Teleosemantics and meta-ethics

Metaethics is the branch of philosophy that tries to elucidate, not what is morally good or bad, or how can we know it (this would be the function of ethics), but what is the nature of ethics, of moral language, of moral judgments, and of moral attitudes. The scientific study of morals ¹ could be considered as an investigation closely related to meta-ethics, only that the latter would be ‘philosophical’… though from a naturalistic point of view there is perhaps not a sharp distinction between philosophical and scientific research. One of the most important questions in metaethics is ‘what is the content of moral judgments?’, i.e., what do things like ‘though shall not kill’, or ‘incest is bad’, really say, what do they mean? Of course, every philosophical theory of language can be put to the service of these type of questions, but one particular approach which is perhaps more science-friendly than other philosophical brands is teleosemantics. This is the approach followed by Neil Sinclair in his recent paper entitled “Metaethics, teleosemantics and the function of moral judgements”², which is applied specifically to the following problem: is the content of moral judgments descriptive, or expressive?

Moral judgments are descriptive if they can be true or false, i.e., if there is something like an ‘objective moral fact’ the judgment can correctly or incorrectly describe. If descriptivism is right, we could justifiably say things like ‘Peter thinks that gay marriage is morally wrong, but he is really wrong about it’. By the way, descriptivism is compatible with the claim that all affirmative moral judgments are false, i.e., that every time that someone says or thinks something like ‘this action is morally good’ or ‘this action is morally bad’, he or she is mistaken; this would happen if the only objective moral fact were that there are no moral properties, i.e., if all moral properties are fictional. So, moral fictionalism would be a (peculiar) variety of descriptivism. Expressivism, instead, affirms that moral judgments are not like descriptive statements, but more like imperatives. A judgment like ‘murder is wrong’ would mean something like “don’t murder; disapprove murder; and disapprove all those who don’t disapprove murder”. Interminable debates have populated the story of philosophy regarding this question, with rationalist philosophers tending more to the descriptivist extreme, and empiricists more towards the expressivist position.

Sinclair examines whether teleosemantics can be of some help in this discussion. Teleosemantics is the theory that takes meaning as basically a kind of biological fact, something that some living beings do; in particular, meaning would be a type of biological function, on the lines (e.g.) of the following schema: “signal S, ‘consumed’ by agent A, has semantic content p if S has the function of guiding A’s mechanisms to fulfil A’s functions, and p is the normal condition of fulfilment of those functions”³. ‘S’ can be a statement belonging to a human language, but also can be any other state or action that has a semantic function for some plants or animals, like honeybees dancing, the colours of flowers, or the presence of some chemical ‘signals’. In the case of moral statements, the question obviously is what can be the related functions of the human ‘consumers’ of those statements, and what are the ‘conditions of fulfilment’ of these functions. According to Sinclair, moral language (together with moral cognition and moral attitudes) would have evolved in order to promote co-ordination in the case of co-operation problems (as, for example, those social problems that have a ‘prisonner’s dilemma’ or ‘tragedy of commons’ structure):

“It would therefore be biologically advantageous for a group of individuals to have a quick and flexible method of generating, testing and sustaining patterns of mutually beneficial cooperation in response to novel bargaining problems … (It) would be even more advantageous, therefore, is a system that advertises a preparedness to engage in these robust sanctioning behaviours without actually doing so. The ability to form moral judgements makes possible a particular type of shared linguistic evaluation whose ultimate goal is to produce and sustain mutually beneficial patterns of cooperation. In short, the function of the moral habit is to produce and sustain such patterns through a distinctive, linguistic, means of collective problem-solving, viz. moral discussion.” (op. cit., pp. 650-1)

At first sight, this seems to favour the descriptivist position, for after all, saying that ‘stealing is wrong’ would be appropriate if and only if avoiding theaft were actually a way to promote the biological fitness of the social group, i.e., if it were truly a solution to a social dilemma. On the other hand, this view of moral meaning is also sympathetic to expressivism, for the way in which moral language fulfils its biological function is by helping to co-ordinate behaviour. Sinclair, however, leans towards the expressivist endpoint, for, as he explains, in order to talk about the ‘truth value’ or a moral statement, we must distinguish (1) the fact that makes the behaviour recommended by the statement a real, biologically sensible solution to a social co-ordination problem, and (2) the fact that it is morally good to solve this kind of problems. As philosophers know very well since David Hume showed in the XVIIIth century that no ‘ought’ can be logically derive from an ‘is’, one thing is to acknowledge that obeying certain moral norms would, as a matter of fact, solve a particular co-operation problem, and a different thing is to assert that we ought solve co-operation problems. As long as we cannot infer the latter from the former, we are not justified (at least on the basis of the teleosemantic explication of the meaning of moral language) to assert that moral judgments have descriptive content.