Selmer Bringsjord (1995) Agnosticism Re-revisited. Psycoloquy: 6(30) Robot Consciousness (16)

Volume: 6 (next, prev) Issue: 30 (next, prev) Article: 16 (next prev first) Alternate versions: ASCII Summary
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 6(30): Agnosticism Re-revisited

Reply to Brown and O'Rourke on Robot-Consciousness

Selmer Bringsjord
Dept. of Philosophy, Psychology & Cognitive Science
Department of Computer Science
Rensselaer Polytechnic Institute
Troy NY 12180 (USA)


Brown and O'Rourke (1995) state that their argument against my first version of the Arbitrary Realization Argument (ARA1) is based on a rule of inference other than the ones I previously (1995) showed to be fallacious. Unfortunately, their argument is not based on this new rule. The rule of inference they do deploy implies that their argument has at least one false premise.


behaviorism, Chinese Room Argument, cognition, consciousness, finite automata, free will, functionalism, introspection, mind, story generation, Turing machines, Turing Test.
1. The Arbitrary Realization Argument (ARA1), detailed in What Robots Can & Can't Be (1992), involves a Turing machine, M, built out of 4 billion Norwegians all working on railroad tracks in boxcars with chalk and erasers, etc., across the state of Texas. One of the premises in ARA1 is

    (1-VI) There is no agent m constituted by M that has genuine
           conscious states (such as those involving fears, hopes,
           pains, etc.).

2. Brown and O'Rourke (1994) argue against this premise in particular, and ARA1 in general, via the following argument:

    (1) Bringsjord admits to being agnostic about low-level
        functionalism (LLF).

    (2) Low-level functionalism is plausible.

    (3) Low-level functionalism, with a few more plausible assumptions,
	implies claim (1-VI) is false.

    (4) So Bringsjord should be an agnostic about (1-VI).

    (5) So Bringsjord should not find ARA1 compelling.

3. I do indeed admit to being agnostic about "neuron-level functionalism" (NLF, one form of LLF), the view -- defended by Cole & Foelber (1984) -- that gradually replacing neurons with "silicon-based workalikes" will leave the associated mentation intact. But Brown and O'Rourke are wrong that this agnosticism is at odds with premise (1-VI) in ARA1.

4. The rule of inference they say they are using is not the provably fallacious

    R-a If s is agnostic about p, and p implies q, then s should be
        agnostic about q,

as I had earlier read them as employing (on the strength of statements like "NLF implies (1-VI). Since Bringsjord is agnostic about NLF, he should be agnostic about its consequences."), but rather

    R-b If s is agnostic about p, and s knows that p implies not-q,
        then s should not accept q.

The instantiations are: s = Selmer, p = NLF, q = (1-VI). But now a problem arises: Selmer doesn't know that NLF implies (1-VI). This is true because if some person s' knows p', then s' believes p' (because an agent's knowing some proposition consists, at least, in the agent believing that proposition, that proposition being true, and the agent having justification for the belief), but Selmer doesn't believe that NLF implies (1-VI) (because for starters, he appreciates the fact that there is no way to construct a derivation, in any known logic, of not-(1-VI) from LLF).

5. Brown and O'Rourke don't believe that NLF implies (1-VI) either. That they don't is clear from their argument, for step (3), as seen above, contains the crucial phrase "with a few more plausible assumptions." So R-b can't be the rule of inference their argument employs. Suppose, then, that they suggest the following modification:

    R-b' If s is agnostic about p, and s knows that p implies not-q
         given some propositions r-1, r-2, ..., r-n s believes (or
         believes to be plausible), then s should not accept q.

This is a very promising rule of inference. But what are the r-i propositions to be instantiated to? In their original review, Brown and O'Rourke wrote that one of these propositions is what they call "Hypothesis (a): The mental aspects of the brain are determined by its pattern of neuron activity." Another is "Hypothesis (c): The functionality of a computer design is not dependent on the physical material from which it is composed." But the problem is that I believe neither of these hypotheses (in fact, I'm agnostic about both of them). Hence, the antecedent in R-b' (appropriately instantiated) isn't true. And so Brown and O'Rourke's argument fails.

6. Finally, note that it's essential that the agent in the relevant rule of inference have beliefs about the background propositions which enable the implication. This is easy to see by considering the result of stripping R-b' of this feature -- a result which reads

    R-b" If s is agnostic about p, and s knows that p implies not-q
         given some propositions r-1, r-2, ..., r-n, then s should not
         accept q.

This rule of inference is easily counter-exampled: Suppose that Chicago-based Jones is agnostic about whether there is a penny in his piggybank produced before 1967 (p"). Now consider the proposition (r) that if there is such a penny in Jones's piggybank, the Empire State Building (ESB) is not standing. If Jones is swift enough to grasp modus ponens (from if p then q, and p, infer to q), he will certainly know that p" implies that (not-q) ESB is not standing given r. But it hardly follows that Jones should not accept the proposition that ESB is standing. In fact, ceteris paribus, Jones ought to hold that ESB still graces the greatest skyline on the planet.


Bringsjord, S. (1992) What Robots Can and Can't Be. Dordrecht, The Netherlands: Kluwer Academic Publishers.

Bringsjord, S. (1994) Precis of: What Robots Can and Can't Be. PSYCOLOQUY 5(59) robot-consciousness.1.bringsjord.

Bringsjord, S. (1995) Agnosticism About Neuron-Level Functionalism. PSYCOLOQUY 6(20) robot-consciousness.13.bringsjord.

Brown, M. & O'Rourke, J. (1994) Agnosticism About the Arbitrary Realization Argument. PSYCOLOQUY 5(83) robot-consciousness.3.brown.

Brown, M. & O'Rourke, J. (1995) Agnosticism Revisited. PSYCOLOQUY 6(21) robot-consciousness.14.brown.

Cole, D. & Foelber, R. (1984) Contingent Materialism. Pacific Philosophical Quarterly 65.1: 74-85.

Volume: 6 (next, prev) Issue: 30 (next, prev) Article: 16 (next prev first) Alternate versions: ASCII Summary