11 Over-precision: "logic chopping"

Some persons, apparently impressed with the rule of reduction and other modern techniques, would reject the extreme-case criterion for permissible vagueness as far too generous an application of Rule 4 (p. 53). It is not enough for them that all would agree on x,: there must be agreement on all cases in the range. That nobody is likely to deny Peter's baldness does not qualify "bald" for precise usage unless unanimity is also possible on Paul and Henry, whatever the state of their scalps. Some writers blame the vagueness of terms in common usage for an alleged lack of progress in social science and for much of the confusion in politics and other activities of everyday life. They demand of every general term that it carry with its use the method of its reduction. As to marginal cases, they are to be eliminated by formal definition or by exact distinction.

Vagueness, they say, encourages reification and Word Magic (see #12). The existence of a certain usage in the language does not guarantee anything about any entities in the world. The name "Zeus" is a proper name, but there is no god. The difficulty with "the good" or "chivalry" is that the use of these terms often commits the speaker to a belief in Platonic entities, since the tendency in language to give some content to words leads speakers to reify when no content from reduction is possible. Moreover, a belief in a Platonic super-child does not matter for communication: the term can always on demand be reduced to this young person or that. But abstractions such as "chivalry" cannot be reduced since the concept is not well enough constructed to permit its users to state the conditions for applying it. They can, of course, provide an ad hoc definition, but then they might better have avoided the term altogether and with it the confusion with common usage and its likely reification.

For these reasons the extreme-case criterion is held to be too broad. It is not enough to eliminate the merely directional usages of terms such as "chivalry," but such terms as "bald" must also go. The fact that there are clear-cut cases misses the objection to vagueness in language: this is that there are any cases not clear-cut. Let speakers adopt the practice of defining each term as it occurs: this will eliminate marginal cases as they arise. There is no reason to put up with indecision about them and to tolerate vague uses. So goes the positivist argument.

Most of these strictures are directed against the vagueness in certain social sciences, psychology, sociology, history. Many of the key concepts in these sciences seem to the critics merely evaluative and literary. Can psychologists precisely state the characteristics of projection, of repression, of sanity itself? On the contrary, they are unable even to specify the level of abstraction for these constructions, much less show how to reduce them to particulars. Can any anthropologist state with clarity the conditions that will indicate for every human trait whether or not that trait is "culturally determined?" These criticisms imply that it is always desirable to give precise rules for the usage of terms, and that it is always possible to do so.

Sigmund Freud and Margaret Mead could, indeed, define their terms and be done with it. The definitions would perhaps be arbitrary, but they could be clear. But what is the problem such writers face? Is it merely to define, say, "repressions," that is, to provide an exact list of characteristics by which psychologists can judge whether or not any given human activity can be arbitrarily designated "repressive"? Or is it to explore a situation believed to involve a basic psychological mechanism? Reflection will show that the task of the scientist here is not to set up artificial systems, but to found tentative constructions on the description of some clear (extreme) cases. In the bewildering complications of fact, these early constructions may be found ultimately to be as basic as was originally thought -- in which case they will by degrees become entirely explicit and uniquely determined in scope -- or they will be found unproductive -- in which case they will be discarded or replaced. In the meanwhile, the investigator is usually aware of the vague and tentative character of his key concepts.

It is a form of perfectionism (see #38) to reject the best available on the grounds that it is not the best conceivable. What is happening seems to be a rejection of serviceable usages for the sole reason that there are cases that are marginal or doubtful. We have stated that the function of classification is to separate, and the criticism of the marginal case seems well taken. How can the rules of exhaustion and exclusion operate if the construction is littered with doubtful cases? But we have also stated that where these are crucial they must be decided. We now add that this decision can be recognized as tentative where a formal division would not further the understanding of the situation, but on the contrary would be arbitrary and artificial. In a discussion where the marginal cases are not at issue, it is captious to reject a construction simply because such cases exist. This amounts to refusing to allow the term "bald" when Old Marbletop is being pointed to, simply because there is a gent, Old Brushacross, who is nowhere around.

In such instances we say that the criticism is fallacious. There is nothing wrong with the extreme-case uses of construction, when the alternative (" 'bald' means 'fewer than 5000 hairs on the head'") is purely arbitrary. If there are some clear cases, as with "baldness," "repression," "culturally determined traits," the historian's "chivalry," then the uses are permissible unless greater precision is possible without artificiality. This is admittedly itself a vague criterion of permissible vagueness, but it must serve the present purpose -- which is only to characterize the fallacy of requiring an unnecessary precision. The fallacy is easily recognized in its extreme forms: wholesale rejection of vast areas of human inquiry as "nonsense," pedantic hair-splitting and logic-chopping over constructions that are the best available and good enough for the purpose, the ironical bark in discussion, "I can't understand that!" "What does that mean!"

EXAMPLE COMMENT
Professor Peter writes in the Historians Quarterly, "It is bad enough to see this term romantic in undiminished circulation amongst music critics and literary historians. What does it mean to characterize Beethoven's Opus 28 as romantic? Or Byron's Don Juan? The sonata contains the notes, chords, movements it contains. How can these add up to some non-musical entity to be designated as 'romantic and thus automatically grouped with Don Juan, as well as with Gothic revival, the agony of a perverse love, trap doors and ghosts? But when an historian, who should know better, seriously sets out to specify clear-cut, extreme cases of the romantic impulse in diplomacy and political agitation, then we have seen everything. I can only say that Professor Paul's latest tome may be program notes or literary obiter dicta, but it is hardly history." It is true that terms such as "romantic," "classical," even Professor Peter's "history," are frequently used in literary and popular contexts in a merely directional way. No precise content can be given to them, and it is not even clear if there is an extreme case. Yet Professor Peter's attack itself refers to fairly definite areas of content (trap doors, ghosts) -- almost all literary critics agree in regarding the appearance of Gothic props in novels and poems as marks of what they call the romantic movement. Moreover, Professor Paul apparently is at pains to find "clear-cut, extreme" cases of practices in diplomacy and politics which he is proposing to call romantic, since they lie in the general direction of what is ordinarily called "the romantic movement." At any rate, he must use some term to cover the phenomena he is describing, and all that can reasonably be required is that he provide some clear cases, even if he must leave it open whether or not other cases are to be so characterized.
Overheard in a bull session: "You say that whatever morality involves, it's bound to include people helping each other in time of need. You cite rescues at sea and loans to friends, even considerate advice to troubled spirits. This is the way all you moralists talk. You list noble deeds that, of course, everybody approves of, then treat them as if they were a class of similar things, give them a name, like "helpful," and think you have proved something, I don't know what. All you have done is cite cases, which may or may not have something in common, and wrapped them up in a vague meaningless name. I still don't know what it means to be helpful, and I don't think you do. Look! What if the man at sea were dying of cancer and didn't want to be saved, or if your alcoholic or whatever didn't want your advice -- maybe he needs his vice, not advice! How would it be helpful to interfere then? Answer me that!" The providing of particular cases was apparently designed to free the term "helpful" of some of its every-day ambiguity. The critical attack does attempt to test the extreme-case instances by imagining counter-examples, but even though such may occur, surely there are many rescues that are desired just as eagerly as the rare one is not. The intention obviously refers to such cases. Is the designation of such cases as "helpful" unjustified? It will be noted that the critic seems himself to have derived from them a clear enough notion of "helpful" to pose counter-cases; he asks, "How would it be helpful to interfere?" If he would accept the term as now specified in the light of rescues, advice, loans, the critic might then find out what his interlocutor thinks he has proved -- presumably that being helpful is a moral good.