Reifying bundles of bits in Pursuit of Truth

In the Pursuit of Truth[1], must one reify? Quine says that “substantial reification is theoretical” (25), beyond the simple observation sentences. We emerge from the “feature-placing” stage (24) and in the persistent recurrence of certain qualitative features we develop a theory that the set of features we regularly observe belong to some consistent, persistent object. These theories are purported to be useful. For the child it helps to recognize Mama as something other than the recurrence of Mama-features, if only for a sense of security. Endless metaphysical thought experiments have us imagining non-identical doubles of things we know, such as twin-Earths and doppelgangers, all with the impression that we’ll recognize that the identity of an object over time is important – seemingly identical objects with the same extension but different causes are purported to be non-identical.

Natural selection seems to have produced reification but is it truth proper to imagine the persistence of a single object between instances of its being observed? The question for the child is how much qualitative difference is acceptable between instances of Mama? If Mama changes her hair or her clothes, has cosmetic surgery or injures her vocal chords, will she still be recognizable as Mama to the child? She’s not strictly identical to the entire set of features previously observed. Observation sentences about her appearance will differ. The theory places her as the same object, but the stimulus meaning of “there’s Mama” has changed drastically since its last utterance. If “Mama” has changed enough, the child may not assent to the observation sentence.

This is where Quine later drew a distinction between perceptual and full reification. In a passage from a late manuscript cited by Peter Hylton in his text on Quine[2], by way of Leonardi and Santambrogio[3], Quine clarifies the terms:

Prior cognition of a recurrent body—a ball, or Mama, or Fido—is on par with our recognition of any qualitative recurrence: warmth, thunder, a cool breeze. So long as no sense is made of the distinction between its being the same ball and its being another like it, the reification of the ball is perceptual rather than full.

For Quine, the child is capable only of perceptual reification, rather than the full reification one that is present in a theory. I take this to be a sort of family resemblance between stimulus meanings. A child can recognize certain sets of stimuli that, in a criss-crossing mesh include various patterns that trigger recognition, representing something perceptually for that child even if the child lacks a theory of persons, minds, bodies or persistence of an object.

Hylton takes the question of reference to be “how we can get from observation sentences to a mastery of language that is clearly about particular entities”[4] My question is, in a world of bits where language refers to various arrangements of those bits, why must language be about entities in the strict sense? Why must we pass from the perceptual reification that occurs when those bits are arranged in familiar ways to the full reification that says that there is some entity there? It seems to me that we can preserve a tasty desert landscape without reification of anything beyond a fundamental bit. At the end of this chapter, Quine expresses his preparedness to bail on the traditional concept of existence (36). If existence can go by the board, the entity can follow.

I know I’m taking a step away from the point of this chapter when I zero in on this, and I had prepared a commentary on Quine’s indistinguishable, isomorphic ontologies but I read it over and wasn’t sure there was sense to any of it. Indeed we can build fairly successful theories with varying versions of the ontology. But I find the reification game a little superfluous, especially if we’re taking a behaviourist account of things. We can build our ontology around dogs, around sets of {dog} or their absolute complements like  ∁{dog}, and each adopt our own isomorphic ontology and I can talk about dogs meaning dogs, and Smith can talk about dogs meaning {dog} singletons and Jones can play around with his ∁{dog} complements – just as my red could be your blue and, I don’t know, maybe there’s magic. Perhaps I need to be reminded that reference even needs to be successful – why can’t it just be a game of resemblance and recognition? What seem to be entities over time swap out their bits with other bits and it barely matters if the bits are the same bits so long as whatever bits are there are behaving in similar ways. We’ll quickly assent to the same observations when the bits move around. Twin Earth is Earth or not Earth and, sure, XYZ can be water if you want it to be and now I sound like a pragmatist so I’ll fall just short of that and say that I don’t see a difference made.

We can all have our own brain-bits differently-arranged and equally assent to the same observation sentences and whenever we generally agree on the borders around sets of bits we’ll call it an entity. Or, better still, we’ll each list off all of the types of entities that we think actually exist as fundamental and with each one take a sip of scotch. Whoever is left standing wins.


[1] Quine, W. V.  Pursuit of Truth. Revised Edition. Cambridge, MA: Harvard University Press, 1992. Print.

[2] Hylton, Peter. Quine. New York: Routledge, 2007. §5, V.

[3] Leonardi, Paolo and Marco Santambrogio, eds. On Quine. Cambridge: Cambridge University Press, 1995). p. 350.

[4] Hylton. Quine. §5, V. #