Milner's (1999a) book has one principal goal that is pursued through repeated soundings of a number of themes. The goal is to hasten a change within what has come to be called cognitive neuroscience from an emphasis on sensory systems to one on sensory-motor integration. The themes are various but include: the importance of the contributions of Lashley and Hebb; the need to update those contributions, particularly the concept of the "engram" or "cell assembly"; the importance of lessons to be learnt from the olfactory system; the need to recognise that (at least 'higher') organisms have a self, which is manifest in motivational and motor response systems; the prevalence of innate structure within the connectivity of the brain; and the importance of reciprocal pathways. This is a small book but it brings together many ideas. The authoritative manner in which they are described and integrated ensures that the attention is held and the imagination stirred.
2. The names of Lashley and Hebb recur throughout the book, unsurprisingly given that Milner was a student of the latter, himself of course a disciple of the former. There has recently been debate as to the appropriate division of accolades between Hebb and Lashley (e.g. Orbach, 1999; Milner, 1999c) but this is not at issue in Milner's book. Here, he seeks to highlight the important contributions made by both his mentors to the understanding of brain and behaviour. By critically assessing and modernising some of their central concepts - ideas they raised against the then prevalent conditioned reflex model of behaviour - he aims to advance the tradition to which he has been heir.
3. For Milner, a particularly appealing feature of Hebb's notion of the cell assembly, or engram, is that it allows the brain to be active independently of sensory input. A cell assembly does not require the presence of the stimulus it represents in order to become active, it may be aroused by association with other cell assemblies representing other stimuli, ideas, motivational states or intentions. This provides a powerful means for understanding the acquisition and association of ideas by the mind, and so also the integration of behaviour. However, certain features of Hebb's formulation of the cell assembly have proved inconsistent with subsequent anatomical and physiological knowledge, as well as with various logical considerations. The concept of reverberation pre-dated discovery of the chemical nature of synaptic transmission, and the principles by which reverberation might be propagated and damped have never been satisfactorily articulated. Milner believes it to be an unworkable concept. He favours instead temporary priming of neurons "produced by sensory or associational input, so that they can be fired for a time by the non-specific arousal system". Additionally, and less controversially, Milner advocates that cell assemblies can enter into a variety of kinds of association. Especially important are associations between assemblies representing stimuli or ideas and those representing response plans or intentions. These associations supposedly provide a basis for selective attention while also preventing the uncontrolled spread of excitation between assemblies. However, while an impressive account of attention forms a central pillar of the book, the latter idea remains somewhat fuzzy.
4. Given that the concept of the cell assembly allows the brain to be active independently of sensory input, Milner early on postulates a non-vitalistic self, which accounts for our sense of behavioural autonomy and hence also for the title of the book. The self is to be thought of as a complex neural mechanism with a predominantly genetic basis. This sounds promising but the idea undergoes very little development in the remainder of the book. The self is identified in an early diagram (p12) with a goal selector "with inputs from needs (usually internal stimuli) and satisfiers of the needs (external stimuli), often called reinforcers. The goal selector has evolved to activate responses that, in the animal's natural environment, are most likely to reduce the need". However, Milner has already pointed out that animals usually thought too simple to have a self must also have a goal selector to cope with approach/approach and approach/avoidance conflicts. There is no discussion of how complex the goal selector has to be to make the leap from non-selfhood to selfhood, or of whether selfhood is a matter of degree. The idea of the self serves mainly to block behaviourist notions of organisms as passively responding to environmental stimuli. Whilst external stimuli may sometimes initiate behaviour, activity often results from an urge to perform that may be due to internal stimuli or the priming of response plans by the non-specific arousal system. This leads back to the central role of selective attention since the goal selector must ensure that only those stimuli relevant to current motivation will gain control of behaviour.
5. Ascription of a genetic basis to the goal selector serves to announce another theme that runs through the book. Hebb had eschewed any inherent organisation of cell assemblies, leaving it to chance associations to wire the pathways of the brain. Milner will have none of this, noting that chance associations could never account for the degree of uniformity found in, for example, the visual and olfactory systems of mammals. For this same reason he is sharply critical of behaviour modelling with artificial neural networks that starts from an assumption of minimal organized structure. Learning should not be taken as a substitute for innate behaviour but understood as an evolutionary development that is itself an inherited ability. The learning of associations and the formation of new cell assemblies are possible only on the back of innately specified and properly functioning assemblies. Indeed, the capacity to learn is an intrinsic feature of such low-level systems as those involved in habituation and rapid storage of new information.
6. Like other contemporary commentators (e.g. Zeki, 1993; Enns and Di Lollo, 2000), Milner stresses the importance of reciprocal, or recurrent, pathways in perception and in the organisation of behaviour. In particular, reciprocal pathways are central to Milner's model of attention, providing the means by which cell assemblies representing intentions and action plans come to control both selectivity and sensitivity in sensory systems. In the olfactory system reciprocal connections project back from cortical and subcortical sites to the olfactory bulb via the lateral olfactory tract (LOT). Much of this LOT feedback is inhibitory, acting to suppress activity in weakly stimulated glomeruli in the bulb. Milner hypothesises that excitatory LOT feedback serves to enhance odour signals likely to be relevant to current motivation. Some part of the activity in the motivation related assemblies must contain a description of relevant olfactory stimulation, such as the odour profile of potential mates. Reciprocal connections play a similar role at various levels within the visual and visuo-motor systems. Inferotemporal neurons that recognise objects must act back to modulate earlier, retinotopic neurons in V1 and V2 that pass to the motor system positional information for manipulation. These early cortical neurons equally feed back to still earlier sensory processes. Reciprocal paths between response release and inhibition control systems are also posited to underlie phenomena like the learning of active avoidance, which has long been problematic for neural theories of behaviour.
7. Many of the above ideas already enjoy considerable currency within cognitive neuroscience but Milner draws them together elegantly and combines them in unique fashion with new or less well known propositions. Ideas are densely packed between its covers, but the book is written with verve and a lightly worn depth of knowledge that make it an engaging and stimulating read.
Enns, J.T. & Di Lollo,V. (2000) What's new in visual masking? Trends in Cognitive Neuroscience, 4(9), 345-352.
Milner, P.M. (1999a) The Autonomous Brain. Erlbaum, Mahwah NJ
Milner, P.M. (1999b) Precis of "The Autonomous Brain" PSYCOLOQUY 10(071) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.071.autonomous-brain.1.milner http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?10.071
Milner, P. M. (1999c) Cell Assemblies: Whose Idea? Book review of Orbach on Lashley-Hebb. Psycoloquy, 10(053).
Orbach, Jack (1998) The Neuropsychological Theories of Lashley and Hebb. MD: University Press of America.
Orbach, J. (1999) Precis of: The Neuropsychological Theories of Lashley and Hebb. PSYCOLOQUY 10(029) ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.029.lashley-hebb.1.orbach http://www.cogsci.soton.ac.uk/psyc-bin/newpsy?10.029
Zeki, S. (1993) A Vision of the Brain. Oxford: Blackwell.