What does lnb drift mean recognized


modified Occam’s razor: Grice was obsessed with sense, ’and thought Oxonian philosophers were multiplying it otiosely - notably L. J. Cohen (“ The diversity of meaning ”). The original razor is what Grice would have as ‘ontological,’ to which he opposes with in his ontological marxism ’. Entities should not be multiplied beyond the necessity of needing them as honest working entities. He keeps open house provided they come in help with the work. This restriction explains what Grice means by "necessity" in the third lecture - a second sense does not do any work. The implicatum does. Grice loved a razor, and being into analogy and focal meaning, if he HAD to have semantic multiplicity, for the case of 'is,' (being) or 'good,' it had to be a UNIFIED semantic multiplicity, as displayed by paronymy . 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) jumped from the attention by Grice to that infamous Bristol symposium between Austin and Strawson. Cf. Moore's 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 formalize 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 Strawson's 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 mutan dis for Let Smith be happy, and buletic satisfactoriness. The move is Tarskian. TBy 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.' The only one who took up Grice's challenge at Oxford was LJ Cohen, "Grice on the particles of natural language," which got a great response by Oxonian RCS Walker (citing D. Bostock, a tutee of Grice), to which Cohen again responded "Can the conversationalist hypothesis be defended." Cohen clearly centers his criticism on the razor. He had an early essay, citing Grice, on the DIVERSITY of meaning. Cohen opposes Grice’s conversationalist hypothesis to his own ‘semantic hypothesis’ (“Multiply senses all you want.”). T. D. Bontly explores the topic of Grice’s MOR. “Ancestors of this essay were presented at meetings of the Society for Philosophy and Psychology (Edmonton, Alberta), of the Pacific Division of the American Philosophical Association (San Francisco, CA) and at the University of Connecticut. I am indebted to all three groups and particularly to the commentators D. Sanford (at the Society for Philosophy and Psychology) and M. Reimer (at the APA). Thanks also to the following for helpful comments or discussion (inclusive): F. Adams, A. Ariew, P. Bloom, M. Devitt, B. Enc, C. Gaulker, M. Lynch, R. Millikan, J. Pust, E. Sober, RC Stalnaker, DW Stampe, and S. Wheeler. "Bontly writes, more or less (I have paraphrased him a little, with good intentions, always!)" Some philosophers have appealed to a principle which HP Grice, in his third William James lecture, dubs Modified Occam's Razor (henceforth, “MOR”): “Senses - rather than 'entities,' as the inceptor from Ockham more boringly has it - are not to be multiplied beyond necessity '.” What is 'necessity'? Bontly: “Superficially, Grice's“ M. O. R. " seems a routine application of Ockham’s principle of parsimony: entities are not to be multiplied beyond necessity. Now, parsimony arguments, though common in science, are notoriously problematic, and their use by Grice faces one objection or two. Grice’s "M. O. R. " makes considerably more sense in light of certain assumptions about the psychological processes involved in language development, learning, and acquisition, and it describes recent * empirical *, if not philosophical or conceptual, of the type Grice seems mainly interested in - findings that bear these assumptions out. [My] resulting account solves several difficulties that otherwise confront Grice’s "M. OR ”, and it draws attention to problematic assumptions involved in using parsimony to argue for pragmatic accounts of the type of phenomena 'ordinary-language' philosophers were interested in. In more general terms, when an expression E has two or more uses - U1 and U2, say - enabling its users to express two or more different meanings - M1 and M2, say - one is tempted to assume that E is semantically (ie lexically) ambiguous, or polysemous, ie, that some convention, constituting the language L, assign E these two meanings M1 and M2 corresponding to its two uses U1 and U2. One hears, for instance, that 'or' is ambiguous (polysemous) between a weak (inclusive) ('pv q') and a strong (exclusive) sense, 'pw q.'Grice actually feels that speaking of the meaning or sense of 'or' sounds harsh (“Like if I were asked what the meaning of 'to' is!”). But in one note from a seminar from Strawson he writes: "Jones is between Smith and Williams." “I wouldn't say that 'between' is ambiguous, even if we interpret the sentence in a physical sense, or in an ordering of merit, say.” Bontly: “Used exclusively, an utterance of 'p or q' (pvq ) entails that 'p' and 'q' are NOT both true. Used inclusively, it does not. Still, ambiguity is not the only possible explanation. " (This reminds me of Atlas, “Philosophy WITHOUT ambiguity!” - ambitious title!). The phenomenon can also be approached pragmatically, from within the framework of a general theory conversation alla Grice. One could, e. g., first, maintain that 'p or q' is unambiguously monosemous inclusive and, second, apply Grice's idea of ​​an 'implicatum' to explain the exclusive. ”I actually traced this, and found that OP Wood in an odd review of a logic textbook (by Faris) in “Mind,” in the 1950s, makes the point about the inclusive-exclusive distinction, pre-Griceianly! Grice seems more interested, as you later consider, the implicatum: “Utterer U has non-truth-functional grounds for uttering‘ p or q. Not really the 'inclusive-exclusive' distinction. Jennings deals with this in “The genealogy of disjunction,” and elsewhere, and indeed notes that 'or' may be a dead metaphor from 'another.' Bontly goes on: “On any such account, 'p or q' would have two uses U1 and U2 and two standard interpretations, I1 and I2, but NEVER two 'conventional' meanings, ”M1 and M2 Or take 'and' (pq) which (when used as a sentential connective) ordinarily stands for truth-functional conjunction ( as in 1a, below). Often enough, though, "p and q" seems to imply temporal priority (1b), while in other cases it suggests causal priority (1c). (1) a. Bill bought a shirt and Christy bought a sweater. b. Adam took off his shoes and [he] got into bed. c. "Jack fell down and [he] broke his crown, and Jill came tumbling o: ter." (to rhyme with 'water' in an earlier line. ”Apparently Grice loved this nursery rhyme too,“ Jack is an Englishman; he must, therefore, be brave, ”Jill says.” (Grice, “Aspects of reason.”) Bontly: "Again, one suspects an ambiguity, M1 and M2, but Grice argues that a 'conversational' explanation is available and preferable. According to the 'pragmatist' or 'conversationalist' hypothesis' (as I shall call it), a temporal or a causal reading of “and” (pq) may be part of what the UTTERER means, but such a reading I2, are not part of what the sentence means, or the word _and_ means, and thus belong in a general theory of conversation , not the grammar of a specific language. ”Oddly, I once noticed that Chomsky, of all people, and since you speak of 'grammar,' competence, etc. refers to“ A. ”Albert? P. Grice in his 1966! Aspects of the theory of syntax. “AP Grice wants to say that the temporal succession is not part of the meaning of 'and.'” I suspect one of Grice's tutees at Oxford was spre ading the unauthorized word! Bontly: “Many an alleged ambiguity seems amenable to Grice’s conversationalist hypothesis. Besides the sentential connectives or truth-functors, a pragmatic explanation has been applied fruitfully to quantifiers (Grice lists 'all' and 'some (at least one'), definite descriptions (Grice lists 'the,' 'the murderer'), the indefinite description ('a finger', much discussed by Grice, “He's meeting a woman this evening.”), the genitive construction ('Peter's bat'), and the indirect speech act ('Can you pass the salt?') - to mention just a few. The literature on the Griceian treatment of these phenomena is extensive. Some classic treatments are found in the oeuvre of philosophers like Grice, Bach, Harnish, and Davis, and linguists like Horn, Gazdar, and Levinson. But the availability of a pragmatic explanation poses an interesting methodological problem. Prima facie, the alleged 'ambiguity' M1 and M2, can now be explained either semantically (by positing two or more senses S1 and S2, or M1 and M2, of expression E) or pragmatically (by positing just one sense (S) plus one super-impo sed implicature, I). Sometimes, of course, one approach or the other is transparently inadequate. When the ‘use’ of E cannot be derived from a general conversational principle, the pragmatic explanation seems a non-starter. " Not for a radically radical pragmatist like Atlas! Ambitious! Similarly, an ambiguity- or polysemy- based explanation seems out of the question where the interpretation of E at issue is highly context-dependent. " (My favorite is Grice on “a,” that you analyze in term of 'developmental' or ontogenetical pragmatics - versus Millikan's phylogenetical! But, in many cases, a semantic, or polysemy, and a conversational explanations both appear plausible, and the usual data - Grice's intuitions about how the expression can and cannot be used, should or shouldn't be used - appear to leave the choice of one of the two hypotheses under-determined. These were the cases that most interest Grice, the philosopher, since they impinge on various projects in philosophical analysis. (Cf. Grice, 1989, pp. 3-21 and passim). ”Notably the 'ordinary-language' philosophy 'project,' I would think. I love the fact that in the inventory of philosophers who are loose about this (as in the reference you mention above, pp. 3-21, he includes himself in “Causal theory of perception”! “To adjudicate these border-line cases, Grice (1978) proposes a methodological principle which he dubs “Modified Occam's Razor,” MOR ”'Senses are not to be multiplied beyond necessity. ’(1978, pp. 118-119)" (I follow Grice in using the Latinate ‘Occam’ rather than the Anglo-Saxon ‘Ockham’ which is currently preferred). More fully, the idea is that one should not posit an alleged special, stronger SENSE S2, for an expression E when a general conversational principle suffices to explain why E, which bears only Sense 1, S1, receives a certain interpretation or carries implicatum I. Thus, if the 'use' (or an 'use') of E can be explained pragmatically, other things being equal, the use should be explained pragmatically. ” Griceians appeal to M. O. R. quite often, "pragmatically bearded or not! (I love Quine's idea that Occam's razor was created to shave Plato's beard. Cfr. Schiffer's anti-shave! It is affirmed, in spirit if not letter, by philosopher / linguist Atlas and Levinson, philosopher / linguist Bach, Bach and philosopher Harnish, Horn, Levinson, Morgan, linguist / philosopher Neale, philosopher Searle, philosopher Stalnaker, philosopher Walker (of Oxford), and philosopher Ziff ”(I LOVE Ziff's use, seeing that he could be otherwise so anti-Griceian, vide Martinich,“ On Ziff on Grice on meaning, ”and indeed Stampe (that you mention) on Ziff on Grice on meaning. One particularly forceful statement is found in“ of all people ”Kripke, who derides the ambiguity hypothesis as 'the lazy man's approach in philosophy' and issues a strong warning. "When I read that, I was reminded that Stampe, in some unpublished manuscripts, deals with the loose use of Griceian ideas by Kripke. Stampe discusses at length," Let's get out of here, the cops are coming . ”Stampe thinks Kripke is only sup erficially a Griceian! Kripke: “‘ Do not posit an ambiguity unless you are really forced to, unless there are really compelling theoretical (or intuitive) grounds to suppose that an ambiguity really is present ’(1977, p. 20). A similar idea surfaces in Ruhl’s principle of "mono-semic" bias ". One’s initial effort is directed toward determining a UNITARY meaning S1 for a lexical item E, trying to attribute apparent variations (S2) in meaning to other factors. If such an effort fails, one tries to discover a means of relating the distinct meanings S1 and S2. If this effort fails, there are several words: E1 and E2 (1989, p. 4). " Grice’s ‘vice’ and Grice’s ‘vyse,’ different words in English, same in Old Roman (“violent.”). Ruhl’s position differs from Grice’s approach. Whereas Grice takes word-meaning to be its WEAKEST exhibited meaning, Ruhl argues that word-meaning can be so highly abstract or schematic as to provide only a CORE of meaning, making EVEN the weakest familiar reading a pragmatic specialization. ”Loved that! Ruhl as more Griceian than Grice! Indeed, Grice is freely using the very abstract notion of a Fregeian 'sense,' with the delicacy you would treat a brick! "The difference between Grice's and Ruhl's positions raises issues beyond the scope of the present essay (though see Atlas, 1989, for further discussion). " I will! Atlas knows everything you wanted to know, and more, especially when it comes to linguists! He has a later book with 'implicature' in its subtitle. "Considering the central role that" M. O.R. " plays in Grice's programs, one is thus surprised to find barely any attention paid to whether it is a good principle - to whether it is true that a pragmatic explanation, when available, is in general more likely to be true than its 'ambiguity' or polysemy, or bi-semy, or aequi-vocal rival. ”Trying to play with this, I see that Grice loves' aequi-vocal. 'He thinks that' must 'is' aequi-vocal' between an alethic and a practical ' use. 'It took me some time to process that! He means that since it's the ‘same,’ ‘aequi’, ‘voice’, vox ’. So ‘aequi-vocal’ IS ‘uni-vocal.’ The Aristotelian in Grice, I guess! "Grice himself offers vanishingly little argument." How extended is a Harvard philosophical audience’s attention-span? “Examining just two (out of the blue, unphilosophical) cases where we seem happy to attribute a secondary or derivative sense S2 to one word or expression E, but not another, Grice notes that, in both cases, the supposition that the expression E has an additional sense S2 is not superfluous, or unparsimonious, accounting for certain facets of the use of E that cannot, apparently, be explained pragmatically. " I wonder if a radically radical pragmatist would agree! I never met a polysemous expression! Grice concludes that, therefore, ‘there is as yet no reason NOT to accept M. O. R.’ (1978, p. 120) - faint praise for a principle so important to his philosophical program! Besides this weak argument for “M. O. R., "Grice (1978) also mentions a few independent, rather loose, tests for alleged ambiguity." (“And how to fail them,” as Zwicky would have it!) But Grice’s rationale for “M. OR, "presumably, is a thought Grice does not bother to articulate, thinking perhaps that the principle's name, its kinship with Occam's famous razor, 'Do not multiply entities beyond necessity,' made its epistemic credentials sufficiently obvious already." Plus, Harvard is very Occamist! “To lay it out, though, the thought is surely that parsimony - and other such qualities as simplicity, generality, and unification - are always prized in scientific (and philosophical?) explanation, the more parsimonious (etc .) of two otherwise equally adequate theories being ipso facto more likely to be true. If, as would seem to be the case, a pragmatic explanation were more parsimonious than its semantic, or 'conventionalist,' or ambiguity, or polysemic, or polysemy or bi-semic rival, the conversational explanation would be supported by an established, received , general principle of scientific inference. ”I love your exploration of Newton on this below! Hypotheses non fingo! “Certainly, some such argument is on Grice's mind when he names his principle as he does, and much the same thought surely lies behind Kripke's references to 'general methodological considerations' and 'considerations of economy' Other 'Griceian' appeals to these theoretical virtues are even more transparent. Linguist JL Morgan tells us, for instance, that 'Occam's Razor dictates that we take a Gricean account of an indirect speech act as the correct analysis, lacking strong evidence to the contrary' Philosopher Stalnaker argues that a major advantage accrues to a pragmatic treatment of Strawson's presupposition in that 'there will then be no need to complicate the semantics or the lexicon' ”or introduce metaphysically dubious truth-value gaps! Linguist S. C. Levinson suggests that a major selling point for a conversational theory in general is that such a theory promises to ‘effect a radical simplification of the semantics’ and ‘approximately halve the size of the lexicon’. " So we don't need to learn two words, 'vyse' and 'vice.' There can be little doubt, therefore, that a Griceian takes parsimony to argue for the pragmatic approach. ”I use the rather pedantic and awful spelling“ Griceian , ”So that I can keep the pronunciation / grais / and also because Fodor used it! And non-philosophers, too! "But a parsimony argument is notoriously problematic, and the argument for" M. O. R. " is no exception. The preference for a parsimonious theory is surprisingly difficult to justify, as is the assumption that a pragmatic explanation IS more parsimonious. This does not mean Grice’s "M. O. R. " is entirely without merit. On the contrary, Grice is right to hold that senses should not be multiplied, if a conversational principle will do. " But the justification for M. O. R. need have nothing to do with the idea that parsimony is, always and everywhere, a virtue in scientific theories. " Also because we are dealing with philosophy, not science, here? What makes Grice’s "M. O. R. " reasonable, rather, is a set of assumptions about the psychological processes involved in language learning, development, and acquisition, and I will report some empirical (rather than conceptual, as Grice does) evidence that these assumptions are, at least, roughly correct. One disclaimer. While I shall defend Grice’s "M. O. R., "and therefore the research program initiated by Grice, it is not my goal here to vindicate any specific pragmatic account, nor to argue that any given linguistic phenomenon requires a pragmatic explanation." This reminds me of Kilgariff, a Longman linguist. He has a lovely piece, "I don't believe in word SENSE!" I think he found that Longman had, under "horse": 1. Quadruped animal. 2. Painting of a horse, notably by Stubbs! He did not like that! Why would 'sense,' a Fregeian notion, have a place in something like 'lexicography,' that deals with corpuses and statistics? “The task is, rather, to understand the logic of a particular type of inference, a type of Griceian inference that can be and has been employed by a philosopher such as Grice who disagree on many other points of theory. Since it would be impossible within the confines of this essay to discuss these disagreements, or to do justice to the many ways in which Grice's paradigm or program has been revised and extended (palaeo-Griceians, neo-Griceians, post-Griceians), my Discussion is confined to a few hackneyed examples hackneyed by Grice himself, and to Grice's orthodox theory, if a departure therefrom will be noted where relevant. The conversational explanation of an alleged ambiguity or polysemy or bi-semy aims to show how an utterer U can take an expression E with one conventional meaning and use it as if it had other meanings as well. Typically, this requires showing how the utterer U’s intended message can be ‘inferred,’ with the aid of a general principle of communicative behavior, from the conventional meaning or sense of the word E that U utters. In Grice's pioneering account, for instance, the idea is that speech is subject to a Principle of Conversational Co-Operation (In earlier Oxford seminars, where he introduced 'implicature' he speaks of two principles in conflict: the principle of conversational self-interest , and the principle of conversational benevolence! I much love THAT than the rather artificial Kant scheme at Harvard). ‘Make your conversational contribution, or move, such as is required, at the stage at which it occurs, by the accepted purpose or direction of the conversational exchange in which you are engaged.’ (1975, p. 44). “Sub-ordinate to the Principle of Conversational Co-Operation are four conversational maxims (he was jocularly‘ echoing ’Kant!) Falling under the four Kantian conversational categories of Quality, Quantity, Relation, and Modus. Roughly: Make your contribution true. Kant’s quality has to do with affirmation and negation, rather. Make your contribution informative. Kant’s quantity has to do with ‘all’ and ‘one,’ rather. Make your contribution relevant. Kant’s relation God knows what it has to do with. Make your contribution perspicuous [sic]. Kant’s mode has to do with ‘necessary’ and ‘contingent.’Grice actually has‘ sic ’in the original" Logic and Conversation. " It's like the self-refuting Kantian. Also in 'be brief (avoid unnecessary prolixity' ”'proguard obfuscation,' sort of thing?“… Further specifying what cooperation entails (pp. 45–46). ”It's sad Grice did not remember about the principle of conversational benevolence clashing with the principle of conversational self-interest, or dismissed the idea, when he wrote that 'retro-spective' epilogue about the maxims, etc. Bontly: “Unlike the constitutive (to use Anscombe and Searle, not regulative) principles of a grammar, the Principle of Conversational Co-Operation and the conversational, universalisable, maxims are to be thought of not as an arbitrary convention - vide Lewis - but rather as a rational STRATEGY or guideline (if 'strategy' is too strong) for achieving one's communicative ends. "I DO think 'strategy' is too strong. A strategist is a general: it's a zero-sum game, war. I think Grice's idea is that U is a rational agent dealing with his addressee A, another rational agent. So , it's not strategic rationality, but c communicative rationality. But then I'm being an etymologist! Surely chess players speak of "strategies," but then they also speak of "check mate," kill the king! Bontly quotes from Grice: "‘ [A] nyone who cares about the goals that are central to conversation, ’says Grice, ought to find the principle of conversational cooperation eminently reasonable (p. 49)." If not rational! I love Grice’s /: rational / reasonable. He explores on this later, "The price of that pair of shoes is not reasonable, but hardly irrational!" Bontly: “Like a grammar, however, the principle of conversational co-operation is (supposedly) tacitly known (or assumed) by conversationalists, who can thus call on it to interpret each other’s conversational moves.” Exactly. Parents teach their children well, not to lie, etc. “These interpretive practices being mutual 'knowledge,' or common ground, moreover, an utterer U can plan on his co-conversationalist B using the principle of conversational cooperation, to interpret his own utterances, enabling him to convey a good deal of information (and influencing) implicitly by relying on others to infer his intended meaning. ”INFORMING seems to do, because, although Grice makes a distinction between 'informing' and 'influencing,' he takes an 'exhibitive' approach. So "Close the door!" means "I WANT YOU To believe that I want you to close the door." I.e. I'm informing - influencing VIA informing. “Detailed discussions of Grice's principle of conversational cooperation are found in many of the essays collected in Grice (1989), as well as in the work by linguists like Levinson (1983) and linguist / philosopher Neale (1992). Extensions and refinements of Grice's approach are developed by linguist Horn (1972), linguist / philosopher Bach and philosopher Harnish (1979), linguist Gazdar (1979), linguist / philosopher Atlas and philosopher Levinson (1981), anthropologist Sperber and linguist Wilson (1986 ), linguist / philosopher Bach (1994), linguisdt Levinson (2000), and linguist Carston (2002).). The Principle of conversational cooperation and its conversational maxims allow Grice to draw a distinction between two dimensions of an utterer's meaning within the total significance. ”I never liked that Grice uses“ signification, ”here when in“ Meaning ”he had said:“ Words , for all that Locke said, are NOT signs. " “We apply 'sign' to traffic signals, not to 'dog'.” Bontly: “That which is 'closely related to the conventional meaning of the word' uttered is what the utterer has SAID (1975, p. 44),” or the explicatum, or explicitum. That which must instead be inferred with the aid of the principle of conversational cooperation is what the utterer U has conversationally implicated, the IMPLICATUM (pp. 49-50), or implicitum. This dichotomy is oversimplified in several ways. First, Grice (1975, 1978) also makes room for 'conventional' implicatures (“She was poor BUT she was honest”) and non-conversational non-conventional implicatures (“Thank you,” abiding with the maxim, 'be polite' ), although these dimensions are both somewhat controversial (cf. Bach's attack on conventional implicature) and can be set aside here. Also controversial is the precise delineation of Grice’s notion of what is said. " He grants he is using ‘say’ ‘artifiicially,’ which means, “natural TO ME !.” Some (anthropologist Sperber and linguist Wilson, 1986; linguist Carston, 1988, 2002; philosopher Recanati, 1993) hold that 'what is said,' the DICTUM, the explicatum, or explicitum, is significantly underdetermined by the conventional meaning of the word uttered , with the result that considerable pragmatic intrusive processing must occur even to recover what the utterer said. "And Grice allows that an implicatum can occur within the scope of an operator." Linguist / philosopher Bach disagrees, though he does add an 'intermediate' dimension (that of conversational 'impliciture') which is, in part, pragmatically determined, enriched, or intruded. For my purpose, the important distinction is between that element of meaning which is conventional or 'encoded' and that element which is 'inferred,' ab-duced, or pragmatically determined, whether or not it is properly considered part of what is said, ”In Grice's admittedly artificial use of this overused verb! (“A horse says neigh!”) A conversational implicatum can itself be either particularized (henceforth, PCIs) or generalized (GCIs) (56). ” Most familiar examples of implicature are particularized, where the inference to the utterer U's intended meaning relies on a specific assumption regarding the context of utterance. "Grice's first example, possibly," Jones has beautiful handwriting "(Grice 1961)." Alter that context much at all and the implicatum will simply disappear, perhaps to be replaced by another. With a generalized implicatum, on the other hand, the inference or abduction to U's intended interpretation is relatively context-independent, going through unless special clues to the contrary are provided to defeat it. "Love the 'defeat.' Levinson cites one of Grice's unpublications as “Probability, defeasibility, and mood operators,” where Grice is actually writing, “desirability.”! “For instance, an utterance of the sentence” 'SOME residents survived the earth-quake,' would quite generally, absent any special clues to the contrary, seem to implicate that not all survived. All survived, alas, seems to be, to some, no news. Cruel world. No special "stage-setting" has to be provided to make the implicatum appreciable. No particular context needs to be assumed in order to calculate the likely intended meaning. All one needs to know is that an utterer U who thought that everyone, all residents survived the earthquake (or that none did?) Would probably make this stronger assertion (in keeping with Grice's first sub-maxim of Quantity: 'Make your contribution as informative as required '). ”Perhaps it's best to deal with buildings. “Some - some 75%, I would say - of the buildings did not collapse after the earth-quake on the tiny island, and fortunately, no fatalities need be reported. It wasn't such a big earth-quake as pessimist had predicted. ”“ A Gricean should maintain that the 'ambiguity' of “some” -> “not all” canvassed at the outset can all be explained in terms of a generalized conversational implicatum. For instance, linguist Horn shows, in his PhD on English, how an exclusive use of ‘or’ can be treated as a consequence of the maxim of quantity. Roughly, since p AND q ’is always‘ more informative, ’stronger, than‘ p or q ’, an utterer U’s choosing to assert only the disjunction would ordinarily indicate that he takes one or the other disjunct to be false. He could assert the conjunction anyway, but then he would be violating Grice's first submaxim of Quality: 'Do not say what you believe to be false' For similar reasons, the assertion of a disjunction would ordinarily seem to implicate that the utterer U does not know which disjunct is true (otherwise he would assert that disjunct rather than the entire disjunct) and hence, and this is the way Grice puts it, which is technically, the best way, that the utterer wants to be 'interpreted' as having some 'Non-truth-functional grounds' for believing the disjunction (philosopher Grice, 1978; linguist Gazdar, 1979). For recall that this all goes under the scope of a psychological attitude. In “Method in psychological philosophy: from the banal to the bizarre,” repr. in "The conception of value," Grice considers proper disjunctions: "The eagle is not sure whether to attack the rabbit or the dove." I think Loar plays with this too in his book for Cambridge on meaning and mind and Grice. “Grice (1981) takes a similar line with regard to asymmetric uses of 'and'.” Indeed, I loved his “Jones got into bed and took off his clothes, but I do not want to suggest in that order. " "Is that a linguistic open?" Don't think so! ”“ The fourth submaxim of Manner, ”'be orderly' - I tend to think this is ad-hoc and that Grice had this maxim JUST to explain away the oddity of“ She got a children and married , "By Strawson in Strawson 1952." says that utterers should be 'orderly,' and when describing a sequence of events, an orderly presentation would normally describe the events in the order in which they occurred.So an utterance of (1b) ('Jones took off his trousers - he had taken off his shoes already - and got into bed.' “Would ordinarily (unless the utterer U 'indicates' otherwise) implicate that Jones did so in that order, hence the temporal reading of 'and'. ”“ (Grice's (1981) account of asymmetric 'and' seems NOT to account for causal interpretations like (1c). ”Ryle says in“ Informal logic, ”1953, in Dilemmas, "She felt ill and took arsenic," has the conscript 'and' of Whitehead and Russell, not the 'civil' 'and' of the informalist. "Oxonian philosopher RCS Walker - what took him to respond to Cohen? Walker quotes from Bostock , who was Grice's tutee at St. John's - (1975, p. 136) suggests that the causal reading can be derived from the maxim of Relation. "Nowell-Smith had spoken of 'be relevant' in Ethics. But Grice HAD to be a Kantian! “Since conversationalists are expected to make their utterances relevant, one expects that conjoined sentences will 'have some bearing to one another ’, Often a causal bearing. More nearly adequate accounts of the temporal and causal uses of ‘and’ (so-called conjunction buttressing ’) are found in linguist / philosopher Atlas and linguist Levinson (1981) and in linguist Levinson (1983, 2000). Linguist Carston (1988, 2002) develops a rival pragmatic account within the framework of anthropologist Sperber’s and linguist Wilson’s Relevance Theory, on which temporal and causal readings are explicatures rather than implicatures. For the purposes of this essay, it is immaterial which of these accounts best accords with the data. In these and many other cases, it seems that a general principle regarding communicative RATIONALITY can provide an alternative to positing a semantic ambiguity. "Williamson is lecturing at Yale that 'rationality' has little to do with it!" But a Gricean goes a step further and claims that the implicatum account (when available) is BETTER than an ambiguity or polysemy account. One possible argument for the stronger thesis is that the various specialized uses of ‘or’ (etc.) bear all the usual hallmarks of a conversational implicatum. An implicatum is: calculable (i.e. derivable from what is said or dictum or explicatum or explicitum via the Principle of conversational cooperation and the conversational maxims); cancellable (retractable without contradiction), and; non-detachable (incapable of being paraphrased away) Grice, 1975, pp. 50 and 57-58). They ought also to be, sort of, universal. " (Cf. Elinor Keenan Ochs, “The universality of conversational implicature.” I hope Williamson considers this. In Madagascar, they have other 'norms' of conversation: since speakers are guarded, implicata to the effect, “I don't know” are never invited! Unlike the true lexical ambiguity that arises from a language-specific convention, an implicatum derives rather from general features of communicative RATIONALITY and should thus be similar across different languages ​​(philosopher Kripke, 1977; linguist Levinson, 1983). ”I. 'm not sure. Cfr. Ochs in Madagascar. But she is a linguist / anthropologist, rather than a philosopher? From a philosophical point of view, perhaps the best who treated these issues is English philosopher Martin Hollis in his essays on' rationality ' and 'relativism' (keywords!) “Since the 'ambiguity' in question here has all these features, at least to some degree, the implicatum approach may well seem irresistible. It is well known, however, that none of the features listed on various occasions by Grice are sufficient (individually or jointly) to establish the presence of a conversational implicatum (Grice, 1978; linguist Sadock, 1978). Take calculability. " Or how to ‘work it out,’ to keep it Anglo-Saxon, as pretentious Grice would not! The main difficulty is that a conversationalimplicatum can become fossilized, or ‘conventionalized’ over time but remain calculable nonetheless, as happens with some ‘dead’ metaphors - one-time non-literal uses which congealed into a new conventional meaning. " A linguist at Berkeley worked on this, Traugott, on items in the history of the English language, or H-E-L, for short, H.O.T.E.L, history of the English language. I don’t think Grice considers this. He sticks with old Roman ‘animal’ -> ‘non-human’, strictly, having a ‘soul,’ or animus, anima. (I think Traugott's focus was on verb forms, like “I have eaten,” meaning, literally, “I possess eating,” or something. But she does quote Grice and speaks of fossilization. “For instance, the expression.” 'S went to the bathroom '(Jones?) could, for obvious reasons, be used with its original, compositional, meaning to implicate that S' relieved himself '. ”“ The intended meaning would still be calculable today. ”Or“ went to powder her nose? ”(Or consider the pre-Griceian (?) child's overinformative, standing from table at dinner,“ I'm going to the bathroom to do number 2 (unless he is flouting the maxim). “But the use has been absorbed, or encoded into some people's grammar, as witnessed by the fact that 'S went to the bathroom on the living room carpet.' is not contradictory (linguist JL Morgan, 1978; linguist Sadock, 1978). ”I wonder what some contextualists at Yale (De Rose) would say about that !? Cf. Jason Stanley, enfant terrible. “Grice's cancellability is similarly problematic. While one may cancel the exclusive interpretation of 'p or q' (eg by adding 'or possibly both'), the added remark could just as well be disambiguating an ambiguous utterance as canceling the implicatum (philosopher Walker, 1975; linguist Sadock, 1978 ). ”Excellent POINT! Walker would be fascinated to see that Grice once coined "disimplicature" for some loose uses. "Macbeth saw Banquo." "That tie is yellow under that light, but orange under this one." Actually, Grice creates "disimplicature" to refute Davidson on intending: "Jones intends to climb Mt Everest next weekend." Intending DOES entail BELIEF, but people abuse ‘intend’ and use it loosely, ’with one sense dropped. Similarly, Grice says, with “You're the cream in my coffee,” where the ‘disimplicature’ is TOTAL! “Non-detachability fares no better. When two sentences are synonymous (if there is, pace Quine, such a thing), utterances of them ought to generate the same implicatum. But they will also have the same semantic implications, so the non-detachability of an alleged implicatum shows very little if anything at all (linguist Sadock, 1978). ”I never liked non-detachability, because it ENTAILS that there MUST be a synonym expression: cfr. God? Divinity? "Universality is perhaps the best test of the four." I agree. When linguists like Elinor Keenan disregard this, I tend to think: “the cunning of conversational reason,” alla Hollis. Grice was a member of Austin’s playgroup, and the conversational MAXIMS were "universalisable" within THAT group. That seems okay for both Kant AND Hegel! “Since an implicatum can fossilize into a conventional meanings, however, it is always possible for a cross-linguistic alleged 'ambiguity' to be pragmatic in some language though lexical in another.” Is that ' f * rnication '? Or is it Grice on ‘pushing up the daisies’ as an “established idiom” for ‘… is dead’ in WJ5? Austin and Grice would I think take for granted THREE languages: Greek and Roman, that they studied at their public schools - and this is important, because Grice says his method of analysis is somehow grounded on his classical education - and, well, English. Donald Davidson, in the New World, would object to the ‘substantiation’ that speaking of “Greek” as a language, say, may entail. “So while Grice’s tests are suggestive, they supply no clear verdict on the presence of an implicatum. Besides these inconclusive tests for implicature, Grice could also appeal to various diagnostic tests for alleged ambiguity. " “And how to fail them,” to echo Zwicky. Grice himself suggests three, although none of them prove terribly helpful. ”Loved your terrible. Cfr. Terrific ’. And the king entering St. Paul’s cathedral: “Aweful!” meaning ‘awe-some!’ “First, Grice points out that each alleged sense Sn of an allegedly ambiguous word E ought to be expressible‘ in a reasonably wide range of linguistic environments ’(1978, p. 117). The fact that the strong implicatum of "or" is UNavailable within the scope of a negation, for instance, would seem to count AGAINST alleged ambiguity or polysemy. On the other hand, the strong implicatum of or ’IS available within the scope of a propositional-attitude verb. A strong implicatum of 'and' is arguably available in both environments, within the scope of a negation, and within the scope of a psychological-attitude verb. So the first test seems a wash. ”Metaphorically, or implicaturally. J “Second, Grice says, if the expression E is ambiguous with one sense S2 being derived (somehow) from the initial or original or etymological sense S1, that derivative sense S2 'ought to conform to whatever principle there may be which governs the generation of derivative senses' (pp. 117–118). ”GRICE AT HIS BEST! I think he is trying to irritate Quine, who is seating on second row at Harvard! (After all Quine thought he was a field linguist!) Bontly, charmingly: “Not knowing the content of thi principle Grice invokes— and Grice gives us no hint as to what it might be - we cannot bring it, alas, to bear here ! ”I THINK he was thinking Ullman. At Oxford, linguists were working on ‘semantics,’ cfr. Gardiner. And he just thought that it would be Unphilosophical on his part to bore his philosophical Harvard audience with 'facts.' At one point he does mention that the facts of the history of the English language (how 'disc' can be used, etc. ) are not part of the philosopher's toolkit? "Third and finally, Grice says, we must 'give due (but not undue) weight to MY INTUITIONS about the existence (or indeed non-existence) of a putative sense S2 of a word E . '(p. 120). ”Emphasis on' my 'mine! - As I say, I never had any intuition about an expression having an extra-putative sense. Not even 'bank,' - since in Old Germanic, it's all etymologically related! Bontly: “But, even granting the point that 'or' is NON-INTUITIVELY ambiguous in quite the same way that 'bank' IS, allegedly, INTUITIVELY ambiguous , the source of our present difficulty is precisely the fact that 'p or q' often * seems * intuitively to imply that one or the other disjunct is false. ”Grice apparently uses 'intuition' and 'introspection' interchangeably, if that helps? Continental phenomenological philosophers would make MUCH of this! For Grice’s intuitions, HIS is own. In a lecture at Wellesley, of all places (in Grice 1989) he writes: "My problems with my use of E arise from MY intuitions about the use of E. I don’t care how YOU use E. Philosophy is personal." Much criticized, but authentic, in a way! “Since he discounts the latter intuition, Grice cannot place much weight on the former!” As I say, Grice’s intuitions are hard to fathom! So are his introspections! Actually, I think that Grice’s sticking with introspections and intuitions save him, as Suppes shows in PGRICE ed Grandy and Warner, from being a behaviorist. He is, rather, an intentionalist! “While a complete review of ambiguity tests is beyond the scope of this essay, we have perhaps seen enough to motivate the methodological problem with which we began: viz., That an, intuitive, alleged, ambiguity seems fit to be explained either semantically (ambiguity thesis, polysemy, bi-semy) or pragmatically / conversationally, with little by way of direct evidence to tell us which is which! ”“ If philosophy generated no problems, it would be dead! ” - Grice. J “Linguists Zwicky and Sadock review several linguistic tests for ambiguity (eg conjunction reduction) and point out that most are ill-suited to detect ambiguities where the meanings in question are privative opposites,” Oddly, Grice's first publication ever was on “Negation and privation, ”1938! Bontly:“ ie where one meaning is a specialization or specification of the other (as for instance with the female and neutral senses of ‘goose’). "Or cf. Urmson," There is an animal in the backyard. " “You mean Aunt Matilda?” Bontly: “Since the putative ambiguities of‘ or ’and the like are all of this sort, it seems inevitable that these tests will fail us here as well. For further discussion, see linguist Horn (1989, pp. 317–18 and 365–66) and linguist Carston (2002, pp. 274–77). It is hardly surprising, therefore, to find that a Gricean typically falls back on a methodological argument like parsimony, as instantiated in “M. O. R. "Let's now turn to Parsimony and Its Problems. It may, at first, be less than obvious why an ambiguity or polysemy or bi-semy account should be deemed less parsimonious than its Gricean rival. " Where the conventionalist or ambiguist posits an additional sense S2, Grice adds, to S1, a conversational implicatum, I ”. Cheap, but no free lunch! (Grice saves) Bontly: "Superficially, little seems to be gained." Ah, the surfaces of Oxford superficiality! “Looking closer, however, the methodological virtues of the Grice’s approach seem fairly clear.” Good! Bontly: “First, the principles and inference patterns that a pragmatic or conversational account utilizes are independently motivated. The principles and inference patterns are needed in any case to account for the relatively un-controversial class of particularized implicata, and they provide an elegant approach to phenomena like figures of rhetoric, or speech - metaphor, irony, meiosis, litotes, understatement, sarcasm - cfr. Holdcroft - and tricks like Strawson's presupposition. So it would seem that Grice can do with explanatory material already on hand, whereas the ambiguity or polysemy theorist must posit a new semantic rule in each and every case. Furthermore, the explanatory material has an independent grounding in considerations of rationality. "I love that evening when Grice received a phonecall at Berkeley:" Professor Grice: You have been appointed the Immanuel Kant Memorial Lecturer at Stanford. " He gave the lectures on aspects of reason and reasoning! Bontly: “Since conversation is typically a goal-directed activity, it makes sense for conversationalists to abide by the Principle of Conversational Cooperation (something like Kant's categorical imperative, in conversational format) and its (universalisable) conversational maxims, and so it makes sense for a co-conversationalist to interpret the conversationalist accordingly. A pragmatic explanation is therefore CHEAP - hence Occam on ‘aeconomicus’ - the principle it calls on being explainable by - and perhaps even reducible to - facts about rational behavior in general. "I loved your" REDUCE. " B. F. Loar indeed thought, and correctly, that the maxims are ‘empirical generalizations over functional states.’ Genius! Bontly: A pragmatic account is not only more economic, or cheaper. It also reveals an orderliness or systematicity that positing a separate lexical ambiguity or polysemy or bisemy in each and every case would seem to miss (linguist / philosopher Bach). To a Griceian, it is no accident that a sentential connective or truth-functor (“not,” “and,” “or,” and “if”), a quantified expression (Grice's “all” and “some (at least one ) ”) And a description (Grice's“ the ”) all lend themselves to a weak and a stronger interpretation” Cf. Holdcroft, "Weak or strong?" in “Words and deeds.” Bontly: “Note, for instance, that a sentence with the logical form 'Some Fs are Gs', and the pleonethetic, to use Geach's and Altham's coinage,' Most Fs are Gs', and 'A few Fs are Gs' are all allegedly 'ambiguous' in the SAME way. Each of those expressions has an obvious weak reading in addition to a stronger reading: ‘Not all Fs are Gs’.Good because Grice’s first examination was:" That pillar-box seems red to me. " And he analyzes the oddness in terms of "strength." (Grice 1961). He tries to analyze this 'strength' in terms of 'entailment,' but fails (“Neither 'The pillar-box IS red' NOR 'The pillar-box SEEMS red' entail each other.”) Bontly: For the conventionalist or polysemy theorist, there is no apparent reason why this should be so. There is no reason, that is, why three etymologically unrelated words (“some,” “most,” and “few”) should display the SAME pattern of alleged ambiguity. The Gricean, on the other hand, explains each the SAME way, by appealing to some rational principle of conversation. The implicata are all ‘scalar’ quantity implicata, attributable to the utterer U’s having uttered a weaker, less informative, sentence than he might have. " Linguist Levinson, 1983). Together, these considerations make a persuasive case for the Grice’s approach. A pragmatic explanation is more economical, and the resulting view of conversation is more natural and unified. Since economy and unification are both presumably virtues to be sought in a scientific or philosophical explanation - virtues which for brevity I lump together under Occamist 'parsimony' - it would NOT be unreasonable to conclude that a pragmatic explanation is (ceteris paribus) a better explanation . So it seems that Grice’s principle, the “M. O. R. " is correct. Senses ought not to be multiplied when pragmatics will do. Still, there are several reasons to be suspicious of the parsimony argument.“I lay out three. It bears emphasis that none of these are objections to the pragmatic approach per se. " I have no quarrel with the theory of conversation or particular attempts to apply it to conversational phenomena. The objections focus rather on the role that parsimony (or simplicity, or generality, etc.) plays in arguments PRO the implicatum and CONTRA ambiguity or polysemy. " Then, there’s dead metaphors. First is a worry that parsimony is too blunt an instrument, generalizing to unwanted conclusions. Versions of this objection appear in philosopher Walker (1975), linguist Morgan (1978), and linguist Sadock (1978). " More recently, Reimer (1998) and Devitt (forthcoming) use it to argue against a Gricean treatment of the referential / attributive distinction. "But have they read Grice’s VACUOUS NAMES? I know you did! Grice notes: "My distinction has nothing to do with Donnellan's!" Grice’s approach is syntactic: ‘the’ and “THE,” identificatory and non-identificatory uses. R. M. Sainsbury and D. E. Over have worked on this. Fascinating. Bontly: “For as with the afore-mentioned so-called‘ dead ’metaphor, it can happen that a word has a secondary use that is pragmatically predictable, and yet fully conventional. In many such cases, of course, the original, etymological meaning is long forgotten: e. G. the contemporary use of "fornication", originally a euphemism for activities done in fornice (that is, in the vaulted underground dwellings that once served as brothels in Rome). (I owe this [delightful] example to Sam Wheeler). Few speakers recall the original meaning, so the metaphor can no longer be ‘calculated,’ as Grice’s “You’re the cream in my coffee!” (title of song) can! ” The metaphor is both dead _and buried _. ”Still un-buriable?“ In other cases, however, speakers do possess the information to construct a Gricean explanation, and yet the metaphor is dead anyway. ”Reimer's (1998) example of the verb ' incense 'is a case in point. One conventional meaning (to make or become angry ’) began life as a metaphorical extension of the other (to make fragrant with incense’). The reason for the extension is fairly transparent (resting on familiar comparisons of burning and emotion), but the use allegedly represents an additional sense nonetheless. ”What dictionaries have as 'fig.' But are we sure that when the dictionaries list things like 1 ., 2., 3., they are listing SENSES !? Cf. Grice, "I don't give a hoot what the dictionary says," to Austin, "And that's where you make your big mistake." Once Grice actually opened the dictionary (he was studying 'feeling + adj.' - he got to 'byzantine,' finding that MOST adjectives did, and got bored! Bontly: Such examples suggest that an implicatum makes up an important source of semantic— and, according to linguist Levinson (2000), syntactic — innovation. A linguistic phenomenon can begin life as a pragmatic specialization or an extension and subsequently become conventionalized by stages, making it difficult to determine at what point (and for which 'utterers') a use has become fully conventional. One consequence is that an expression E can have, allegedly, a second sense S2, even when a pragmatic explanation appears to make it explanatorily superfluous, and parsimony can therefore mislead. ”I'm not sure dictionary readers read 'fig.' as a different 'sense,' and lexicographers need not be Griceian in style! Bontly: “A related point is that an ambiguity account needn't be LESS unified than an implicatum account after all. If pragmatic cons iderations can explain the origin and development of new linguistic conventions, the ambiguity or polysemy theorist can provide a unified dia-chronic account of how several un-related expressions came to exhibit similar patterns of alleged 'ambiguity.' Quantifiers like 'some', ' most ', and' a few 'may be similarly allegedly ambiguous today because they generated similar implicatures in the past (cf. Millikan, 2001). "OKAY, so that's the right way to go then? Diachrony and evolution, right? Bontly: “Then, there’s tradeoffs. A ‘dead’ metaphor suggests that parsimony is too strong for the pragmatist’s purposes, but as a pragmatic account could have hidden costs to offset the semantic savings, parsimony may also be too weak! E. g. an implicatum account looks, at least superficially, to multiply (to use Occam’s term) inferential labor, leaving it to the addressee to infer the utterer’s intended meaning from the words uttered, the context, and the conversational principle. Thus there are trade-offs involved, and the account which is semantically more parsimonious may be less parsimonious all things considered. ”Grice once invited the“ P. E. R. E., "Principle of economy of rational effort, though. Things which seem to be psychologically UNREAL are just DEEMED, tacitly, to occur.Bontly: “To be clear, this is not to suggest that the ambiguity or polysemy account can dispense with inference entirely. Were the exclusive and inclusive senses of 'or' BOTH lexically encoded (as they were in Old Roman, 'vel' and 'aut,' hence Whitehead's choice of 'v' for 'pv q') still hearers would need to infer from contextual clues which meaning were intended. The worry is not, therefore, so much that the implicatum account increases the number of inferences which conversants or conversationalists have to perform. The issue concerns rather the complexity of these inferences. Alleged dis-ambiguation is a highly constrained process. In principle, one need only choose the relevant sense Sn, from a finite list represented in the so-called ‘mental lexicon’. Implicature calculation, on the other hand, is a matter of finding the best explanation (abductively, alla Hanson) for an utterer’s utterance, the utterer’s meaning being introduced as an explanatory hypothesis, answering to a ‘why’ question. Unlike dis-ambiguation, where the various possible readings are known in advance, in the conversational explanation, the only constraints are provided by the addressee’s understanding of the context and the conversational principle. So it appears that Grice’s approach saves on the lexical semantics by placing a greater inferential burden on utterer and addressee. ”But Grice played bridge, and loved those burdens. Stampe actually gives a lovely bridge alleged counter-example to Grice (in Grice 1989). Bontly: “Now, a Gricean can try to lessen this load in various ways. Grice can argue, for instance, that the inference used to recover a generalized implicatum is less demanding than that for a particularized one, that familiarity with types of generalized implicate can “stream-line” the inferential process, and so on. ”Love that , PERE, or principle of economy of rational effort, above?! Bontly: “We examine these moves. There's justification. Another difficulty with Grice’s appeals to parsimony is the most fundamental. On the one hand, it can hardly be denied that parsimony plays a role in scientific, if not philosophical, inference. " Across the sciences, if not in philosophy, it is standard practice to cite parsimony (simplicity, generality, etc.) as a reason to choose one hypothesis over another; philosophers often do the same. ”Bontly’s‘ often ’implicates,‘ often not ’! Grice became an opponent of his own minimalism at a later stage of his life, vide his “Prejudices and predilections; which become, the life and opinions of Paul Grice, ”by Paul Grice! Bontly:“ At the same time, however, it remains quite mysterious, if that's the word, why parsimony (etc.) should be given such weight by Occamists like Grice. If it were safe to assume that nature is simple and economical, the preference for theories with these qualities would make perfect sense. Sir Isaac Newton offers such an ontological rationale for parsimony in the "Principia." Sir Isaac writes (in Roman?) "I am to admit no more cause of a natural thing than such as are true and sufficient to explain its appearance." "To this purpose, the philosopher says that Nature does nothing in vain, and more is in vain when less serves." "For Nature is pleased with simplicity, and affects not the pomp of a superfluous cause." "While a blanket assertion about the simplicity of nature is hardly uncommon in the history of science, today it is viewed with suspicion." Bontly: "Newton’s reasons were presumably theological." "If I knew that the Creator values ​​simplicity and economy, I should expect the creatION to display these qualities as well." “Lacking much information about the Creator’s tastes, however, the assumption becomes quite difficult, if not impossible, to support.” Cfr. literature on 'biological diversity.' Bontly: "(Sober discusses several objections to an ontological justification for the principle of parsimony. Philosopher of science Mary Hesse surveys several other attempts to justify the use of parsimony and simplicity in scientific inference. Philosophers of science today are largely persuaded that the role of parsimony is 'purely methodological' epistemological, pragmatist, rather than ontological - that it is rational to reject unnecessary posits (or complex, dis-unified theories) no matter what nature is like. One might argue, for instance, that the principle of parsimony is really just a principle of minimum risk. The more existence claims one accepts, the greater the chance of accepting a falsehood. Better, then, to do without any existence claim one does not need. Philosopher JJC Smart attributes this view to John Stuart Mill. ”Cf. Grice:“ Not to bring more Grice to the Mill. ”Bontly:“ Now, risk minimization may be a reasonable methodolo gical principle, but it does not suffice to explain the role of parsimony in natural science. When a theoretical posit is deemed explanatorily superfluous, the accepted practice is not merely to withhold belief in its existence but to conclude positively that it does not exist. As Sober notes, 'Occam's razor preaches atheism about unnecessary entities, not just a-gnosticism.' ”Similarly, Grice's razor tells us that we should believe an expression E to be unambiguous, aequi-vocal, monosemous, unless we have evidence for a second meaning. The absence of evidence for this alleged additional, "multiplied" "sense" is presumed to count as evidence that this alleged second, additional, multiplied, sense is absent, does not exist. But an absence of evidence is not the same thing as evidence of an absence. " The difficult question about scientific methodology is why we should count one as the other. Why, that is, should a lack of evidence for an existence claim count as evidence for a non-existence claim? The minimum risk argument leaves this question unanswered. Indeed, philosophers of science have had so little success in explaining why parsimony should be a guide to truth that many are tempted to conclude that it and the other ‘super-empirical virtues’ have no epistemic value whatsoever. Their role is rather pragmatic, or aesthetic. ”This is in part Strawson’s reply in his“ If and the horseshoe ”(1968), repr. in PGRICE, in Grandy / Warner. He says words to the effect: "Grice’s theory may be more BEAUTIFUL than mine, but that’s that!" (Strawson thinks that 'if' acts as 'so' or 'therefore' but in UNASSERTED clauses. So it's a matter of a 'conventional' IMPLICATUM to the inferrability of “if p, q” or “p; so, q.” I agree with Strawson that Grice's account of 'conventional' implicatum is not precisely too beautiful? Bontly: “Parsimony can make a theory easier to understand or apply, and it pleases those of us with a taste for desert landscapes, but (according to these skeptics) they do not make the theory any more likely to be true. ”The reference to the 'desert landscape' is genial. Cfr. Strawson's“ A logician's landscape. ”Later in life, Grice indeed found it unfair that an explanation of cherry trees blooming in spring should be explained as a 'desert landscape.' “That's impoverishing it!” Bontly: “van Fraassen, for instance, tells us that a super-empirical virtue 'does not concern the relation between the theory and the world, but rather the use and usefulness of the theory; it provide reasons to prefer the theory inde pendently of questions of truth. "" If that were correct, it would be doubtful that parsimony can shoulder the burden Grice places on it. "" For then the conventionalist may happily grant that a pragmatic explanation is clever and elegant, and beautiful. " "The conventionalist can agree that an implicature account comprehends a maximum of phenomena with a minimum of theoretical apparatus." "But when it comes to truth, or alethic satisfactoriness, as Grice would prefer, a conventionalist may insist that parsimony is simply irrelevant." “One Gricean sympathizer who apparently accepts the 'aesthetic' view of parsimony is the philosopher of science RCS Walker (1975), who claims that the '[c] hoice between Grice's and Cohen's theories is an aesthetic matter' and concludes that 'we should not regard either the Conversationalist Hypothesis or its [conventionalist] rivals as definitely right or wrong. '”Cfr. Strawson in Grandy / Warner, but Strawson is no Griceian sympathizer! "Now asking Grice to justify the principle of parsimony may seem a bit unfair." "Grice also assumes the reality of the external world, the existence of intentional mental states, and the validity of modus ponens." "Need Grice justify these assumptions as well?" "Of course not!" "But even if the epistemic value of parsimony is taken entirely for granted, it is unclear why it should even count in semantics." “All sides agree, after all, that many, perhaps even most, expressions of natural language are allegedly 'ambiguous.'” “There are both poly-semies, where one word has multiple, though related, meanings ('horn', ' trunk '), and homo-nymies, where two distinct words have converged on a single phonological form (' bat ',' pole '). " "The distinction between poly-semy and homo-nymy is notoriously difficult to draw with any precision, chiefly because we lack clear criteria for the identity of words (Bach)." "If words are individuated phono-logically, there would be no homo-nyms." "If words are individuated semantically, there would be no poly-semies." “Individuating words historically leads to some odd consequences: eg, that 'bank' is poly-semous rather than homo-nymous, since the 'sense' in which it means financial institution and the 'sense' in which it means edge of a river are derived from a common source. " “I owe this example to David Sanford. For further discussion, see Jackendoff. ”Soon at Hartford. And Sanford is right! Bontly: "Given that ambiguity is hardly rare, then, one wonders whether a semantic theory ought really to minimize it (cf. Stampe, 1974)." "One might indeed argue that the burden of proof here is on the pragmatist, not the ambiguity or polysemy theorist." "Perhaps we ought to assume, ceteris paribus, that every regular use of an expression represents a SPECIAL sense." “Such a methodological policy may be less economical than Grice's, but it does extend the same pattern of explanation to all alleged ambiguities, and it might even accord better with the haphazard ways in which natural languages ​​are prone to evolve (Millikan, 2001). "Yes, the evolutionary is the way to go! Bontly:" So Grice owe us some reason to think that parsimony and the like should count in semantics. " "He needn’t claim, of course, that parsimony is always and everywhere a reason to believe a hypothesis true." "He needn’t produce a global justification for Occam’s Razor, that is — a local justification, one specific to language, would suffice." "I propose to set aside the larger issue about parsimony in general, therefore, and argue that Modified Occam’s Razor can be justified by considerations peculiar to the study of language." "Now for A Developmental Account of Semantic Parsimony." “My approach to parsimony in linguistics is inspired by Sober's work on parsimony arguments in evolutionary biology.” And Grice was an evolutionary philosopher of sorts.Bontly: “In Sober's view, philosophers have misunderstood the role of parsimony in scientific inference, taking it to function as a global, domain-general principle of scientific reasoning (akin perhaps to an axiom of the probability calculus). " "A more realistic analysis, Sober claims, shows that parsimony arguments function as tacit references to domain-specific process assumptions - to assumptions (whether clearly articulated or not) about the process (es) that generate the phenomena under study." "Where these processes tend to be frugal, parsimony is a reasonable principle of theory-choice." "Where they are apt to be profligate, it is not." "What makes parsimony reasonable in one area of ​​inquiry may, on Sober’s view, be quite unrelated to the reasons it counts in another." "Parsimony arguments in the units of selection controversy, for instance, rest on one set of process assumptions (i.e. assumptions about the conditions necessary for‘ group ’selection to occur)." "The application of parsimony to‘ phylogenetic ’inference rests on a completely different set of assumptions (about rates of evolutionary change)." “As Sober notes, in either case the assumptions are empirically testable, and it could turn out that parsimony is a reliable principle of inference in one, both, or neither of these areas. Sober’s approach amounts to a thorough-going local reductionism about parsimony.It counts in theory-choice if and only if there are domain-specific reasons to think the theory which is more economical (in some specifiable respect) is more likely to be true.The "only if" claim is the more controversial part of the bi-conditional, and I need not defend it here. For present purposes I need only the weaker claim that domain-specific assumptions can be sufficient to justify using parsimony - that parsimony is a sensible principle of inference if the phenomena in question result from processes themselves biased, as it were, towards parsimony. Now, in natural-language semantics, the phenomena in question are ordinarily taken to be the semantic rules or conventions shared by a community of speakers. ”Cf. Peacocke on Grice as applied to ‘community of utterers,’ in Evans / McDowell, Truth and meaning, Oxford. Bontly: “The task is to uncover the‘ arbitrary ’mappings between a sound and a meaning (or concepts or referent) of which utterers have tacit knowledge. This "semantic competence" is shaped by both the inputs that language learners encounter and the cognitive processes that guide language acquisition from infancy through adulthood. So the question is whether that input and these processes are themselves biased toward semantic parsimony and against the acquisition of multiple meanings for single phonological forms. As I shall now argue, there are several reasons to suspect that such a bias should exist. Psychologists often conceptualize learning in general and word learning in particular as a process of generating and testing hypotheses. A child (or, in many cases, an adult) encounters an unfamiliar word, forms one or more hypotheses as to its possible meaning, checks the hypotheses against the ways in which he hears the word used, and finally adopts one such hypothesis. This child-as-scientist ’model is plainly short on details, but whatever mechanism implements the generating and testing, it would seem that the process cannot be repeated with every subsequent exposure to a word. Once a hypothesis is accepted - a word learned - the process effectively halts, so that the next time the child hears that word, he doesn’t have to hypothesize. Instead, the child can access the known meaning and use it to grasp the intended message. For that reason, an unfamiliar word ought to be the only one to trigger the learning process, and that of course makes ambiguity problematic. Take a person who knows one meaning of an ambiguous word, but not the other. To him, the word is not unfamiliar, even when used with an unfamiliar meaning. At least, it will not sound unfamiliar. So, the learning process will not kick in unless some other source of evidence suggests another, as-yet-unknown meaning. Presumably the evidence will come from "anomalous" utterances: i.e. uses that are contextually absurd, given only the familiar meaning. This is not to say, of course, that hearing one anomalous utterance would be sufficient to re-start the learning process. Since there are other reasons why an utterance may seem anomalous (e.g. the utterer simply misspoke), it might take several anomalies to convince one that the word has another meaning. In the absence of anomalies, however, it seems highly unlikely that learners would seriously entertain the possibility of a second sense. A related point is that acquisition involves, or is at least thought to involve, a variety of 'boot-strapping' operations where the learner uses what he knows of the language in order to learn more. ”Oddly Grice has a bootstrap principle (it relates to having one's metalanguage as rich as one's object-language.Bontly: “It has been argued, for instance, that children use semantic information to constrain hypotheses about words' syntactic features (Pinker) and, conversely, syntactic information to constrain hypotheses about words 'semantic features (Gleitman). Likewise, children must surely use their knowledge of some words' meanings to constrain hypotheses as to the meanings of others, thus inferring the meanings of unfamiliar words from context. However, that process only works insofar as one can safely assume that the familiar words in an utterance are typically used with their familiar meanings. If it were assumed that familiar words are typically used with unknown meanings, the boots traps would be too weak. Together, these considerations point to the hypothesis that language acquisition is semantically conservative. Children will posit new meanings for familiar words only when necessary — only when they encounter utterances that make no sense to them, even though all the words are familiar. Interestingly, experimental work in language acquisition provides empirical evidence for much the same conclusion. Psychologists have long observed that children have considerable difficulties learning and using homo-nyms (Peters and Zaidel), leading many to suspect that young children operate under the helpful, though mistaken, assumption that a word can have but one meaning (Slobin). Children have similar difficulties acquiring synonyms and may likewise assume that a given meaning can be represented by at most one word. (Markman & Wachtel, see Bloom for a different explanation). I cannot here survey the many experimental studies bearing on this hypothesis, but one series of experiments conducted by Michele Mazzocco is particularly germane. Mazzocco presents children from several age groups, as well as adults, with stories designed to mimic one’s first encounter with the secondary meaning of an ambiguous word. To control the effects of antecedent familiarity with secondary meanings, the stories used familiar words (eg, 'rope') as if they had further unknown meanings — as 'pseudo-homo-nyms'. For comparison, other stories included a non-sense word (eg 'blus') used as if it had a conventional meaning - as a 'pseudo-word' - to mimic one's first encounter with an entirely unfamiliar word. ”Cf. Grice's seminar at Berkeley: “How pirots karulise elatically: some simpler ways.” “A pirot can be said to potch or cotch an obble as fang or feng or fid with another obble.” “A person can be said to perceive or cognize an object as having the property f or f2 or being in a relation R with another object. ”Bontly: Some stories, finally, used only genuine words with only their familiar meanings. After hearing a story, subjects are presented with a series of illustrations and asked to pick out the item referred to in the story. In a subsequent experiment, subjects had to act out their interpretations of the stories. In the pseudo-homo-nym condition, one picture would always illustrate the word’s conventional but contextually inappropriate meaning, one would depict the unfamiliar but contextually appropriate meaning, and the rest would be distractors. As one would expect, adults and older children (10- to 12-year-olds) performed equally well on these tasks, reliably picking out the intended meanings for familiar words, non-sense words and pseudo-homonyms alike. Young children (3- to 5-year-olds), on the other hand, could understand the stories where familiar words were used conventionally, and they were reasonably good at inferring the intended meanings of non-sense words from context, but they could not do so for pseudo-homonyms. Instead, they reliably chose the picture illustrating the familiar meaning, even though the story made that meaning quite inappropriate. These results are noteworthy for several reasons. It is significant, first of all, that spontaneous positing of ambiguities did not occur. As long as the known meaning of a word comported with its use in a story, subjects show not the slightest tendency to assign that word a new, secondary meaning — just as one would expect if the acquisition process were semantically conservative. Second, note that performance in the nonsense word condition confirms the familiar finding that young children can acquire the meanings of novel words from context - just as the bootstrapping procedure suggests. Unlike older children and adults, however, these young children are unable to determine the meanings of pseudo-homo-nyms from context, even though they could do so for pseudo-words - exactly what one would expect if young children assumed that words can have one meaning only. Why young children would have such a conservative bias remains controversial. Unfortunately it would take us too far afield to delve into this debate here. Doherty finds evidence that the understanding of ambiguity is strongly correlated with a grasp of synonymy, suggesting that these biases have a common source. ” Doherty also finds evidence that the understanding of ambiguity / synonymy is strongly predicted by the ability to reason about false beliefs, suggesting the intriguing hypothesis that young children’s biases are due to their lack of a representational ‘theory of mind’). "