Published in

Dittrich WH (1998) Book review of Allen & Bekoff (1997) on Cognitive Ethology

Animal Behaviour 56 1305-1307

---------------------------------------------------------------

Species of Mind. The Philosophy and Biology of Cognitive Ethology. By Colin

Allen & Marc Bekoff. London: MIT Press (1997). Pp.xxi+207. Price 29.50.

 

The authors of this book, a philosopher and an ethologist, champion Darwin

in an attempt to strengthen the foundations of a discipline pioneered by

Griffin (1976), namely 'cognitive ethology'. Unsurprisingly, they take the

nuts-and- bolts view that animals have states of mind which are real like

any other bodily features and which have evolved like such organs as

kidneys or stomachs. In this view, animals' mental experiences are

intrinsically intentional and form the basis for their individual

experience of consciousness. In the authors' conception of mental

experience it is the 'aboutness' of mental images that makes

intentionality a necessary ingredient whereas they remain uncommitted

about consciousness.

 

In chapter 1 Allen & Bekoff attempt to argue that Darwin's thesis of

mental continuity can be assessed only by adopting a mentalistic

viewpoint. They claim that organisms have "mental lives" and the best way

to understand "mental attributions across species" is to use the framework

of cognitive ethology, which they define as "the comparative,

evolutionary, and ecological study of animal thought processes, beliefs,

rationality, information processing, and consciousness" (page ix). This

introductory chapter already makes clear that this book is more a

philosophy-of-mind book than a book about animal behaviour. Their first

subheading summarizes well the main aim of the book, to contribute to the

"naturalism about the mind" debate. This view rejects the Cartesian

insight that the mind is distinct from things around us but assures us

that it is part of the physical world. In this sense, mental states are

physically real entities and one would not be surprised if Allen & Bekoff

support neurobiological attempts to describe such mental states. However,

they consider an alternative approach to such a reductionist view: they

favour the approach to compare mental states of animals and they look for

credibility by adopting the Darwinian idea of mental continuity. They

explicitly link their approach with Griffin's cognitive ethology, shifting

the emphasis though. First, they are aware of the empirical deficiency as

well as primatocentrism of cognitive ethology, a trend they try to

overcome. Second, because of their view about real mental states they have

to establish intentionality as the focus of cognitive ethology replacing

Griffin's emphasis on consciousness. In Allen & Bekoff's approach, mainly

following Millikan's (1984) concept of behaviour and intentions,

intentionality is divorced from consciousness altogether.

Their claim that Darwin had already successfully used such an

anthropomorphic approach seems ill conceived for two reasons. (1) Although

Darwin was concerned with the study of animals' mental abilities and

stressed the continuity of mental abilities between animals and humans,

this area was just another arena for him to demonstrate the continuity of

physiological and behavioural functions in organic life, in particular in

the highly controversial field of 'the descent of man'. In this context of

proving the generality of the principles of natural selection one has also

to understand his often neglected work about 'the expresssions of

emotions', in which he brilliantly linked functional properties with

underlying structures shaped by evolutionary processes. Darwin's ongoing

contribution is to have demonstrated how the principles of evolution can

be used to understand organic phenomena which until then were merely

anecdotal curiosities. (2) Of course, his use of language was often

affected more by lyrical intuitions than scientific clarity and his

cultural background was that of the mid 19th century. Ghiselin (1969) has

no doubt that he was aware of these ambiguities when using metaphorical

language but Darwin's main aim was not to found a new discipline, ethology

or evolutionary psychology (but see his excellent work on earthworms), but

to apply evolutionary theory in fields other than taxonomy. Therefore, his

approach can be seen as very different from that of cognitive ethology and

his descriptions cannot be construed as contributing to "naturalizing the

mind by natural selection".

 

In chapter 2 the brief historical account of classical ethology and

cognitive ethology is rather disappointing, from taking literally Darwin's

anecdotal descriptions of behaviour, to a confusing interpretation of the

Clever Hans phenomenon (the issue was whether a horse was able to count

because of its clever conscious mind and not the nature of behavioural

studies as suggested in the book), to a highly selective characterization

of classical ethology (e.g. von Frisch's brilliant methodological

contributions are neglected). This chapter consists of disconnected

topics, whose tiresome confusions of internal states, mental states or

internal variables serve the place of argument. Statements such as "Hull

recognized only a small set of intervening variables, such as thirst and

habit...Hull also sometimes wrote as though the intervening variables were

merely convenient fictions, not be accorded any real status in the

psychology of the organism. Tolman recognized a much richer set of

internal processes" (page 28) seem to confuse important distinctions in

methodology between constructs, variables and inner states by simply

referring to the importance of animals' inner states as variables.

Postulating constructs and variables does not allow any conclusions about

mental states or, similarly, the reality of inner states to be drawn.

Allen & Bekoff resist this implication and claim in circular fashion

"cognitive ethology is also an important extension of ethology because it

explicitly licenses hypotheses about the internal states of animals" (page 37).

 

A brief but precise discussion, in chapter 3, of the very notion of

behaviour reminds the reader of the importance of conceptual

considerations even in merely descriptive studies on animal behaviour.

 

In chapter 4 one finds a casual discussion about the scientific

nature of cognitive ethology under the headings "cognitive awakenings",

"mental privacy" and "the behaviourist challenge". The initial discussion

seems narrowly informed by recent debates on the mind (e.g. on

higher-order intentionality, subjectivity, language) rather than

epistemological concepts (e.g. causality or scientific proof). In the end,

the authors' main argument seems that because one can distinguish between

stimulus-bound and stimulus-free behaviours, and behaviourists are unable

to explain stimulus-free behaviours, the latter are the natural domain for

cognitive ethology. Only cognitive ethology seems able to explain

stimulus-free behaviours by considering mental states as reason for such

behaviours. At this point the reader is left with the puzzling question:

what is, in principle, the difference between Allen & Bekoff's

stimulus-bound and stimulus-free approach to mental states and Lorenz's

(1950; internal drive model) and Lehrman's (1953; S-R model) approach to

motor behaviour as these approaches both rely on the distinction between

internal and external stimulation?

 

In chapter 5 the authors discuss the usefulness of folk psychology

and its smooth transition to an empirically based cognitive ethology.

Using a rather rhetorical style they challenge Stich's (1983) and

Dennett's (1991) attacks on the role of folk-psychological notions in

cognitive science. Stich and Dennett have argued that the contents of

mental states are necessarily underspecified and therefore, data based on

mental states would not qualify to feature in predictions and scientific

explanations of behaviour. Allen & Bekoff's main argument seems that this

issue cannot be solved conceptually, as Stich and Dennett argue, but that

it is an empirical question about the quality of data on mental states,

and is therefore unresolved. Do Allen & Bekoff believe that up to now

attempts to provide rich enough data from investigating mental states have

failed only because of the behaviourists' 'mindless' challenge?

 

In chapter 6 the whole dilemma of cognitive ethology seems nicely

captured in the opening paragraph on play in ants. Disappointingly, the

crucial question "How could Huber have seen or inferred pretense from the

behavior of the ants?" (Page 88) remains unanswered (not only for ants) if

one is looking for conceptually-based criteria supporting a mental-state

approach. After a brief, but informative discussion about the definition

of play the above question about play, pretense and intentionality is

applied to the social play of canids. At this point, I must confess a

bias: it has always been a delight to read Bekoff's studies on social play

which are stimulating, instructive and a rich source of information. One

must ask, however, whether the data really afford the kind of intentional

interpretations the authors suggest in this chapter. The topic of social

play is taken as a prime example to provide an empirical underpinning of

Millikan's (1984) concept of behaviour and intentionality. Millikan's

notion is strongly based on the historical role of behavioural traits and

finally the functional aspect of behaviour. Clearly, the design question

in natural selection and its usefulness is the underlying topic. For

example, the notion of the 'sign- stimulus' is replaced by Millikan's

notion of the 'intentional icon'. The main core of the book seems an

attempt to provide an empirical underpinning to Millikan's ideas in order

to fulfil the promise 'to naturalize the mind'. Certainly, providing a

conceptual foundation for studying animals' mental experiences would be an

important step for the development of cognitive ethology.

 

In chapter 7 a discussion of antipredatory behaviour is based for the

first time on animals' cognitive abilities. Unfortunately, the important

question in animal cognition, that is, whether animals use internal

representations is hopelessly confused with the crucial question in

cognitive ethology, that is, whether animals can make use of their own or

others' mental experiences. The author's confusing use of the term concept

may illustrate the problem. The view to attribute 'concept use' to

pigeons can be justified merely on heuristic principles. However, it is

completely different to postulate that pigeons can experience their mental

concepts and use them as stimuli, that is to subscribe to the ontological

position that the mind consists of real mental states as the authors do.

The notion of 'receiver psychology' (Guilford & Dawkins 1991) would

capture perfectly well the message of the chapter without relying on folk

psychology or mental states.

 

In chapter 8 the question is asked whether animal consciousness is

essential or dispensable for a future cognitive ethology. The reader

expecting an overview will find that, instead, the discussion, not

surprisingly from a theory-of-mind viewpoint, focuses on three problems:

qualia, sensation, and self-consciousness. Qualia refers to the

qualitative nature of the conscious experience, for example, the pleasure

experienced while bird watching. Allen & Bekoff now claim that we can also

say something sensible about the same birds' mental experience when being

watched even if we might not be able to share their experiences, rejecting

Wittgenstein's (1953) wisdom that even if a lion could speak, we would not

understand it.

 

In their final chapter the authors present cognitive ethology as an

eclectic discipline that is not restricted to field studies but takes the

challenge of laboratory scientists (=behaviourists) seriously. Thus, the

whole chapter is based on an earlier debate between the authors and Heyes

& Dickinson as proponents of the behaviourist challenge. Certainly, Heyes

& Dickinson (1990) present good arguments (e.g. reproducability of

evidence, causality) that proof of intentionality cannot be gained from

mere observational studies and that experimental manipulations are

required. However, Allen & Bekoff have a strong point when discussing

shortcomings (e.g. narrowness of intentionality concept, ecological

validity) of Heyes & Dickinson's neobehaviouristic modelling of animals'

'intentions', based on the belief criterion (instrumental belief of the

form that a specific behaviour R causes access to some desired object) and

the desire criterion (desire for access to object is causally relevant for

behaviour R) in typical Tolmanesque tradition. In this debate, the whole

dilemma of, on the one hand, Allen & Bekoff's approach and, on the other,

Heyes & Dickinson's behaviouristic approach comes to light and the reader

can see that it is neobehaviouristic terminology and their use of it that

can as easily lead to false assumptions in this field as can the use of

mentalistic terms.

 

Many of us believed that the debate about animals' mental states, at

least in the form revived in this book, had disappeared. The promise seems

that cognitive ethology is promoting new insights in the study of animal

behaviour by the adoption of mentalistic tools. The authors' make most of

the objections they consider into strawman arguments and they often use

rhetorical figures which can best be illustrated as follows. (1) All

animals are linked by continuity. (2) Humans are intentional animals. (3)

Non-human animals possess kinds of intentional minds. Would their

conclusion best be labelled as a typical post-hoc fallacy (since human

minds have appeared, they have done so because of the pre-existence of

minds in a continuous lineage of evolution) or a fallacy of composition

(since one can label one species of animal as intentional, one can

therefore find intentionality in the whole animal kingdom linked by

continuity)?

 

On the overall evidence, I would say rather that Allen & Bekoff are

obscuring present insight about cognitive processes in animals by

introducing conceptual confusions (e.g. empirical approach (see index),

stimulus-free versus stimulus-bound behaviour), adaptationist just-so

stories (chapters 6, 7) and mis- characterizations of evolutionary

explanations (e.g. pp. 72, 180). There are too many arguments by

assertion, too many unsubstantiated claims, and too many vague promises.

The book makes accessible the wide variety of problematic and

controversial issues in cognitive ethology, if read from a natural science

point of view, but is unevenly referenced and unreliably indexed. In the

index, the foremost popularizer of Darwinism, Richard Dawkins, is confused

with Marian Dawkins, a leading proponent of one variety of mental

continuity, namely 'emotional consciousness'. Despite the criticism, Allen

& Bekoff have done a service, however, by getting students of animal

behaviour to think about animal cognition in suitably complex terms, and

by offering what I think is a fine list of important issues in the study

of animal behaviour from Uexkull's (1920) 'Umwelt' point of view. They

remind us of the richness and multifariousness of animal behaviour and the

everlasting search for new questions in studying it. Yet has the book

increased our understanding of how animals behave? Regrettably, the answer

has to be 'No', or perhaps 'Not yet' for the theory-of-mind ethologists

telling their never-ending story.

 

WHD is supported by the Wellcome Trust.

 

Winand H Dittrich

Imperial College School of Medicine,

Division of Neuroscience at Charing Cross Hospital,

London W6 8RF, U.K.

 

References

Dennett, D.C. 1991. Consciousness explained. Boston: Little, Brown.

Ghiselin, M. 1969. The Triumph of the Darwinian Method. Berkeley: The

University of California Press.

Griffin, D.R. 1976. The question of animal awareness. New York: The

Rockefeller Press.

Guilford, T. & Dawkins, M.S. 1991. Receiver psychology and the evolution of

animal signals. Anim.Behav., 42, 1-14.

Heyes, C. & Dickinson, A. 1990. The intentionality of animal action. Mind Lang.,

5, 87-104.

Lehrman, D.S. 1953. A critique of Konrad Lorenz's theory of instinctive

behaviour. Quart.Rev.Biol., 28, 337-363.

Lorenz, K. 1950. The comparative method in studying innate behaviour patterns.

Symp.Soc.Exp.Biol., 4, 221-268.

Millikan, R.G. 1984. The language of thought and other biological categories.

Cambridge, Massachusetts: MIT Press.

Stich, S. 1983. From Folk Psychology to Cognitive Science. Cambridge,

Massachusetts: MIT Press.

Uexkull, J. 1920. Theoretische Biologie. Berlin: Paetel.

Wittgenstein, L. 1953. Philosophical Investigations. Oxford: Blackwell.