Derek Bolton (2002) The Role of Encoded Information in Explanation. Psycoloquy: 13(025) Mental Disorder (4)

Volume: 13 (next, prev) Issue: 025 (next, prev) Article: 4 (next prev first) Alternate versions: ASCII Summary
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 13(025): The Role of Encoded Information in Explanation

Reply to Griffin and Clune on Bolton and Hill on Mental Disorder

Derek Bolton
Psychology Department
Institute of Psychiatry
London SE5 8AF

Jonathan Hill
University Child Mental Health
Royal Liverpool Children's Hospital
Mulberry House
Eaton Road
Liverpool L12 2AP


In their reviews of Bolton and Hill (1996) (MMMD) Griffin (2002) and Clune (2002) identify some potential problems. Griffin points out that some aspects of folk psychology may turn out to be unsupported by the science. Generally, however, MMMD does not rely on folk psychology, and indeed its theory of intentionality is explicitly applied to biological systems not only psychological, where folk or any other psychological concepts would be out of place. Clune suggests that the evolutionary analysis together with advancing knowledge and technology may make the concept of intentionality redundant. We reply that the case for intentional causality is embedded in an evolutionary perspective and that the case for intentionality in biological systems does not depend on inadequate knowledge.


mental disorder, causality, intentionality, meaning, reduction
1. Griffin is correct to identify as central to Mind, Meaning and Mental Disorder (MMMD) the idea that the cognitive-behavioural sciences need a theory of cognition based on organism-environment interactions, and right to pick up for example on the promising if unfashionable use of Wittgenstein in explicating this idea. Griffin's expressed misgivings (sections 7-9) have to do mainly with our apparent realism about the 'theory of mind' as embodied in folk psychology. Folk psychology may be wrong about cognitive processes and functional architecture, and some accumulating evidence suggests that in some circumstances it is wrong. For example, invoking propositional-attitude based theory to model social behaviour may be wrong. One version of the general problem, to do with connectionist-based criticisms of Fodor's language of thought hypothesis, is considered in MMMD (ch.2), but our general reply is as follows. The position sketched in MMMD is committed to the following:

(a) the science of behaviour (the science that wants to explain and predict animal including human behaviour) - where behaviour is understood in terms of effective, goal-directed behaviour-in the-environment - needs to posit that the organism (among other things) encodes information about the environment; (b) the attribution of specific informational content depends on the (range of) behaviours (actual but including also hypothetical and counter-factual) that is to be explained; (c) how the organism encodes the information it needs in order to behave at it does (or might or would) is up to the science to find out - no a priori claims here; (d) specification in any particular (kind of) case of the form of information (e.g. as a propositional attitude, or theory-like network of propositional attitudes), or of content, might be wrong.

2. There is no a priori given form and content of information that regulates behaviour - it is all an empirical matter. Examples are diverse, and the theory of intentionality in MMMD is explicitly applied not only to psychological systems, but also to biological systems, where folk or any other psychological concepts would be out of place. As things stand, on the other hand, it is difficult to avoid - even if one tries to - use of folk psychological concepts to specify the form and content of cognition and behaviour, and this applies when doing cognitive psychology and neuropsychology also. But this reliance on a folk theory should not be mistaken for the view - and it was not the intention of MMMD to claim - that some its specific characteristics, such as invoking propositional-attitude based theory, are fixed once and for all. The validity of folk psychology is not supposed to be a fundamental matter for the position in MMMD. What is fundamental, as in (a) and (b) above, is the proposition that science needs to posit information-carrying, intentional states in order to explain and predict behaviour. The real problem for the position in MMMD would therefore not be that the cognitive mechanisms underlying our navigation of the social landscape are not theory-like, but rather would be that they involve no encoded information about the landscape.

3. In most respects Clune has provided a very fair summary of our thesis. We will point out some differences in emphasis between his and our accounts, before taking on his objections. Clune gives a lot of weight to our discussion of the way systems go wrong, and less to our views on how they operate successfully. However our analyses of disorder depend on the success of our account of adaptive functioning. Clune does not bring out that at the heart of our argument is the case that intentional causality has to be invoked to explain how all biological systems work, and that no matter how 'simple' these processes may be, they have properties that provide the basis for psychological functioning. In many respects, processes in all biological systems more resemble those in psychological systems, than they do those in physical systems. Equally we do not argue that bacteria can think. There is a massive distance between the operation of simple biological systems and the mind, which we discuss.

4. In his point 13 Clune says, 'The authors ... cannot conceive of how physical states have intentionality, except through encoding' and then proposes that this may reflect a limitation in our current state of knowledge. We argue that it is not that we cannot conceive of it, but rather we can show that physical and intentional states are radically different. What do we mean by a physical state? It is a state defined by parameters such as mass, force, velocity, energy or electrical charge. These are used in equations that account for the way the state has been achieved, and what its future will be. The equations summarise the physical laws covering the ways in which these properties interact. By contrast the intentional state is defined by the information in the system about a state of affairs elsewhere, within or outside the system, with implications for action in relation to that state of affairs. The links are specified by rules of correspondence. These link physical events to the state of affairs in one way, but other ways can be envisaged that preserve the same informational content. That is to say rules are defined by a convention, rather than a natural state of affairs defined by physical equations. The equations specifying the physical state simply do not refer to these intentional processes.

5. Then Clune questions whether we need to invoke these intentional processes in our explanations of disorder. Our aim is first to show that we cannot provide a causal account of biological processes without intentionality, and then to show that this opens up the possibility that psychiatric disorders may be explained by non-intentional, physical disruption of function, or by an undermining of intentional processes from within. Clune proposes that this can be bypassed through the development of evolutionary psychiatry. We have absolutely no problem with an evolutionary model, indeed we argue that it is because events in a biological organism are not predicted or determined by physics, that the setting for the transmission of information and adaptation is created. For example the nucleotide sequence in DNA is determined by the template provided by previous generations. Its causal role in relation to protein synthesis depends on the fact that it codes for amino acid sequences that lead to the production of functional proteins. If the sequence of nucleotides were determined by their physics, it would not be available to carry that information. Furthermore, the occurrence of chance differences in the code, amounting to mistakes (defined as differences that have implications for function), is the motor of evolution. The same physical difference can contribute to disease or health. We use the example of sickle cell trait that under some conditions increases risk of episodes of intense pain, and under others to protection from malaria (Bolton & Hill, 1996, p. 236). The functional implications of a physical difference may be advantageous or disadvantageous, depending on the fit between the organism and its environment.

6. So when Clune asks 'why not work on explaining how patterns of neurones firing are supposed to fit (behaviour and) the environment,...' we answer that this is what are doing, and when he continues, '.... and forget about intentionality altogether?' we say it can't be done. Patterns of neurones firing, for example in the optic nerve, have intentionality with respect to events in the external world. The 'equation' links the firing to those events specifies a correspondence between changes in light intensity, location, movement, and so on. It is not defined by the physical qualities of the stimulus, but rather by the convention used by that organism to encode changes of interest. We have used the example of the visual system of the horseshoe crab, Limulus, to make the point (Bolton & Hill, 1996, p. 241-2). The frequency of impulses in the optic nerve of the crab is roughly proportional to the logarithm of light intensity in its surroundings. This is adaptive because the crab is exposed to light intensities that vary by a factor of ten million. If the frequencies bore a linear relationship to these variations, the range of frequencies required to encode the information would, at the lower end of the range, require impulses every several thousand seconds, which could not result in effective signalling in relation to action in that environment. The physics of the light intensity does not determine the frequency of the impulses. These are determined by a rule encoded in the crab's DNA. The changes in light intensity could be encoded in many different ways. The scope for the rule to vary, in contrast to the physics that cannot, creates the conditions for the adaptation of the organism to the environment.

7. Once the nature and the necessity of intentional causal explanations throughout biology are accepted, the next step is to describe the ways in which intentional processes in humans differ from those in simpler organisms. Essentially this entails a development of the properties of intentional processes, such as that if events in the external world can be interpreted under one set of rules, then they may be interpreted under many others. This we argue provides the basis for a flexible and complex social intelligence, and also for disorder. A full summary of our ideas on psychiatric disorder is, however, beyond the scope of this response.


Bolton D. and Hill J. (1996) Mind, Meaning, and Mental Disorder: The Nature of Causal Explanation in Psychology and Psychiatry. Oxford University Press.

Bolton, D. & Hill, J. (2001). Mind, Meaning & Mental Disorder: The Nature of Causal Explanation in Psychology & Psychiatry PSYCOLOQUY 12(018) Precis of D.Bolton & J.Hill on Mental-Disorder (1)

Clune, A.C. (2002). Mental disorder and its cause. PSYCOLOQUY 13 (018) Book Review of Bolton and Hill on Mental-Disorder (3)

Griffin, R. (2002). Mind, Meaning and Cause: so what if the mind doesn't fit in the head? PSYCOLOQUY 13 (015) Book Review of Bolton & Hill on Mental-Disorder (2).

Volume: 13 (next, prev) Issue: 025 (next, prev) Article: 4 (next prev first) Alternate versions: ASCII Summary