Like other contemporary realists, Millikan conceives of truth's correspondence relation as a mapping function from sentences to conditions in the world (1986, 353-54). There are "rules that project, from the parts and structures of sentences in a language, the conditions under which these sentences would be true" (Millikan 1990, 344). Call these rules truth rules. We understand the character of the correspondence relation---we have a theory of correspondence truth---just so far as we understand the character of truth rules.
Any representation in one's mind will have some sort of mappings onto the world, some sort of connections to the world. What is the character of the truth mappings, or truth rules, for those representations that are true? Such rules cannot be ultimately determined by one's intentions (Millikan 1986, 455). As Peirce observed, we certainly intend all our authentic beliefs to be true, yet some are false.
Peirce would say that truth rules are distinguished by their reliability. Millikan contends, on the contrary, that truth rules cannot be simply whichever rules project our beliefs or sentences most reliably onto our world (1986, 455-56; see also Nozick 1993, 65-67). True and false are normative notions. "A false sentence or a false belief is defective. A look at statistics, no matter how thoroughly these statistics are accounted for, hence how projectable they are, can turn up only statistical norms, never normative norms" (Millikan 1986, 456).
Millikan argues that understanding one's own mental representations must be a practical skill or ability (ibid., 456-57). It is a know-how, not just a know-that. We could not simply manufacture in thought a representation of X knowing only that it is a representation of X. To understand one's own thought, one must know how to manufacture a representation of X that maps onto X's in accordance with prescriptive truth rules. And one could know how to do this without always succeeding in conforming to those rules.
How could there be an ability (a know-how) to conform to truth rules? Would not conformance to a rule always entail knowing-that? Millikan observes that one's having an ability to do A is not the same as the fact that one sometimes does A. Moreover, abilities cannot be analyzed simply in terms of dispositions (Millikan 1986, 457-59; 1990, 327-29). One can be competent to do something and have no disposition to do it and vice-versa. Abilities must have something to do with intentions or purposes. "Purposiveness permeates every corner of our thinking processes; the ability to think is no exception to the rule that abilities are such only within the context of purposive doing" (Millikan 1986, 459). To understand how conformity to truth rules could be a know-how, we need a theory of purposeful doing "that does not make reference to explicit intentions---that does not entail that one has to know or represent to oneself everything that one has as one's purpose to do" (ibid.). The truth rules that we consciously strive to follow are constituted by some nonrepresented purposes of ours.
Before saying more of these nonrepresented purposes, we should notice some further things about abilities. Abilities in general depend not only upon workings within a person, but upon the structure of the outside world (ibid., 459-60). Linguistic ability in particular is an ability relative to the world in which the speaker lives. Even privately understanding one's own thought is something done against the background of the actual world. "Understanding is not something that I do in my head; I do it in my world" (ibid., 460). Notice also that abilities need not be infallible. One may have the ability to walk yet sometimes trip. The same goes for ability to speak or think.
Things we might take to be inside the head (say, dispositions and states) to help account for the ability to abide by truth rules do not determine truth rules, at least not when those things are taken alone and apart from the world. Such things in the head---call them intensions---are only means to the end of abiding by truth rules. Intensions are not criterial in the sense of being a standard by which extensions could be fixed. Truth rules, for the correspondence theorist, are made determinate in "a relation between the head and the world or between sentences and the world" (ibid., 463).
According to Millikan, I have said, truth rules are constituted by some of our nonrepresented purposes. Millikan elaborated this idea into a theory in 1990. In her presentation, she uses to intend for pursuits of explicitly thought-about, represented purposes and to purpose for pursuits of unexpressed, nonrepresented purposes (1990, 329).
When a person means to perform a mental activity, such as adding two numbers, there is an ingredient coming from the person, yet that ingredient is not given to his consciousness (ibid., 326). Our explicitly intending to follow mental rules, specifically truth rules, is rooted in our unexpressed purposing (ibid., 329). In what does such purposing consist?
Millikan distinguishes three ways of conforming to a rule: merely coinciding with a rule, purposefully following an explicit or expressed rule, and purposefully conforming to an implicit or unexpressed rule. That last way of the three "involves having an unexpressed purpose to follow a rule and succeeding in this purpose. It is the same as displaying a competence in conforming to the unexpressed rule, . . . [and this] differs from a mere disposition to coincide with the rule" (ibid.).
Her thesis is that the unexpressed purposes that lie behind acts of purposing are biological ones; the competence displayed is a biological one (ibid., 330; cf. Binswanger 1990, 7-13). Millikan wishes to argue that it is possible to have a biological competence to follow correspondence truth rules (cf. Devitt 1984, 73-99). To this end, she distinguishes between proximal rules and distal rules, using the hoverfly to illustrate (cf. Davidson 1990, 321, and Kelley 1986, 37, 138-41, 144-45, 148-51).
Male hoverflies spend much of the day hovering in one spot, keeping ever-ready to dart instantly after any passing female hoverfly. Their chasing behavior is hair-trigger enough that all sorts of inappropriate targets elicit pursuit (pebbles, birds, and other insects) as well as the occasional female hoverfly. Evolution has apparently favored a speedy response over the suitability of the target. Moreover, the male hoverfly does not dart directly at the female, but attempts to intercept her in a roundabout way. Millikan describes the male's immediate response as conforming to a proximal rule (turn 180° away from target minus about 1/10 of the vector angular velocity of target's image across retina) and its intercepting the female as conforming to a distal rule. The male hoverfly, of course, cannot calculate over some inner representation of its proximal rule in order to follow its proximal rule. "Rather, the hoverfly has an unexpressed biological purpose to conform to this rule" (Millikan 1990, 331).
Conformity to the proximal rule is a means to conformity to the distal rule. The well-functioning male hoverfly has an ability, a know-how, an unexpressed biological purpose, to conform to the distal rule attending the proximal rule. Whether the male hoverfly succeeds in following the proximal rule depends only on his internal functioning. Whether or how often he conforms to the distal rule depends on conditions outside him. His competence to conform to the distal rule is less reliable than his competence to conform to the proximal rule (ibid., 331-37). Notice with Millikan that "if the hoverfly ends up coinciding with the [proximal] rule not because his nerves and muscles work in a normal way but only because the wind serendipitously blows him around to face the right direction, he fails to express a competence" (ibid., 332).
Return to humans. Millikan observes that the unexpressed rules that humans purposefully conform to when using inner or outer language are not hard-wired but learned. Animals that learn can acquire biological purposes peculiar to them as individuals, tailored to their own circumstances or histories. Part of our human behavior is that we not only learn, but form new ideas and draw inferences from what we learn. All the same, "there must be a finite number of proximal and distal 'Homo sapiens rules' that we have as biological purposes to follow" (ibid., 342). Any bit of behavior of persons with well-functioning body and brain can be described "as a biological activity, a description that tells what proximal and distal biological purposes, and what biological competences if any, the behavior expresses" (ibid.). What, then, is the relation "of ordinary human purposes, of human intentions and meanings, to biological purposes? . . . Do ordinary human intentions merely, accidentally, cohabit with biological purposes?" (ibid.).
Millikan contends that ordinary human purposes, or intentions, are a species of biological purpose (and not a biological epiphenomenon). One's "explicit intending rests on biological purposing---biologically purposing to be guided by, to react this way rather than that to, one's representations" (ibid., 343).
We shall bring the subject of proximal and distal biological purposing to bear on truth rules shortly. First I ought to sketch some contemporary verificationist views on language and on truth rules, views which Millikan aims to overturn.
The verificationist, preeminently Michael Dummett, holds that whatever connection there is between sentences and that which determines their truth must be a connection established by the actual employment of language. Whether truth rules are verificationist or realist (correspondence) in nature, the practical abilities of speakers who understand the language would have to reflect those rules. Indeed, practical abilities of speakers who understand the language would have to determine the content of those rules. The verificationists hold that it is not possible that any truth rules we exhibit practically could be realist rules, which assign correspondence truth conditions; any exhibited truth rules must be verificationist rules, which assign assertability conditions to sentences (Millikan 1990, 344, 323).
There is a defective background supposition about language, according to Millikan, behind all the verificationists' tortured arguments for the impossibility of realist truth rules. It is supposed that language understanding is a disposition. Language understanding is no such thing in Millikan's analysis.
While one is speaking sincerely, one is trying to conform one's sentences to their truth rules. The way in which "the actual practice of a language embodies truth rules is that these are the rules in accordance with which the competent speaker (or thinker), when sincere, purposes to make (or think) assertions" (ibid., 345). At some deep enough level, these rules must be unexpressed rules. We cannot simply introspect and view these primal rules. To know (know-that) the rules, we need a theory, "an explanatory hypothesis about what rules we are purposing to follow when we make sincere assertions" (ibid.).
From the biological perspective, we can ask how proximal, as opposed to distal, truth rules might be. A realist, such as Millikan, contends they are for the most part distal. Correspondence truth rules are distal; they concern the distant world, the world beyond the boundaries of one's body. Distal truth rules, of course, are backed by proximal rules. These proximal rules, Millikan calls proximal assertability rules. Conformity to proximal assertability rules has, as a biological purpose, conformity to distal rules. Conformity with proximal rules does not guarantee conformity with distal truth rules. Neither does the male hoverfly's conformity to its proximal rule (the retina rule) guarantee interception of a female hoverfly.
The verificationist would insist that our truth rules have to be proximal. There cannot be such a thing as distal truth rules because language is a disposition, and dispositions cannot concern anything beyond the boundaries of the organism. Millikan counters that language is not merely a disposition. Truth (and meaningfulness) is a normative notion. No mere disposition, no mere performance, can determine a standard or norm. Language ability is a competence, and no set of mere dispositions equals a competence. "My disposition to fall if left unsupported is no competence, nor is the hoverfly's disposition to chase birds" (ibid., 351).
Competence in language and in conformity to truth rules is a know-how, and most know-how involves distal action. Because the verificationist locates truth rules wholly within the proximal, Millikan denominates that view myopic.
[Without fail, see also John Haugeland's "Truth and Rule-Following" in Having Thought (Harvard, 1998).]
