where
sense is present, as in this sentence from Paul's First Epistle to the
Corinthians: "And though I bestow all my goods to feed the poor and have
not charity, it profiteth me nothing." This is an apparent contradiction,
for "charity" means almsgiving. But in an older sense, charity means
love (from the Latin caritas), and certainly charity in the modern sense is
possible without love.
Ambiguity may also result in fallacious reasoning, as in this example of a bad
argument: Science has discovered many laws of nature. This is proof that there
is a God, for a law implies the existence of a lawgiver, and God is the great
Lawgiver of the universe." This argument is vitiated by the ambiguity of
the word "law." In "laws of nature," law is used in its
scientific sense. It means "a description of the uniform behavior of
natural events." In another sense, that is, in the sense of "legal
law," law does imply the existence of a lawgiver, for law in this sense
means "regulations emanating from legislatures and courts, rules
enforceable by the state, etc.,' A simple way of differentiating the two is to
say that a natural law is a description; a civil law is a prescription. We need
a lawgiver to prescribe, or issue a command, but scientists use the term
"law of nature" to mean that they have found some kind of invariant
behavior in nature. Their description of this behavior has nothing to do with
commands. To use an ambiguous term in two different senses, as this argument
does, and to draw a conclusion based on this confusion of two senses, is to be
guilty of the fallacy of "equivocation."
The first type of ambiguity, the ambiguity of single words, is also a prolific
source of humor: If I had a mind to," Wordsworth said, "I could write
like Shakespeare." the following example, the humor was unintentional: It
was reported that at Oxford and Cambridge, "half of the school population
are poor students on scholarships." "Poor": in money or studies?
Good, bad, and indifferent puns are based on the same kind of ambiguity. Sydney
Smith, the wittiest of English divines, once described how two women used to
lean out of their windows, on opposite sides of the street, and argue with each
other. "They will never agree," he said, "for they are arguing
from different premises." Benjamin Franklin coined a famous pun when he
warned his fellow colonists that they had better work in cooperation against
England. "We must indeed all hang together," -he said, "or, most
assuredly, we shall all hang separately."
2. The ambiguity of sentences
Let us now look at the second type of ambiguity, the ambiguity of sentences, as
distinguished from the ambiguity of single words or expressions. Ambiguous
sentences are statements whose grammatical construction may lead to possible
misinterpretation. Such sentences are called "amphibolous." This is
the kind of ambiguity involved in the "help wanted" sign at the
factory entrance: "Wanted: Young Girls to sew Lace Trimmings on the 4th
Floor." This "grammatical ambiguity" is the kind employed in the
messages which diplomats like to send to each other. The ambiguous language
permits the writer to claim that he did not mean what the reader thought he
meant. Literally and strictly interpreted, his language may commit him to
nothing. For example, a diplomatic editor wrote a short note to a would-be
author: "I shall lose no time in reading your manuscript."
This type of ambiguity was used by the famous oracle at Delphi in ancient
Greece. The oracle made predictions of things to come, and the predictions
always came true, because of the form in which they were cast. For example: If
the oracle were questioned concerning the outcome of a battle between the
Greeks and the Persians, the oracle would deliberately cast its prediction in
amphibolous language, something like this: "The God Apollo says that the
Greeks the Persians shall subdue."
3. The ambiguity of emphasis
Our third type of ambiguity we call the "ambiguity of emphasis." It
occurs when we are uncertain as to the emphasis which words require. This is
one of the reasons why it is so much more interesting to hear a play performed
by competent actors than to read the play, for the actors give the words their
proper emphasis. An example of this type of ambiguity is found in "Nothing
is too good for her." Consider how the meaning will vary with the
emphasis!
Errors in emphasis occur when we stress the wrong words in a sentence and thus
distort the meaning of the writer. Consider the Ninth Commandment: "Thou
shall not bear false witness against thy neighbor." Now, if one reads this
commandment with the accent on neighbor, this suggests that it is permissible
to bear false witness against those who are not our neighbors, a meaning
obviously not intended.
Errors in emphasis also occur in our writing when we misrepresent another
author's meaning by making biased selections or quotations from his work. Such
improper emphases, however, are usually due either to carelessness or to
deliberate design, rather than to actual ambiguity. For example, a professor
made the following comments on a student's thesis:
"Your thesis is both good and original. Unfortunately, the good things in
it are not original and the original things are not good." The student, an
expert excerpt-lifter, quoted his professor's remark: "Your thesis is both
good and original."
4. The ambiguity of significance
Our fourth and final type of ambiguity is called the "ambiguity of
significance"; ambiguity, that is, concerning the significance of what is
being said. This type of ambiguity occurs when a perfectly true statement has
misleading connotations, as if one were to say, "John didn't beat his wife
last week." (Was this his usual practice?) Or this one: "British
statesmen always put the interests of Britain first." True, but what is
its significance? It insinuates that British statesmen are more selfish than
others. But every statesman puts the interests of his country first. It is not
the statesman's job to sacrifice his country's interest to the interests of
other countries. And when a characteristic is true of everyone, as this one is,
specific individuals or nations deserve neither credit nor discredit for
sharing it with everyone else.
An old story about a sea captain and his first mate will serve as a final
illustration. The captain and his mate alternated in writing the happenings of
each day in the ship's log. One day the mate drank too much, and the next day
he found the entry, "The mate was drunk today." Very much annoyed,
the mate asked the captain why he had made that entry. t's true, isn't
it?" the captain asked. The mate admitted it was. "Well, then,"
said the captain, "if it's true then it is properly entered in the
log." The next day the captain (who was a sober man) opened the log and
found the mate's revenge in the notation, "The captain was not drunk
today.,"
So much, then, for the subject of ambiguity and its four forms, ambiguity of
single words, of sentences, of emphasis, and of significance. Is there anything
we can do to avoid these fertile sources of misunderstanding and thus improve
the process of communication and our writing? There is. Whenever we find
ourselves in a dispute, we can ask whether a key word is being used in
different senses. Whenever we read editorials or other discourse containing
opinions and arguments, we can check to determine whether any of the four forms
of ambiguity are present, and if so, whether our first interpretation of the
meaning is the only possible one that makes sense in the given context.
The cure for the troublesome aspects of ambiguity lies in making our ideas
clear. And this means that we must define our terms. The general problem of
definition will be discussed in the next chapter.
FOR DISCUSSION AND WRITING
1. Meanings of words are based on arbitrary human choices. As a result, few
words have fixed meanings but are in a constant state of flux. For example, the
word "silly" has taken on a somewhat derogatory meaning in recent
times, but it originally meant 'blessed" or 'holy"; later, it came to
mean "defenseless," and now signifies "foolish." Using a
good dictionary (the Oxford English Dictionary if available), trace the
histories of the following words:
propaganda
liberal
brave
counterfeit
nice
democracy
crafty
fret
villain
buxom
What conclusion can you draw concerning the reliability of etymology as a clue
to present meaning?
2. As we have seen, the meaning of a word is based upon convention; that is,
the speakers of a given language accept by general agreement that a certain
symbol signifies a particular thing. Speakers of English will accept
'how-wow" as the word which describes the sound of a barking dog. Indeed,
it is generally agreed among speakers of English that 'how-wow" is what
they hear when a dog barks. But even in such a case where the word is apparently
based on onomatopoeia, convention enters in. For example, a Chinese dog goes
"wang-wang, and a Netherlandish dog "waf-waf." In the following
exercise can you match the words which signify the sound with the animal that
makes that sound?
1. Polish dog |
a. miih |
2. Latvian dog |
b. liau-liau |
3. Japanese dog |
c. kikeriki |
4. American cow |
d. mu-bu |
5. German cow |
e. snof-snof |
6. French cow |
f. meck-meck |
7. Hungarian cow
|
g. vau-vau |
8. Portuguese
pig |
h. meu |
9. Polish pig |
i. kukeleku |
10. Hungarian
pig |
j. cue-cue |
11. Finnish pig |
k. be |
12. French sheep
|
l. ming-ming |
13. Russian
sheep |
m. bo-bo |
14. German goat |
n. roff-roff |
15. Turkish goat
|
o. moo |
16. Dutch
rooster |
p. kwick-kwick |
17. German
rooster |
q. beh |
Answers: 1-b; 2-g; 3-1; 4-o; 5-a; 6-h; 7-d; 8-j;
9-p; 10-n; 11-e; 12-k; 13-q; 14-f; 15-m; 16-i; 17-c
3. In the following statements a word becomes more than an arbitrary symbol for
a thing. In some instances the word takes on magical power; in others, the word
attempts to give us a guarantee as to the worth (or lack of worth) of a thing.
In still others, the word is considered as necessarily connected with a thing.
Analyze the statements, attempting to describe the author's assumptions in
using his words. Does he recognize that words are merely symbols? Is he
consciously trying to establish a connection between the word and the thing?
For what purpose?
a. Imperial margarine is fit for a king.
b. Professor: If you were to create a new language, which of the
following words would you use to designate "iron": "sig"
or "sug"? Student: "Sug"! The word sounds like
"iron."
c. I baptize you in the name of the Father, the Son, and the Holy Ghost. Amen.
d. The allied forces made a protective reaction strike yesterday as they bombed
enemy territory.
e. Dr. Henry Gibbons described a kiss as "the anatomical juxtaposition of
two orbicularis oris muscles in a state of contraction." Of this
definition a newspaper editor remarked, "A kiss may be one of those
things, but it doesn't taste like it."
f. Capri. The first sexy European under $2600.
g. Notice to all department heads: In all reports, the work designation
"garbage collector" will be changed to "sanitation
engineer."
h. I refuse to salute the flag. It represents all those things I consider
anti-American.
i. "Who could find anything better than hum, or buzz, or whir?
Who could think of anything more sloppy than slop? Is not the word sweet
a kiss in itself, and what could suggest a more peremptory obstacle than stop?"
(From Englishmen, Frenchmen, and Spaniards, by Salvador de Madariaga.
New York: Oxford University Press, 1937)
4. Consider the following statements. What does the author of each suggest
about the nature of language? Is he correct in his assessment?
a. "Dilapidated: said of a building or other structure. But the word is
from the Latin lapis, a stone, and cannot properly be used of any but a
stone structure." (Ambrose Bierce)
b. "As the language is, so also is the nation." (Otto Jespersen)
c. What's in a name? That which we call a rose by any other name would smell as
sweet." (Shakespeare, Romeo and Juliet)
d. "Names belong to things by nature and the user of words must keep in
mind that the name of a thing belongs necessarily to the thing itself."
(Plato, Gratylus)
e. "So God formed out of the ground all the wild animals and all the birds
of heaven. He brought them to the man to see what he would call them, and
whatever the man called each living creature, that was its name. Thus the man
gave names to all cattle, to the birds of heaven, and to every wild
animal." (Genesis 2, 19-20)
5. Verbal disputes usually involve the ambiguity of a single word or
expression. One of the ways in which we can detect the presence of this kind of
ambiguity is to ask a question containing the suspected word and phrase the
question so that it can be answered by Yes or No. If a Yes or No answer
requires a specification of the sense, then the question is ambiguous. In the
following sentences, apply the above test. Which words in the statement require
clarification before a Yes or No answer can be given?
a. God is dead.
b. Welfare programs destroy the recipients' desire to succeed.
c. Anti-pornography laws should be strengthened.
d. Love makes the world go round.
e. There is no such thing as a moral war.
6. Select one of the above statements and analyze the crucial words, their
various possible meanings, their purpose in the context of the statement, and
the author's intention in using those words. In a paragraph, attempt to
describe the conditions under which one could answer Yes or No to the
statement.
7. The following newspaper headlines are ambiguous because of their grammar or
diction. Rephrase each headline, indicating the two meanings which it may
suggest.
a. Student rates high
b. Man eating shark found
c. Nude swims tonight
d. Player shoots ace in tournament
e. Students stoned in Hollywood
f. Men's trousers half off Tuesday only
g. Race held inferior
h. Giraffe tastes sweet
8. The ambiguity of significance occurs in a statement in which the semantic
value of the words is clear, but the full significance of the factual statement
is not. For example, in the sentence, 'e number of heroin addicts in the United
States rose 20 percent during the present administration," the fact is
verifiable, but its significance is ambiguous. Does the figure 20 percent
represent a drastic increase when compared with other comparable periods? Or
does it represent a drastic decrease? How was the figure arrived at? Does the
speaker have a special interest in making the statement, that is, is he
pro-administration or anti-administration? Analyze the possible ambiguities in
the following statements. what questions need to be asked about the fact
contained in each?
a. The enemy suffered 735 fatalities last week; the friendly forces suffered
seven.
b. Professionally performed abortions are about eight times safer than
full-term pregnancies.
c. Unemployment is up to 2.5 million in the first quarter of this year.
d. They were soundly beaten in the roll call vote by 246 votes.
e. The 14 percent increase in hospital room rates this year indicates that
hospitals are doing everything in their power to hold prices down.
9. In a short paragraph, write an analysis of one of the preceding statements,
indicating the possible ambiguities which may be present.
CHAPTER 3
Define Your Terms!
The great practical problem of semantics, as we saw in chapter 2, is
communication. We have seen how ambiguity is responsible for failures in
communication. But communication fails for other reasons, too: we don't know
what the other fellow's words mean; he doesn't understand what we mean; and we
may not even understand what we ourselves are saying.
The point is that we ought to know what we are talking about. When we use
high-level abstractions, words like "democracy," "freedom,"
"capitalism," or "communism," we should make our meanings
clear. We should never forget that a word is like a check drawn against the
world of experience, and that it has no meaning for us unless we can "cash
it in" by pointing eventually to that to which it refers. When we speak or
write, let us beware of "glittering generalities" which we do not
understand, and let us not speak unless we know what we are Saying. It is this
necessity for making our meanings clear that was in the mind of the French
philosopher Voltaire when he said, "Before I will discuss anything with
you, you must define your terms."
Definition is one of the convenient and natural techniques for developing your
ideas in an essay. In your college writing you will frequently need to use
words and phrases that have various meanings, or shades of meanings. To avoid
vagueness or ambiguity, you will need to define your terms. Some words can be
defined by a synonym or a phrase. Others, however, are too complex or abstract,
and you will want to extend their definitions for a paragraph or longer. In any
event, once you have faced the need to understand or explain a term-once you
have defined your subject you will see the direction your discussion or theme
should take. Equally important, your reader or listener will understand what
you mean.
Vagueness
Communication often fails because words are ambiguous, or vague, or because
they are used loosely, or carelessly, or without meaning. Ambiguity is not
quite the same thing as vagueness. An ambiguous word is one capable of being
understood in more than one sense in a given context. We are not sure which
sense is intended. By a vague word, we mean one whose meaning may be fairly
well understood, but whose limitations are unclear. Thus, the term
"obscene" is vague, for we don't know where to "draw the
line." A statute which forbids vehicles in a public park obviously applies
to motorcycles. Bicycles also? Tricycles, Roller skates?
The use of most of the words in ordinary language becomes vague when we are
confronted with "fringe" applications, but each word also refers to
clear-cut examples, called "paradigm" cases, to which the word
typically refers. There is no question that automobiles, airplanes, and railroad
trains are vehicles, as are boats.
In mathematics and the sciences, and also in law, terms are precisely defined,
but such words are taken out of everyday circulation. Words in ordinary speech
cannot be defined with such exactness.
Vagueness can sometimes be eliminated by "drawing a' line." The
expression "the West," as used in American history, is a vague term.
Bernard DeVoto gave this term a specific meaning when he said that "the
West begins where the average rainfall drops below twenty inches. when you
reach the line which marks that drop-for convenience, the one hundredth
meridian-you have reached the West."
There are some important words which are both ambiguous and vague, that is,
they have several unclear meanings. This is a kind of
"double-barreled" ambiguity. The words "freedom" and
"liberty" fall into this category. Consider the famous
"Declaration of the Four Freedoms" which mentions Freedom of Speech,
Freedom of Religion, Freedom from Hunger, and Freedom from Fear. Note that the
first two of these Freedoms have a "negative" connotation, referring
to the limitation of the power of the state with respect to the rights of the
individual. The third and fourth freedoms, on the other hand, imply that the
state must take positive action to provide the people with food and security.
Or consider words such as "liberal" or conservative." There are
no universally accepted definitions of these terms; dictionaries can merely
list their various uses.
Defining your terms
In serious discussions, when key words are subject to varying interpretations,
i.e., when words like "liberal" or "conservative" are used,
the speaker or writer who wishes to make his meaning clear should stipulate
definitions of his terms. By "stipulation" we mean "specifying
or particularizing": The speaker announces the precise sense in which he
will use the word. lie should begin by saying, "In this discussion the
word X will mean . . ." This is flow the meaning the word will have
whenever he uses it. Three possibilities are open to the stipulator: (1) He may
find one of the several customary meanings of a word adequate, (2) he may
choose to stipulate a modified version of a customary meaning, or (3) he may
stipulate a brand-new meaning.
Stipulations would be unnecessary if words had one and only one meaning. But
since this is not the case, stipulations of definitions are often an
indispensable element in making our meanings clear. The stipulator has a great
deal of freedom, but great caution should be used in stipulating new meanings.
Readers and audiences find it difficult to follow a speaker who uses words in
unfamiliar ways, and the speaker's freedom is limited by his desire to hold his
audience. Confusion is likely to result when new meanings are used, for old
habits are hard to get rid of and we usually persist in giving familiar words
their customary meanings. And worse: The stipulation of new meanings may
involve dangerous traps for the innocent. There ought to be a code of
linguistic ethics for all speakers and writers; when they stipulate new
meanings, they should warn the reader that they are doing just that. The danger
is that the stipulator may give us a new meaning as if he were merely giving us
one of the customary meanings that all of us accept, and thereby get us to
believe things we otherwise would not.
As an example of the stipulation of a new meaning without prior notification
that this is being done, let us examine the following paragraph:
Many readers are quick to dismiss "confession" magazines as
containing little of permanent literary worth. They forget that the confession
story has provided us with some of the greatest realistic, revealing, personal
literature of the ages-Cellini's Autobiography, The Confessions of Jean Jacques
Rousseau, De Quincey's Opium Eater. Boil down to its essence any of the great
enduring classics and you will find-a confession. Anna Karenina, Madame Bovary,
Tess of the D'Urbervilles, Sister Carrie, Of Human Bondage-these are all
confession stories-all based on human emotional problems, personal conflict,
human desire, human greed, human passion.
The net effect of this passage is to make the reader think that the stories he
finds in confession magazines on any newsstand-with their predictable plots,
one-dimensional characters, banal dialogue, and superficial treatment of the
human condition-are in the same genre and of the same quality as the world's
greatest literature. How does the writer of this passage do this? By implying,
in his last sentence, that any literary work based on "human emotional
problems, personal conflict, human desire, human greed, human passion"
fulfills the requirements of the confession story. In other words, because
confession stories deal with the same topics as the classics he cites, they
belong to the same literary classification. But such a definition ignores the
differences in technique, depth of feeling and emotion, use of language,
originality of plot, portrayal of characters, and all of the other
characteristics we associate with great literature.
There is another type of trick that may be used in the stipulation of
definition. The stipulator may take familiar words having favorable
connotations-like "democracy," or "freedom"-and redefine
them with a quite different content than is customary. The unwary may fall into
the trap of carrying the favorable connotations of the old word to the new
content, thinking that the new must be as good as the old because it is called
by the same name. This, of course, is the fallacy of thinking that words are
guarantees of things.
For example, we sometimes hear people say, "The Russians (or Chinese) have
democracy, just as we have democracy, except in a different way. They have
democracy as they define the word, namely, as a system in which 'the
people,' rather than private capitalists, own the industrial plant." Now,
of course, the Russians (or Chinese) may call their system a "people's
democracy," and they have "democracy" in accordance with their
stipulated meaning. There is no international law which forbids this kind of linguistic
behavior. But they mean something quite different from what we mean, and they
adopted this term "democracy" for very special purposes. Having their
own aims in view, and knowing that the slogan of "democracy has great
appeal to people because of the libertarian and egalitarian ideals it usually
stands for, they adopted this word, but gave it a new meaning. The fact that
the same word is used makes people feel that the thing is the same. And since
we regard democracy as good, and they use the same word, the innocent may conclude
that their system must be as good as ours, though in a different way.
When confusions of this kind occur in a discussion it is better to discard the
word "democracy" altogether, and thus avoid the possibility of these
confusions. The important thing is to compare the actual practices of the rival
systems, regardless of what they are called.
We have been discussing some problems with respect to the stipulation of
meanings. It is desirable for speakers and writers to tell us how they will use
important words. They may stipulate a customary meaning or a new one. But when
they stipulate brand new meanings without informing us that these are new
meanings, we may be misled. Danger lurks in the failure to give us such
warning.
Definitions, then, contain more than meets the eye. The definer's purpose, in
other words, may be other than informative. lie may have propagandistic aims,
and his definitions may be slanted according to his purposes.
There is also a type of expression which looks like a definition but which
really aims at imaginative insight, to capture the "essence" of
things. "Architecture is frozen music" is an example. "Poetry is
music plus imagination" would not pass muster as an exact definition,
although it may be more enlightening for some purposes than a more literal
statement. And inventive imaginations have coined many amusing
"definitions" which contain sharp social satire: "A politician
is a man who sits on a fence with his ear on the ground"; "A wedding
ring is a matrimonial tourniquet designed to stop circulation"; "An
explorer is a bum with an excuse.
A definition is a statement that says, "For this word A, substitute these
words, X, Y, Z." For the word 'perjury," substitute its legal
definition "the wilful and voluntary giving of false testimony under oath
or affirmation, with regard to a material matter, in a judicial
proceeding." The single word is substitutable for the long expression, and
vice versa. But there are different ways in which a definition may clarify
meanings. When we turn to a dictionary for the meaning of an unfamiliar word,
we find three kinds of definitions. We find synonyms, examples, and analyses of
meanings. For example, if we look up the word "troglodyte," we may
get a synonym: "hermit." Or suppose we look up "oxymoron."
The American Heritage Dictionary tells us that an oxymoron is "a
rhetorical figure in which an epigrammatic effect is created by the conjunction
of incongruous or contradictory terms." This is an analysis of the
meaning. But since this analysis is somewhat obscure, our dictionary helps us
out by furnishing us with an example of an oxymoron: "for example, 'a
mournful optimist."' (Other examples are, "To make haste
slowly," "His kindness was cruel.")
Definition by analysis
Definitions, then, may consist of synonyms, examples, or analyses. Any one of
these will do, provided that it makes the meaning clear and helps us to
understand what the other fellow is trying to say. But the most enlightening
type of definition is the analytical. This type is sometimes called definition
per genus et differentiam, that is, it states the general class of things to
which a term belongs, and then shows its differentiating characteristics within
that class. When we define a triangle as a "plane figure having three sides,"
we have noted the class of things to which triangles belong: plane figures; and
we have noted how they differ from other plane figures, that is, in being
three-sided.
The most important logical problem in connection with analytical definitions
concerns the nature of an adequate definition of this type. "A triangle is
a plane figure having three sides" is a perfect definition of this type
because it has the attribute of "convertibility." A convertible
definition is one that can be "turned around" and still yield a true
statement. Thus, we can say, "Any plane figure having three sides is a
triangle." Convertible definitions show an "equivalence" between
the definition and the word being defined. Thus, in Aristotle's definition of
man as a "rational ,animal," if we agree that all men are rational
animals and that all rational animals are men, this definition will exhibit
equivalence and convertibility. Another way of checking for convertibility is
to use the "All and Only" test. Can we say all men are rational
animals and that only men are? If we can, then the definition is convertible.
A definition is inadequate if it lacks convertibility. It is then either too
broad or too narrow. "Too broad" means that it covers too much
ground, as in a definition of propaganda as "any talk or action which
influences anyone toward some predetermined end." This definition covers
things it does not mean to cover, as when I say to my neighbor at dinner:
"Please pass the salt." I have influenced him; have I thereby become
a propagandist? (We cannot say that only propaganda involves influencing people
or that all influence is propaganda; we may influence people with-out being
propagandists.) On the other hand, a definition may be too restrictive, and not
cover enough ground. But most faulty definitions will be found to be too broad.
We find some element common to the items we are interested in and hastily
define the thing in terms of those common elements without pausing to notice
that our net has swept all sorts of other things into our definition.
Definition by synonym and example
We noted earlier that definitions by synonym or by example are often quite
adequate. This is the case when all we desire is a reference to the sort of
thing for which a word stands. But synonyms and examples will be wholly
inadequate when an analytic definition is called for. For example: In a
discussion of religion, someone may raise the question, "Just what do you
mean by God?" An answer like "By the term, God, I mean the
Deity" would be inadequate in most cases, for the questioner wanted an
explication or analysis of the term. Definition by synonym is often as
insufficient as was Polonius' response to the King and Queen when they
questioned him concerning Hamlet's strange condition:
Your noble son is mad:
Mad call I it; for, to define true madness,
What is it but to be nothing else but mad?
Definition by synonyms is sometimes called "circular" definition, and
results in a "begging of the question," a matter to which we will
return in Chapter 7. An example: "A morally good man is one who acts
virtuously." "Morally good" and "virtuous" are
synonymous terms in this context, so that the definition merely repeats the
word that is being defined. Circular definitions are of course
"convertible equivalents," but they are faulty in that they offer no
clarification of the meaning of the word being defined. If we are looking for
clarification of the meaning of "morally good" so that we may know to
what kinds of conduct it refers, it is not helpful to be told that
"morally good" refers to "virtuous actions." This is like
saying that 'virtuous actions are virtuous actions." When Hamlet tells his
friends that he brings "wonderful news," namely, that "there's
ne'er a villain dwelling in all Denmark, but he's an arrant knave,"
Horatio answers: "There needs no ghost, my lord, come from the grave to
tell us this."
Closely resembling these faults in definition are such things as pleonasms and
rhetorical tautologies, as in saying, "lie is writing his own
autobiography," or "I have one small son, a boy." True, but
foolishly superfluous. Even more amusing, usually, is the "Irish
bull," a good example which occurred in the movie The Quiet Man:
"He'll regret this to his dying day, if he lives that long." The
absurdity of this lies in the fact that it denies a tautology, and a tautology,
though vacuous, is necessarily true. And here is one more example, from Pliny,
the ancient Roman: "It is better to be idle than to do nothing."
Synonyms, then, do not always satisfy our demand for a definition.
"Definition by example" may also be inadequate: "What is
poetry'?" "Milton's Paradise Lost is an example of a poem." This
does not tell us much about the nature of poetry. Or suppose that we are asked
to define "free enterprise," and we point to the United States as an
example of a nation having free enterprise. We shall give more than we intend
by this example of "pointing," for a stranger might conclude that
legislation in aid of farmers was an essential part of a free-enterprise
system. The vagueness of the reference that may accompany the gesture of
pointing is well illustrated by a story, that has become a classic among
students of language. It is narrated by J. H. Weeks, in his Among Congo
Cannibals:
I remember on one occasion wanting the word for Table. There were five or
six boys standing round, and, tapping the table with my forefinger, I asked,
"What is this?" One boy said it was a dodela, another that it was an
etanda, a third stated that it was bokali, a fourth that it was elamba, and the
fifth said it was meza. These various words we wrote in our notebook, and
congratulated ourselves that we were working among a people who possessed so
rich a language that they had five words for one article.
But later Weeks discovered that
one lad thought we wanted the word for tapping; another understood that we
were seeking the word for the material of which the table was made; another had
an idea that we required the word for hardness; another thought we wished a
name for that which covered the table; and the last, not being able, perhaps,
to think of anything else, gave us the word meza, table--the very word we were
seeking.
There are of course some situations in which an analytic definition cannot even
be attempted, and where only a definition by example (or pointing) is possible.
If we are asked what "chrome yellow" means, no words can designate
its sense qualities. If this is not obvious, ask your-self how you would
explain what any color is, to a person who had been blind from birth. We can
explain what sound and light waves are to a blind person, for he can understand
these things in terms of his sense of touch, but he cannot understand what we
mean by color, for this depends on a sense he lacks.
In concluding our discussion of definition we must issue one final warning. The
history of philosophy may be regarded as the record of man's search for
adequate analytical definitions of the key terms in human discourse, words such
as "truth," "beauty," and "goodness." Let us not
hastily assume that "the last word" has been spoken in defining these
terms. The "last word" has not been spoken on these matters, and
probably never will be. And let us not dogmatically assume that we have
adequate definitions of any term whose meaning is a matter of controversy.
Nor should we demand precise definitions of that which lacks precisely
determined characteristics. "Art" is in this category. After
attending a few exhibitions of "modern art" many spectators
experience a sense of bewilderment, and they are apt to raise the question:
"Just what is art?" The dictionary will not be very helpful here.
According to the Random House Dictionary, art is "the class of objects
subject to aesthetic criteria; works or objects belonging to this realm, as paintings,
drawings, etc.: a museum of art." Not a very useful instrument for
distinguishing art from not-art! There is no definition which can precisely
delimit art from that which is not art. And the same problem will be found to
arise in connection with many familiar words. To insist on formulating and
applying precise definitions in situations in which they are inappropriate is
to fail to make sense.
FOR DISCUSSION AND WRITING
1. The meaning of a word is determined largely by its context. As the context
changes, so too a word's signification may change, however slightly.
Interpretation of a writer's or speaker's definition of a word often depends
upon its use in the sentence, paragraph, or larger unit of communication. If
you were to define the word grable as it is used in the following five
statements, what definition could you provide? Write a twenty-word definition,
paying careful attention to the possible shades of meaning the word may
suggest.
a. To grable before the enemy is hardly heroic.
b. She grabled for several years, but her psychiatrist finally was able to help
her.
c. I am not sick; I grable because I am naturally high-strung.
d. The tender leaf grabled before the wind's force.
e. Transmission grabling? See Honest Doug!
In a similar manner, define krafic on the basis of its use in the following
sentences:
a. It was krafic enough to see the mountains.
b. His argument was krafic but his language was atrocious.
c. I have never seen a more krafic young lady. I would like my son to marry
her.
d. Did you krafic the inside as well?
e. The vernal glade kraficked my soul.
2. How is democracy used in each of the following sentences?
a. The United States is a democracy.
b. In a democracy any man or woman can speak freely.
c. In spite of the fact that he is a member of the nobility, he takes a very
democratic view of the man in the street.
d. The Democratic party uses the donkey as a symbol.
e. The Students for a Democratic Society did not allow the speaker to continue.
3. Which of the following offer a definition of some term? What method is used
to explain the meaning of the word (definition by synonym, analysis, example,
or other method)?
a. Happiness is an activity of the soul in accordance with virtue in a complete
life. (Aristotle, The Nicomachean Ethics)
b. "babble": chatter
c. Science is what we know and philosophy is what we don't know. (Bertrand
Russell, Bertrand Russell Speaks His Mind)
d. "Network" means anything reticulated or decussated, at equal
distances, with interstices between the intersections. (Samuel Johnson's
Dictionary)
e. "cosmetic": rouge, lipstick, face powder
f. A liberal is a man who cultivates the skills that make freedom operational.
He is always a man on special assignment. (Max Ascoli, editor and publisher of
The Reporter magazine)
g. The root difference between the Conservatives and the Liberals of today is
that Conservatives take account of the whole man, while the Liberals tend to
look only at the material side of man's nature. (Barry Goldwater, The
Conscience of a Conservative)
h. "awkward": clumsy
i. Philosophy is a battle against the bewitchment of our intelligence by means
of language. (Ludwig Wittgenstein, Philosophical Investigations)
j. "philosopher": such men as Aristotle, Plato, Kant, Dewey, and
Jaspers
4. Select one of the following words or phrases and, in a paragraph of no more
than 200 words, define it so that your reader knows exactly the limit of your
use of the word.
a. adolescence
b. university
c. the common man
d. conservative
e. religion
f. law
g. liberty
h. idealism
i. the American way of life
j. love
5. For each of the following terms supply a synonym, an example, and an
analysis.
a. pop art
b. automobile
c. sin
d. honor
e. a delicious meal
6. Give examples of definitions that are circular, tautological, and lacking convertibility.
You may cite those you encounter in your reading or supply your own.
CHAPTER 4
What Kind of Language Are You Using?
A misunderstanding may occur because we fail to grasp the meaning of a word, or
the thoughts embodied in a sentence. But there is another obstacle to
communication: we misunderstand the purpose of speech. The most typical form of
this mistake is to treat every use of language as if it were intended to convey
information. For example, many people read a poem as if it were a scientific
treatise. But the criteria of truth and falsity may be irrelevant to the poet's
purpose, so far as his actual statements are concerned. He may be trying to
evoke a mood, or a state of feeling, or attitude, rather than to give us
literal truths. When Shelley writes of the skylark,
Hail to thee, blithe spirit!
Bird thou never wert,
he did not really mean to deny that the skylark is a bird. His language was
"emotive" rather than scientific in its purpose. What we are
emphasizing here is that we shall do well to look for the purposes and
intentions of speakers and writers. In this chapter we shall consider the
different functions, purposes, and uses of language.
The functions of language
Language, in other words, has more than one purpose. We might say that language
operates on different levels, except that the word "levels" suggests
higher and lower planes in a scale of value, and this is not intended here. We
shall deal with three functions: the informative the expressive, and the directive.
To say that language has three functions is to say that there arc three
different reasons for speaking. One reason, or purpose, is to communicate
factual information. This is the informative function, probably the most
important for our purposes. We speak also in order to express our feelings, to
"blow off steam," or to stir the feelings and attitudes of the person
we are talking to. We shall call this the expressive or "emotive"
function. And, finally, we speak in order to get people to act. This is the directive
function.
Some illustrations are in order. A book on astronomy describes the solar system
and the stars. We learn that the diameter of the earth is about 8,000 miles,
that of the sun, about 800,000 miles, a ratio of 100 to 1. We learn that the
star Betelgeuse has a diameter three hundred times that of the sun. This means
that if the earth is represented by a baseball, about three inches in diameter,
then Betelgeuse would have a diameter of almost a mile and a half. We may learn
that there are as many stars in the heavens as there are grains of sand on all
the seashores of the world . . . I have just been using language to communicate
information. This use of language tries to avoid words with private
connotations or words that unnecessarily stir up feelings. If you want your
writing to be clear, do not use words that distract your readers' attention
from what you are saying to how you are saying it.
Expressive language is a second type. When I talk about the United States
senator I like least, I may let off some steam and relieve my pent-up feelings.
I may even infect you with my feelings, making you feel as I feel. The poet, of
course, is a specialist in expressive language, as in the lines:
Comes the blind Fury with th' abhorred shears
And slits the thin-spun life.
These lines give expression to John Milton's feelings and perhaps make us feel
as he felt. When we tell our friends a funny story to get a laugh, we express
our feelings too and affect theirs. Expressive language often occurs in
political speeches, in plays, and in advertising. When you are writing an
argument, you should avoid whenever possible the use of expressive language.
The third type, directive or action-provoking speech, is illustrated by
examples like: "Do unto others as you would have others do unto you, or
"Drink to me only with thine eyes." We say these things to get
action. Ceremonial language, such as "I am happy to meet you, "What a
beautiful baby!" and conversation about the weather, also have a directive
purpose: to establish social rapport and to get a friendly response. We use
directive language when we are explicitly and overtly trying to get some-one to
act a certain way or to refrain from acting in that way.
There are, then, at least three different purposes of discourse. We may also
make a somewhat similar classification for words, that is, for words taken by
themselves. A basic distinction here is between what we shall call neutral
words and emotive words. Neutral words merely convey ideas to us, as when I
say, "The sun rose at six this morning." The words in this sentence
do not arouse our emotions. But words like "God," "love,"
"freedom," and "family," are so closely connected with our
total attitudes to life that they are likely to arouse emotional reactions. This
division of words into neutral and emotive, however, is relative to our
personal experiences, for there is nothing in the world itself which makes it
neutral or emotive. If a word conveys nothing but an idea to you, then it is
neutral to you; if it arouses your emotions, then it is emotive to you. The
word "bread" is a neutral word to me, but to a "fat boy" or
a starving man, it may be fraught with emotion. Nevertheless there are some
words which can be counted on to make almost everyone "see red," so
to speak, like the word "traitor.
This classification of words is independent of our classification of the
functions of language, for those who wish to inform may use either type, as may
those who want to express their feelings, or to get action. In general,
however, neutral words will be used when we wish merely to inform; emotive
words when we wish to be expressive.
Let us return to our classification of the purposes of language. And let us
avoid the vice of oversimplification. In life, or living speech, the functions
of language are seldom found in a pure or unmixed form. In life things are
rarely simple and never pure. Speech and writing usually present mixtures of
the informative, expressive, and directive functions of language. Consider the
informative item concerning the diameters of the sun and the earth. Though this
language informs, and though it is not its primary purpose to stir our
emotions, our feelings may nevertheless be stirred when we learn from Sir James
Jeans that there are as many stars as there are grains of sand on all the
seashores of the world, for this knowledge may make us realize how infinitely
vast is space and how infinitesimally small and feeble is man, crawling for his
brief day upon an insignificant planet.
Nevertheless, informative language is the type most likely to be found in a
relatively pure form. The writing of scientists is apt to be purely
informative, especially in the physical sciences, like physics and chemistry,
though less so in political science. But expressive language is rarely, if
ever, used exclusively. Expressive language is usually mixed with something
else. The lines from Milton's "Lycidas" did not aim primarily at
giving us information, but they do say something that has the ring of truth in
it. They tell us that men are mortal and that there are forces beyond our
control at work upon us, which give us the helpless feeling that we are the
pawns of fate. Alexander Pope's "Essay on Man" is philosophical
discourse in rhyming couplets and contains a developed system of thought, as in
the closing lines of the First Epistle:
All Nature is but art unknown to thee,
All chance, direction which thou canst not see;
All discord, harmony not understood,
All partial evil, universal good.
And, spite of pride, in erring Reason's spite,
One thing is clear, whatever is, is right.
The poet may also mix a directive purpose with the expressive one. He may want
us to do something, as in so-called "inspirational poetry." Clough's
lines are an example:
Say not, the struggle naught availeth,
The labour and the wounds are vain
The enemy faints not, nor faileth,
And as things have been, they remain.
If hopes were dupes, fears may be liars;
It may be, in yon smoke concealed,
Your comrades chase e'en now the fliers,
And, but for you, possess the field.
The last line of this poem, "But westward, look, the land is bright!"
was quoted very effectively by Winston Churchill during the dark days of the
Battle of Britain in 1940. This poem mixes the directive and expressive types
of discourse. But one should not expect all poems to give us a moral message,
or practical advice, or scientific information. Poets are not necessarily
preachers, though some are, and they are not necessarily scientists or
philosophers, though some may be. Wordsworth's "Daffodils" should not
be read as if it were a botanical treatise on the "Narcissus
pseudo-narcissus of the amaryllis family," to use the technical name for
the daffodil. One of the aims of poetry is to communicate feelings and
attitudes toward life, to convey to us the poet's feelings concerning his
experiences, and to make us aware of life's mystery and wonder.
More about directive language
We shall now examine the directive type of language in some detail. When a
speaker wants action from his audience, he may tell them to do what he wants
them to do. But, as every parent knows, it is often better to use an indirect
approach to get action. Instead of saying, Johnny, eat your spinach" (or
whatever other torture-food happens to be the vogue at the moment), we say,
"My, what a gorgeous-looking dish of spinach that is." (The method
suggested here cannot, unfortunately, carry a guarantee with it.) Similarly a
politician may ask us to work for him, to help him win election; but he may
also use an indirect approach. He may stir our emotions by painting a vivid
picture of the horrible crimes committed by the opposition party. He arouses
our emotions, and these emotions, the psychologists tell us, will demand an
outlet. The person thus aroused wants to go out and do something.
Now, this technique of getting action by working on men's emotions has been
known to mankind since time immemorial and has been practiced by politicians
ever since they made themselves indispensable. An excellent illustration of
this technique is found in Mark Antony's funeral oration in Shakespeare's
Julius Caesar. Brutus, you will recall, assassinated Caesar because he feared
that Caesar had ambitions to make himself dictator of Rome. Brutus is now in
power, but graciously permits Mark Antony, Caesar's friend, to make the funeral
oration over Caesar's dead body. Antony, however, is not primarily concerned
with eulogizing Caesar. He uses the occasion as a step toward seizing power for
himself and seeks to turn the Roman populace from hatred of Caesar to hatred of
Brutus and his fellow conspirators. The oration begins with the famous lines:
Friends, Romans, countrymen, lend me your ears:
I come to bury Caesar, not to praise him...
He was my friend, faithful and just to me.
But Brutus says he was ambitious;
And Brutus is an honorable man....
You all did love him once, not without cause.
What cause withholds you then to mourn for him? ...
My heart is in the coffin there with Caesar,
And I must pause till it come back to me.
Antony is beginning to awaken the emotions of his hearers. He quotes Brutus,
sarcastically. He mentions the sacred bonds of friendship which bound him to
Caesar and which once bound the crowd to Caesar. Antony then goes on to mention
Caesar's will and says that it leaves generous bequests to the people, but, he
adds, he cannot possibly read this will aloud. If the people but knew what
Caesar had done for them, he says, they would not be able to control
themselves, for they are not made of "wood" or "stones" but
of flesh and blood. The mob is now inflamed with expectancy and demands that
the will be read, but Antony puts them off, first showing them the cloak which
Caesar wore when he was killed. He points to the holes made by the stabbing
daggers:
Look! in this place ran Cassius' dagger through...
Through this, the well-beloved Brutus stabb'd...
And, as he pluck'd his cursed steel away,
Mark how the blood of Caesar follow'd it.
For Brutus, as you know, was Caesar's angel:
Judge, O you gods, how dearly Caesar lov'd him.
This was the most unkindest cut of all.
Tears begin to flow, and Antony now reads the will, with its generous bequests
to the people. The crowd leaves Antony in a fury, resolved to destroy Brutus
and the assassins who killed their beloved Caesar. As they leave, Antony,
knowing that he has accomplished his purpose, mutters to himself:
Now let it work: mischief, thou are afoot,
Take thou what course thou wilt!
Antony, like other demagogues, is a master of practical human psychology who
knows how to move the masses to his own ends. This kind of psychological
insight is often turned to bad uses. But it should also be part of the
equipment of anyone who wishes to be effective in moving people to action. Even
in the best of causes human beings need some stimulus to action.
In the next chapter we shall distinguish between legitimate and illegitimate
appeals to emotion. It should be obvious, however, that emotional appeals are
sometimes quite proper, as when the Community Fund and organized charities get
us to contribute by appealing to our hearts.
It is well to know when speakers are trying to get us to do something, rather
to get us to inform us. And also-to know when they are trying to get us to do
nothing. This is a reverse kind of directive language. And just as ordinary
directive language seeks to arouse emotions, so the reverse type seeks to
neutralize emotions, or to de-emotionalize a situation. For example, a shocking
crime occurs in our city: an important public official is arrested for
embezzlement. There will be a great public clamor for drastic action. Those who
fear for the status quo will seek to d am pen public indignation. We will be
told that "the authorities are investigating," etc.
The negative use of directive language often characterizes official government
papers" concerning actions which seem immorally brutal. The actions will
be described in a calm, unemotional manner. Neutral words will prevail, and
emotive terms will be avoided. This use of language was often resorted to by
the Johnson and Nixon administration spokesmen when they sought to gloss over
the harsh realities of the Vietnam war. "In our time," George Orwell
says, "political speech and writing are largely the defense of the
indefensible." Neutral language helps to immobilize our emotions.
Propaganda techniques
Our distinctions between types of language may help to clarify some aspects of
"propaganda analysis." The meaning of propaganda is extraordinarily
vague, at least so far as common usage is concerned. Most people use the word
"propaganda" without having any precisely defined meaning in mind. In
everyday speech this word may mean anything from "a pack of lies" to
"the attempt to influence anyone about anything." But these are
inadequate definitions. Not every liar is a propagandist, and we do not usually
think of a hostess urging us to have another helping of her culinary specialty
as a propagandist. She is trying to influence us about something; but unless
the word "propaganda" is more limited in its meaning, it will be
quite useless, and ought to be dropped from the language. "Trying to
influence others" is almost coterminous with speaking.
The usual dictionary definition is better: "Propaganda is a systematic
effort for the gaining of public support for a course of action." But even
this definition is not accepted by everyone. This lack of agreement concerning
the meaning of propaganda may be demonstrated by the following experiment: Co
to any library and find twenty books in the fields of government, political
science, sociology, and social psychology which contain the word
"propaganda" in the indices. You will find twenty different
definitions of the word. Some years ago a public-spirited citizen of New York
offered a thousand-dollar prize to anyone who would define the word in such a way
as to win general acceptance. There were no applicants.
We shall not offer a twenty-first definition here. But we can point out some of
the things we ought to have in mind when we suspect that we are on the
receiving end of propaganda. The word, after all, is of little consequence; it
is the thing that counts. Just what is it that we are suspicious of, when we
think of propaganda as being in some fashion dishonest, as most people do? This
is what concerns us here.
As thoughtful citizens, we want to know the facts, so that we can come to
intelligent decisions. We are afraid that there are organized efforts to
deceive us so that we will act blindly and unintelligently. What can we do
about this? The answer is quite a simple one: become as well-informed as
possible. Knowledge is needed to see through a lie, whether a big or a small
one, and nothing else will do the job.
In recent years, with the public's increasing awareness that it is being
victimized by "propaganda"-understood as a form of organized deception-many
writers have sought to help the public in its search for protection against
deception. These analysts have sought to give us a method for detecting
propaganda and thereby to achieve security against its harmful effects. We are
told to watch for such techniques or "tags" as
"name-calling," "glittering generalities,"
"testimonials," "band-wagon appeals," etc. Let us look at
the meaning of these categories.
Name-calling, or "labeling," refers to the practice or attaching
"bad" names to individuals or groups or ideas: names such as
"Fascist, reactionary," "totalitarian, racist. These words stir
feelings of fear and hatred for the persons and ideas to whom they are
attached. The abusive terminology makes people forget about the necessity for
knowing what the facts are. But name-calling, though a popular instrument of
propaganda, is not confined to propagandists. We all indulge to some extent)
even when we call a man a "name-caller."
We ought not to call names, of course, as a substitute for giving evidence. And
the rules of courtesy (not to speak of the laws against slander and libel) tell
us that calling names is a boorish as well as an illogical way of making a
point. But it is also very easy to exaggerate these warnings. Some people use
the category "name-calling" for the use of any uncomplimentary term
applied to anyone. But it is not improper to characterize a man as a Fascist
when this word is adequately defined and there is proof that the statement is
true. When a man is called a Fascist merely because he espouses views that are
less liberal than our own, however, we exhibit intellectual irresponsibility.
Similar considerations apply to the other techniques. A "glittering
generality" is a smug generalization, illustrated by "Woman's place
is in the home," or "What is good for business is good for the
country." There is a wisecrack which says that generalizations are always
false, and statements like the ones just quoted undoubtedly ought to be
qualified. But certainly propagandists have no monopoly on the use of such
generalizations. Finally, consider the "testimonial." This may be
worthless as evidence, especially when it comes from a non-expert or
unqualified source, but the opinions of qualified persons, on the other hand,
are worthy of respect. We should always consider the source, that is, the
competence of the testifier, the probability of his being prejudiced, etc. But
testimonials are not necessarily illegitimate.
The analysis of these "propaganda techniques" calls for our attention
to the possibility that we may be permitting emotion to sway us, and we are
reminded that we ought to look into the evidence. But language alone does not
distinguish propaganda from other forms of discourse. Propagandists may avoid
these techniques altogether, and non-propagandists may use them.
What is meant by propaganda in this discussion? Its meaning may be made clearer
by contrasting the propagandist and the educator. There are some, of course,
who deny the distinction and who tell us that "the advocacy of what we
believe in is education; the advocacy of what we don't believe in is
propaganda." But this is not what most of us mean by these words. An
educator is one who, in the phrase of Robert Hutchins, seeks to teach people
how to think for themselves." An educator wants people to seek the truth.
To this end he will present them with facts; he will appeal to their reason; he
will follow an argument to whatever conclusions may be warranted by the
evidence. An educator will, or ought to, have his own point of view, his own
preferences, and he will recommend his personal beliefs to his audience, but he
will state the grounds on which he holds these beliefs, and he will state the
major objections to them. Thus his students will be able to judge for
themselves concerning the validity of his arguments and the truth or falsity of
his beliefs.
We have been speaking of an ideal educator, a truth-seeker. But a propagandist,
in the strict sense, is not interested in the truth for its own sake, or in
spreading it. his purpose is different. He wants a certain kind of action from
us. He doesn't want people to think for themselves. He seeks to mold their
minds so that they will think as he wants them to think and act as he wants
them to act. lie prefers that they should not think for themselves. If the
knowledge of certain facts will cast doubts in the minds of his hearers, he
will conceal or ignore these facts.
It may be said that there are no educators in this ideal sense and that,
really, everyone is a propagandist. "The propaganda with which we agree is
called education; the propaganda with which we disagree is called
propaganda." If we accept this notion, we are forced to deny the
distinction between tricksters and truth-seekers. This confusion of categories
may be fostered by those who are afraid of the truth and who therefore want us
to disbelieve whatever we read about them. If they can get us to believe that
"everything is propaganda," we will believe nothing, including the
truth about them. But to accept the wholesale skepticism suggested by the
phrase "everything is propaganda" is just as foolish an attitude as
to be completely unskeptical. There are two errors we ought to avoid: to be too
trusting and to be too skeptical. Some people believe everything they read in the
papers, and others believe nothing. We must learn to be discriminating-to
distinguish between what it is reasonable to believe, what it is reasonable to
doubt, and what we ought to dismiss as probably false.
To sum up. We have noted that there are three kinds of language, or purposes in
speaking, and we examined the distinction between neutral and emotive words. We
saw how emotion is employed in order to get action. Propaganda, we saw, has a
directive purpose, for the propagandist wants us to act. The question,
"What shall we do in order to protect ourselves against propaganda?"
is misleading. The question assumes that propaganda is bad, and we cannot say
that this is so without making certain distinctions. Insofar as propaganda
seeks to get us to act by emotional appeals coupled with a concealment of
facts-facts that might make us think about the merits of the proposal-it is
"bad" as a method. It is possible of course that the propagandist may
have our best interests at heart, so that his goal may be a good one.
Another reason why it is misleading to speak of "protection against
propaganda" is that this implies that there is a special kind of defense
against propaganda There is no magic amulet whereby one may exorcise its evils.
The only defense against harmful propaganda is to add to our knowledge and to
sharpen our critical abilities. We shall then know how to protect ourselves
against the various forms of hokum.
The Chinese have a proverb which says that there are three sides to every
question: my side, your side, and the right side. Though the right side is hard
to find, to seek for it is legitimate, and we should have the confidence that
we may find it if we try hard enough.
FOR DISCUSSION AND WRITING
1. This chapter describes the three functions of language: informative,
expressive, and directive. Using these terms, clarify the following passages
according to their function.
a. He ran the last mile in a blistering four minutes eight seconds, finished in
13:22.8-and, in the process, clipped seven seconds off his old record.
b. Go and catch a falling star,
Get with child a mandrake root,
Tell me where all past years are,
Or who cleft the Devil's foot,
Teach me to hear mermaids singing,
Or to keep off envy's stinging,
And find what wind
Serves to advance an honest maid. (John Donne, Song)
c. "I was obligated to hire a team and a man for the plowing though I held
the plow myself. My farm outgoes for the first season were, for implements,
seed, work, etc., $14.7232'. The seed corn was given me. This never costs
anything to speak of, unless you plant more than enough. I got twelve bushels
of beans, and eighteen bushels of potatoes, beside some peas and sweet corn.
The yellow corn and turnips were too late to come to anything." (Henry
David Thoreau, Walden)
d. Light up with a Lucky.
e. For America to survive at all in the future it must develop representative
proportional government and evolve to equitable distribution of wealth. That
is, simply, democracy and cooperative economics.
2. You have been asked to evaluate a product for a consumer protection
organization. Write an informative report describing the article, its
performance, durability, value, etc.
3. You have been asked to write a description for the product above (#2) which
will be used in a sales brochure to be used by salesmen. Write a persuasive
(directive) report.
4. In the following advertisements, classify the copy writer's use of language
according to the divisions established in this chapter. Are the distinctions
clearly delineated, or blurred? To what purpose?
a. Nearly half of the new cars sold in America last year were hard-tops. The
public, it seems, is in love with hardtops. At Volvo, we're not. As far as
we're concerned, the best way to build a safe car is to build a strong car. So
Volvos have six steel pillars holding up the roof. Each one is strong enough to
support the weight of the entire car. These pillars are part of a box
construction that surrounds and protects the passenger compartment. A Volvo's
body is fused together by 10,000 spot welds. And when you build this kind of
strength into a car's body, it holds up. Are you in the market for a hardtop?
Or is what you really want a hard top?
b. Salem uses only natural menthol, not the kind made in a laboratory. Like our
rich, full-flavored tobaccos, our menthol is naturally grown. Then we blend
natural menthol with our superb golden tobaccos. It is a unique blend found in
no other cigarette. A blend that gives Salem a taste that's never harsh or hot
. . . . A taste as naturally cool and fresh as Springtime.
c. Ever since the U.S. Government began testing the tar and nicotine levels of
cigarettes, True has been lower in both tar and nicotine than 99% of all other
cigarettes sold. It was True in 1969. True in 1970. True in 1971. And it's
still True in 1972.
d. This isn't just a new model of an old favorite. This Jeep Commando is a
whole new vehicle. Take its strength. Jeep guts make this Commando the
strongest one ever built. With a hefty 232 CID 6-cylinder engine as standard
equipment. And a 304 CID V-8 as a mighty option. That's power. They'll take you
places you never dared go before. Take its looks That new front end makes the
Commando more stylish than ever-along with the nine bright, up-to-date colors
you have to choose from. This 4-wheel drive vehicle looks at home-at home! And
take the interior. The Commando adds more of everything you want . . . . And
more luxury, too, with extra trims and options. Altogether, it's the new Jeep
Commando-the most exciting new 4-wheel drive vehicle in America.
5. In the following pairs of words, each word signifies approximately the same
meaning as its mate. However, one word has a pejorative connotation, and the
other an ameliorative or more pleasant meaning. Using a dictionary, describe
the differences in meaning.
a. satisfaction-surfeit
b. statesman-politician
c. indoctrinate-educate
d. learned-sophistic
e. foreign-alien
f. beliefs-dogma
g. innocent-naive
h. provincial-local
i. hint-insinuate
6. The following sentences are emotionally neutral. Rewrite each sentence
keeping its original meaning, but giving the sentence (a) an unfavorable
connotation, and (b) a favorable connotation.
a. Frank enjoys words, and his vocabulary reflects his enjoyment.
b. The mayor asked the protesters to allow him to speak.
c. The cost of living rose three percent last month.
d. Professor Hogins always spends three weeks discussing human reproduction
with his biology class.
e. The businessmen of Collinsville petitioned the state legislature to lower
corporate taxes.
7. Write a description of an automobile accident from these different
perspectives: from the point of view of the newspaper reporter who covered the
story, from the view of a legislator who cites the accident to support his
thesis that fifteen-year-olds are too young to drive, and from the point of
view of the fifteen-year-old driver who believes that he is innocent.
PART TWO
The Argument
CHAPTER 5
How Not to Argue
Much of your college writing will be in the form of argument--that is, a series
of statements, some of which serve as reasons for others, leading to a
conclusion based on those reasons. Good argumentative writing depends, to a
large extent, on logic. But this raises an important question: must all writing
and speech rely exclusively on logic and reason? Is there no place for emotion
or feeling? The following historical sketch presents this dilemma rather
succinctly.
In the spring of the year 399 B.C., a famous Greek philosopher was put on trial
for having committed two crimes. One was impiety to the gods of the state; the
other was the corruption of youth, by teaching them impiety. The penalty for a
conviction on these charges was a severe one, possibly death. The prisoner's
name was Socrates, and he was seventy years old at the time.
There were other reasons, political reasons, for trying Socrates. He had been
associated with the old aristocratic regime, now overthrown by the democracy,
and he was held in suspicion as a critic of the democracy. Among other things,
he became unpopular for his strange doctrine that even politicians ought to
know what they are doing.
Socrates was reputed to be the wisest man of his time. This reputation
surprised him, he said, for he considered himself to be an ignorant man,
ignorant of the answers to the supreme questions concerning human happiness and
human destiny. But he was also sure that no one else knew the answers to these
questions, and this furnished him with an explanation of his reputation as a
wise man. Though he was ignorant, he alone knew that he was ignorant, whereas
other ignorant men did not know that they were, thinking they had all the
answers.
The court which tried Socrates was made up of 501 of his fellow Athenians, the
501st man being added in order to avoid a tie vote. As Plato reports the trial
in his short masterpiece "The Apology," Socrates vigorously denied
that he had done anything wrong. He denied that he was guilty of impiety to the
gods, and he denied that it was wrong to carry on free and open discussions
with young men. He was convinced that he was right in what he was doing, and he
was sure that he could convince his judges, by logic and reason, of the justice
of his cause. Throughout his life he acted on the principle that clear thinking
is in-dispensable for right living, and that human life without the joy of
thinking is a life not worth living. Let us look at one passage from his speech
to the judges at his trial:
"Perhaps some of you," he says to his judges, "when you appeared
before the judges in a similar situation, begged and besought the judges with
many tears, and perhaps you brought your children into court to arouse the
compassion of the judges. But I will do none of these things, though I am in
peril of losing my life. I too have children, but nevertheless I shall not
bring any of them into court to beg you to acquit me. This is not because I am
stubborn, my fellow Athenians, or because I lack respect for you, but because I
think it disgraceful for respected people to act in that manner. But, apart
from the question of reputation, gentlemen, I think it is not right to implore
the judge or to get acquitted by begging; we ought rather to inform the judge
and convince him. For the judge is not here to grant favors in matters of
justice, but to give judgment; and his oath binds him not to grant favors
according to his pleasure, but to judge according to the laws."
Socrates, in other words, used a rational and logical approach in presenting
his case to the jury. This was the speech of a rational man. Socrates presented
the evidence, and refused to indulge in an emotional harangue in his own
defense. In other parts of his speech,. however, he did present an eloquent
defense of the free pursuit of truth, and he also goaded his accusers with
withering sarcasm. Socrates' spirit of reasonableness and, incidentally, his
sense of humor, did not desert him even after he heard the sentence of the
court that he be condemned to die by drinking a poison made of hemlock. An
anecdote bears witness to this. After the verdict of guilty was returned, one
of his disciples, Apollodorus, who was at the trial with him, exclaimed,
"But what I find it hardest to bear, Socrates, is that I see you being put
to death unjustly." Socrates replied, "Was it your preference,
Apollodorus, to see me put to death justly?'.
Logic or emotion?
The question may be raised: Was Socrates' rational approach the proper one
under the circumstances? Should one use a "logical" approach, or
should one use an emotional appeal in presenting a case to a jury? We are all of
us familiar with that favorite of the cartoonists, the glamorous blonde witness
who sits in the witness chair with her beautiful legs crossed provocatively.
This, one might say, is the proper technique for presenting a case to a jury,
especially when the jury is an all-male one.
Before we try to answer our question, it will be instructive to compare
Socrates' approach with that of Mark Antony (see the example). The funeral oration
was an emotional appeal. Antony achieved the goal he sought. Now, can one say
that Socrates used the wrong technique and Antony the right one? If Socrates
had had you as his lawyer, would you have advised him to appeal to the emotions
of the court, consisting, as it did, of a large crowd? It all depends, of
course, on what one is after. The question, "Did Socrates use the proper
approach?" is somewhat ambiguous. "Proper approach" has at least
two meanings. "Proper" may refer to an efficient technique for
achieving a goal, whatever it may be, or it may mean what we ought (in a moral
sense) to do.
Taking people as they are, and desiring to mold them to ones purpose, an
emotional appeal may be more effective than a rational one. Unscrupulous
demagoguery may get results. One might wish it were other-wise, but in the real
world, as distinguished from an ideal society, such is often the case. But talk
also has a moral aspect. There is a moral obligation to tell the truth. And
there is also the clement we call honorable conduct." Though few would
have condemned Socrates if he had obtained an acquittal by an appeal to the
court's emotions, he preferred death to what he considered dishonorable
conduct. His sense of moral integrity did not permit him to compromise with his
principles, and he became a martyr to the principle of unswerving devotion to
the truth. To the truth, that is, as he saw it.
Thus the question, "Who used the proper approach?" depends upon the
meaning we give to "proper." But there is also a larger issue. Logic
has its proper place, and so has emotion. A purely intellectual approach to
life is as insufficient as a purely emotional one. The activities of life may
be divided into two broad categories, the logical and the nonlogical. By
"nonlogical" we do not mean 'illogical," but rather activities
that have nothing to do with logic. There are times when we reason and argue
and draw inferences. But, for the most part, we are engaged in nonlogical
activities, like eating and sleeping or narrating the events of the day, and so
on. Logic enters only when we give reasons for our beliefs.
When we give reasons for our beliefs, we are reasoning. Reasoning is either
logical or illogical. Illogical reasoning is bad reasoning, but the nonlogical
has nothing to do with reasoning. When we seek to prove that something is or is
not the case, then we engage in argument, in which we say: This is true because
that is true, or this is so because that is so. When the reasoning is adequate,
we say it is logical, when not, illogical.
The law of rationality
When we assert beliefs which may be questioned, then we have an obligation to
be rational. This common human obligation may be stated in the form of a
"law of rationality" or "law of argument," that we ought to
support our beliefs by adequate evidence. When we say that we know that
something is true, we ought to be able to justify our belief by adequate
evidence. What is adequate evidence? This term is best defined by example, and
we shall give examples as we go along, but we shall assume here that we agree
pretty well as to the distinction between evidence that is good and sufficient
and that which is not. In the end there is only one court of final appeal in
settling a problem concerning what is rational and what is not-the community of
reasonable men. Fortunately, the human race has always agreed pretty well on
which of its members are reason-able and which not.
Logic is not all, then, but we have a common obligation to be logical when
logic is relevant. In the last chapter we saw how emotional language is used to
get action, and we raised the question as to when emotional appeals are
appropriate and when not. An emotional appeal under circumstances like those
portrayed by Shakespeare is highly improper, for the mob did not know what the
facts were. They came believing that Caesar was an evildoer, and they left
convinced that Brutus was. But they did not revise their judgment on the basis
of evidence. When the facts' are in dispute, we ought to demand information and
evidence rather than emotion. The Roman mob thus fell short of the human
obligation to be rational. Before acting they should have asked themselves some
questions concerning the facts: Did Caesar really aspire to be a dictator? Did
Brutus seek to save their freedom? What evidence is there for or against the
is-sues involved in these questions? But Antony foreclosed the inquiry by
substituting emotion and passion for reason. And when we act on emotion without
concerning ourselves with the facts, we are likely to rush into disaster.
Usually, when a politician does what Antony did, that is, when he substitutes
emotional appeals for proof, propaganda for rational persuasion, when he
inflames rather than informs, we shall find that he does so for one of two
reasons. Either he has a contempt for the people, treating them as if they were
children, incapable of understanding the issues, or he doesn't want them to
know the truth.
We are not saying that emotional appeals are never appropriate. On the
contrary. When the facts are not in question, and action is desired, then an
emotional appeal is appropriate, even indispensable. In the critical days
during the "Battle of Britain," Prime Minister Winston Churchill made
his great 'blood, toil, tears, and sweat" speech to the British people. He
inspired his people and spurred them to heroic efforts. Emotion is the best
fuel for this kind of energy, and this kind of stimulus is needed even the best
of causes.
How not to argue
Let us pause for a moment to get our bearings. The law of ration-is the central
core of the rational approach: When we assert that a belief is true, we should
be prepared to support our belief by adequate evidence. But the law of
rationality is frequently violated, and it may be evaded. All of us hold many
beliefs that are unsupported by evidence, and we sometimes argue as if evidence
were unnecessary. One of the most common violations of the law is one that we
have been discussing in this chapter: to make an emotional appeal at a time
when evidence is required. We have not condemned emotional appeals under all
circumstances, but only when we substitute emotion for proof when proof is
required. The latter form of behavior is the essence of what is meant by
"How Not to Argue."
Students of logic were provided many examples of the use of emotion substituted
for evidence during the 1950s (the "McCarthy era"), when public
officials were often attacked for grotesquely irrelevant reasons. In some
instances, they, their parents, or wives had been born in a
Communist-occupied country; in others, they were attacked because friends or
associates had once belonged to organizations which numbered among their
members individuals who had belonged to the Communist party. Such tactics were
attempts to divert attention from the is-sues at hand.
Appeal to emotion
The appeal to emotion sometimes takes the special form called "the appeal
to laughter." If one is unable to refute an opponent's arguments by
evidence, it is always possible to make him the butt of a joke and thereby
evade the necessity of presenting evidence. A notorious example of sort of
thing, which apparently misfired, occurred in a celebrated debate over the
theory of evolution in 1860. Bishop Wilberforce scored when he asked Thomas
Huxley, who was defending the Darwinian theory, whether it was through his
grandfather or his grandmother that he claimed descent from a monkey? Huxley,
who was in no mood to appreciate the Bishop's humor, retorted that he preferred
descent from monkey to descent from a man who used his great gifts and versatile
intellect to distract the attention of his hearers from the real point at issue
by eloquent digressions and skilled appeals to prejudice.
We have been discussing bad logical behavior on the part of speakers and
writers who try to divert our attention from the need for evidence by working
on our emotions. They fool us in this way. But we also fool ourselves. We
rationalize; we engage in a wishful thinking"; we may accept unfounded
beliefs because they satisfy us emotionally. For example, do we find ourselves
saying, "I believe thus and so because it makes me feel good so to
believe"? Or do we say, "I must believe as I do because I just
couldn't bear to think my belief false"? We deceive ourselves if we
believe that our emotions guarantee truth. A beautiful passion, which makes its
object appear not only handsome or beautiful, but also good, reliable, noble,
and intelligent, really guarantees nothing of the kind. Perhaps it is not
wholly undesirable to use a little logic even in love.
When one says, 'This must be true because I feel so strongly about it, and if
it were not true I could not feel as I do," he may be misleading himself.
For alas, wishes are fathers to thoughts that just aren't so. The fact that we
want something very strongly apparently does not guarantee that it will come
our way. This would be a much nicer world, of course, if our wishes could make
things come true. There would then be no broken hearts, unsatisfied ambitions,
or even lack of the wherewithal to own yachts, including the cost of the
upkeep.
Our emotions, in other words, may interfere with our logic and prevent us from
seeing the truth. This is why we fail to see ourselves as others see us.
"I am firm; you are stubborn; he is pigheaded." And do you know women
whose attitude might be expressed in the following way: I am beautiful; you
have quite good-looking features; while she isn't bad looking, if you like that
type"?
It is our emotions that make us adopt a double standard of intellectual
morality--one for ourselves, another for the other fellow. The Democrats are
naturally enraged when they are unjustly attacked by Republican speakers; the
Republicans may candidly acknowledge that the spokesmen of their party have
been guilty of some exaggerations, but they will say that such attacks are
justified in political debate. The Republicans, on 'the other hand, find unjust
Democratic charges unbelievably vicious, while the Democrats will say of their
own speakers that, though they may have made it a little strong, nevertheless
all's fair in love and politics.
And then there are those who think that they have transcended their emotions,
who "see both sides," but all too often what they tell us is, in the
words of an unknown poet,
In matters controversial,
My perception's rather fine.
I always see both points of view,
The one that's wrong, and mine.
So much for one of the major ways in which we evade what we have called the law
of rationality" or the law of argument." We should aim to support our
beliefs by adequate evidence. The form of the evasion we have been discussing
is called the "appeal to emotion." There are of course many other
ways in which the law of rationality is evaded. To draw an analogy from ethics:
Aristotle once said that good men are good m one way, but the evil are evil in
many ways; that is, good men resemble each other in their actions, but there is
great variety in wickedness. Perhaps that is why we read so much more crime
fiction than stories about virtuous men. Variety is more interesting. Aristotle
also uses the image of the archer shooting his arrow at the target: there is
just one way to get a bull's-eye, but many different ways in which to miss.
So with arguments. A good argument must hit the point exactly, but there are
many ways in which we can miss. Logicians have catalogued many types of errors
of reasoning, but it would be impossible to list every possible kind of error,
for there are an infinite number of ways in which we can miss the target. In
arguments, too, it sometimes seems as if the archer has turned his back on the
target and shot in the opposite direction!
Argumentum ad hominem
Let us now consider another major evasion of the law of rationality, the
"argumentum ad hominem." This term, from the Latin, means "an
argument directed to the man." To the man, that is, as distinguished from
the point at issue.3 For example, let us suppose that we disagree with what a
speaker says. Now, we may try to disprove what he says by presenting the
evidence. Instead, we simply verbally attack the speaker to cast doubt on his
statement. Such attacks are often on the behavior, motives, family, dress, or
other characteristics of the person who disagrees with us.
If we believe that a statement is false, we ought to attack the statement, not
the man who utters it. A speaker, let us say, attacks the so-called
"right-to-work" laws, which forbid compulsory union membership. He
argues that the law unfairly discriminates against unions) on the ground that
workers who benefit from union activities ought to pay for these benefits. Now,
if you disagree with the speaker, you should support the position that the law
does not unfairly discriminate against labor. But suppose, instead, you say to
the speaker, "By the way, you re a union man, aren't you?" The
question implies that the speaker's views must be false, on the ground that his
union membership makes him so biased and prejudiced that it would be a waste of
time to take his remarks seriously-they simply must be false.
As another example of this sort of thing, let us examine some remarks made by
the German philosopher Arthur Schopenhauer in his "Essay on Women."
In reading what follows, it may be helpful to remember that Schopenhauer was a
pessimist who believed that life is a painful and very sad affair. here are a few
lines from the essay:
It is only the man whose intellect is clouded by his sexual impulses that
could give the name of the fair sex to that undersized, narrow-shouldered,
broad-hipped and short-legged race: for the whole beauty of the sex is bound up
with this impulse. Instead of calling them beautiful, there would be more
warrant for describing women as the unaesthetic sex. Neither for music, nor for
poetry, nor for fine art, have they really and truly any sense or
susceptibility; it is a mere mockery if they make a pretense of it in order to
please. Hence, as a result of this, they are incapable of taking a purely
objective interest in anything.
Schopenhauer continues on and on in the same vein. He tells us that women are
interested only in acquiring husbands and that to this end they develop their
real interests-in cosmetics, in clothing, and in jewelry, to the exclusion of
all higher interests.
Perhaps at this point I should emphasize, as strongly as I can, that I happen
to disagree with Schopenhauer. I am using this example for illustrative
purposes only. The point is this: How do many women react to these remarks? Do
they present evidence to disprove what he says, as required by the fundamental
law of rationality? No. Rather, they attack Schopenhauer himself, with remarks
like these: "That fellow must have ad very little success in his love
life"; "He must have been refused by very woman he proposed to";
"He must have been psychologically frustrated, and suffered from an
anxiety neurosis"; "He should have been psychoanalyzed."
Now, these remarks give us an example of the argumentum ad hominem, for the
argument attacks the man instead of disproving what he says. But even an
unpleasant fellow like Schopenhauer may be stating the truth, so if you
disagree with what he says, present the evidence. It is not true, for instance,
that all women have short legs.
Perhaps we ought not to take Schopenhauer's remarks so seriously, for they are
of the nature of an emotional diatribe. His purpose may have been merely to
express his splenetic feelings. The point is that a man's statements are
logically independent of who the man is, or what he is, and that we do not
disprove what he says by raising doubts concerning his parentage. Logically, a
statement stands or falls on its own merits, regardless of who makes it. Truth
and falsity are determined only by evidence. Personalities do not determine
logical issues, and discussions should not degenerate into name-calling.
The reader may have a question at this point: Is it always wrong to attack the
speaker personally? Is it wrong to cite the speaker's history, background, and
associations in order to discredit what he says? This question requires a
distinction between the argumentum ad hominem and a different kind of attack
against a speaker. A witness testifies in a courtroom and the other side
believes that he is lying. If they cannot directly disprove his alleged
eyewitness testimony, then they will seek to attack his character. They show
that he was once convicted of perjury, after testimony in another trial. Now,
what effect should this have on a jury? The jury learns that the witness once
lied under oath, but this does not prove that he is lying now. He may be
telling the truth. To say that his conviction for perjury proves that he is
lying now would be to commit the argumentum ad hominem. But, though there is no
proof that he is lying now, he has been shown unworthy of trust, and the jury
should therefore refuse to give much weight to his testimony.
In other words, it is quite legitimate to show that a speaker is unworthy of
trust, or that he is prejudiced, or biased, or that special interests have paid
him to say what he is now' saying, that he is insincere, and so on. "what
you are," we say, "speaks so loudly that it is difficult to hear what
you are saying, even though what you are saying may be true." The
important thing, however, is that we should clearly distinguish between
convicting a speaker of prejudice, on the one hand, and disproving what he has
specifically said, on the other. We customarily give a speaker our trust and
faith; we assume that he is telling the truth as he sees it. But if. the
speaker has open or concealed affiliations, such as paid or unpaid connections
with propaganda organizations, or other special interests, which make it
impossible for him to tell the whole truth, then we should not give him our
trust. He may be telling the truth, but we should not rely on anything he says
merely because he says it, since we do not believe in his sincerity.
The history of warfare shows that every new offensive weapon encourages the
development of new defensive weapons. The same is true of arguments. The ad
hominem is an attack, and this attack often calls forth a counterattack.
Logicians call this counterattack, or defensive weapon, the "tu
quoque." Translated into less dignified language, this means "You're
another." This counterattack is appropriate only when one has been
unjustly and irrelevantly attacked with an ad hominem. Here is a simple example:
A man in his forties argued that the drafting of men into the army was
desirable, since it would make the United States ready for any emergency m the
dangerous world situation. A young man, instead of trying to prove that the
draft law was unnecessary, used the ad hominem attack. He said to the speaker,
"You favor the draft because you are past the draft age and won't have to
serve." This ad hominem approach calls for an obvious tu quoque. The older
man replied, "By the same token, the only reason you are against the draft
is because you are afraid you will have to serve." But the real question
should have been, "Is the draft in the best interests of our country? The
answer to this question does not depend on who says what. An attack against the
speaker proves nothing concerning the merits of what is being discussed.
There is a variant of the ad hominem which furnishes a useful clarification of
the tu quoque. Speakers sometimes try to discredit theories by calling them
old-fashioned. Now, "old-fashioned" may be a devastating criticism in
the field of the exact sciences when experiments have disproved an old theory.
But in the field of social and political ideas, most theories have some merit,
regardless of their age. Let us assume that a speaker wishes to refute the theory
of "free-enterprise." "'That idea is old fashioned," he
says, "It goes back to the eighteenth century." Two answers are
possible. One may point out that this "criticism" is like the ad
hominem approach, for the date of a theory is irrelevant to its validity. Only
evidence can disprove a theory. Or one may use the tu quoque here. "If the
theory of free-enterprise is discredited because it goes back to the eighteenth
century," we may say, "then the theory of government regulation is
even more conclusively discredited, for it goes back to the seventeenth
century."
One of the rules of a good discussion is that the participants should stick to
the point Their remarks should be relevant to the matter at hand. In the
following pages we shall be concerned with relevance in the sense of logical
rather than causal connection. For example, suppose that I should say that the
great majority of Americans enjoy, on the average, the highest living standards
of any people in the whole world. An objector says that my statement is false
and that he can disprove it. He points to the fact that there are many people
in the United States who are quite poor, having yearly incomes of less than
$4,000 per year. Now, the facts cited by the objector are logically irrelevant.
They are not to the point. Why not? Let us grant the truth of the facts he
cites. There are many poor people in the United States. But I did not say that
every American was well off." Nor did I say that there were no poor in the
United States. I said that the majority of our people have the highest living
standards, on average, in the world. To disprove my statement, the objector
would have to cite a country in which the masses of the people enjoy a higher
living standard. There may be such countries, but it is irrelevant to point to
poor people in our country as a disproof of a statement which refers
comparative average living standards.
To argue in the manner of the objector involves a slipping away from the point.
The objector has committed a "diversion". This practice is also
refereed to as "drawing a red herring across the trail of an
argument."
A speaker or writer may also slip away from his own point and thus create a
diversion. This usually happens when he has undertaken to support a difficult
case. He may seek to create the impression that it is sufficient if he proves a
point that can more easily be established. But the latter may represent a quite
different issue. For example, a senator may speak in behalf of a bill requiring
"100 percent parity prices" for farmers He may dwell at length (or
even exclusively) on the economic suffering of the farmers before the days of
government-supported farm prices. But instead of answering the question,
"Is 100 percent parity desirable? he answers one that may be stated as
"Is some form of government aid to farmers desirable?" It is much
easier to support the latter question than the former. When speakers set out to
prove the obvious, look for a diversion
Diversions are traps for the unwary, and one should be constantly on the alert
for them. Let us imagine a conversation between a pacifist and a non-pacifist.
The pacifist argues that "all wars are morally evil, no matter for what
purpose they may be fought." "No war," he goes on, "is ever
justified, so we should sternly refuse to distinguish between just wars and
unjust wars. It is always wrong to kill a fellow human being." His
opponent then asks him whether it would be immoral for a man to fight in
defense of his country in the event his country is invaded. To this the pacifist
responds. "I can assure you that no one is going to invade us." His
opponent then proceeds to argue that there is a real danger that their country
may be invaded.
Let us analyze what happened here. The pacifist was asked whether his
principles require that a man should refuse to fight to repel an invader. The
pacifist slipped away from the point when he said there will be no invasion.
This was a diversion. But his opponent did not notice it and fell into the
trap, arguing the irrelevant issue as to whether there might be an invasion,
and the original question was forgotten.
The diversion sometimes takes the special form called "extension. A
speaker says that some corporation executives sympathize with the aims of
organized labor. A critic argues that it is false to say that all corporation
executives are friendly to labor. But the critic is not attacking the speaker's
statement, which said "some." The critic has extended
"some" into "all" and attacked his own extended version of
the original statement. Similarly, if the issue in debate were, "Are all
corporation executives opposed to labor unions?" it would be irrelevant to
prove the falsity of the statement that all executives are friendly. For even
if it is false that all executives are friendly, this would not necessarily
mean that all were unfriendly. Some are and some are not. And the fact that
some are unfriendly does not prove that none are friendly. Debaters use this
trick of extension because it is so much easier to disprove an extreme statement
than a moderate one.
Begging the question
So much for one type of irrelevancy in argument. But before we discuss another
type of failure to "stick to the point," we shall discuss an error
which may be called an "overdoing" of this business of sticking to
the point. This new error, in fact, is opposite in form to the error of
slipping away from the point. In this new error we never get away from the
point. We merely repeat it, over and over again. Examples: "why do I say
that every human being believes in God? How do I know that? The proof is that
the belief is universal in the human race." Now, this is sticking to the
point with grim tenacity. But it proves exactly as much-no more, no less-as we
prove when we argue that a rose is a rose. The speaker says that everyone
believes in God, and when asked for proof tells us that everyone does.
"How do I know that it is so? Because it is so." In an argument a
reason should be given for a belief, and the reason should be a fact from which
the belief can be inferred. But in this "argument" the reason (or
proof) is exactly the same as the original belief. The reason ("'The
belief is universal in the human race") merely repeats the original
statement in different words. This is not proof. This kind of
"reasoning" is on the level of the child's response to the question
"Why?" The child responds with a "because" and nothing
more. The demand for a reason is unsatisfied.
This fault in reasoning is called 'begging the question." Its older Latin
name is "petitio principii." The error consists in our pretending to
prove something when actually we assume, in the "proof," that which
we are supposed to prove. "Why do I believe that the Chinese can't be
trusted? Because they can't be." (Am I doing all right or am I doing all right?)
Now, no formal logical error is committed in these examples, for we do not
infer conclusions unwarranted by the facts. We infer no conclusion at all-we
merely make an assertion. If the belief in God is universal, then it is surely
the case that everyone believes in God. But the repetition of a belief is not
the same as proof that it is true.
Here is another simple example. A guardian (self-appointed) of the public
morals tells us that it is morally wrong for topless dancers to perform in
public bars. We ask, "why is it immoral?" and receive the answer,
"Because it isn't right." The answer begs the question. It repeats,
in different words; what was supposed to be proved. And if we now asked,
"But why isn't it right?" the answer we should expect to receive
would be, "Because it's wrong."
When we beg the question, we make a pretense of proving a point but actually
merely repeat it. Nietzsche once said that all mankind was corrupted, and when
challenged for a proof answered, "The mere fact that you disagree with me
is in itself proof that you are corrupted." And here are two passably
humorous illustrations. The first is a little story about two men who
approached a teller in a bank One of them wished to cash a check. The man was
unknown to the banker, who asked if some-one could identify him,
"Yes," the man said, "my friend will identify me. "But I
don't know your friend, either," the banker objected. "Oh, that S all
right," the man replied, "I'll be glad to introduce you to him."
The second illustration, a very old tale, is about two medieval Jews who were
engaged in a dispute concerning the respective spiritual gifts of their rabbis.
To clinch his case one of them said, "And now I'll give you proof positive
that my rabbi is the most wonderful rabbi in the whole world Is there another
rabbi who dances with angels every night after he falls asleep?" The other
was somewhat skeptical of this. "But how do you know," he asked,
"that your rabbi really dances with angels?" "Why," replied
the other, "because he told us so himself." "But can you believe
him?" "What!" the other retorted in indignation, "would a
rabbi who dances with angels every night tell a lie?"
"Reasoning in a circle" is a "drawn-out" form of begging
the question. It contains intermediate steps. A man says that classical music
is better than modern music. Challenged for proof, he answers, "The best
critics agree that this is so." Who are the best critics? "Those who
prefer classical to modern music." And here is a more complex example: The
founder of a new religion tells his followers that he is inspired, so that they
may believe whatever he tells them. Now, in the unlikely case that he should be
challenged for proof of his inspiration, he might answer, "Because I am
inspired." That would be the simplest form of begging the question. But if
he "reasoned in a circle," the argument might go like this: "Why
do I say that I am inspired? Because here. is a book which says that I speak in
God's name. Why should we believe this book?" he is asked. "Because it
comes from God," he answers. 'How can we know this?" "Because
you can take my word for it." "And why should we take your
word?" "Because I am inspired." If we now ask, "How can we
know you are?" the circle will start all over again.
The so-called "argument by definition" is a special form of begging
the question. Jones asserts that all Christians are good men. Brown disagrees
and points to Thwackum, who is a Christian but whose conduct falls very much
short of that of a model of virtue. "Ah," answers Jones,
"Thwackum may attend church regularly, but he is no real Christian, for if
he were, then he would be a completely virtuous man. I reiterate, all
Christians are virtuous men." This argument begs the question. What Jones
meant was this: "I define a Christian as a good man. Thus I can assert
without fear of contradiction that every Christian [defined as a good man) is a
good man." Jones' original statement cannot be proved false, for it is not
a statement about facts but a "stipulative definition." A stipulation,
insofar as it is nothing but a declaration of intention as to how a word is
going to be used, is neither true nor false. The question, "Is this
statement true?" can be raised only with respect to assertions that
purport to describe facts. Observation or experiment may confirm or disprove
factual statements-they are true or false-but we cannot raise the question of
truth or falsity concerning a man's declaration of his intention to define a
word in a certain way. If Jones had said that every churchgoer is faithful to
his spouse, that statement would be true or false. But Jones merely tells us
that he is going to use the words "Christian" and "virtuous
man" interchangeably. By definition, then, no bad man can possibly be a
Christian. (If we stipulate a definition of a square as a four-sided figure, it
is senseless to ask, "Can a square have more than four sides?")
Jones' statement, that "all Christians are good men," is thus
tautological. He is saying that all good men are good men. But this is not what
Brown took him to mean. Brown understood him to say that a Christian, defined
as a member of a church that worships Christ, can always be depended on for his
trustworthiness. This might be false. But a definition taken as a stipulation
concerning word usage cannot be false.
And here is one more variety of question-begging: the "question begging
epithet." The previous varieties were cast in the form of arguments, with
a pretense at giving proof. Repetition was substituted for evidence. 'Our new
variety merely uses epithets. For example, expressions like "the stupid
conservative point of view," or "wild-eyed radicalism," contain
question-begging epithets which assume something that may require proof,
without even a pretense at proof. Instead of proving, first of all, that
Moriarity is a crook, we ask, "What do you think of that crook?" A
ready-made conclusion is put into the hearer's mind. This variety of
question-begging is perhaps more dangerous than the other forms, for it
operates on our reluctance to question a positive assertion that is
"unquestionable," especially when asserted in a strong manner. The
intimidated listener may also be at a loss to know just what it is that he
should question.
Argumentum ad ignorantiam
Another mistake commonly made with respect to the burden of proof in an
argument is called the "argumentum ad ignorantiam." As the name
suggests, this means an argument based on ignorance, or on an appeal to our
lack of knowledge. For example, a man states his belief that God selected Mohammed
as the final prophet of His Word. When challenged for a justification of this
belief, he asks: "Can you disprove it? If you deny what I say, it's up to
you to disprove my statement." Or let us say that an atheist denies the
existence of God. When asked for evidence; he answers: "Can you prove that
God exists?" But just as failure to disprove is not proof of the opposite
(the Mohammed example) so failure to prove is not equivalent to disproof. It
all depends on who has the burden of proof. If a man makes the claim that God
does not exist, then the burden is on him to prove his point. Atheism is quite
different from agnosticism, which merely says, "I don't know."
Similarly, the per son who argues that God does exist has the burden of proof
for that thesis.
This, then, is the appeal to ignorance. Instead of proving a statement by
positive evidence in its favor, we appeal to the fact that our opponent hasn't
disproved it. But the law of rationality tells us that we should furnish
positive evidence for our beliefs.
An amusing variant on the ad ignorantiam argument is the old story about the
justice of the peace who heard a case concerning a man who was accused of
stealing a horse. Two witnesses testified that they saw the accused unhitch the
horse from a post and lead it away, but five witnesses testified that they had
not seen the defendant steal the horse. The J. P. said that since there were
more people who had not seen the alleged theft than there were persons who
claimed to have seen it, and since he believed that the majority were always
right, he was forced to dismiss the case.
"Either-or"
Let us consider a final example of "how not to argue": the
"either or" fallacy. These two little words "either" and
"or" make trouble for us unless we watch them carefully. They are
useful tools in thinking, but a great deal of bad and confused thinking falls
into an either-or pattern. A The worst of these vices-the error of insufficient
options-is the assumption that there are only two possibilities in a situation,
or only two choices, when there are in fact more than two. This can be a very
serious matter, indeed. Consider, for example, an application of this
assumption to our international problems. There are people who say,
"Either other nations are for us or they are against us; either they will
take sides with us against our enemies or they will take sides with our enemies
against us," But this ignores the possibility of neutrality as in the case
of Switzerland, or of other spheres of influence, as in the case of the 'Third
World nations that have coalesced around economic and cultural interests in
recent years.
Similarly, it is not the case that I am either for you 100 percent or I am
against you." I may be for you 99 percent, or I may be neither for you nor
against you. There is a middle ground between love and hate called
indifference, just as between indifference and hate there is a region called
dislike. We cannot say of a man that he is either an angel or a devil, either a
god or a beast. As Aristotle said long ago, we are in between, superior to the
beasts and inferior to the gods. Let us stop using just two categories for
people: the saints and the devils. Many of us are middle-of-the-roaders.
The kind of thinking we have just described is sometimes confused with a law or
principle of logic called the "law of the excluded middle." This law
tells us that a thing either has a particular characteristic or it does not
have that particular characteristic: A man either has a million dollars or he
does not; he either owns a home of his own, or he doesn't. Anything, the law
tells us, is either A or it is not A; either it has characteristic A or it does
not. These alternatives do exhaust the possibilities; the options are
sufficient. Note that the examples of "insufficient alternatives" we
considered above do not illustrate the law of the excluded middle. It is false
that a car's color must be either red or yellow; it may be black or blue or
green. But the law of the excluded middle says only that a car is either red or
it is not red. A chemist analyzes a solution to determine the presence or
absence of arsenic. Either at least one molecule of arsenic is present or no
arsenic at all is present. This is in accordance with the law.
The most interesting and troublesome problems involving the use of
"either-or," however, concern a type of application in which the law
of the excluded middle is not violated-the two alternatives actually do exhaust
the possibilities-but in which the "either-or" may be misleading or
unrealistic. This type of thinking requires a more extended discussion. One of
the most, striking illustrations of this sort of thing is found in the bad
habit of thinking which 'we shall call "moral perfectionism." The
perfectionist sets up an ideal or standard of moral perfection, and then thinks
in terms of only two alternatives, "Either you are a good man, or you are
not." Though the perfectionist does not always say so, he usually implies
that he himself has achieved perfection and that the rest of the world has not,
and he judges everyone else as falling short of the standard. The perfectionist
thinks of every man as falling into one or the other of the two classes, the
good and the not-good.
There are many illustrations of this kind of perfectionism. To the ancient
Stoics there were no degrees in vice; you were either perfectly virtuous or you
were not virtuous at all. The theft of a piece of firewood, they said, is just
as much a violation of the moral law as ruining a man in a swindle. As one of
them put it, "The man who is a hundred miles from Canopus, and the man who
is only one mile from Canopus, are both equally not in Canopus." A similar
thought is expressed in the New Testament, in the words of James: "For
whosoever shall keep the whole law, and yet offend in one point, he is guilty
of all." In other words, for the perfectionist you are either good or you
are not good-and there is no middle ground.
This kind of thinking permits of no compromise. Either you fulfill the moral
law completely or you do not fulfill it completely. Either you are perfectly
good or you are classed undiscriminatingly with all the backsliders. And herein
lies the most serious fault of this attitude. Though the perfectionist says,
"all-or not-all," in practice he means, "all-or nothing,"
for the man who falls just short of perfection is regarded as being in the same
class as the most vicious and hardened criminal. The perfectionist is not
interested in the degree in which one falls short.
The perfectionist attitude is found in all sorts of places. Consider the
perfectionist's attitude toward philanthropic benefactions. Either .a person is
a "perfect giver," they say, or he is not. A perfect giver always
gives anonymously, for if it is known that he has made a contribution, then we
must assume that he expects to receive applause for what he has done. And if he
desires public acclaim, then he deserves no credit for giving, for he didn't
give solely for the joy of giving. Those who don't give at all are apparently
no more ungenerous than the donors whose names appear in the lists. But surely
we should distinguish between the man who gives, hoping for some expression of
gratitude, and the man who refuses to give anything at all.
So much for perfectionism, which uses the "either-or" to formulate
unrealistic dichotomies. Let us now restate the three types of usage of the
"either-or" formulation that we have considered. There is first the
"either A or B" statement, in which one says, "He is either a
member of the bourgeoisie or he is a member of the proletariat. He must be one
or the other." This is the error of "insufficient options." He
need not be either; there are other possibilities. This type of usage is
characteristically at variance with the facts. The second usage is in the form
of the law of the excluded middle: "Either A or not-A." Here we can
properly say, "It must be one or the other, A or not-A," for these
exhaust the possibilities; either he has a particular characteristic or he does
not have it. We then drew a distinction between realistic and unrealistic
applications of the law of the excluded middle. "Either A or not-A,"
we noted, is quite adequate where degrees are not involved, as when the chemist
says, "Either arsenic is present or it is not present." But when
degrees are involved,' then the expression "either A or not-A" may be
unrealistic and misleading. This is the third type of usage we discussed:
"He is perfect or he is 4 not perfect" Now, this statement is in
accordance with the principle of the excluded middle, and so not incorrect when
we understand it is a precise formulation concerning those who have a
characteristic and those who do not. Nevertheless, the statement is misleading
because the important thing about human conduct is the degree with which
conduct approaches a standard of perfection. "He is either perfect or
not-perfect" makes us lose sight of the degrees of imperfection.
Let us look at some further illustrations of the application of
"either-or" to cases involving degrees. Where there is a continuum of
degrees-as in rating the intelligence of human beings-the subjects do not
divide into two sharply contrasted opposites: the intelligent and the
unintelligent. It would be misleading to use the principle of the excluded
middle here, to say, "Every human being is either intelligent or
unintelligent," as if there were just two classes, into one of which every
human being falls. Or at least this would be a very arbitrary thing to do.
But-and this is the other side of the coin-there are occasions when it is
necessary to make such arbitrary distinctions between the two classes, the
intelligent and the unintelligent, as in the Armed Forces Qualification Tests.
It may even be necessary to draw a sharp and arbitrary distinction between sane
and insane, as in a trial for murder. Our criminal law draws a distinction
between offenders who are sane and those who are not, on the principle that we
ought not to treat the two classes in the same manner. And judges need
clear-cut definitions, or at least some kind of arbitrary dividing line. In the
state of Illinois, for example, the statutes define sanity as the ability to
distinguish between right and wrong. Though most psychiatrists consider this
definition inadequate, it is at least workable, in a rough sort of way, and gives
us an arbitrary dividing line.
Or consider the matter of academic grades. In our schools we use an
"either-or" for passing or not-passing. Passing means that a student
has mastered the subject matter to the required degree. Thus we can say of any
student, "Either he has mastered the course or he hasn't." But, you
may protest, this is unrealistic, for mastery is a matter of degree, and it is
wrong to divide all students into only two classes. The differences between
students lie in a continuum of almost imperceptible differences in degrees.
Let us say that a grade of 65 percent is set as the minimum passing grade. But
how much difference is there, after all, between the student who makes
sixty-five and the one who makes sixty-four? The latter fails the course, and
it seems grossly unfair that a 1 percent difference should have such enormous
consequences. Shall we pass the sixty-four percenter? But now, the sixty-three
comes to claim equal justice. We pass him too. The logic of this procedure is
that we must finally pass even the person who makes zero, for zero is only 1
percent below one. The grades. from zero to 100 are connected by small
graduations in degrees. But school administrations need a dividing line, and
they arbitrary set it at 65 percent or some other definite figure. If there
were no such dividing lines, there could be no definite standards of
competence. There could be no bar examinations or C.P.A. examinations. Though
we recognize the dividing line as arbitrary, still we must have one. And even
the kindest-hearted teacher, who passes the sixty-fours and the sixty-threes
and the sixty-twos, must draw a line somewhere. Certainly the twelves and
thirteens should not pass.
There are two extremes to avoid with respect to the continuum. One is the danger
of making too sharp divisions in reality, of thinking that all people can be
divided into just two classes-into capitalists and Communists, or into the good
and the not-good. The other extreme is to deny the existence of all
distinctions simply because one class passes into another by imperceptible
degrees. This leads to a fuzzy-mindedness, which says that the good are really
not-good, for they are connected by imperceptible degrees to the not-good, or,
conversely, that the not-good are really good, for exactly the same reason. In
making the first error we think in terms of two sharply divided natural
classes; in the other we see no classes at all. Justice Holmes once referred to
the question, "where are you going to draw the line?" as "the
tyro's question." We must draw lines, he added, for all life involves
"the marking of grades between white and black." There are classes of
things even though they merge into each other by imperceptible degrees. The
mistake we too often make is in thinking that there can be only two.
There are times when we must make decisions, and the etymology of the word is
instructive. It comes from the Latin, meaning "to cut off." We make a
sharp break when we decide, for we must decide in one way or another, no matter
with how many qualifications. In an election we must decide whether an
officeholder has or has not satisfied our standards. Though our minimum
standards of decency are vague and ill-defined, they are there. There are times
also when we cannot evade a yes or no answer, when we must choose between
alternatives. People who dislike yes or no answers dislike saying
"Everything is either A or not-A." Some people would even like to
"abolish" the law of the excluded middle.
This chapter has emphasized a fundamental paradox, which may be called the
"it is and it isn't" situation. There are no sharp divisions between
white and black, and yet there are. There are continuities, and there is also
the necessity for arbitrary standards. The lesson of this chapter, if there is
one, is this: Let us not deny or forget the facts of continuity, and let us not
make sharp divisions where these are inappropriate. And let us not deny the
distinctions between classes of things.
FOR DISCUSSION AND WRITING
1. The following statements contain errors in logic. Find each error, showing
how it violates the law of rationality.
a. As a lawyer, Frank, you are obviously against no-fault automobile insurance
because such legislation would cut into your income severely.
b. America: love it or leave it!
c. Your attitude toward morality is founded on an outmoded nineteenth-century
notion that there is a double standard for men.
d. You might know Sally would give you advice about your marriage. After all,
she's had three husbands herself.
e. In your heart you know he's right.
f. Of course, as a Democrat I'm pro-labor, con, but better that than being a
follower of big business like all you Republicans.
g. But Mom, it's a very "now" thing to do.
h. His views on social reform are very radical, but what can you expect from a
professor of sociology?
i. You've learned only theory in college; you can't be expected to understand
the problems of the man in the street.
j. I suspect his fitness as a candidate for that office. He underwent
psychiatric treatment nine years ago.
k. Her proposals for welfare programs are indeed radical; I remember when she
entertained the Socialist Club in her own home.
l. Don't explain his voting record in the Senate to me. The fact that he is a
Mormon is all I need to know.
m. It's just one of those things you can't explain, but I feel so strong about
his proposal I know it must be a good one.
n. His conception of foreign aid is influenced considerably by the hopelessly
antiquated theory which was popular at the end of World War II.
2. Can you think of circumstances in which the use of emotion in a speech or
argument would be acceptable? Should advertising, for example, be based totally
on reason? Can it be?
3. Using the following examples as models, construct your own list of
adjectives: "I am firm; you are stubborn; he is pigheaded"; "I
am slim; you are thin; she is skinny."
4. Analyze the following statements. In each case, determine whether the
argument violates logical thinking.
a. Either you are for us or against us.
b. Either an animal is a canine or it is not.
c. Whosoever shall walk in the paths of righteousness will be saved; he who
swerves from that path even for a moment is damned.
d. Either Hamlet is a classic or it is not.
e. Either Jones receives the nomination or the country is in serious trouble.
f. Either you vote for what is good for this country or you will vote for the
Democrats.
5. Analyze the following arguments, showing whether they are valid or invalid.
a. While the strike was an effective means of helping the working man in the
nineteenth century, it no longer is a necessary tool for labor. In fact, it is
a dangerous tool. If we allow dock workers to strike or teachers or the police,
nothing will prevent others from doing the same. Imagine what would happen to
the country if state employees went on strike or the army or the doctors or
even politicians.
b. The argument that our welfare programs will lead to a socialistic state is
groundless. We already give aid to the poor, the' aged, the sick. why not then
develop the program merely a step further. I would pro-. pose that we increase
our aid to the sick by providing them with free medical aid. The state could do
this easily if it followed my plan. Because there is little difference between
free medical aid financed by the state and state control of medical aid, let
the federal government take over alt hospitals. Because there is little or no
difference between health insurance for the ill and directly paying the salary
of the individual doctors, let the government control all hiring, training, and
locating of doctors; in that way we could all be assured of prompt and high
quality medical attention.
6. Given the following statements as your sole source of information, how would
you define the "law of the excluded middle"?
a. All girls are fickle. No girls are not fickle.
b. All lakes are polluted. No lakes are unpolluted.
7. Write sentences that contain the following violations of the law of
rationality:
a. argumentum ad hominem
b. begging the question
c. argumentum ad ignorantiam
d. "either-or" fallacy
CHAPTER 6
Putting Up a Logical Argument
Whether we are reading or writing argumentative prose, we should know how to
take it apart, that is, we should be able to identify its conclusions and
premises. We should always ask-even of our own writing-the following questions:
What is the author's point? What reasons does he give to support his point? Is
the argument valid? Is it true?
The argument and its parts
But first, a warning. The word "argument" has several senses. To most
who hear the term, it is a contest in reasoning in which one person wins and
another loses. In contentious arguments one person tries to prove another
wrong. An argument in this sense is often contrasted with
"discussion," an interchange of ideas in which there is no attempt to
defeat an opponent.
Arguments, in the popular sense, often become verbal slugfests, m which one
person tries to beat another down. All too often contentious arguments are
wrangles in which tempers rise and in which the arguers often put up a
mule-like resistance against anything their opponents may say. But we shall not
be concerned here with such arguments.
By "argument" we shall mean the basic unit of reasoning. The proof
any statement or belief is always presented in the form of argument, defined as
"a unit of discourse in which beliefs are supported by reasons." Our
interest henceforth is in argument, not for purposes of contention, but insofar
as arguments are an indispensable element in the quest of truth. Argument in
this sense is the heart and soul of the rational enterprise.
Premises and conclusions
Arguments, then, are not things that are either lost or won, but units proof
(or attempted proof) that something is or is not the case. Here is an example:
"Only citizens who have registered can vote, and you haven't . registered,
so you can't vote." In this argument the speaker seeks to prove that
"you can't vote." This statement is supported by reasons. A statement
supported by reasons is known technically as the "conclusion" of the
argument. This is its "point," what it is "driving at." The
statements which sup-port the conclusions (the reasons) are called the
"premises" of the argument. The premises are the evidence, or facts,
or assumptions, or reasons, on which the conclusion is based. Note that a
statement is a premise only in the context of an argument. The mere assertion,
"Dust thou art," becomes a premise when it is used in an argument:
'Dust thou art; therefore the body is inferior to the spirit." A statement
becomes a premise by virtue of the role it plays in an argument.
In an argument, then, we say, "This, because of that," or "'This
is so, therefore that is so." This process is called
"inference." But before we engage in further analysis of argument,
let us contrast an argument with what we shall call a "mere
assertion." Here is an example of the latter:
'There are thousands of young people whose lives are being ruined by marijuana.
Marijuana should not be legalized."
By a "mere assertion" we mean any statement for which no
"justifying reasons are given. In an argument there are several
statements, one of which is supported by others. The supported statement is the
conclusion; the supporting statements are the reasons or premises. In the
"marijuana" quotation there are no supported and supporting
statements. And when. you read, "Dust thou art, and unto dust shalt thou
return"-period, this too is an unsupported statement, thus a mere
assertion.
The argument, in the sense of "discourse containing inference," is
the central core of logic. Only arguments can be called logical or illogical.
Not all discourse is argument; perhaps most is not. We ask of a friend,
"What's new?" and he tells us. Narration is not argument, and so
there is little argument in newspaper reporting and historical writing. But
when we read a newspaper editorial, we are likely to find argument. The writer
will be trying to prove something, such as the error in present public policy
or the desirability of a new course of action. We support our beliefs by
argument when we expect to be met with a challenging "Why?"
"Why do you believe that?" "Why do you think so?" "Why
ought we to?" We said earlier that it was the mark of a rational man to
support his beliefs by adequate evidence. This Is especially the case when his
beliefs are of a controversial nature.
Now, when we read argumentative discourse, it is well to know how to "take
it apart" with the critical eye of logic. There is perhaps no more
important lesson for skilled reading than this: When you read argumentative
discourse, find and identify its conclusion, and then note its supporting
reasons, or premises. There are two questions which should always be in the
forefront of the reader's mind: (1) What is the writer's point, exactly what is
he trying to prove, or disprove; what is he trying to "put across"?
(2) What reasons does he present to persuade me that he is right, on what basis
am I expected to agree with or accept the conclusion? These two questions, of
course, constitute only the first lesson in logical analysis. The next lesson
will take up the question whether the argument presented by the writer is sound
or unsound. But one thing at a time. What we are now concerned with is the
analysis of an argument into its parts, and the reader who makes the two
questions part of his normal response to argumentative discourse is already on
the road to becoming a more critical and intelligent reader.
An argument, then, has two parts: the premises (or premise) and the conclusion.
The premises may be stated before the conclusion, or they may be stated after
the conclusion. There are certain words, called "logical indicators,"
which connect the premises and the conclusion. When the premises are stated
first, the word "therefore" (or a synonym) will be used: "You
haven't registered; therefore, you can't vote." The word
"therefore" always precedes the conclusion of an argument, and it
always follows the premises. On the other hand, when the conclusion comes
first, we use "because" (or a synonym) to connect the parts of the
argument:
"You can't vote, because you haven't registered." The word
"because" always precedes a premise. The conclusion may also be
sandwiched in between two premises: "Only those who have registered can
vote, so you can't vote, for you haven't registered." If you are
henceforth in doubt as to what the conclusion of an argument is, look for the
logical indicators. They may, of course, not be present in expressed form (they
may be understood), and one should then look to see where they can be inserted.
There are many synonyms for "therefore" and "because." For
example, if you substitute words like "so," "hence," and
"consequently" for "therefore," you Will see that they have
the same meaning. Sometimes the logical indicator is spelled out more fully, as
in "which shows that," "which indicates that," "and so
we may conclude that," etc. Synonyms for "because," are
"for," "since," or phrases like "in view of the fact
that," or "for the reason that," and so on.
In an argument, then, we find the two elements, premises and conclusion. There
is no rule as to whether we should state the premise be-fore the conclusion, or
vice versa. In an extended argument which aims at persuasion, however, certain
strategical considerations may influence the positions of the premises and the
conclusion. For example, let w suppose that you are going to argue in behalf of
a conclusion which will sound reasonable to your audience. It will then be well
to state your conclusion at the outset, before giving your reasons for
accepting it. You will have won the confidence of your hearers by the
'soundness" of the conclusion which you are going to prove. On the other
hand, let us assume that you wish to attack the popular view. Now, if you were
to state your conclusion at the outset, this would arouse strong opposition and
perhaps resentment. Many hearers would regard you as so unreasonable that they
would not listen carefully to the rest of what you had to say. In this case,
may be advisable to build up the evidence with facts that your hearers accept,
and then show how these facts logically require a conclusion different from the
one they accepted previously. But these strategical considerations have nothing
to do with the logic; they relate solely to the strategy of the argument.
Arguments may have more than one premise, and they may also have more than one
conclusion. "I believe that Police Captain Blue takes bribes. He
associates with gangsters and has become very wealthy." Two premises for'
one conclusion here. Note that there are no explicit logical indicators in this
argument, but the sense indicates the structure. "The farmers voted for
quotas. This will increase government expenditures, and it also signals the end
of free enterprise." Two conclusions drawn I here from a single fact.
In a "serial argument," we find a conclusion drawn from a reason, and
this conclusion then serves as a reason for drawing a second conclusion. The
final conclusion is the main point of the argument. Here is an example: "I
am not as wealthy as I used to be, because of the decline in the stock market.
And, since I am not as wealthy as I used to be, I shall be unable to buy that
new house."
Proof and disproof
An argument has been defined as "discourse containing inference."
Inference is used in proof, which we shall now consider. Proof may be used in a
broad or in a strict sense. In the broad sense proof means "enough
evidence" to justify a conclusion, as in the legal expressions,
"proof by preponderance of the evidence" or "proved guilty
beyond a reasonable doubt." In the strict sense, as used in the proof of a
theorem in geometry, proof means "demonstration," that is, the
logically necessary relation between axioms and theorem. We shall use the
broader sense unless otherwise noted.
"Proof" also covers "disproof," for a disproof is simply
proof that a statement is false. This must be distinguished from "failure
to prove, just as a verdict of "not proven guilty" differs from proof
of innocence. (We suspect that many guilty defendants have been acquitted in
criminal trials.) Disproofs (or refutations) often take a characteristic
pattern which resembles the "reductio ad absurdum" (reduction to an
absurdity). The principle of the reductio is that, if a statement implies
absurd (or false) consequences, then the statement must be false. Here is a
sample of the reductio ad absurdum:
It is a common notion that morality simply means conformity to the customs of
one's group. But this cannot be the case. If it were, we could never criticize
and improve the morals of our group, at least we would have no moral basis for
doing so. However superstitious, or stupid, or cruel the customs of our
community are, they would be, by definition, morally right-for us. The
unthinking conformist would be the moral man, the moral reformer the immoral
man. There would be no moral progress. But no one really believes this. We all
constantly criticize the morals of our group. (Adapted from an argument by
Durant Drake, in Invitation to Philosophy.)
This argument seeks to disprove the theory that morality means conformity to
the customs of one's group. The theory is disproved by showing that it entails
absurd or false consequences. If this theory were true, the argument runs, then
we could never be justified in criticizing the morals of our group, and it
would be senseless to speak of "improving" them. But, as the last
sentence in the argument notes, we do criticize the morals of our group and
assume that we are justified in doing so. The consequences being false, the theory
must be false.
The last argument may raise some unanswered questions: Is the argument sound?
Do the conclusions really follow from the facts cited? But such questions must
be postponed until later in this chapter. What we are here concerned with is simply
to illustrate the structure of an argument, not to assess its validity. The
criticism of an argument for faulty reasoning must come after we understand
exactly what the argument says. To understand what it says we must know what to
look for. Thus the understanding of an argument often requires a great deal of
cooperation from the reader. It is the duty of a writer to make his meanings
clear, and also to make the argumentative structure clear, but there are bad
readers as well as bad writers.
The syllogism
We shall now analyze the formal structure of the syllogism, a common form of
argument. A syllogism is an argument consisting of two premises and a
conclusion: "Only those who have registered can vote, and you haven't
registered, so you can't vote." The most famous syllogism of all time is
one used in logic texts for the past two thousand years:
All men are mortal
And Socrates is a man;
Therefore, Socrates is mortal.
The syllogism has two premises, one major and the other minor. major premise is
simply the premise which contains the "major term," which is defined
as the "predicate of the conclusion." "Mortal" is the
predicate of the conclusion, so "All men are mortal" is the major
premise. The subject of the conclusion is called the "minor" term:
Socrates. There is one more term, which appears in both premises but not in the
conclusion. This is called the "middle" term: men. The "middle
term" is so-called because it is the connecting link between the other two
terms. "Man" connects Socrates and mortality.
The syllogism we have just analyzed is only one type of syllogism. It is
sometimes called the "Aristotelian" type, because Aristotle was the
first logician who analyzed syllogisms having this type of structure. Later
logicians have examined other kinds of syllogisms. Here is a different type:
"If prices continue to rise, then the unions will ask for further wage
increases." Prices are continuing to go up, so we may be confident that
the unions will demand further wage increases." And another: "Either
a world government with an international police force will be established, or
the world will continue in a state of tension. But there are no signs that such
a government will be established, hence we can expect tension to
continue." Now these last two arguments are syllogisms, for each has two
premises leading to a conclusion. But they do not contain major, minor, and
middle terms. Different kinds of analyses are required for these types, as we
shall see in Chapter 7.
There are many confusions concerning the role of the syllogism in thinking. We
are sometimes told that the syllogism is artificial and outmoded," or that
no one ever reasons in accordance with the form of the Socrates example above.
But the logician does not present the syllogism as a model to be imitated. His
point is, rather, that it is a form we actually do use in our reasonings. This
may seem surprising to the reader, for no one would normally think in the
pattern of the Socrates example. This example is truly in an artificial form,
but it is deliberately put into that form for purposes of analysis, so that we
may get a clear picture of what we are talking about. In "real life"
this is the way the reasoning might look: "Socrates must die: we must lose
him some day. For he. is but a man, and mortality is a doom which none of us
can escape." If we eliminate the rhetoric here, we shall find the familiar
syllogism, which gives us only the bare bones of the argument.
A syllogistic pattern may sometimes add more to our knowledge than the Socrates
example does. The American philosopher W. P. Montague once described how, many
years ago, he had puzzled over whether women should have the right to vote. He
was a firm believer in that famous historical slogan: "No taxation without
representation." He suddenly saw the answer to his problem: Taxpayers
should have the right to vote, and women pay taxes, so women obviously should
have the right to vote. He had found his middle term! Needless to say, his
initial puzzlement would have made him a target for today's women's rights
groups.
Enthymemes
There is another reason for our surprise when we learn that we reason in
syllogisms. This is because few syllogisms are stated completely in everyday
talk. The obvious should not be belabored, and so, to avoid boredom and tedium,
we leave something to the imagination of the listener. When what is clearly
indicated is too obvious to mention, we may omit a premise from a syllogism, or
we may even omit the conclusion. The following example is typical of ordinary
reasoning: "Joe must be doing well this year, for he took his family to
Florida this winter." This is a syllogism, but it is incompletely stated.
There was an additional premise in the speaker's mind: "People who take
their families for winter vacations in Florida are 'doing well."'
Incompletely stated syllogisms are called "enthymemes" (rhymes with
"Bentham-eems"), from the Greek en (in) and thymos (mind). Here are
some more examples: "Naturally I consider him an intelligent man; he's an
independent voter, isn't he?" This assumes that all independent voters are
intelligent. "Our police should not carry guns. This type of police
practice has worked well in England." This argument may sound more
plausible in its incomplete form than it does when spelled out For the missing
premise would be something like "what works well in England will work well
in the United States." This last argument closely resembles the next one:
"This cough medicine ought to be good for my cough, for, according to the
advertisement, it helped a man in Minneapolis." Will whatever helps one
man m Minneapolis help you?
And here is an enthymeme with its conclusion missing: "An old Abbot,
talking among a party of intimate friends, happened to say, 'A priest has
strange experiences; why, ladies, my first penitent was a murderer.' Upon this,
the principal nobleman of the neighborhood enters the room. 'Ah, Abbe', here
you are; do you know, ladies, I was the Abbot's first penitent, and I may
promise you my confession astonished him!"' (From a story by Thackeray)
"Chain arguments"
We shall note one further type of structure here. This is the "chain
argument." A chain argument is a serial argument (mentioned earlier), in
which a conclusion becomes a premise for a further conclusion. Let us look at a
fairly complex example taken from Leibniz:
The human soul is a thing whose activity is thinking. A thing whose activity is
thinking is one whose activity is immediately apprehended, and without any
representation of parts therein. A thing whose activity is immediately
apprehended without any representation of parts therein is a thing whose
activity does not contain parts. A thing whose activity does not contain parts
is one whose activity is not motion. A thing whose activity is not motion is
not a body. What is not a body is not in space. What is not in space is
insusceptible of motion. What is insusceptible of motion is indissoluble (for
dissolution is a movement of parts). What is indissoluble is incorruptible.
What is incorruptible is immortal. Therefore, the human soul is immortal.
There are several variations in chain arguments, but this sample will indicate
the general idea. The validity of arguments of this type depends on a property
of what logicians call "transitive relations." A transitive relation
is one such that, if A has it to B, and B has it to C, then A must have it to
C. "Ancestor of " is a transitive relation. If A is the ancestor of
B, and B is the ancestor of C, then A is the ancestor of C. WI' en we speak of
one class of things as being "included in" another class of things,
we are also dealing with a transitive relation, for if class A is included in
class B, and B is included in C, then A must be included in the class C. If the
"class" of whales is included in that of mammals, and the class of
mammals is included in that of warm-blooded creatures, then whales must be
warm-blooded. In the Leibniz argument we find this relation of class inclusion.
Each sentence can be interpreted in terms of the relations of two classes to
each other: The class of "souls" is included in the class of
"things whose activity is thinking." The latter class is in-eluded in
the class of "things whose activity is immediately apprehended," and
so on. The argument is valid because "class inclusion" is a
transitive relation.
Not all relations are transitive, of course. Some are "intransitive,"
such as "being the father of." An intransitive relation means that if
A has a certain relation to B and B has it to C, then A cannot have it to C.
"Ten percent larger than" is another such relation. Then there are
"non-transitive" relations, such as "being a friend of": If
A has a relation of this kind to B, and B has it to C, then A may or may not
have it to C.
Logicians also classify relations as "symmetrical,"
"asymmetrical,,' and "non-symmetrical." A
"symmetrical" relation is one such that if A has it to B, then B must
have it to A: "equal to." (If A equals B, B equals A.) An
"asymmetrical" relation is one such that if A has it to B, B cannot
have it to A: "mother of." A "non-symmetrical" relation,
obviously, is one where B may or may not have the relation to A, when A has it
to B: lover of." The reader may find it amusing to discover additional
examples for the nine possible combinations of the relations of transitivity
and symmetry:
1. Transitive-symmetrical: equal to
2. Transitive-asymmetrical: greater than
3. Transitive-non-symmetrical: included in the class of
4. Intransitive-symmetrical: polygynous spouse of
5. Intransitive-asymmetrical: father of
6. Intransitive-non-symmetrical: nearest blood relative of
7. Nontransitive-symmetrical: cousin of
8. Nontransitive-asymmetrical: unrequited lover of
9. Nontransitive-non-symmetrical: lover of
In this chapter we have been concerned with understanding what an argument is,
how to identify one, and how to break it up into its parts Two questions should
be asked whenever we find an argument: What is its conclusion? What reasons are
presented in support of the conclusion?
But there are other questions we must ask, These new questions center around
the goodness or badness of the argument What kinds of arguments do you consider
good ones? When you say that an argument is "good," do you mean that
you agree with the conclusion? Does it make a difference, in your estimate of
an argument, whether or not you agree with the premises? Can you refuse to
grant the truth of the premises, and yet accept the truth of the conclusion?
Can an argument be a good one though every statement in it is false? Can it be
a bad argument though every statement in it is true?
Before we try to answer these questions, let us note an ambiguity in the word
"good" when applied to arguments. A "good" argument may
mean one which is valid in form, that is, an argument whose structure is such
that, if the premises are true, the conclusion must necessarily be true. Such
an argument is valid even if the premises are not true. On the other hand,
"good argument" may mean one that is completely satisfactory: valid
in form and containing true statements. A valid argument, then, may not be
completely satisfactory.
Let us now examine the principles which will help us develop and recognize
logical arguments.
PROOF
Necessary and probable proofs
Are there two persons in the city of Chicago who have exactly the same number
of hairs on their heads? Perhaps you think this highly unlikely, or perhaps you
think it likely, but can we prove it one way or the other? It would be highly
desirable if we could decide the issue without having to count the hairs on
thousands of heads. Logic comes to our aid here. There are two well-known facts
that make our answer certain. First, we must acknowledge the fact that a human
head can have, as a maximum, about a quarter of a million hairs. A second fact
is that there are close to four million persons in Chicago. Now, let us put
these facts together, and we see that there must be two persons with the same
number of hairs. For suppose that we actually did start counting the number of
hairs on people's heads. And suppose that in the first 250,000 heads we counted,
each head had a different number of hairs, so that no duplication occurs in the
first 250,000 heads. In other words, there will be one head with one hair, a
second with two, and so on up to 250,000, the maximum possible. We then come to
the 250,001st head, that is, the first head beyond a quarter of a million. We
must now duplicate one of the numbers of the earlier subjects, since no one can
have a number greater than 250,000. How useful logic is in sparing us tedious
investigations!
This was an example of a logical proof. Let us now look at a different kind of
"proof," the kind we find in a law court. A man is tried for the
murder of his business partner. The accused was the beneficiary of a large
insurance policy made out by the victim, and an additional motive is
established in the fact that the victim was in love with the accused's wife.
Ballistics experts establish that the bullet which killed the deceased was
fired from the defendant's gun. The accused man claims that he is innocent, but
cannot establish an alibi. His only defense is his claim that he is innocent.
The jury must weight his denial against the evidence presented by the state.
The jury finds that the state has proved his guilt beyond a reasonable doubt.
We have examined two arguments, each of which uses the term proof." But
there is an important difference between them. If it is true that there are x
numbers of hairs on a human head and more than that number of persons in
Chicago, then it must necessarily be true that there are two persons with the
same number of hairs. This is a formally valid, or necessary, argument. But in
the second case, if we grant the truth of the premises (the evidence given),
the conclusion-that the accused is guilty-may or may not be true. There is a
probability that he is, but he is not necessarily guilty. There is a
possibility that he is innocent.
In the strictest sense, "proof" means an argument in which the
conclusion necessarily follows from the premises. This is "formal"
proof, or "demonstration." If we accept the premises in the
hair-counting argument, we must accept the conclusion of that argument; if we
grant the truth of the premises of that argument, then we must accept the truth
of its conclusion. On the other hand, when we say "proved beyond a reason
able doubt," we do not mean that the conclusion must be accepted or that
it necessarily follows from the premises, but only that it would be
unreasonable" to accept the premises and not accept the conclusion. This
is "proof' in a less precise sense of the term. We shall use the word
"proof" in both senses, but we should recognize the difference
between the strict and the looser sense of the word. We shall also use the term
valid" for an argument that involves logical necessity, and we shall call
the other type a "probable argument." The distinction depends on the
relationship of the premises to the conclusion. Granted the truth of the
premises, must the conclusion be true? If it is impossible that the conclusion
should be false when the premises are true, then the argument is a valid one.
If it is true that all reformers are idealists, and that all idealists are
nonconformists-if these premises are true-then it is impossible for "all
reformers are nonconformists" not to be true, and so this is a valid argument.
But if the premises can be true while the conclusion can be false-as in the
murder example above-then the argument is not valid, though the premises may
make the conclusion highly probable. In a "probable argument," of
course, the degree of the probability, whether high) low, or moderate, will
depend on the quantity and quality of the evidence.5
Truth and validity
Thus far we have been discussing the distinction between a necessary and a
probable argument. We shall now discuss the relationship between truth and
validity. The main point is this: The truth or falsity of the premises (or the
conclusion) has nothing whatsoever to do with the validity of the argument.
This point is perhaps the most important lesson that one can learn about
logical thinking: the distinction between the logical structure of an argument,
on the one hand, and the truth or falsity of its evidence, on the other.
In connection with this distinction it is well to bear in mind the precise
definitions logicians give the terms "truth" and
"validity." Statements are true or false: A true statement is one
which describes the facts correctly. Arguments are valid or invalid: A valid
argument is one in which I the conclusion is necessitated by the premises. Note
in particular that only statements are true or false; only arguments are valid
or invalid. Thus logicians never (or almost never) say "a true
argument" or "a valid statement."
Let us illustrate the distinction between truth and validity. "Human
beings can't live on the moon, for there is no oxygen on the moon, and human
beings can live only in places that contain oxygen." This is a valid
argument, for if the premises are true then it would be impossible for the
conclusion to be false. But we may question the truth of the second premise:
Astronauts can bring their own oxygen with them when they land on the moon. So,
though formally valid, this is not a satisfactory argument.
Further, an argument may be valid though all its assertions are preposterous:
"If Eskimos were Cubans, and if every Cuban were an atomic scientist, then
every Eskimo would be an atomic scientist." Validity is concerned with
form or structure alone. If we symbolize Eskimos by E, Cubans by C, and atomic
scientists by A, the form of the argument is: If E's are C's, and C's are A's,
then E's are A's. The content of the substitutions for E, C, and A is
irrelevant to validity. The rules of logic in relation to the substance of an
argument are like the rules of arithmetic in relation to examples of
application: "If you had twelve purple cows and seven of them were
kidnapped, you would then have five in your possession." It would be an
irrelevance, insofar as we were interested in the correctness of the arithmetic
here, to tell us that you, speaking for your-self alone, never saw a purple
cow.
Let us now apply the same considerations to a probable, or inductive, argument.
Another murder trial. The prosecution presents three witnesses who testify that
they saw the accused murder the victim. A motive is established, opportunity is
proved, and other incriminating circumstances weigh heavily against the
accused. The jury brings in a verdict of guilty beyond a reasonable doubt. We
say that the jury acted reasonably, that is, they reasoned quite soundly in
returning the verdict of guilty. But later we learn that the three witnesses
were conspiring against the accused; they had framed him with his own gun, etc.
The logical question:
Does this new information affect our previous decision that the jury acted
reasonably? It does not. It was reasonable to believe that the accused was
guilty on the basis of all the evidence known at the time of the verdict. The
quality of the reasoning, then, is independent of the truth of the assertions
on which the conclusion is based.
Now we can see that the truth or falsity of the premises is irrelevant to the
"logic" of an argument. Just as an argument may be valid though its
premises are false, so a conclusion may be probable on the basis of accepted
premises, even though these premises turn out to be false or questionable.
The principles we just discussed will perhaps explain some familiar
experiences, in which we listen to arguments in which each step follows from
the preceding one without a logical flaw-but in the end we are of the same
opinion still," unconvinced. This sort of thing often happens when someone
is trying to convince us of the errors of our political or religious ways.
These unconvincing arguments may actually have been quite sound from a purely
logical point of view. But this means only that the conclusion did actually
follow from the premises assumed by the speaker. Your refusal to accept his
conclusion, then, indicates that you refuse to accept the truth of his
premises, and if you examine the premises. carefully, you will be able to spot
the place where your disbelief or doubt arises. (We are assuming that the
argument is understandable.) Arguments of this kind are obviously
unsatisfactory, for a completely satisfactory argument is not only correctly
reasoned in the purely logical sense, but it is also one in which the premises
are acceptable to us.
If the reader will turn back to the chain argument quoted from Leibniz (click here), he will find that the
logical form of this argument is impeccable. You may or may not agree with
Leibniz' conclusion that the soul is immortal. But if you do accept the truth
of his premises, you must accept the truth of his conclusion, for the argument
is valid. It is not the mark of a rational mind to say, "Your argument is
valid, and your premises are true, but I refuse to grant the truth of your
conclusion." This statement would be self-contradictory, for it first
grants the validity of the argument and then denies it. It is like saying,
"This is a square, but it has five sides." if an argument is valid,
and the premises are true, the conclusion must be true, for this is what
validity means: a structure such that if the premises are true, the conclusion
must also be true.
But let us suppose that a reader disagrees with Leibniz' conclusion that the
soul is immortal, or that he remains uncertain of its truth, although he does
see that the argument is a valid one. This must mean, then, that he disagrees
with, or is at least uncertain concerning the truth of, at least one of the
premises. He may be able to point to the premise he disagrees with. But a
reader untrained in metaphysics may find the meaning of the premises in
Leibniz' argument obscure, or even unintelligible. A properly worded objection
to the argument, in this case, might go like this:
"I am unconvinced of the truth of the conclusion, even though it
follows logically from the premises. I must therefore be unconvinced of the
truth of a premise or premises. I can't tell you which I consider wrong, for I
don't understand them well enough to criticize, but I feel that there must be
one premise that I would consider uncertain even if I understood it, for I
don't accept the conclusion."
Now, this is not an unreasonable position to take, for, just as it is the mark
of a boor to criticize an argument he doesn't understand, so it is the mark of
an intellectually irresponsible person to accept an argument 6-he doesn't
understand. If we wish to make friends and influence people, we may find
intellectual irresponsibility less irritating to others than candor, but an
intellectually responsible person will refuse to agree or disagree until he
understands.
We have devoted a good deal of attention to the distinction between validity
and truth, for this is a much misunderstood point. And there is a special kind
of confusion concerning the matter of accepting an unproved premise "for
the sake of argument." Suppose someone says, "If the need for large
military expenditures should diminish within the next year . . ." and an
objector breaks in at this point. "Stop right there," he says.
"There's no point in your going on with your argument, for I don't accept
your basic premise, and therefore can't accept whatever conclusion you will
draw." But this is a refusal to be rational, for it may be very
en-lightening to deduce the logical consequences of uncertain, or even false,
premises. What the speaker was going to say was: "If the need for military
expenditures should diminish within the next year, and we are not prepared with
plans for immediate tax reductions and other stimulants to the domestic
economy, then we shall have a severe depression." It is worthwhile
considering the logic of this argument even if we do not admit the truth of its
major assumption. For if the premises should turn out to be correct, it is
certainly useful to know the consequences they entail, and thus prepare
ourselves accordingly.
The good thinker, then, must often entertain unproved or even false assumption
"for the sake of the argument." Scientists do this as a matter of
course. Sir Isaac Newton's "first law of motion" tell us that if a
moving body is not influenced by outside forces, it will continue in motion
forever. The first part of this law contains an assumption which is contrary to
fact, for there are no bodies which are "not influenced by outside
forces." But physicists find this law useful, for it implies that a body
will continue in motion for a longer and longer period as friction is reduced.
Our basic distinction is between is between the truth of the premises or
conclusion, on the one hand, and the logical validity of the form, on the
other. The distinction may be put in another way. It is like the difference, in
an audit, between the soundness of the evaluations of items and the arithmetic
used in adding up the totals. I may evaluate a pretzel factory building at
$50,000, the machinery at $50,000, and the pretzels on hand at $900,000. Total
assets: $1,000,000. The arithmetic is faultless, but the evaluations may be
unrealistic.
Invalid argument
The significance of what the logician means by validity may become clearer when
we understand the meaning of an "invalid argument." A formally
invalid argument is one which is deductive in form, but in which the conclusion
is not necessitated by the premises. The conclusion does not "follow"
from the premises. This is the meaning of "non sequitur." Two
examples:
(1) Cats climb trees
And squirrels climb trees;
Therefore, cats are squirrels.
(2) Manitobans live in the northern part of North America
And Canadians live in the northern part of North America
Therefore, Manitobans are Canadians.
Note that these arguments are similar in form. In each, the premises compare
two things, cats with squirrels, and Manitobans with Canadians. In both, the
compared entities have common characteristics. In (1) the common characteristic
is tree-climbing; in (2) it is living in the northern part of North America.
The conclusions are similar in that each tells us that one of the compared
entities is identical with, or at least included within, the other.
Both of these arguments are invalid. In neither case does the conclusion follow
from the premises. The relevant principle of logic is this:
The mere fact that two things have one or more characteristics in common does
not justify us in concluding that the two things are identical, or even that
one is included within the other. The fact that cats and fox terriers suckle
their young does not justify the inference that cats are fox terriers; the fact
that the Chinese and Japanese eat rice does not justify the inference that
Chinese, after all, are nothing but Japanese.
Note that the conclusion in (1) is false; that in (2) true. Many people will
think that (2) is a better argument than (1), but the arguments are equally bad
from a strictly formal point of view. The second appears more plausible because
the conclusion is true, but in order to see how bad the logic is, let the
reader substitute the word "Alaskans" for "Manitobans" in
(2). The new premise: "Alaskans live in the northern part of North
America." The new conclusion: Therefore, Alaskans are Canadians."
The error we have just described may appear to be a very simple one, too
obvious to mention, but we commit it often. The error occurs most often in complex
contexts, and in subject matter that involves our emotions. We are particularly
apt to overlook the badness of an argument when we believe the conclusion to be
true, and particularly so when we derive emotional gratification from it.
Consider the following: Joe Doakes must be a Communist, for he believes that
Communists should be permitted to speak on university campuses, and we all know
that Communists are in favor of permitting them to speak. This argument will
sound plausible to many.
But, once more, the same error as above. Joe agrees with the Communists in one
respect, we are told: both believe that Communists should have permission to
speak. A common characteristic! But this does not prove that Joe is a
Communist, for "the fact that two things have a characteristic in common
does not prove that they are identical." Joe may be violently opposed to
communism, and yet believe that students should be permitted to hear all points
of view.
But a warning signal should be posted at this point. Though an argument is
invalid, it may have considerable merit. An invalid argument is one in which
the conclusion does not necessarily follow from the premises. We remarked
earlier that the phrase "not necessarily" covers a lot of ground. The
conclusion that cats are squirrels does not follow necessarily from the
premises we noted above. Nor does the conclusion follow necessarily in this
argument: "True theories are confirmed by careful experiments, and
Einstein's theory has been confirmed by careful experiments, so his theory must
be true." This argument, like the cats and squirrels argument, is invalid,
and for the same reason. The premises tell us that Einstein's theory shares a
characteristic with true theories, namely, that both are confirmed by careful
experiments. But this does not guarantee that Einstein's theories must be
trite, for the sharing of a characteristic does not prove identity. Scientists
will agree that this is a correct analysis of the argument. For it is a
well-known fact that many theories have been confirmed by experiments, only to
be disproved by later experiments. This is the basic reason why scientists
disclaim absolute certainty for their findings.
We are not, of course, singling out Einstein's theory as a special case, but
are using it only as an illustration. Every experimental proof in science takes
the same form. Nor are we casting the doubt of skepticism over scientific
findings, for though the argument we just considered is in-valid from a
technical, formal point of view, it differs from the foolish arguments
considered in this chapter as sense differs from nonsense. This point requires
careful consideration.
"Invalid" means that the conclusion does not follow necessarily from
the premises. But in some cases the premises of an invalid argument seem to
make the conclusion highly probable, as in the Einstein example; in others they
do nothing of the sort. Why this difference? It all depends on the nature of
the characteristic in which the two things agree. We say that the sharing of a
common characteristic does not prove that two things are identical, or that one
of these things is included within the other, but we come nearer to proving
this inclusion in some cases than in others. For example, anarchists eat food,
and so does Joe. Agreement in "eating food" proves absolutely nothing
concerning the political similarities of food-eaters. But if we say, "The
anarchists believe in abolishing all government controls, including the police
force, and so does Joe," here the shared characteristic is highly
significant. The argument is invalid, if we conclude that Joe is necessarily an
anarchist, for the premises may be true and the conclusion false. Joe may
merely be a "rebel without a cause." But the probability is high that
he is an anarchist.
In other words, if the shared characteristic is one that is possessed only by
anarchists, or if there is a high probability that anyone having the shared
characteristic is an anarchist, then we can translate our invalid argument into
a valid one. If we can say, "Anyone who believes in the abolition of all
government controls, including the police force, is probably an
anarchist," and we find that Joe so believes, then obviously Joe is V,
probably an anarchist. This argument is valid for the conclusion necessarily
follows from its premises.
If we return now to the argument that led to this discussion, the premise
"True theories are confirmed by careful experiments" can be
translated into "If a theory is confirmed by careful experiments, then it
is probably true." (Note that we did not say "necessarily.") And
thus, since Einstein's theory was so confirmed, we can derive the valid
conclusion that it is probably true. But this kind of translation is possible
only when the shared characteristics are of such a nature as to make it
probable that one thing is included within another. We cannot translate our
"cats and squirrels" argument in this way, for "Cats climb
trees" can't be translated into "If it climbs trees it probably is a
cat" (or a squirrel). Thus the conclusion of this argument is not a
probable one. The point is this: when we note an argument based on shared
characteristics hereafter, let us V-also note the significance of the shared
characteristics. Some shared characteristics may yield a probable conclusion.
To repeat, "not necessarily' covers a lot of ground, from a highly
probable conclusion to a worthless one.
The fallacy of the "shared characteristic," known technically as the
fallacy of the "undistributed middle term," resembles the idea of
"guilt by association." Just as the sharing of a characteristic does
not prove identity, so the fact that a man knows a Communist does not prove
that he is one. But just as the significance of the characteristic, or
characteristics, is important in establishing probability, so with one's
associates. If a state's attorney has gangsters as his constant companions,
there may be a justifiable suspicion as to his honesty, and we may want state's
attorneys who are not only unconvicted of crimes, but also above suspicion.
In concluding our discussion of validity and its relation to the truth of
assertions-the subject matter of "deductive" logic-we shall sum up
the matter schematically. There are four possible combinations of premises and
conclusion with respect to their truth and falsity:
(1) the premises may be true, and the conclusion true
(2) the premises may be false and the conclusion true
(3) the premises false and the conclusion false
(4) the premises true and the conclusion false.
We shall now illustrate these combinations with examples of invalid and valid
arguments, respectively. In arguments which are invalid, we can find each of
the four combinations listed above:
(1) The invalid Manitobans argument. (click here)
(2) "Socialists are capitalists, and those who wish to abolish private
property are capitalists, so Socialists wish to abolish private property."
(The sharing of characteristics does not prove identity.)
(3) Too obvious to illustrate.
(4) The "cats are squirrels" argument. (click here)
In valid arguments, on the other hand, only the first three of these
combinations of truth and falsity can be illustrated:
(1) The familiar "All men are mortal" syllogism.
(2) "Socialists are capitalists, and capitalists favor the abolition of
private property; therefore, Socialists favor the abolition of private
property." This is a valid argument, for if the premises are true, then
the conclusion would have to be true. It may be helpful to compare this
illustration of a valid argument for combination (2) with the illustration of
the invalid argument for combination (2) given above. Each has false premises
and a true conclusion. But in the invalid argument the conclusion is not
necessitated by the premises; that is, acceptance of the premises as true would
not require us to accept the conclusion as true. The conclusion of that
argument happens to be true, but we cannot say: If these premises are true,
then the conclusion must be true. But this is precisely what we must say of the
valid argument illustrating combination (2) in this paragraph.
(3) The valid Eskimos argument (click here)
(4) This combination is impossible when an argument is valid. For consider the
meaning of the term "valid argument": one in which the truth of the
premises requires us to accept the truth of the conclusion. If we say that an
argument has true premises and a false conclusion, we thereby declare that it
is not a valid argument
The main points of this chapter may be summed up.
(1) By formal logic alone we cannot prove the truth of any assertion. What
formal logic tells us is that if we start with true premises and reason
logically from these premises, then our conclusion must be true.
(2) If our premises are false, or even uncertain, then even when we reason
logically our conclusion has not been proved to be true.
(3) When the logical form is invalid, that is, when the reasoning is illogical,
then even true premises cannot guarantee a true conclusion.
(4) In arguments like the Einstein example, we found that some invalid
arguments can be translated into valid arguments yielding probable conclusions.
We also recall our earlier discussion of the ambiguity of "good" as
applied to arguments. A completely satisfactory argument, we said, was not only
valid in form, but also contained true premises. A bad argument, then, is one
which is either invalid or which lacks truth. We have also seen that true
conclusions may be supported by either true or false premises in invalid
arguments.
A final point. A bad argument, that is, one invalid in form or containing false
premises, cannot prove a true thesis. But it is also important to remember that
a bad argument does not discredit a true thesis, though it may sometimes appear
to, as when a weak premise is attacked in a debate. The refutation of the
premise may then seem to be a refutation of the thesis. In general, it is
better to present a few good arguments for a thesis, rather than a great many,
one or more of which may be weak, for the opponents are apt to seize on the
weak premise and, by discrediting it, appear to discredit the conclusion. But
the logical person will consider the merits of ideas regardless of the bad
arguments used to support them.
FOR DISCUSSION AND WRITING
1. Analyze the following arguments. In each, identify the premises, conclusion,
and (if present) logical indicators.
a. I think that Professor Dwyer is a particularly good teacher. His classes are
always full, and lie is known to have excellent rapport with his students.
b. I am not as happy as I was before Roberta left me. And because I am unhappy,
I will never be able to love again.
c. The candidate has claimed that the United States is a nation of neglected
poor. If this were the case, we would suppose that our welfare programs were
seriously lacking. But this is not the case. The federal and state governments
spend 20 percent of their funds on welfare. Therefore, this cannot be a nation
of neglected poor.
d. Con McAuliffe is a successful businessman. He always buys luxury
automobiles, and his house is a mansion.
e. The Jets have not played as well since Joe Namath's knees gave out. Because
their play is so poor, they will never win the Super Bowl.
2. Analyze the formal structure of the following syllogisms. What are the major
and minor premises, the major and minor terms, and the middle term? If any
syllogisms are incomplete, furnish the missing parts.
a. All A's are B; C is an A; therefore C is a B.
b. All girls are made of sugar and spice. Alice is a girl. Therefore, Alice is
made of sugar and spice.
c. All athletes are well conditioned. Frank is an athlete. Therefore, Frank is
well conditioned.
d. Fred must have accepted a promotion, because his salary has in-creased
twofold.
e. All newspapers are slanted in their coverage. The Times is a newspaper.
Hence, the Times must be slanted.
f. The editorials of that newspaper have a Republican bias; the owner is a
Republican, isn't he?
g. All artists are nonconformists. Larry is an artist; thus Larry is a
nonconformist.
h. Mary has always been precocious; she started reading at the age of three.
i. That automobile ought to be the best one on the market. All the race drivers
recommend it.
3. Analyze the following syllogisms. Are their premises true? Is the conclusion
true? Valid?
a. All monkeys have tails; Mr. Smith is a monkey; therefore, Mr. Smith has a
tail
b. All petunias have petals; all roses have petals; therefore, all roses are
petunias.
c. If all men were viola players, and if every viola player was bow-legged,
every man would be bowlegged.
d. Francis lives in Belgium, and Belgium is a European country; therefore,
Francis is a European.
e. Gloria is a Belgian, and Belgium is a European country; therefore, Gloria is
a European.
f. Birds live in trees, and squirrels live in trees; therefore birds are
squirrels.
g. Woodfinches live in trees, and robins live in trees; therefore woodfinches
are robins.
h. Woodfinches live in trees, and birds live in trees; therefore woodfinches
are birds.
i. A classic is a piece of literature that has undergone the test of time;
Shakespeare's King Lear has undergone the test of time; therefore, King Lear is
a classic.
j. Communists believe all property should be owned by the state; Burl believes
that all property should be owned by the state; therefore Burl is a communist.
k. All Scotsmen are stingy; Bill is Scotch; therefore, Bill is stingy.
l. Most people are basically selfish; selfishness is an evil; therefore, most
people are evil.
4. Write valid syllogisms which have the following statements as conclusions.
a. Narcotics are poison.
b. Education can take place outside of school.
c. Some ice cream is delicious.
d. Some professors are interesting.
e. Manitobans are Canadians.
f. A thunderstorm is dangerous.
g. Only education can alleviate prejudice.
h. intelligent women never leap to unwarranted conclusions.
5. Write a syllogism for each of the following combinations of premises and
conclusions:
a. The premises are true, and the conclusion is true.
b. The premises (or one premise) may be false and the conclusion true.
c. The premises are false, and the conclusion is false.
d. The premises are true, and the conclusion is false.
6. Test the validity of the following arguments by rewriting the argument as a
syllogism.
a. The study of religion always includes the study of how men should act toward
each other. Our schools also have the responsibility of teaching the young how
to live in society; it is logical to conclude, there-fore, that religion should
be taught in our schools.
b. It is a truism that a political candidate should be very wary of declaring
his stand on controversial issues. In fact, the successful candidate is often
the one who appeals to the largest number of voters by promising everyone what
he or she wants. Our candidate, therefore, should avoid controversial issues.
c. The president has a moral obligation to do for the country what is best for
the majority. But special interest groups are constantly attempting to persuade
him to act in their behalf. To be morally just, however, he should not listen
to their appeals.
CHAPTER 7
Some Patterns of Reasoning
The ability to construct a valid argument is extremely important when writing
convincing argumentative prose. It is time now to examine he notion of validity
somewhat more closely and to explain why some structures of reasoning permit us
to draw valid deductions whereas others do not. In carrying out this task we
shall have to indulge in a technicality or two.
Four patterns of reasoning
We begin by considering a lion, his cage, and a zoo. We shall construct four
very simple arguments which reveal some of the basic patterns of reasoning:
1. The lion is in his cage, and the cage is in the zoo. We draw the
'conclusion: The lion is in the zoo. Obviously this conclusion follows
necessarily from the premises. The argument is valid.
2. The lion is in his cage, but the cage is not in the zoo. The conclusion:
"The lion is not in the zoo,' follows necessarily from these premises.
Valid.
3. The cage is in the zoo, but the lion is not in his cage. Now, can we
conclude with certainty concerning the whereabouts of the lion? Obviously not,
for the lion may or may not be in the zoo. If we draw either one of these
conclusions: "He is in the zoo" or "He is not in the zoo,"
we will draw a conclusion not warranted by the premises, and the argument will
be invalid. Neither conclusion follows necessarily from these premises.
4. The cage is in the zoo, and the lion is in the zoo. We cannot draw a
necessary conclusion concerning the lion's relationship to his cage. He may be
inside it, and he may not be. To draw either conclusion as following from the
information given would be illogical.
In these simple examples we find four basic patterns of reasoning, two of which
are sound, and two unsound. The information given to us by the premises permits
us to draw valid conclusions in patterns 1 and 2; not in 3 and 4. We shall now
draw diagrams to illustrate these patterns in order to get a visual picture of
the difference between valid and invalid structures of reasoning. We shall
combine the premises of each argument in diagrams, in order to see why the
conclusions are necessitated by patterns 1 and 2 and not by 3 and 4.
1. Premises: The lion is in his cage, and the cage is in the zoo.
This diagram exhibits the premises: the lion in his cage and the cage in the
zoo. The conclusion. "The lion must be in the zoo" is unavoidable.
2. Premises: The lion is in his cage, but the cage is not in the zoo.
This is the only way in which it is possible to diagram the information given
by the two premises. The conclusion: The lion cannot be in the zoo.
3. Premises: The cage is in the zoo, but the lion is not in his cage.
Difficulties arise when we try to diagram these premises. Let us start with the
first premise:
We must now show "the lion is not in his cage." There are two ways of
showing this, in conjunction with the first premise:'
Diagrams "a" and "b" are both faithful to the premises, but
neither one is necessarily required by the premises. Diagram "a"
shows the lion in the zoo; "b" shows him outside. An argument which
drew either conclusion would be invalid.
4. Premises: The cage is in the zoo and the lion is in the zoo. Similar
difficulties arise. We begin by diagramming the first premise:
The second premise tells us that the lion is in the zoo, Two ways to show this:
Each diagram shows us both the cage and the lion in the zoo. But the premises
give us no information concerning the relation of the lion to his cage. To conclude
that either "a" or "b" necessarily follows from these
premises is to draw an unjustified inference.
Our first set of illustrations is oversimplified, for we were talking about a
particular lion, a particular cage, and a particular zoo. Most reasoning
concerns classes of things (cabbage, kings, etc.) rather than exclusively
individual objects. Our next four illustrations will exhibit the same general
patterns for classes of things. Note the responsive similarities in formal
structures for the arguments bearing the same numbers:
1. Whales are mammals, and mammals are animals, so whales are animals. Diagram
to illustrate the premises:
If the premises are accepted, then the conclusion must be accepted. Valid.
2. All monarchists are conservatives, and no conservatives are Utopians. We may
properly conclude that no monarchists are Utopians. The diagram shows that the
argument is valid:
3. All Hindus are vegetarians, and no Sikhs are Hindus, so no Sikhs are
vegetarians. The conclusion "No Sikhs are vegetarians" is not
justified by the evidence presented. The argument is invalid. As before, we
draw a diagram for the first premise:
The second premise tells us to draw Sikhs outside of Hindus, but it tells us
nothing about the position of Sikhs with respect to vegetarians. We can draw at
least two diagrams to show the possibilities:
Diagram "a" shows Sikhs outside the class of vegetarians. This was
conclusion of the argument But "b" is also a possibility, and
"b" shows Sikhs inside the vegetarian box. To say that "a"
(or "b") follows necessarily from the premises is to draw an
unjustified inference.
4. Iranians live in the Near East, and Kurds live in the Near East, so Kurds
must be Iranians. Invalid. From the information given in the premises, Kurds
may or may not be Iranians. We cannot conclude that they must be.
The first premise gives us:
We must now draw Kurds inside the box "People who live in the Near
East" But they may be inside the Iranian box or outside. Neither
conclusion is necessitated, so the argument is invalid.
We may sum up our discussion of the four patterns of reasoning in symbolic
form. We shall use the symbol "A," "B," "C" for
any three classes of things:
1. A is in B and B is in C. That A is in C follows necessarily.
2. A is in B and B is outside of C. A must be outside of C.
3. A is in B and C is outside of A. We cannot conclude that C must be outside
of B, for C may be outside of A and inside of B. Nor does any other conclusion
follow necessarily.
4. A is in B and C is in B. We cannot conclude that C is in A (or A in C), for,
though both must be in B, they may be outside of each other.
Before we continue, a further word or two should be said concerning what tile
diagrams show. When we draw diagrams for the premises in a valid argument, the
conclusion is seen to be inescapable. But when we draw diagrams for the
premises in an invalid argument, we see that no definite conclusion is
necessitated. We can draw the premises of an invalid argument without
exhibiting the particular conclusion which the argument drew. In other words in
an invalid argument we need not accept the conclusion even though we accept the
premises. The conclusion may be true, but it is not proven true by the
premises.
Let us sum up for a moment. Arguments 1 and 2 were valid in each set; 3 and 4
were invalid. Let us now examine the technical rules of logic violated by
patterns 3 and 4. Argument 3 is an example of the fallacy called illicit
distribution," and argument 4 illustrates the fallacy of the
"undistributed middle term." These are two of the most frequently
encountered errors in reasoning. In order to understand the meaning of these
fallacies, let us examine the technical logical term known as
"distribution."
"Distribution"
The logician speaks of the "distribution" of words that designate
classes of things (apples, mortals, emotions). To say that' a term is
distributed means that we have said something about all members of the class to
which it refers, that we have asserted something about each and every member of
that class. Thus, in "All dogs are animals," "dogs" is a
distributed term, for we said all dogs. But we did not say anything about all
animals in this particular sentence. Dogs constitute only part of the class of
animals, so our sentence refers only to some animals. "Animals" is an
undistributed term in this sentence. Similarly, if we had said "Some dogs
are hunters," both dogs and hunters would be undistributed.
In the sentence "No men are angels" both terms are distributed. The
sentence asserts that each and every man is outside the class of an-gels, and
it also says that each and every angel is outside the class of men.
To sum up. The distribution of the subject term in a sentence depends on
whether its quantifier is "all," "no," or "some."
The distribution of the predicate term in a sentence is dependent on whether
the sentence is affirmative or negative, and this distinction in turn depends
on the copula of the sentence. The copula, a form of the verb "to
be," connects the subject and predicate. When the copula is "is"
or "are," the sentence is affirmative (S is P); when it is "is
not," the sentence is negative (S is not P). "No S are P" is
classified as negative, since it states that S are not P, that is, all of S are
excluded from all of P. The sentence "All non-S are non-P" is
affirmative, since the copula is "are."
The distribution of the predicate term may now be summed up in two rules: (1)
Affirmative sentences never distribute the predicate term (P): All S are P,
Some S are P; (2) Negative sentences always distribute the predicate term: No S
are P, Some S are not P.
The distribution of S and P in the four possible types of subject-predicate
sentences are shown below:
All S (distributed) are P (undistributed)
No S (distributed) are P (distributed)
Some S (undistributed) are P (undistributed)
Some S (undistributed) are not P (distributed).
This discussion of "distribution" by no means exhausts the subject,
but it is sufficient for our purposes. We are now ready to explain the
fallacies of "illicit distribution" (illustrated by argument 3) and
"undistributed middle term" (illustrated by argument 4). Let us take
number 4 first, since this fallacy is already familiar to us. The reader will
recall the many examples of this error-the error involving "shared
characteristics"-in the last chapter: the arguments concerning cats and
squirrels both being tree-climbers; the argument that told us that Chinese must
be Japanese since both eat rice; that Manitobans must be Canadians since both
live in North America; that Einstein's theory must be true since it has been
confirmed by careful experiments, and all true theories are so confirmed. We
shall see, in a moment, how we may apply this particular technique of logic-the
"distribution" idea-to these arguments.
Just one further preliminary comment: the "middle term" of a
syllogism is the term that appears in the two premises, serving as a connecting
link between the other two terms,
Validity of the syllogism
The fallacy of the "undistributed middle term" refers to a rule of
logic which tells us that the middle term of a syllogism must be
"distributed at least once" in order to permit the drawing of a valid
conclusion. This means that, if the middle term is not distributed at all, no
valid conclusion is possible. Argument 4 illustrates this fallacy. Let us
examine it:
(All) Iranians live in the Near East (middle term)
And (All) Kurds live In the Near East (middle term)
Therefore, Kurds are Iranians.
"Live in the Near East" is the "middle term" (defined as
the term which appears in both premises). This term was distributed in neither
premise. "Kurds" and "Iranians" were distributed in their respective
premises, for we assume that "all" was understood for each, but the
premises refer only to "some" people who live in the Near East. Thus
this syllogism violates a rule of logic. The reader will find the same
technical error in the other examples of this fallacy in the previous chapter.
The diagrams above (click here)
show us the sense of the rule.
In the patterns of reasoning illustrated in those diagrams, only example number
4 violates this rule. In argument number 1 the middle term is
"mammals." This term is not distributed in the premise, "All
whales are mammals," but it is distributed in "Mammals are
animals," for "MI mammals" is intended here. So the rule is
satisfied. In argument number 2 the middle term is "conservatives."
It is distributed in "No conservatives are Utopians." In argument
number 3 the middle term is "Hindus." This term is distributed twice.
This argument, then, does not violate the rule concerning the distribution of
the middle term, but it violates a different rule involving distribution. Let
us examine it.
A second rule of logic tells us that "if a term is undistributed in the
premises, then that term must not be distributed in the conclusion." This
is the sense of the rule: when a term is distributed, information is given
concerning each and every member of the class referred to. When a term is
undistributed, information is given only about some of its members. From
information concerning "some" we can draw no conclusion concerning
"all." For example, if experience teaches me that some women are
fickle, I cannot logically conclude that all are. We commit the fallacy of
going from "some" to "all" when the conclusion distributes
a term that was not distributed in the premises.
Let us restate number 3, which illustrates the fallacy of "illicit
distribution":
All Hindus are vegetarians
And No Sikhs are Hindus
Therefore, No Sikhs are vegetarians.
We examine the distributed terms in the conclusion to detect a possible violation
of the rule against illicit distribution. If a term is distributed in the
conclusion, then it should have been distributed in the premises. We find that
both of the terms in the conclusion are distributed. Now examine these terms in
the premises. "Sikhs" was distributed in the second premise. But
"vegetarians" was not distributed in its premise, and so the rule is
violated. When we conclude that "no Sikhs are vegetarians," we assert
something about all vegetarians; "All of them," we say, are outside
the class of Sikhs." But the premise gave us information only about some
vegetarians.
(All) Senators have traveling privileges
And (All) Senators are politicians
Therefore, (All) politicians have traveling privileges.
"Politicians" was undistributed in the second premise and distributed
in the conclusion.
Let us examine the conclusions of arguments 1, 2, and 4, above, in which this second rule is
not violated.
Number 1: Whales are animals. Whales is distributed in the premise as well as
in the conclusion.
Number 2: No monarchists are Utopians. Both terms distributed in the premises
as well as in the conclusion.
Number 4: All Kurds are Iranians. Kurds is distributed in the premises. This
last argument, how-ever, violates the rule concerning the distribution of the
middle term.
Negative premises and conclusions
So much for two of the basic rules of validity for syllogisms. There are also
three additional rules, each of which concerns negative premises or a negative
conclusion:
(1) A valid conclusion cannot be drawn from two negative premises;
(2) A negative premise requires a negative conclusion;
(3) A negative conclusion requires a negative premise.
These five rules are like the axioms in Euclidean geometry. They are necessary
and sufficient to test the validity of any syllogism involving
subject-predicate sentences in ordinary language.
With practice you should now be able to test the validity of almost any
deductive argument, to determine whether the conclusion follows from the
premises. This skill will be helpful in both writing and reading argumentative
prose. By using the syllogism as a model or form, your paper will be compelling
in its logic. And in reading someone else's deductive argument, you will be
able to reduce it to a syllogism and test it.
FOR DISCUSSION AND WRITING
1. Analyze the following syllogisms by diagraming their premises and
conclusions. Which arguments are valid? How many possible diagrams can be drawn
to illustrate the invalid arguments?
a. All men are animals.
Emil is a man.
Emil is an animal.
b. All men are animals.
Emil is an animal.
Emil is a man.
c. All men are animals.
All men are reasonable.
Animals are reasonable.
d. All animals are sentient.
Emil is an animal.
Emil is sentient.
e. All students are activists.
Al is an activist.
Al is a student.
f. All students of government are activists.
All politicians are students of government.
All activists are politicians.
g. All students are activists.
All activists are radicals.
All radicals are students.
h. All students are activists.
Al is a student.
Al is an activist.
i. Some students are activists.
Al is a student.
Al is an activist.
j. No student is an activist.
Al is a student.
Al is not an activist.
k. No professor is careless.
Muriel is not a professor.
Muriel is careless.
l. No animal is rational
Larry is an animal.
Larry is not rational.
m. Some animals are dogs.
All dogs are four legged.
All four-legged creatures are animals.
n. Some animals are four-legged
Dogs are four legged.
Dogs are animals.
o. Some animals are not carnivorous.
Don is an animal.
Don is not carnivorous.
p. Some animals are carnivorous.
Don is carnivorous.
Don is an animal.
q. Some humans are not evil
Politicians are human.
Politicians are not evil.
2. In the preceding syllogisms, which terms were distributed? Which were
undistributed? Indicate the middle term.
3. Below are five rules of logic governing the syllogism. Apply them to the
syllogisms in question 1. If the argument in any syllogism is not valid,
indicate the particular rule it violates.
a. The middle term of the syllogism must be distributed at least once.
b. If a term is undistributed in the premises, then that term must not be
distributed in the conclusion.
c. A valid conclusion cannot be drawn from two negative premises.
d. A negative premise requires a negative conclusion.
e. A negative conclusion requires a negative premise.
PART THREE
Truth and Falsity
CHAPTER 8
Truth and Evidence
The truth is something that all of us would like to know, except, of course,
when it displeases us. But what is it?
We have been told that the question, "What is truth?" was asked by
jesting Pilate, but that he "stayed not for an answer." His hasty
departure implied that the question was an unanswerable one. We have been told
that the truth is something which, when crushed to earth, is sure to rise
again; that it is something which can never lose out when engaged in a grapple
with falsehood; that it is something which, when told, shames the devil; that
it is folly to tell women the truth; that there is nothing so powerful as
truth, and nothing so strange; and that the truth will make us free. But what
is the truth?
Mankind has been struggling to know the truth for some years now, and at last
accounts the final results are not yet in. There are some who insist that the
truth is absolute; others say it is relative. Some believe that mankind can
achieve the truth; others, the skeptics, deny this possibility. Empiricists
believe that science can give us reliable knowledge, but they are skeptical
with respect to metaphysical truths, that is, the "ultimate" truths
concerning man's origin and destiny. Others are skeptical even concerning the
evidence of their senses.
Our primary concern is with a more prosaic kind of truth, the kind we shall
call "factual." By "factual truth" we mean only this: a
truth relating to the facts of human experience. And, prosaically, let us
define truth. What do we mean when we say that a statement is true? We mean that
a statement agrees with, or corresponds with, the facts. This is the definition
Cardinal Newman had in mind when he said, "Truth means facts and their
relations, which stand towards each other pretty much as. subjects and
predicates in logic." For example: The statement, "Lyndon Johnson was
President of the United States from 1963 to 1968" is true, for it agrees
with the facts. A true statement describes the facts correctly, something like
the way in which a map pictures a territory. When a map shows the relations of
towns, rivers, mountains, and valleys just as they exist in reality, then the
map is a true map. A true statement is like a true map.
"Logical relativism"
Let us now consider some of the implications of this definition of truth. These
implications may be best brought out by considering the theory of "logical
relativism." A relativist, in the logical sense, denies the possibility of
universal truths. He holds that what is true for one man may be false for
another, what is true at one time may be false at another time, and what is
true in one place may be false in another. Truth, he holds, is relative to the
circumstances of the viewer, to his "frame of reference.
The theory of logical relativism is sometimes associated with the theory of
relativity in physics--a familiar theory since Einstein. But they have little
in common, though both emphasize the "frame of reference." In physics
relativity refers to the importance of establishing a "frame of
reference" whenever we describe motion. For example, if we ask, "Is
the furniture in your home moving at the present moment?" the answer
should be "Yes and no, depending on your frame of reference." Since
the earth is moving around the sun at a speed of eighteen miles a second, your
furniture is moving at the same speed relative to the sun. But, relative to the
earth, your furniture is at rest. In other words, the furniture is moving in
one frame of reference, but it is at rest in another. A seated passenger in a
moving car is at rest but also in motion.
But the physicist does not say that truth is relative. He believes that it is
really or "absolutely" the case that motion is relative to a frame of
reference. He does not say that the relativity theory is true for some
scientists and false for others. And this is precisely what the logical
relativist declares. He says that what is true for anyone will depend on his
past experience, his training, his education, and the ideas accepted by his
time and environment. Another man, with different experiences, will find a
different truth. One man's food, in other words, is another man's poison in
matters of truth as well as diet
But when the physicist says that motion is relative, he does not think of the
principle or law of relativity as relative. He believes that it must be true
for everyone that motion is relative to the frame of reference.
In support of his position the logical relativist points to the differences of
opinion that divide the human race in time and place. People once believed that
the sun moved around the earth; today we believe other-wise. But was not the
former belief true for the people of the middle ages? The relativist notes the
different points of view in the Orient and Occident, and in the countries on
different sides of the Iron Curtain. He quotes Pascal's aphorism: "Truth
on one side of the Pyrenees, error on the other side."
Now, there is a good deal to say in behalf of this relativistic position, for
human beliefs do differ. The candid observer will in fact be deeply impressed
by the actual variety of opinions that men hold on all matters of real
importance, on politics, religion, and so on. Consider the different versions
of events that lead to the breaking up of friendships! Each person seems to be
right from his point of view. But none of these considerations requires the
conclusion that truth is relative, if we accept our definition of truth: a true
statement is one that agrees with the facts.
The issue between the relativist and the nonrelativist may be a terminological
one. The dispute will then be verbal, depending on how we define the word
"truth." By "true" the relativist may mean "believed
to be true." Let us try to clear up this semantical confusion. In this
discussion we shall assume that it is possible to know whether some factual
statements do or do not correspond to the facts. We assume that the statement
"Water boils at 212? F. at sea level" does correspond with the facts,
and that this is not "just an opinion," or mere "belief."
It is statements of this kind that we shall refer to as factual truth, or truth
in the strict sense of the term.
There are also looser usages. Thus we sometimes speak of religious or moral
truth. "What is right for one person must be right for all similar persons
in similar circumstances" is a principle that may appeal to us as a
"true" moral principle. What we usually mean is that this moral
belief is one that every rational person ought to subscribe to, though we admit
that it cannot be verified by the eye, or other senses. It is also difficult to
speak of literal truth in assessing the "causes" of great social
events, such as the origin of "capitalism" or the causes of a great
war. Different theories will appeal to us partly because of our prior
sympathies and interests and national origins. Historical facts, as Dwight
MacDonald has put it, are not "solid, concrete (and discrete) objects like
marbles. Rather are they subtle essences, full of mystery and metaphysics, that
change their color and shape, their meanings, according to the context in which
they are presented."
When people argue the respective merits of unrestricted free enterprise versus
government controls in producing higher living standards, value judgments will
influence opinion. Some people prefer freedom, others prefer security, just as
some enjoy a touch of danger, while others abhor danger. Our reception of
theories may also depend on whether we stand to gain or lose by change. It is
difficult to attain literal truth in such discussion, even when
"factual" predictions are made. Prophecies of ultimate disaster
cannot be verified. It is always safe to make a prediction if one does not set
a time limit.
These considerations would seem to support the relativistic position. But even
in the complex issues and predictions just discussed some views will seem to be
more reasonable than others, some will seem plausible or probable and others
preposterous. Our primary interest, however, is in clarifying the meaning of
truth in the literal sense. Let us return to our example of a map which purports
to delineate an area. A true map will correctly describe; a false map will
describe incorrectly. In a largely uncharted area, on the other hand, it may be
impossible to determine whether a tentative map is correct or not. Its truth or
falsity will then be 'uncertain.
Now consider a map of the world printed in the year 1492 which did not show the
continents of North and South America. The people of that time believed that
their map was a true map of the world. Shall we say, rather, that in 1492 men
were ignorant of the true nature of the globe, for some of its continents were
as yet undiscovered, and so, since this map corresponded with their ignorance,
this map was false? Though believed true, the map was never true, though it
corresponded with the known facts. But it did not correspond with the actual
facts. Instead of saying that what was true at one time became false at another
time, let us say rather that what was believed to be true was actually false,
or at least that what was believed to be true at one time is believed to be
false at a different time.
It is quite inappropriate, then, to say that anything can be true here and
false elsewhere, if by true we mean that it correctly describes the facts. We
would never think of saying that a map of the United States is true for
Americans but false for Russians.
But the relativist has another type of argument Take the statement: "It is
warm today." Is this statement true at all times and all places?
Obviously, the answer is No. The statement, "It is warm today, may be true
in July but not in January; it may be true in San Diego but not in the northern
woods of Wisconsin. But this does not prove that truth is relative. The
plausibility of the relativist's point is based on the fact that "It is
warm today" is phrased in the vague language of every-day speech. In order
to test the truth of a statement, however, we must first give it the precision
of a scientific sentence. A scientific statement is dated and located. A
properly formulated scientific statement concerning the weather would read
something like this: "The temperature was 89 Fahrenheit on September 28,
1973 at 3:00 P.M., Pacific Daylight Saving Time, at the meteorological station
in San Diego, California Now, stated in this form, we have a statement that may
or may not correspond to the facts. That it was 89? F. in San Diego at 3:00
P.M. on September 28, however-q true-must be true for the Chinese as well as
for the Australians, and it must be true forever, for our statement was dated
and located for a specific time and place.
Similarly, when we look at a map in a contemporary history book showing
"Europe in 1800," it would be incorrect to say that the map was true
in 1800 but false today. It is the case, of course, that the map corresponds to
the political boundaries of Europe in 1800, but not to those of today. But when
a map is drawn, it carries with it the qualification of a specific time-this is
the way the boundaries looked in 1800. And if the map was true in 1800, then it
will be true forever, for the year 1800.
Let us examine another type of case that seems to support the relativist's
argument-the relativity of perceptions. I look at a book and say that it has a
blue cover. A color-blind man says that it has a gray cover. Are not both of us
correct, though we state contrary ideas? Is not one man's truth error for
another, then? But if we analyze this situation properly, we shall find that it
does not support the relativist's position. We should distinguish between two
types of statements. When I said, "The cover is blue" I may have
meant that the cover will appear blue to a person with normal vision; that is,
its pigment reflects light rays measuring about 485 millionths of a millimeter.
If the color-blind man denies that it is "blue" in this sense, then
at least one of us must be wrong. Whether or not such rays are being reflected
cannot be true at a certain time and place, and also be false. It cannot both
be true for one physicist and false for another, regardless of whether one is
color-blind. But I may have meant something else by "It is blue."
Perhaps I meant, "I have the experience of seeing a blue cover." And
if the color-blind man meant that he has the experience of seeing a gray cover,
then both of us would be right. There is no inconsistency in saying that two
persons have different experiences. It is common, even proverbial, knowledge
that people feel differently when an ox is gored, depending on the ownership of
the ox.
There is a relativity of feeling and of experience, then, but this is not a
relativity of truth. People have different experiences, depending upon the
physical conditions of their bodies, their past experiences, their conditioning
and reconditioning, so that one's perceptions and responses will be relative to
one's "frame of reference." A man cannot be wrong if he correctly
reports what he has himself experienced. But when we talk about the speed of
light, or the light waves reflected from the cover of a book, we are talking
about something other than our own experiences, and if our physical frames of
reference are the same, then conflicting descriptions cannot both be true.
We do not wish to minimize the great importance of the subjective element in
perception. This element is often overlooked, and we frequently objectify
subjective experiences. There are many parables and stories that illustrate
this point. There is the parable about the blind men and the elephant. The
blind men encountered an elephant along the highway and compared notes after
examining him. "'The elephant," said one, is something like a stone,
cool and smooth, and shaped like a curved cylinder." "No," said
another, "the elephant is like a hairy rope." A third said, "No,
the elephant is something very massive and solid, full of little hills and
valleys." Now, obviously, each of the blind men was right in saying that
the elephant had the characteristics he mentioned, though no one had the whole
picture. Each had a different conception of the elephant, and each idea was
consistent with the other ideas when put together, for one man had come into
contact with the tusks, another with the tail, and the third with the side wall
of the body. But each was wrong in thinking that the elephant was nothing but
what be had experienced it to be.
So much for the doctrine of logical relativity." We have argued that { if
a statement is true, then it is always true, and true for everybody. But a new
kind of problem arises: How can we be sure that a statement does correspond to
the facts? For example, the statement, "'There is buried treasure beneath
the house in which I am writing," is either true or false, but no one
knows which, for certain. It is one thing to define truth as correspondence
with the facts, but another to determine whether or not a statement does so
correspond. It is true or false that some world leaders intend to start a third
world war, but we do not know whether this statement actually corresponds to
the facts. Let us examine the new problem: Can we be certain about the truth of
any statement?
Truth and certainty
Are we ever justified in saying, "This belief of mine is absolutely
true?" Are there any beliefs of which we can be absolutely certain? Can we
ever assert with complete confidence that we are completely right and the other
fellow completely wrong? We may deceive ourselves. It is a common experience,
in a law court, to find several honest and sincere witnesses giving different
reports of how an accident occurred. A person sometimes quite positive that he
saw something he did not see. And then there are the distortions, conscious or
unconscious, based on personal interest. In the Japanese film Rashomon a murder
occurs, and three different versions of the incident are presented by those
involved in it as participants. One naturally expects that each of these will
give a biased version in order to justify his own behavior. But there is also a
fourth witness, a supposedly impartial observer, and he too presents a version
that is distorted and subject to doubt.
The question, then, is whether we can ever be certain that any statement is
true. Even if a statement is verifiable by observation, and we make the claim
of truth in the strict sense, can we be positive that we are not mistaken?
The French philosopher Rene' Descartes raised the question: Can we ever be
certain that we really know? Descartes had been brought up with a
traditionalist education, and he later came to believe that many of the ancient
beliefs he had been taught were really false. If some of these beliefs are
proved false, he asked himself, "How can I be sure that any of them is
true?" On what basis can I say, "'This or that belief is certainly
true"? He then initiated a new method of philosophical study, the method
of doubt, and thereby gave a new direction to the course of modern philosophy.
He said that he would doubt everything without exception, and then see whether
there was anything that could successfully withstand critical scrutiny.
Descartes even doubted the evidence of his senses. Perhaps, he said, they
deceive me. The traveler on the desert sees a mirage in the distance, but the
wooded oasis he thinks he sees is only an optical illusion. Descartes said that
he could even doubt whether he was actually awake at a given moment, for he
might merely be dreaming that he was awake. Now do you think that you can be
absolutely certain that you are not dreaming at this moment? Is it conceivable
to you-I do not mean, is it likely, but is it even conceivable--that your alarm
clock will ring shortly, that you will awake, and say: "What a vivid dream
that was-I dreamt that I was reading a book-though I forget what it was
about"? And perhaps you have heard about the college professor who dreamed
that he was lecturing to his college class-and when he awoke he found that that
was exactly what he was doing!
In the end, of course, Descartes found that there was one belief he could not
doubt, and this was the belief that he himself existed. For, he said, if I
doubt my own experience, my doubt implies that there must be a doubter. If I
did not exist, then I could not even doubt that I existed.
Skepticism
In its philosophical form, skepticism is a doctrine which denies that the human
mind is capable of attaining genuine knowledge about any-thing. In the history
of mankind there have been many skeptics who carried doubt very far. In ancient
Greece, in the fourth century B.C., there was a sect of skeptics who maintained
that they could not be certain of anything whatsoever. The senses are
deceivers, they said; they affect us according to the way we feel, and their
reports are always uncertain. We see differently in sickness or health, and in
joy or sorrow. Nothing seems to be so true, their leader Pyrrho said, but that
it has not some-where been thought false, and nothing seems so false but that
it has not somewhere been thought true. The Greeks liked to tell amusing
anecdotes about Pyrrho, but they undoubtedly embroidered on his behavior and
beliefs, for, if the following story were true, it is hardly likely that he
could have lived until the ripe old age of ninety. Pyrrho, one story goes, was
crossing a road, when he saw a chariot approaching. "That looks like a
chariot," he said, 'but how can I be sure that my senses are not deceiving
me?" As he considered the matter, the chariot came by, knocking him down.
His loyal disciples picked him up and dusted him off. They were always around
for just such emergencies. It was also said that when Pyrrho died, his
disciples did not mourn him, for they could not be positive that he was dead.
Perhaps we should note here that Pyrrho was really certain about at least one
thing, namely, that his senses sometimes deceived him. And this inconsistency
confronts all skeptics, for when they say that they know nothing, surely they
must know that they know nothing, and this is a self-contradiction. A person
who says that he knows nothing, in other words, is thereby saying that he does
know something. A consistent skeptic will keep his mouth shut and say nothing
at all.
Bertrand Russell in A History of Western Philosophy, neatly disposes of this
thoroughgoing type of skepticism. "It should be observed," he writes,
"that Skepticism as a philosophy is not merely doubt, but what may be
called dogmatic doubt. The man of science says 'I think it is so-and-so, but I
am not sure.' The man of intellectual curiosity says 'I don't know how it is,
but I hope to find out.' The philosophical Skeptic says 'nobody knows, and
nobody ever can know.' It is this element of dogmatism that makes the system
vulnerable."
Let us sum up for a moment. We began by defining a true statement as one that
corresponds with the facts, and this means that if a statement is true, then,
when stated with proper precision, it must be true for all. When we say,
"The truth changes," or "The truth of one age is the falsehood
of another" what we mean is that beliefs concerning the truth change. We
then discussed the difficulties in determining whether a statement really does
or does not correspond with the facts. Relativism and skepticism properly
emphasize the formidable obstacles to the doctrine that it is possible for us
to know the truth. For what one age considers absolutely true, another age
often rejects. And are not the ideas of today just as much subject to error as
those of the past? How, then, can we ever claim to know the truth about
anything?
Probability
The best answer to this question is found in the theory of probability, which
harmonizes the doctrines of relativism and skepticism with the search for
truth. The scientist, for example, seeks the truth, but he is also very much
aware of the difficulties attendant on this search. He thinks of the actual
attainment of truth-the actual correspondence of a statement with the facts,
where this correspondence is established once and for all-as an ideal which we can
never attain, for it is always possible to question any belief whatsoever. We
can never be certain that no error has occurred or that some factor has not
been overlooked. And to say that a belief may be questioned means that it is
not absolutely certain. No scientific statement, then, is exempt from the
requirement of proof, and no proof can be final.
But though the scientist believes that the truth can never be completely
attained-he believes that we can never be sure that we actually have arrived at
a final answer in the sense of a real correspondence of statement and
reality-nevertheless he believes that we can come closer and closer to the true
answers. By "closer and closer" he means that our answers may acquire
higher and higher degrees of probability.
Before we continue, it is important to note that we are concerned here with
empirical probability, and not with the a priori probabilities of mathematical
calculation. Past experience is of the essence here. We. can have certainty in
mathematical probability provided we make certain assumptions. We can be
certain that an ideal coin will fall heads or tails with equal probabilities.
There are also other instances of logical certainty, where we deal not with the
world of experience, but with the analysis of concepts. Thus we can be certain
that no one will ever see a square with five sides, for a five-sided figure is
not a square. I also know for certain that 2+2=4, by the meaning of the
concepts. But this is pure mathematics, and we cannot have the same certainty
when we apply mathematics to real things in the world of experience. For two
gallons of alcohol mixed with two gallons of water will add up to less than
four gallons of liquid, because of the chemical changes resulting from the
mixing.
The scientist, then, thinks in terms of empirical probabilities. He says that
we can know the probabilities, and that probability is the guide of life. Some
beliefs are warranted, and some are not, depending on the evidence, which
establishes probabilities. And he thinks of probability in terms of degrees.
The notion of degrees of probability may be clarified by a diagram showing the
'line" of probabilities:
0 .01 .25 .50 .75 .99 1
On this line "1" stands for unity, or certainty that something is the
case, as when we can say flatly: "'This is so." "0" means
certainty that something is not the case: "'This is not so," or
"'The belief is false." ".50" is the middle state, when we
say: "It may or may not be the case, and I don't know which."
".75" means something like "It is very likely," and as we
move toward unity the probability increases. ".99" stands for an
overwhelmingly high probability. As we go from ".50" to "0"
we use statements like "It is improbable," or "It is highly
unlikely," and when we reach ".01"- we mean "'There is just
a theoretical possibility." These are not to be taken as technically exact
descriptions of these numbers, but the general idea should be clear.
If we could be sure that we have reached unity or zero, we could be sure that
we have discovered what is really true or false) without any question. Now, in
common-sense terms, there are many statements about which such certainty seems
justified. I know that I am not sitting in a jet bomber as I write these lines,
and I know that at least some motion pictures are not exhibited at the bottom
of the ocean. I know that there is at least one professor of English who cannot
run around the circumference of the earth in less than ten seconds. But in what
follows we shall ignore such simple certainties. For when the scientist says
that "probability is all: we never can be certain," he usually refers
to statements that go beyond our immediate experience; he refers to judgments
in which error really is possible. I can be certain that the water I am now drinking
is cold--this is an immediate experience--but did I have water with my lunch on
Monday of last week? A judgment that I did goes beyond immediate experience; it
relies on memory, and memory is notoriously I tricky. Not many of us, of
course, are like the fabled university professor who was walking near the
campus one day around the noon hour. He stopped a student: "Would you
please tell me whether I am walking north or south?" "You are walking
north, sir," the student replied, "Ah," said the professor,
"then I've had my lunch." And waiters, as Jacques Barzun has
observed, are often even more absentminded than professors.
Memory goes beyond immediate experience, but so does every inference from
things observed to things unobserved. And scientific generalizations and
predictions not only go beyond immediate experience, but beyond all past
experience of any kind whatsoever. And so, whenever we deal with a scientific
law or prediction, we can have no more than probabilities. It is not certain,
for example, that the sun will rise tomorrow; some sort of cosmic cataclysm may
occur five minutes from now. That "all men are mortal'- is also only
highly probable. When we carefully state what we mean by this generalization,
which says that all human beings must someday die, we find that we are making a
prediction that all human beings now alive will die before reaching a certain
age, say 200. To be testable, a determinate time limit must be placed on a
prediction. And this prediction, of course, is not absolutely certain.
Similarly we cannot be absolutely certain that some things are impossible.
There is a possibility that life exists elsewhere in the universe. But it is
unlikely that men as we know them exist there, and still less likely that they
are listening to "acid rock" bands. But even the last fantasy cannot
be ruled out as completely impossible. Strange things do occur, and also
strange coincidences. There is just enough possibility in the following story
to make it amusing: The story is told of two friends who were patients of the
same psychoanalyst One day they decided to play a practical joke on him. Each
would tell the analyst the details of a fantastic dream they invented, every
phrase of which would be narrated in exactly the same way. One patient told his
story, and then, two hours later, his friend repeated exactly the same story.
Toward the end of the second recital the analyst could no longer contain his
amazement, and blurted out, "What a coincidence! In the last twenty-four
hours I have heard exactly the same dream told to me three times!"
Truth and probability
There is an important difference between the meaning of truth and the meaning
of probability. This difference can be brought out very simply by showing the
possible relations of truth and probability: A statement can be true and
probable or false and improbable, but it can also be true and improbable or
false and probable. By truth, in other words, we mean the actual correspondence
of a statement with the facts, but probability is relative to the evidence
available to us. The statement, "'The earth is motionless," once
appeared highly probable, in the light of the evidence then available, but
today we say it is false; "The earth moves, once seemed improbable in the
light of the evidence of the senses, but today we think it is true.
The statement, "'There is buried treasure beneath the house in which I
write" is either true or false at this moment. Either the facts
correspond, in which case the statement is true, or they don't, in which case the
statement is false. The truth of a statement is not relative to the
evidence-the statement either corresponds with the facts or it doesn't, even
though we do not know which. This follows from our definition of truth. But
when I say, "There probably is treasure," or "'There probably is
no treasure," I mean relative to a certain body of evidence.
A statement cannot be true for one man and false for another, but it can be
probable for one and improbable for another, depending upon the available
evidence. On the basis of the evidence available to me, it may be reasonable
for me to suppose that Jones is guilty of cheating at cards; on the basis of
the evidence available to you, it may appear highly unlikely. He is or he is
not, but neither of us may really know for certain which. The truth of a
statement cannot change, but probability judgments vary with every change in
the evidence. In May it may appear likely that there will soon be an end to
tension in the Middle East; in July this may appear improbable. At any given
moment we make our estimate of probabilities on the basis of the evidence
available to us at that time. And, the scientist adds, we can never reach more
than a highly probable conclusion, for all the facts can never be known.
What practical applications can we make of these matters? We can try to assess
the probabilities of the facts upon which we rely whenever we make decisions.
We should be exceedingly careful before we claim we know something for certain.
We should be certain when we have the right to be, of course. A reliable
observer can give an accurate report of what he actually witnessed. In general,
however, we ought to abandon the use of the expression "absolutely
true," except for matters within our immediate experience. And where there
is controversy, let us use more modest expressions, such as, "The evidence
indicates that this or that is probably the case." And let us also
remember Jefferson's distinctions between truths, probabilities, possibilities,
and lies. There is no golden touchstone to guide us in each case. We can simply
try to avoid two extremes. One is a too uncritical attitude, whereby we jump to
conclusions whenever we hear an idle rumor; the other is a too skeptical
attitude which refuses to believe despite good evidence.
Though it happens that different opinions are sometimes reasonably held by
different observers, this does not mean that "everything is just a matter
of opinion." This is a form of skepticism which denies both the
possibility of knowledge, on the one hand, and of ignorance; on the other. Nor
is it the case that everyone is "entitled" to his own opinions,
except in a legal sense. When we say "not entitled," we mean from a
rational or logical point of view. For consider: A man says that every
professor at a certain university is a Communist and an atheist, and then adds
the words, "in my opinion." If you disagree with him, he says that
simply means that his opinion is not your opinion, but that everyone is
"entitled" to his own opinion on the matter. But the facts may make
it overwhelmingly probable that his opinion is a false opinion. Unless one is
intellectually irresponsible, he will inquire into the truth or probability of
his opinions. One is not forgiven for making a dangerously irresponsible
statement merely because he tells us that such is his opinion. We ought to
examine the evidence before we talk. The question the careful thinker will
always ask himself is: "What is the evidence?"
FOR DISCUSSION AND WRITING
1. Read the following statements carefully. In each, how would you define the
author's attitudes concerning the possibility of attaining truth? How would you
describe his method?
a. The improver of natural knowledge (science) absolutely refuses to
acknowledge authority, as such. For him, skepticism is the highest of duties;
blind faith the one unpardonable sin. . . . The man of science has learned to
believe in justification, not by faith, but by verification. (Thomas Huxley)
b. Strong Son of God, immortal Love,
Whom we, that have not seen thy face, By faith, and faith alone, embrace,
Believing where we cannot prove. (Alfred, Lord Tennyson)
c. A struggle for existence inevitably follows from the high rate at which all
organic beings tend to increase. Every being, which during its natural life-time
produces several eggs or seeds, must suffer destruction during some period of
its life, and during some season or occasional year; otherwise, on the
principle of geometrical increase, its numbers would quickly become so
inordinately great that no country could support the product . . . . Linnaetis
has calculated that if an annual plant produced only two seeds-and there is no
plant so unproductive as this-and their seedlings next year produced two, and
so on, then in twenty years there would be a million plants. (Charles Darwin)
d. We say that the general phenomena of the universe are explained-as far as
they can be--by the Newtonian Law of Gravitation. On the one hand, this
admirable theory shows us all the immense variety of astronomical facts as only
a single fact looked at from different points of view; that fact being the
constant tendency of all molecules toward each other, in direct proportion to
their masses and inversely as the squares of their distances .... As to
determining what attraction and weight are in themselves or what their causes
are, these are questions which we regard as insoluble and outside the domain of
the Positive Philosophy; we, therefore, rightly abandon them to the
imaginations of the theologians or the subtleties of the metaphysicians.
(Auguste Comte)
2. This chapter distinguishes between "empirical probability" and a
priori probability." How would you classify each of the following
examples?
a. If I throw this die, chances are one in six that I will throw a two.
b. The chances of the Republican party nominating a woman for president are
mighty slim.
c. It is highly improbable that an ordinary woman could run the mile in three
minutes.
d. Thirty-five percent of today's elementary school students will go to
college.
e. It is probably true to say that what is highly probable is going to happen
sometimes.
f. His batting average is .500; therefore, the chances are one in two that he
will get a hit today.
g. If one were to fly continuously for ten years, the chances are very high that
he would be involved in an airplane crash.
3. The theory of logical relativism is useful in impressing the observer with
the variety of opinions that are held about a controversial topic. It is also
useful in that it reminds us of how easily we might close our minds to another
point of view. Unless we can appreciate the other side of an argument, we
cannot approach any kind of objective evaluation. Construct a statement which
expresses you own point of view. Then, changing only the particulars of the statement
(persons, places, etc.) reconstruct exactly the same statement. Now you are
able to judge the validity of your opinion by seeing it in a new context. For
example, we might read of a Russian intellectual who, in a series of novels,
attacks the communist system. We might approve of such an action because it
fits our own opinions. But, because the statement is recast to describe an
American intellectual who, in a series of novels attacks the capitalistic
system, our attitude might shift. What does such an exercise illustrate about
the relativistic position?
4. The possible relations of truth and probability are four:
a. A statement can be true and probable.
b. A statement can be false and improbable.
c. A statement can be true and improbable.
d. A statement can be false and probable.
Construct a statement for each of these situations, explaining how the meaning
of truth and the meaning of probability are related in each instance.
CHAPTER 9
Knowing the Causes of Things
In 1926 The Nobel Prize for physiology and medicine was awarded to Johannes
Fibiger, a Danish pathologist, for being the first to produce cancer
experimentally in a laboratory animal. In a series of postmortem examinations
of tubercular rats, he noticed that several had suffered from stomach cancers,
a highly unusual occurrence. He learned that the dealer who had supplied his
laboratory with the rats had secured them from a sugar refinery. Investigation
of the place revealed a high infestation with cockroaches, which formed a large
part of the rats' diet. Fibiger suspected there was a connection between
roaches, rats, and cancer. He collected thousands of roaches from the refinery
and fed them to rats from another breeding establishment. When the rats died,
three years later, Fibiger opened them up. To his astonishment, he found many
stomach cancers. A microscopic study of the growths revealed in every case that
they had formed around a parasitic worm, the same worm to which the roach had
been host before is was fed to the rat The larva of the worm, coiled up in the
muscles of the cockroach, later developed into an adult worm in the rat's
stomach. Around this the tumorous growth had appeared.
Fibiger's study illustrates the "cause and effect" relation.
Something unpredicted or unexpected happens; we ask "Why?" We then
search for the cause. This is what happens when a housewife finds that a recipe
did not turn out as expected. Her problem is also one of finding the cause, and
her thinking resembles that of Fibiger, or the thinking of a biologist when he
searches for the cause of a disease. The chief difference between the
scientists and the housewife is that the former are specialists who have vast
stores of knowledge to aid them in solving such problems.
When Francis Bacon said, "Knowledge is power," he meant that science
could give man mastery over the forces of nature by discoveries and inventions.
Bacon's aphorism is a modern echo of the ancient poet Virgil's saying,
"Happy is he who knows the causes of things." For, in large measure,
man's power and control over nature have their origins in his understanding of
the causal connections among events. Our under-standing of causal connections
enables us to improve the fertility of the soil; it is because we know the
causes of diseases and other unpleasant occurrences that we can eliminate them.
Perhaps some day we will know, more precisely, the causes of business
depressions, and then the fear of widespread unemployment will disappear
forever.
Before we proceed further, let us note the assumption which is taken for
granted by Fibiger, the housewife, and the biologist, in our examples. This
assumption is that everything has a cause. We believe that "things don't
just happen by themselves," but that something is responsible for every
single thing that happens. This assumption is sometimes called the
"principle of determinism," and it is a postulate of rational
thinking about the events of human experience. It is called an
"assumption" or "postulate" because it is obviously
impossible to prove that everything has a cause, including all future events.
We don't even know the causes of all sorts of contemporary events-cancer, for
example-but we are sure that there is a cause. Why are we sure? Because, we
say, there must be a cause! The point of the principle is this: We have made up
our minds not to regard any events as being beyond explanation. No events in
human experience are inherently unexplainable.
Much bad writing stems from careless cause and effect reasoning. Merely making
bold assertions about alleged causes and effects will persuade no one; indeed,
such assertions usually hide uncertainty and confusion about one's experience
and ideas. Carefully marshalling evidence and limiting your topic will force
you to scale down your generalizations, and therefore lessen the likelihood of
this common fallacy.
Let us look at the semantical aspects of our problem: Exactly what we mean by
"cause"? In popular speech a cause means "that which is
responsible for a thing's happening," or "the power that produces an
event." These are question-begging definitions, however, for we
immediately ask: What is meant by "responsible" and
"produces"? The notion of cause is involved in the meaning of these
words. Nevertheless we shall begin with the common-sense meanings of the term,
and try to refine them as best we can.
What is a cause?
Let us also examine the way in which the scientist uses the words cause and
effect." When a biologist tells us that yellow fever is caused by a
filterable virus transmitted by the bite of a certain kind of mosquito, he
means that when this virus enters a host body, the host will suffer from yellow
fever, and if an individual suffers from yellow fever, then we know that this
virus has entered his body. By "cause" the scientist means the necessary
and the sufficient conditions for the occurrence of an event. These terms
require definition.
A catarrhal affection of the respiratory tract, commonly called "a
cold," is probably due to a virus. But this virus is presumably present in
many persons all of the time, without their having "colds." The cold
occurs when body resistance is lowered, owing to fatigue or exposure to low
temperatures, dampness, or drafts. If we can assume that this is the correct
theory concerning some types of colds-there is no universal agreement on the
matter-then the virus is not strictly the cause of the cold, but only an
indispensable prerequisite. A necessary condition is defined as a condition
without which an effect cannot occur. We have assumed that colds cannot occur
unless the virus is present, so the virus Is a necessary condition of the
cold's occurrence.
A sufficient condition, on the other hand, is illustrated by the following: A
murderer administers cyanide of potassium to his victim, and the victim dies. We
say that the poison was the cause of death. But the relation of poison to death
is quite different from the relation of the virus to the cold. Death may occur
in other ways than by poisoning: old age, to give just one example. Poison (of
a certain amount under certain conditions) is sufficient in itself to cause
death, but it is not a necessary condition of death. In the former example, the
virus is not sufficient to cause a cold, but there can be no cold without the
virus.
In other words, some conditions are necessary but not sufficient for a given
effect: the virus and colds. Some conditions are sufficient but not necessary
for a given effect: poison and death. If a person has the virus in his body, he
may or may not have a cold, but if he has a cold, then we know he has the
virus. If a certain amount of poison has been administered to man, then we know
that he will die, but if all we know is that a person is dead, we do not know
whether or not he was poisoned.
Though scientists often speak of a sufficient condition as "the
cause"-poison is a cause of death-and though they may even speak of a
necessary condition as the cause, the scientist seeks for something more than
either a necessary or a sufficient condition. He has an ideal conception of a
cause as that set of conditions which are both necessary and sufficient to
bring about a certain effect. The scientist seeks to know the entire
constellation of conditions which will always result in the effect and without
which the effect will never occur. C is the cause of E (in the ideal sense)
when E always occurs following on C's occurrence, and E never occurs unless C
has occurred.
This ideal notion of "cause" may be illustrated by considering a
forest fire. A lighted cigarette is thrown into the brush by a camper. A forest
fire results. Can we say that the cigarette was the cause of the fire? Common
sense tells us that it is quite proper to speak in this way, but the scientist
seeks a more accurate type of statement. He notes that no fire would have
resulted if the leaves had been damp from a recent rain. Dry leaves, then, are
a necessary condition for a forest fire. Dry leaves, plus an igniting element,
plus sufficient wind make up "the cause," for these are necessary and
sufficient conditions of a forest fire. Without any one of these conditions the
forest fire would not occur; when all occur together, a forest fire will always
occur.
In law, and in the ordinary affairs of life, of course, the smoker would be
held responsible, for in law and ordinary affairs we seek the causes of
individual events in order to ascribe responsibility. The lawyer thinks of the
cause as some identifiable act or event without which the result would not have
occurred. The scientist, on the other hand, does not seek to ascribe "responsibility."
He seeks for general causal connections, that is, repeatable patterns in
nature, and so he looks for the conditions that are both necessary and
sufficient.
In practice, however, even scientists are often satisfied with less than the
"ideal" statement of the cause. Nor is it always necessary to know
the cause in the strict sense. It may be enough to learn either the necessary
or the sufficient condition in order to achieve the purpose m mind. Practical
considerations are involved here. It all depends on what one is after: to
produce or to prevent. If scientists wish to produce some-thing, such as a
specific to cure a disease, or synthetic leather, or a stimulus to business
activity, they need only know the sufficient conditions of these effects. If
they wish to prevent or eliminate an effect, such as a disease, it is very
helpful if they know the necessary conditions without which the disease cannot
occur.
Discovering causes
We shall now examine a famous example of the search for the cause of a disease.
In the year 1910 the disease known as pellagra was widespread in many of our
southern states. Pellagra is characterized by skin eruptions, gastric
disturbances, and nervous derangement. The problem became so serious that the
United States Public Health Service sent its Dr. Goldberger to the state of
Mississippi to search for the cause of the disease.
In the early years of the twentieth century it was the prevailing doctrine
among biologists that there was only one possible cause of disease, namely,
germs or microbes. This is the "germ theory" of disease. It was
therefore assumed, at that time, that pellagra must also be caused by a
microbe. The Webster International Unabridged Dictionary, published in 1900,
defined pellagra as "a disease caused by a microbic parasite," and
added that it was probably carried by a fly. Now, one of the implications of
the germ theory is that the disease will be "catching." that is,
infectious, and that it will be transmitted from one person to another through
contact. Goldberger began with the hypothesis that the disease of pellagra was
caused by microbes, and he first investigated to see whether there was personal
contact among the victims. There was.
Goldberger began with a preliminary hypothesis which told him what to look for.
His first observations seemed to confirm his hypothesis. But he was a man who
refused to take anything for granted. and he decided that it would be wise to
test the microbe theory further before undertaking a full-scale search for the
particular microbe which might be the cause of the disease. And so he reasoned
as follows: If pellagra is catching, then whenever we find healthy people
coming into contact with victims, under normal circumstances, the hitherto
healthy persons will acquire the disease. He decided to test this implication
by investigation. He visited a hospital, and he observed that the nurses,
orderlies, and doctors were in close contact with the patients, and that they
made no efforts to avoid such contacts. But none of the nurses, orderlies, or
doctors had ever caught the disease from the patients. Goldberger thereupon
decided that he was on the wrong track, and he rejected the hypothesis that
microbes were the cause of pellagra.
Goldberger had predicted, on the basis of the then prevailing theory, that
contact would cause the acquisition of the disease. The facts were against this
theory, and so he abandoned it. For the scientist there is no such thing as
"a good theory which does not work." If "the stubborn and
irreducible facts" are against a theory, so much the worse for the theory.
If a theory does not work, then it is a bad theory and must be discarded. The
apparent exception to this rule is the case where the facts are misinterpreted.
Goldberger now had to make a completely fresh start. He was sure now that
pellagra was not caused by microbes. He had to look elsewhere for the cause. He
investigated further, and noted that only poor people seemed to suffer from the
disease. He carefully observed their habits of living, especially their diets.
He noted that the pellagra victims lived on a rather uniform diet, consisting
of cornmeal mush, hominy grits, and similar foods. On one occasion, while
visiting an orphanage, he found that some children suffered from pellagra, and
others not. The pellagra victims, further, were all in the six to twelve age
group. This was a surprising situation, and he made further inquiries. He was
told that the children who were over twelve were required to work on the farm,
and they were fed meat to provide them with the necessary energy. The children
under six were given milk, since they were regarded as being in the
"baby" stage. The orphanage's funds were very limited, and so the six
to twelve group received no milk or fresh meat. They were too young to get meat
and too old to get milk.
Goldberger now developed a new hypothesis-that a dietary deficiency,
specifically, the absence of milk and fresh meat, is the cause of pellagra. He
predicted that if the children in the six to twelve age group were supplied
with milk or meat, their pellagra would disappear. The Public Health Service
supplemented the children's diet with the missing ingredients, and all cases of
pellagra in the orphanage were cured. His hypothesis was confirmed.
But science is a never-ending search for truth, and no proof is ever final.
Even confirmed hypotheses will be subjected to retesting, again and again, for
confirmation in previous experiments may have been due to the presence of
special factors. Unknown factors may produce exceptional results in a specific
case. So Goldberger did not stop at this point' And there was another reason
why further tests were called for: Many scientists were still unconvinced that
his investigations had actually disproved the germ theory. After all, they
said, the sufferers from pellagra were in contact with each other, even though
the attendants did not catch the disease. They may have been blessed with
immunity. The critics wanted further proof. And we have often seen that
stubborn refusal to give up a theory in the face of contrary evidence may in
the end be justified.
Goldberger proceeded with new tests. He was given permission to by out a
dietary experiment at one of the state's prison farms. He proposed to feed a
group of convicts a special diet for six months, to test his hypothesis that a
deficiency in diet is the cause of pellagra. Volunteers were asked for; the
reward for participation was to be freedom after the experiment was finished.
One or two lifers accepted, then several others, and he finally had twelve
subjects, all of them in good health. The twelve convicts were isolated from
the others and fed almost nothing but white bread, corn pone, grits, sweet
potatoes, salt pork, cane syrup, and cooked cabbage for six months. They
received no milk or fresh meat' After several months of this diet, beginning in
April, 1915, the twelve convicts became listless; they began to develop severe
abdominal pains and finally developed skin eruptions of the pellagra type. But
the rest of the convicts at the prison farm suffered no such disorders.
Goldberger had now established a solid confirmation of his hypothesis that
nutritional deficiencies alone are sufficient to cause pellagra. The convicts
were now fed a proper diet and, as expected, all of them recovered and were
given their freedom. No further proof of his hypothesis seemed to be necessary,
but there were still some skeptics among the microbe theorists, and so
Goldberger and his assistants decided to do one further experiment to convince
the doubters. The experimenters performed this test on themselves. They
injected blood from pellagra victims into their own bloods streams, and no ill
effects followed. This was a crucial experiment, for if pellagra was caused by
microbes, then they should have become infected. After this test it was an
accepted fact among all biologists that pellagra was caused by nutritional
deficiencies. Later research
has confirmed Goldberger's findings; our knowledge today is simply more precise
than his. It is not the absence of milk and fresh meat C's such that causes
pellagra, but rather the absence from the diet of certain factors in the
vitamin B complex These factors are found in muscle meats, milk, liver, kidney,
fish, and green vegetables.
Goldberger's experiment with the convicts illustrates one of the most reliable
scientific techniques for discovering causes. A careful analysis of the method
employed in this case will enlighten us concerning the nature of a scientific
proof that one kind of thing is the cause of another. This method requires the
setting up of two situations which are identical except for the presence or
absence of the factor which is being investigated. There were two sets of
convicts at the prison farm-those who received a special diet and those who did
not' The living conditions of both groups were identical except for the diet,
and only those who were on the restricted diet got pellagra.
Goldberger's method may be made clearer by a somewhat more detailed example.
Let us assume that a young woman, call her Susan, suffers from skin irritation
and inflammation of the face. She seeks for the cause. The hypothesis occurs to
her that it may be due to her use of face powder. She stops using the powder
for a period of time, and the irritation disappears. She then uses the powder
again, and the irritation reappears. She has solved her problem: She knows the
cause of the irritation and how to avoid it' Face powder seems to be a
necessary and sufficient condition for the occurrence of the irritation.
But this newfound knowledge does not make Susan completely happy. She hates to
give up the use of face powder, for she isn't quite so attractive when she
doesn't use it' The thought occurs to her that it may not be the powder as such
that is the culprit but rather one of the ingredients contained in face powder.
So she consults a chemist friend and asks him for a solution to her problem.
The chemist makes an analysis of the powder and finds that it contains six
ingredients: talc, kaolin, magnesium carbonate, zinc oxide, ochre (for
coloring), and perfume. The chemist now formulates the hypothesis that the
perfume may be the cause of the irritation. To test this hypothesis he prepares
a batch of face powder containing all the ingredients except the perfume, and
he then divides this batch into two parts, to one of which he adds perfume.
Susan now uses the powder without perfume and suffers no ill effects. Then she
tries the part with the perfume added, and irritation appears. This is
sufficient proof that she is allergic to the perfume in the face powder, and
not to the powder itself. (Or to some ingredient in the perfume.)
"The method of difference"
The method we have been illustrating is sometimes called the laboratory or
"controlled experiment" method for determining cause and effect' The
English philosopher John Stuart Mill, who pioneered in the field of scientific
logic, called it the "method of difference." The basic idea is to use
two cases, identical in all respects except one. In the face powder example
this method was used twice, first to determine that the original powder was the
cause of the irritation and then to determine that it was a specific element in
the powder and not the powder as such that was the cause. Goldberger also used
this method, for he kept all living conditions for his two groups of convicts
identical except for one factor-the diet' "Identical," of course,
means "the same in all respects that are considered relevant to the
experiment"' This kind of identity is achieved most perfectly in the sciences
of physics and chemistry, where we can be fairly certain that there are no
relevant factors outside the conditions of the experiment. This method can also
be applied with a large degree of precision in the biological sciences.
The most important point to remember in applying this method is to keep the two
groups of things identical (or substantially the same) except for one element.
If the "addition" of this element results in the effect, and if the
effect never occurs when it is absent, then we have found the probable cause.
But it is not always possible to apply this method, for it may be impossible to
isolate one factor and keep all others the same. This is especially true in the
social sciences. The variables involved in human actions are exceedingly large
in number, and human behavior is vastly more complex than is the career of a
germ. Since we cannot be sure that we have accounted for all the relevant
factors in a situation, we cannot be sure that we have an identity of all
factors except one. An illustration or two may be helpful here. Suppose we try
to isolate a single factor to explain why X defeated Y in an election. Let us
assume that X favored capital punishment and Y did not. If it is claimed that
this was the reason for his victory, we must look to see whether all other
factors were substantially the same. Were the men equally able, and did they
agree pretty much on all other issues? Did they have the same number of
influential friends and enemies; did they have equal campaign funds, etc? The
attitude toward capital punishment may have been a factor in X's victory, but
not necessarily the cause, or even the most important factor.
Let us now examine a sociological problem. Consider the difficulties in
applying the differential method to the problem of juvenile delinquency. Can we
find two individuals, one a delinquent and one a good junior citizen, in whom
all characteristics except one are alike? It is unlikely that their antecedents
and experience will differ in one respect and one respect only. No two
individuals are identical except for a single difference. Similar
considerations apply to such problems as finding the causes of divorce or war
or dope addiction. But many people apply the Method of Difference carelessly,
forgetting that other conditions are not the same. The next time someone tells
you that he knows "the cause" of war or juvenile delinquency or some
other social evil, check to see whether he has isolated a single causal factor,
all others remaining the same. If not, has he used one of the other causal
methods to be described shortly?
These difficulties have led some social scientists to abandon the search for
the "causes" of social behavior and to limit themselves to a search
for "tendencies" stated in statistical form. Children from broken
homes, for example, may show a greater tendency toward delinquency than those
from stable families. If so, that would be useful knowledge. A tendency,
however, is simply a modest way of indicating a possible causal connection.
When appropriate, and when properly applied, the method of difference, or the
differential method, is the most convincing possible kind of proof that we have
found the cause. But sometimes it is impossible to apply this method.
"The method of variations"
We turn now to a second method for determining the causes of events. This
method is called the "method of variations," known in statistics as
the "method of correlations." Examples: A manufacturer of cosmetics
uses newspaper advertising to sell his product. Each ad contains a coupon
offering the reader a free sample. The advertiser finds that an increase in the
lineage of the ad brings in a larger number of coupons; a decrease brings a
smaller number. An increase in the crop of oranges (other things 'being equal")
is followed by lower prices; a smaller crop by higher prices. In other words,
if two kinds of factors vary "directly," so that an increase in one
factor is always followed by an increase in a second, and a decrease by a
decrease (the cosmetics example), or if they vary "inversely," so
that an increase is always followed by a decrease, and a decrease by an
increase (the orange crop), then there is a good reason to suspect a causal
relation between the two factors. They vary together, or concomitantly.
There are, of course, possibilities of error in applying this method. It has
been pointed out that there is a tendency for women's dresses to be shortened
during periods of "prosperity" (the 1920s, the period of the Second
World War, and, as we know, they reached new heights in the unparalleled boom
of the '60s) and lengthened in periods of depression (1932). But the
correlation may be an "accidental" one. It was once discovered that
over a period of time the number of storks in Sweden varied in precise
proportion to the number of human births in the United States, but this does
not prove a causal relation between the two factors. We should also seek other
kinds of evidence which make it reasonable to believe that a causal relation
does in fact exist.
"The method of agreement"
A third method, called the "method of agreement," seeks a common
factor in the conditions which precede the effect that interests us. If a
single common factor is discovered, this often gives us some probability that
we have found the cause. For example, the public health authorities in a small
town were confronted with an outbreak of typhoid fever. The authorities
investigated the food and beverages consumed by the victims just prior to the
outbreak of the fever and found that all of the victims had just one thing in
common: they had all attended a picnic and had drunk from the water in a well
at the picnic grounds. Since this was the only common factor, it was a
reasonable inference that the well was contaminated. Laboratory tests showed that
the water contained the typhoid bacillus.
The method of agreement, or "common factor method," is not so
convincing as the differential method, but it is useful when the more precise
method cannot be applied. The method of agreement also is subject to careless
applications. As an example of the dangers involved in the use of this method,
consider the anecdote concerning the man who wished to find the cause of his
becoming intoxicated every time he attended a cocktail party. This man had
heard of the success of the public health authorities in using the "common
factor method," and he decided to emulate them. He looked for the single
common factor in the beverages he had imbibed on each occasion when he became
intoxicated. After making an exhaustive study of the matter, he found that
intoxication followed after drinking bourbon and soda, Scotch and soda, rye and
soda, rum and soda, brandy and soda, and vodka and soda. Since soda was the
only common factor in every instance, he concluded that soda was the cause, and
at the next party he attended he insisted on drinking his whiskey straight!
"Negative" tests for hypotheses
Thus far we have examined methods which tell us how to search for causes and
how to discover them. It is also well to remember two negative tests which must
be "passed" by hypotheses which assert a causal connection. The tests
are these: (1) Nothing can be the cause if the effect fails to occur in its
presence, and (2) Nothing can be the cause if the effect occurs in its absence.
Let us illustrate.
Dr. Goldberger eliminated microbes as a possible cause of pellagra, for,
although microbes are always carried in the bloodstream, pellagra failed to
occur when the blood of victims was injected into the blood-streams of the
investigators. Microbes are not the cause of pellagra, then, for the effect
(pellagra) failed to occur in their presence. An ancient Roman, Pliny the Elder
(A.D. 23-79), once disproved the claims of the astrologers by. using the same
test' "If a man's destiny is caused by the star under which he is
born," Pliny wrote, "then all men born under the same star should
have the same fortune. But masters and slaves, kings and beggars, are born
under the same star at the same time." In other words, the star cannot be
the cause of a particular kind of destiny, for in its presence that particular
kind of destiny fails to occur. One more example: the belief that a
"broken home" is the cause of juvenile delinquency. But not all
children from broken homes become delinquents. There may be some connection, of
course, but we cannot yet speak in terms of cause and effect"
The second negative test may also be illustrated by the delinquency case.
Children from stable homes sometimes become delinquents. The effect has
occurred in the absence of the factor "broken home," so this cannot
be the "cause" of delinquency. All too often we forget to apply the
negative test' We often jump to the conclusion that one thing is the cause of
another because we forget that there may be negative evidence. This happens most
frequently when our emotions cause us to try to prove what our hearts desire.
Sir Francis Bacon called this tendency to ignore evidence that does not suit
our purposes an "Idol of the Tribe," by which he meant a faulty habit
of thinking, common to the human race. In Aphorism 45 of his Novum Organum he
gave a striking example of this error:
And therefore it was a good answer that was made by one who, when they showed
him hanging in a temple a picture of those who had paid their vows and then
escaped shipwreck, and would have him say whether he did not now acknowledge
the power of the gods-"Aye," asked he, 'but where are they painted
that were drowned after their vows?" And such is the way of all
superstition, whether in astrology,, dreams, omens, divine judgments, or the
like; wherein men, having a delight in such vanities mark the events where they
are fulfilled, but where they fail, though this happens much oftener, neglect
and pass them by.
Fallacies in causal reasoning
The search for causes, as we have seen, is beset with numerous forms of
fallacious reasoning. Perhaps the most important of the general fallacies in
causal reasoning is the "post hoc," an abbreviation for the ##Latin
expression "post hoc, ergo propter hoc": "after this, therefore
because of this." This means: The fact that one thing follows upon another
is no proof that the first is the cause of the second. For example, I have a
pain in my shoulder and take a pink pill. A little later the pain disappears. I
say that the pill was the cause of the disappearance of the pain. Why? Because
I took the pill and then the pain disappeared. But the mere fact that one thing
follows another is no proof that the first is the cause of the second. To prove
a causal connection we must use one of the procedures discussed earlier. It may
be that the pain would have disappeared even if I had not taken the pill. An
effect, of course, always follows the cause. When one thing follows another the
first may be the cause, but more proof must be forthcoming before we can say
"proved." Mere succession in time is not proof. However, succession
of events may involve a causal connection, as in the sharp increase in births
in New York City following upon the Great Blackout of October, 1965. The fact
that the increase came nine months later indicates the likelihood of some
connection. In this case the inference is based on more than mere succession in
time.
And here are some more post hocs: Old man Jones celebrates his one hundredth
birthday, and the newspaper reporters, as usual, are on hand, curious to know
just how he did it; "Well," says old man Jones, "I drink a pint
of beer every day." He drank a pint of beer, and he lived another day, and
he did this again and again. But there are teetotalers who live until one hundred,
and some beer drinkers have been known to die young. Post hoc reasoning is
common also at baseball parks. A fan yells, "Hit a home run!" and the
batter hits one over the fence. It will be difficult indeed to convince this
fan that his yell was not the cause of the home run. Baseball players are
notoriously superstitious, and their superstitions are based on post hoc
reasoning. Their managers are superstitious too. If a manager fails to shave on
the day when his team ends a losing streak, he will probably assume that his
failure to shave was the cause of the victory, and allow his' beard to grow
until his team loses in. A recent example of post hoc reasoning: In February,
1971, sixty-five people were killed in a southern California earthquake. Prior
to the earthquake a fundamentalist preacher had warned that southern
Californians were sinners and that God planned to punish them. This was
sufficient evidence for many people. They were absolutely convinced that the
earthquake was God's punishment for their wicked ways.
Another error in causal reasoning is that of reversing the connection between
the cause and the effect. This error is often called
"over-simplification." An English writer once argued that since those
among the English poor who had cows were the most industrious, the way to make
the others industrious was to give them cows. Though the writer's reasoning was
fallacious, he may, nevertheless, have a point. Is it not possible that, with
cows to care for, many people would become more industrious and more concerned
with their own welfare? The next example is a clearer example of the fallacy of
reversing cause and effect: We find that students who major in mathematics
generally rank high scholastically. This is considered proof that the study of mathematics
makes students bright, but perhaps only bright students major in mathematics.
The error of reversing cause and effect puts the cart before the horse. An
interesting application of this reasoning occurs in the frequent controversies
concerning the "10w cultural level" of television programming. We
accuse the television industry of debasing the taste of the public by a
"vast wasteland" of low-quality shows. The industry retorts that it
gives the public what it wants; if it aims higher, it will lose money. The
industry claims that it is the low taste of the public that is responsible for
a situation that they, too, deplore. This situation is one of reciprocal
causation. The low taste affects the industry, and the industry may further
debase an already low taste. This is the familiar "vicious" circle,
but there is also a 'beneficial" circle. If the programs improve, taste
will improve, and then the spiral will be reversed.
One final point. Suppose we find that students who smoke heavily are less successful
in their studies than students who don't smoke. It would be a mistake to
conclude from this that smoking is the cause of low grades. It may be so, but
the facts cited are not proof that it is, for it may be that both the heavy
smoking and the low grades are due to other factors, such as personality
traits, lack of academic interests, or extracurricular activities.
A summary of this chapter may be welcome. We defined a "cause" as the
set of necessary and sufficient conditions of an event. When we say that X is
the cause of Y, we mean: (1) If X occurs, Y will always occur; that is, X is
the sufficient condition of Y, and (2) if X does not occur, Y will not occur;
that is, X is the necessary condition of Y. This is the meaning of
"cause" in the ideal sense of the term; in practice, scientists are
often satisfied when they know either the sufficient or the necessary
conditions.
We then noted three methods used by scientists in discovering causes: the
methods of difference, variations, and agreement. The first is the most
rigorous of these methods, but it demands complete control over all of the
factors in a situation, and so is difficult to apply outside of the physical
and biological sciences. Two situations must be exactly alike except for the
presence or absence of a single factor, the effect occurring when this single
factor is present.
These three methods tell us what the cause is. There are also methods for
eliminating "false causes," which tells us what the cause is not:
"Nothing can be the cause in whose absence the effect occurs."
"Nothing can be the cause in whose presence the effect fails to
occur."
We noted some of the errors or fallacies which result from the careless use or
application of the methods, and we concluded with a discussion of some of the
fallacies with which the search for causes is beset The major fallacy is called
the "post hoc": the assumption that if Y follows X, this in itself is
sufficient proof that X is the cause of Y. We also reverse causes and effects,
and we sometimes assume that when two things are associated with each other,
one must be the cause of the other, when actually both may have a common cause.
FOR DISCUSSION AND WRITING
1. In the following statements, causes and results are described. In each,
decide whether the cause was necessary, or sufficient, or both.
a. The life expectancy of American men ranks twenty-fourth in the world. The
reason is that American men tend to overdo things. They consume too much in the
way of calories, cholesterol, nicotine, and alcohol.
b. Beverly found herself allergic to candy wafers. A chemist friend analyzed
the contents of the candy and found six elements: sugar, corn syrup, retsyn,
copper gluconate, vanilla, and a coloring agent. By isolating elements and
testing Beverly's reactions, it was found that vanilla was the cause of the
allergy.
c. After years of testing, medical researchers have isolated the gumbuba fly as
the sole cause of the rare tropical disease synchromose.
d. Frank smoked cigarettes for twenty-seven years. He died of lung cancer last
fall
e. Because of the extreme rainfall this spring, floods occurred throughout the
South.
2. Analyze the following evidence and conclusion. How reliable is the
conclusion?
a. New York and New Jersey are heavily industrialized.
b. Sixty percent of the population of those two states are directly involved in
some way with light or heavy industry.
c. Eighty-five percent of the industries surveyed are housed in plants in which
a residual noxious fume of some kind is emitted.
d. A recent survey shows that the rate of cancer in New York and New Jersey is
hi her,, on a per capita basis, than in any other state. Conclusion: The high
cancer rate in New York and New Jersey is the result of industrial pollution.
3. Which of the following statements represent logical cause-effect reasoning?
a. Man has always gone to war, and therefore it is impossible for us to ever
end war.
b. The 1974 Buick must be well constructed, because Buick has a reputation for
quality.
c. The United States is a major world power because of the strong moral fiber
of its citizens.
d. I failed my English exam today because I stayed out late last night and
neglected to review.
e. A copper bracelet cured my rash.
4. Select one of the following statements and write a paragraph in which you
analyze the relationships between the causes and the effects:
a The man who reads fast is the man who is on his way up in this world. In a
recent survey, 500 men and women were tested for reading speed. Those who were
capable of reading 1,000 words a minute or faster had an average income of
$15,000. Would you like to make $15,000? Write for our brochure on the Hogins
Speed Reading Course.
b. The spring of 1973 was one of the coldest in recorded history. This unusual
cold snap followed immediately upon the French testing of a new hydrogen bomb
in the Pacific Ocean. The editorial pages of American newspapers were filled
with the cries of "Stop the French bomb testing, or it will destroy the
balance of nature!"
c. At a midwestern university 1,000 students were interviewed concerning their
dating habits. Ninety-one percent of those who were going steady were in the
bottom half of their class. Those who went out on dates with two or more
persons were in the top half of their class. Those who seldom dated were in the
top 15 percent of their class. The conclusion is inescapable: because of the
obligations of steady dating, the student who dates one person solely will
suffer scholastically. Indeed, there seems to be a direct relationship between
the number of dates and one's academic standing.
5. This chapter describes three methods for determining cause and effect: the
"method of difference," the "method of variations," and the
"method of agreement." For each of these methods write a paragraph in
which you apply the method to an effect and an alleged cause.
CHAPTER 10
Are All Generalizations False?
We begin with a generalization: human beings are great generalizers. Every race
has its proverbs, and proverbs are generalizations. "It never rains but it
pours." "Faint heart never won fair lady." "Familiarity
breeds contempt." Sometimes, of course, these proverbs are incompatible
with each other, as in "Absence makes the heart grow fonder," and
"Out of sight, Out of mind." Listen attentively to those around you,
and note the generalizations that float into every conversation: Women drivers
are the most careless. Professors are absentminded. The Irish are alcoholics.
Gentlemen prefer blondes. Politicians are crooks. The French are great lovers.
People on welfare don't want to work. And so on. After more of the same we may
be tempted to agree with Justice Holmes that "the chief end of man is to
frame general propositions, and no general proposition is worth a damn,"
Our awareness of the inadequacy of "sweeping generalizations" may
lead us to say that all generalizations are false. But this is truly a sweeping
generalization! And worse: if it is true, then the witticism that "all
generalizations are false, including this one" would appear to be
justified. But this will not do either, for this generalization asserts that it
itself is false, from which it follows that it is not the case that all
generalizations are false. Or perhaps we should say that "all
generalizations are half-truths-including this one"? But this is not much
better. The fact of the matter is that some generalizations are true, others
are false, and still others are uncertain or doubtful. The deadliness of this
platitude may be forgiven because of its truth.
By a "generalization" is meant a general law or principle which is
inferred from particular facts. As a sample of the way in which we arrive at
such generalizations, consider the following: Some years ago I saw my first
Italian movie. The directing, the acting, the dialogue, the lighting-all were
superior. Encouraged by this initial experience, I saw another Italian movie.
It, too, was enjoyable. I saw other Italian movies, always with the same
results-comedies, dramas, "Westerns," thrillers. I generalized: All
Italian movies are enjoyable.
A generalization is a statement that goes beyond what is actually observed, to
a rule or law covering both the observed cases and those that have not as yet
been observed. This going-beyond is called the "inductive leap." An
inductive leap is a "1eap in the dark," for the generalization may
not be true, even though the observations on which it is based are true. Thus,
there may be a bad Italian movie-happily I have not seen it-but if so, then I
should not say that all are good.
A generalization involves an "inductive leap." The word induction,
from Latin roots meaning "to lead in," means that we examine
particular cases and "lead in" to a generalization. Induction is the
method we use when we learn lessons from our experience; we generalize from
particular cases. Deduction, on the other hand, refers to the process of
"drawing out" the logical consequences of what we already know (or
assume) to be true. By induction we learn that Italian movies are enjoyable. If
a friend tells us that he saw a bad movie, then by deduction we know that he
did not see an Italian movie. Both induction and deduction are essential
characteristics of rational thinking.
A generalization is a statement of the form: "All A's are B's."
"All" means exactly what it says: all without exception. A single
exception overthrows a generalization of this kind. Before we proceed further
we must first dispose of a popular confusion concerning the expression,
"The exception proves the rule." This is a sensible statement when
properly interpreted, but it is sometimes understood in a manner that makes it
nonsense. If I say that "all A's are B's," a single exception will
make my statement false. Now, suppose that someone says, 'The fact that there
is a bad Italian movie proves that all are good because it is an exception, and
the exception proves the rule" Does a wicked woman prove that all worn-en
are saints? The sensible interpretation of the expression, 'The exception
proves the rule," is this: When we say that a certain case is an 4'ex-ception,"
we imply that there is a rule which generally holds. When a mother tells her
daughter, "Have a good time at the prom, and, for tonight, you have my
permission to stay out until 3 A.M.," she implies that this is an
exception to the rule which requires earlier reporting. A statement that
creates an exception implies a rule for all nonexceptional cases, but a
generalization that is stated as a rule without exceptions (all A's are B's)
would be overthrown by a single exception.
Scientific laws, stated in the form "All A's are B's," or some
variation thereof, are never "violated." When an exception to a law
is definitely established, the law in its previous form is abandoned, but it
may be p05sible to revise it to exclude the "exception" as a special
case because of special circumstances. The revised law: "All A's, under
such and such conditions, are B's." Water freezes at 320 F. at sea level.
All too often "general propositions are not worth a damn," as Holmes
remarked. This is because we generalize too hastily on the basis of
insufficient evidence.. The fallacy called the 'hasty generalization"
simply refers to the fact that we jump too quickly to conclusions concerning
"all." For example, we see a woman driving carelessly, and
generalize:
"All women are poor drivers." We see a car weaving in and out of
traffic, and note that it has a California license: "Wouldn't you
know," we say. "A California driver. That's the way they all drive
out there." Anita LOOS'S gay heroine thought that gentlemen preferred
blondes because she was a blonde and men were attracted to her.
We learn that Napoleon got along on five hours of sleep. From this we may
conclude that 'five hours of sleep is all that anybody really needs." Our
assumption is that what Napoleon could do, anybody can do, until we learn that
we are not Napoleons. (If we don't learn this eventually, we aren't permitted
to circulate freely.) The next example is undoubtedly the worst example of
generalizing ever committed: A man declared that all Indians walk in single
file. When challenged for his evidence, he replied, "How do I know that? I
once saw an Indian walk that way.
Hasty generalizing is perhaps the most important of popular vices in thinking.
It is interesting to speculate on some of the reasons for this kind of bad
thinking. One important factor is prejudice. If we are already prejudiced
against unions or businessmen or lawyers or doctors or Jews or Negroes or
whites or gentiles, then one or two instances of bad conduct
by members of these groups will give us the unshakable conviction that
"they're all like that." It is very difficult for a prejudiced person
to say, "Some are, and some aren't." A prejudice is a judgment formed
before examining the evidence.
A psychological reason for asserting "wild" generalizations is
exhibitionism: The exhibitionist desires to attract attention to himself. No
one pays much attention to such undramatic statements as "Some women are
fickle," or "Some politicians are no better than they ought to
be." But when one says that "all men are liars," this
immediately attracts notice. Goethe once said that it is easy to appear
brilliant if one respects nothing, not even the truth.
Let us avoid careless and hasty generalizing. The fault of bad generalizing,
however, need not make us take refuge in the opposite error-the refusal to
generalize. This error is illustrated in the anecdote concerning the student
who wrote an essay on labor relations, in which he argued for equal pay for
women. Women, he wrote, work hard; they need the money; they are the foundation
of the family; and, above all, they are the mothers of most of the human race!
There is an old anecdote about the cautious man whose friend pointed to a flock
of sheep with the remark, "Those sheep seem to have been sheared recently."
"Yes," said the cautious man, "at least on this side."
Generalizations are dangerous, but we must generalize. To quote Justice Holmes
once more: he said that he welcomed "anything that will discourage men
from believing general propositions." But, he added, he welcomed that
"only less than he welcomed anything that would en-courage men to make
such propositions"! For generalizations are indispensable guides. One of
the values of knowledge lies in its predictive power-its power to predict the
future. Such knowledge is stated in generalizations. It is of little help to me
to know that water froze at 320 F. yesterday unless this information serves as
a warning to put antifreeze in my car radiator before winter comes. History, in
the "pure" sense of this term, merely tells us what has happened in
the past, but science furnishes us with general laws, and general laws tell us
what always happens under certain specified conditions.
Science is interested in the general, rather than in the particular or individual.
When Newton saw an apple fall from a tree in his orchard-even if this story is
a fable, and therefore false in a literal sense, it is true in its insight-he
was not interested in the size and shape of the apple. Its fall suggested an
abstract law to him, the law of gravity. He framed this law in general terms:
Every particle of matter attracts every other particle of matter with a force
directly proportional to the product of their masses and inversely proportional
to the square of their distances. Chemists seek general laws concerning the
behavior of matter. The physician wants to know the general characteristics of
the disease called myxedema, so that when he has a case he will recognize it
and know exactly how to treat it. The finding of general laws, then, is the aim
of all science-including history insofar as it is a science.
The problem of the scientist is one of achieving sound generalizations. The
scientist is careful not to make assertions which outrun his evidence, and he
refuses to outtalk his information. He generalizes, but recognizes that no
generalization can be more than probable, for we can never be certain that all
the evidence is in, nor can the future be guaranteed absolutely-not even future
eclipses of the sun and moon. But the scientist knows that certain laws have a
very high degree of probability.
Let us look at the logic involved in forming sound generalizations. The number
of cases investigated in the course of formulating a scientific law is a factor
in establishing the truth of the law, but it is by no means the most important
one. Obviously, if we observed one hundred swans, all of which are white, our
generalization that "all swans are white" does not have the same
probability it would have if we observed one thousand swans. But no matter how
great the number of specimens involved in this type of observation, no more
than a moderately high degree of probability is ever established. Countless
numbers of white swans were observed throughout the ages (without any
exceptions), and then in the nineteenth century black swans were observed in
Australia.
The weakness of the method of "induction by simple enumeration of
cases" is amusingly illustrated by Bertrand Russell's parable in his
History of Western Philosophy:
There was once upon a time a census officer who had to record the names of all
householders in a certain Welsh village. The first that he questioned was
called William Williams; so were the second, third, fourth. . . . At last he
said to himself. "This is tedious; evidently they are all called William
Williams. I shall put them down so and take a holiday." But he w as wrong;
there was just one whose name was John Jones."
Scientific generalizations based on other types of evidence than simple
enumeration often acquire a much higher degree of probability after only a few
observations. When a chemist finds that pure sulphur melts at 125 degrees C. in
an experiment in which every factor is accurately analyzed and controlled, the
law concerning the melting point of sulphur achieves as great a degree of
certainty as is humanly attainable. Accurate control of every element of one
case, then, is more important in establishing probabilities than is mere
enumeration of many cases.
A single carefully controlled experiment, such as the sulphur experiment, can
give us a much higher degree of probability than the mere observation of
thousands of swans. The reason is that we also know that no chemical element
thus far observed has a variable melting point under conditions of constant pressure.
The chemical law is thus consistent with and is borne out by the rest of
chemical knowledge, whereas the 'law" holding that all swans are white was
based on an "accidental factor. Or consider the generalization concerning
the mortality of mankind. This law is based not merely on the fact that
countless numbers of human beings have died in the past, but also on the fact
that all living beings must, by reason of physiological limitations, die, and
that all matter wears out in time. So the harmony of a particular
generalization with the rest of our knowledge is also a factor in giving it a
high degree of probability.
So much for the logical analysis of generalizations. Thus far, we have been
concerned with "uniform" generalizations, which take the form:
"All A's are B's." A generalization, we have seen, is a statement
that says something about "all" of a group, the evidence consisting
of observations of items in which we always find a single characteristic. The
observed cases are taken as a sample of the whole group or population with
which we are concerned. We observe a number of swans and take these as a sample
of all swans, past, present, and future. We find that all are white, and make
the inductive leap: Swans are always white, everywhere.
"Statistical" statements
We shall now examine "statistical" statements. Statistical statements
give us information, not about characteristics possessed by all of a group or
population, but about those possessed by a definite proportion (or most) of the
group or population, as when we say, "Most A's are B's," or
"Sixty-five percent of all A's are B's." The first thing to note here
is that statistical statements may, in fact, be generalizations and thus
involve the notion of "all." This point involves very important (and
common) misunderstandings.
In order to make this point clear, let us reinterpret our "uniform"
generalizations. We say: "The sample is so-and-so (all observed swans are
uniformly white) -therefore, the whole population of swans is uniformly
white." Now, we do the same sort of thing in statistical generalizations.
We say, "In the sample of redheads we examined, 53 percent were
hot-tempered; therefore, 53 percent of all redheads are hot-tempered." (Or
53 percent of the whole population of redheads is hot-tempered.) Logically,
both examples, uniform and statistical, are of the same type, for in each we
make the inductive leap from the sample to the whole population. The only
difference between them is that in the one case we assert a uniform character in
the whole population, while in the other we assert that a characteristic holds
in a certain proportion in the whole population.
This fundamental point will help us to evaluate the degree of probability of a
statistical generalization. We saw earlier that uniform generalizations can
never be absolutely certain-though for practical purposes we often consider
them so, especially in the physical sciences. The probability of a
generalization depends especially on the quality and also on the quantity of
the cases that constitute the sample. The same holds for statistical
generalizations, which may have a high probability, depending on the character
of the evidence. Though the inductive leap is involved m all generalizations,
in some cases the leap is justified. Let us examine the criteria of
justification for the leap.
Before we proceed we shall discuss an important distinction: that between the
sample and the inference we draw from it. It is one thing to describe a sample
accurately and quite another to draw an accurate inference. If I say, "I
have observed ten swans (the sample) and all were white," we may assume
that the sample is accurately described. But if I now go on to generalize (that
is, draw the inference) concerning all swans, my inference may not be a good
one. A generalization always involves a leap in the dark," sometimes
justified and sometimes not Similarly, if I say, "I have talked to ten
friends concerning their income, and six [60 percent] told me that they earned
more than $20,000 a year," the description of the sample may be accepted
as true. But suppose I now go on to make the following inference:
"Therefore, 60 percent of all Americans earn more than $20,000 a
year." This would be a hasty generalization indeed.
We distinguish, then, between the sample and the inference. A study of I. Q.
scores of 18,782 Mr. Force enlisted men revealed that those who were
accountants in their civilian lives had the highest median (128.1), while those
who were miners were among the very lowest (92.0). Now, these figures involve
no inferences. They simply describe the actual facts in the sample. We draw an
inference, on the other hand, when we assume that all accountants and all
miners in the United States would have shown the same kinds of averages as the
sample. In our discussion, henceforth, we shall be concerned only with the
logical problems involved in statistical inferences.
Suppose that a public opinion poll was recently taken. The polling I
organization tells us that 58 percent of the American people approve of the
record of the present administration in Washington. How do they know this? Let
us examine the evidence on which this finding is based. Obviously not everyone
was consulted. A sample was taken. There were 3,000 interviews. Since there are
approximately 150 million adults in the United States, each individual in this
sample is taken as representative of 50,000 adults. Further, in the sample,
1,000 persons said that they had no opinion." Eleven hundred and sixty
said that they "approved," and 840 said they did not. Thus 58 percent
of those with opinions approved, and this means, we are told) that
approximately 87 million Americans approve. The pollsters assume that the
undecided individuals will probably divide in the same proportion as the others
when they make up their minds.
Now, we are not raising any questions concerning the truth of the report made
of the sample. But is the inductive leap from the sample to the generalization
concerning 150 million people justified? It may be. It all depends upon the
reliability of the sample. What makes a sample reliable? It must be fair,
unbiased, and representative of the whole. But the crucial problem is to
determine whether or not it has these characteristics.
The size of the sample is obviously important. A sample of 100 would not be so
reliable as one of 1,000, and 1,000 would not be so reliable as one of a
million. But large numbers in themselves may not be the most important factor
in establishing the reliability of generalizations or inferences.
The unimportance of large numbers as such is best illustrated by the ill-fated
Literary Digest presidential election poll in 1936. The magazine sent
pre-election ballots to 10 million persons and received over 2 million
responses. The responses showed Landon running ahead of Roosevelt. In the
election in November, however, Roosevelt got about 28 million votes, Landon
around 18 million.
The reason for this colossal failure was the unrepresentative character of the
sample. The Digest took names "at random" from telephone directories
and lists of registered owners of automobiles. These were relatively well-to-do
folk. The lower income groups, however, were completely, or almost completely,
unrepresented.
An ideal sample is one taken "at random" from the entire population,
and not from a selected portion of the population being studied. The Gallup
poll, for example, uses a special kind of random sampling, and, barring a
spectacular failure in 1948, has been far more successful than the Literary
Digest poll. Let us see how the Gallup poll operates. A sample of 3,000
individuals is taken, but with great care to make the sample representative.
The population is classified into subgroups by geographic regions, by rural or
urban residence, economic status, age, education, and declared politics. In
1948, for example, Gallup estimated that 28 percent of the American people
lived in the Middle Atlantic states, 10 percent on the West Coast; that 34
percent lived in cities of over 100,000 population; that 23 percent were of an
"average" economic station; that 43 percent were between the ages of
thirty and forty-nine; that 42 percent had gone to high school; and that 38
percent called themselves Democrats, 36 percent Republicans, and 26 percent
independents or members of smaller parties. The 3,000 interviews in the sample
were distributed so that each geographic area, each economic group, etc., would
be represented in its appropriate numerical strength.
Individuals are then chosen "at random," rather than by selection,
from within each subgroup, and the resulting sample is highly representative of
the whole population. The Gallup poll enjoys a successful record, on the whole,
except for 1948. In other words, the method works, and one must respect its
findings. But no poll can ever eliminate the possibility of error or guarantee
accuracy except within a margin of error of several percentage points. And in a
presidential election forecast the pollster is either completely right or
completely wrong in predicting who will win. Odds of ten to one against a
candidate of one of the major parties are probably not justified even if all
the polls are confidently unanimous as to the final results. These were the
odds against Harry Truman in the presidential election of 1948!
An election prediction can be judged by the election results, and a long series
of successful predictions gives us confidence in the methods of the pollsters.
This check cannot be made on polls which tabulate public opinion on issues of
the day, for the whole population is never counted. Similarly for polls which
rate television shows, for the whole audience is not counted. Such polls, of
course, also generalize on the basis of samples. To illustrate the logical
problems in assessing the reliability of a statistical study of the
"public opinion poll" type we shall comment on Sexual Behavior in the
Human Female, by Alfred C. Kinsey and his staff.
Kinsey's study, published in 1950, tabulates and classifies data concerning
5,940 white American females, ages two to ninety. He did not claim that his
averages necessarily apply to all human females, despite the title of his book,
nor even to all American women, of whom there were approximately seventy
million in 1950. It is inevitable, however, that such inferences will be drawn,
and our question is: Are such inferences justified? This depends entirely on
the representativeness of Kinsey's sample.
Critics of Kinsey's report have emphasized the unrepresentiveness of his
sample. His subjects were not distributed proportionately in geographic areas;
most were from Illinois, Florida, and California. They were more highly
educated than a representative cross-section of the population; 75 percent of
his subjects went to college, as compared with a national average of 13
percent. Three percent of his women did not go beyond grade school as compared
with the national average of 37 percent. A larger than average proportion were
from middle and upper economic groups. Very few of the women were Roman
Catholics or orthodox Jews.
Critics have also argued that the very nature of the study involves a kind of
bias, for many women will refuse to discuss matters of such "delicate
privacy" with interviewers, so that his volunteers must be
unrepresentative of women in general. And there is also the problem of
credibility. Critics have said that people who like to talk about such things
tend to understate or overstate, and even to embroider a little.
Kinsey, of course, recognized the limitations and incompleteness of his sample,
and, as noted, did not claim that it was representative of the whole
population. But it will be interpreted in this way, and if Kinsey wished to
avoid such interpretations, he should have called his study "Sexual
Behavior of 5,940 Women." Inferences would probably be drawn, however,
even if he had so titled his study.
The elements of distortion in Kinsey's sample detract from its reliability as a
basis for generalizing. On the other hand, as a review of the book in Life put
it, though the statistics are not perfect, they are, at any rate, "the
only statistics in town." His study is by no means worthless as an index
of sexual behavior. We must not use an "all or nothing" approach
here. The reliability of his sample with respect to university women as a
single group, for example, is certainly much higher than that for the female
population as a whole. But we cannot conclude that the whole female population
resembles the sample since the sample is not a representative one.
Generalizations in statistics, then, are judged by the same logical criteria we
use in judging any generalizations. Fallacies, however, are more common in
statistical than they are in uniform generalizations. For it is easier to check
on the reliability of a uniform generalization: one exception overthrows the
general rule or 'law." In statistics, however, since nothing is said about
any specific individual, an "exception" is a meaning-less term. An
exceptional individual does not disprove an "average." But there is,
as we have already noted, a method for checking the reliability of a
statistical generalization concerning a population, and that is to count the
whole voting population in an election. But even a test of this kind is not
conclusive, for many of the voters do not vote on election day, because of
laziness, overconfidence, or some other reason.
Errors of inference in statistics are frequently overlooked because of the
mathematical language in which statistics are presented. The spell which
numbers weave often prevents us from seeing errors in arguments--errors which
would be obvious were they not clothed in mathematical garb. And many dishonest
reasoners take advantage of this fact and present highly selected data for
purposes of propaganda rather than information. Misuses of the science of
statistics have resulted in such jibes as, "Figures don't lie, but liars
figure," and there are three kinds of lies: ordinary lies, damnable lies,
and statistics." But these cynical remarks should not be taken as
criticisms of statistics. The fault never lies with the figures, or with the
science, but with their careless use. It is simply not the case that "you
can prove anything with figures" (or statistics), just as it is never the
case that "you can prove anything by logic." To the uninitiated, it just
seems that you can.
FOR DISCUSSION AND WRITING
1. Below are some generalizations followed by the samples on which they are
based. In each, explain why the sample may not be typical of the class being
investigated.
a. Generalization: Water always makes me ill. Sample: During the last week I
drank a glass of water after drinking a highball, a glass of wine, a six-pack
of beer, a shot of vodka, and a half-pint of whiskey.
b. Generalization: Child geniuses usually turn out to be disappointing when
they grow up. Sample: A recent article in Reader's Digest described seven cases
in which children rated "genius" became alcoholics, irresponsible, or
mentally ill.
c. Generalization: Watching television causes lower grades. Sam-pie: Two
students in my history class have failed all of their tests, and blame their
long hours of television viewing, rather than studying, for their low
performance.
d. Generalization: Smoking marijuana is harmful. Sample: Captain Fleming of the
police department says that the use of marijuana was admitted by seven suspects
recently captured in a heroin ring.
e. Generalization: If you want to avoid tooth trouble, you should avoid candy.
Sample: My sister stayed away from candy for three years, and during that time
she had no cavities.
2. Read each of the following paragraphs and follow the directions that
accompany them.
a. In a survey covering twenty-three states, the research team of a Democratic
candidate for the presidency discovered the following statistics. Of the
registered Democrats in those states, 43 percent had salaries below $7,000; 37
percent had salaries between $7,000 and $10,000; and 20 percent had salaries
above $10,000. Imagine that you are the candidate's adviser. What advice would
you give him on his campaign strategy with respect to economic issues, in the
light of these statistics. What factors must be considered in your analysis'?
b. In a recent survey, English majors at a northeastern university were
studied. In the years 1950-1970, 42 percent took jobs as teachers, either in
junior high or senior high schools; 18 percent found jobs with publishers or in
publisher-related industries; 3 percent either found no job at all or were
engaged in occupations for which English skills had no particular value;' and
37 percent continued their education by entering graduate school. If you were
given the opportunity to redesign the English curriculum at that university,
what use would you make of those statistics? What conclusions could you draw
concerning the present curriculum in terms of its meeting the needs of English
majors?
c. A university committee on student affairs examined class attendance during
the academic year 1972-1973. They found that class attendance on Mondays was 80
percent of the enrollment; on Tuesdays, 84 percent; on Wednesdays, 91 percent;
on Thursdays, 83 percent; and on Fridays, 71 percent. What generalization could
the committee make about the tendencies of college students to cut classes?
3. Most-if not all-of us repeat and accept generalizations in all areas of our
lives. List several that you find yourself repeating without usually stopping
to examine them carefully. How valid are they? How often do you use them as the
basis of action?
CHAPTER 11
On Matters of Taste and Opinion
We have been engaged in analyzing matters of logic and science. Science, we all
agree, statements must be proved by evidence. In physics and chemistry,
certainly, scientific laws are not just "matters of opinion." A
successful. laboratory test is something quite definite and convincing. The law
court, too, requires proof. In a criminal trial the evidence might prove a man
guilty 'beyond a reasonable doubt" There may be differences of opinion
concerning some verdicts-miscarriages of justice occurring in acquittals as
well as in convictions-but the rule is that it is the evidence that counts. The
"reasonable man's judgment" is the ultimate criterion for questions
of fact in a law court, and reasonable men the world over would probably reach
the same verdict in a given dispute. In the social sciences, too, we try to get
beyond matters of mere opinion. Carefully tabulated statistics tell w what
percentage of paroled convicts will probably "go straight"
thereafter, within a given "margin of error."
Value judgments
We agree that logic is relevant in questions concerning facts, for these
questions involve evidence and proof. There is an important class of
statements, however, which we have not yet examined and which you will often be
called upon to write in essays, reviews, and reports in college. These are
"value judgments," which many people regard as exempt from the
requirement of proof or as incapable of proof. Examples: "I'. S. Eliot was
a great poet." "The wartime bombing of cities is morally wrong."
"The Beatles were the greatest pop music group of the 1960s." A value
judgment, as we shall use the term in this discussion, is an assertion that
something is either good or bad in an aesthetic or a moral sense. This
restriction means that we shall exclude the purely technical sense of good (or
bad) in this discussion, as when we speak of "a good automobile tire"
(one that will run a long distance) or "a good repair job." Value
judgments, then, are statements such as "X is beautiful" (possessing
aesthetic excellence) or "X is morally right." Value judgments
obviously refer also to statements that things are ugly or that actions are
morally wrong.
Value judgments are usually contrasted with "factual statements,"
which make assertions about events that can be observed in the world of space
and time. By a factual statement we do not necessarily mean a true statement. A
factual statement, in the sense in which we use this term, refers to one that
is about facts. Factual statements are true or false, for they may describe the
facts correctly or incorrectly. "The Gateway Arch in Saint Louis is higher
than the Leaning Tower of Pisa" is a factual type of statement. So also is
"The Leaning Tower is higher than the Gateway Arch." "The
Gateway Arch is more beautiful than the Leaning Tower," on the other hand,
is a value judgment.
"It is illegal to grow and possess marijuana in the state of
California" is a statement of fact, which happens to be true. We can
verify this statement by looking up the law. "One ought not to smoke
marijuana" is a value judgment which asserts that such conduct is morally
wrong. Now, there is general agreement that it is possible to specify the kind
of evidence which would prove factual statements true or false, but many people
think that value judgments are incapable of proof. Value judgments, it is said,
are "mere matters of opinion." It is important to note the precise
sense m which this ambiguous phrase is meant. An "opinion" sometimes
means a judgment that has a certain measure of probability, but not certainty,
as when a man expresses the opinion that higher taxes are necessary to stave
off inflation. When the evidence is conflicting, as in a case of this kind, we
may speak of 'legitimate differences of opinion." But the
"opinion" which asserts that higher taxes are necessary may be true,
and the opposite opinion may be false. When people say that value judgments are
mere matters of opinion, on the other hand, they usually mean "a matter of
personal feeling or preference" or a "matter of taste." When we
qualify the word "opinion" by mere in this discussion, this is the
view to which we refer. According to this view, value judgments are incapable
of proof and thus outside the realm of logical or scientific criticism.
If the "mere opinion" point of view is correct, then one value
judgment is "as good as another," and proof is not only impossible
but irrelevant If one value judgment is as incapable of justification as
another, then reason and intelligence are irrelevant in the discussion of such
matters. In this chapter we shall endeavor to show that logic is relevant in
the discussion of value judgments as well as in the realm of scientific or
"factual" statements. We shall first discuss two theories which hold
that value judgements are incapable of logical or scientific justification. One
of these we shall call the "taste" theory; the other the
"approval" theory. We shall then discuss the sense in which logic is
relevant in value judgments.
The "taste" theory
There is an ancient adage which tells us that "of matters of taste there
is no disputing." This is sensible advice. If you prefer red wine and I
prefer white, this establishes a basis for harmony such as prevailed in the
famous Spratt family, and it would seem fruitless to argue the question as to
which really tastes better. We may grant that it is impossible to prove that
the taste of black caviar is superior to that of red, even though most
"epicures" prefer the former. Some people may argue that it is a
fact, not a "matter of taste," that French cuisine is superior to
English cooking, but mankind has wisely decided that these are matters that
ought not to be disputed. But how far, and to what kinds of things, can this
principle be applied? "Matter of taste" is frequently used for things
other than gustatory flavors. It often covers individual preferences and likes
or dislikes in the arts as well, and it is sometimes held that ethical
judgments are matters of taste. Let us examine the "taste theory" as
applied in the fields of aesthetics and ethics.
No logical problem arises when one says that he prefers the sound piano to that
of a violin or when he says "I prefer Tchaikovsky to Brahms." He is
merely describing his personal taste. The interesting problem for logic arises
when he says, "Tchaikovsky is a better composer than Brahms," or
"Tchaikovsky's symphonies are more beautiful than those of -Brahms."
Most literary critics regard Kahlil Gibran, author of The Prophet, as a
second-rate writer, despite his phenomenal popularity all over the world? and
perhaps an equal number say that Rod McKuen
writes bad poetry. Are these judgments true, or false, or neither? The
"taste" theory holds that they are neither. When we assert value
judgments of this kind, the taste theory tells us, the judgments merely express
the preferences of the speaker, so that "second-rate,"
"bad," etc., are not words with any 6bjective reference. "X is
beautiful," this theory tells us, means nothing more than "I like
X," and "X is better than Y" means only "I prefer X to
Y." If we grant that the speaker is telling the truth, and not lying about
his actual preference, that is the end of the matter.
Is there no disputing "matters of taste" in the arts? In practice, of
course, we do dispute such matters. The word "taste" is often used in
a sense other than "personal preference"-for "keenness of
discernment, or insight." Immanuel Kant, in one of his rare flashes of
humor, once played on the double meaning of "taste" when he said
"Art is a matter of taste, but there is no point in arguing matters of
taste with the tasteless." Even when a person says that he prefers
Tchaikovsky to Brahms, a Brahmsian is apt to accuse him of having a perverse
taste. The Brahmsian believes that a person of "genuine" taste will
prefer Brahms to Tchaikovsky. When some people tell us what they prefer, it thus
appears, they regard their preferences as authoritative!
In practice, moreover, most persons will place limits on the taste theory. They
may believe that one's preference for Brahms or Tchaikovsky is a matter of
taste, but they will hesitate to say that the judgment Most male opera singers
at the Met sing more beautifully than some truck drivers" is simply a
matter of opinion, incapable of justification. Most of us do believe that there
are shared standards of merit which go beyond merely individual liking and
disliking. Semantically, then, it would appear that "I like X" is an
inadequate translation of "X is beautiful." We never ask others to
justify their feelings, as when they say, "I just happen not to like
Picasso's Guernica." But we do think it is reasonable to ask them why they
think a painting lacks aesthetic merit.
We have been discussing the taste theory in aesthetics. It may also be applied
to ethical judgments. "X is morally right" (or "ought to be
done") is held to be translatable into "I like X," and such
judgments are also held to be mere matters of opinion, hence unarguable. But
here again it seems that no one who understood the meaning of the words could
fail to agree with the judgment "A teacher ought not to flunk a student in
English 101 on the sole ground that he wears a beard." This sentence means
a good deal more than -"I dislike teachers who do such things."
Consciously or unconsciously we carry in our minds a definition of
"wrong," or a standard of justice, and when we say the action of the
teacher is wrong, we have classified it under our conception of wrongful or
unjust actions. Again, to say that this judgment is a "matter of
taste" and to attempt its translation into "I dislike X" is
inadequate.
The "approval" theory
The second theory which holds that value judgments. cannot be logically
supported or justified is the "approval theory." This theory tells us
that judgments concerning right and wrong can be translated into, "My
group approves or disapproves." This view is usually associated with the
doctrine of "ethical relativism."15 The relativist holds that nothing
is always right or wrong but that these terms are relative to time, place, and
circumstance. The relativist is impressed by the great variety of contrary customs
in different parts of the world. Monogamy is customary in western countries,
polygamy in Arab countries, and in some places polyandry is the custom. The
customary, the relativist notes, is considered morally right; the uncustomary
morally wr6ng. And so what one country considers morally right, another
considers wrong. And their customs of today may not be their? customs of
tomorrow.
Thus far the relativist merely cites commonplaces known since the beginning of
tourism. He now makes his distinctive contribution: What a group of people
consider right, he says, is right. Thus, monogamy is right for Americans, and
polygamy is right for Saudi Arabians. Right, of course, means right for them.
When we say that polygamy is wrong in the United States, the relativist says,
we mean only that Americans disapprove of this matrimonial system. Right and
wrong are relative to group customs and group approvals.
But we noted earlier that the mere fact that people believe something is not
sufficient to make it true, and similarly the fact that people practice certain
customs doesn't make them right. Let us sum up the relativist position in a
quotation from Pascal:
Three degrees of latitude reverse all jurisprudence; a meridian decides the
truth. Fundamental laws change after a few years of possession; right has its
epochs. . . . A strange justice that is bounded by a riven Truth on this side
of the Pyrenees, error on the other side.
This theory has an appealing plausibility when we consider the variable customs
throughout the world. It appears presumptuous for one nation to tell another
that its customs are "immoral." It would be too easy to return the
compliment. But the group-approval theory also seems quite inadequate as a
reflection of what we mean by right and wrong. If group approval makes an
action right, then it would be nonsensical, if not meaningless, to say: "I
think this action is wrong, but I am in a minority." For "wrong"
can have no meaning in this remark if the approval of the majority makes the
action right. But does anyone seriously believe that the majority can never be
mistaken in its judgment about right and wrong?
Let us now consider some of the implications of the two
"translations" of aesthetic and moral judgments that we have been
discussing. The taste and approval theories deny the possibility of genuine
disagreements over values. What appear to be disagreements, they hold, are
merely "verbal disputes." Consider a difference over moral values.
When one man says that selfishness is morally wrong and another preaches that
it is a virtue, if the first means "I dislike selfishness" and the
second "I like it," then they are not really disagreeing with each
other. They are merely uttering confessions about their feelings or making
disguised autobiographical statements. It is as if I say I like tennis and you
say you don't. Similarly, when one man says that polygamy is wrong, and another
says it is right, if what each means is that his group disapproves on the one
hand and approves on the other, then each is reporting a sociological fact
about his society, and one statement does not contradict the other.
The approval theory also makes moral discussion impossible within a given
group. If a moral problem arises, say over the denial of civil rights to a
minority group, all the approval theory can tell us is: Take a "Gallup
poll"; find out who has the votes. If 51 percent vote "right,"
then it is right. But "This has the vote of the majority" seems quite
different from saying, "This is morally right." The majority have the
power to decide what the civil laws shall be, but legality and morality are not
the same; they sometimes diverge from each other. Would the abolition of the
freedom of religious worship in the United States be right if 51 percent
approved? Is the morality of such a matter determined by making an accurate
count of the votes?
The "value standard" theory
We shall now discuss a new approach to the problem of values, the "value
standard" theory, which makes logical analysis relevant in matters of
values. We shall illustrate the theory by applying it to a problem in social
ethics: Ought we to legalize gambling? My first reaction may be, "Yes, I
don't see why not. Let people gamble if they wish to," or "No
gambling is wrong." But there is a counterpart of the "1aw of
rationality" m the field of values. The law of rationality tells us that
we ought to justify our beliefs by evidence and reasons, instead of asserting
them dogmatically. Similarly, if I am a reflective person, I will seek to
justify my value judgments. (A value judgment which gives no reasons can
scarcely be called a judgment.) I will think about my reasons for approving or
disapproving of gambling, instead of saying, "I like (or dislike)
gambling," or "The community approves (or disapproves) of gambling."
I ought to consider the logical consequences of my choice. My thinking may
proceed along the following lines: "It is impossible to suppress the human
desire to gamble. If this desire is denied a legal outlet, it will find an
illegal one. Illegal gambling funnels vast sums of money into the hands 6f
undesirable elements in the community and gives these elements great power.
They corrupt the police force and may even control the political machines in
our great cities." These considerations make me lean toward legalization.
Now I consider the other side: "If gambling is legalized, it may become
respectable to gamble, and many more people may take to this vice. It is also
possible that undesirable elements may manage to obtain control of legalized gambling."
Implicit in my thinking about this problem is the notion that the "public
good" ought to be served. If I finally decide that, considering all the
consequences, we ought to legalize gambling, I may justify my decision in some
such fashion as this: "We ought to promote the general welfare. This duty
requires us to choose the greater of two possible goods, or the lesser of two
evils. The consequences that will follow from the legalization of gambling
will, on the whole, diminish the general welfare less than the alternatives.
Therefore we ought to legalize gambling." This "is a logical
proof" of a value judgment. My major premise is a value standard which I
assume as a basis for value decisions. My minor premise is a factual assertion
to the effect that legalized gambling will diminish the general welfare least.
("General welfare," of course, is a term whose minimal meaning
connotes freedom and respect for law, and I must define it in any actual
discussion.) I then arrive at my conclusion.
Similarly, if I say that "government regulation of television is
wrong," I ought to define what I mean by "wrong" in terms of a
value standard. I should make my value standard explicit-whatever that value
standard may be. I should then show how my standard applies to the matter of
government regulation. If by "wrong" I mean "that which
diminishes the general welfare," I must show how regulation will have that
effect.
The value-standard analysis of value judgments makes logical discussion
possible. A discussion of the legalization of gambling is not only a legitimate
procedure but very useful when each disputant accepts the same value standard.
Public discussion of such questions helps to clarify the consequences that
follow from one or another course of action and thus leads to a thoughtful
decision. Most value disputes, actually, involve disagreements concerning the
appropriate means to mutually accepted ends. This is, of course, not
necessarily the case: Some may have standards other than that of the general
good, and not everyone who says that he seeks the general good actually does.
Some may pay lip service to this standard, but belie their acceptance by their
actual behavior. "Hypocrisy," as La Rochefoucauld said, "is the
homage which vice pays to virtue."
A similar type of value standard analysis may be applied to aesthetic
judgments. When I say that a novel is "good"-not merely that I like
it-I must justify my judgment. I should define what I mean by "good,"
and I thus presuppose an aesthetic or critical standard of excellence in
fiction. Perhaps my standard is that a good novel should have characters who
are "real people," a significant human conflict situation properly
developed and resolved, and an interesting story. One may quarrel with my
standard, or with its application to a particular work, but such questions are
not entirely outside the realm of rational discussion, as the taste theory
presupposes. The thoughtful application of a standard is quite different from a
snap-judgment made on the basis of surface liking or disliking, such as
"This is a good novel; I always like to read stories about artists."
We have presented some examples of the way in which the value standard theory
is used to justify value judgments. This approach is quite different from that
of the taste theory. We shall now contrast the value-standard type of analysis
with the type of analysis based on the theory of moral relativism.
The "untouchable" taboo in India is probably approved by the majority
of the Indian people. According to the approval theory, if the people of a
foreign country approve of a custom, then that custom must be right for them.
It is thus not only improper for us to criticize the Indians for their caste
system, but it must also be wrong for Indians to criticize, for the mere fact
of approval makes the system morally right in India. The present government in
India, however, seeks to abolish this system. This government obviously does
not accept the approval theory. It believes that the system is wrong in India
because it believes that the outcast system diminishes the general welfare of
the Indians, or because it violates the principle that every human being should
be treated as an end-in-himself and never as a mere means to an end.
To many, it would seem outrageous to condemn the Indian government for seeking
to abolish untouchability. If we believe that the Indian government is right,
then we do not accept the approval theory. And if Indian critics of
untouchability are at least logically justified in their criticism of the
system, then it is logically permissible for Americans to join in these
criticisms, at least on the same grounds as those used by the Indian
government. It is, of course, an extremely complex and difficult problem to
decide whether a given custom is right or wrong. In judging untouchability we
must consider the history of India, its traditions, religion, and customs of
life, and we should also consider the adequacy of the standard we employ in
making our value judgment. But these are matters open to logical discussion.
Ethical relativism has the insight that different customs may be equally right,
depending on the history, traditions, and circumstances of life in different
countries, and that we ought not to judge the customs of other countries by a
blind application of the customs that happen to prevail in our own. Time,
place, and circumstance do alter cases. But the relativist theory fails to
distinguish customs and local rules, on the one hand, from standards and
principles on the other. The principle that every human being ought to be
treated as an end-in-himself is not on the same level as that of a marriage
custom. And the relativists also deny that a custom may be wrong, as when a
practice violates a standard that may be used to judge customs. In the United
States, for example, the principle of equality has been violated in the custom
of racial segregation.
Approval theorists are, of course, seldom consistent in applying their theory.
This inconsistency indicates that most approval theorists do not themselves
really believe that it is approval alone that makes an action right. They, too,
assume that there are reasons for value judgments. They may say that polygamy
must be right in a particular community "because the people approve of
this custom." But if we ask them: "Why do the people approve?"
we may get an answer something like this: "Large numbers of men were
killed off in constant wars, or in dangerous occupations like seal-hunting, and
this resulted in a surplus of women." A reason of this kind, however,
indicates that this community judged polygamy right because it was thought
desirable for the general good, and this is the use of a standard.
Anthropologists, indeed, tell us that most customs are based on the necessity
(or the presumed necessity) for particular kinds of behavior. If a group is to
cope successfully with its environment, it must, of course, adapt its customs
to that environment. Obviously approval has its basis in reasons, and the
reasons are conceived as those which make the action right.
We have been seeking to justify the value standard theory in ethical and
aesthetic value judgments. Reflection, we believe, will reveal that most of us
do apply standards even though we may not be conscious of doing so.
But we must now deal with a question which may have been in the mind of the
reader throughout this discussion: Can value standards themselves be justified
by logical reasoning? This is indeed a difficult question, and its difficulty
accounts for the continuing popularity of the taste and approval theories.
We have been justifying value judgments by the use of value standards without
attempting to prove these standards. Value standards, of course, are neither
true nor false in a literal sense, for they tell us what ought to be" and
are not mere descriptions of facts. We have simply been assuming these
standards in our proofs. At this point a more sophisticated 'taste"
theorist may say, 'That is my real point. I grant that we can use logic in
proving that the means are efficacious or not efficacious in achieving a given
end, as in the discussion of legalized gambling, or in a marriage custom in
relation to human happiness, but the end, as formulated in the standard, must
simply be assumed. Whether we accept it or not is a matter of taste,
unprovable."
This sophisticated version of the taste theory has a wide following among
philosophers. But it is a long step beyond the simple form of the theory, for
it synthesizes tastes and the use of standards. We may make the following
comments on this version of the theory which asserts that the acceptance of
standards is in the end an arbitrary matter:
1. Many, if not most, differences over values are differences with respect to
the ways in which a certain end may be achieved, not differences with respect
to the standards, and so discussion may turn disagreement into agreement. This
is also true when we disagree over the application of a commonly accepted
standard.
2. Some standards are more basic than others. We may say that abortion is
wrong, because it involves the taking of a life. This indicates that we have
taken the right to live as our standard, rather than the mother's right to
terminate her pregnancy. But if someone were' to challenge our acceptance of
this standard, we may wish to justify it by a more basic standard, such as
"the natural law," believing that abortion is a violation of that
law.
And further, we ought never to consider even our basic standards exempt from
discussion) for no standard can be known to be absolutely final. As we develop
in maturity, we may see beyond our present "ultimates."
3. The thoughtful person chooses his standards, not by mere liking," but
by a personal commitment after surveying the consequences of his commitment
with respect to everything he desires out of life. It is a question of how we
really want to live. If this is in the end an arbitrary choice, it is one based
on consideration of all relevant factors. It is not arbitrary in the way in
which likes and dislikes may be.
Before we close this discussion, we shall briefly examine a different kind of
philosophical theory of ethics which seeks to prove that some standards ought
to be accepted and others not. It seeks to use reason and rational thinking to
justify standards.
The philosopher Immanuel Kant believed that there was a supreme moral principle
which could be established by reason itself. Men are rational beings, said
Kant, and the essence of rationality is consistency and the avoidance of
self-contradiction. A rational principle, furthermore, must be capable of
universalization, that is, it must apply to all persons in all situations.
His basic principle is that "we should act only according to a maxim which
can be universalized." This basic principle resembles the Golden Rule,
which, as we know, tells us to "Do unto others as we would have them do
unto us." Kant shows how the violation of this principle involves
inconsistency.
Consider the person who makes a lying promise, that is, one who makes a promise
that he does not intend to keep. The "maxim" of his act is that
"one may tell a lie when it is convenient to do so." Can this maxim
be universalized? Can the liar say: Let everyone do what I am doing? No, says
Kant. The liar "gets away" with his lie because others believe him.
But if everyone lied when it was convenient to do so, then mutual trust would
disappear and no one would believe anyone, so that a liar could derive no
benefit from his lie.
Do not lie or steal, says Kant, not because "Honesty is the best
policy" (though it probably is), or because "Crime doesn't pay"
(though it usually doesn't), but because such conduct is morally wrong.
Morality demands that we universalize our conduct, that we give others the
privileges we claim for ourselves (insofar as they are similar to us in relevant
characteristics and in similar situations), and the liar and thief cannot do
this without self-contradiction. For lying and stealing, if universalized,
would frustrate the goals of the liar and thief. The one would not be believed,
and the other would find that the rights of property had disappeared, so that
he could not enjoy his ill-gotten gains in peace.
Slavery (or any form of coercion except for crime) is wrong. A slave is treated
as if he were a purposeless thing, and his dignity as a fellow human being is
disregarded. Can anyone be willing for everyone to be treated in like fashion,
including himself? The contradiction is manifest, for no one can wish that his
own wishes be frustrated. Slavery is wrong, says Kant, even though a minority
of slaves may contribute to "the greatest happiness of the greatest
number," for nothing that violates the basic principle of morality or
justice can be right. And the rightness of these principles, he argues, is a
necessary consequence of rational thinking. If the universalization of one's
act leads to the defeat of one's own purposes, Kant holds, the act involves
inconsistencies and self-contradiction, and the maxim of such an action cannot
be an acceptable principle to a rational man.
In this chapter we have tried to show the relevance of logic in value
judgments. The law of rationality is relevant here. We ought to justify our
value judgments by reasons. We have shown how value standards may be used in
making such judgments, and we have discussed the possibility of justifying the
value standards by logic. Whether such justification can finally be achieved or
not, the acceptance of such standards is far from being an arbitrary matter. We
shall now sum up the different ways in which logic helps us to clarify our standards
in the field of ethics.
1. Logic clarifies what we mean by right and wrong, in terms of our moral
standards.
2. Logic helps us to determine whether a particular judgment makes sense in
terms of these standards. We ought to work out the consequences that follow
from our choices. We may find that we are defeating our own ends. Many people
have supported the mandatory death penalty for kidnapping on the ground that it
will effectively discourage the commission of this crime. But one should also consider
some of the undesirable consequences that result from a law of this kind. The
death penalty-for kidnapping as such-eliminates any inducement to the kidnapper
to bring his victim back alive, and juries will also be reluctant to render a
verdict of guilty in cases not involving murder, for a guilty verdict will mean
the death penalty. If we want victims brought back alive, and if we want juries
to do their duty, we may wish to revise our approval of the mandatory death
penalty for kidnapping.
3. Logical analysis may also show us that we are striving for the wrong ends,
as many prohibitionists in the 1920s discovered when they saw that the Volstead
Act did not bring about the elimination of the problems associated with the
consumption of alcoholic beverages. In fact, they discovered that prohibition
created new problems and magnified others.
4. Logical analysis may also reveal self-consistencies in our own thin king.
Few of us, perhaps, are as inconsistent as was the Mexican Society for the
Prevention of Cruelty to Animals when it raised a consider-able sum of money in
a benefit performance at the bull ring. But we are sometimes as inconsistent as
was King David-though in other contexts-when he condemned the rich man (in
Nathan's parable) for taking away the poor man's little ewe lamb. David himself
had sent Uriah to his death in order to possess his beautiful wife, Bathsheba.
We see, then, that logic is relevant in all human problems-in problems of
values as well as those of a scientific nature. To dismiss judgments of value
as mere matters of opinion, and not subject to discussion, is to invite an
irrationalist attitude, an attitude which dismisses the criteria of logical
analysis and the possibility of testing of our opinions by experience.
Reason is opposed to dogmatism, fanaticism, and obscurantism. The rational way
of life offers no panaceas; the rational individual recognizes the complexities
of human problems and the difficulties in proving one side or the other in
controversial issues. But most human problems are soluble, and if we try hard
enough, using the best methods that the human race has so far devised for
thinking about these problems, we are justified in the faith that we shall
solve them, one by one. The best methods, as we have seen, require working out
the probable consequences of our ideas and testing them in experience.
FOR DISCUSSION AND WRITING
1. Classify each of the following statements as a "value judgment" or
as a "factual statement.
a. Cavelier pipe tobacco has a smooth taste.
b. The New York Times has more subscribers than any other newspaper.
c. Heavy petting before marriage is sinful.
d. Maxwell House coffee is made from a blend of ten varieties of coffee beans.
e. The new El Dorado is the finest Cadillac made.
f. White's toothpaste is better than Score toothpaste.
2. As this chapter makes clear, value judgments should be recognized for what
they are--neither true nor false in a literal sense. With this in mind, write a
500-word essay in which you set forth a value judgment and support it with
factual statements.
3. Read the following statement carefully. Then, in a 300-word essay, analyze
its conclusion in the light of the following questions: What is the major
premise? What is the value standard underlying the premise? What terms must be
defined?
Capital offenders, it has been observed, are often the victims of their
environment, and to execute them is a cruel and unusual punishment Therefore,
the death penalty should be abolished.
Glossary
Ad hominem argument
An argument directed "to the man" rather than to the issue. Ad
hominem arguments frequently attack the character, religion, or nationality of
an opponent.
Ad ignorantiam argument
An argument based on ignorance, or an appeal to our lack of knowledge.
Ambiguity
The state of having two or more possible meanings.
Argument
Discourse that contains at least two asserted statements and the claim that one
statement ought to be believed because another is true.
Begging the question
An error in reasoning In which we pretend to prove something when actually we
assume In the "proof' that which we are supposed to prove.
Cause
The necessary and the sufficient conditions for the occurrence of an event.
Conclusion
In a syllogism, the proposition that must follow from the major and minor
premises; a deduction.
Deduction
The process of "drawing out" the logical consequences of what we
already know or assume to be true.
Deductive reasoning
The application of generalizations and premises to particular Instances.
Definition
A verbal equivalent.
Enthymeme
An incompletely stated syllogism, that is, one with a missing premise.
Generalization
A general law or principle which is inferred from particular facts.
Hasty generalization
A generalization based on insufficient or inadequate evidence.
Hypothesis
A tentative answer to a problem to test its truth.
Induction
The process of formulating a conclusion by going from specific instances or
examples already known.
Inductive reasoning
The formulation of generalizations or conclusions as a result of individual
examples or instances.
Law of contradiction
The principle of logic which says that if we have asserted two contradictory
statements, we must abandon one or abandon all claims of rationality.
Law of excluded middle
The principle of logic which tells us that a thing either has a particular
characteristic or does not have that particular characteristic.
Law of identity
The principle of logic which states that A is A, that is, that "anything
is liself."
Laws of thought
The principles of identity, excluded middle, and contradiction.
Logic
The science of proof, the process by which statements are supported with
adequate proof by being tested against the right amount and kind of evidence.
Major premise
The premise in a syllogism which contains the major term.
Middle term
A term in the syllogism which appeals in both premises but not In the
conclusion.
Minor premise
The premise in a syllogism which contains the minor term.
Post hoc fallacy
An error in reasoning in which an effect is incorrectly attributed to a cause.
solely on the basis of chronological order.
Proof
Sufficient evidence to justify a conclusion.
Semantics
The study of meaning; in logic, the study of the relationships between signs
and symbols and what they represent.
Stipulation
A specific or particular sense or way in which a word will be used.
Syllogism
An argument consisting of two premises and a conclusion, viz.,
- All Armenians are left-handed;
- Joe is Armenian.
- Joe is left-handed.
Tautology
A statement that is true because it provides for all logical possibilities, and
is valid as demonstrated by logical analysis.
Truth
A statement proven to be or accepted as true and in correspondence with fact,
actuality, or logic.
Valid argument
An argument in which the conclusion is necessitated by the premises.
Value judgment
An assertion that something is either good or bad in an aesthetic or a moral
sense.