The Definition Fallacy
Or, Humpty Dumpty meets Rush Limbaugh
(Note: No, this entry does not argue for or against Rush Limbaugh's actual political views. It only uses one of his popular phrases as an illustration of a linguistic principle. Sorry if the subtitle got anyone's hopes up….)
* * *
'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.'
`The question is,' said Alice, `whether you can make words mean so many different things.'
`The question is,' said Humpty Dumpty, `which is to be master - - that's all.'
-- Lewis Carroll, Through the Looking Glass
* * *
Fallacies in Literary Studies
In the 1940's, literary scholars W.K. Wimsatt and Monroe Beardsley wrote two essays identifying what they perceived as logical errors commonly made in literary studies at that time. The essays were titled "The Intentional Fallacy" and "The Affective Fallacy." The "intentional fallacy," as they called it, was the assumption that the meaning of a literary work was necessarily what its author intended. The mirror image of it was the "affective fallacy," in which the critic assumes that the meaning of the work is defined by its effect on the reader.
In recent years, as reader-oriented studies have become predominant in much of academic literary criticism, Wimsatt and Beardsley's paper on the "intentional fallacy" is often pointed to as the turning point in the abandonment of author-centered studies. The charge of committing it is often leveled as an accusation against anyone who looks to the author's statements for help in interpreting a book. But the lack of similar attention paid to the "affective fallacy" may perhaps fairly be taken as a hint that present-day theorists aren't really in Wimsatt and Beardsley's camp on these matters at all. Wimsatt and Beardsley's interest was in treating a literary work as a "verbal icon," being guided strictly by the words on the page, without regard for the personal interests either of the author or of the reader. Their objection to the twin "fallacies" was because they drew attention away from the verbal object itself as a work of literary art.
As for me: My own interest, in the literary discussions in which I've participated, has been largely in the promotion of rational discussion. The goal of such a discussion would be to bring out our ideas about the world of the HP books, to see if we can agree on them, and if not, to at least understand what we're disagreeing about (and, if possible, why). To throw out a string of random ideas and conclude, "well, I guess we've all got our own opinions," does not interest or attract me at all. And I have found that there are two fallacies, two logical errors, that often inhibit the discussions that might otherwise have become rational. One of these I call "the Definition Fallacy"; the other is the "Valuation Fallacy." In the Definition Fallacy we incorrectly treat that a word as if it meant the same thing to everyone else that it does to us; in the Valuation Fallacy we assume that everyone else should hold the same artistic value judgments that we do.
Trouble with Definitions
In last Friday's entry I discussed a particularly blatant example of the Definition Fallacy, in which certain parties were reportedly insisting that when Trelawney referred to Harry's "mean stature" she must have meant it in the statistical sense of "average." A more subtle example of it is when people assume that "ambition," as applied to Slytherin House by the Sorting Hat, is not necessarily a bad thing (in the moral structure of the books); even if that interpretation is possible, it can't just be assumed-- the opposite definition (that "ambition" is a bad thing after all) is certainly possible too.
It was about a year ago that I had a bit of a clash with certain others over the meaning of "interpretation." My chief opponent was insisting that an interpretation could be valid even if the author explicitly contradicted it. I replied to the effect that she'd have to explain what she meant by "interpretation"; most of us took it to mean "a good-faith attempt to understand the author's intended meaning," and that if she used it otherwise without explaining herself she'd only cause confusion. She reacted angrily, questioning my honesty (thinking that I should have known what she meant; I suspect that she was overestimating the extent of my reading-- I don't know, maybe I should have been flattered). But she never did give her definition; and to this day I still don't know how she would have defined it. Indeed, the reading I've done since then has led me to believe that there does not exist a consensus definition of "interpretation" in literary circles.
At another point in my HP fandom career, I was accused of making the same mistake as Humpty Dumpty. (That was in response to my ill-fated "Through JKR's Glasses" thread, in which I had argued for interpreting literature from the author's point of view.) Let's consider Humpty's view and see what we can make of it.
In the text quoted above, Humpty says, "There's glory for you," in a context where Alice has no clue what on earth he means by "glory." Humpty then insists that the only way she can know his meaning is if he tells her, because when he uses a word, it means just what he chooses it to mean-- "neither more nor less."
We know something's faulty with Humpty's use of language-- but what?
Or consider Rush Limbaugh (don't worry, we'll get back to Humpty). "Words mean things," says Rush. When he says that, he's in effect accusing his opponents of playing Humpty Dumpty-type games with words, inventing and insisting on their own meanings. (I don't listen to his show, and don't know how often he's correct in so thinking; and please, people, let's not sidetrack this with a debate for or against Rush!) But whether or not it's valid in a particular political context, it certainly doesn't work as an overall linguistic principle, because words indisputably do mean different things to different people at different times.
A third possibility might be to say that, if the speaker doesn't get to define his own words (contrary to what Humpty suggests), and if words don't have pure, fixed meanings (contrary to what Rush's principle could be taken to imply), then perhaps it's the hearer who gets to make up the definitions. But this doesn't quite work either; if Humpty had said "there's a nice knock-down argument for you," and Alice had insisted on applying her own private definition of "argument," the discussion would have gotten no further ahead.
What's going on in all these examples-- whether between Humpty and Alice, or between Rush and his opponents-- is a failure of communication. Because the parties to the discussion are using words to mean different things, no communication is taking place.
What is Language, Anyway?
Last spring I had an interesting conversation with a guy who was majoring in linguistics at the University of Washington. One of the questions I had for him was how they define a "language," as opposed to a "dialect." His answer was more or less what I expected: that they define "languages" by whether people can communicate with each other. If I speak with somebody from England, for example, there might be minor points on which would misunderstand each other ("sweater" vs. "jumper" and that sort of thing), but for the most part we'll still be able to communicate. We speak the same language. But if I were in the presence of somebody from Russia who hadn't studied English, then we wouldn't be able to communicate at all-- because our languages are different.
The point is: Language is a matter of communication. It's a two-way street (at least in its ordinary uses; there may be exceptions, like if someone is keeping a private diary). And if we want to know what a word "means" in a given context, perhaps a clearer way of asking the question might be: "What does it communicate?"
Let us suppose, for example, that Humpty Dumpty and Alice were old friends, and that the use of "there's glory for you!" to mean "there's a nice knock-down argument" was a long-running private joke between them. Would it then be fair to say that, in the context of a discussion between those two individuals, "glory" really did mean "a nice knock-down argument"? Yes, of course it would. It might not mean that to an eavesdropper overhearing the conversation; but between the two parties to the discussion, meaning would be communicated. And that's what language is for.
(Note: in order to avoid accusations of the "valuation fallacy" once I explain that in a later essay, I should acknowledge that it may be possible for some to view the purpose of language differently-- that language is for seeking meaning in life, or for bringing about the Revolution, or whatever. I take the language-as-communication position, however, in the confidence that most of my readers will find that to be the most sensible and satisfying view of the matter.)
We might even suppose that Humpty, writing in a private diary which no one else will read, uses the phrase "there's glory for you" in the unorthodox private sense of "knock-down argument." Then can we say that it has the "meaning" he associates with it? I think so, in that limited context. The only "communication" is between Humpty before he writes it and Humpty after, but he's still reminding himself of what his thoughts were, and we can still see the words as preserving meaning (that is, mental content) for him.
But if Humpty is talking to Alice, and Alice has no clue what Humpty means by "glory," then can we say that the word means anything between Humpty and Alice? No, we can't. It means one thing to Humpty and another to Alice, but it means nothing between them-- because no communication has taken place. This, then, was Humpty's "error" (as presented by Carroll): the use of private meanings in a context where communication was expected.
And How Does This Apply to Literature?
The matter sometimes becomes more complex in discussions of literature. This is because not everyone views communication (between author and reader) as a purpose of literature. Some (especially in discussions of poetry) have taken the view that meaning is irrelevant: that "a poem should not mean, but be"-- that it shouldn't be viewed as communicating meaning, only as being a work of art.
But (as I believe C.S. Lewis said, although I can't find the reference at the moment), whether the work as a whole has a "meaning" beyond the covers of the book or not, the words and phrases of it must certainly have a meaning within those book-covers. Otherwise we wouldn't use words. We'd just say "dooby-dooby-doo" or "shoo-ba-doo-wah" or some such nonsense. Whether or not we say that the HP series as a whole has any particular social or philosophical meaning (or whatever), it cannot sensibly be denied that it describes an imaginary world involving a boy with dark hair, glasses, and a scar on his forehead. The words of the books mean at least that much.
However, we may still ask questions about cases where literature fails to convey meaning even at these lower levels. For example, there are evidently some people who take "mean stature" to have a meaning different from the meaning JKR associated with it. Should we say that it "means" one thing for JKR and some readers, but another thing for other readers-- in which case there's no unique "meaning" after all? Or what about higher-level issues, as when one reader claims that JKR has failed in her attempt to present Ron Weasley to her as a likable character?
What we can say then is this: To some readers, meaning has been communicated. To others, it has not. In some cases, readers who have not experienced communication with the author on some points then go on to construct their own meaning (e.g. "average height"). Is this a meaning of the text, even for them?
I would find it more sensible to take that as a meaning suggested by the text for them. It is no longer the same story that the author told; at least on the specific point in question, it is a parallel story built up by our own imaginations, which just so happens to have the same words as the original. Whether that is worth discussing is a value judgement that each of us must make; I for one do not find it so. In any case, communication has not taken place; and in cases where a substantial number of readers do share the author's interpretation, it would seem that the failure to communicate may not be on the author's part. And it is certainly true that the readers who do perceive JKR's intended implications have communicated with her more fully than those who have not.
There are, of course, further questions that could be discussed. But (besides the fact that I need to go to bed now!) those deal with questions of artistic value judgment, and discussions of them are more likely to fall victim to the Valuation Fallacy; and so that discussion must wait for my future essay on that topic.
Summary of Definition Fallacies
So then, what is a "definition fallacy"?
I have already defined it as "incorrectly treating a word as if it meant the same to everyone else as it does to us." Now we can also illustrate it by some examples:
(1) Humpty Dumpty committed a definition fallacy by treating "glory" as if it meant "a nice knock-down argument," when (by Humpty's own admission) Alice could have no possible way of associating the word with that particular content.
(2) Rush Limbaugh in effect accuses his opponents of definition fallacies when he insists that "words mean things." But it may well be that both sides are guilty of the same error, each in failing to understand their opponents' terms.
(3) In my "Through JKR's Glasses" thread, my opponent and I were each guilty of definition fallacies in failing to recognize and honestly deal with each other's definition of "interpretation." (Near the end of that discussion I realized on what point we were miscommunicating, but by that time it was too late to do any good.)
(4) Six months later, in my next major clash (thankfully, my last thus far), I avoided the fallacy by realizing that my opponent was using the word "interpretation" in a sense different from mine (and from that of most Quillers). My opponent did not avoid the fallacy; she continued to insist that I was misusing the term but was not willing to explain her own definition.
This is not a fallacy to which Quillers are immune. In my observation, many disagreements are due to opposing parties failing to realize that they're using their terms differently. If we insist that "of course" a word means what we think it does, while our opponents think that "of course" it means something else, we will continue to miscommunicate. But if we notice the disconnect and ask, "okay, what do you mean when you say that?", they may or may not give us a straight answer-- but in either case, we've done our part to help make the discussion a rational one.
Unfortunately, it takes two to accomplish that.
(Note: No, this entry does not argue for or against Rush Limbaugh's actual political views. It only uses one of his popular phrases as an illustration of a linguistic principle. Sorry if the subtitle got anyone's hopes up….)
* * *
'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.'
`The question is,' said Alice, `whether you can make words mean so many different things.'
`The question is,' said Humpty Dumpty, `which is to be master - - that's all.'
-- Lewis Carroll, Through the Looking Glass
* * *
Fallacies in Literary Studies
In the 1940's, literary scholars W.K. Wimsatt and Monroe Beardsley wrote two essays identifying what they perceived as logical errors commonly made in literary studies at that time. The essays were titled "The Intentional Fallacy" and "The Affective Fallacy." The "intentional fallacy," as they called it, was the assumption that the meaning of a literary work was necessarily what its author intended. The mirror image of it was the "affective fallacy," in which the critic assumes that the meaning of the work is defined by its effect on the reader.
In recent years, as reader-oriented studies have become predominant in much of academic literary criticism, Wimsatt and Beardsley's paper on the "intentional fallacy" is often pointed to as the turning point in the abandonment of author-centered studies. The charge of committing it is often leveled as an accusation against anyone who looks to the author's statements for help in interpreting a book. But the lack of similar attention paid to the "affective fallacy" may perhaps fairly be taken as a hint that present-day theorists aren't really in Wimsatt and Beardsley's camp on these matters at all. Wimsatt and Beardsley's interest was in treating a literary work as a "verbal icon," being guided strictly by the words on the page, without regard for the personal interests either of the author or of the reader. Their objection to the twin "fallacies" was because they drew attention away from the verbal object itself as a work of literary art.
As for me: My own interest, in the literary discussions in which I've participated, has been largely in the promotion of rational discussion. The goal of such a discussion would be to bring out our ideas about the world of the HP books, to see if we can agree on them, and if not, to at least understand what we're disagreeing about (and, if possible, why). To throw out a string of random ideas and conclude, "well, I guess we've all got our own opinions," does not interest or attract me at all. And I have found that there are two fallacies, two logical errors, that often inhibit the discussions that might otherwise have become rational. One of these I call "the Definition Fallacy"; the other is the "Valuation Fallacy." In the Definition Fallacy we incorrectly treat that a word as if it meant the same thing to everyone else that it does to us; in the Valuation Fallacy we assume that everyone else should hold the same artistic value judgments that we do.
Trouble with Definitions
In last Friday's entry I discussed a particularly blatant example of the Definition Fallacy, in which certain parties were reportedly insisting that when Trelawney referred to Harry's "mean stature" she must have meant it in the statistical sense of "average." A more subtle example of it is when people assume that "ambition," as applied to Slytherin House by the Sorting Hat, is not necessarily a bad thing (in the moral structure of the books); even if that interpretation is possible, it can't just be assumed-- the opposite definition (that "ambition" is a bad thing after all) is certainly possible too.
It was about a year ago that I had a bit of a clash with certain others over the meaning of "interpretation." My chief opponent was insisting that an interpretation could be valid even if the author explicitly contradicted it. I replied to the effect that she'd have to explain what she meant by "interpretation"; most of us took it to mean "a good-faith attempt to understand the author's intended meaning," and that if she used it otherwise without explaining herself she'd only cause confusion. She reacted angrily, questioning my honesty (thinking that I should have known what she meant; I suspect that she was overestimating the extent of my reading-- I don't know, maybe I should have been flattered). But she never did give her definition; and to this day I still don't know how she would have defined it. Indeed, the reading I've done since then has led me to believe that there does not exist a consensus definition of "interpretation" in literary circles.
At another point in my HP fandom career, I was accused of making the same mistake as Humpty Dumpty. (That was in response to my ill-fated "Through JKR's Glasses" thread, in which I had argued for interpreting literature from the author's point of view.) Let's consider Humpty's view and see what we can make of it.
In the text quoted above, Humpty says, "There's glory for you," in a context where Alice has no clue what on earth he means by "glory." Humpty then insists that the only way she can know his meaning is if he tells her, because when he uses a word, it means just what he chooses it to mean-- "neither more nor less."
We know something's faulty with Humpty's use of language-- but what?
Or consider Rush Limbaugh (don't worry, we'll get back to Humpty). "Words mean things," says Rush. When he says that, he's in effect accusing his opponents of playing Humpty Dumpty-type games with words, inventing and insisting on their own meanings. (I don't listen to his show, and don't know how often he's correct in so thinking; and please, people, let's not sidetrack this with a debate for or against Rush!) But whether or not it's valid in a particular political context, it certainly doesn't work as an overall linguistic principle, because words indisputably do mean different things to different people at different times.
A third possibility might be to say that, if the speaker doesn't get to define his own words (contrary to what Humpty suggests), and if words don't have pure, fixed meanings (contrary to what Rush's principle could be taken to imply), then perhaps it's the hearer who gets to make up the definitions. But this doesn't quite work either; if Humpty had said "there's a nice knock-down argument for you," and Alice had insisted on applying her own private definition of "argument," the discussion would have gotten no further ahead.
What's going on in all these examples-- whether between Humpty and Alice, or between Rush and his opponents-- is a failure of communication. Because the parties to the discussion are using words to mean different things, no communication is taking place.
What is Language, Anyway?
Last spring I had an interesting conversation with a guy who was majoring in linguistics at the University of Washington. One of the questions I had for him was how they define a "language," as opposed to a "dialect." His answer was more or less what I expected: that they define "languages" by whether people can communicate with each other. If I speak with somebody from England, for example, there might be minor points on which would misunderstand each other ("sweater" vs. "jumper" and that sort of thing), but for the most part we'll still be able to communicate. We speak the same language. But if I were in the presence of somebody from Russia who hadn't studied English, then we wouldn't be able to communicate at all-- because our languages are different.
The point is: Language is a matter of communication. It's a two-way street (at least in its ordinary uses; there may be exceptions, like if someone is keeping a private diary). And if we want to know what a word "means" in a given context, perhaps a clearer way of asking the question might be: "What does it communicate?"
Let us suppose, for example, that Humpty Dumpty and Alice were old friends, and that the use of "there's glory for you!" to mean "there's a nice knock-down argument" was a long-running private joke between them. Would it then be fair to say that, in the context of a discussion between those two individuals, "glory" really did mean "a nice knock-down argument"? Yes, of course it would. It might not mean that to an eavesdropper overhearing the conversation; but between the two parties to the discussion, meaning would be communicated. And that's what language is for.
(Note: in order to avoid accusations of the "valuation fallacy" once I explain that in a later essay, I should acknowledge that it may be possible for some to view the purpose of language differently-- that language is for seeking meaning in life, or for bringing about the Revolution, or whatever. I take the language-as-communication position, however, in the confidence that most of my readers will find that to be the most sensible and satisfying view of the matter.)
We might even suppose that Humpty, writing in a private diary which no one else will read, uses the phrase "there's glory for you" in the unorthodox private sense of "knock-down argument." Then can we say that it has the "meaning" he associates with it? I think so, in that limited context. The only "communication" is between Humpty before he writes it and Humpty after, but he's still reminding himself of what his thoughts were, and we can still see the words as preserving meaning (that is, mental content) for him.
But if Humpty is talking to Alice, and Alice has no clue what Humpty means by "glory," then can we say that the word means anything between Humpty and Alice? No, we can't. It means one thing to Humpty and another to Alice, but it means nothing between them-- because no communication has taken place. This, then, was Humpty's "error" (as presented by Carroll): the use of private meanings in a context where communication was expected.
And How Does This Apply to Literature?
The matter sometimes becomes more complex in discussions of literature. This is because not everyone views communication (between author and reader) as a purpose of literature. Some (especially in discussions of poetry) have taken the view that meaning is irrelevant: that "a poem should not mean, but be"-- that it shouldn't be viewed as communicating meaning, only as being a work of art.
But (as I believe C.S. Lewis said, although I can't find the reference at the moment), whether the work as a whole has a "meaning" beyond the covers of the book or not, the words and phrases of it must certainly have a meaning within those book-covers. Otherwise we wouldn't use words. We'd just say "dooby-dooby-doo" or "shoo-ba-doo-wah" or some such nonsense. Whether or not we say that the HP series as a whole has any particular social or philosophical meaning (or whatever), it cannot sensibly be denied that it describes an imaginary world involving a boy with dark hair, glasses, and a scar on his forehead. The words of the books mean at least that much.
However, we may still ask questions about cases where literature fails to convey meaning even at these lower levels. For example, there are evidently some people who take "mean stature" to have a meaning different from the meaning JKR associated with it. Should we say that it "means" one thing for JKR and some readers, but another thing for other readers-- in which case there's no unique "meaning" after all? Or what about higher-level issues, as when one reader claims that JKR has failed in her attempt to present Ron Weasley to her as a likable character?
What we can say then is this: To some readers, meaning has been communicated. To others, it has not. In some cases, readers who have not experienced communication with the author on some points then go on to construct their own meaning (e.g. "average height"). Is this a meaning of the text, even for them?
I would find it more sensible to take that as a meaning suggested by the text for them. It is no longer the same story that the author told; at least on the specific point in question, it is a parallel story built up by our own imaginations, which just so happens to have the same words as the original. Whether that is worth discussing is a value judgement that each of us must make; I for one do not find it so. In any case, communication has not taken place; and in cases where a substantial number of readers do share the author's interpretation, it would seem that the failure to communicate may not be on the author's part. And it is certainly true that the readers who do perceive JKR's intended implications have communicated with her more fully than those who have not.
There are, of course, further questions that could be discussed. But (besides the fact that I need to go to bed now!) those deal with questions of artistic value judgment, and discussions of them are more likely to fall victim to the Valuation Fallacy; and so that discussion must wait for my future essay on that topic.
Summary of Definition Fallacies
So then, what is a "definition fallacy"?
I have already defined it as "incorrectly treating a word as if it meant the same to everyone else as it does to us." Now we can also illustrate it by some examples:
(1) Humpty Dumpty committed a definition fallacy by treating "glory" as if it meant "a nice knock-down argument," when (by Humpty's own admission) Alice could have no possible way of associating the word with that particular content.
(2) Rush Limbaugh in effect accuses his opponents of definition fallacies when he insists that "words mean things." But it may well be that both sides are guilty of the same error, each in failing to understand their opponents' terms.
(3) In my "Through JKR's Glasses" thread, my opponent and I were each guilty of definition fallacies in failing to recognize and honestly deal with each other's definition of "interpretation." (Near the end of that discussion I realized on what point we were miscommunicating, but by that time it was too late to do any good.)
(4) Six months later, in my next major clash (thankfully, my last thus far), I avoided the fallacy by realizing that my opponent was using the word "interpretation" in a sense different from mine (and from that of most Quillers). My opponent did not avoid the fallacy; she continued to insist that I was misusing the term but was not willing to explain her own definition.
This is not a fallacy to which Quillers are immune. In my observation, many disagreements are due to opposing parties failing to realize that they're using their terms differently. If we insist that "of course" a word means what we think it does, while our opponents think that "of course" it means something else, we will continue to miscommunicate. But if we notice the disconnect and ask, "okay, what do you mean when you say that?", they may or may not give us a straight answer-- but in either case, we've done our part to help make the discussion a rational one.
Unfortunately, it takes two to accomplish that.