Leowontin reviews Steven Rose's latest book on the brain

[Chris Burford <cburford@xxxxxxxxxx>]

> From the Lancet March 12 2005

The 21st-Century Brain: Explaining, Mending and Manipulating the Mind


--------------------------------------------------------------------------------
Steven Rose. Jonathan Cape, 2005. Pp 344. £20·00.
ISBN 0-224-06254-9.

In their attempts to make the mysterious complexities of the natural 
world understandable, scientists have over and over again used 
metaphors. Descartes likened the animal body to a machine, and 
physicists tell us that if we are to understand the way molecules of a 
gas behave we should think of them as billiard balls that, unlike any 
real ball, undergo perfectly elastic collisions. Hanging on the wall 
of my study is a German scroll from the 1920s, The Human Factory, 
which depicts the inner workings of the human torso below the neck as 
an interconnected series of assembly lines, supply pipes, spray guns, 
crushers, storage vats, wheels, and pulleys. Then, as one's gaze rises 
to the head, the form of the metaphor changes. At the base of the 
skull, connecting the factory floor to the offices on the upper 
stories, is a telephone exchange with women operators plugging and 
unplugging wires in a switchboard. And in those offices, labelled 
"Willpower", "Intelligence", and "Judgement" are men sitting at tables 
and desks arguing, consulting, and--thinking. Even the electric line 
metaphor of pathways of communication requires operators to decide 
whom to connect with whom. The metaphor for thought is just more 
thought. Technology had not yet provided devices that could serve as 
models for memory, consciousness, and rationality.

We have changed all that. In the three-quarters of a century since 
Fricke and Company produced The Human Factory, science has produced 
elaborate devices for the manipulation and storage of information, and 
these have been seized upon by scientists desperate to make the mental 
concrete. There are computer models in which various regions of the 
brain are analogised to disk memory, in which both data and program 
instructions are stored, chip processors that manipulate the 
information and modify the programs, and input-output circuits for 
connections to the rest of the body. When it became clear that much 
mental information is diffusely located rather than concentrated at 
discrete spots, the brain as hologram was substituted for brain as 
electronic computer. But these devices have not succeeded in capturing 
what is now known about mental processes. The optimism that we will 
understand the brain as just another form of some object we have 
already invented has given way to a renewed puzzlement about how to 
connect the mental with the physical. The naive reductionist 
20th-century Mental Machine has given way to the undoubtedly physical 
yet mysterious entity, The 21st-Century Brain.

Steven Rose, an accomplished investigator of the neurobiology of 
memory and producer of careful essays and books on what is surely 
biology's most difficult question, has not fallen into the temptation 
that has ensnared so many "big thinkers" about thinking. Rose does not 
try to produce the sort of grand theory that has been so attractive to 
scientists anxious to be remembered as the Newton of the mind. Rather, 
he presents us with the outcome of experiments, recorded anatomical 
traumata, and surgical interventions in human beings and other animals 
that have led us to reject simple mechanical models, and he makes 
clear what phenomena must be incorporated (in both senses of that 
word) in a correct physical picture of mental processes. In so doing, 
Rose necessarily destroys the basis for every simple physical model of 
mental function that has been constructed and leaves us hanging where 
we belong for the present, in perplexity.

So what are the essential facts? First, the brain as a whole is 
broadly regionally differentiated and brain imaging studies certainly 
show specific areas of the brain that light up when specific mental 
functions are in process. However, the relation between these areas 
and particular functions is complex. There is, for example, no "memory 
bank" corresponding to a computer memory. Damage to the hippocampus 
interferes with the ability to retain new long-term memories, but 
those memories are not encoded in specific neuronal connections. Over 
time other brain regions and connections become involved in specific 
memories. Second, there is no process of the build-up and then slow 
loss of fixed wiring connections of neurons in the brain that 
corresponds to fixed mental states. Certainly new neurons and synaptic 
connections are being produced and are dying throughout life with, 
alas, a preponderance of losses as we grow older, but these births and 
deaths are not in some one-to-one connection with events remembered or 
individual mental processes. There is not some fixed physical module 
corresponding to the ability to do long division or remembering Pi to 
six decimal places. Third, just as new neurons become physically 
involved in old functions and memories, neurons that are already 
present can increase the number of connections to other neurons over 
time, and thus become involved in a new multiplicity of pathways. We 
do not have neuron-by-neuron catalogues that relate particular cells 
to particular mental states, so it may well be that such multiple 
pathways are not related in any categorisation that makes logical 
sense. Could the dozens of synaptic connections that a single neuron 
makes with dozens of other different neurons be involved in such 
diverse functions as remembering a telephone number and an instant 
ability to construct a grammatically correct sentence in French? As 
the most common phrase in The 21st-Century Brain has it, "Nobody 
knows".

Among the curiosities discussed by Rose is the finding that when 
long-term memories are recalled, it is not the original memory of the 
event that is being referenced, but the most recent recall of it. We 
have memories of memories. I have a vivid pictorial memory of having 
shared the stage with Carl Sagan at a debate about evolution with some 
creationists. I also have indisputable documentary and eye-witness 
evidence that Sagan and I appeared on different days of the event and 
did not overlap. Although convinced of the historical fact, I am 
unable to expunge from my memory that image of our joint appearance. 
The implications of such a phenomenon for the validation of false 
memories are considerable.

The 21st-Century Brain takes up, in its last chapters, the problem of 
the reification of behaviours as concrete psychological pathologies. 
One of the best known cases is attention deficit/ hyperactivity 
disorder. Is there a common neuroendocrinological phenomenon that lies 
at the basis of this claimed entity? The belief that it has a 
corporeal, rather than a purely definitional, existence invites 
chemical intervention. Rose is sceptical of its physical reality and 
the reality of various other newly defined behavioural pathologies. 
But, then again, he may not be a disinterested observer. The behaviour 
that he and I engaged in during the days of the American invasion of 
Vietnam could easily be seen as a manifestation of the disease now 
listed in the authoritative Diagnostic and Statistical Manual of 
Mental Disorders as "oppositional defiance disorder".

After finishing The 21st-Century Brain, the reflective readers may 
find themselves seriously considering the possibility that the human 
species may be extinct before we understand the central nervous 
system.
Richard Lewontin