Holy Cyclops

In Paul Boghossian’s Fear of Knowledge:  Against Relativism and Constructivism, a book I’ve recently started reading and am enjoying, he presents an argument against what he calls “cookie-cutter fact-constructivism.”  He is arguing against the sort of view advocated by Nelson Goodman and Hilary Putnam, and illustrated by Goodman’s “constellation” example:  We draw the lines around groups of stars and call them “constellations”; the groups of stars aren’t constellations until we notice them as such.  There is no such fact as “The Big Dipper is a constellation” until we say there is.  That fact is socially constructed.  Moreover, all such facts are socially constructed:  We use our concepts to “carve up” reality, just as a cookie cutter cuts up dough into shaped pieces, and since this “carving up” isn’t done until we do it, we construct the facts about reality; moreover, no way of conceptually “carving up” reality is closer to the way things really are than any other, for there is no way things really are.

Boghossian argues that this “cookie-cutter” view contradicts itself, because even if one says that constellations are delineated by human concepts, and that their constituent stars are similarly delineated by human concepts, and that their molecules (well, plasma, but I don’t think that’s important here) are similarly delineated by human concepts, and so on, one still requires, at some basic level, a “dough” which our concepts can operate on, and there has to be a way that that “dough” really is, unless one wants to insist that the “carving up” goes on infinitely, level after level, which clearly seems absurd.

The problem I see is that this argument only works on the reality-objectivist fact-constructivist—the one who says that there really is an objectively existing reality, and whose fact-constructivism consists of applying human concepts to that reality.  That is to say, his argument only works to establish the reality of an underlying objectively real “dough” against those fact-constructivists who already agree that there is one.  (Such fact-constructivists might then have to agree that there is a way things are at the fundamental level—the level of the “dough”—and simply reserve their fact-constructivism for all higher levels of fact; they might also adopt the view I’m about to describe for everything except the “dough.”)  The reality-subjectivist fact-constructivist, on the other hand, may simply insist that concepts do not apply to objectively existing reality (or to elements of it), but instead apply to one’s own mental life (or to elements of it)—that, for example, my concept of stripedness is not applicable to objectively real tigers but is instead applicable to my ideas of tigers.  (He might or might not agree that if there happened to actually be objectively existing tigers, then my application of the concept of stripedness to my ideas of tigers could be extended, so that the concept of stripedness would then be applicable to actual tigers.)  As long as it is not agreed that there are actual tigers, and as long as it is maintained that the concept of stripedness applies to my ideas of tigers rather than to actual tigers, the only “dough” the reality-nonobjectivist fact-constructivist is committed to is the “dough” of his own ideas—his own thoughts and mental images and the like. 

Perhaps Boghossian will address that sort of position later in his book.  I’ll find out.