Discussions of “Brute facts” often treat of facts that are merely unexplained or not necessary for some explanation when in fact such a fact must be one not merely unexplained but inexplicable. This difference is as wide as that between something unread and unreadable, or an unsolved problem (i.e. any problem in an untouched math book) and an insolvable problem (which is, one hopes, no problem in a math book).

Again, “the unexplained” can be itself a fact – I don’t need to explain everything that is involved in X to explain it. The inexplicable is a much stronger claim that requires a theory of explanation, and a proof for why some fact cannot be an object of it.


1 Comment

  1. D.S. Thorne said,

    March 18, 2014 at 4:24 pm

    I myself am suspicious of ‘brute facts’. Appeals to brute facts carry with them an interpretive framework in every case, even if tacitly. Some data sets might only make sense on one interpretation, but an interpretation is involved nevertheless.

    There is a kinship here with the syntactic/semantic dichotomy in computational theories of mind – all syntax is underdetermined to the point of pure meaninglessness without a semantic engine (i.e. mind) to read that syntax.

    A similar problem arises in positivist history, that would reduce history to ‘facts’ and statistics, when in fact the very choice of facts and statistics is driven by what one considers important at the outset – this is what Windelband called the problem of “valuation”. And then, of course, the facts and statistics thus gleaned are multiply charged, standing in need of disambiguation, again from an interpretive standpoint.

    “Red, here now”: the protocol sentence of the analytic philosopher is predicated on his previous abdication of moral responsibility, his failure to acknowledge that our situation in a moral landscape with better and worse regions is constitutive of being human. Marcuse once remarked that it is the dream of a totalitarian regime when their intellectuals voluntarily do this…

%d bloggers like this: