Customer Reviews

Neural Darwinism: Theory of Neuronal Group Selection (Oxford paperbacks)

5 star:		(5)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(1)

 

 

The distinguished Nobel Laureate proposes a global brain theory that demonstrates that the brain does not work like a computer but rather operates under principles of selection that assure individuality, autonomy, imagination, etc. Since this book was published in 1986, the essentials of its proposals have been confirmed and absorbed at almost all levels of neurobiological and psychological inquiry. More accessible are two subsequent books, "The Remembered Present" and "Bright Air, Brilliant Fire"

Mike Vanier's experience with Edelman's prose gave the typical bioassay result: its hard to read Edelman's books. I often try to imagine the state of mind of people in 1875 who tried to wade through Darwin's "Origin of Species" or someone who came across the work of Gregor Mendel in the 1890's. Unfortunately for the Science of Mind, Mike is just the kind of person Edelman might have hoped to be able to reach. Well, Mike, did you read right through the Bible (or substitute "Your First Calculus Textbook" for "Bible") the first time you picked it up? There really is a forest in "Neural Darwinism" once you get past the trees.

The claim "his ideas are neither new, nor original, nor correct" is one of the standard put-downs of the academic world. Anyone who works on non-trivial scientific issues and is intelectually honest will admit that his work in based on ideas taken from others and that his work is incomplete and contains errors. Edelman makes these admissions. Edelman's ideas about how brains can learn and function to produce what we experience as minds are positive contributions to science and worth getting to know.

Nobelist Gerald Edelman "theory of neuronal group selection" can be taken to provide a neurological understanding for psychoanalytic theory and experience. Because of the dense overlapping of dendrites and
axons in gray matter, a given area of cortex is capable of a varying
array of responses to a given input. Of the many possible responses,
one inevitably leads to the strongest, most adaptive, or rewarding
output. Suppose that this "fittest" response were "selected" for

synaptic changes enhancing the likelihood of future firing of just
that pattern of response when the same or similar input next arrives.
That pattern of function would have "won" in a Darwinian competition
to dominate the activity of that group of neurons when those same or
similar experiential conditions occur again. (I hope the reader can
hear in this a basis for transference experience: the neuronal response previously selected, perhaps by childhood experience, will be reactivated again in the future under specifically evocative conditions.)

We can thus anticipate a direct neurophysiologic account for how
"object relations" may in part derive from internalized__introjected__ experiences with objects and with their functions. Each experience in present real time consists of, is generated by, and resides in the activation of neural groups, interconnected in an ad-hoc network, distributed throughout the brain anatomically, and thus involving many functions of sensation, perception, motor function, emotion and cognition. The specific functioning of that network just then may be in effect "selected" by facilitative synaptic changes; its components might be be predisposed to fire together again under the right conditions.

Reentrant signalling over anatomically complete loops probably allows
anatomically distributed (distant) neuronal groups continuously to
communicate with one another and to synchronize their activities,
according to Edelman. Groups active in one area (eg, visual) can thus
reciprocally excite and cause synaptic enhancement of groups in a
separate area (eg, tactile). Edelman points out that particular
patterns of response in one brain "map" (consisting of many groups)
thus can become functionally associated with particular patterns in
other(s). Responses to present inputs can thus be linked to previous
patterns , across maps. Thus, a sight can stimulate the memory of a
touch. A sound, perhaps a spoken word or melody (obviously lots of unaddressed complexity here) can stimulate and reactivate a visual, operational, or cognitive memory, and, "inscribing" on it in the present, change it.

A mapping of many such selected groups could thus provide the neurophysiologic correlate to and substrate for unconscious object representations, which could potentially, if later reactivated, be expressed as affect-full object-related fantasies, impulses, and behaviors, potentially conscious, yet, perhaps exerting a neurologic influence that could remain unconscious.

A potentially responsive neural network representing past experience
will be activated, Edelman theorizes, if current input is a "good
enough match" to previous input which led to the original synaptic
enhancement. That is, if enough of its neighboring, or synaptically
linked, groups are activated ("the present context"), it will be
activated too. Thus, the present experience is categorized according to previous experience. Thus, presentation of a current object (eg, the
analyst) may activate a sufficient number of groups previously
functionally related in experience of the original object ( eg, a
parent), and thus a "memory", specific feeling state, or a behavioral repetition may arise.

This perhaps provides some neurophysiologic correlate to Charles Spezzano's recent observation that, "the immediate
situation is idiosyncratically and unconsciously put by us into a category of
situations and it is this idiosyncratic and unconscious categorizing that
leads to the feeling we have... Still, the patient may hear us saying that,
while we see them putting the immediate event into a category that leads to
a feeling, we don't understand why it has to be put into that category (don't
understand in the sense that we see how it could just as easily be put into
other unconscious meaning categories and is categorized by the patient in an

idiosyncratic way, maybe one we trace historically..."

How does the daily free-associative process work? The superolateral prefrontal cortex, "connected to a multitude of (remote) sensory association areas and limbic and paralimbic cortex," apparently may address and retrieve representations of memories from many fields distributed in the brain. It has been shown in pet scan studies consistently to be activated in tasks which are not run by external stimuli, ie, it is activated in tasks (eg, "visualize walking down your street and turning left at the corner") requiring organization by the brain itself of itself. (ROland, 1993)

In Edelman's model, the continuous reciprocal communication via reentrant signalling between maps of neural groups activated in current experience and those previously enhanced and synchronized in past experience (eg, with the original objects) allows the two sets reciprocally to "translate" or to "inscribe" their activities upon one another. This furthers our account of the brain activity in free association. The functional interaction between maps may permit continuous rearrangement according to unpredictable variations in the environment, (eg, the analyst's challenging an accustomed assumption, or "destabilizing a compromise formation", with a question, or action.) THus, past and present are recategorized in terms of one another, producing change in neurologic and thus in psychic structure. The present real experience with and of the analyst thus can "insinuate itself" into the prior arrangements ( the prior neuro-psychic structure), necessarily reorganizing the object relations constellation (transmuting internalization).

"We do not simply store images or bits but become more richly endowed with the capacity to categorize in connected ways." (Rosenfeld, 1986) Increasing the number of neural arrangements by which one experience is compared to, contrasted with, and categorized according to prior experience or fantasy allows greater richness and flexibility in functioning, one result of a successful psychoanalysis.

Samuel T. Goldberg, M.D.
Baltimore-Washington Institute for PSA

I read this book about 15 years ago and whilst it was exciting and inspiring I was frustrated by not being able to understand how on Earth this was Darwinism. What a relief when I found that Francis Crick had said of Gerald Edelman's theory of Neural Darwinism that because it didn't propose any true replicators in the brain it was more appropriately called Neural Edelmanism. 150 years after Darwin, why do two Nobel Prize winners disagree as to what is and is not a Darwinian natural selection algorithm? Recently we've developed a new theory of Darwinian Neurodynamics which attempts to correct the conceptual errors that underlie Edelman's Neural Darwinism by proposing how true units of evolution could exist in the brain. See [..] for some papers that critically review this book and Edelman's theory. I hope these serve to clarify some confusions.

It seems complicated, but it is not. Edelman deserves to be read. Precise and humble writing. Strong arguments.

Superlative. Outstanding scientific thinking and a must read for every neuroscientist. A few parts are too scantily discussed, but the text is (overall) of the highest caliber

I got about one-third of the way through this book, and then just couldn't continue. This is without a doubt the worst science book I've ever (tried to) read. Edelman goes out of his way to use unnecessarily long, ponderous phrases to describe simple concepts. This is presumably meant to impress you, but personally it makes me wonder what the author is trying to hide. How about this: that his ideas are neither new, nor original, nor correct.