The Reluctant Cannibal

In WoW:i, Grice considers Hare's dictum

"To say 'x is good' is to commend x"

Grice says that Hare cannot get away with it -- or if he can, he shouldn't. For surely if the Reluctant Cannibal is, er, ... reluctant as to whether

"Eating people is healthy"

There's more to do with his claim that embed it in a conditional!

--- "If eating people is wrong, I'm vegetarian".

Grice complicates things by having them _in context_. If the cannibal mother teaches her daughter:

"Eating people is _not_ wrong"

the question is -- what does she mean? Is she trying to merely exhibit her belief, or desire, or pro-attitude, that eating people is not, for her, wrong; or is she projecting into the wild world, alla Hume, a quality that she expects her daughter will pick up in the air and make her own?

These things I have discussed with M. Silcox, and he learned about the protreptic/exhibitive thing into the bargain...

My reactions on M. Silcox on R. M. Hare and Co. were motivated by M. Silcox having recently offered a most illuminating set of 'radical reflections' on Expressivism & Geach's puzzle. But more of this below.

Cheers,

JL

---

It seems as tho' gist of it all is best summed by MS's remark in the defense of moral expressivism/projectivism. "None of the philosophers who have written about RI have speculated in any kind of detail however on the extent to which a moral utterance can usefully be thought of as being radically interpretable". About time too! It all reminded me of M Flanders & D Swann's song The Reluctant Cannibal ('Eating People is Wrong'), from their Londrevue At the drop of a hat (I'll see if I can find the lyrics to it, as it may have more philosophical relevance than I first thought)

M Silcox writes: "Moral Expressivism (henceforth ME) explains the meaning of a moral term ('good', 'right', 'rational') in terms of the utterer's STATE OF MIND it expresses. Those who advocate it have long recognised the need to respond to P Geach's criticism to an ASSERTION involving a moral term as interpreted as a NON-TRUTH-EVALUABLE utterance of an AFFECTIVE [desiderative, boulomaic?] response to the situation which prompts the utterance. This, he objects, would create a problem for
understanding a moral term in a NON-ASSERTORIC utterance, as in the antecedent of a conditional in the apparently straightforward argument (GEACH's MODUS PONENS): 'It's good (right, rational) to protect one's own children; if it's good (right, rational) to protect one's own children, then it's good (right/rational) to protect others' children. Thus, it's good (right/rational) to protect others' children'".

Two remarks.

1. I wonder if Geach's equivocity puzzle has been posed (or what its consequeces would be if it held) re nonmoral contexts? (I take MS's view that Expressivism is not necessarity a MORAL stance as it may apply eg to a doxastic construal of truth). By 'nonmoral context' I mean to the alleged symmetry of 'good' ('right') and 'true': 'It's true he protects his children. If it's true he protect his children, then it's true he protects other people's children. Thus it's true he protects other people's children'.

2. One wonders if Geach's puzzle is ultimately MORAL? If it's just the formal side he's discussing, I guess it's neither moral nor amoral? (while not necessarily immoral). A moralist would try & eg grasp the CONTENT of the 2nd premise to fully dub it 'moral' (Cf Eating people is wrong & if eating people is wrong then eating Jones is wrong, thus eating Jones is wrong').

"Geach's claim is that ME is based solely on an utterance within which something is ASSERTED [to be good/right/rational] it provides no way of assuring that 'good' [right, rational, etc] means the same in the first and 2n hence whether this is in fact a valid instance of modus ponens [But isn't such a thing as condi]. Blackburn suggests the proper reading of a moral conditional utterance is as an endorsement for the 'sensibility' of recognising a link between [goodnesses]. The obvious difficulty with this is that, after all, there's surely no inconsistency in recognising a link between two [goodnesses] while at the same time being prepared to adopt a permissive attitude toward a transgressor against the first than one does
against a transgressors of the 2nd".

Another (more basic?) difficulty may be that, pace H P Grice, 'if' may be just material and no 'link' inovlved, moral or other?

"Yet we intuitively think of a person who accepts 1-2 but not 3 as having
inconsistent BELIEFS rather than some sort of defect in 'sensibility'. Most of the arguments re ME, both sympathetic (Blackburn, Gibbard) and critical (Wright, Johnston) have centrered around whether it provides a plausible account of the 'logical form' of a moral utterance, one supporting the intuition that an inference like Geach's IS valid. Blackburn defends a PROJECTIVIST construal of a moral utterance as "mimicking" a truth-evaluable utterance. Asking us to imagine a moral utterance on the model of a calculus consisting of truth-evaluable utterances together with all of the classical truth-functors as well as a pair of intent[s?]ional operators:

Hp (Hooray for p)

and

Tp (It's tolerable that p).

He then provides rules for the interpretation of formulae containing H and T based on model-theoretic principles familiar from possible worlds semantics.
MS writes 'intentional', which R Helzeman has rephrased in his comments as 'intensional'. I wonder btw what's exactly projective about 'projectivism'. HP Grice speaks of The Humean Projection, but I guess *this*
is different?
"According to Blackburn an INCONSISTENT utterance ["Hooray for p, and if booray for p then hooray for q, yet not-hooray for q". Or: Non-hooray for eating people. If Non-hooray for eating people, then non-hooray for eating Jones. Yet Hooray for eating Jones] is subject to logical as well as psychological censure, because such an attitude would make it impossible for an agent to achieve a 'consistent realisation of aims'. Yet this suggestion has NOT won a great deal of widespread support. Two problems are obvious. 1) because the constructed language itself contains all the truth-conditional connectives (notably 'if') its in-principle comprehensibility clearly depends on there being a near-univocity with English. But the meaning of 'if' in English IS traditionally supposed to consist precisely in the roles it plays in a TRUTH-preserving inference. To the extent that ME commits to
truth NOT providing an appropriate criterion for the assessment for a moral
utterances, such proposals must be viewed as hostage to the fortunes of a highly controversial thesis in the philosophy of language; to wit, that one can provide an adequate account of consequence that makes no indispensable use of TRUTH".

Well, Blackburn only said 'mimicking', didn't he? Perhaps one is 'mimicked' (Moral) Truth. And then, there's also the parallel Oxford tradition of H P Grice, A J P Kenny, J D Atlas, et al & their logic of 'satisfaction', with satisfaction as a neutral term covering boulomaic, doxastic, alethic and deontic versions?

"A second obvious problem is WRIGHT'S DILEMMA. He takes issue with ME's claim to have provided a REDUCTIVE analysis of a moral utterance. Success for a projective proposal can 'only consist in establishing rules of transformation from an assertoric moral utterance to a NON-Assertoric moral utterance. A dilemma arises: are the rules guaranteed to be content-preserving? If not, nothing will have been shown. If so, how is it supposed to be clear that no assertion is made in the reductive utterance?

If assertoric syntax can mislead, so can non-assertoric syntax. Even if the language used to reveal the structure of a moral utterance is something we have stipulated into existence, as soon as we come to make use of such a language we must reckon with the possibility that we ourselves are being deceived about its content by the appearance of its surface syntax. CW suggests the only acceptable way out of the dilemma is to abandon trying to construe a moral utterance as NON-truth-evaluable, proposing instead a method for accommodating the intuitions that motivate ME (eg the unmediated influence that a moral utterance has on motivation"

Shouldn't that be the other way round, with motivation influencing a moral utterance? Or perhaps you mean the motivation on the recipient of the moral utterance? Anyways, btw, Wright's argument re surface syntax is less than crystal clear to me. One good thing about a surface syntax transfomation is that one does HAVE a criterion to say a certain feature is there or not.

"and the supposed absence of an irreducibly moral state-of-affairs from the natural world [The is/ought question?]. One should pay this respects to these intuitions about a moral utterance by providing its 'SUBSTANTIAL TRUTH-CONDITIONS' with an essential reference made to a response such as approbation, enthusiasm, or disapproval. It's worth mentioning that neither Blackburn nor Gibbard describe the relation between a moral utterance and their constructed languages as being straightforwardly REDUCTIVE in nature.

Gibbard explicitly rejects his theory of expressivism is a reductive analysis of moral utterances. Blackburn's view is more equivocal and involves 'fast' vs 'slow track quasirealism. Hooray for Blackburn's 'Hooray'. What, with C Wright's 'substantial [how much so?] truth-conditions', and having recently experienced with Analytic some of Dummett's caveats with of canonical truth-conditions (and his proposal of Assertability Defeasible Conditions in Non-Bivalent Contexts, instead) I tremble at Wright's suggestion.
"Wright's suggestion HAS attracted widespread support among those whose moral views often have little else in common (eg R Boyd, and J Mcdowell have both defended a TRUTH-CONDITIONAL account of a moral utterance. eg McDowell's 'Values & Secondary Properties' in Morality & Objectivity, ed T Honderich, London, RKP). My aim here is to argue that there are powerful reasons why ME better reject a truth-conditional interpretation of a moral utterance. I shall defend a PROJECTIVISM re moral assertion and inference on the basis of the radical interpretation (henceforth RI)"
I always wondered what IS radical about RADICAL interpretation, anyways.
I've always taken Davidson to be meaning the 'native' (Americans, of 'Gavagai' fame, so radically unlike us). But given that such intepretation may well occur within one's language, I fail to see the radicalism of it all? Perhaps this is connected to Gibbard's Radical Interpretation of his own thought?
"The support Projectivism gets from RI does NOT depend on the success of a
coherent normative 'logic of attitudes' ala Blackburn or Gibbard. It is important to try & get as clear a sense as possible of precisely what ME has to GIVE UP in order to accept the prima facie harmless suggestion that there is at least a sense in which a moral utterance can have a DETERMINATE TRUTH-CONDITION. Here it will be helpful to call to mind some of the more important historical antecedents of ME. It is widely acknowledged that the most important historical forefather was D Hume. His insistence on a distinction between the influence of reason and passion on motivation..."
Shouldn't that rather be 'on action' rather than 'on motivation'? (I'd take motivation to be a sort of passion). Also, isn't it interesting Hume's use of 'reason' here? MS's definition of Expressivism lists 'rational' is as yet another evaluative term, along 'good' and 'right'. So I guess either MS or DH is wrong?
"has been seen as leading more or less directly to the view that a moral
utterance is best explained by appeal to passion. If a moral utterance is NOT TRUTH-EVALUABLE this can only be because passion cannot properly be thought of as being directed toward attaining as many true beliefs about the world as possible".
I guess this may need some qualification? After all there IS a connection between a DESIRE (including your morally good intenSion) and the 'true' predicate (cf MS's remarks on the optimism of the prophets in his reply to M Murphy). If I don't desire to eat people (ie, if I hold that eating people is wrong), then I guess I hold (by definition) that it's preferrable that it's FALSE (not true) that people are going to be eaten.
"It's fair to speculate that for someone who comes to ME the suggestion that a reference to 'response' is simply built into an account of the TRUTH-CONDITIONS of a moral utterance may be perfectly acceptable. So long as one acknowledges the impossibility of assessing a moral utterance without reference to 'passion' upon thought there would be no impediment to extending 'true' to a moral utterance as being a matter of mere convention. Those who have been impressed however by G Moore's 'open question' (in Principia Ethica, C Stevenson, R Hare) have been attracted to ME for quite different reasons. ME has often represented for those the best way to explain the recalcitrance to definition that Moore thinks is exhibited by 'good', 'the 'unanalysable predicate of value'. Now, it's highly doubtful that ME would accept a description of 'good' as an unanalysable predicate standin[irreductive, non-supervenient] and yet accept a characterisaion of the moral utterance's TRUTH-CONDITIONS in terms of a disposition to 'respond' to phenomena in one's natural environment. ME should NOT require any views about what precise passion or 'responsive' disposition is 'expressed'.
Well, I guess there are variants, although they may fall within your
Reductive (Reductionist) Analyses (BTW, I owe the reductive/reductionist
distinction to Grice's Studies in the Way of Words. Not everything reductive
is reductionist).

What about the following substantial truth-condition. 'Hooray for p' is true iff Utterer hoorays for p ('Hx' is true iff Utterer Expresses Attitude H towards p). Now, of course, an asymmetry would result with your standard non-moral assertion, as 'Snow is white' would be true (contra Tarski) iff speaker BELIEVED that snow is white, rather than iff snow is white!

"The most wholesome lesson to be learned from this is that in seeking to understand the meaning of a moral utterance one should strive a NEUTRAL attitude with respect to the attitude of the moral passion. My insistence on the separateness of these two traditions within ME might initially seem a rather shallow artifice of historiography. In Blackburn and Gibbard we do after all find evidence for a strong commitment to Humean BELIEF/DESIRE PSYCHOLOGY AND to a Moorean scepticism about the possibility of defining 'good' [or 'right']. But the question whether a semantic distinction is drawn on the basis of independently arrived-at theses in psychology has been a divisive one among philosophers of language".
If the obscure historicist may himself say so, I think Silcox's historiograpical construal is very correct (Hooray for it). As for your 'divisive' supposition, I guess I'm glad I know where I stand. Cf Grice, author of 'Meaning' and 'Method in Philosophical Psychology in (respectively) Studies in the Way of Words & The Conception of Value. But now let the radical interpreter enter.
"Now, regarding this dispute of how much psychological meaning is, arguments about the possibility of RI have been central: Quine, Davidson, Rorty & Lewis have argued against the intrusion of what they take to be illicit psychological considerations into semantics, trying to provide support for their views by describing a situation in which understanding is achieved in a limited epistemological circumstance"
One _may_ disagree. I, for one, always thought Davidson's motivation was trying to prove the interrelation of belief and meaning (if not, a la Grice, the primary of the former) rather than forbidding belief 'intrude' with meaning. I'm less sure why Quine introduced his discussion of 'gavagai' in the first place. Also, I guess the semantics/pragmatics distinction within the 'theory of meaning may apply here, discussing the meaning of moral utterances: psychology may be said to be irrelevant to semantics but is said to be CONSTITUTIVE of pragmatics.
"Davidson describes RI as one in which one 'CANNOT hope to take as evidence for the MEANING of an utterance an account of the delicately discriminated INTENTIONS with which it is typically uttered'. None have speculated in any kind of detail however on the extent to which a moral utterance can be thought of as being radically interpretable (Gibbard does make the mysterious remark that the aim of his representation 'normative content' is to solve a 'problem' with "the RI of a person's thoughts" (Wise Choices p94). He says nothing at all however about the significance his solution to this 'problem' might have for satisfying the need of a clear account of the relation between semantics and psychology). Yet, by asking
what the link might be between these two quite disparate tasks (the RI of a straightforwardly nonevaluative fact-stating utterance vs an moral and 'expressive' one) we may remove some of the obstacles that block ME from adopting PROJECTIVISM".
It's not clear what difference there ultimately is between Expressivism and Projectivism. They seem to mean pretty much the same to me. Aren't all expressivists, by definition, projectivistic? "One plausible answer would be that a moral utterance could simply NOT be understood in the epistemically restricted circumstance of RI. This denial seems however only plausible to someone who is addicted to a peculiarly extreme or dogmatic form of ethical relativism".
But one thing I was told a Relativist is NOT is dogmatic, is she? I don't see the connection you point to. I take it NIHILISM may be a better label (or agnosticism in the case of religion?), or better, Ethical Inescrutability. I guess the issue is tricky because one can endorse RI and very well be an Ethical Relativist: RI is not concerned with ACCEPTING the native's moral utterances. Only with trying to
UNDERSTAND/interpret them (cf Habermas's dialogue logic).
"A more reasonable approach might be that RI is in a position to engage in
well-founded, detailed speculation about the informants' beliefs..."
AND DESIRES. I guess the addition is relevant when talking of Expressivism/Projectivism. cf Eating people is not wrong. Does your reluctant cannibal believe that, or is it a mere expression of his desires? In his reply to Murphy, MS wonders if there's an equivalent to CHARITY for moral utterances. Well, it seems CHARITY (as per Christian charity) is all about MORALS, rather than epistemics (Forget about Davidson).
"...no later than one could begin the task of figuring out what substantial view the informant may have about right and wrong [or good and bad]. Here it's worth noting that one of the few features shared by most accounts of RI is that in order to even get it off the ground, it must be possible to recognise antics in the informant's behaviour that manifest a pro/anti attitude towards an assertoric utterance [eg 'Gavagai?' Cf Blackburn's 'Hooray for Gavagai'].
"Now, there are different ways in which one could complicate the picture. One such would be to develop different criteria to identify instances of moral vs nonmoral assertoric utterance (Interpretational Dualism, henceforth ID). An alternative strategy would be that behavioural patterns whereby the utterer manifests the attitude associated with both moral and nonmoral assertoric assertion may well be perfectly INDISTINGUISHABLE (INTERPRETATIONAL MONISM, henceforth IM)."
I guess H P Grice's lectures on Aspects of Reason and Reasoning may be
useful here (cited in The Conception of Value). I guess one may need to distinguish an 'attitude-less' ALETHIC modus ponens (p, if p then q, thus q) from a doxastic (A believes London is nice, A believes nice is true, A believes London is true), a BOULEMAIC version (A desires London. etc), and a deontic one. I also guess one may pretty well mix desire and believe (in a reasoning or discoursive pattern), even if you behaviourally manifest them differently (I believe the cat is on the mat, I want it on the bed though. Thus I believe that a bed is not a mat). I guess the question here may also concern PRIVILEGED access (so-called), rather than mere RI? Cf Eating People is wrong. Either you feel it or you don't).
Etc.