In my last post I began speaking about something that is more important, I think, than people realise, and that raises all the thorny issues of the foundational nature of science, as well as questions about scientism, which I will interpret here as the belief that the only rational foundation for our beliefs is to be found in one or other of the positive sciences, especially physics, chemistry and biology. The social sciences, dealing much more intimately with questions of interpretation, while attempting to provide empirical foundations for their conclusions, are much more closely involved with questions of interpretation, and therefore to the kinds of uncertainty built into our interpretive conclusions. The reason this increasingly seems to me an important field for investigation is simply that some of the internet disputes that involve questions as to how we can underwrite our conclusions increasingly seem to give favoured status to the positive sciences as the only pursuit likely to provide the kinds of certainty we are seeking, especially when we are dealing with a widespread human pursuit that, on the face of it, seems simply to be swallowed up by interpretation, and to provide no point at which it touches down in any determinate way in human experience.
By referring to something that is simply swallowed up by interpretation, I am referring to theology, of course, but within the scope of that term may be included ethics (or moral philosophy), and philosophy itself, which, for many internet atheists, is assumed to be a largely useless pursuit closely related to theology, because providing almost nothing in the way of empirical (observational) foundations. And while most new atheists think (rightly) that it is possible to be good without god, very few seem to have an understanding of morality that distinguishes it from emotivism (the expression of feelings and personal preferences) and relativism, which makes the question of goodness without god largely irrelevant, since moral goodness can be reduced, without remainder, if emotivism is true, to subjective feelings of value.
At this point I want to quote something from a commenter on my last post, Daniel Lafave, not because, I hasten to add, I want to hold it up to ridicule, but because it expresses in a fairly narrow compass where I think some of our problems lie. He had just said, to do him justice, that philosophy has an important role in clarifying the theoretical conclusions to which scientists arrive, based on observational evidence, and that, without such clarification, observational evidence is (or at least may be) of little value. And then this:
The converse problem in philosophy is that too many philosophers think that philosophy has special methods of justification other than evidence, either intuition or conceptual analysis, by which claims can be justified independent of evidence. For some people this is what makes philosophy different from science. I think that’s just absurd. Intuition is just a word for a mental itch, and conceptual analysis at best just reveals something about our existing conceptualization but nothing about the world we are trying to study. The good thing is that this intuitionist perspective is slowly dying out in philosophy. Fewer and fewer people think their job can be done while ignoring evidence.
The problem here lies in the reference to “special methods of justification other than evidence,” and then the very rapid segue to what these methods might be, whether “intuition” or “conceptual analysis,” as though that exhausted the field. And this in response to a post in which I introduced the concept of “interpretive evidence,” because it has begun to seem to me that many of the disputes over “ways of knowing” are getting hung up at precisely this point.
The problem here is that the word ‘evidence’ does not provide us with criteria of what will count as evidence. Indeed, what will count as evidence will differ, depending upon what kind of investigation we are undertaking. I think that Daniel Lafave has in mind empirical evidence, the kind of evidence that can be discerned by the senses, and able to be repeated in systematic ways in the context of experimentation or observation. That is what I think he has in mind, but it is not what he says. There seems to be a sense that it is somehow obvious what constitutes evidence, and that, when we have it, we are somehow observing the world without the mediation of theory, what Ronald Dworkin in his book Justice for Hedgehogs calls “bare truth,” as though bare truths were simply “out there” in the world waiting to be observed and recorded.
The truth, however, seems to be otherwise. There are no “bare” truths to be observed, for we observe things always through a network of theory. As Kant says (and I’m not sure that I’ve got the words exactly): “Experience without theory is blind, and theory without experience is empty.” Or as some philosophers of science have said, it is interpretation all the way down. We never get to a level of observation which is not already somehow encapsulated in theory, and when scientists make new observations, and claim to have discovered new knowledge about the world, all the theories up to that point — since science is composed of sets of models (or theories) — shift (sometimes ever so slightly) to make room for the new information that is captured by that new knowledge (obtained by observation or experiment). Which, of course, I might add, is what makes creationist nonsense pure nonsense, since creationists do not engage with the theories, but think that their convictions are enough to eliminate the theories or models of science, in terms of what the sciences say makes sense, as a whole, a procedure which is quite simply absurd. You cannot shift the totality of science by sniping from the outside, because, when you try to do so, the models remain unaffected, and the truths of science remain intact.
Now, it’s at this point that the question of interpretive truth arises. First of all, science already deals with interpretive truth. New observations in biology, for example, which, let us say, call some of the accepted findings of biology into question, take place within the scope of the models in terms of which observations make sense. Since I know virtually nothing about theories in evolutionary biology, I can take, as an example of the kind of thing I have in mind, a recent article which claims that evolutionary theory is in crisis. (See John Dupré’s “Evolutionary Theory’s Welcome Crisis.”) It doesn’t matter, for my purposes, whether Dupré is right or wrong about the crisis in evolutionary theory. The point is that, from the point of view of theory, crisis must always be welcome in any case, for crisis means that the theory has to be expanded in order to accommodate new information; and as it does (that is, if Dupré is right), the rest of evolutionary theory will have to shift to accommodate the new information, and the models that are used to identify evidence as evidence will undergo change and expansion in the process. The end result will be that knowledge will increase and our understanding of the processes involved will be enhanced.
Now, it is just at this point that the question (and sometimes the accusation) of scientism arises. Daniel Lafave wants to say that we should restrict the term ‘scientism’ to its pejorative role as a term of abuse, and substitute, in the place where scientists are claiming the priority of scientific “ways of knowing,” the term ‘evidentialism’ instead. From the point of view that I am developing here, this would not really help, for the question of what constitutes evidence is to a large degree still dependent upon interpretation and model construction. The model-dependent realism that Hawking and Mlodinow highlight in The Grand Design as their particular way of accounting for their ontological commitments, is still an interpretation, a model in terms of which what will count as evidence is determined, and much as they might like to think that this is not a philosophical (or even metaphysical) commitment, the truth is that they do not have evidence for the model itself; for it is the model itself that determines what will count (for the model) as evidence. However, within those prescribed limits, science does deliver the goods, a kind of certainty and stability that other disciplines can only dream of. And even when they disagree, as Dworkin points out, differences of opinion in the sciences are “generally small compared to what [all scientists] think in common.” (155) And it could scarcely be otherwise.
But this does not make disciplines that depend more substantially upon interpretation completely different ways of knowing, where ascriptions of truth or falsity no longer apply. As Dworkin points out, the claim that whereas
scientific claims are true or false, interpretive judgments are something different … [--] sound or unsound, or more or less reasonable, or something of that sort. 
“These distinctions,” he points out, “are empty.” We could of course limit the use of ‘true’ and ‘false’ to scientific judgements, “[b]ut that stipulation would be pointless because we can claim no utility for it.” (loc. cit.) And then he goes on, in more detail, to say the following:
We cannot map the distinction onto any more familiar distinction by explaining, for instance, that “true” indicates objectivity while “most reasonable” indicates only subjectivity, or that “true” marks a cognitive judgement while “most reasonable” marks some form of noncognitive expression. [151-2]
On the contrary, both the expression ‘true’ and the expression ‘most reasonable’ signify a special kind of unique success, and though in each case the content may be different, both claims may rightfully be thought eligible to be accorded the accolade of truth, and to qualify, therefore, as objective.
This does not mean, of course, that the stability of the kinds of truth involved in each case is the same. Interpretive truths may be less secure, and, in a way very different from the judgements of science, their justifying goals and intrinsic goals merge. The justifying reasons for carrying out a particular scientific investigation are irrelevant to the conclusions reached, but the reasons justifying an interpretive inquiry are not independent of, and may bring to bear concerns which in fact merge with the intrinsic goals of the interpretive process. This happens in historical inquiry, for example, as is demonstrated with exemplary clarity by Richard Evans’ account of his participation in the defence of Penguin Books and Deborah Lipstadt against the charge of libel made by David Irving, who defended himself before Mr. Justice Gray against the accusation that he was a Holocaust denier, and a falsifier of evidence. As Evans shows in his book Lying about History: History, Holocaust, and the David Irving Trial, the reasons for addressing the question of the justice of Lipstadt’s accusation against Irving, merged seamlessly with the intrinsic historical interest of the understanding and interpretation of the events commonly brought together under the description “The Holocaust”, as well as with the goals and procedures of the historian’s craft. The book itself is a fascinating insight into how the particular interpretive discipline known as history functions to provide the truth about the past, and how that discipline differs, in important respects, from what is understood under the term “natural science”.
It is, in fact, the fudging of this distinction between interpretive and scientific discourse and investigation which largely vitiates many of the claims made by Sam Harris in his book, The Moral Landscape. Hidden in one of the first footnotes in the back of the book, Harris says that
[f]or the purpose of this discussion, I do not intend to make a hard distinction between “science” and other intellectual contexts in which we discuss “facts” — e.g., history. … I think “science,” therefore, should be considered a specialized branch of a larger effort to form true beliefs about events in our world. [fn 2, 195]
This merging of science and other forms of inquiry which “form” true beliefs about “events” tends to conceal the great difference between the methods of science, on the one hand – the word ’science’ usually understood in English in terms of what are often called the “hard” sciences of physics, chemistry, and, increasingly, biology – and interpretive disciplines like history, on the other, where much more than just “events in our world” are subject to exploration and interpretation. Indeed, the statement in which he forces these different kinds of inquiry into his Procrustean bed, so that, in some sense, he can make the claim that there can be a “science” of morality, is itself an interpretive move in the game of trying to understand what we know and how we know it, which cannot itself stand as a hard datum on the basis of which other things can be determined to be true or false. Without addressing the presuppositions included in what amounts to a foundational claim whose justification is never explored, and is thus more or less a cuckoo’s egg in the nest presided over by philosophy, an interpretive discipline for which Harris shows some contempt, his programme of creating a “science” of morality is hopelessly compromised.
This is why it is important for us to listen to philosophers like Philip Kitcher, whose New Republic article, “The Trouble with Scientism: Why history and the humanities are also a form of knowledge ,” raises serious questions about the self-sufficient nature of science to deal with the issues that face us. For Kitcher, the problem is that it simply leaves too much out of account, and by denying the rational nature of interpretive disciplines like history, literary and aesthetic criticism, ethics, and so forth, our ability to deal rationally and adequately with so many of the problems that face us is severely straitened and limited. That does not mean that commentators like Jerry Coyne, for example, do not take literature and music with great seriousness, for he clearly does. What it means is that, by failing to acknowledge a role for reason, and therefore for truth, interpretive truth, to distinguish it, if that seems necessary, from the type of truth to which science lays claim, the limits of knowledge are seriously impoverished, and the humanity, the humanness or humaneness, if you like, of our conclusions is threatened. The trouble with scientism, for people like Kitcher and Pigliucci (who also deprecates scientism), is that it tends to make science the only arbiter of what constitutes the truth. Here is how Pigliucci defines scientism:
Scientism. This is the pernicious tendency to believe that science is the only paragon of knowledge and the ultimate arbiter of what counts as knowledge.
The point that I hear both Pigliucci and Kitcher making is that to confine knowledge to the conclusions of science is to ignore so much that is simply, as Kitcher says in his New Republic article,
too large, too complex, too imprecise, and too important to be addressed by blundering over-simplification.
This is actually a matter which he has discussed in some detail in his book Vaulting Ambition: Sociobiology and the Quest for Human Nature, where he addresses himself to the oversimplifications that happen when very complex human systems, interpretive systems, to stick with that term, are forced into a scientifically explanatory context without dealing adequately with the interpretive issues that inform them.
To take but one example, Kitcher takes E.O. Wilson to task for his naivety when it comes to his understanding of ethics, and his failure to take into consideration important dimensions of ethical discourse, apparently because he does not understand it:
There is no subtle new argument [writes Kitcher] for debunking ethical objectivity. Wilson reaches his conclusion by ignoring the serious alternatives. He ignores them because, apparently, he does not understand them. 
Nonetheless, Wilson goes on to make claims about providing “the physical basis of moral thought and considering its evolutionary meaning,” which will, he thinks, enable people to control their own lives. But, as Kitcher justly asks:
If there is no sense to moral correctness, then what exactly is meant by the claim that greater self-knowledge might place us in a better position to choose ethical precepts? 
The point is important. Wilson simply dismisses the serious alternatives to the kind of emotivism he favours, and, in doing so, empties of content any suggestion that there is a basis upon which control of one’s life can be premised, precisely the kind of interpretive understanding that he has already so quickly dismissed. And the point is that the interpretive context of human action cannot simply be ignored by over-simplifying the context in which decisions are made. It is this kind of impoverishment that Kitcher and Pigliucci are concerned about in their anxiety about scientism. It seems to me that some of this anxiety is justified, and should be taken with greater seriousness, especially by the new atheist community which, with some justice, emphasises the success of science in contrast to the failure of religion to provide foundations for its various beliefs in supernatural beings, revelations and guidance, but then includes as quasi-religious other aspects of the human project which include sources of knowledge and understanding which cannot so easily be subsumed under the rubric of science.