Thursday, April 19, 2007

Feelings and fitness

I have been reading Sean Carroll’s book “Endless forms, Most beautiful”, which tracks recent progress in the field of evolutionary developmental biology. One of the themes of the book is the pivotal role of 'genetic switches’ within the genome, which determine when and where in the course of development of the embryo genes will be expressed. The activation of these switches is itself determined by proteins which themselves had been translated as a result of the action of other switches and so forth. Carrol writes:



“ Ultimately, the beginning of spatial information in the embryo often traces back to asymmetrically distributed molecules deposited in the egg during its production in the ovary that initiate the formation of the two main axes of the embryo”. (p.116)

The role of switches in the coordination and choreography of the unfolding of developing organisms from single fertilised eggs to complex, multicellular and multifunctional creatures struck me as an example of what an intricate, amazing process evolution is.

Which brings me to the topic of natural selection and the evolution of consciousness.

Graham Cairns-Smith in his excellent book “Evolving the Mind : On the Nature of Matter and the Origin of Consciousness” argues that consciousness ( which he roughly equates with subjective feelings) would not have evolved if it did not have adaptive physical effects:

“..if it was natural selection that was the engineer then there must be consequences to feelings; they must have actual physical effects or there is no reason to expect that the means to produce them would have evolved.” (p.200)

An epiphenomenalist wishing to counter this argument could argue that consciousness evolved as a by-product of other brain processes. However, Cairns Smith responds that this does not help explain that feelings are generally appropriately “nice or nasty” in ways that enhance survival.

This argument seems to me to be a strong one for the casual efficacy of consciousness . I find it inconceivable that the fantastically sophisticated but blind machinery of random mutation, natural selection, genetics and embryology would produce phenomena which appear to be adaptive but which in reality do not do anything.

Note that this position does not imply that all feelings must be adaptive, only that the overall reproductive fitness of organisms with feelings would be greater than those without. Whilst there are undoubtedly many feelings which may not be adaptive (such as drug cravings, suicidal thoughts, destructive impulses, depression and so forth), on the whole feelings such as hunger, pain, pleasure and sexual attraction do, on the face of it, appear to enhance the chances of an organism surviving and producing off spring.

There is one possibility which could cohere with consciousness being epiphenomenal which Cairns-Smith briefly considers. This is the possibility that experientiality is concomitant to most or all physical processes, rather than a phenomena produced through natural selection. But Cairns-Smith argues that this option would still leave problems in relation to the fitness of feelings:

“..this would not explain why feelings are appropriately produced. It would not explain why pleasures on the whole go with acts that promote our survival and pains with the opposite.” (p.201)

However, a possible response to this view is that feelings are appropriately “nice or nasty” as a result of the development of fundamental principles operative in nature. So, for example, one could speculate that at some basic level gravitational, electromagnetic and nuclear forces are accompanied by primitive feelings or “protofeelings” (perhaps with attractive or repulsive forces being associated with incipient pleasures or pains).

To speculate further, these basic protofeelings could then have developed in organisms as a by-product of the increasing complexification of organised matter. So, for example, the movement of a bacterium toward a nutrient may be accompanied by some basic feeling of attraction. Moving up the evolutionary scale, when organisms with brains are in a position to create representations of the body and external world, a diversity of feelings would accompany such representations. This would occur not through any selective advantage, but through the incidental development of feelings which are fundamental in nature.

This option no doubt has problems of it own (with the foremost being how feelings at the elemental level interact with or relate to feelings at higher levels such as that of human consciousness). I imagine that many would find it even more unpalatable than the view that consciousness is an emergent phenomena that is causally efficacious.

Nevertheless, it seems to me that it if one accepts that feelings are “real” and a product of natural processes then the options are that either:

- Feelings are an evolved adaptation which generally enhance reproductive fitness and therefore have physical effects on the behaviour of the organism.
- Feelings are fundamental but epiphenomenal with no physical effects. They are appropriately produced in organisms because of the incidental development of their fundamental features, such as their attractive or repulsive character.
- Feelings are fundamental but also have physical effects, such that feelings are fundamental to at least some material processes, with natural selection ensuring they are appropriately produced in conscious organisms because they have survival value.

free web page counters

3 comments:

Anonymous said...

I'd say it's important to remember that evolution can break certain "meta frames". Notably, if some variation "works better", it doesn't matter much whether the change is purely biochemical, developmental, an "accidental" interaction with the environment, or even an epiphenomenon of some other process. All that matters is that it's heritable in some fashion.

Now consider this account of how "emotions" can make sense even in reductionistic terms....

1) "Appropriate response to the environment" is selectable on every level. That starts with primary senses and reflex responses.

2) Modal behavior is another early advance -- initially recognizing whether current circumstances are "hostile", "crowded", "feast", etc. Anything that makes it more accurate works better, and associative memory is an early contender.

3) Both senses and the processing thereof continue to advance -- but so does the internal state-machine representing the modalities, and that association network. The "reptile brain" is a bit of a cliche, but that limbic system is still the core of our own minds.

4) At some point, the two advances merge to produce modeling of the external world. This allows not only planning and problem-solving, but also complex social interaction. Social identities start to emerge.

5) The social complexities both encourage and support better communications and signalling. When this feeds back into the world-modelling capacities, it allows for abstract thought and reflective self-awareness.

Justin said...

I think that's a very good account of how higher order information processing and cognitive functioning might have evolved.

However, I meant "epiphenomneal" in the sense of not having any causal influence on the body or behaviour.

The argument of Cairns-Smith is that if subjective consciousness had no effects on behaviour then there would be no evolutionary reason why, generally speaking, subjective feelings are appropriately associated with behaviours in ways which enhance survival (eg destructive pains "feel bad", sex feels good, the feeling of hunger is unpleasant until satiated etc) .

Anonymous said...

what is yoga