Medical Humanities Lexicon: Fact-Value Problem

Sadly, it has been quite awhile since we produced an installment of the Lexicon.  There are myriad reasons for this, but I do admit to a failure of imagination in some sense.  Please, noble readers, do not hesitate to suggest suitable topics for consideration in a future Lexicon!

Today's entry is the fact-value problem, which is perhaps more a genuine philosophical problem than a object of inquiry within the medical humanities.  Yet it is difficult for me to imagine that any serious student or scholar of the medical humanities could be unaware of the implications of the fact-value problem in thinking about medicine, health, and illness.

The fact-value problem arises because of the apparent difficulty in disentangling facts and values.  Both within and without academic discourse, it is extremely common, in my experience, to hear calls to present the facts alone, without bias or prejudice.  This, of course, assumes that such a presentation is in fact possible, and that a phenomenon such as facts exist apart from any bias or prejudice.

There are, however, very good reasons for doubting both of these premises.  These reasons are cross-disciplinary in nature, and I will try to at least mention some of them here, though, as always, thoroughness is difficult if not impossible in blog posts.  One of the most compelling critiques of a brightline distinction between facts and values stems from linguistics and philosophy of language.  Prior to the 1960s, a traditional conception of language was that it was the medium utilized to convey information pertaining to so-called brute facts, propositions that obtain without reference to language.  Oversimplified, this conception posits the existence of a fairly inert world of brute facts, upon which language operates to convey sense and reference, to borrow Frege's term.

However, a number of linguists and philosophers -- Chomsky, Davidson, and Quine, among many others -- pointed out a number of problems with this conception.  Two of these problems center on the indeterminacy of translation (Quine) and/or the problem of incommensurability (Kuhn).  Virtually anyone who speaks two languages will have noticed an interesting phenomenon: some concepts, phrases, and propositions cannot be translated into the other language, no matter how much vocabulary is learned, no matter how fluent the speaker is.  This is a key observation -- some concepts seem untranslatable; they can only be indeterminately conveyed in any language other than the original.

The argument therefrom, which is generally accepted though remains somewhat controversial in some respects, is simply that, far from being a medium which conveys sense and reference about a world of brute facts, language actively creates our world.  To speak a language is to engage in a conceptual scheme -- or, if you prefer Wittgenstein, to engage in a form of life -- that actively shapes the worlds we inhabit.  Thus a particular concept in, for example, Japanese, may reflect a way of conceiving the world that simply does not exist in an American English frame of meaning.  Some concepts cannot be translated.

But if this is so, it follows that the notion of brute, uninterpreted facts seems highly problematic.  Indeed, anthropologists and sociologists, among others, have observed for many decades that our social worlds are irreducibly local and particular.  This follows because it is our linguistic and social practices that shape the world.  Geertz, of course, is the seminal (though hardly the only) source here.

The implications for facts and values should be obvious: if there is merit to the picture sketched above, it is fair to question in what sense facts and values can be separated.  Other philosophers and thinkers, far from philosophy of language, have argued time and again that we are fundamentally interpretive creatures, and that, to borrow from Kantian epistemology, interpreting the world of facts is a condition of our understanding.  Thinkers friendly to the hermeneutic tradition contend that we can do nothing but interpret the local worlds we inhabit, and therefore that the notion of a fact totally independent of subjective values and impressions is itself incoherent.

Similarly, phenomenologists tend to eschew metaphysical questions such as 'are there facts' in favor of focusing on the undeniable point that we do experience and interpret the world, and that it is these experiences and interpretations that are the most interesting desiderata.

However, arguing that facts and values can never be entirely disentangled is certainly not equivalent to arguing that facts and values are the same phenomena.  Denying, a la Wittgenstein, the existence of any essence to "facts" and "values" that would act as necessary and sufficient criteria for each does not imply that no meaningful distinctions can be drawn between facts and values.

Quite the contrary; there are crucial differences between facts and values.  Most obvious is Hume's admonition that "oughts" cannot be inferred from "is's."  In slightly different formulations, this appears as Moore's account of the naturalistic fallacy, which I have repeatedly invoked on MH Blog, and which I personally see as a common sin committed in thinking about the medical humanities, bioethics, and health policy.  Moreover, there is nothing about the fact-value problem that prohibits epistemic agents from working to present conceptions of the world that seem as fair, as accurate, and as 'balanced' as possible.  Criteria for judging the evidence, however, cannot depend on a world of brute facts untouched by human perception and thought.  The alert reader is no doubt making connections between the implications of the fact-value problem and the hoary, discredited but still-powerful notion of the view from nowhere.

This segueway permits us to briefly review the implications of the fact-value problem for the medical humanities.  First, the notion of presenting brute facts, unencumbered by values, biases, subjectivity, and presuppositions, is deeply problematic if not entirely incoherent.  To the extent we are linguistic creatures, our social behavior acts to categorize and conceptualize the local worlds we inhabit, including so-called "brute facts."  As such, particularly when thinking about science and medicine, it is vital to keep uppermost the value-laden nature of even the most benign "scientific" and "medical" facts.  Second, as entangled as facts and values may be, it remains important to think about the differences between facts and valeus, even if none of those differences constitute necessary and sufficient criteria for membership.  There are very good reasons to avoid inferring normative propositions from descriptive propositions; there are differences between the two kinds of propositions, even if there is an aspect of normativity in virtually all descriptive propositions.

See also the Lexicon entry on hermeneutics.

_________________________________________________________________

Bibliography

David Hume, Treatise on Human Nature.

W.v.O. Quine, Word and Object (Cambridge, MA: The MIT Press, 1960).

Donald Davidson, "On the Very Idea of a Conceptual Scheme," Proceedings and Addresses of the American Philosophical Association, Vol. 47, (1973 - 1974), pp. 5-20.   

Thomas Kuhn, The Structure of Scientific Revolutions, 3d ed. (Chicago: University of Chicago Press, 1996).

Paul Feyerabend, Farewell to Reason (New York: Verso Books, 1987).

Clifford Geertz, The Interpretation of Cultures: Selected Essays (New York: Basic Books, 1977).

Interpretive Social Science: A Second Look, Paul Rabinow and William M. Sullivan, eds. (Berkely, CA: University of California Press, 1988).         

Comments

[Paul Gowder]

I fear this is why I'm not a thinker friendly to the hermeneutic tradition. The idea that difficulties expressing P (e.g., across languages) implies that "and therefore that the notion of [P] totally independent of subjective values and impressions is itself incoherent" just strikes me as lunacy. This is so for several reasons, two of them chief:

1. There are Moorean facts. That is, there are some facts that we know (intuit) a lot stronger than we know (intuit) all these claims about languages. I am sitting down as I type this. Even were I to try and type this in a language that didn't have a concept of sitting down, I'd be sitting down as I typed this -- if only because sitting down is engangled with a lot of other concepts and observations, some of which surely must be in any well-functioning language (my legs are crossed, my torso is close to my feet, light is reflecting from my body in a certain fashion, the soles of my feet do not touch the ground). Indeed, we can recognize facts that aren't translatable into any non-mathematical language, like the facts of quantum physics.

2. The position is self-defeating: the argument for the incoherence of facts is premised on fact claims about language and human cognition.

[Daniel S. Goldberg]

Hey Paul,

Do you think that we can draw brightlines between facts and values? That is -- what precisely do you find to be lunacy about the hermeneutic critique?

Note that a healthy portion of the argument sketched here has absolutely nothing to do with hermeneutic-style thinkers. Much of it comes from heavy philosophy of language/philosophy of science itself.

The indeterminacy of translation argument comes from Quine, not from Gadamer or his ilk.

Re your specific points:

(1) I think you know my general feeling about Moorean epistemology (it's borderline absurd and totally unconvincing to me). That said, it is not at all clear to me how the existence of many totally uncontroversial facts establishes a brightline criterion or set of criteria we could use to effectively distinguish between facts and values. Nothing about the criticisms sketched above seem to me to deny the possibility of a great many uncontroversial facts that the vast majority of epistemic agents are justified in relying on.

What does the issue of justified belief have to do with the metaphysical issue of whether brute facts unencumbered by normative commitments exist?

2. Which position is self-defeating? The argument is not that facts per se is incoherent; the argument is that a metaphysical notion of facts as maintaining an essence that suffices to distinguish it from values is incoherent. It's not at all clear to me why that is incoherent or self-defeating.

You seem to be suggesting that an argument concluding that facts are incoherent cannot thereby rely on a factual premise. But you've got the conclusion wrong; it is not that facts are incoherent. It is, rather, that an idea of facts existing in any essential sense apart from values is incoherent.

For purposes of this post, I have no trouble assuming that there are facts, and that these facts are subject to varying degrees of justificatory force. None of these premises imply, however, that facts are discrete metaphysical entities existing apart from the values of the interpretations we use to apprehend these facts.

[Paul Gowder]

Really briefly (running out the door), I think the chief problem is that we're dancing back and forth between metaphysics and epistemology here. That is, I'm not sure how any premises about language, regardless of the philosophical credentials of its source, can establish any claims about the metaphysics of facts. At most, such premises can establish claims about what can be known and expressed about facts -- that is, about epistemology. So I don't particularly feel obligated to defend anything, against a linguistic challenge, stronger than that agents can have well-justified beliefs about facts.

That, I think, answers the question in 1). As for 2), I think that one implication of this alleged value-entanglement is that the strength of arguments relying on factual premises is undermined. That's so because if anything at all follows from the supposed value-entanglement, it's that arguments for factual claims necessarily have normative premises, and that seriously threatens factual claims, because normative claims are necessarily more subject to dispute, less amenable to final resolution, etc., than factual claims.

Anyway, yes, I think there are brute facts apart from values. That doesn't necessarily mean we can communicate them without some interpretive work, and bits of that interpretive work may be normative (a la Davidsonian charity), but none of that even touches the metaphysical question.

[Daniel S. Goldberg]

Ah, that clarifies things. And you can chalk up another proposition we happen to agree about: "agents can have well-justified beliefs about facts." There is nothing about the fact-value problem that commits me to the belief that well-justified beliefs about facts are impossible.

Of course, since beliefs posit the existence of some subject, a believer, as it were, this proposition is entirely friendly to those, like myself, who remain exceedingly dubious about an essential difference between facts and values.

As for #2, I tend to agree with you -- the value-laden nature of all facts tends to undermine a positivist hope in the compelling force of those facts. But so what? This does not mean that we cannot have better or worse justifications for belief regarding certain facts; nor does it imply the nonexistence of facts.

Rather, it describes some features about how we come to know facts, and whether those facts are knowable without interpretation, construction, and bias. And there are, IMO, very good reasons for doubting that we can know as such.