Markman & Dietrich still want to hold that all representational states carry information about their contents, despite the well-known perils of the error problem.
2. First, it could mean that information is the basic representational relation, nonetheless not all representational states carry information about their contents. This is the view of information semanticists such as Dretske (1983, 1988) and Fodor (1987, 1990) (see Clapin 1998, para. 5). This view is, of course, plausible, and there has been a very vigorous debate about exactly how to make it work. But on this reading, not all representational states carry information about their contents, and so not all representational states are mediating states.
3. Second, it could mean that ALL representational states carry information about their contents. "Information" is defined thus: "an event, e, at a receive,r carries the information that something, s, has property P, P(s), if and only if the conditional probability of P(s) given e is 1" (Markman & Dietrich, 1998a, para. 9). On this second reading, which appears to be the one Markman & Dietrich are defending, a perceptual judgement that there is a dog carries the information that there is a dog if and only if the conditional probability of there being a dog, given the presence of the perceptual judgement that "there is a dog," is 1. Given that all representational states are mediating states, and thus carry information about their contents, it follows that any perceptual judgement of the form "there is a dog" only occurs when there is in fact a dog in the vicinity.
4. But this is inconsistent with what Markman & Dietrich WANT to say: "At this moment, there is some mediating state in the system which carries information that there is a dog on the floor. When a light is turned on, it may become evident that this 'dog' was actually some rumpled clothes on the floor" (Markman & Dietrich 1998b, para. 7). They want to say that there can exist a representation which means X but is not in "lawful connection" (Markman & Dietrich 1998b, para 9) with X.
5. This is in essence the problem with their position. They want to hold two inconsistent views: (a) that all representational states carry information (in the technical sense) about their contents; and (b) that some representational states (the errors, at least) don't carry information (in the technical sense) about their contents.
6. This problem -- the problem of misrepresentation (also known as the "error problem" or the "disjunction problem") -- is nothing new, of course. Those who want to put forward an informational semantics have been grappling with it for decades. But Markman & Dietrich don't appear to appreciate that it is THEIR problem too. It is their problem because they claim that (i) all representational states are mediating states; (ii) all mediating states carry information (in the technical sense); and (iii) if representation R is about content C, then qua mediating state it carries information about C.
7. Markman & Dietrich could deny (iii), of course, and if they did so they would no longer be caught up in the error problem. But denying (iii) means that the INFORMATION carried by a misrepresentation (or other problem case) has nothing to do with the CONTENT of the (mis-)representation. Thus, in the dog example, the content of the perceptual judgement that there's a dog nearby involves dogs and not clothes, whereas, qua mediating state, it carries information about clothes (since it was caused by a pile of clothes). But this is inconsistent with the way they tell the story about the dog ("At this moment, there is some mediating state in the system which carries information that there is a dog on the floor" (Markman & Dietrich 1998b, para 7).
8. The error problem shows us that not all representations can carry information about their contents. Markman & Dietrich are right to emphasise mediating states as representation-like; by doing so they present a useful challenge to representational sceptics whose models make use of mediating states. However, the notion of mediating state fails to capture any property common to all representations.
Clapin, H. (1998) Representation is not Information. PSYCOLOQUY 98(9) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/psyc.98.9.64.representation-mediation.8.markman
Dretske, F. (1981). Knowledge and the Flow of Information. Cambridge, Massachusetts: MIT Press.
Dretske, F. (1988). Explaining Behavior. Cambridge, Massachusetts: MIT Press.
Fodor, J. A. (1987). Psychosemantics. Cambridge, Massachusetts: MIT Press.
Fodor, J. A. (1990). A Theory of Content and Other Essays. Cambridge, Massachusetts: MIT Press.
Markman, A.B. & Dietrich, E. (1998a). In Defense of Representation as Mediation. PSYCOLOQUY 98(9) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/psyc.98.9.48.representation-mediation.1.markman
Markman, A.B. & Dietrich, E. (1998b). Mediating States, Information and Representation: Reply to Clapin on Representation-Mediation. PSYCOLOQUY 98(9) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1998.volume.9/psyc.98.9.66.representation-mediation.10.markman