Commentary on: Block, N. (1995). On a confusion about a function of consciousness. Behavioral and Brain Sciences,18, 227-287.

Published in: Behavioral and Brain Sciences, 18, 248-249, 1995.

Consciousness without conflation

Anthony P. Atkinson and Martin Davies
Department of Experimental Psychology, University of Oxford, Oxford  OX1 3UD, England.;

Abstract: Although information-processing theories cannot provide a full explanatory account of P-consciousness, there is less conflation and confusion in cognitive psychology than Block suspects. Some of the reasoning that Block criticises can be i nterpreted plausibly in the light of a folk psychological view of the relation between P-consciousness and A-consciousness.

The history of cognitive psychology is studded with attempts to associate or identify various aspects of consciousness w ith information-processing constructs.  Consciousness is putatively explained in terms of the operation of, for example, short-term memory, attentional, and central executive systems, often with a (usually serial) "limited capacity".  According to Block, much of this psychological theorising is undermined by a failure to distinguish between P-consciousness and A-consciousness, but we do not think that what is going on here is a simple conflation.

We agree with Block that there is an importan t distinction to be drawn between P-consciousness and A-consciousness, and that P-consciousness leaves us (at least in our current state of understanding) with an explanatory gap (Davies, 1995; Davies and Humphreys, 1993).  Nagel (1974) says that "st ructural features" of experience "might be better candidates for objective explanations of a more familiar sort." But even if these structural features are explained in information-processing terms, we are still left with the question of why there should be something rather than nothing that it is like to have certain processes going on in our brains. Current information-processing theories of P-consciousness are bound to be incomplete whereas (we assume with Block) there is no similar obstacle in the way of an information-processing explanation of A-consciousness.  To that extent, standard cognitive psychological accounts of consciousness are more appropriate to A-consciousness than to P-consciousness.  But that is not to say that there is a sy stematic tendency toward confused theorising grounded in the failure to distinguish between the two notions of consciousness.

In the context of an investigation of "automatic" and "controlled" processing, Shiffrin and Schneider (1977, p.157) speculate , for example, that "the phenomenological feeling of consciousness may lie in a subset of STS [the short term store], particularly in the subset that is attended to and given controlled processing."  It is easy to imagine an objection: what purports to be an account of an aspect of phenomenal experience is cast in terms of storage and processing, terms that are appropriate for a theory of A-consciousness (cf. what Block says about Baars, Shallice, and others).  But although Shiffrin and Schneide r do not fully settle the question of the causal relation between being in the special subset of STS and being subject to controlled processing, there is no real evidence of conflation here.

A more sympathetic view of the cognitive psychological liter ature is possible if we begin from the plausible idea that P-consciousness may plausibly figure in the causal explanation of A-consciousness. Why do we say that the idea of a causal relation running from P-consciousness to A-consciousness is plausible? Br iefly, Block’s notion of A-consciousness is a dispositional notion; and when a state has a dispositional property, it is natural to seek a more intrinsic property of the state in virtue of which it has that disposition.  So, we can ask, in virtue of what property of my pain state am I in a position to report that I have a pain? Or, in virtue of what property of the pain in my leg is it the case that the content I have a pain in my leg is poised for rational control of my actions?  The int uitive folk psychological answer is that these dispositions are grounded in my pain’s being a phenomenally conscious state. It is because the pain is P-conscious that it is A-conscious.

An A-conscious belief likewise has dispositional properties, and once again we may ask for a property of the belief state that explains why the content of the belief is poised to figure in theoretical and practical reasoning, and why I am able to express and report the belief.  On some accounts of the distinction between P-consciousness and A-consciousness, on which beliefs are not P-conscious states, this question proves to be problematic.  But Block is explicit that P-consciousness extends to thoughts, so he can allow the answer that it is in virtue of bein g a P-conscious state that a belief has the dispositional properties characteristic of A-consciousness.

This folk psychological view of the relation between P-consciousness and A-consciousness cannot currently be fully reflected in information process ing psychology, since — given the explanatory gap — we cannot give a full explanation of P-consciousness in information processing terms.  But a partial reflection would be seen in the idea that the conditions that explain "structural features" of ph enomenal experience might be found among the immediate antecedents of the processing that underpins reasoning, decision taking, and reporting.  This view would also encourage the thought that there is an asymmetric dependence relation between P-consc iousness and A-consciousness, as follows.  If, as we are actually constituted, P-consciousness is the categorical and relatively intrinsic basis for the dispositional and relatively relational A-consciousness, then we should expect there to be actual cases of P-consciousness without A-consciousness produced when crucial relational links are missing.  But, we should not expect to find actual cases of A-consciousness without P-consciousness.  This is just the asymmetry to which Block points ( Sections 4.1, 4.2).

Given the possibility of this more sympathetic reading of the cognitive psychological literature, what are we to make of the target reasoning that Block criticises?  We shall surely agree that some arguments are nonstarters.&n bsp; If P-consciousness is actually present when flexibility in behaviour is absent (as in the epileptic seizure case), then any argument for the addition of flexibility as a function of P-consciousness is undercut. In the cases of prosopagnosia and blind sight, however, important aspects of normal conscious experience are plausibly absent: there is a P-consciousness deficit. And even if there is covert knowledge of the identity or profession of the person whose face is presented, this information is not a t the service of rational decision taking: there is an A-consciousness deficit. We agree with Block that it would be a mistake to infer anything about one-way causal dependence, or about the function of P-consciousness, given only the association between these two deficits. But still, their cooccurrence is consistent with, and makes sense in the light of, the folk psychological view of the relationship between P-consciousness and A-consciousness.


Davies, M. (1995). Consciousness and the varieties of aboutness.  In: The Philosophy of Psychology: Debates on Psychological Explanation, ed. C. Macdonald G. Macdonald.  Oxford: Blackwell Publishers.

Davies, M. and Humphreys, G.W. (1993). Editors’ Introduction.  In: Consciousness: Psychological and     PhilosophicalEssays, ed. M. Davies & G.W. Humphreys.  Oxford: Blackwell Publishers

Nagel, T. (1974). What is it like to be a bat? Philosophical Review, 83, 435-450.
Shiffrin, R. M. & Schneider, W. (1977). Controlled and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory.  Psychological Review, 84: 127-190.