The philosophy of non-human animal mind (henceforth “animal”) can be understood as a methodological reflection guided by the following metaphysical questions: (1) Is there an animal mind? If the answer to (1) is no, then we can stop our reflection. If the answer is yes, then a further question arises naturally: (2) What kind of mind do they have? Is their mind in some sense similar to or different from the human mind? In the current discussion the two competing models are the language-of-thought hypothesis and the iconic-thought hypothesis. A third question also addressed in this book is (3): Are animals endowed with the capacities of mindreading, metacognition, self-consciousness and emotion?

The contemporary consensus concerning the metaphysics of animal mind reflected in this book is that there are good reasons to answer (1) positively and accept that animals possess a mind. Philosophers do not believe anymore in the classic argument that having a mind requires possessing a public language, whose history goes from Aristotle to Donald Davidson. They consider that it only requires a representational capacity to represent distal objects and events and act according to such representations. Furthermore, philosophers have been forced to accept the evidence coming from ethological and experimental studies concerning the complexity of animal behavior and they also had to acknowledge that explanations in terms of release mechanisms and innate programs are no longer satisfactory to account for much of animal behavior.

The book begins with two reflections on question (1):  Is there an animal mind? The first essay of the book by Jamieson (chap.1) adopts the Humean view. It tries to solve the tension between (a) our tendency to attribute thoughts to animals and (b) our inability to characterize their thoughts. Its strategy is to strengthen Hume’s argument by analogy by means of the intentional systems theory. Given that an intentional stance permits us to predict and explain animal, as well as human behavior, in a reliable way and to fulfill our pragmatic purposes of mental attributions, then we should accept the positive answer to question (1), i.e. that animals do have a mind.

The second essay by Saidel (chap.2) also follows the Humean path, but with a realist and functionalist lens; in other words, it supposes that behavior can only be explained postulating the existence of discrete internal states that play causal and functional roles. Saidel argues that animal goal-directed behavior, such as the chimpanzees’ complex strategy to open a nut and eat the meat inside, can only be explained as being caused by distinct mental representations of the goal and the means to do it, since only beings capable of having distinct representations are capable of problem solving and learning from their experience. He cites a good number of data in order to provide support to this thesis.

After accepting the positive answer to (1), that animals are endowed with mind, the biggest divergence reflected in this book concerns (2) the characterization of the form or structure of animal mind. The main answers to this question are the language-of-thought hypothesis and the alternative iconic-thought view. Some articles in the book hold that animal mind works mainly by means of a kind of language of thought with a logical form (Tezlaff & Rey, chap.4; Carruthers, chap.5), whereas others argue that does is not, but rather constitutes an iconic or feature-based way of thinking (Rescola, chap.3; Camp, chap.6; Bermúdez, chap.8; Proust, chap.9). This hot discussion is introduced by Rescorla’s essay. He analyzes the classic theme of Chrysippus’ dog, which was supposedly able to execute logical reasoning, and the different accounts that aim to explain it. He claims that an account in terms of Bayesian calculations over cognitive maps (whose structure is neither logical nor sentential, but cartographic) provides a better explanation of animal behavior than the classic view that ascribes logic-deductive capacities to animals in order to perform deductive reasoning by means of a language of thought. However, even if Rescorla’s model in terms of cognitive maps and Bayesian calculation looks promising, it seems to me that he failed to provide decisive arguments to privilege his account over the language-of-thought hypothesis. At best, he has shown that his account is coherent, useful and may be developed further to explain other cognitive phenomena.

The following two chapters defend the language-of-thought hypothesis for explaining animal mind. The chapter by Tezlaff & Rey examines bees’ abilities to navigate and convey information. Their claim is that the only way to explain honeybees’ ability to transfer and use information about time, distance and direction of a place relative to another, as well as their ability to recognize locations and recall appropriate directions, is to attribute to them structured representations that display a variety of systematicities. The authors interpret these facts as evidence that honeybees are endowed with a language of thought, syntactically structured, over which they compute calculations that guide their behavior.

Peter Carruthers criticizes the strong version of the generality constraint, according to which the possession of genuine concepts (instead of proto-concepts) requires that a creature must be able to systematically combine all concepts with the other concepts they possess. He claims for a weaker generality constraint that genuine concepts must be merely recombinable at least with some others according to the right categories. He argues that it fits better our normal concept use and allows us to attribute genuine concepts to animals. Carruthers claims that the behavior of bees illustrates this flexible use of information since they are able to entertain mental representations with interchangeable contents that interact among them to direct their behavior. This analysis leads him not only to conclude that animals are real full-fledged thinkers, but also that what approximates human beings to the strong generality constrain is their capacity to entertain creative suppositions or faux-thoughts — the capacity to entertain thoughts without commitment to their truth. He concludes his paper with the suggestion that “distinctively human thinking consists of mere faux-thoughts” (p. 107).

Elisabeth Camp takes a critical point of view in this discussion. She examines Cheney and Seyfarth’s (2007) recent account of vocal communicative behavior of baboons in terms of a language-of-thought. They claim that baboons’ thought has a language-like structure because it exhibits some properties of language such as being representational, propositional, independent of sensory modalities, compositional, and hierarchically structured. However, Camp convincingly demonstrates that none of the properties cited by Cheney and Seyfarth is sufficient to establish that baboon cognition is distinctively language-like; these properties can also be exhibited by other representational systems such as maps and diagrams. Her analysis highlights the main difficulties of whatever theory that tries to determine the form or the structure of thought from its content, since the same content can be conveyed very well by different representational formats: “thought can be implemented in many formats, with different degrees of expressive generality” (p. 122). Likewise, the paper shows that representational formats other than language, such as maps and diagrams, have received little attention from philosophical reflection.

McAninch, Goodrich and Allen (chap.7) re-analyze the debate between the emotional and referential interpretations of the meaning of animal communicative vocalizations and try to show that neo-expressivism is the better account for the data on cognitive and emotive elements of animal calls. According to this third view, an utterance is understood to have both expressive and referential functions. So their account aims to unveil the systematicity and semantic structure of animal calls.

The next essays continue the discussion about the form of animal mind, but also start the consideration of question (3): Are animals capable of mindreading, metacognition and self-awareness and emotion?

Bermúdez (chap.8) gives an interesting account of how to understand mindreading capacities. He claims that there is a fundamental distinction between: minimal mindreading –when a creature’s behavior covaries with the psychological states of others — and substantive mindreading — when a creature’s behavior is caused by representations of others’ mental states and predictions based on such representations. Substantive mindreading is further divided into perceptual mindreading — the capacity to predict others’ behavior based on what they perceive and some generic, universal principles of behavior — and propositional attitude mindreading — the capacity to think and reason about others’ mental states as mental, taking into account their background psychological profiles. The difference between perceptual and propositional mindreading is also that the former can work implicitly, exploiting direct connections between perception and action, whereas the latter always requires an explicit representation of an agent’s psychological profile in order to reason about how it produces action. Finally, Bermúdez claims that all animal and many human social interactions can be explained in terms of perceptual mindreading without appeal to the propositional attitude one, and that the latter requires the possession of language.

Proust (chap.9) presents some experimental evidence on animal metacognition as their capacity to monitor and control cognitive activities such as remembering and perceiving. Then she proceeds to elaborate an attractive account of the representational format needed to allow this capacity in animals that lack mindreading in terms of non-propositional, feature-based representations that she calls “mental affordances”. This paper constitutes a remarkable attempt to develop a theory of metacognition and self-awareness that does not imply higher-order thought or metarepresentations in animals and in human beings.

Contrary to Proust’s non-propositional account, Gennaro (chap.10), DeGrazia (chap.11) and Roberts (chap.12) adopt a metarepresentational view. Gennaro claims that recent experimental data strongly suggest that we should acknowledge introspective self-awareness to animals. In other words, animals are not only aware of the world and their body by means of their attitudes (such as perceptions, sensations and feelings), but they are also introspectively aware of their attitudes; so they are aware that they perceive something, feel something or believe something. To adjust this with some other empirical data, the author is driven to claim that (i) there can be first-person metarepresentation of one’s attitudes even in the absence of the ability to metarepresent others’ attitudes (mindreading). And (ii) he defends the claim that there are degrees of mental concepts and self-concepts consisting in a partial understanding of what it is like to be in such a mental state, and therefore we should weaken the generality constraint for concept possession and accept that animals have at least minimal mental and self-concepts that allow them to metarepresent their attitudes.

DeGrazia proposes a distinction between three kinds of self-awareness: bodily self-awareness, social self-awareness and introspective self-awareness. He claims that some capacities apparently present in some animals — such as intentional action, emotions, anticipation, memory, imitation, self-recognition in mirrors, taking into account other’s special perspective, deception, complex social cognition and metacognition — demonstrate the presence of bodily self-awareness in most animals and also suggest the presence of the other two kinds of self-awareness in some species (monkeys, apes and cetaceans).

Roberts analyzes to what extent animals experience emotions according to his concern-based construal theory of emotion. From this point of view, emotions are perception-like states with a factual and evaluative character independent of thought. Animals have emotions insofar as they undergo fear, grief, anxiety and other emotions about something, but the scope of their emotions is limited since they lack language and thought, and especially because they are incapable of second-order attitudes.

The agreement in the last three papers on the metareprestational view of experiences seems too much optimistic face of the data they appeal to sustain this claim and the alternative accounts that may explain animal behavior. Not only it seems unlikely that animals have stricto sensu second-order thoughts about their attitudes (e.g., Bermúdez, chap.8; Carruthers 2008), but it seems that less demanding explanations are available (e.g., Proust, chap. 9). Animals may be aware of themselves and others in ways that deeply differ from ours and do not require second-order thought for it.

The last two papers by Sober (chap.13) and Fitzpatrick (chap.14) examine some epistemological concerns of the study of animal mind. They mainly focus on the principle of conservatism (also known as the principle of simplicity, parsimony or Morgan’s canon) and the justified or unjustified use of this principle by some scientists for ascribing mindreading to animals.

The essays contained in this book provide a notable state of the art of the philosophy of animal minds. They show some major agreement among theorists concerning several topics — e.g. the existence of an animal mind –, but the main merit of this volume consists in drawing attention to some core disagreements — e.g. about the form or structure of animal thought — inviting further work on those problems. Moreover, the book illustrates the fruitful interaction between different disciplines, methods and perspectives for the understanding of complex phenomena such as human or animal minds. However, an interesting question not directly addressed by any of these essays is whether it makes sense to speak of the animal mind instead of thinking that there may be several kinds of animal minds (see Dennett 1996).

 

References

Cheney, D. and Seyfarth, R. 2007. Baboon Metaphysics. Chicago: Chicago University Press.

Carruthers, P. 2008. “Metacognition in animals: A skeptical look”. Mind & Language, 23: 58–89.

Dennett, D. 1996. Kinds of Minds. New York: Basic books.

 

© 2010 Santiago Arango Muñoz

 

Santiago Arango Muñoz, Werner Reichardt Center for Integrative Neuroscience (CIN), Universität Tübingen, Germany

Jean-Nicod Institut, Paris (IJN), France

http://uni-tuebingen.academia.edu/SantiagoArangoMunoz