(sent to conlawprof)
Holmes was father to both realism and positivism. At least the positivism that emerged in American legal culture following the rise of the administrative state in the 1900s. His essential contribution is the claim that judging is not a self-contained craft. It isn't an a priori orthodoxy. It isn't, in short, a mathematics. Remember that Holmes comes along ahead of his time during the moral sciences period of intellectual history. Inasmuch as judging is not the imposition of its own logic, what is it then? Is it subjective? No it is not. Holmes, like most of the realists who followed him intellectually, was an objectivist in terms of his epistemology. "Law" was simply selecting the most empirically-appropriate policy configuration. One biographer of his calls this sort of ideology about law the new "policy science." (It was rather understandable, of course, that "policy science" would emerge in legal culture at the same time that empirical
positivism (positivism in science) had reached such a high ascendancy in intellectual history). So now, instead of law being a mathematics, it was more in the nature of an empirical growth science. But because judges were not equipped to be the governors of such a venture, deference had to be given to the legislative organs -- because they were more equipped to configure good (and even correct) public policy. Although there was a segment of realist thought that tried to say that judges should interpret laws consistent with empirical science -- lawyers tended to call it "sociological jurisprudence" -- it failed for obvious reasons. Instead, the king of jurisprudence in American legal culture became positivism. Holmes was therefore an integral figure in both realism (that judging is not an a priori craft) and positivism (that legislatures have the ultimate legitimacy in this respect).
If one treats the constitutional convention as an assembly, one can see why Holmes would defer to constitutional commands that have more of a clear meaning rather than deferring to legislative and pragmatic goals when the words are the especially fluffy ones. So Holmes very much would defer to the clear commands of the positive law, even where he disagreed with them, because legislatures write the law (it being only policy) and because judges are not to impose their systematicized constructs upon the legal system under warrant of finding truth through legal orthodoxy (which is a myth).
Holmes really was an interesting American thinker. Every bit as important in his circles as James, Dewey or Rorty was to theirs (although I am not a big fan of Rorty myself).