When are the neighbours in?

These days there is an interdisciplinary spirit abroad in the philosophy of mind. Arguably much of the most interesting work being done in this field draws extensively, and in an informed way, upon what is going on in other disciplines. Both of these books bear witness in different ways to that spirit of collaborative engagement.

Theories of Theories of Mind collects together papers by philosophers, psychologists and primatologists based upon their contributions to a series of interdisciplinary workshops sponsored by the Hang Seng Centre for Cognitive Studies at Sheffield. Here "theory of mind" is a term of art used to designate the ability, possessed by most human beings and, possibly, some other species, to make sense of the minds of others.

As a domain of research this is a particularly good example of the fertility of interdisciplinary collaboration: the question "do chimpanzees have a theory of mind?" posed by primatologists, led, via suggestions from the American philosophers Daniel Dennett and Gilbert Harman, to the development of the "false-belief task" used by developmental psychologists to investigate the acquisition of a theory of mind by children. The idea behind the task is that the correct attribution of a false belief to another person is evidence of the ability to think of others as having beliefs about the world which are distinct from one's own. Normally children become able to succeed at the task between the ages of four and five years old. In autism, however, success with the task is significantly delayed, if it is attained at all. This has led to the proposal that the primary deficit underlying autism is the absence of a theory of mind in the autistic subject.

These elements are all represented in the collection. The longest section, however, concerns the issue of what underpins the theory of mind abilities of mature human adults - how do we explain and predict the actions of ourselves and others? Here the field is occupied by two broad alternatives. "Theory"-theory attributes "mind reading" abilities to the possession of a theory of the psychology of others that sustains explanation and prediction of their behaviour. Simulation theory, on the other hand, denies that these abilities are due to the grasp of a theory, holding instead that they arise from our capacity to project ourselves in imagination into other people's perspectives, simulating their mental activities with our own.

There are papers by champions of each of these positions. Jane Heal, for example, argues against a thorough-going "theory"-theory view on the grounds that it would have to include a theory of relevance allowing us to predict which of the myriad beliefs which comprise a person's world-view it is appropriate to consider in any particular circumstance - such a theory would be massively complex and, insofar as we have no idea what form it might take, deeply tacit.

Nichols et al argue against simulation theory that it cannot adequately account for our failures to predict behaviour in some situations. On the simulation view such a failure must be due either to the subject of prediction being different to the predictor, or because the wrong beliefs and desires are fed into the simulation, and it is claimed that there are cases of mistaken prediction that fall into neither category. Only if prediction is based on a theory, it is argued, subject as it is to error in its theorems, can we explain these mistakes.

These papers, while they do not resolve the debate, suggest that a consensus may be emerging involving elements of both broad views. The collection is valuable both as an introduction to an exciting area of research and as a snapshot of the current state of the debate.

The Postmodern Brain seeks to forge different kinds of interdisciplinary connections, leaping the apparent chasm between contemporary philosophical reflection upon cognitive science and the postmodern philosophy of Heidegger and Derrida. Gordon Globus holds that we are bound to make no progress in understanding how the brain is the physical substratum of the mind so long as we see the brain as a computer engaged in the manipulation of representations. He sees in biologically realistic neural networks the resources for a noncomputational conception of brain functioning, the characteristics of which are the basis of the connections he strives to establish between this radical connectionist view and some central concepts in postmodern continental philosophy. A deconstructive approach is applied to types of narrative on aspects of the mental, including the computational theory of the mind and classical dynamic psychotherapy, which essentially embody the "metaphysics" that postmodern philosophy rejects. And in parallel, the postmodern, connectionist picture of brain functioning is applied to the explanation of mental illness and dreams.

In the field of cognitive science there is as yet no consensus on how connectionist models relate to computational theories of cognition. Are neural networks simply implementations of computational processes by a brain-like structure, or, more radically, do connectionist models rival and replace classical computational models? Globus sketches some arguments from the literature for taking the latter position.

The central contention of this book is that there is some interesting connection, or "resonance", between postmodernism and the radical connectionist view of brain functioning. Echoes there may indeed be though I found them difficult to assess, but they seem insufficient to sustain claims about the postmodern brain. The attempted assimilation seems to me to fail to sufficiently address the challenge that from a postmodern perspective the radical connectionist model of brain functioning is itself engrossed in "metaphysics", and as such is grist to the deconstructionist mill. Indeed a related methodological tension pervades the whole book, with the writer moving untroubled between "constructive", linear accounts setting out the connectionist model and making the connections with postmodernist concepts, and the deconstructive treatments elsewhere. On the whole I found this book rather thinly argued and, in places, decidedly opaque.

David Elwell is a psychiatrist and graduate student in philosophy, University of Oxford.