A Possible Dilemma for Situation Semanticists

This paper examines the concept of information in situation semantics. For this purpose the most fundamental principles of situation semantics are classified into three groups: (1) principles of the more fundamental kind, (2) principles related to regularity, and (3) principles governing incremental information. Fodor’s well-known criticisms of situation semanticists’ concepts of information target the first group. Interestingly, situation semanticists have been anxious to articulate either the principles of the second group or the principles of the third group in order to meet these criticisms. Based on these observations, I will launch a dilemma for situation semanticists. Either they fail to handle information about individuals or they fail to present any acceptable account of the laws of nature. Millikan’s version of situation semantics, I shall argue, is not the exception to the rule.

1. According to him, “[s]ome versions of conceptual role semantics, and theories based on biological function, are competitors, but none of these is as highly developed as informational semantics [Ibid.].

2. Throughout this section, I shall largely depend on Thagard’s report, because it rather succinctly summarizes the development of theories of information until the time when situation semantics was introduced. For more updated surveys, see Chemero (2009), Bremer and Cohnitz (2004), Shea (2013), and Floridi (2008).

3. Some part of this subsection (esp. up to 3.2.1) is drawn from Park (2014). One anonymous reviewer criticized my discussion of Fodor’s criticism is not well-organized and includes unnecessary elements, which were unhelpful and somewhat confusing. Though I concede that the reviewer’s criticism is warranted, I left the relevant paragraphs as they stood, for the possible confusions are not entirely mine, and the confusions themselves could be informative.

4. (Fodor 1986; Barwise 1986; Fodor 1987; Barwise 1989). Fodor thinks that the theory of information held by Dretske and articulated by Barwise and Perry treats the information content of a situation as the basic notion. In this theory, the information content is non-perspectival, receiver neutral, intrinsic, and non-intentional. On the other hand, he claims, what are needed in cognitive science are the notions of information displayed by a signal, the information encoded by the signal, and the information in the signal that is available to the agent. They are perspectival, receiver relative, non-intrinsic, and intentional. Barwise sensitively responds to such a characterization of their theory, especially as behavioristic. He also launched a counterattack on Fodor s ideas of lingua mentis, which is not my present concern.

5. No doubt, Fodor is simply adopting Goodman’s influential criticism of resemblance theory here. According to Goodman, unlike representation, resemblance is reflexive and symmetric. Without any further argument other than this seemingly undisputable observation, Goodman claims that resemblance is neither necessary nor sufficient for representation in art. Examples Goodman exploits in favor of the insufficiency of resemblance for representation are the automobiles off an assembly line and the twin brothers. In these cases, “neither one of a pair of very like objects represents the other” (Goodman (1967, p. 4). In order to prove that resemblance is not necessary for representation, Goodman simply resorts to the widely accepted fact that “almost anything may stand for almost anything else”. [Ibid., p. 5] Insofar as we can accept that “denotation is core of representation”, Goodman seems to have an extremely strong, "if not conclusive” case for his conclusion that representation is “independent of resemblance”. [Ibid.] Elsewhere, I showed how formidable has been the power of Goodman’s terse argument for the independence of representation from resemblance, by giving a few examples from philosophy of mind, aesthetics, and philosophy of science (Fodor 1984, 233; Lopes 1996, p. 18; Bailer-Jones 2009, p. 183; van Fraassen 2008, p. 16).

6. It is well-known that informational theories of mental representation is bound to face a fatal problem of error. For, as Godfrey-Smith explains, they “apparently cannot explain how a representation can acquire a determinate content and yet be false” (Godfrey-Smith 1989, 534). In order to understand why it appears to be impossible for informational theories of mental representation to explain the cases of mental misrepresentation, possibly the best way is to introduce “Fodor’s disjunction problem”, which Godfrey-Smith believes to be “the error problem’s most troublesome form” [Ibid.].

7. Godfrey-Smith even counts Fodor as a convert to informational semantics in the following historical report about informational semantics: “The central work is Dretske (1981), and the theory was largely developed at the University of Wisconsin by Fred Dretske, Dennis Stampe, and Berent Enc. Recently, informational semantics has roamed far beyond its Wisconsin home, and built a sizeable collection of followers. Converts include Jerry Fodor (1987), Robert Stalnaker (1984) and, less faithfully, Churchland and Churchland (1983) and Field (1986)” (Godfrey-Smith 1989, 533).

8. It should be clear that the alleged dilemma is not intended to be a knockdown argument against situation semanticists. It is merely a device by which we might find the weaknesses of situation semanticists’ ideas. Indeed two anonymous reviewers questioned the appropriateness of presenting my criticism of situation semanticists in dilemma form.

9. One anonymous reviewer correctly pointed out that Millikan’s target was Barwise and Perry, not Barwise and Seligman. Part of the reason why Barwise and Seligman attempt to offer an account of regularity in the paper was Barwise’s dissatisfaction with the treatment of constraints in Barwise and Perry’s Situations and Attitudes. Another anonymous reviewer perceptively pointed out the relevance of the connection between Peirce and Millikan for the issues dealt with here.

10. Millikan notes: “Failure to account for our capacity to represent individuals in language and thought has been, perhaps, the most serious failing common to contemporary naturalist theories of content” (Millikan 2004a, p. 43).

11. It is simply beyond the scope of this paper to examine and assess her ambitious attempts. In addition to the issues dealt with so far, there are some other significant points to note. (1) Situation semanticists are probably exploiting the current dominance of direct reference theories in the philosophy of language in their excuse not to deal with the problem seriously; (2) In close connection with their discussion of incremental information, situation semanticists invoke the age-old distinctions and theories of British empiricists. We saw above how they approximate their theory of meaning to that of Hume. More prominent is the distinction between ideas and notions. For example, we read; “Among the representations that figure in ideational beliefs, I will call the representations of things notions, and the representations of properties and relations, ideas” (Crimmins 1992, 75). I think that the idea/notion distinction must be closely related to the scholastic discussion of the knowledge of singulars, and that thus it is highly dangerous and/or anachronistic to adopt it without any discussion of its historical background; (3) Situation semanticists’ talk about the scheme of Individuation is another problem. Ontological issues of individuation are thereby completely ignored by them, since their concern in the scheme of individuation is just epistemological and/or semantical; (4) All this may be due to their reluctance to discuss ontology, in particular the dialectic between universals and particulars. Barwise’s extensive appeal to the type/token distinction is another sign of such reluctance. Also telling is the fact that their principle (C) merely claims that the information a fact carries is relative to a constraint without asserting that the information of a constraint is relative to the individuals of the system. After articulating the conception of constraint, the lack of a principle governing constraint and individuals reminiscent of Aristotelian hylomorphism or the principle of exemplification seems more sorely felt; (5) If they spell out their ontology, I surmise that it might be a kind of trope theory. If so, I wonder whether it can support their theory of information that is meant to be of service to cognitive sciences. Probably, they are too mindful of naturalization without paying due respect to intentionality.

12. I am here indebted to an anonymous reviewer.

I am indebted enormously to anonymous reviewers and the editors of this journal for their constructive criticisms.

Authors and Affiliations

1. Humanities and Social Sciences, KAIST, Daejeon, South Korea

   Woosuk Park

Corresponding author

Correspondence to Woosuk Park.

Reprints and permissions