I'll summarize our class discussion and add a few objections.

    Objection 1.  Long novel sentences.  (a):  They don't have conventional uses in the same way that "Hello," "You're on," "Excuse me" etc. do.

    Reply (Sellars, Brandom):  They have inferential roles.  The meaning of a sentence is determined, at least in part, by what other sentences the given sentence can properly be inferred from and by what further sentences we can infer from it.

    (b):  But for a long novel sentence, no convention has ever been directed upon utterances of it.  Some additional mechanism must be introduced to extrapolate the Use theory to such sentences.  The Wittgensteinian must grant that we understand novel sentences compositionally, in virtue of understanding the individual words that occur in them and working out the sentences' overall meanings from the way in which the individual words are strung together.  It follows that what is understood, that is, a sentence's meaning, is not simply a matter of there being conventional norms directed upon that sentence's deployment, for the sentence's meaning is in large part a function of its internal structure as well.

    Objection 2.  Twin Earth.  Use is exactly the same on Twin Earth as on Earth, but meanings differ.  Use is exactly the same on New Earth as on Earth, but truth-values even differ.  Thus, use does not determine meaning.

    First reply:  Remember, even the Putnamian cannot maintain that "water" means H2O.  Remember the pre-chemistry meaning of "water."  And Putnam's indexical model, whatever its deficiencies, explains how Earth sentences and Twin/New Earth sentences can differ in truth-value without differing in meaning.

    Second reply (Sellars):  Uses do differ as between Earth and Twin Earth.  They are characterized in part by reference to environmental things.  On Earth, if I say "Water, waiter," the waiter is supposed to bring me water (H2O); on Twin Earth, the waiter is supposed to bring Twin Lycan Twin water (XYZ).

    Objection 3.  Could I not know the use of an expression and fall in with it, mechanically, but without understanding it?  I have known undergraduates who are geniuses at picking up academic jargon of one sort or another and slinging it with great facility, but without understanding.  As an undergraduate himself, my former colleague Steven Boer took a phenomenology course taught by a visiting Parisian, and understood none of it, but learned the knack of stringing the jargon expressions together so well that his term paper earned or "earned" an A.  Use perfect (or at least graded A); meaning nil.

    Reply (Brandom):  If Boer's facility with the jargon had really been entirely reliable, so that he could not be crashed or detected as a phony even under the most demanding conversational situations (which was surely not the actual fact), then what's left out?  That he did not have a warm, fuzzy feeling of understanding seems irrelevant.

    Objection 4.  Many rule-governed social activities -- sports and games themselves in particular -- do not centrally involve the kind of meaning that linguistic expressions have.  Certainly chess moves and tennis shots do not have meaning of that sort.  (Contrast the case where spies are using chess moves as an actual secret code; for example, N-Q3 may have conventionally been stipulated to mean "Take the zircon to Foppa and tell him we move tonight.")  What, then is supposed to distinguish language-games from ordinary games?
    Suppose that some community agrees to use certain words -- or at any rate sounds and marks -- in a peculiar way; say they decide to put only "words" with the same number of syllables next to each other in threes, or they utter "sentences" only in rhyming pairs, where each string begins with a one-letter word and adds one letter successively to each ensuing item.  (This might be a sort of community-wide parlor game.)  Or, again, consider the Placard Game.  If a newcomer happened upon this whimsical society and knew nothing of the arrangement, s/he would not understand what was going on.  The newcomer might, in time, dope out all the rules according to which the various tokens were being used, and yet have no notion what, if anything, was being said.   And in the simple cases, at least, nothing is being said.  Someone might suggest that such a game, like the builders’ language, is just too simple and/or primitive to qualify.  But it is hard to see how the mere addition of further complexity would help.
    Just after class, Sophie brought up a great further example.  For those of you who have had Symbolic Logic:  Think of doing proofs in the predicate calculus when the constants and predicates have not been interpreted.  A highly rule-regulated social activity, but without interpretation (assigning of--you guessed it--referents to the constants and predicates), none of the formulas you write down are meaningful, and nothing gets said; it's just a calculus.

    First reply (Freidrich Waismann, who anticipated this objection):  Genuine language-games are "integrat[ed]... into life."  By contrast, the placard movements, like chess moves and tennis shots, "bear a far less close relation to life than words used in earnest."  A language-game cannot be encapsulated, something that we keep at arm's length and play just when we feel like it.

    But, rejoinder:  Some language-games, such as the telling of shaggy-dog jokes, are encapsulated and played only occasionally and at will.  Also, even if we agree that more serious, multi-purpose language-games are thoroughly integrated into life, we usually think of that close, integrative relation as that of referring.  The Wittgensteinian does not agree that meaning essentially involves referring, and so Waismann needs to say what the "integration" is instead.  The idea seems to be that language-games are integrated with other social practices.  But it is hard to see how the Wittgensteinian can spell that out (a) in such a way as to explain how the linguistic moves take on propositional content, but (b) without secretly introducing referring.

    Second reply:   Someone might argue that if its rules are rich enough and advert often enough to ambient conditions, reference and predication will be recoverable from the game description.  Suppose there is a rule that whenever the waiter comes in, every third player shouts "Here, waiter," and is given a martini; whenever any player says "Mix please," s/he is passed the bowl of snacks by whomever is nearest it; and the like.  One would then be tempted to conclude that "waiter" refers to the waiter and "mix" means snack food.   So the game moves would have meaning after all.

    Rejoinder:  Perhaps in that case, the utterances specified by the game rules would have meanings -- but only because they do stand for or refer to things and not just because of their conventional deployment behavior.
    Let us therefore firmly stipulate that, no matter how complex  the game becomes, the players’ utterances do not refer to things external to the game; they are only moves in the game.  Call this the Pure Placard Game.  But then it seems even more obvious that the game is not even the beginning of an actual language, and that the moves do not have meaning in the same way that utterances of English sentences do.  So the Use theorist's explicit conditions are not sufficient for something's being a language.

    That's as far as we got in class.  Here are a few more considerations.

    Objection 5.  One clear sense in which a social practice qualifies as an actual language is that, according to it, one can make noises or inscribe marks and thereby say that P for some suitable sentence replacing "P."  And one of the things that is surely essential to language is that we can say things in it.  But no such indirect discourse is licensed just in virtue of some people's playing chess or the Pure Placard Game; none of the players has said or asked or requested or suggested...that anything at all.  There is something missing.  We are playing a game, and using tokens according to a set of conventional rules, and engaging in a social practice that may not only be fun but have some larger point; it might even be in some way vital to our way of life.  The things the players in these various games have done may have significance in some sense, but nobody has made any assertions or asked whether anything or advised anyone to do anything.

    Objection 6.  A point closely related to Objection 5 is that declarative sentences can differ in meaning from each other even though they are all used to assert and even though they are very similar in surface-grammatical form, e.g.,

            (R) The ball is red.
            (B) The ball is blue.

How do these differ in "use"?  Only, it seems, in the contents of the assertions they are used to make.  But the Use theorist does not want to appeal to assertion "contents" (= meanings or propositions).  S/he might try to avoid reference to "contents" by using "that"-clauses: "(R) is used to assert that the ball is red."  But "that"-clauses already presuppose a notion of sameness of meaning.  ("Adolf asserted that Eva is faithful" means something like, "Adolf asserted something that in his language means the same as: Eva is faithful.")  On pain of circularity, the Use theorist cannot presuppose a notion of sameness of meaning in offering an analysis of sameness of meaning.

    Objection 7.  What is (in one obvious sense) the same sentence can be cast in different grammatical moods:

            Mother will eat the oyster.
            Will mother eat the oyster?
            Mother, eat the oyster!

Plainly those (in an equally obvious sense) three sentences are closely related in meaning, and any adequate theory of meaning will have to reflect that.  But the Use theorist appeals only to use-potential in determining sameness or difference of meaning.  And our three sentences differ entirely in use-potential.  (The first would be used to predict or to guarantee, the second to ask a question, and the third to influence someone's behavior).  It seems the Use theorist is committed, repugnantly, to saying that the three sentences differ in meaning from each other as sharply as it is possible for any sentences to differ in meaning.

    Objection 8.  It is hard to see how a theory that took "Hello" or "slab" as its paradigms could succeed in explaining some of the more refined of the meaning facts.  Meaningfulness, synonymy and ambiguity are not a problem; but what of entailment between complex sentences?  The sentence "Chuck buttered the toast with a paring knife in the bathroom at 2:00 a.m." entails each of the sentences "Chuck buttered something," "Chuck used a knife at 2:00 a.m.," "Someone buttered something in the bathroom," "The toast got buttered with a paring knife," and other such.  How might a Use theorist explain those entailments?

    Objection 9.  Proper names pose a problem for the Use theorist.  Try stating a rule of use for the name "William G. Lycan," or for the name of your best friend.  Remember, it has to be a rule that every competent speaker of your local dialect actually obeys without exception, because it is supposed to constitute the public linguistic meaning of the name.  The only candidate rules that occur to me push the Use theorist into a Description theory of meaning for names.  As before, arguably some names to have fixed descriptions associated with them ("Jack the Ripper"), which would give us some handle on their Wittgensteinian uses, but most don't.  (Wittgenstein himself found Descriptivism congenial, but he had not read Kripke.)