Rodrigo Ventura wrote:
>> Hi. I'm a researcher in the field of AI (artif. intelligence),
> my work is about artificial emotions (or emotion-based agents, if you
> prefer the name), but I am extremely interested in neuroscience. In
> our AI discussion group we have been debating the following issue, but
> we are kind-of stuck... Here is my question to this forum:
>> Let's assume a functional division of the CNS in three layers
> [MacLean]: a "reptile" brain which is based in reactive behavior, a
> "mamalian" brain which involves emotions (limbic system), and the
> neocortex associated with reasoning, etc. The question is: how can the
> reactive layer (e.g., the spinal cord response to a burning stimulus)
> and emotions (e.g., fear) be functionally distinguished. I mean, how
> can one describe these two layers without resorting to morpholigic
> arguments (such as this involves the thalamus, or that involves just
> the spinal cord)?
>> Another question that is related with this is: when did the
> limbic system appeared, in the species evolutio, in the brain
> (reptile? mamalian? birds? ants? bees? spiders?)? And more
> imposrtanty, what was the _qualitatively_ step that evolution took
> when emotions appeared? What was the shortcomings of the non-emotional
> animals that lead to the appearence of emotions?
>> I'd appreciate any response/discussion around these questions,
> either by email or in this newsgroup.
I think that is a _very_ interesting model. The tri-partite brain might
be too simple for real life (and for LeFever PhD...) but we have to
start with simplification to understand the bits.
If you look at the developmental stages of the brain from the neural
tube, then you will see that the tri-partitioning is related to the
spatial location on the "tube".
The "reptile brain" may cover the area of the tube from the spinal cord
until the diencephalon. Areas in between include the hypothalamus (for
autonomic control). Most part are thus here for (as you said) reactive
behaviour, or _reflexes_. After this part there is the limbic structure
associated with emotions, and behind this is the neo-cortex with sensory
functions and fine motor control.
OK. I think it should be clear that these partition correspond to reflex
behaviour, emotional behaviour and controlled (or conscious?) behaviour.
Assume that all sensory input and outgoing motor behaviour _must_ go
through the spinal cord.
For reflexes to occur, something is sensed and immediately the "reptile
part" induces a behaviour. No emotion nor consciousness is involved
because to neuronal information reaches the "mammalian" or "human" part.
When the information does reach the mammalian part, then this interact
with the emotions. IPO the function of emotion is to _amplify_ or
_suppress_ reflex behaviour.
When the information reaches the "human" part (neocortex) we can
manipulate the information in a "conscious" way, make relation, estimate
its effects and mix it with experience. IPO the function of the
neo-cortex is to _induce_ (conscious) voluntary behaviour.
The problem with even conscious behaviour is that is _has_ to travel
through the "mammalian" (emotional) and "reptile" (reflex) part. So in
the end human behaviour is a mixture of conscious, emotional and reflex
components. It cannot be seperated, unless we will be able to shortcut
motor-control directly to the cortex.
On the evolution, I belief that the development of the midbrain and
cortex are not on purpose (refer to Dawkins), it just _happened_ and
happened to be good. Simply, we could say that the tube grew longer and
thicker during the course of evolution. While in the spinal cord the
neurons were nicely seperated in parallel bundels, the cells in longer
and thicker tubes just got incredibly mixed up in a huge mess (that grey
stuff in your brain). But the advantage was that cells were now
connected which opened a lot of possibilities.
I look forward to reflections.
Department of Medical Physics and Bioengineering,
University College London.
Office: 0171 - 209 6415 Fax: 0171 - 209 6269
Home: 0171 - 794 5243 E-mail: kwee at medphys.ucl.ac.uk