Robert W. Lurz (Ed.) The Philosophy of Animal Minds. Cambridge (CUP), 2009, 308pp.
The Philosophy of Animal Minds is not an introduction to the expanding field of cognitive ethology and the study of animal cognition in the cognitive sciences, but a collection of fourteen ‘state of the art’ papers on several aspects of animal cognition. The collection thus presupposes knowledge of recent discussions and (experimental) results. For those familiar with the basic topics and threads of controversy The Philosophy of Animal Minds expands their understanding of the field and opens some new horizons.
The Philosophy of Animal Minds is also a ‘state of the art’ collection in the sense that explicitly or implicitly methodological issues in the debate on animal cognition appear centre stage. It has become crucial to state the framework of some problem or issue before entering in the discussion of ethological observations. With respect to the raised standard of methodological reflection one may well consider the study of animal cognition a test case for the methodology of the cognitive sciences in general, as some authors explicitly note, for instance: ‘studying animal metacognition should provide new perspectives on the structure of mental content and on mental activity in general’ (166).
Theories of intermediate faculties have become a focus of discussion, since often it seems unwarranted to ascribe (full blown) human cognitive faculties to animals. Focussing on intermediate faculties helps excluding overstated claims (like Camp’s paper denying that even Baboons have a ‘language of thought’ in the full sense) without excluding some fairly elaborated cognitive faculties in animals which resemble the faculties we know from the human case. One way to do this is to weaken one or more of the defining characteristics of some faculty. Carruther’s paper proceeds this way with respect to sign usage meeting a ‘weak version’ of the ‘generality constraint’ (that one commands a sign only if one can apply it in all the grammatically possible sentences of one’s language), so that he can claim that even invertebrate sign usage exhibits some generality. Bermudez distinguishes several forms of metacognition, so that one form of mindreading might be present in apes.
Following this strategy requires developing improved theories of intermediate faculties. Rescorla presents a more complex theory of cognitive maps, as maps may be an alternative representational system for animals not using language or not having sentence like representations. Allen et al. take up the discussion on propositional attitudes and argue for two-faced attitudes (i.e. attitudes that combine an expression of  volition with a claim of adequately representing present descriptive/cognitive content – as popularized in Milikan’spushmipulliyou’-theory of animal attitudes), and apply it to monkey alarm calls. Roberts argues that emotions in animals have characteristics of such two-faced attitudes.
To place some species of animal on a scale of (intermediate) faculties it is also crucial to distinguish an occurrence (or anecdotic attribution) of a representation and the occurrence of that representation as an expression of a more general faculty. Several authors are sensitive to this distinction, which prevents ascribing full blown general capacities where domain specific representations are present in animals, for instance ‘that baboons don’t manifest an ability to think hierarchically structured thoughts in any domain except dominance’ (124), or with respect to mindreading: ‘Social interactions that are sufficiently stereotypical to be modelled in terms of frames can proceed without propositional mindreading’ (154). Denying the domain specific abilities would be an error as well as over generous attribution of human abilities. Theories of domain specific abilities may help in seeing an evolutionary road to human general abilities. This is more promising than weakening the criteria for some human capacity (like mindreading) and then ascribing general metacognition or even mindreading to animals (like in Proust’s paper); this only invites the problem of explaining why such supposedly present faculties, say of higher order thought, are so seldom employed as witnessed by the absence of conventional language with these animals.
The methodology might be called a ‘divide and conquer’-strategy. We understand animal cognition better by dividing the more general human capacities into domain specific capacities in animals.
As cognitive capacities involve concepts theories of concepts or – given the methodology just outlined – theories of proto-concepts form the main building block of theories of animal cognition. These theories then have to be connected to theories of representational systems or theories of cognitive coding.
A second approach related to the study of intermediate faculties is also present in The Philosophy of Animal Minds, namely: separating several faculties which are (traditionally) grouped together under one heading, but which might and thus should be kept apart. If several capacities are grouped under one ambiguous heading it may turn out that animals have some full blown capacities, but not all the full blown capacities related to that heading. DeGrazia and Gennaro in their papers try to separate several faculties from each others all of which are related to self-awareness. Given this separation they claim that at least some animals are capable of ‘I’-thoughts, just as human reasoners have a representation of oneself being the reasoning agent or possess some form of self-awareness.
A broader methodological issue often present in the discussion of animal cognition is dealt with in the last two chapters by Sober and Fitzpatrick: parsimony or simplicity of explanation and description. Often a claim on animal cognition is backed up by claiming it to be more parsimonious or it being simpler than rival theories. Interestingly such claims have been made both by authors attributing and by authors denying cognitive capacities to animals – as one observes in reading the other papers in the collection! Sober expands his favourite theory of simplicity, whereas Fitzpatrick assents to a common scepticism of simplicity, simplicity being rather a side effect of other theoretical virtues (like simplicity being a form of explanatory power or simplicity arguments being just ‘poverty of the stimulus’-arguments).
The editor rightly asserts that meanwhile ‘the philosophy of animal minds is a field in its own right’ (1). And the collection The Philosophy of Animal Minds adds new insights to this field.
Manuel Bremer, University of  Düsseldorf, Germany