Grice's M

meta-ethics: For Grice it is complicated, since there is an ethical or practical side even to an eschatological argument. Grice’s views on ethics are Oxonian. At Oxford, meta-ethics is a generational thing: there’s Grice, and the palaeo-Gricieans, and the post-Gricieans. There’s Hampshire, and Hare, and Nowell-Smith, and Warnock. P. H. Nowell Smith felt overwhelmed by Grice’s cleverness and they would hardly engage in meta-ethical questions. But Nowell Smith felt that Grice was ‘too clever.’ Grice objected Hare’s use of descriptivism and Strawsons use of definite descriptor. Grice preferred to say “the the.”. “Surely Hare is wrong when sticking with his anti-descriptivist diatribe. Even his dictum is descriptive!” Grice was amused that it all started with Abbott BEFORE 1879, since Abbott’s first attempt was entitled, “Kant’s theory of ethics, or practical philosophy” (1873). ”! Grices explorations on morals are language based. With a substantial knowledge of the classical languages (that are so good at verb systems and modes like the optative, that English lacks), Grice explores modals like should, (Hampshire) ought to (Hare) and, must (Grice ‒ necessity). Grice is well aware of Hares reflections on the neustic qualifications on the phrastic. The imperative has usually been one source for the philosophers concern with the language of morals. Grice attempts to balance this with a similar exploration on good, now regarded as the value-paradeigmatic notion par excellence. We cannot understand, to echo Strawson, the concept of a person unless we understand the concept of a good person, i.e. the philosopher’s conception of a good person.   Morals is very Oxonian. There were in Grices time only three chairs of philosophy at Oxford: the three W: the Waynflete chair of metaphysical philosophy, the Wykeham chair of logic (not philosophy, really), and the White chair of moral philosophy. Later, the Wilde chair of philosophical psychology was created. Grice was familiar with Austin’s cavalier attitude to morals as Whites professor of moral philosophy, succeeding Kneale. When Hare succeeds Austin, Grice knows that it is time to play with the neustic implicatum! Grices approach to morals is very meta-ethical and starts with a fastidious (to use Blackburns characterisation, not mine!) exploration of modes related to propositional phrases involving should, ought to, and must. For Hampshire, should is the moral word par excellence. For Hare, it is ought. For Grice, it is only must that preserves that sort of necessity that, as a Kantian rationalist, he is looking for. However, Grice hastens to add that whatever hell say about the buletic, practical or boulomaic must must also apply to the doxastic must, as in What goes up must come down. That he did not hesitate to use necessity operators is clear from his axiomatic treatment, undertaken with Code, on Aristotelian categories of izzing and hazzing. To understand Grices view on ethics, we should return to the idea of creature construction in more detail. Suppose we are genitors-demigods-designing living creatures, creatures Grice calls Ps. To design a type of P is to specify a diagram and table for that type plus evaluative procedures, if any. The design is implemented in animal stuff-flesh and bones typically. Let us focus on one type of P-a very sophisticated type that Grice, borrowing from Locke, calls very intelligent rational Ps. Think of them very roughly as creatures with the capacities for thought and action characteristic of persons. Being benevolent genitors, we want to design these Ps so as to maximize their chances for survival. As Grice recently pointed out in conversation-by talk of survival, he does not, in the case of very intelligent rational Ps, mean simply staying alive. A full explanation of what Grice has in mind here would require an account of his views on teleology; however, for our purposes a full explanation is unnecessary. We need note only the following points. First, in constructing Ps we build in certain ends, and for our purposes we may imagine ourselves as having a fairly free hand in deciding what ends to select. To build in an end is to construct the diagram and table so that the Ps have that end as a standing, constant end-an end where they strive to realize in all appropriate circumstances. The restriction to appropriate circumstances is necessary for two reasons. First, we will want to endow the Ps with a variety of ends, and we will not want a P to try to realize each end at each moment of time. We want them to schedule their pursuit of ends in a way that maximizes the realization of the whole array in the long run. Second, we will, in the case of very intelligent rational Ps, want to give them the (limited) ability to eliminate (or inhibit for a long time the pursuit of) built-in ends should circumstances prove especially inappropriate. Now we can explain what, for present purposes, we mean by survival: to maximize chances for survival is to maximize chances for the realization of built-in ends. How are we to design the Ps so as to maximize their chances for realizing the built-in ends? The answer would be easy if we could take as given a very detailed specification of the environment in which the Ps live. Then we could tailor the diagram and table to that specific environment by building in exactly the responses that the environment demands. But we cannot assume such a specific description of the environment; on the contrary, we know that the Ps will face a variety of changing environments. So we need to design the Ps to function effectively in the widest possible range of environments. We could, of course, avoid this if we were willing to descend periodically from Olympus in order to redesign the Ps in response to each significant change in the environment. But there is a more efficient way to achieve the same result: we give the Ps the ability to redesign themselves. There are two aspects to this ability. First among the ends we build in is the end of being an end-setter. To be an end-setter requires that one have the (limited) ability to adopt new ends and to eliminate ends one already has. To have the end of being an end-setter is to have the end of employing this ability to adopt and eliminate ends. This is not, as we will see, a complete specification of what it is to be an end-setter, but it will suffice for the moment. By making the Ps end-setters we will enable them to redesign themselves by altering what they aim at. Second, to enable Ps to determine when to use their end-setting ability, we have given them an appropriate set of evaluative principles. These principles incorporate in the Ps some of our wisdom as genitors. We do not need to descend periodically to redesign them because in a sense we are always present-having endowed them with some of our divine knowledge. What does this have to do with ethics? Grice answers this question in Method in Philosophical Psychology. To interpret the reference to rational capacities and dispositions in the following passage, recall that, given the connection between evaluative principles and rationality Grice spells it out, we have, in giving the Ps evaluative principles, given them a capacity for rational evaluation. Let me be a little more explicit, and a great deal more speculative, about the possible relation to ethics of my programme for philosophical psychology. I shall suppose that the genitorial programme has been realized to the point at which we have designed a class of Ps which, nearly following Locke, I might call very intelligent rational Ps. These Ps will be capable of putting themselves in the genitorial position, of asking how, if they were constructing themselves with a view to their own survival, they would execute this task; and, if we have done our work aright, their answer will be the same as ours . We might, indeed, envisage the contents of a highly general practical manual, which these Ps would be in a position to compile. The contents of the initial manual would have various kinds of generality which are connected with familiar discussions of universalizability. The Ps have, so far, been endowed only with the characteristics which belong to the genitorial justified psychological theory; so the manual will have to be formulated in terms of that theory, together with the concepts involved in the very general description of livingconditions which have been used to set up that theory; the manual will therefore have conceptual generality. There will be no way of singling out a special subclass of addressees, so the injunctions of the manual will have to be addressed, indifferently, to any very intelligent rational P, and will thus have generality of form. And since the manual can be thought of as being composed by each of the so far indistinguishable Ps, no P would include in the manual injunctions prescribing a certain line of conduct in circumstances to which he was not likely to be Subjects; nor indeed could he do so even if he would. So the circumstances for which conduct is prescribed could be presumed to be such as to be satisfied, from time to time, by any addressee; the manual, then, will have generality of application. Such a manual might, perhaps, without ineptitude be called an immanuel; and the very intelligent rational Ps, each of whom both composes it and from time to time heeds it, might indeed be ourselves (in our better moments, of course). We can both explain and motivate this approach to ethics by considering three objections. First, one may complain that the above remarks are extremely vague. In particular, what are the evaluative principles-the rational capacities and dispositions-with which we endow the Ps? These principles play a central role in compiling the manual (Immanuel). How can we evaluate the suggested approach to ethics until we are told what these evaluative principles are? This complaint is somewhat unjust-in the context of “Method” at least, for there Grice labels his remarks as speculative. But, more importantly, Grice has done a considerable amount of work directed toward providing this objection with the information it demands; this work includes investigations of happiness, freedom, reasoning, and teleology. While the examination of these projects is unfortunately beyond the scope of our introduction, we should comment briefly on Grices work on happiness. In Some Reflections about Ends and Happiness, Grice develops an account of happiness, and on this account it is clear that the conception of happiness could certainly function as a central evaluative principle in endsetting. It is also worth remarking here that Grices views on happiness are very Aristotelian; Grice emphasizes the Kantian aspect of his view in the passage quoted, but when the views are worked out, one finds a blend of Kantian and Aristotelian themes. The second objection is that Grices approach makes it too easy to escape the demands of morality. What can Grice say to a personor P-who rejects the manual, rejects moral demands and constraints? Suppose, for example, that a person reasons as follows: If I continue to heed the voice of morality, I will continue on occasion to sacrifice my welfare and interests in favor of anothers welfare and interests. Why should I be such a fool? After all, what am I after except getting as much as I can of what I want. Thorough-going egoism is the path to take; Ill have to resist these impulses to help others, in the way I resist sweets when I am dieting. Perhaps I will be able to condition such impulses out of myself in time. Does Grices approach have a reply to the consistent thorough-going egoist? It does-as Grice pointed out in a recent conversation; the considerations which follow are based on that conversation. First we need to provide a more detailed account of end-setting. When we give our Ps the end of end-setting we have a good reason for giving them each of the evaluative principles in order to build in the capacity to redesign themselves, and we build in that capacity in order to maximize their chances of realizing their ends over the widest possible range of environments. So we have a good reason for giving them each of the end-setting evaluative principles: Namesly, each one contributes to the capacity of redesigning in a way that maximizes the chances of realizing endls. The Ps themselves are capable of recognizing that the evaluative principles make such a contribution, so each P has (or can have) a reason for having the evaluative principles. (We are assuming that contributing to the maximization of the realization of ends constitutes a good reason; a defence of this assumption would require an examination of Grices view on teleology.) A second essential point is that we design the Ps so that they do not simply adopt or eliminate ends at will; rather, they do so only when they have good reasons to do so-good reasons derived from the evaluative principles that govern end-setting. We design them this way in order to maximize their chances for the realization of their ends. We want them to use their ability for end-setting only when the evaluative principles we have built in determine that a change of ends is called for in order to maximize the overall realization of ends. (In the typical case at least, an end-setter will only alter some of his ends as to maximize the realization of all his (remaining and newly adopted) ends.) An end-setter then has the end of adopting or eliminating ends when he has good reasons to do so-where these reasons are provided by evaluative principles; and these evaluative principles are such that he has a good reason for having each of those principles. Let us call such an end-setter a Griceian end-setter. Returning now to egoism, we can distinguish three different situations in which one might try to reject the demands of morality. Before going on, one may insist on knowing what we mean by the demands of morality, but it is enough for present purposes that we agree that morality demands at least that one does not always treat others purely as means to ones own ends. It is this demand that the egoist described earlier rejects. First, if the egoist is a Griceian end-setter who wishes to remain a Griceian end-setter, then he cannot abandon the non-egotistical principles since they are self-justifying and do not depend on other premisses. Second, if the egoist envisioned is one who would cease to be a Griceian end-setter, this too is impossible for a rational agent. Being a Griceian end-setter is itself one of the self-justifying ends, and thus it can be abandoned only if one abandons reasoning. Finally, there is the question of whether an agent who is not a Griceian end-setter can be an egoist. Again the answer appears to be no, if the agent is rational and considers the question. For being a Griceian end-setter can be seen on reflection to be a self-justifying end, and thus must be adopted by any reflective rational agent. Let this suffice as a brief indication of Grices approach to the second objection, and let us turn to the third and last objection. This objection concerns what we have been calling the demands of morality; the objection is that the notion of demand is vague. What do we mean by demand when we talk of the demands of morality? What kind of demand is this? What sort of claim is it that morality has on us? Grice has done a considerable amount of work relevant to this question including Probability, Desirability, and Mood Operators, the Locke Lectures, and recent work on Kant. In explaining the claim morality has on us, Grice employs distinctions and notation provided by his theory of meaning. We can begin with the sentence Pay Jones the money! Grice assigns this sentence the following structure: !+I pay Jones the money where ! is the imperative mood operator and I pay Jones the money is a moodless sentence radical. This structure is embeddable in other sentences. In particular, it occurs in both I should pay Jones the money and I should not pay Jones the money. Grice assigns these the following structures: Acc+!+I pay Jones the money; Not+ Acc+! +I pay Jones the money, where Ace may be read as it is acceptable that. So if we read ! as let it be the case that, the whole string, Acc+! I pay Jones the money may be read as: It is acceptable that (let) it be the case that I pay Jones the money (whole Not+ Acc.+! +I pay Jones the money may be read as It is not the case that it is acceptable that (let) it be the case that I pay Jones the money). In Probability, Desirability, and Mode Operators Grice motivates this assignment of structures by arguing, in effect, that the sentence I should pay Jones the money means-on the central and important reading-that it is acceptable that (let) it be the case that I pay Jones the money. The argument rests on an analysis of practical reasoning and on the analysis of sentence meaning. Actually, Grice does not say that I should pay Jones the money means what we just said it means. In Probability, Desirability, and Mode Operators he is much more circumspect. After discussing probability inferences, Grice notes that, bearing in mind the variety of interpretations to which sentences containing ought and should are susceptible, he finds it natural to take, as practical analogues to sentences like an invalid is likely to be in retirement, sentences like it is desirable for an invalid to keep in touch with his doctor. For expositional purposes, he uses should-sentences since the interpretation we want these sentences to bear is clear, and the use of should-sentences highlights the connections with ordinary moral reasoning. Suppose morality demands that I pay Jones the money; that is, I act morally only if I pay Jones the money. Grice holds that this is true only if an appropriate sentence (or thought) is derivable from my evaluative principles-a sentence (or thought) whose underlying structure is Acc+!+I pay Jones the money. I can, that is, derive that it is acceptable that (let) it be the case that I pay Jones the money; in other words, that I should pay Jones the money. Grice holds that since I derive this from evaluative principles, it is necessary; that is, it is necessary that I should pay Jones the money. There are two points to note in order to explain the claim morality has on us. First, Grice holds that the self-justifying evaluative principles are necessarily true, and he holds that I can show, e.g. that it is necessarily true that I should pay Jones the money, by constructing a suitable derivation of I should pay Jones the money from my self-justifying evaluative principles. These claims follow from a general view Grice has of the nature of necessity, a view that he considers elsewhere. To be more precise, what I derive from my evaluative principles is a sentence with the underlying structure: Acc+!I pay Jones the money, which we read as It is acceptable that (let) it be the case that I pay Jones the money. Since it is possible to construct an appropriate derivation it is necessary that it is acceptable that (let) it be the case that I pay Jones the money. This is how we should understand attaching necessary to a should-statement. The sentence Necessarily, I should pay Jones the money expresses the necessary acceptability of the imperative Pay Jones the money! (Since my derivation will involve contingent information about the circumstances C, we should represent what I derive as I should in these circumstances C pay Jones the money; this will be what is necessary. We ignore this detail.) Second, it does not follow from the fact that it is necessary that I should pay Jones the money that I will pay him the money. Even if it is necessary that it is acceptable that (let) it be the case that I pay Jones the money, and even if I derive this, I may not act on it. It is true that I cannot have a good reason not to act on it; after all, I have derived the necessity of accepting the imperative, Pay Jones the money!; and as a Griceian end-setter I am committed to acting on such reasons; but this does not mean I will. A person is capable of irrationality-even in the face of acknowledged necessity. Now we are in a position to explain what we mean by talk of the demands of morality. The demands of morality are expressed by necessary should-statements. Or perhaps we may want to say that they are expressed by a special subset of such statements. We need not investigate this possibility since it would not alter the point we are making here-which is that the demands of morality express the necessity of rational agents accepting and acting on certain imperatives (in so far as they act rationally). Consider the role elements of Grices theory of meaning play in the above discussion of ethics, we have in a way returned to the startingpoint of our exposition of Grices views. And it is certainly high time we let the discoverer of M-intentions formulate some in response to what we have written. High time but not quite time. For one thing, we should note that the discussion of ethics resolves an issue we suppressed when discussing psychological explanation. At one point in that section, we wrote, with respect to M-intending, Given our ends and our environment, there is good and decisive reason to have such a pre-rational structure. We did not raise the question of what makes those considerations into a reason; we tacitly assumed that relations to happiness and survival secured that the considerations counted as reasons. The ethics discussion points the way to detailed and informative treatment of this issue. Not that the discussion suggests that we were wrong to appeal tacitly to happiness and survival; on the contrary, it indicates that we should explain the reason-giving force of such considerations by examining the role they play for a Griceian-end-setter. Refs.: Most of Grice’s theorizing on ethics counts as ‘meta-ethic,’ especially in connection with R. M. Hare, but also with less prescriptivist Oxonian philosophers such as Nowell-Smith, with his bestseller for Penguin, Austin, Warnock, and Hampshire. Keywords then are ‘ethic,’ and ‘moral.’ There are many essays on both Kantotle, i.e. on Aristotle and Kant. The H. P. Grice Papers, BANC.

mode: Grice is a modista. He sometimes did use ‘modus,’ after Abbott. The earliest record is of course “Meaning.” After elucidating what he calls ‘informative cases,’ he moves to ‘imperative’ ones. Grice agreed with Thomas Urquhart that English needed a few more moods! Grice’s seven modes.Thirteenthly, In lieu of six moods which other languages have at most, this one injoyeth seven in its conjugable words. Ayer had said that non-indicative utterances are hardly significant. Grice had been freely using the very English not Latinate ‘mood’ until Moravcsik, of all people, corrects him: What you mean ain’t a mood. I shall call it mode just to please you, J. M. E. The sergeant is to muster the men at dawn is a perfect imperative. They shall not pass is a perfect intentional. A version of this essay was presented in a conference whose proceedings were published, except for Grices essay, due to technical complications, viz. his idiosyncratic use of idiosyncratic symbology! By mode Grice means indicative or imperative. Following Davidson, Grice attaches probability to the indicative, via the doxastic, and desirability to the indicative, via the buletic-boulomaic.  He also allows for mixed utterances. Probability is qualified with a suboperator indicating a degree d; ditto for desirability, degree d. In some of the drafts, Grice kept using mode until Moravsik suggested to him that mode was a better choice, seeing that Grices modality had little to do with what other authors were referring to as mood. Probability, desirability, and modality, modality, desirability, and probability; modality, probability, desirability. He would use mode operator. Modality is the more correct term, for things like should, ought, and must, in that order. One sense. The doxastic modals are correlated to probability. The buletic or boulomaic modals are correlated to desirability. There is probability to a degree d. But there is also desirability to a degree d.  They both combine in Grices attempt to show how Kants categorical imperative reduces to the hypothetical or suppositional. Kant uses modality in a way that Grice disfavours, preferring modus. Grice is aware of the use by Kant of modality qua category in the reduction by Kant to four of the original ten categories in Aristotle). The Jeffrey-style entitled Probability, desirability, and mode operators finds Grice at his formal-dress best. It predates the Kant lectures and it got into so much detail that Grice had to leave it at that. So abstract it hurts. Going further than Davidson, Grice argues that structures expressing probability and desirability are not merely analogous. They can both be replaced by more complex structures containing a common element. Generalising over attitudes using the symbol ψ, which he had used before, repr. WoW:v, Grice proposes G ψ that p. Further, Grice uses i as a dummy for sub-divisions of psychological attitudes. Grice uses Op supra i sub α, read: operation supra i sub alpha, as Grice was fastidious enough to provide reading versions for these, and where α is a dummy taking the place of either A or B, i. e. Davidsons prima facie or desirably, and probably. In all this, Grice keeps using the primitive !, where a more detailed symbolism would have it correspond exactly to Freges composite turnstile (horizontal stroke of thought and vertical stroke of assertoric force, Urteilstrich) that Grice of course also uses, and for which it is proposed, then: !─p. There are generalising movements here but also merely specificatory ones. α is not generalised. α is a dummy to serve as a blanket for this or that specifications. On the other hand, ψ is indeed generalised. As for i, is it generalising or specificatory? i is a dummy for specifications, so it is not really generalising. But Grice generalises over specifications. Grice wants to find buletic, boulomaic or volitive as he prefers when he does not prefer the Greek root for both his protreptic and exhibitive versions (operator supra exhibitive, autophoric, and operator supra protreptic, or hetero-phoric). Note that Grice (WoW:110) uses the asterisk * as a dummy for either assertoric, i.e., Freges turnstile, and non-assertoric, the !─ the imperative turnstile, if you wish. The operators A are not mode operators; they are such that they represent some degree (d) or measure of acceptability or justification. Grice prefers acceptability because it connects with accepting that which is a psychological, souly attitude, if a general one. Thus, Grice wants to have It is desirable that p and It is believable that p as understood, each, by the concatenation of three elements. The first element is the A-type operator. The second element is the protreptic-type operator. The third element is the phrastic, root, content, or proposition itself. It is desirable that p and It is believable that p share the utterer-oriented-type operator and the neustic or proposition. They only differ at the protreptic-type operator (buletic/volitive/boulomaic or judicative/doxastic). Grice uses + for concatenation, but it is best to use ^, just to echo who knows who. Grice speaks in that mimeo (which he delivers in Texas, and is known as Grices Performadillo talk ‒ Armadillo + Performative) of various things. Grice speaks, transparently enough, of acceptance: V-acceptance and J-acceptance. V not for Victory but for volitional, and J for judicative. The fact that both end with -acceptance would accept you to believe that both are forms of acceptance. Grice irritatingly uses 1 to mean the doxastic, and 2 to mean the bulematic. At Princeton in Method, he defines the doxastic in terms of the buletic and cares to do otherwise, i. e. define the buletic in terms of the doxastic. So whenever he wrote buletic read doxastic, and vice versa. One may omits this arithmetic when reporting on Grices use. Grice uses two further numerals, though: 3 and 4. These, one may decipher – one finds oneself as an archeologist in Tutankamons burial ground, as this or that relexive attitude. Thus, 3, i. e. ψ3, where we need the general operator ψ, not just specificatory dummy, but the idea that we accept something simpliciter. ψ3 stands for the attitude of buletically accepting an or utterance: doxastically accepting that p or doxastically accepting that ~p. Why we should be concerned with ~p is something to consider.  G wants to decide whether to believe p or not. I find that very Griceian. Suppose I am told that there is a volcano in Iceland. Why would I not want to believe it? It seems that one may want to decide whether to believe p or not when p involves a tacit appeal to value. But, as Grice notes, even when it does not involve value, Grice still needs trust and volition to reign supreme. On the other hand, theres 4, as attached to an attitude, ψ4. This stands for an attitude of buletically accepting an or utterance: buletically accepting that p, or G buletically accepting that ~p, i. e. G wants to decide whether to will, now that p or not. This indeed is crucial, since, for Grice, morality, as with Kantotle, does cash in desire, the buletic. Grice smokes. He wills to smoke. But does he will to will to smoke? Possibly yes. Does he will to will to will to smoke? Regardless of what Grice wills, one may claim this holds for a serious imperatives (not Thou shalt not reek, but Thou shalt not kill, say) or for any p if you must (because if you know that p causes cancer (p stands for a proposition involving cigarette) you should know you are killing yourself. But then time also kills, so what gives? So I would submit that, for Kant, the categoric imperative is one which allows for an indefinite chain, not of chain-smokers, but of good-willers. If, for some p, we find that at some stage, the P does not will that he wills that he wills that he wills that, p can not be universalisable. This is proposed in an essay referred to in The Philosophers Index but Marlboro Cigarettes took no notice. One may go on to note Grices obsession on make believe. If I say, I utter expression e because the utterer wants his addressee to believe that the utterer believes that p, there is utterer and addresse, i. e. there are two people here  ‒ or any soul-endowed creature  ‒ for Grices squarrel means things to Grice. It even implicates. It miaows to me while I was in bed. He utters miaow. He means that he is hungry, he means (via implicatum) that he wants a nut (as provided by me). On another occasion he miaowes explicating, The door is closed, and implicating Open it, idiot. On the other hand, an Andy-Capps cartoon read: When budgies get sarcastic Wild-life programmes are repeating One may note that one can want some other person to hold an attitude. Grice uses U or G1 for utterer and A or G2 for addressee. These are merely roles. The important formalism is indeed G1 and G2. G1 is a Griceish utterer-person; G2 is the other person, G1s addressee. Grice dislikes a menage a trois, apparently, for he seldom symbolises a third party, G3. So, G ψ-3-A that p is 1 just in case G ψ2(G ψ1 that p) or G ψ1 that ~p is 1. And here the utterers addressee, G2 features: G1 ψ³ protreptically that p is 1 just in case G buletically accepts ψ² (G buletically accepts ψ² (G doxastically accepts ψ1 that p, or G doxastically accepts ψ1 that ~p))) is 1. Grice seems to be happy with having reached four sets of operators, corresponding to four sets of propositional attitudes, and for which Grice provides the paraphrases. The first set is the doxastic proper. It is what Grice has as doxastic,and which is, strictly, either indicative, of the utterers doxastic, exhibitive state, as it were, or properly informative, if addressed to the addressee A, which is different from U himself, for surely one rarely informs oneself. The second is the buletic proper. What Grice dubs volitive, but sometimes he prefers the Grecian root. This is again either self- or utterer-addressed, or utterer-oriented, or auto-phoric, and it is intentional, or it is other-addressed, or addressee-addressed, or addressee- oriented, or hetero-phoric, and it is imperative, for surely one may not always say to oneself, Dont smoke, idiot!. The third is the doxastic-interrogative, or doxastic-erotetic. One may expand on ? here is minimal compared to the vagaries of what I called the !─ (non-doxastic or buletic turnstile), and which may be symbolised by ?─p, where ?─ stands for the erotetic turnstile. Geachs and Althams erotetic somehow Grice ignores, as he more often uses the Latinate interrogative. Lewis and Short have “interrŏgātĭo,” which they render as “a questioning, inquiry, examination, interrogation;” “sententia per interrogationem, Quint. 8, 5, 5; instare interrogation; testium; insidiosa; litteris inclusæ; verbis obligatio fit ex interrogatione et responsione; as rhet. fig., Quint. 9, 2, 15; 9, 3, 98. B. A syllogism: recte genus hoc interrogationis ignavum ac iners nominatum est, Cic. Fat. 13; Sen. Ep. 87 med. Surely more people know what interrogative means what erotetic means, he would not say ‒ but he would. This attitude comes again in two varieties: self-addressed or utterer-oriented, reflective (Should I go?) or again, addresee-addressed, or addressee-oriented, imperative, as in Should you go?, with a strong hint that the utterer is expecting is addressee to make up his mind in the proceeding, not just inform the utterer. Last but not least, there is the fourth kind, the buletic-cum-erotetic. Here again, there is one varietiy which is reflective, autophoric, as Grice prefers, utterer-addressed, or utterer-oriented, or inquisitive (for which Ill think of a Greek pantomime), or addressee-addressed, or addressee-oriented. Grice regrets that Greek (and Latin, of which he had less ‒ cfr. Shakespeare who had none) fares better in this respect the Oxonian that would please Austen, if not Austin, or Maucalay, and certainly not Urquhart -- his language has twelve parts of speech: each declinable in eleven cases, four numbers, eleven genders (including god, goddess, man, woman, animal, etc.); and conjugable in eleven tenses, seven moods, and four voices.These vocal mannerisms will result in the production of some pretty barbarous English sentences; but we must remember that what I shall be trying to do, in uttering such sentences, will be to represent supposedly underlying structure; if that is ones aim, one can hardly expect that ones speech-forms will be such as to excite the approval of, let us say, Jane Austen or Lord Macaulay. Cf. the quessertive, or quessertion, possibly iterable, that Grice cherished. But then you cant have everything. Where would you put it? Grice: The modal implicatum. Grice sees two different, though connected questions about mode. First, there is the obvious demand for a characterisation, or partial characterisation, of this or that mode as it emerges in this or that conversational move, which is plausible to regard as modes primary habitat) both at the level of the explicatum or the implicatum, for surely an indicative conversational move may be the vehicle of an imperatival implicatum. A second, question is how, and to what extent, the representation of mode (Hares neustic) which is suitable for application to this or that conversational move may be legitimately exported into philosophical psychology, or rather, may be grounded on questions of philosophical psychology, matters of this or that psychological state, stance, or attitude (notably desire and belief, and their species). We need to consider the second question, the philosophico- psychological question, since, if the general rationality operator is to read as something like acceptability, as in U accepts, or A accepts, the appearance of this or that mode within its scope of accepting is proper only if it may properly occur within the scope of a generic psychological verb I accept that . Lewis and Short have “accepto,” “v. freq. a. accipio,” which Short and Lewis render as “to take, receive, accept,” “argentum,” Plaut. Ps. 2, 2, 32; so Quint. 12, 7, 9; Curt. 4, 6, 5; Dig. 34, 1, 9: “jugum,” to submit to, Sil. Ital. 7, 41. But in Plin. 36, 25, 64, the correct read. is coeptavere; v. Sillig. a. h. l. The easiest way Grice finds to expound his ideas on the first question is by reference to a schematic table or diagram (Some have complained that I seldom use a board, but I will today. Grice at this point reiterates his temporary contempt for the use/mention distinction, which which Strawson is obsessed. Perhaps Grices contempt is due to Strawsons obsession. Grices exposition would make the hair stand on end in the soul of a person especially sensitive in this area. And Im talking to you, Sir Peter! (He is on the second row). But Grices guess is that the only historical philosophical mistake properly attributable to use/mention confusion is Russells argument against Frege in On denoting, and that there is virtually always an acceptable way of eliminating disregard of the use-mention distinction in a particular case, though the substitutes are usually lengthy, obscure, and tedious. Grice makes three initial assumptions. He avails himself of two species of acceptance, Namesly, volitive acceptance and judicative acceptance, which he, on occasion, calls respectively willing that p and willing that p.  These are to be thought of as technical or semi-technical, theoretical or semi-theoretical, though each is a state which approximates to what we vulgarly call thinking that p and wanting that p, especially in the way in which we can speak of a beast such as a little squarrel as thinking or wanting something  ‒ a nut, poor darling little thing. Grice here treats each will and judge (and accept) as a primitive. The proper interpretation would be determined by the role of each in a folk-psychological theory (or sequence of folk-psychological theories), of the type the Wilde reader in mental philosophy favours at Oxford, designed to account for the behaviours of members of the animal kingdom, at different levels of psychological complexity (some classes of creatures being more complex than others, of course). As Grice suggests in Us meaning, sentence-meaning, and word-meaning, at least at the point at which (Schema Of Procedure-Specifiers For Mood-Operators) in ones syntactico-semantical theory of Pirotese or Griceish, one is introducing this or that mode (and possibly earlier), the proper form to use is a specifier for this or that resultant procedure. Such a specifier is of the general form, For the utterer U to utter x if C, where the blank is replaced by the appropriate condition. Since in the preceding scheme x represents an utterance or expression, and not a sentence or open sentence, there is no guarantee that this or that actual sentence in Pirotese or Griceish contains a perspicuous and unambiguous modal representation. A sentence may correspond to more than one modal structure. The sentence is structurally ambiguous (multiplex in meaning  ‒ under the proviso that senses are not to be multiplied beyond necessity) and will have more than one reading, or parsing, as every schoolboy at Clifton knows when translating viva voce from Greek or Latin, as the case might be. The general form of a procedure-specifier for a modal operator involves a main clause and an antecedent clause, which follows if. In the schematic representation of the main clause, U represents an utterer, A his addressee, p the radix or neustic; and Opi represents that operator whose number is i (1, 2, 3, or 4), e.g g., Op3A represents Operator 3A, which, since ?⊢ appears in the Operator column for 3A) would be ?A ⊢ p. This reminds one of Grandys quessertions, for he did think they were iterable (possibly)). The antecedent clause consists of a sequence whose elements are a preamble, as it were, or preface, or prefix, a supplement to a differential (which is present only in a B-type, or addressee-oriented case), a differential, and a radix. The preamble, which is always present, is invariant, and reads: The U U wills (that) A A judges (that) U  (For surely meaning is a species of intending is a species of willing that, alla Prichard, Whites professor, Corpus). The supplement, if present, is also invariant. And the idea behind its varying presence or absence is connected, in the first instance, with the volitive mode. The difference between an ordinary expression of intention  ‒ such as I shall not fail, or They shall not pass  ‒  and an ordinary imperative (Like Be a little kinder to him) is accommodated by treating each as a sub-mode of the volitive mode, relates to willing that p) In the intentional case (I shall not fail), the utterer U is concerned to reveal to his addressee A that he (the utterer U) wills that p. In the imperative case (They shall not pass), the utterer U is concerned to reveal to his addressee A that the utterer U wills that the addresee A will that p.  In each case, of course, it is to be presumed that willing that p will have its standard outcome, viz., the actualization, or realisation, or direction of fit, of the radix (from expression to world, downwards). There is a corresponding distinction between two uses of an indicative. The utterer U may be declaring or affirming that p, in an exhibitive way, with the primary intention to get his addressee A to judge that the utterer judges that p. Or the U is telling (in a protreptic way) ones addressee that p, that is to say, hoping to get his addressee to judge that p. In the case of an indicative, unlike that of a volitive, there is no explicit pair of devices which would ordinarily be thought of as sub-mode marker. The recognition of the sub-mode is implicated, and comes from context, from the vocative use of the Names of the addressee, from the presence of a speech-act verb, or from a sentence-adverbial phrase (like for your information, so that you know, etc.). But Grice has already, in his initial assumptions, allowed for such a situation. The exhibitive-protreptic distinction or autophoric-heterophoric distinction, seems to Grice to be also discernible in the interrogative mode (?). Each differentials is associated with, and serve to distinguish, each of the two basic modes (volitive or judicative) and, apart from one detail in the case of the interrogative mode, is invariant between autophoric-exhibitive) and heterophoric-protreptic sub-modes of any of the two basic modes. They are merely unsupplemented or supplemented, the former for an exhibitive sub-mode and the latter for a protreptic sub-mode. The radix needs (one hopes) no further explanation, except that it might be useful to bear in mind that Grice does not stipulated that the radix for an intentional (buletic exhibitive utterer-based) incorporate a reference to the utterer, or be in the first person, nor that the radix for an imperative (buletic protreptic addressee-based) incorporate a reference of the addresee, and be in the second person. They shall not pass is a legitimate intentional, as is You shall not get away with it; and The sergeant is to muster the men at dawn, as uttered said by the captain to the lieutenant) is a perfectly good imperative. Grice gives in full the two specifiers derived from the schema. U to utter to A autophoric-exhibitive ⊢ p if U wills that A judges that U judges p. Again, U to utter to A ! heterophoric-protreptic p if U wills that A A judges that U wills that A wills that p. Since, of the states denoted by each differential, only willing that p and judging that p are strictly cases of accepting that p, and Grices ultimate purpose of his introducing this characterization of mode is to reach a general account of expressions which are to be conjoined, according to his proposal, with an acceptability operator, the first two numbered rows of the figure are (at most) what he has a direct use for. But since it is of some importance to Grice that his treatment of mode should be (and should be thought to be) on the right lines, he adds a partial account of the interrogative mode. There are two varieties of interrogatives, a yes/no interrogatives (e. g. Is his face clean? Is the king of France bald? Is virtue a fire-shovel?) and x-interrogatives, on which Grice qua philosopher was particularly interested, v. his The that and the why.  (Who killed Cock Robin?, Where has my beloved gone?, How did he fix it?). The specifiers derivable from the schema provide only for yes/no interrogatives, though the figure could be quite easily amended so as to yield a restricted but very large class of x-interrogatives. Grice indicates how this could be done. The distinction between a buletic and a doxastic interrogative corresponds with the difference between a case in which the utterer U indicates that he is, in one way or another, concerned to obtain information (Is he at home?), and a case in which the utterer U indicates that he is concerned to settle a problem about what he is to do ‒ Am I to leave the door open?, Shall I go on reading? or, with an heterophoric Subjects, Is the prisoner to be released? This difference is fairly well represented in grammar, and much better represented in the grammars of some other languages. The hetero-phoric-cum-protreptic/auto-phoric-cum- exhibitive difference may not marked at all in this or that grammar, but it should be marked in Pirotese. This or that sub-mode is, however, often quite easily detectable. There is usually a recognizable difference between a case in which the utterer A says, musingly or reflectively, Is he to be trusted?  ‒ a case in which the utterer might say that he is just wondering  ‒ and a case in which he utters a token of the same sentence as an enquiry. Similarly, one can usually tell whether an utterer A who utters Shall I accept the invitation?  is just trying to make up his mind, or is trying to get advice or instruction from his addressee. The employment of the variable α needs to be explained. Grice borrows a little from an obscure branch of logic, once (but maybe no longer) practised, called, Grice thinks, proto-thetic ‒ Why? Because it deals with this or that first principle or axiom, or thesis), the main rite in which is to quantify over, or through, this or that connective. α is to have as its two substituents positively and negatively, which may modify either will or judge, negatively willing or negatively judging that p is judging or willing that ~p. The quantifier (∃1α) . . . has to be treated substitutionally. If, for example, I ask someone whether John killed Cock Robin (protreptic case), I do not want the addressee merely to will that I have a particular logical quality in mind which I believe to apply. I want the addressee to have one of the Qualities in mind which he wants me to believe to apply. To meet this demand, supplementation must drag back the quantifier. To extend the schema so as to provide specifiers for a single x-interrogative (i. e., a question like What did the butler see? rather than a question like Who went where with whom at 4 oclock yesterday afternoon?), we need just a little extra apparatus. We need to be able to superscribe a W in each interrogative operator e.g., together with the proviso that a radix which follows a superscribed operator must be an open radix, which contains one or more occurrences of just one free variable. And we need a chameleon variable λ, to occur only in this or that quantifier. (∃λ).Fx is to be regarded as a way of writing (∃x)Fx. (∃λ)Fy is a way of writing (∃y)Fy. To provide a specifier for a x-superscribed operator, we simply delete the appearances of α in the specifier for the corresponding un-superscribed operator, inserting instead the quantifier (∃1λ) () at the position previously occupied by (∃1α) (). E.g. the specifiers for Who killed Cock Robin?, used as an enquiry, would be: U to utter to A  killed Cock Robin if U wills A to judge U to will that (∃1λ) (A should will that U judges (x killed Cock Robin)); in which (∃1λ) takes on the shape (∃1x) since x is the free variable within its scope. Grice compares his buletic-doxastic distinction to prohairesis/doxa distinction by Aristotle in Ethica Nichomachea. Perhaps his simplest formalisation is via subscripts: I will-b but will-d not. Refs.: The main references are given above under ‘desirability.’ The most systematic treatment is the excursus in “Aspects,” Clarendon. BANC.

modified Occam’s razor: Grice loved a razor. The essay had circulated since the Harvard days, and it was also repr. in Pragmatics, ed. Cole for Academic Press. Personally, I prefer dialectica.  ‒ Grice. This is the third James lecture at Harvard. It is particularly useful for Grices introduction of his razor, M. O. R., or Modified Occams Razor, jocularly expressed by Grice as: Senses are not to be multiplied beyond necessity. An Englishing of the Ockhams Latinate, Entia non sunt multiplicanda præter necessitatem. But what do we mean sense. Surely Occam was right with his Entia non sunt multiplicanda præter necessitatem. We need to translate that alla linguistic turn. Grice jokes: Senses are not be multiplied beyond necessity. He also considers irony, stress (supra-segmental fourth-articulatory phonology), and truth, which the Grice Papers have under a special f. in the s. V . Three topics where the implicatum helps. He is a scoundrel may well be the implicatum of He is a fine friend. But cf. the pretense theory of irony. Grice, being a classicist, loved the etymological connection. With Stress, he was concerned with anti-Gettier uses of emphatic know: I KNOW. (Implicatum: I do have conclusive evidence). Truth (or  is true) sprang from the attention by Grice to that infamous Bristol symposium between Austin and Strawson. Cf. Moores paradox. Grice wants to defend correspondence theory of Austin against the performative approach of Strawson. If  is true implicates someone previously affirmed this, that does not mean a ditto implicatum is part of the entailment of a  is true utterance, further notes on logic and conversation, in Cole, repr. in a revised form, Modified Occams Razor, irony, stress, truth. The preferred citation should be the Harvard. This is originally the third James lecture, in a revised form.In that lecture, Grice introduced the M. O. R., or Modified Occams Razor. Senses are not be multiplied beyond necessity. The point is that entailment-cum-implicatum does the job that multiplied senses should not do! The Grice Papers contains in a different f. the concluding section for that lecture, on irony, stress, and truth. Grice went back to the Modified Occams razor, but was never able to formalise it! It is, as he concedes, almost a vacuous methodological thingy! It is interesting that the way he defines the alethic value of true alrady cites satisfactory. I shall use, to Names such a property, not true but factually satisfactory. Grices sympathies dont lie with Strawsons Ramsey-based redundance theory of truth, but rather with Tarskis theory of correspondence. He goes on to claim his trust in the feasibility of such a theory. It is, indeed, possible to construct a theory which treats truth as (primarily) a property, not true but factually satisfactory. One may see that point above as merely verbal and not involving any serious threat. Lets also assume that it will be a consequence, or theorem, of such a theory that there will be a class C of utterances (utterances of affirmative Subjects-predicate sentences [such as snow is white or the cat is on the mat of the dog is hairy-coated such that each member of C designates or refers to some item and indicates or predicates some class (these verbs to be explained within the theory), and is factually satisfactory if the item belongs to the class. Let us also assume that there can be a method of introducing a form of expression, it is true that /it is buletic that  and linking it with the notion of factually or alethic or doxastic satisfactory, a consequence of which will be that to say it is true that Smith is happy will be equivalent to saying that any utterance of class C which designates Smith and indicates the class of happy people is factually satisfactory (that is, any utterance which assigns Smith to the class of happy people is factually satisfactory. Mutatis mutandis for Let Smith be happy, and buletic satisfactoriness. The move is Tarskian. The two standard truth definitions are at first glance not definitions of truth at all, but definitions of a more complicated relation involving assignments of objects to variables: a satisfies the formula F, (where the symbol F is a placeholder for a Names of a particular formula of the object language). In fact satisfaction reduces to truth in this way: aa satisfies the formula FF if and only if taking each free variable in FF as a Names of the object assigned to it by aa makes the formula FF into a true sentence. So it follows that our intuitions about when a sentence is true can guide our intuitions about when an assignment satisfies a formula. But none of this can enter into the formal definition of truth, because taking a variable as a Names of an object is a semantic notion, and Tarskis truth definition has to be built only on notions from syntax and set theory (together with those in the object-language); In fact Tarskis reduction goes in the other direction: if the formula FF has no free variables, to say that FF is true is to say that every assignment satisfies it. The reason why Tarski defines satisfaction directly, and then deduces a definition of truth, is that satisfaction obeys recursive conditions in the following way. If FF is a compound formula, to know which assignments satisfy FF, its enough to know which assignments satisfy the immediate constituents of FF. Here are two typical examples: The assignment a satisfies the formula F and GG if and only if aa satisfies FF and aa satisfies GG. The assignment aa satisfies the formula For all xx, GG if and only if for every individual ii, if bb is the assignment that assigns ii to the variable xx and is otherwise exactly like aa, then bb satisfies GG. We have to use a different approach for atomic formulas. But for these, at least assuming for simplicity that LL has no function symbols, we can use the metalanguage copies #(R)#(R) of the predicate symbols RR of the object language. Thus The assignment aa SATSIFIES the formula R(x,y)R(x,y) if and only if #(R)(a(x),a(y))#(R)(a(x),a(y)). Warning: the expression ## is in the meta-meta-language, not in the meta-language MM. We may or may not be able to find a formula of MM that expresses ## for predicate symbols; it depends on exactly what the language LL is.). Subjects to this or that mild reservation, Tarskis definition of satisfaction is compositional, meaning that the class of assignments which satisfy a compound formula FF is determined solely by the syntactic rule used to construct FF from its immediate constituents and the classes of assignments that satisfy these immediate constituents. This is sometimes phrased loosely by saying that satisfaction is defined recursively. But this formulation misses the central point, that the above do not contain any syntactic information about the immediate constituents. Compositionality explains why Tarski switches from true to satisfied. You cant define whether For all x,Gx,G is true in terms of whether GG is true, because in general GG has a free variable xx and so it isnt either true or false. The reservation is that Tarskis definition of satisfaction in Tarskis essay doesnt in fact mention the class of assignments that satisfy a formula FF. Instead, as we saw, he defines the relation aa satisfies FF, which determines what that class is. This is probably the main reason why some people (including Tarski himself in conversation have preferred NOT to describe the definition as compositional. But the class format, which is compositional on any reckoning, does appear in an early variant of the truth definition in Tarskis essay on definable sets of real numbers. Tarski had a good reason for preferring the format aa satisfies FF in his essay, viz. that it allowed him to reduce the set-theoretic requirements of the truth definition. He spells out these requirements carefully. Compositionality first appears in an essay by Putnam. In talking about compositionality, we have moved to thinking of Tarskis definition as a semantics, i.e. a way of assigning meanings to formulas. Here we take the meaning of a sentence to be its truth value. Compositionality means essentially that the meanings assigned to formulas give at least enough information to determine the truth values of sentences containing them. One can ask conversely whether Tarskis semantics provides only as much information as we need about each formula, in order to reach the truth values of sentences. If the answer is yes, we say that the semantics is fully abstract (for truth). One can show fairly easily, for any of the standard languages of logic, that Tarskis definition of satisfaction is in fact fully abstract. As it stands, Tarskis definition of satisfaction is not an explicit definition, because satisfaction for one formula is defined in terms of satisfaction for other formulas. So to show that it is formally correct, we need a way of converting it to an explicit definition. One way to do this is as follows, using either higher order logic or set theory. Suppose we write SS for a binary relation between assignments and formulas. We say that SS is a satisfaction relation if for every formula G, SG, S meets the conditions put for satisfaction of GG by Tarskis definition. E.g., if GG is G1G1 and G2G2, SS should satisfy the following condition for every assignment aa: S(a, G) if and only if S(a, G1) and S(a, G2).S (a, G) if and only if S(a, G1) and S(a, G2). We can define satisfaction relation formally, using the recursive clauses and the conditions for atomic formulas in Tarskis recursive definition. Now we prove, by induction on the complexity of formulas, that there is exactly one satisfaction relation SS. There are some technical subtleties, but it can be done. Finally we define aa satisfies FF if and only if: there is a satisfaction relation SS such that S (a, F) S (a, F). It is then a technical exercise to show that this definition of satisfaction is materially adequate. Actually one must first write out the counterpart of Convention TT for satisfaction of formulas, but I leave this to the reader. The remaining truth definition in Tarskis essay  – the third as they appear in the essay – is really a bundle of related truth definitions, all for the same object-language LL but in different interpretations. The quantifiers of LL are assumed to range over a particular class, call it AA; in fact they are second order quantifiers, so that really they range over the collection of subclasses of AA. The class AA is not Namesd explicitly in the object language, and thus one can give separate truth definitions for different values of AA, as Tarski proceeds to do. So for this section of the paper, Tarski allows one and the same sentence to be given different interpretations; this is the exception to the general claim that his object language sentences are fully interpreted. But Tarski stays on the straight and narrow: he talks about truth only in the special case where AA is the class of all individuals. For other values of AA, he speaks not of truth but of correctness in the domain AA. These truth or correctness definitions dont fall out of a definition of satisfaction. In fact they go by a much less direct route, which Tarski describes as a purely accidental possibility that relies on the specific peculiarities of the particular object language. There is no hope of giving a definition of satisfaction by recursion on the complexity of formulas. The remedy is to note that the explicit form of Tarskis truth definition in Section 2.1 above didnt require a recursive definition; it needed only that the conditions on the satisfaction relation SS pin it down uniquely. For Henkins first style of language this is still true, though the reason is no longer the well-foundedness of the syntax. For Henkins second style of language, at least in the notation by Hintikka (independence friendly logic), the syntax is well-founded, but the displacement of the quantifier scopes means that the usual quantifier clauses in the definition of satisfaction no longer work. How can we analyze satisfaction? The answer to this question is in some ways reminiscent to our answer of how to construct a theory of truth for a language with only finitely many sentences. So see how, first suppose that our language has only three Namess and three predicates, Bob, Jane, and Nancy and is nice, is mean, and is lazy. We can then give the following analysis, indeed, definition, of satisfaction. An item I satisfies predicate p ≡df [(p=is nice and i is nice) ∨ (n=is mean and i is mean) ∨ (n=is lazy and i is lazy)] There is an analogy between the material-adequacy constraint which Tarski set on the theory of truth, and similar constraints which we should expect a definition of satisfaction to meet. Just as a theory of truth should imply every instance of S is 1 in L iff S so we should expect our theory of satisfaction to imply every instance of the following schema. I satisfies is F in L iff i is F. We arrive at Tarskis first-order definition as a definition of satisfaction which makes no use of concepts other than those employed in the object-language itself. By stress, Grice means suprasegmental phonology, but he was too much of a philosopher to let that jargon affect him! Refs.: The locus classicus, if that does not sound too pretentious, is Essay 3 in WoW, but there are references elsewhere, such as in “Meaning Revisited,” and under ‘semantics.’

myth: Grice knows a little about Descartess “Discours de la methode,” and he is also aware of similar obsession by Collingwood with philosopical methodology. Grice would joke on midwifery, as the philosopher’s apter method at Oxford: to strangle error at its birth. Grice typifies a generation at Oxford. While he did not socialize with the crème de la crème in pre-war Oxford, he shared some their approach. E.g. a love affair with Russell’s logical construction. After the war, and in retrospect, Grice liked to associate himself with Austin. He obviously felt the need to belong to a group, to make a difference, to make history. Many participants of the play group saw themselves as doing philosophy, rather than reading about it! It was long after that Grice started to note the differences in methodology between Austin and himself. His methodology changed a little. He was enamoured with formalism for a while, and he grants that this love never ceased. In a still later phase, he came to realise that his way of doing philosophy was part of literature (essay writing). And so he started to be slightly more careful about his style – which some found florid. The stylistic concerns were serious. Oxonian philosophers like Holloway had been kept away from philosophy because of the stereotype that the Oxonian philosophers style is pedantic, when it neednt! A philosopher should be allowed, as Plato was, to use a myth, if he thinks his tutee will thank him for that! Grice loved to compare his Oxonian dialectic with Platos Athenian (strictly, Academic) dialectic. Indeed, there is some resemblance of the use of myth in Plato and Grice for philosophical methodological purposes. Grice especially enjoys a myth in his programme in philosophical psychology. In this, he is very much being a philosopher. Non-philosophers usually criticise this methodological use of a myth, but they would, wouldnt they. Grice suggests that a myth has diagogic relevance. Creature construction, the philosopher as demi-god, if mythical, is an easier way for a philosophy don to instil his ideas on his tutee than, say, privileged access and incorrigibility. Refs.: The main source is Grice’s essay on ‘myth’, in The H. P. Grice Papers, BANC.