How the chaffinch learned to sing
Eighth in a series beginning with Art in the family.
I find the conundrum from last time quite fascinating:
On the one hand an artistic canon is a flimsy construct, like the emperor’s new clothes – only there because people say it is there. On the other hand it is rock-solid. Western culture is a huge and irremovable component of the world we all inhabit, with a large part of why our world is the way it is being due to the ‘Western canon’…
For an object to be regarded as a good or great work of art it does not have to satisfy any objective criteria – other than perhaps minimal criteria which might also apply to things which are not ‘works of art’ at all (see eg Art in the family), let alone good or great works of art. The object only needs to be regarded as a good or great work of art – which is a bit circular.
I’d like to suggest a particular way of looking at this conundrum. To an extent it’s inspired by a book I’ve just finished reading: Darwin Machines and the Nature of Knowledge by Henry Plotkin. Plotkin doesn’t address art specifically, but he does present a broadly ‘Darwinian’ account of knowledge, which seems to map in a quite interesting way to the idea of artistic survival.
I won’t try to summarise his thesis, because I’ll get into a mess. Remember also that I’m reflecting what I think he’s saying – and I could well have misunderstood. It will involve quite a detour, but I hope I’ll eventually get back to where I want to be.
Plotkin’s idea, in a nutshell, is that knowledge = (biological) adaptation and (biological) adaptation = knowledge. So for example if we take what we commonly understand as human knowledge, he sees this as a type of biological adaptation, or a subset of biological adaptations. But if we take the whole domain of biological adaptations, these can all be described as ‘knowledge’ of one kind or another.
The trick, if you like, is that knowledge does not have to be conscious knowledge. It does not even have to be possessed by an entity capable of consciousness. Conscious knowledge, and therefore human knowledge, is just a special kind of knowledge. And not all biological adaptations are stored in the genes.
Some examples might help. The apparatus of photosynthesis in plants is a collective adaptation to the existence of light as an energy source. The plant’s internal structure ‘fits’ the environmental phenomenon of light in such a way that the plant can capture the light energy so as to survive and reproduce.
The plant does not ‘know’ (in the sense of human conscious knowledge) what it is doing as it photosynthesises or as it develops its photosynthetic equipment. But something is stored inside the plant’s genetic material which supplies the information to build an appropriate response to the existence of light – appropriate in the sense of using the light to improve the plant’s survival prospects.
Another example is the spines of a hedgehog, and the hedgehog’s related behaviour of rolling into a ball in the presence of a predator.
We could say the photosynthesising plant ‘knows’ about light, and the hedgehog ‘knows’ it has potential predators who will be discouraged it rolls into a spiny ball. But the inverted commas here do not exactly imply analogy or metaphor. Plotkin is not saying it is ‘as if’ the plant knows about light and what to do with it, or it is ‘as if’ the hedgehog knows about predators and how they can be repulsed. He is saying rather that if you treat knowledge as something which may be conscious but does not have to be conscious, which may be in a brain but does not have to be in a brain, then knowledge boils down to an adaptive ‘fit’ or correspondence between an internal state and an environmental – typically but not necessarily external – state or phenomenon.
The internal state may be caused by the environmental state, but not necessarily. There are theoretically two overall ways in which inherited adaptations could arise which correspond to environmental conditions. One, the ‘Lamarkian’ way, is by the environmental state of affairs causing a change in the organism which is then inherited by the organism’s offspring. The other is ‘Darwinian’, in which inheritable variations occur randomly and are then filtered for survival by natural selection. Environmental conditions do not cause specific adaptive variations in the first place, but they select the most adaptive – the ones providing the greatest competitive advantage. For reasons we don’t need to go into here, the ‘Lamarkian’ explanation has been largely discredited at the expense of the ‘Darwinian’ explanation, which for present purposes we can take as read.
Plotkin isn’t claiming anything new about evolution by natural selection in itself. His contribution is to draw the link between evolutionary adaptations and knowledge construed broadly. He then views the kind of (animal, and particularly human) knowledge we are more familiar with in terms of Darwinian selection.
The photosynthesising plant is constrained by its genetic inheritance. Its inner organisation is such as to exploit the existence of light. But if that energy source changed in a way that affected the ability or efficiency of the photosynthetic equipment to harness it, the plant’s response is limited by how it has evolved so far. Many higher plants display phototropism, in that they can change their growing patterns in order to grow towards the light. But if the sun suddenly changed the range of light wavelengths it emitted, or a person dug up the plant and shut it away in a dark room, it would have no survival strategies. The plant’s intricate photosynthetic equipment developed over long, evolutionary, timescales. Its limited mechanical responses (phototropism; autumn leaf fall; opening and closing of stomata) also evolved over those same evolutionary timescales, over which patterns of change (night and day; weather; seasons) were themselves largely unchanging.
With animals (which unlike plants, tend to move about), and in particular the ‘higher’ animals, we get the first inkling of a kind of knowledge – adaptive response – operating at a frequency faster than that of slow evolutionary change. A lot of behavioural response is still hard-wired – eg the migration of birds triggered by day-length. But in a variety of animals (including the primates, but not just the primates) real learning takes place. Plotkin gives the examples of songbirds like the European chaffinch who learn the song they are exposed to as they grow: a chaffinch born in Somerset but raised in Sussex will sing with a Sussex dialect (a Sussex chaffinch dialect that is!), and vice versa. The chaffinch is not born hard-wired with its eventual chaffinch song, but it is born hard-wired with the knowledge of what to learn.
Plotkin calls the evolution of adaptation (as chance variation filtered by natural selection and passed on by genes) the ‘primary heuristic’ – the mechanism by which apparent adaptive ‘design’ in nature gets ‘invented’. He then calls the operation of learning and intelligence in the higher animals the ‘secondary heuristic’.
There is not enough evidence to decide whether this secondary heuristic is a Darwinian (‘selectional’) mechanism, a Lamarckian (‘instructional’) mechanism, or something else entirely. One of Plotkin’s reasons for favouring the Darwinian explanation is that it accounts for creativity better than an instructional approach. Another is the parallel he draws with the development of the immune system within the individual, which does seem to proceed by Darwinian mechanisms. This is another form of (unconscious) knowledge – of how to fight off invading toxins.
If the secondary heuristic is an essentially Darwinian mechanism, it is one operating at a much higher frequency than genetic evolution operates. Also the secondary heuristic would be ‘nested inside’ the primary heuristic. For example the genetic endowment of humans (primary heuristic) determines that they will have intelligence and will learn – and how and what kinds of things they will learn – but the precise content of what an individual will learn (secondary heuristic) is open-ended.
The development of language in humans is obviously key here, and today’s post-Chomskyan consensus sees a similar ‘nested’ pattern: the structure and grammar of language are inherited, along with the child’s ability to develop (almost to ‘grow’) a language (primary heuristic); but the actual words and sentence-formation rules are supplied by the environment, which may be English-speaking, French-speaking, Xhosa-speaking – or indeed home to two or more languages (secondary heuristic).
Evolved design (or ‘design’) is rarely perfect. As far as I know there are no wheels in nature, which would have been a huge leap (roll?) forward had evolution invented them. The engineering compromises in the human body alone are legion. The same applies to animal behaviour and knowledge, including that of the higher animals and ourselves.
Plotkin makes use of a neologism coined by the American psychologist Herbert Simon as a cross between ‘satisfy’ and ‘suffice’: satisfice. Something satisfices when it does the job, when it is good enough even if it is not perfect.
A lot of knowledge and behaviour, both instinctive (‘behaviour without thought’) and learned, fall into the category of ‘satisficing’. Plotkin gives examples of the different nesting behaviours of kittiwakes and black-headed gulls; the human ability to remember and recognise faces; and the surprising gaffes which people commonly make with issues of logic and probability, depending on how the question is presented. People often struggle with questions of applied logic when they are presented abstractly, but when the same issues are presented in terms of social exchange and cheating people solve them much more easily. A classic example is the Wason selection task.
What does all this have to do with works of art and artistic canons? Hmm… I think I’ll leave that for next time!
© Chris Lawrence 2010.