On 2/24/2012 22:20, Terren Suydam wrote:
On Fri, Feb 24, 2012 at 4:47 PM, acw<a...@lavabit.com> wrote:
On 2/24/2012 20:51, Terren Suydam wrote:
On Fri, Feb 24, 2012 at 3:30 PM, Terren Suydam<terren.suy...@gmail.com>
wrote:
On Fri, Feb 24, 2012 at 2:27 PM, meekerdb<meeke...@verizon.net> wrote:
On 2/24/2012 10:26 AM, Terren Suydam wrote:
I certainly will. In the meantime, do you have an example from Damasio
(or any other source) that could shed light on the pain/pleasure
phenomenon?
Terren
http://www.hedweb.com/bgcharlton/damasioreview.html
I think emotions represent something above and beyond the more
fundamental feelings of pleasure and pain. Fear, for example, is
explainable using Damasio's framework as such, and I can translate it
to the way I am asking the question as above:
Question: What kind of organization arose during the evolutionary
process that led directly to the subjective experience of fear?
Answer: A cognitive architecture in which internal body states are
modeled and integrated using the same representational apparatus that
models the external world, so that one's adaptive responses
(fight/flight/freeze) to threatening stimuli become integrated into
the organism's cognitive state of affairs. In short, fear is what it
feels like to have a fear response (as manifest in the body by various
hormonal responses) to some real or imagined stimuli.
Yes, that seems to be mostly it, but it's subtler than that. Those internal
states that we have also include expectations and emotional memories - it
can lead to the memory recall of various past sensations and experiences.
Certain internal states will make certain behaviors more likely and certain
thoughts (other internal states) more likely. We cannot communicate the
exact nature of what internal states actually are - the qualia, but beyond a
certain point we cannot say anything more than that we have them and us
having them will usually correspond to some internal states in our instance
of a cognitive architecture.
You can substitute any emotion for fear, so long as you can identify
the way that emotion manifests in the body/brain in terms of hormonal
or other mechanisms. But when it comes to pain and pleasure, I don't
think that it is necessary to have such an advanced cognitive
architecture, I think. So on a more fundamental level, the question
remains:
What kind of organization arose during the evolutionary process that
led directly to the subjective experience of pain and pleasure?
That's a very interesting question. Pain and fear means aversion towards
certain stimuli - that is, reducing the frequency that some stimuli will be
experienced, which can lead to increased survivability. Pain is
unfortunately a bit more complicated than that, it leads not only to future
aversion, but involuntary action-taking - forcing an immediate quick
response, which may not be backed by conscious thought. It can be seen as
unpleasant, because it combines the memory of constantly being forced to
have to take involuntary actions and the actions being aversive. Such
involuntary actions can also be seen as a huge change in attention
(allocation) - one becomes much less capable of consciously directing their
attention.
All of that makes sense, but pain is more than unpleasant. Pain can be
blindingly horrible... ask any migraine sufferer. What accounts for
the intensity of such experiences? I'm asking this in terms of "how",
not "why". How does it get to be so intense.
Intense pain can make us scream or do things we would never do normally
- irrational responses, but possibly advantageous when they first
evolved. We could make a mechanistic theory for how pain manifests.
Someone might suppress their reactions to pain with effort, but that
doesn't mean that there weren't circuits triggered that would have led
to certain actions if not for conscious effort (attention allocation)
involved in preventing such behavior. Maybe we could see pain as the
intense desire to perform certain immediate actions in response to some
stimuli, against our better judgement. In the mechanistic version (when
we look at the architecture and what it represents) we would see that
the most likely outcome would be such random actions being performed.
Actually accounting for the exact nature of the internal state beyond
communicable parts (intensity of desire, involuntary reactions, etc)
might not even be possible for any such theory. At best we might end up
translating - "X is a locally accessible goal, we expect goal X to lead
to pleasure or fulfillment of subgoals or expectation of state to change
in what we expect to be our favor or ..." as "we desire X". Many similar
translations could be done for other emotional responses and more basic
drives - the body can only "do", but we think we can "want". Thinking
about this in detail in the a mechanistic framework tends to end up as a
deconstruction/explanation for what exactly "will" is.
Pleasure is similar, but in reverse - it makes certain actions more likely
to be performed, possibly even leading to some feedback loops. However, it
seems that in humans, pleasure and compulsion have similar and almost
parallel circuits, but are not identical. Pleasure may also have calming
effects by reducing responses/actions instantly, the opposite of pain, while
also making it more likely that actions that caused pleasure to be performed
again - which is a bit similar to compulsion. In a nutshell, they correspond
to mechanisms which lead to certain actions being more or less likely, and
this eventually leads to complex goals and behavior - I'd say that's a huge
reason for pain/pleasure responses to have evolved.
I have the same issue with this description of pleasure. What accounts
for the intensity of peak pleasure experiences?
How does pleasure affect all your other senses, what parts are just
"distilled" pleasure? Pleasure seems more like a way of modulating other
experiences. Pain seems like a way of causing certain experiences or
having the conflicting desire of doing some things.
Or put another way, what kind of mechanism feels pleasurable or
painful from the inside?
The notion of feeling is more complicated because it involves memories and
complex feedback loops.
Presumably the answer to this question occurred earlier in the
evolutionary process than the emergence of fear, surprise, hunger, and
so on.
I like these articles/videos on how AGIs may get emergent emotions from
simple basic drives:
http://agi-school.org/2009/dr-joscha-bach-understanding-motivation-emotion-and-mental-representation
http://agi-school.org/2009/dr-joscha-bach-understanding-motivation-emotion-and-mental-representation-2
http://agi-school.org/2009/dr-joscha-bach-the-micropsi-architecture
http://www.cognitive-ai.com/
Thanks for the references, I hope for a little more time to read/watch
them some time. :-]
Terren
To go a little further with this, take sexual orgasm. What is
happening during orgasm that makes it so pleasurable?
My guess is that it's a fairly complex emotional and somatic response that
could get broken down into simpler parts. You could ask the same question
differently: what makes some music good? what makes some food delicious?
what makes a picture beautiful? The answers to those questions depend
heavily on the person and are probably not easy at all due to their sheer
complexity.
I chose orgasm because of the simplicity of its conception in terms of
evolution and adaptation. In other words, it seems possible that any
animal that reproduces might experience pleasure in the act of doing
so... in some theory of mechanistic pleasure yet to be articulated.
Maybe look at it as nature's "artwork"? Orgasms themselves don't seem as
simple as one might think - there's a lot of additional sensory
experiences which are "modulated/evaluated" in a different "internal
environment" (the way a particular stimuli is experienced during orgasm
can be very different after or before it! - it changes sense processing
among others)
I wonder if we could one day develop implants or software which would let us
break down our feelings/qualia into simpler parts so we could better
understand what they're made of and what they're caused by. Greg Egan wrote
a short SF novelette about this idea:
http://www.infinityplus.co.uk/stories/tap.htm
Presumably there are special circuits in the brain that get activated,
which correlate to the flush of orgasmic pleasure. But what is special
about those circuits? From a 3p perspective, how is one brain circuit
differentiated from another? It can't be as simple as the
neurotransmitters involved; what would make one neurotransmitter be
causative of pain and another of pleasure? It's shape? That seems
absurd.
If we could represent our cognitive architecture abstractly and be able to
decode our memories, responses, expectations, ... Such questions might be
partially answerable. However, that would only be partially explainable -
you won't be able to explain the color red appearing in your visual field
except in relation to other colors and how they're recognized by your visual
cortex - the non-communicable parts of what "red" is will likely remain
forever incommunicable (except where images/data can be shared, but then how
can you know my red is exactly the same as your red, besides its relative
nature? I can't even know if my red doesn't change all the time and I'm
incapable of knowing. )
I don't have an issue with neutral "content of qualia" experiences per
se. I mean, for qualia like color I am completely OK with mapping that
to the correlative 3p mechanism involved with processing retinal
input, broken down by differing wavelengths. There's a leap there, for
sure, but it's one that is easy to make - one you must make assuming
comp.
I don't think only having retinal input of different wavelengths is
sufficient for color qualia, one has to take into account the entire
visual system and how it differentiates and recognizes patterns.
It seems that the consequence of that neural circuit firing would have
to achieve some kind of systemic effect that is characterized... how?
There would likely be many changes throughout the entire cognitive
architecture instance - changes in attention allocation, memories triggered
and so on.
Those are after-the-fact effects... or are you saying that those are
the kinds of mechanisms you could point to that "feel pleasurable"
from the inside?
Don't they happen almost simultaneously and thus also tint how certain
stimuli (be they external (senses) or internal(such as other emotions))
feel?
Pain is just as mysterious. It's not as simple as "what it feels like
for a system to become damaged". Phantom limbs, for example, are often
excruciatingly painful. Pain is clearly in the mind. What cognitive
mechanism could you characterize as feeling painful from the inside?
Watch the videos I linked, the author goes into great detail about such
emotions and how they emerge. Another interesting (and compatible) view is
that of the neocortex as a pattern predictor, you can find it in Hawkins'
"On Intelligence".
I think emotions are more complex than pain and pleasure. The real
question for me is what is the source of the "valence" of pain and
pleasure, and what accounts for it in mechanistic terms?
They are quite the opposite even behaviorally, aren't they? In internal
states and memory retention they also have opposite roles.
Failure to account for this in mechanistic terms, for me, is a direct
threat to the legitimacy of mechanism.
You can account for all the behavior, but it will be hard to account for
specific instances of qualia - those become indescribable past a certain
point, so much that even we can't say more about them except that we have
them and they are different and related in this and that way.
Again, I'm trying to keep this as simple as possible. For instance, do
you believe that there's a class of animals that do not have emotions
as we are discussing them, but still can perceive pain or pleasure?
Sure, pain and pleasure are likely one of the earliest to appear, likely
in the simplest nervous systems. A system needs basic drives/motivation,
otherwise it can't function correctly in an environment/has reduced
survivability. Even simple multicellular organisms should get it quite
early on.
This seems very likely to me that there is such a category of animal.
If so, how would you characterize it's cognitive architecture, as
distinct from the animals above it that can experience emotion, and
the animals below it, that can't feel pain/pleasure?
Pain - immediate actions, random or not, to specific dangerous stimuli.
Aversion/avoidance in more complex organisms (such as those capable of
expecting or predicting painful stimuli).
Pleasure - reduced or repeated same actions, to specific pelasurable
stimuli. Pleasure seeking behavior in more complex organisms (such as
those capable of expecting or predicting pleasurable stimuli).
Stimuli can be both internal (emotion) or external (senses).
Obviously for beings as complex as humans the nature of certain emotions
can be much more complex than that because they are mixed in with many
others, but I think that's what the simplest behavioral characterization
of pain/pleasure that I know of.
Terren
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to
everything-list+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
