The Hypercomputational Case for Substance Dualism

Introduction

I'm a dualist; in fact, a substance dualist. As you probably know, this places me within a rather small minority, at least among academics, and certainly among professional philosophers.¹ There are of course a number of property dualists about (e.g. Jjacquette 1994, Clarmers 1996),² but those of my ilk are rather hard to find. Why then do I believe what I believe? Well, myriad arguments compel me to believe as I do, some going back to Descartes. (The vast majority of these arguments are elegantly and crisply canvassed by Meixner 2004). But one of these arguments is a new one that I articulate herein; this argument exploits the contemporary computational scene, as well as a long-established logico-mathematical continuum of increasingly powerful information processing, ranging from the processing that devices below Turing machines can muster, to what Turing machines can do, to what ``hypercomputing'' machines can do.

As I soon explain, it's the hypercomputational nature of human cognition which entails that Descartes (with a Chisholmian slant³ was right all along. Encapsulated, the entailment can be charted as follows: If human persons are physical, then they are their brains (plus, perhaps, other central nervous system machinery; denote the composite object by `brains⁺'). But brains⁺, as most in AI and related fields correctly maintain, are information processors no more powerful than Turing machines. Since human persons hypercompute (i.e., they process information in ways beyond the reach of Turing machines), it follows that they aren't physical, that is, substance dualism holds.

The plan for the paper is as follows. After some remarks on the niceties of defining dualism (§ 2), I give (§ 3) enough background from relative computability theory to understand my argument, and then, in section 4, I give a more explicit version of it that can be effortlessly certified as deductively valid. Each premise in the argument is then separately defended (in some cases against objections), with the majority of attention paid to premise (4), which says that human persons hypercompute. In the penultimate section (5), I consider some additional objections, and emphasize that my objective in the present paper is only to present a formidable argument for substance dualism. The fully developed case for substance dualism that the present paper points to includes many previously published arguments for the proposition that human persons hypercompute; and these publications include answers to numerous objections. I thus claim herein not that the main argument expressed in the present paper is conclusive, but rather that, again, it's quite formidable: put another way, that it provides enough ammunition to make being a substance dualist, in our day, perfectly rational. That said, the content herein, plus supporting argumentation published elsewhere (cited below), does by my lights constitute a conclusive case for substance dualism. I end the paper with a brief conclusion (§ 6).