[Commentary on Dawkins and Singer by S.F. Walker]
Behavioral and Brain Sciences, (1990) 13, 43.


Natural and unnatural justice in animal care

Stephen F. Walker

Department of Psychology

Birkbeck College, University of London

London WC1E 7HX
(Current home page)


It is not necessary to concur with Singer’s position of drawing no line whatever between the moral status of humans and other species in order to welcome Dawkins’s carefully considered suggestions for devising behavioural measurements of preference, on the assumption that these could be helpful in minimizing the non-preferred emotional states experienced by captive animals. While I argued for certain kinds of continuity in the psychological processes of humans and other species in the book referred to by Dawkins, there remains ample evidence for significant discontinuities, which undercut the assumption that there is a simple dimension of suffering which can be put into cross-species utilitarian equations of the kind Singer appears to propose. For instance the traditional marker of language as a special determinant of human cognition has resisted all efforts to blur it (Premack, 1987, Chomsky, 1980). The equally venerable view that distinctive of aspects of human mentality are culturally determined (and/or “socially constructed”) is still vigorously promoted (e.g. Harre, 1986).

Therefore, while scientists should be (and are legally) obliged to accept the commonsense view that animals can suffer, Singer’s argument (precommentary, p.5) that we should consider human and non-human suffering as essentially similar to our own defends, as he points out, a much stronger position. It is worth noting that the commonsense view seems to recognize differences in the moral status of animals - many pet owners neuter their charges when they are young and vigorous, and “put them down” when they are old and feeble. Philosophical and empirical analyses may be able to support this distinction: it might be claimed that fear of being castrated, or (resentment at having been so treated) is both socially constructed and linguistically mediated; or following Dawkins it might be possible to show that fear of going to the vets’ (or of a standard signalling stimulus) is no greater for animals after a painless neutering experience that after a control condition involving no loss of natural function.

Dawkins’ article is both moderate and practical, and under “Problems with this approach” she seems to have considered most of the difficulties that could arise from too stringent an application of her main recommendation of assessing preference according to the principle of “inelastic demand”. In now emphasising some of the problems I do not wish to detract from the advantages of her behavioural approach.

A theoretical problem which may be minor in practice is that inelasticity may indicate automaticity of behaviour rather than high emotional value in some species, or in special cases. Suppose for instance that a cockroach (Horridge, 1962) or a decerebrate embryo (Heaton et al, 1981) showed inelasticity of response effort in avoiding aversive stimuli, or that a decorticate mammal showed less sensitivity than normal to response cost in food rewarded behaviour (e.g. Oakley, 1979). Should we assume greater suffering, or the operation of a more mechanical motivational system? A greater difficulty is that high elasticity of demand might conversely indicate more cognitive representations of goals or “declarative” emotional states (McFarland, 1989). Pigs might show elastic demand for social companions by comparison with food (p.11 target article) because of a relatively blind drive for food, and a more considered evaluation of the social benefits used in a particular experiment. For practical purposes this sort of potential problem is covered by Dawkins’ solution of using several measures of welfare (e.g. of the general condition of socially isolated though well fed pigs), but a greater distinction may need to be made between intensity of motivational effects on behaviour and the type of cognitive representation involved in a given motivational effect (Dickinson, 1985, Walker, 1987)

A more global reservation which applies both to parts of the target article and to the precommentary concerns the relation between suffering and natural behaviours. Dawkins’ conclusion is that it is an empirical question whether the absence of the opportunity to perform a natural activity such as migration leads to suffering, but she tends to assume that it is only the prevention of motivated acts, and not the natural performance of such acts, that could be associated with unpleasant emotional states. On the other hand Singer seems to suggest that we have a duty to reduce pain and suffering that may arise in animals’ natural conditions (p.5). There is surely something to be said for the point that natural life itself may involve high levels of stress - for instance during migration, hard winters for non- migrators, disease, drought and famine for all and predation for prey. Thus while supporting Dawkins’ idea that negative motivational states due to the prevention of natural behaviours in captive animals should be empirically assessed, and steps taken to minimize them, I am not convinced of the implied corollary that animals free to engage in natural behaviours always suffer less. Since the main issue is the care of captive animals this is not directly relevant, but it would have serious implications if one accepted Singer’s position that the welfare of wild animals (in the case, for example, of infant mortality rates) is as deserving of concern as is that of our own species.



Chomsky, N. (1980). Rules and representations. Behavioural and Brain Sciences, 3, 1-61.

Dickinson, A. (1985). Actions and habits: the development of behavioural autonomy. Proceedings of the Royal Society, B, 308, 67-78.

Harre, R., ed. (1986) The Social Construction of Emotion. Blackwells.

Heaton, M.B., Galleher, E.L., Baker, R.T., Otero, J.M. and Alvarez, I.M. (1981). Operant escape learning in decerebrate duck embryos. Journal of Comparative and Physiological Psychology, 95, 199-204.

Horridge, G.A. (1962). Learning of leg position by the ventral nerve cord in headless insects. Proceedings of the Royal Society, B, 157, 33-52.

MacFarland, D. (1989). Problems of Animal Behaviour. Longman Scientific and Technical (see Chapter 2, “Suffering in Animals”).

Oakley, D. (1979). Instrumental reversal learning and subsequent fixed ratio performance on simple and go/no-go schedules in neodecorticate rabbits. Physiological Psychology, 7, 29-42.

Premack, D. (1986). Gavagai! or the Future History of the Animal Language Controversy. MIT Press.

Walker, S.F. (1987). Animal Learning. Routledge.