Hilary Putnam: Objectivity and the Science–Ethics Distinction
Hilary Putnam: Objectivity and the Science–Ethics Distinction
Abstract and Keywords
Krüger evaluates whether Putnam succeeds in showing that unified concepts of objectivity and rationality can be developed for science and ethics, questioning whether it is desirable or advisable to insist on the ideal of the unified scientific–ethical enterprise. Krüger objects to Putnam's conclusion of the existence of a unified domain of rationally assertable truths because he sees it as a sweeping extrapolation of the correct observation that entanglements of fact and value are inextricable. Krüger further claims that Putnam's renewed attempt at unifying human reason would work against cultural pluralism, which is not desirable.
What I admire and appreciate very much about Putnam's paper on the science–ethics distinction—indeed about his more recent philosophical work (1981, 1987) in general—is the fresh and original attack on a key problem of the Western scientific–philosophical tradition: the problem of establishing a unified and coherent cognitive enterprise which comprises values and facts, decisions and insights, ethics and science alike. If he should succeed, we would—vis‐à‐vis the problem of quality of life—be in a very different situation, because we could settle disputes on values by co‐operative investigations rather than controversial negotiations.
The commentator, it seems to me, must inevitably ask the following critical question: has Putnam succeeded in showing that unified concepts of objectivity and rationality can be developed for science and ethics? Indeed, is it even desirable or advisable to insist on the ideal of the unified scientific–ethical enterprise?
In pursuing these questions, I shall not adopt a strategy that might seem to be suggested, if not demanded, by the arrangement of Putnam's argument; that is, I shall not try to examine the possible defences of the opponents he attacks in his paper, notably Bernard Williams. I cannot do this, simply because I happen to accept Putnam's main argument against Williams. Like him, I think that the distinction between relative and absolute truth (in the sense he attributes to Williams) is untenable. But, unlike him, I see different reasons for this untenability; hence I am led to adopt different conclusions—conclusions that are at variance with the idea of a unified sphere of objectivity for science and ethics. Leaving other aspects of Putnam's paper aside, I turn immediately to the disputed distinction between the relative and the absolute.
Relative truths are such as will be accepted by some community or other of rational believers but remain dependent on certain limited cultural contexts. (What the limits of these contexts and of these cognitive communities are remains an open question to be explored for each case in its own right.) Absolute truths, however, are context‐independent; and this they are said to be by virtue of their correspondence to the things ‘out there’, or, more precisely, by their corresponding to such features of reality as could invariantly and equivalently be expressed in any other sufficiently elaborate language.
Favoured candidates for the status of such features are, of course, suitably chosen objects as described by physics. Candidates for relative truths, on the other hand, are ordinary statements of everyday life such as ‘the grass is green’, ‘our neighbour is a malicious person’, or ‘the tax laws are just’. What counts as ‘grass’ or ‘green’, let alone as a ‘malicious character’ or a ‘just tax law’, is (p.159) hardly culturally invariant, though such statements may be recognized as true or false, once a cultural context is taken for granted.
Now Putnam opposes this view for at least two important reasons. First, he thinks that the conception of absolute truth is untenable, indeed unintelligible; second, he thinks that relative truth is less than we need and less than we can attain. I fully agree with both points.
My misgivings with Putnam's approach concern his views on the connection between his two critical points and the conclusions he derives from this connection. In my reading of his paper the connection is the following: If you grant the first point, that is, if you drop absolute truth and the correspondence between beliefs and things out there altogether, you are left with nothing but relative truths. Yet these — as far as their epistemic status is concerned — have already been granted to form a homogeneous field. That the neighbour is a malicious person is a truth, or a falsity, of the same kind and epistemological status as the assertion that the grass is green. Putnam, his opponents, and further authors (some cited in his paper, e.g. Iris Murdoch), have accumulated an impressive sample of statements that do not admit of an analysis into a ‘factual’ and an ‘evaluative’ component. The examples range from recognizing a person as pert or polite to condemning human sacrifice as utterly evil. All these cases display an inextricable ‘entanglement of fact and value’. Nevertheless, they are all plausibly presented as acceptable candidates for the assignment of a truth‐value, though Putnam does not deny that the actual ascription of either truth or falsity to such assertions about value‐facts may depend on comprehensive ideologies and forms of life, for example, the Aztec religion, which demanded human sacrifice.
We can now see how the rejection of an admittedly ill‐conceived and untenable partition of the set of truths combines with a positive characterization of those that fall on the relativistic side. As far as Putnam's present paper goes, it is this combination that leads on to further conclusions, conclusions which I believe to be unacceptable. The most important of these is the desired unity of our cognitive field. Another important conclusion (which, if true, would support the first) is the following: given that there is no absolute truth because there are no fixed objects ‘out there’, the notion of correspondence between statements or beliefs and transcendent ‘facts’ becomes an empty or even an unintelligible concept. It is replaced by rational acceptability.
Here are my objections. The first conclusion — the existence of a unified domain of rationally assertable truths — is an all‐too‐sweeping extrapolation of a correct observation: that of inextricable entanglements of fact and value in many life issues. This extrapolation in its turn is encouraged by the second conclusion, which is that truth can never consist in correspondence to external facts. To this I object that the inference from ‘no correspondence to something fixed out there’ to ‘no correspondence at all’ is a non sequitur. Now, if there were some kind of correspondence in some cases and no correspondence of any kind in some other cases, a new form of the fact–value distinction and (p.160) of the ethics–science distinction might be established — a form that is freed from the metaphysical burden of absoluteness.
I take up the second point first, because it lies close to the centre of Putnam's theoretical philosophy, that is, his internal or pragmatic realism. To avoid misunderstandings, I want to stress that I second Putnam's criticism of external, absolute, or metaphysical realism, and that I am all for internal, knower‐related, or pragmatic realism. Moreover, I share Putnam's hope of nevertheless resisting a sweeping cultural (let alone subjective or personal) relativism. I part company with him, however, on at least two issues. First, I do not believe that we can secure a non‐relativistic realism without some notion of ‘correspondence’ to external reality. Second, I do not believe that all rationally acceptable propositions have the same epistemological status, rather that some are acceptable because of their correspondence, others for entirely different reasons. The first kind can be called ‘true’ (or ‘false’) in the standard meaning of the terms, the second kind cannot. If truth talk is applied to the second kind, this has to be done by tactful extension of, or on analogy with, the first type. But there are limits to these extensions or analogies, limits that are to be assessed in a rational analysis. In other words: I think that the field of rationality is much wider than that of truth, and that this circumstance is of vital importance for the survival of rationality.
What is correspondence if it is not to be absolute or metaphysical? It is a relationship between our statements or beliefs, on the one hand, and our perceptions and actions, on the other. Both relata are experiential and free from metaphysical suspicion. Take a simple example, for instance, the statement that there is now an even number of people in this room. If it should happen to be true, that would mean, among other things, that you could pair off all people present, so that no one is left alone. Here is an obvious correspondence between our verbal or mental representation and our material, that is, non‐verbal and non‐representational, contacts with a piece of reality.
Now, it may conceivably be the case that we could have a system of representations that lacks the conceptual distinction between ‘odd’ and ‘even’. One may, therefore, argue — and this is quite in line with internal realism — that even my banal example illustrates the fact that our conceptual system (presumably together with our imaginative possibilities) irrevocably prejudges what can ever become an object for us. But it is equally clear from the example that it would be absurd to collapse the doubleness of representation and material contact.
So far everything seems trivial. But perhaps the next step is already less innocuous. We all believe that people come in countable integer units and that they cannot penetrate walls, so that it lies in the nature of things — that is, of our material contacts with the world — that there is actually no choice between alternative representational systems when it comes to the question of how many people have gathered in a closed room. Statements concerning such matters are simply true or false; and they are that, first, because of a certain (p.161) correspondence with reality and, second, regardless of the cultural context, provided it is developed enough to discuss such matters at all.
Again, no more than a trivial point? Yes, as far as normal people's concerns go. But no, as soon as subtle philosophy is brought into play. What I am implying is no more, but also no less, than an assumption to the effect that the niceties of a logically refined ontology are completely irrelevant to our problem of correspondence. It is part of our material world contact that we do not encounter, and cannot count, time‐slices of people, or experience them as portions of some stuff, of a kind of human mush. Nor do we have a choice between an individualistic ontology à la Carnap and a mereological ontology according to the Polish logicians (Putnam, 1987: 17–21) in this case. Brevity forces me to make an unduly perfunctory statement here: correspondence, and with it the reference of terms, does not suffer in the least from ontological relativity, at least not in our humdrum examples.
But how far do they carry us into more complex and more abstract matters? The lack of serious alternative descriptions smacks of absoluteness again. But here we must guard against the philosopher's professional disease, of turning universal experiences into conceptual necessities. I take the lack of serious alternative representations of simple matters to be no more but also no less than a contingent, if enduring, fact about those matters and about humans who take cognizance of them. Keeping this point in mind, let us now turn to a different example, for instance, the proposition that all the people in this room are independent and honest thinkers. There is — we will probably concede — some fact of the matter concerning this statement, that is, circumstances that justify its assertion or denial, as the case may be. But it is also clear that this statement is much less stable than the one about the even number, especially across time and across cultural boundaries. Now, I submit that the difference is not just one of precision or accessibility to examination or the like, but a difference in subject‐matter and, by implication, in our relation to that subject‐matter. (Here I find myself in partial, though not complete, agreement with Williams, as rendered by Putnam on pp. 146–56.)
What lies behind this apparent difference? It is not of our doing that humans — or shall I artificially say: instantiations of humanity — come in discrete physical units, whereas codes of intellectual independence, still more of honesty, are products of intentional and, within limits, free human life. True, they are far from arbitrary, and, as with everything historical, any single person will usually feel almost powerless to change them. But (pace Plato) numbers, whether in ‘pure’ mathematics or in experience, and honesty are worlds apart in kind. One may also express the contrast in a Kantian language, using a key term of his philosophy: with respect to some matters we are (or could and should be) autonomous, with respect to others we are not. This is what sharpens the point of the famous confrontation of the moral law in us with the starry sky above us. To be sure, we conceive of both within the limits of our intellect and reason, but we do not impose on ourselves, in Kant's terms, the laws of the (p.162) intellect, whereas we do impose on ourselves the laws of our actions. In other words, what I want to say is that the distinction between facts and values can and should be reconstructed within the framework of internal realism.
In the spirit of the rejection of pre‐scientific metaphysics — which I share with Putnam — I feel driven towards and into history for further instruction and examination of this claim. What I believe I see there is the following pattern (I hope the reader will forgive me for its crudeness, which I cannot avoid owing to lack of space and competence): to build the world of values and forms of life according to the order of the cosmos is a fundamental characteristic of our philosophical–scientific tradition — born with Pythagorean speculation, nourished by Plato (though not, or so it seems, by Socrates), pronounced the guideline of social order by the Stoics, surviving in the Christian idea of the participation of the rational human soul in the divine order of things, finally entering into the optimistic research programmes of the modern philosopher–scientists like Galileo, Descartes, Hobbes, and Hume. But precisely in its modern enlightened form the great programme has come to grief. We can only register, or so it seems to me, a deep disappointment with its outcome. The disillusionment is, first, that no scientific ethics, as projected by the philosopher–scientists of the New Age, has been established; second, that the same holds for a scientific insight into ‘the laws’ of history, if there are such at all. On the contrary, current and recurrent attempts at providing a specifically scientific basis for the responsible and rational investigation of history and ethics — sociobiology or certain orthodox versions of Marxism are examples — are at best helpless vis‐à‐vis our problems and at worst dangerously ideological.
My statement of the historical lesson is, of course, a sketch, no more. It lacks the elaboration and the defences it would need to stand up to scrutiny. But it may help to see what I am up to: in our age of science we cannot but connect even the most elementary statements with their actual or potential theoretical context. But if we try to do this with my two examples — the even number on the one hand and the intellectual independence and honesty on the other — we find ourselves confronted with two deeply different sorts of context: on the one side there is that growing web of scientific theories about that part of our experience that we cannot produce, change, or destroy. Let us call it ‘non‐human nature’ (though it comprises, of course, a good part of our physical and mental apparatus). True, we can exploit this nature, but only, as Francis Bacon said, by obeying it. On the other side there is that part of our experience that we can change, one is tempted to say, that part of it that is us. And here obedience does not seem to be the right, perhaps not even a possible, attitude. We want to have, and to an impressive degree do have, disciplined research traditions and institutions that explore this domain of our experienced reality; but they would be ill advised, should they suppose as their object a human nature in any sense comparable to the non‐human nature which is the supposed object of the natural sciences.
(p.163) For easiness of reference, I may perhaps call the view I have just outlined ‘historical realism’. I think of it as an elaboration, but also as a correction, of internal realism. The name is to remind us that we learn what is real in the course of our history. One lesson — the one I have mentioned — appears to be that self‐conscious and potentially autonomous human beings are a peculiar reality of their own — a reality to which we do not react in the same way as that in which we react or respond to non‐human nature, so that our beliefs or statements about this reality cannot correspond to their contents in a way in which our beliefs or statements about non‐human nature can.
The significance of this lesson, if it is one, hardly needs to be pointed out. If we have a claim to autonomy, narrow as its bounds may be, it is important to recognize that, if for beliefs and statements belonging under its rule there is truth or falsity, then it is not truth or falsity that relates to ‘correspondence’ in the sense outlined above. I do not want to dispute anyone's use of ‘true’ or ‘false’. We may find that all the people in this room are independent thinkers, and we may wish to say that a statement to this effect is true. But here ‘true’ will presumably mean something like ‘rationally acceptable’ or even ‘not rationally rejectable’. Any attempt at ascertaining this kind of truth, however, will involve us in explications of what it is to be independent. These explications, in turn, will involve questions about what balance we want, or ought, to strike between individual originality and mutual social adjustment. In short, the answer to the question as to whether or not we are independent thinkers is not independent from the question as to what kind of people we want to be or what type of society we want to have.
For reasons like these, there can be no objectivity in the domain of autonomy, since there is no object. Instead, there is a world of subjects. Needless to say, I am not here denying the possibility of, indeed the urgent need for, intersubjectivity and rationality. What I claim is only that Putnam's renewed attempt at unifying human reason has not really advanced beyond Kant's great divide of theoretical and practical reason. But how could it have? And would that have been desirable? Besides the by now global unification of our substantial science plus its material technology, we will want to preserve cultural pluralism, an option I share with Putnam (see e.g. 1981: 147–9). It is true, he does not seem to fear that the continuity between science and ethics will work against pluralism. But I do fear that it would work against cultural pluralism, if only it were effectively realizable. And that would not be desirable.
Thus, I conclude that to reach reasonable agreement — or reasonable disagreement, wherever the circumstances leave room for it — is one thing; to claim objectivity or truth is another.