A most sophisticated analysis of the song of a single bird was published by Keith Nelson in 1973 (1). If I am not mistaken, it has been surpassed only by the same author's study (1990) on singing structure in the European Song Thrush (2).
In the first-mentioned paper, Nelson breaks a lance for the use of non-linear models - as exemplified by coupled automata - in the study of complex behaviour of animals in general and song structure in particular. At the time, that was a revolutionary idea, which had been considered by very few ethologists indeed.
The relatively simple linear models, then commonly used, that Nelson referred to were Markovian and linear control systems. Considering the common occurrence of non-linear processes in living organisms, he wrote, such models can hardly be adequate. By applying them, one would fail to discover any more complex aspects of the object.
Why, then, did one use such models so often and with so much confidence? Nelson thought this was because scientists are deeply ingrained with rules of parsimony, in particular Ockham's Law, also called Ockham's Razor: Entia non sunt praeter necessitatem multiplicanda or in English: '[In explaining Nature] entities must not be multiplied beyond necessity'.
Ockham's Law is basic to science, but a too scrupulous adherence may have negative effects. Too often, according to Nelson, one begins with a simplest possible model, complicating it progressively as more and more peculiarities of the object come to be known, until the model becomes unwieldy and unnecessarily more complicated than a simple version of a more sophisticated model. This course of events has been called 'accommodation' by Piaget. "Simplex sigillum veri" (Simplicity is the stamp of the true), wrote Hermanus Boerhaave, but this is a misconception. Parsimony is a necessary instrument in the natural sciences, not because Nature should be simple, but in order to keep knowledge manageable and to help scientists communicate about it; and perhaps to gain control over matter. Hence: Simplex scientiae sigillum neque tamen veri (parsimony is a token of science, not of the true).
To find a proper name for the fact that the complex aspects of Nature tend to be overlooked if we study her with inadequately simple means, we may develop the metaphor of the Razor a little further. A common use of a razor is to keep a man's beard down to the surface of the chin. By doing so, we hide a number of small facts like the beard's detailed colours and shape; but above all we shall never discover what its role could be in the self-image and personality of its bearer. We are likely even to never ask this question!
I propose, then, as an homage to Keith Nelson's pioneering work in bird song and some other aspects of animal behaviour, to call this principle: "Nelson's Beard".
"Nelson's Beard" is not a negation of Ockham's Razor, but a necessary element in the balance that should be struck between it and the law of the medieval franciscan monk Willem van Ockham.
(1) K. Nelson 1973. Does the holistic study of behaviour have a future? In: P.P.G. Bateson & P.H. Klopfer (eds.) Perspectives in Ethology I. London, Plenum, pp. 281-328.
(2) K. Nelson 1990. Hierarchical Organisation, revisited. Neth. J. Zool. 40: 585-616.
The relatively simple linear models, then commonly used, that Nelson referred to were Markovian and linear control systems. Considering the common occurrence of non-linear processes in living organisms, he wrote, such models can hardly be adequate. By applying them, one would fail to discover any more complex aspects of the object.
Why, then, did one use such models so often and with so much confidence? Nelson thought this was because scientists are deeply ingrained with rules of parsimony, in particular Ockham's Law, also called Ockham's Razor: Entia non sunt praeter necessitatem multiplicanda or in English: '[In explaining Nature] entities must not be multiplied beyond necessity'.
Ockham's Law is basic to science, but a too scrupulous adherence may have negative effects. Too often, according to Nelson, one begins with a simplest possible model, complicating it progressively as more and more peculiarities of the object come to be known, until the model becomes unwieldy and unnecessarily more complicated than a simple version of a more sophisticated model. This course of events has been called 'accommodation' by Piaget. "Simplex sigillum veri" (Simplicity is the stamp of the true), wrote Hermanus Boerhaave, but this is a misconception. Parsimony is a necessary instrument in the natural sciences, not because Nature should be simple, but in order to keep knowledge manageable and to help scientists communicate about it; and perhaps to gain control over matter. Hence: Simplex scientiae sigillum neque tamen veri (parsimony is a token of science, not of the true).
To find a proper name for the fact that the complex aspects of Nature tend to be overlooked if we study her with inadequately simple means, we may develop the metaphor of the Razor a little further. A common use of a razor is to keep a man's beard down to the surface of the chin. By doing so, we hide a number of small facts like the beard's detailed colours and shape; but above all we shall never discover what its role could be in the self-image and personality of its bearer. We are likely even to never ask this question!
I propose, then, as an homage to Keith Nelson's pioneering work in bird song and some other aspects of animal behaviour, to call this principle: "Nelson's Beard".
"Nelson's Beard" is not a negation of Ockham's Razor, but a necessary element in the balance that should be struck between it and the law of the medieval franciscan monk Willem van Ockham.
(1) K. Nelson 1973. Does the holistic study of behaviour have a future? In: P.P.G. Bateson & P.H. Klopfer (eds.) Perspectives in Ethology I. London, Plenum, pp. 281-328.
(2) K. Nelson 1990. Hierarchical Organisation, revisited. Neth. J. Zool. 40: 585-616.
Geen opmerkingen:
Een reactie posten
Opmerking: Alleen leden van deze blog kunnen een reactie posten.