Philosophical reasoning - the allocation to categories

["Jay Halcomb" <jhalcomb8@comcast.net>]

Hi. At the risk of being accused - rightly - of being philosophical, I want to pass along the following remarks, which I happened to run across and thought germane to SUO. - Cheers, Jay

 

PHILOSOPHICAL REASONING

 

“Our intuitive feeling  that we know the difference between the two [the formal and nonformal] is disturbed by contemplation of the series:

 

(I) Jones is a man
(2) Red is a colour
(3) Pleasure is a process
(4) Knowledge is a relation

 

Does only (4) allocate to a category? Or do both (3) and (4)? Or do all four allocate to a class? Or does (2) bring red under a `determinable' whereas (3) brings pleasure under an `empirical', as distinct from a `formal', category?

 

By now the conventionalist will have lost patience. There is one way, and one way only, of settling such a dispute, he will say: arbitrary decision. For some purposes it is convenient to draw distinctions, for other purposes it is better to ignore distinctions; for some purposes the distinctions are best made at a certain point, for other purposes at a different point: In the end, and not very far from the beginning, philosophical reasoning gets us nowhere: all it does is to confuse us, by making what is really a decision-problem look as if it were a theoretical problem, thus intolerably complicating what ought to be a simple choice.

 

Undoubtedly, conventionalism has a good deal to be said for it. Often enough, one must admit, decisions have been confusingly disguised as deductions; what distinctions we bother to make will certainly depend upon our interests and concerns; if we distinguish sharply between categories and classes, the false and the absurd, the valid and the true, we shall always, in some degree, be deciding to use words in a way in which they are not ordinarily used, and so far `establishing a new convention'. Conventionalism, too, promises us liberation from an endless metaphysical treadmill, in which we establish differences by using a criterion, and then find that to justify the use of that criterion we need to make another differentiation which we do not know how to make. An act of choice, it might seem -`here I draw the line'-is the only way of establishing a secure starting-point.

The difficulty, however, is that we have to distinguish not only cases but classes of cases, and this resuscitates our original problem. If we say-`I intend to describe as "logically false" all sentences like, and only sentences like, "my kangaroo is not a kangaroo" '-we shall have to determine, and this was really our original difficulty, what sentences are like `My kangaroo is not a kangaroo'. `My kangaroo is green' is like `my kangaroo is not a kangaroo' in containing `my kangaroo' and `is', but it would be ridiculous to describe it as being logically false.

 

Perhaps, however, we should work with lists, we should simply lay it down that certain signs are formal constants, and that logically false propositions contain one or the other of such-and-such combinations of constants. `No satisfactory criterion for distinguishing just what is logic from what is not', writes Nelson Goodman, `has been discovered. Rather, logic is specified by listing the signs and principles that are to be called logical: and the lists given by different logicians are not all the same' (`About', Mind, Jan. 196 1, p. 8.) Should I then just say that such-and-such is not a category, just say that such-and-such are not formal consequences, just say (in criticizing Ryle) that such-and-such propositions are not absurd, but false? And if challenged, point to my list of categories, my list of formal consequences, my list of absurd propositions? The last case brings out an obvious difficulty: listing is only possible in certain special instances, where we are convinced that no new candidates could possibly arise; clearly, we could list the absurd propositions only if we identified them with logically false propositions and these in turn with propositions containing certain logical. constants and certain patterns of variables.

 

Quite apart from that practical difficulty, however, a merely arbitrary classification would be quite pointless. It would be quite useless to draw up a list of classes and a list of categories if one merely decided arbitrarily what was to go in each list. In the case of logic, conventionalism has been used as an instrument of conservativism ; there is an established subject-formal logic-and its integrity has been preserved by defining it as being concerned with those signs and principles with which it has traditionally been concerned. But the effect of this is to hinder inquiry, not to advance it, if there are, or could be, other signs or principles which will generate the same type of theoretical construction; if not, the signs and principles ought to be definable in virtue of the fact that a certain type of theory can be constructed around them. It is a real question, not to be settled by decree, whether, for example, modal and deontic logics can be incorporated with traditional logic into a single generalized logic. Conventionalism, if used to short-circuit such investigations, gives us not greater freedom of inquiry, but merely freedom from inquiry-a licence to retreat from criticism.

 

Naturally, we begin from a list-explicit or implicit-of what we take to be classes and what we take to be categories. But if we find no way of distinguishing them, this is as much as to say that the distinction is of no theoretical consequence. We may then modify the lists in what we expect to be a more promising way; there is a certain interplay between list-membership and the progress of our investigation. Quite arbitrarily to stand by the original lists is to cut clean across the established principles of inquiry. If we say that `are' is, but `greater than' is not, a logical constant; that `disposition' is a category but that `colour' is not; that `my kangaroo is not a kangaroo' is logically false, but `my kangaroo is the fifth day of the week' is empirically false, we have to justify these distinctions by showing, for example, that questions about whether something is coloured are settled in a quite different way from, questions about whether `brittle' is a dispositional predicate. And if it is not apparent that the way of settling is different, that, too, has to be argued. Thus, in the present instance, Ryle cannot avoid the question whether we can, and how we are to, distinguish between a false statement and an absurd sentence; for my part, I cannot avoid the question whether, and how, I can distinguish between a formal and an empirical consequence. To treat as a matter of decision the distinction between false and absurd, between formal and empirical, rather than the original distinction between class and category, would be arbitrariness of the second degree, arbitrariness about the point at which to be arbitrary. But the responsibility which is thus incumbent upon me, to give grounds for distinguishing between formal and empirical is too heavy, and would carry me too far from my present theme, to be embarked upon now; it will suffice for my present purpose if I have shown that it is in fact my responsibility, that conventionalism does not provide us with an easy escape route from philosophical reasoning.

 

For the rest, if moral be needed, Fowler has already supplied it in Modern English Usage: `Category should be used by no one who is not prepared to state (1) that he does not mean class and (2) that he knows the difference between the two'. But what is it `not to mean class'? And how are we to persuade others, or even ourselves, that we know the difference between the two?”

 

John Passmore, Philosophical Reasoning, 1961, Charles Scribner's Sons, New York, Ch. 7, “Allocation to Categories”, P. 144-7.