H. P. GRICE E J. L. SPERANZA: LA CONVERSAZIONE -- I VERBALI: PACE

 G.: Let us begin where Pace begins, with interpretation made narrower than it ought to be and yet not wholly unserviceable. S.: You mean his insistence that interpretatio is vox. G.: Exactly. “Interpretatio namque est vox, quae animi sensa, et per ea res ipsas ex instituto significat.” Useful as far as it goes, but it goes only by voice. S.: Whereas you want signification in a wider field. G.: Much wider. Things other than vox signify. A bandaged leg may signify refusal. Smoke may signify fire. A pause may signify annoyance. Silence may signify more than some syllogists. S.: Then Pace’s opening restriction is already too grammatical. G.: Too grammatical and too scholastic in the narrow sense. He wants the parts of syllogism, and so he begins where the schoolroom begins, with nomen, verbum, enuntiatio. S.: Yet he does at least inherit a respectable Latin apparatus. G.: He does. Cicero’s signum, signare, significare are all there in the background, though Pace borrows more than he returns. S.: A humanist acquisitiveness. G.: Precisely. But the acquisitiveness is useful. It lets us pry open his abbreviations. S.: Then let us take your own form first. S(x,y,z). G.: Yes. Let S stand for total signification. x is the utterer, z the addressee, and y the content, the thing reportable by a that-clause. S.: So if x says “Socrates currit” to z, then S(x,y,z) holds where y is that Socrates runs. G.: Correct as a beginning. But the crucial point is that y need not exhaust total signification. S.: Because the dictum may be narrower than the full signification. G.: Exactly. Pace is content, much of the time, with the interpretation as that which expresses animi conceptus. I want the whole communicative scene. S.: Which already means that “animi sensa” has to be expanded. G.: Yes. Not merely conceptions in the mind, but what the utterer means the addressee to gather. S.: So Pace’s route is res to conceptus to vox. G.: Roughly. And mine, in the interesting cases, is utterer to utterance to recognisable intention to addressee’s uptake. S.: Triadic from the start. G.: Entirely. Signification is not a property of a sound alone. It is what one person means to another by means of some vehicle. S.: Then your first complaint against Pace is that he overprivileges the vehicle. G.: Yes. He tells us too quickly that interpretatio is vox, when really vox is only one family of vehicles. S.: Yet in his defence, he is commenting on De interpretatione as part of the Organon. G.: Quite. He has the parts of syllogism before him, and that narrows the field. I do not blame him for narrowing it; I blame readers who take the narrowing for completeness. S.: Then the next step is ex instituto. G.: Yes, and there he is better. He sees that what is in voce is not by nature but by institution, by posit, by accepted use. S.: Against Cratylus. G.: Exactly. A proper anti-Cratylean move. Words do not cling to things by nature like burrs to trousers. S.: Though one must then ask what institution really adds. G.: Quite. Pace says, sensibly enough, that different peoples have different voces and scripturae though the animi conceptiones and the res themselves are the same. S.: Hence Greek and Latin differ grammatically while logic remains one. G.: Yes. “Caelum” in one tongue, something else in another, yet the logical object remains what it is. S.: That is one of his better observations. G.: It is. But still too inward, if left alone. S.: Because he says that names signify immediately the concept, and consequently the thing through the concept. G.: Precisely. A tidy representational chain. But the communicative act is not only representation; it is directed uptake. S.: So we need to widen “significat conceptum” into something like “x by uttering u means y to z.” G.: Exactly. And that is why I prefer S(x,y,z) to any merely dyadic significat relation. S.: Then how do we reform Pace without ruining him? G.: By keeping his distinctions but changing the level. Take his simple interpretation, nomen and verbum, and his composite interpretation, enuntiatio. S.: Simple items signify without truth-value; composite items can be true or false. G.: Yes. That much is perfectly serviceable. “Homo” and “currit” are simple; “homo currit” is composite. S.: And “hircocervus” signifies though there is no such thing. G.: Good. He is sensible on that too. The name can signify without yet being true or false until “est” or “non est” enters. S.: So “hircocervus” is not false by itself. G.: Precisely. A point some moderns would improve by making it worse. S.: Then where do you part company? G.: At the point where he thinks that because logic chiefly regards the internal speech, one may treat the external vehicle as if its role were exhausted by representing conceptions. S.: Whereas for you external use matters. G.: Entirely. Internal conception is not enough to explain what someone means in uttering something to someone else. S.: Then even if we keep his dictum–or rather enuntiatio–we need another layer. G.: Yes. Let us call the dictum d, if you like, the explicit propositional content made available by the utterance. S.: And then the total signification S(x,y,z) may include more than d. G.: Exactly. It may include what is implicated, suggested, allowed to be gathered. S.: So Pace gives us the bare “that”-clause, and you want to ask what else x means z to gather by means of that clause. G.: Correct. Pace’s analysis remains close to what would later be called locutionary content. I want the full communicative economy. S.: Then perhaps we should take one of your standard examples. G.: By all means. Suppose x, a tutor, says to z, another don, “Smith has beautiful handwriting.” S.: The dictum is that Smith has beautiful handwriting. G.: Yes. That is what can be reported under a straightforward “that”-clause. y1, if you like. S.: But in context the total signification includes y2, that Smith is poor at philosophy. G.: Precisely. And that is not a second sense of “beautiful handwriting.” It is a broadened signification under conditions of use. S.: Which Pace’s machinery, left alone, cannot capture. G.: Not comfortably. He would need to say either that the utterance has another enuntiatio hidden within it, or that the hearer moves by some practical reasoning beyond the enuntiatio. S.: You prefer the latter. G.: Entirely. The dictum remains what it is. The significatum in the fuller conversational sense exceeds it. S.: Then perhaps we need levels. y1 for dictum, y2 for implicatum, and S(x,{y1,y2},z) for total signification. G.: Very good. That is already better than Pace without being unfaithful to him. S.: And if we keep his “animi sensa,” we might say that y1 corresponds to the concept explicitly expressed, while y2 corresponds to what the utterer intends the addressee to infer under rational assumptions. G.: Exactly. Though I would not speak of y2 as another “concept in the mind” in Pace’s static way. It is an intended inferential destination. S.: Nicely put. G.: Keep it and flatten it later. S.: Then what of his claim that the same enuntiatio may occur across languages because what is in voce is the same insofar as it represents the same concept? G.: Good as far as it goes. “Omnis homo est animal” and its Greek counterpart can indeed be the same enuntiatio in one respectable sense. S.: Because the proposition is the same though the voces differ. G.: Yes. But again, conversationally, the same proposition uttered in Greek and in Latin may not have the same total signification in the same scene. S.: Because the choice of language itself may signify something. G.: Precisely. Choice of idiom, register, language, timing, order, all may enter into what is meant. S.: So Pace abstracts away from pragmatic atmosphere. G.: Entirely. He has to, to do the schoolwork he is doing. But we must not inherit the abstraction as ontology. S.: Then there is his distinction between the simple interpretation as nomen and verbum, the same items as subjectum and attributum within enunciation, and as major, minor, or middle term within syllogism. G.: Yes. A nice set of role distinctions. The same item can be considered per se, then as part of a proposition, then as part of a syllogism. S.: You like that, surely. G.: Very much. It shows a decent awareness that what something is communicatively depends on the larger whole in which it functions. S.: Which is almost Gricean already. G.: Structurally, yes. A word said alone, a word in an enuntiatio, and a word as term in an argument are not different sounds but different functional standings. S.: So one might extend his insight beyond syllogism into conversation. G.: Exactly. A sentence considered per se, a sentence as answer, a sentence as refusal, a sentence as hint, a sentence as irony. S.: Then his role distinctions become the seed of pragmatic role distinctions. G.: Quite so. “Smith has beautiful handwriting” per se is praise of handwriting. As answer in a tutorial report it may be faint praise. As reply to a question about philosophical promise it may be damnation in gloves. S.: Then the same enuntiatio changes role without changing sense. G.: Precisely. That is one of the central lessons. S.: Which means Pace helps most where he is least ambitious. G.: Usually the fate of commentators. S.: Then what about his opposition between simple conceptions, expertes of truth and falsity, and composite conceptions where truth and falsity arise? G.: Entirely sound, so long as one keeps clear what level is in question. A bare term or name is not true or false. Truth-value enters with composition or division. S.: Homo est animal. Homo est lapis. G.: Exactly. And he rightly sees that composition and division operate both in mente and in voce. S.: You approve that too? G.: Yes, though again I wish he had looked harder at use. For in conversation the composition may be explicit while the division is implicated, or the reverse. S.: An example? G.: Suppose x says, “He is certainly original.” The explicit composition is praise. The implicated division may be from the class of the competent. S.: Very Oxford. G.: Entirely. We divide while appearing only to compose. S.: Then Pace’s compositio and divisio can be pragmatically retooled. G.: Very much so. Not merely affirmation and negation in grammar, but modes of placing and separating under communicative purpose. S.: So one might say that x composes a predicate with a subject explicitly while dividing the subject from some expected evaluation implicitly. G.: Admirably put. S.: Thank you. G.: Do not become too pleased. S.: I shall remain only syntactically vain. G.: Worse and worse. Now, what of his insistence that logic concerns the internal speech rather than the external? S.: You think that overdone. G.: Entirely overdone, if made foundational. Internal speech may be useful for some explanatory purposes, but what interests me is public meaning in use. S.: Because a meaning not available to uptake is hardly your concern. G.: Precisely. A merely inwardly murmured conception is not yet a conversational act. S.: Then S(x,y,z) is public from the start. G.: Entirely. It involves an utterer, an addressee, and a meant content. S.: And that meant content may outrun what Pace would call the enuntiatio taken per se. G.: Exactly. That is why his scope is narrow. S.: Yet not contemptibly narrow. G.: No. Let us be fair. He gives a disciplined account of the dictum. That is already something. S.: Then perhaps the right line is this. Pace provides a semantics of the dictum; Grice requires a pragmatics of total signification. G.: Precisely. Keep that, too. S.: So if we revisit his famous definition, “interpretatio est vox quae animi sensa et per ea res ipsas ex instituto significat,” we might rewrite it. G.: Yes. Try. S.: “Interpretatio, in Pace’s narrow logical sense, is a vocal sign instituted to express mental content and thereby represent things; but in the fuller conversational sense, an utterance is a vehicle by which an utterer x means some content y to an addressee z, often by allowing z rationally to infer more than the explicit dictum contains.” G.: Excellent. Too long for a scholastic margin, but very serviceable. S.: Then should we keep vox at all? G.: Only with a warning. Voice is one chief vehicle in De interpretatione, but not the only medium of signification in life. S.: So the bandaged leg returns. G.: Always. Philosophers who forget the bandaged leg deserve a limp. S.: And what of scriptura? G.: The same point. Writing is not merely notation of voice. It may also do its own pragmatic work. S.: Because the choice to write, rather than say, itself signifies. G.: Exactly. Delay, permanence, distance, publicity, deniability, all that. S.: Pace treats scriptura chiefly as sign of voice. G.: Which is too thin for later purposes. A note slipped under the door and a sentence spoken across a table do not mean in the same way merely because the propositional content matches. S.: So the external media are not philosophically negligible. G.: Far from it. They are often the whole game. S.: Then perhaps his best legacy is his layered nomenclature. Nomen, verbum, enuntiatio, propositio, problema, conclusio. G.: Yes. He shows that one and the same item changes philosophical character according to functional setting. S.: Which encourages your own treatment of utterances as answer, hint, rebuke, refusal, and so on. G.: Exactly. Conversational role is the pragmatic counterpart of his logical role. S.: Then the transition from Pace to you runs not through words as such, but through role within a larger rational whole. G.: Splendid. That is exactly right. S.: So when he says that dictiones become enunciationes and enunciationes become syllogismi, you would say that utterances become moves and moves become exchanges. G.: Very good indeed. Pace’s ladder is logical; mine is conversational. But both are ladders of function. S.: Then let us come back to hircocervus. G.: A dear old friend. S.: Pace says it signifies though there is no such thing, and that only with “est” or “non est” does truth-value arise. G.: Quite sound. And useful against crude referentialism. S.: But you would add that in conversation “hircocervus non est” may do more than state a falsehood or truth about non-being. G.: Exactly. It may be jest, correction, irony, scholastic display, annoyance, or all four before luncheon. S.: So again the dictum is not the whole of signification. G.: Precisely. Pace gives us the logical minimum. Conversation supplies the humane excess. S.: Humane excess sounds suspiciously like rhetoric. G.: Because it is. But disciplined rhetoric, not bad upholstery. S.: Then would you say that Pace has no place for implicatum at all? G.: Not explicitly. But he has spaces in which it can later be inserted: ex instituto, role distinctions, composition and division, the priority of concept over vehicle, and the recognition that grammar and logic do not coincide. S.: Because grammar differs by language, while logic remains one. G.: Yes. And one may then add: while pragmatics varies with occasion, institution, and speaker intention. S.: So the full picture would be something like this. Grammar concerns voces and their forms. Logic concerns what can be true, false, inferred, denied, and composed. Pragmatics concerns what one person thereby manages to mean to another here and now. G.: Admirably put. Pace largely handles the first two. I insist on the third. S.: Then perhaps the 100th move must be this. Pace tells us what sort of thing can count as a dictum in the logical building. You tell us what sort of thing can be done with it once a human being puts it to use before another. G.: Exactly. And between the two lies all the difference between De interpretatione read in Examinations Hall and conversation overheard in a corridor. S.: Dry enough? G.: Sufficiently Paduan, with just enough Oxford draught.