display | more...

There are often a number of different levels of description, or explanation, at which we can look at any given problem. In many cases each level of explanation makes a lot more sense when you are aware of one or more of the more fundamental ways of describing a phenomenon. Hence the basic laws of chemistry follow very naturally from a physics-based description of atoms, molecules and so on; much of biology becomes comprehensible only when you start to understand the chemistry that it is based on; psychology starts to make more sense once you understand a bit about the biology of the brain; theories of economics that take into account realistic ideas about human psychology have started supplanting simpler theories that abstract people as rational self-interested actors, and so on.

One perspective on the relationship between different levels of explanation is reductionism - the idea that chemistry is really just applied physics, biology is just applied chemistry, and so on. It's a very powerful approach which has given scientists a great many fresh insights over the centuries.

In practice, however, each new layer of description tends to include approaches and concepts which don't seem to follow from the previous layer in any obvious way. It is true for example that the whole periodic table follows very naturally from quantum physics, but if you insisted on always looking at things from a strictly physics-based perspective, you would probably miss an awful lot of interesting chemistry, and make a great deal of extra work for yourself in the process. Part of the reason for this is emergence - the tendency for behaviour to manifest as a result of known laws which may well be deterministic in principle, but from which, in practice, nobody would have predicted that behaviour without exceedingly careful consideration.

Having a description of a phenomenon in terms of the more basic things that give rise to it is satisfying, and sometimes has great practical implications, but there is also a danger of being led astray by applying what is supposed to be a more fundamental level of description to things which are actually better explained at a higher level. It might be technically accurate, for instance, to say that a violent incident occurred thanks to an excess of adrenaline and cortisol in the assailant's brain, but it would probably be a lot more helpful to say that the fight occurred because the other guy was going out of his way to wind him up. Interestingly, it has been shown that people tend to rate an explanation of some observed behaviour as a great deal more convincing if it is accompanied by a bit of neuroscience, even if the neuroscience in question is completely irrelevant (link).

This difficulty in choosing the best level of explanation to work on has extremely important practical consequences, largely because your perspective on a problem constrains the sorts of strategies you are likely to consider for dealing with it. If you view depression as a 'chemical imbalance', the obvious thing to do is to try re-balancing the chemicals in someone's head. If you take a neurophysiological perspective, you might be more interested in what structures and systems within the brain might be leading to undesirable levels of those neurotransmitters occurring in the patient's brain on an ongoing basis. If you see it through the lens of psychology, you are likely to want to try looking for a psychological fix, perhaps trying to see if there are root causes in their life events or the attitudes they take to them that might be made better. A sociological perspective would look at the social relations between the person and those around them - perhaps they are unhappy because modern society lacks the structures that would fulfil their needs.

The different levels of description feed into each other, sometimes in very subtle ways, and it is not always easy or even desirable to pick them apart. Neither does the applicability of one level of explanation imply that solutions based on quite different perspectives will not also be helpful. Attacking symptoms can be very valuable, after all, and while aiming to fix things at root might be the best long-term solution in principle, it only works if you can both correctly identify and do something about the basic problem, so it is not always clear where our efforts are best spent.

The sheer quantity of learning available to modern science makes it very difficult for researchers to get a handle on every possible angle on a problem - there is an inevitable tension between specialisation and interdisciplinarity, closely related to the tension between reductionism and holism. We might like to see all of nature in terms of an overarching scheme or pattern of organisation, and historically many scientists and philosophers have been driven - sometimes fruitfully, sometimes misleadingly - by the urge to reveal such a scheme. However, while there are often lessons to learn and principles to apply when shifting between perspectives on the world, it now seems very unlikely that any detailed 'theory of everything' will manage to be a theory of music and economics as well as subatomic particles and things moving through space.

The availability of different levels of description has a tendency to lead people astray when it is not obvious how to marry up the different levels, which has notoriously been the case when it comes to the mind/body problem. We can describe the brain (at least somewhat) and we can describe the mind (as we experience it), and it's clear that they're connected in some way, but the specifics of how one might give rise to the other are really not at all obvious, leading some commentators to insist that they must be fundamentally different sorts of thing. In philosophical terms, this is confusing ontology with epistemology - drawing conclusions about what actually is from what we seem to be able to know. It's a seductive trap, that people fall into in all sorts of contexts - 'I can't imagine how this could possibly happen, therefore it can't possibly happen.' Science has repeatedly shown that the unimaginable can become plausible - and ultimately unavoidable - when new ways of looking at things present themselves. I believe that cognitive scientists and others are steadily closing this particular explanatory gap, but navigating the different levels of explanation available for mental/neurological phenomena will be a problem of practical as well as philosophical interest for many years to come.

Further reading: The best discussion of this that I've come across is in Douglas Hofstadter's 'I am a Strange Loop', mainly in chapter 2, although I hadn't seen that when I wrote this. Another few months later, I came across this paper by Uri Wilensky and Mitchel Resnick on the topic.

Log in or register to write something here or to contact authors.