Page 137 - Encyclopedia of Philosophy of Language

However, Sects. 84-7 contain a very carefully struc- tured discussion. Each paragraph with a remark tend- ing in the above direction contains a counterbalancing remark to block any skeptical interpretation of the point. For example, in Sect. 84, 'Can't we imagine a rule determining the application of a rule, and a doubt which it removes—and so on?' is balanced by 'But that is not to say that we are in doubt because it is possible for us to imagine a doubt.' People are gen- erally capable of perceiving what rules, or at least familiar ones, require at this or that point. That people are—and there is no reasonable doubt that that is what they are doing—is sufficient guarantee that there are facts as to what rules require; all the guarantee one should expect. Very roughly, those facts are what we perceive them to be, except where the rational course, for a specific, special reason, is to take us to be misperceiving. If facts about what rules require need to be constituted by anything, then it is by our natural reactions, plus the surroundings in which we have them—those in which the rule might require this or that.
2.2 Fitting Concepts to Objects
It is remarkable that all these points are made within and in aid of the family resemblance discussion. With that in mind, return to the problem. Pick a concept and an object. For the object to fit the concept is for it to satisfy the rule: count an object as fitting that concept just in case such-and-such, where that 'such- and-such' would obtain just where an object did fit the concept. But now, independent of the question of what such a rule might be—how the dummy 'such- and-such' might be filled in—one may question what itwouldbeforittobeafactthatthusandsowasin conformity, or not, with such a rule. This question now directs attention, in seeking the core point of family resemblance, away from the form that some definition might take to a more fundamental level concerning the application of concepts in conformity with rules.
2.3 Explanations of Meaning
Before attempting an answer, some more features of the text should be noted. Another point that Wittg- enstein emphasizes here is that what we know about what words mean, or which concept any given one is, and (so?) what words do mean, and which properties concepts have, is actually stated, with no inexpressible residue, in those explanations of meaning that we can and do give (see Sects. 69, 71, 75). Wittgenstein concentrates on cases where we explain meanings in terms of examples. But the point is: such explanations are as good as explanations by general formulae (e.g., those providing necessary and sufficient conditions). Either style of explanation is equally, and just as fully and explicitly, an actual explanation of what the rel- evant words mean. Whatever shortcomings beset
explanation by example also beset explanation by for- mula, and vice versa. Either sort of explanation might be adequate; but there is no point in saying this if one has a proof that explanations of the latter sort could not be correct, as per the initial imagery-driven read- ing (in Sect. 1 above).
2.4 Family Resemblanceand Proper Names
In Sect. 79, Wittgenstein makes a crucial application of the idea of family resemblance to the case of proper names, or individual concepts, primarily to the Bib- lical name 'Moses.' On one interpretation, suggested by Kripke (1980), and perhaps due to John Searle, the application consists in a 'loose-backing-of- descriptions' theory of names. On such a theory, a name, in a particular use, is to be understood in terms of a specific, though perhaps loosely bounded, set of descriptions, or general properties. Their function is this: the name, so used, refers to that unique indi- vidual, if there is exactly one, of whom most of those descriptions are true. Wittgenstein does consider such a theory in the first sentence of the second paragraph of Sect. 79, but immediately rejects it. It occurs as one of a series of trial balloons he punctures en route to his goal.
The remarkable features of Sect. 79 lie elsewhere. The simplest is that in the course of that section he applies the idea of family resemblance to concepts of all types ('The fluctuation of scientific definitions'), including cases where he obviously recognizes that one can give definitions in terms of general formulae. So the 'impossibility' of doing that cannot be the message of family resemblance.
Second, the initial imagery limps badly and obvi- ously when applied to proper names. General concepts, simply conceived, as per the initial picture, have extensions which, as a rule, contain multitudes, at least in principle. Overlapping similarities between cases may then be overlapping similarities between items in the extension. Here, we at least know where we stand. However, 'Moses' applies to, if anything, only one thing. If there are overlapping similarities, it is not clear between which items they hold.
'Moses' refocuses attention away from objects that are to fit a concept or not, towards the surroundings in which they are to do so. There are different cir- cumstances in which one may be confronted with a person who either is or is not Moses (the bearer of that name). In such different circumstances, different things will count as required for being that referent— so says Sect. 79. (This, incidentally, provides an alter- native treatment of all the phenomena that led Kripke and others to apply a notion of 'direct reference' to names.) Note that this is not the 'loose-backing-of- descriptions' view. There is no fixed stock of descrip- tions, the majority of which must fit a referent. Requirements on a referent merely vary. It would be equally out of step with Sect. 79 to see a general con-
Family Resemblance
115