Gibbard on Normative Logic

Citation preview

Gibbard on Normative Logic Simon Blackburn Philosophical Issues, Vol. 4, Naturalism and Normativity. (1993), pp. 60-66. Stable URL: http://links.jstor.org/sici?sici=1533-6077%281993%294%3C60%3AGONL%3E2.0.CO%3B2-X Philosophical Issues is currently published by Ridgeview Publishing Company.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/rpc.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Sun Jun 17 04:43:27 2007

PHILOSOPHICAL ISSUES, 4 Naturalism and Normativity, 1993

Gibbard on Normative Logic Simon Blackburn

Evaluative judgements fit into all the classical contexts: negation and disjunction as well as conditionals, and when we consider evaluative predicates, quantification as well: Anthony did something wrong, or whenever Anthony does something wrong he also does something right. If we call the good behaviour of evaluative sentences in all these contexts the propositional surface of evaluative discourse, then the challenge for the expressivist is explaining the propositional surface, and this is one of the problems Gibbard claims to solve in Chapter 5 of Wise Choices, Apt Feelings. To those impressed by the Frege-Geach problem it is as if there is an abyss between the simple states of mind the expressivist relies upon, and real judgement. Putting it this way also reveals that the problem is an example of the more general problem of understanding the relationship between judgements and simpler things like pictures or representations, such as animals may manage, but which apparently cannot be negative, or disjunctive, or quanficational. Indeed, it was negation itself that Frege used to point out the depth of the abyss. Images and other substitutes for judgements cannot be negated. They are not in the space of logic at all. The same seems to be true of the motivational states or pressures on action that expressivists wish to build upon.

Like many philosophical puzzles this of indirect contexts has a tantalizing aspect. On the one hand it seems as if we habitually allow ourselves a propositional surface even when an expressivist starting point seems almost irresistible: it makes no difference whether I say "Yummy!", or "That's nice!", yet if I say the latter I can go on to disjoin, negate, hypothesize and quantify. So it can seem as if the phenomenon should be easily explicable, and that the Geach-Frege point cannot really mark an important obstacle to expressivism. On the other hand it can seem to be absolutely critical. For expressivism it seems that there is nothing n o proposition or j u d g e m e n t to hypothesize, or no real predicate to quantify over. When one starts off "If killing is wrong.. . " there is nothing being supposed to be the case; no fact imagined being put into place. To make the absence vivid, we can play it through with another example. Suppose we interpret Hume, as I think we should, as holding an expressivist or projectivist theory of causation: one asserting a causal connexion between two kinds of event is dignifying their concurrence in a certain way (e.g. avowing a policy of relying on the concurrence, or integrating it into the system with which to meet the future or conduct practical reasoning). What then is meant by talking of "hidden springs and principles" : unknown causes that escape human scrutiny? It's as if there is nothing there: no fact waiting to be uncovered. Yet if Hume cannot make sense of this range of thought he will have contradicted himself, and his approach fails. I believe that R.M. Hare (1970) first tried to neutralize the FregeGeach objection. But for the last decade I thought I was almost the only expressivist attempting to meet it. The swift victory claimed by Gibbard is therefore welcome but a little shaming; I feel like a n elder climber who has struggled with a pitch watching a younger one saunter up it with no difficulty. But is the victory all it seems? A solution to the Frege-Geach problem must explain, and make legitimate, the propositional surface. But it must do this without invoking properties of evaluative discourse that go beyond the expressivist starting point. The constraint is quite delicate. For example it would not be met if we simply invoked the fact that evaluative judgements can be negated: negation is itself in need of a theory -it is itself part of the propositional surface, and the wrong side of the Fregean abyss. The constraint would not be met if we allow ourselves a notion of consistency and inconsistency, since that too is on the wrong side. All these notions must be built, if they can be, on the more primitive basis. With the problem thus identified, a doubt may arise whether Gibbard has in fact confined himself to legitimate means in solving it, or

whether at some point material from the wrong side of the abyss has been surreptitiously used. In Chapter Five itself my doubts focus upon two ingredients. The first is the very notion of a "system of norms". And the second is the relationship of some of the states of mind needed for the analysis. First, I shall display why these might give rise to doubts, and then discuss whether material from the rest of the work dispels the doubts. The notion of a system of norms is introduced on pp. 86-89. Norms are described as outweighing or overriding others, and systems disagree. Any system can be characterized: by a family of basic predicates 'N-forbidden', 'N-optional', and 'Nrequired'. Here 'N-forbidden' simply means "forbidden by system of norms N", and likewise for its siblings. Other predicates can be constructed from these basic ones; in particular 'N-permitted' will mean "either N-optional or N-required" . When a system of norms "applies in a definite way to an alternative, that results in the alternative's having exactly one of the three basic properties, being N-forbidden, N-optional, or N-required". Is this invoking material from the wrong side? Compare a less compromised starting point. If, instead of starting with acceptance of a system of norms we start with the notion of expressing an attitude, we need to work to generate a notion of consistency or inconsistency. We know (we think) what is wrong with believing that p and believing simultaneously that -p. One belief must be false; the world cannot conform with both beliefs. But what is wrong with commending or approving of p, and commending or approving of -p? Nothing to do with falsity, evidently, so how does a charge of inconsistency get off the ground? After all I might wish that p and wish that -p without logical error. I do not think this problem is insoluble, if we start with simple attitude. The solution, I believe, must rely upon a notion of consistently realizable attitude. If you promote p and promote -p you are in a practical analogue of an inconsistent position, in which not all your goals can be realized. But the problem seems to be only sidestepped if we start off by representing attitudes in terms of a structure defined by notions of consistency and inconsistency. Yet that is what a system of norms is. It is as if in response to Frege's charge that primitive representations cannot be negated we decided that what they represent are systems of facts that themselves already stand in logical relations. Whereas if we describe a creature as having certain representations, thought of as rather like pictures, our description needs the crucial transformation before we arrive at

negatable content, and it is explaining that transformation that is difficult. T h e second point at which we might worry whether Gibbard has appropriated material from the wrong side of the abyss is hidden deeper within the theory. Once we have systems of norms in place, Gibbard can talk of combinations of norm and belief, and a piece of normative content is eventually identified by the set of such combinations that it rules out. Gibbard appropriates possible world semantics to formalize this notion. In the development a statement with normative terms in it "holds" or not at a factual-normative world (w, n ) , and this is defined in terms of whether a related statement in which the normative terms are replaced by a term that n-corresponds with them, holds at (w). The n-corresponding term is purely descriptive, so this last is a classical semantic evaluation. Gibbard gives us an example: Take for example, Cleopatra's claim, Whenever Antony does anything irrational, he sticks to his purpose stubbornly. It holds in a factual-normative world (w, n) if and only if Whenever Antony does anything forbidden by n, he clings to his purposes stubbornly holds in w. By substituting n-corresponding predicates for normative predicates, we substitute a proposition capable of normal evaluation for the expressivist original. The content of a normative statement is then the set of factual normative worlds for which it holds, or, more perspicuously from the standpoint of a theory of meaning, it is given by the set of combinations of normative systems and factual possibilities that it rules out (p. 99). In such an area we must be careful of ambiguity. Naturally, the content of the statement 'Freddy did not behave badly' is not only given by the set it rules out, namely the set of combinations of normative systems and factual possibilities, such that the system applied to the facts delivers that Freddy behaved badly. The content is only given by this set and the fact that the statement does rule them out, that is, that it stands in a determinate relation of negation to them. So the doubts raised already surface here as well, since negation and contradiction come from the wrong side of the abyss. But further we must ask how the full theory actually relates to the original norm expressivism. Consider the disjunction: "either

Freddy behaved badly or Jane did". In saying this I express no attitude to Freddy's actions, and none to Jane's. Equally I do not express acceptance of a system of norms that condemns Freddy's actions, nor of one that condemns Jane's. So what do I do? Gibbard's answer is that I rule out an indefinite variety of combinations of factual belief and norm, such that the norms applied to the beliefs yield both that Freddy did not behave badly and that Jane did not. It's as if I stand ready to oppose any of these indefinite combinations. So we have it that in 'Freddy did not behave badly' I express acceptance of a system of norms that allow Freddy's behaviour, and in the disjunction I rule out systems that yield that Freddy did not behave badly and that Jane did not. Naturally, then, I am committed to supposing that Jane behaved badly. Or am I? What do we say to someone who refuses to hear the wrong combination as ruled out by logic? If we accept (p V q) and rule out p, naturally we must accept q. But that is already describing our state of mind propositionally. If we take the more expressivist descriptions -I express acceptance of a system allowing Freddy's behaviour, and I stand ready to oppose an indefinite combination of systems that allows both their behaviours- it will not be so apparent why it is a defect of logic in me to also tolerate Jane. It might be a defect more like that of being fickle, or impractical. The dilemma seems to be that if states of mind are each described in their own functional terms we lose a logical conception of what is wrong with their combination, whereas to describe them propositionally is already to draw material from across the Fregean abyss. The problem here is the same as one that was urged by critics of my treatment in Spreading the Word (e.g. Hale, 1986). There I suggested a state of mind for someone who asserted a conditional with evaluative components, and argued that it would be a clear defect to assent to the antecedent, assent to the conditional, but refuse to draw the consequent (I called it possessing a "fractured sensibility"). Critics admitted that it would be a defect, but denied that it would be one of logic alone (rather, the defect was like that of being a nuisance). My response was to slightly reformat the way in which assent to a conditional was described, and to provide a conception of logic that drew its fundamental notion of inconsistency from the idea of the realization of a set of attitudes (Blackburn, 1988, p. 508). Gibbard, I think, has two resources to draw upon. One is his general conception of logic. Giving propositional content to someone's states is subject to two constraints. One is to connect them to experience, and the other is to interpret inferences the person

takes as "immediate and unproblematical" in terms of entailments (p. 101). T h e propositional surface of our discourse is there to do the latter. On this account if people simply do readily pass from the one combination of states to another, or to the rejection of another, then a n interpretation that assigns contents which stand in the parallel logical relations is justified. Logic becomes a construction from the pressures we put on combinations of endorsement and rejection. I think this is fine, except that I would add that the utility of a propositional surface is not purely for interpreters. We move to a propositional form because of our own practical needs, notably for the discussion and evaluation of these complex inferential propensities. In my 1988 paper I talked of acceptance of a disjunction in terms of being tied to a tree, that is, being ready to move to the one limb on being removed from the other, and the idea is that when a state has this functional role, for it to be communicated and defended and made public it is expedient to invent a disjunction with this as its content. This gives a "conceptual role" theory of logic, but not a role that is only there to smooth the path of interpretation. The other resource Gibbard can draw upon is to further defend the role that inconsistency (or negation) is to play. I have already queried whether the package which brings it in together with normative system is uncontaminated by Frege's standards. What is needed is a more basic story about the states of mind expressed that makes it plain how they can be candidates for opposition and denial, and thence, by the mechanisms of the last paragraph, why we can naturally and justifiably invent pieces of content -p and -p respectively- as the focus for those oppositions. As I have said, my own attempt to do this includes pressing the parallel between someone with inconsistent beliefs, and someone with inconsistent desires, or unrealizable preferences. Gibbard's section (pp. 289-91) emphasizes rather the pragmatic costs of inconsistency, although he is well aware (p. 288) that the defect of inconsistency does not much feel like a "mere" defect of meanslend reasoning. I marginally prefer my own emphasis, partly for this reason, and partly for a reason Gibbard also acknowledges, which is that it is nice to emphasize the parallel with ordinary beliefs. At least this reminds us that we do not have such a marvellous story (free of pragmatic elements?) about how we cross the Fregean abyss in the case of non-evaluative representations. In all, I do not think there is anything in the theory of Chapter 5 that cannot be saved by the means I have sketched. But it is reassuring to me to reflect that the abyss was crossed so swiftly and elegantly only because other material is available, and can be called upon when span looks too large.

Blackburn, S. 'Attitudes and Contents', Ethics, 1988.

Geach, P.T., 'Assertion', Philosophical Review, 1964.

Hale, R.L.V. 'The Compleat Projectivist', Philosophical Quarterly, 1986.

Hare, R.M., 'Meaning and Speech Acts', Philosophical Review, 1970.