Abstract
In this chapter we refute computationalism via specification and defense of the following argument: Computation is reversible; cognition isn’t; ergo, cognition isn’t computation. The specification of the argument involves a quartet: (i) certain elementary theorems from com- putability theory, according to which computation is reversible; (ii) the doctrine of agent materialism, according to which, contrary to any sort of dualistic view, human agents (= human persons) are physical things whose psychological histories are physical processes; (iii) the introspection- and physics-supported fact that human cognition is not reversible; and (iv) the by-now-familiar claim — put roughly for now — that cognition is computation. The basic structure of the argument is straightforward: the conjunction of (i), (ii) and (iii) entails the falsity of (iv).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Footnotes
See (Bringsjord 1991) for a discussion of the consequences of this fact for AI.
The interested reader can consult an octet of books we find useful: For broad coverage of the basic material, see (Lewis & Papadimitriou 1981, Ebbinghaus et al. 1984, Boolos & Jeffrey 1989, Hopcroft & Ullman 1979). Once again: For a nice comprehensive discussion of computability theory that includes succinct coverage of uncomputability, including the Arithmetic Hierarchy, see (Davis et al. 1994), the text we drew from in Chapter 1. There is also the difficult but rewarding (Soare 1980). Partee et al. (1990) contains a very nice discussion of the Chomsky Hierarchy. And, of course, there’s always the classic (Rogers 1967).
The “Turing’s World”™ software comes with (Barwise & Etchemendy 1993).
This is borne out by, among other things, looking at proposed comprehensive models of cognition in cognitive science: these models now often include some component claimed to account for consciousness. (Recall that we pointed out in Chapter 1 that these days consciousness, even phenomenal consciousness, is regarded to be cognitive.) For example, Schacter (1989) gives us the picture of cognition shown in Figure 6.4. For a penetrating analysis of such models (including Schacter’s) see (Block 1995). For those interested in charting the first-order formalization of our argument, the relevant sentence in first-order logic would be a full symbolization of ∀x(x is cognizing ⇒. x is conscious), where ‘cognizing’ is understood to indicate the full scope of human cognition as purportedly captured in models like Schacter’s.
Compare this sort of indivisibility with the type Descartes famously ascribed (perhaps incorrectly) to the mind when he said: In order to begin this examination, then, I here say, in the first place, that there is a great difference between mind and body, inasmuch as body is by nature always divisible, and the mind is entirely indivisible. For, as a matter of fact, when I consider the mind, that is to say, myself inasmuch as I am only a thinking thing, I cannot distinguish in myself any parts, but apprehend myself to be clearly one and entire; and although the whole mind seems to be united to the whole body, yet if a foot, or an arm, or some other part, is separated from my body, I am aware that nothing has been taken away from my mind. (Descartes 1911, p. 196)
Bringsjord does personally conduct a lot of “Weak” AI, in the form of an attempt to engineer systems capable of passing determinate tests. (This brand of “Weak” AI is discussed in the final chapter of this volume.) See, for example, the programs featured in (Bringsjord & Ferrucci 2000), which are designed to produce an agent capable of passing S3G (Bringsjord 1998a).
In (Bringsjord & Zenzen 1991) we devise and exploit a variation on the classic brain-in-a-vat gedanken-experiment in order to make the case lor this view. Bringsjord (1995a) argues, contra Harnad (1991), that Turing Testing, in order to test for consciousness, needn’t include a test for the ability of a would-be AI to interact with its environment.
One method for such encoding comes via Gödel numbers; we resorted to this method earlier in the book. For an example of such coding explained with help from convenient tables, see (Boolos & Jeffrey 1989). Some of the philosophical issues related to this particular form of coding are discussed in (Bringsjord 1998b).
We place scare quotes around ‘Theorem’ because we describe the result very informally. For a more precise statement of the theorem, as well as a more precise account of the proof than what we provide below, see (Lewis & Papadimitriou 1981).
We leave aside a complexity-based construal of ‘power.’ Obviously, a physical TM with multiple tapes could sometimes solve problems faster than an ordinary one-tape TM. However, there are no problems which are unsolvable by a standard TM yet solvable by a multi-tape machine. Likewise, a quantum computer can solve certain problems very fast, but can’t solve any Turing-unsolvable problems (Deutsch 1985). See, however, (Stannett forthcoming).
Again, for a detailed discussion of Max, ballistic computers, and other, similar devices, see Bennett (1973, 1984).
Bringsjord (1991) distinguishes between various sorts of computationalists and connectionists.
This paper is based on the classic statement of connectionism given by Paul Smolensky (1988a, 1988b).
McCulloch and Pitts (1943) showed long ago that such a simple activation function allows for the representation of the basic Boolean functions of AND, OR and NOT.
Readers wanting to see some of them are encouraged to consult Poundstone’s (1985) classic discussion of Conway’s (1982) Game of Life.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Bringsjord, S., Zenzen, M. (2003). The Argument from Irreversibility. In: Superminds. Studies in Cognitive Systems, vol 29. Springer, Dordrecht. https://doi.org/10.1007/978-94-010-0283-7_6
Download citation
DOI: https://doi.org/10.1007/978-94-010-0283-7_6
Publisher Name: Springer, Dordrecht
Print ISBN: 978-1-4020-1095-8
Online ISBN: 978-94-010-0283-7
eBook Packages: Springer Book Archive