In this book, Lawrence Shapiro confronts an issue that is commonly acknowledged as one of the most significant metaphysical question in philosophy: the mind-body problem.

In an agreement with contemporary naturalistic approaches, he does not share the famous Cartesian dogma on mind that Gilbert Ryle labelled as the "Ghost in the Machine". Nevertheless, by testing two of the main approaches about the mind-brain relationship (the multiple realizability thesis and the separability thesis), he aims to show that the "ghost" dualist conception survives in these apparently physicalist conceptions.

The two hypotheses are treated as empirical and evaluated in terms of relative likelihood. In fact, Shapiro joins scientific works to philosophical analysis to tighten the tether between minds and bodies in order to test the degree of support the evidence provides for each hypothesis. Each hypothesis is then tested against a competitor. The multiple realizability thesis (MRT) is analysed against what he calls mental constraints thesis (MCT), which states that there are few ways to realize the mind. The separability thesis (ST) is tested against the embodied mind thesis (EMT), which states that bodies are more thoroughly integrated with minds than is typically acknowledged.

The basic idea of this book is that hypotheses on mind make predictions about what we can predict about the brain (the mind's realizer) and the body (the mind's container). Shapiro tries to show that, if MRT and ST are true, it is not possible to predict, from a description of the mind's capacities, what properties the mind's realizer and container must possess. But if MCT and EMT are true, as he thinks, there will be few ways to realize a humanlike mind.  

 

The first chapter is dedicated to multiple realizability and considers why philosophers have invested in this thesis, its scope, and conceptual and empirical arguments in its support. A general introduction to the discussion that will take place in other chapters is given.

MRT for minds is "the thesis that a mind with this uniquely human psychological profile can be built in distinct ways" (p.1). This is an obvious idea, according to Shapiro, for things other than minds (watches, corkscrews).

Shapiro describes the origin of such a thesis as ideological in a certain sense. First, some philosophers claim that MRT has profound implications for the relationship between biology and physics (Kitcher, Rosenberg). Second, MRT introduces the token-identity theory as a theory in the middle between extreme dualism and radical-type identity theory. Third, it is conceived as offering salvation from the possibility of intertheoretic reduction, thus allowing the autonomy of psychology (Fodor). But except for these ideological reasons, no theoretical reasons seem to arise from Shapiro's reflections.

Shapiro explores the possibilities of building a humanlike mind: the nomological or physical possibility (laws of nature), the logical possibility (logical truth) and the circumstantial or historical possibility (that is, the importance of initial conditions in predictions of what can happen). From Shapiro's analysis it emerges that the circumstantial or historical possibility is the most interesting amongst them.

He examines conceptual arguments for MRT, the Turing Machine Functionalism and the Functional Analysis Functionalism.

In the first case, he thinks that the popular analogy traced by Putnam between human minds and Turing machines is not an empirical hypothesis, as Putnam thinks. It is impossible to imagine evidence that might bear on this question. Moreover, this analogy suffers from at least two shortcomings: first, it might not to be testable within our lifetimes and maybe ever, and second, our experience with such systems suggests that it is false.

The emergent idea is that MRT arose first as an objection to strict materialism and radical identity theory. Nevertheless, firstly, functionalism is not incompatible with dualism in its own right. Secondly, Putnam's concept of functional isomorphism (correspondence between the states of one system and the states of the other that preserves functional relations) is an unlikely hypothesis because all systems functionally isomorphic to mind would be classified as minds.

The Functional Analysis Functionalism, offered by Fodor, is intended as much richer in a teleological sense: functions are contributions towards a goal that becomes apparent in the course of a functional analysis of a system. MRT is conceived in this case as a consequence: if various physical kinds can all exhibit the same characteristic activity, then mind as a goal-directed system is likely to be multiply realizable. Nevertheless Shapiro finds a gap. He shows that there are cases in which a functional description may apply only to a single kind of physical object. And, furthermore, that Fodorian Functionalism seems to explore only the logical possibility of MRT, which is the less interesting.

Among empirical arguments for MRT, Shapiro examines Putnam's empirical argument about pain, which induced him to refuse type-identity theory, and general empirical considerations proposed by Block and Fodor. Putnam's argument seems likely. It is obviously ambitious to affirm that pain has one and the same physical correlate in terrestrial and (conceivable) extraterrestrial brains. Nevertheless, even though it is a good argument against type-identity theory, it supports all kinds of functionalism. Moreover Shapiro thinks that the argument is not as strong. We don't know if brains of mammals, reptiles and so on are or not relevantly different. What we need is empirical investigation.

Shapiro is right in thinking that at present empirical considerations are not conclusive. We have a plastic structure of the brain and convergent evolution of traits in distinct species, but we do not share a definition of "realized" or a distinction between sameness and difference in realization. And we have to admit, following Shapiro, that in thirty years AI could not build a machine psychologically similar but physiologically different.

 

The second chapter examines philosophical ideas on realization by the means of an accurate philosophical analysis of concepts used in the debate. Shapiro proposes his own conception to distinguish how two structures can be conceived as the same or distinct realizations of a single kind.

An interesting distinction between realization and causation is proposed. Realization is a synchronic relation between the realized kind and its realizer, while causation is a diachronic relation. The realizer, he says, determines that which it realizes.

Two standard views of realization are examined and rejected because they cannot accommodate illustrations of realization that appear in the sciences.

What Shapiro suggests is the use of R-properties to decide whether two realizations of a kind ought to count as two different kinds of realization. R-properties is a label for those properties of realizations whose differences suffice to explain realizations. In his opinion, R-properties are those that make a causal contribution to relevant functional capacities, assuming a weak conception of function that he takes from Cummins (1975). The R-properties of a realization are those in virtue of which the realization is able to achieve that capacity which makes it realize the functional kind that it does.

What Shapiro tries to show in this chapter is that, if this idea of R-properties is reasonable, then the empirical evidence for MRT (as neural plasticity) are not decisive.

 

Chapter three deals with constraints on the mind, nervous system and body. The importance of a discussion on constraints is in how they impose themselves on a system and limit the variation that the system can exhibit. The thesis explored in this chapter is the thesis that competes with MRT, MCT, namely the possibility to infer facts about nervous systems from facts about psychology. By the use of the mathematical construct of morphospace, Shapiro remarks that MRT and MCT (the mental constraint thesis, according to which minds are not multiply realizable) must be intended not as precise doctrines, but as different ends on a continuum.

The main question is what kind of structures are needed for the performance of a function. The presence of constraints, in fact, allows predictions from structure to function.

The constraints Shapiro suggests as common properties of humanlike mind are historical constraints. The distinction between universal and historical constraints corresponds to the distinction traced in the first chapter between nomological and circumstantial possibility. That is, the distinction between laws of nature versus features acquired by evolution. Again here the main focus is the plausibility of terrestrial MRT and therefore Shapiro is interested in constraints that explain what must have happened for things to be as they are now (even if in some occasions some universal constraints of humanlike minds are suggested). As examples of how universal and historical constraints impose a limit on the number of structures that can possibly produce some function Shapiro proposes mammalian homeostasis and image-forming eyes. At the end, he shows how some traits in various species are convergent, that is, the fact that selection has caused unrelated species to converge on a similar solution to some problem. Convergence makes the multiple realizability thesis less likely than alternatives.

 

The fourth chapter musters evidence from neuroscience and brain evolution in order to compare the

likelihoods of the multiple realizability thesis and the mental constraint thesis. In this chapter, Shapiro considers various constraints on the brain design. The presence of brain constraints seems to support the claim for MCT. Again, he uses evolutionary convergence as a source of evidence. He discusses sensory systems, sensory topographical organisation, brain wiring, brain components, myelin and trends in brain evolution. He tries to show that constraints, against MRT, allows predictions about the properties of the organs that realizes a humanlike mind.

 

The fifth chapter challenges the possibility of multiple realization for the prospect of intertheoretic reduction and tries to show that the issues of autonomy of mental is not so strictly linked to realizability thesis as many philosophers claim. He presents Fodor's argument against the tenability of both bridge laws and laws from which higher-level laws can be derived. The argument, according to Shapiro, displays an exegetical misunderstanding of positions that propose reduction. Shapiro thinks that unitary science is a more attractive prospect and that reduction does not necessarily conflict with autonomy of the special sciences.

 

In the sixth chapter, Shapiro considers the separability thesis via two trends in cognitive science: body neutrality and envatment. ST is the thesis according to which "from knowledge of mental properties it is impossible to predict properties of the body" (p. 167). He shows how evidence about body neutrality and envatment might present a difficulty for ST and support for EMT. He then concludes that investigation of the mind would reveal much about the body.

 

In the seventh chapter, he illustrates the various forms of embodiment and its three main claims (embodied thought, embodied conceptualization and extended mind). What he aims to show is that the body is profoundly involved in mental operations. This claim concerns not only how the mind does what it does (how it perceives, attends, recognizes, reasons, and so on), but also the content of what is thought. To explain how the body influences what we think he proposes a broad section on three branches of research. He discusses the Turing test, metaphor and language. In the end, he suggests that mind conception is extended beyond brains and body. Therefore, he seems to say that, if it is possible to predict properties of the brain and mind from the knowledge of mental properties, it is also possible to predict properties of the world.

 

In the last short chapter the author offers conclusions and suggestions. He concludes that the questions he has been considering, which remain open, might be answered by further empirical research.

 

To conclude, The Mind Incarnate is a very complex and technical philosophical text. It explores in detail metaphysical and epistemological issues from conceptual and empirical perspectives, by the means of a careful philosophical discussion. This heavily theoretical book is surely not for a general public, but it is directed to scientists and philosophers. It is perfectly in line with the contemporary popular approach of embodiment in cognition and it is directed to thinkers who accept the challenge of reconsidering pervasive assumptions about mind as naturalized descendants of the Cartesian dogma of the "Ghost in the Machine".

Each chapter is significantly rich of philosophical arguments, references, and makes much use of empirical data. Nevertheless, although the book is rich in its explanations, it seems obscure in some passages. Moreover, although it uses a lot of empirical data, the discussion often appears more evocative than convincing. Firstly, the link between MRT and ST is not clear. He claims the former entails the latter, but nowhere offers any clear argument for this. Secondly, Shapiro seems to treat ST as a metaphysical thesis, while it is only a claim on how the mind should be described. The claim that the properties of the mind can be investigated in isolation from those of the body does not coincide with the claim that the mind is like the occupant of a house. That is, the disembodied description of mind does not entail the disembodied conception of mind. The former is an epistemological claim, while the latter is a metaphysical one.

However, even though this book is not conclusive and seems to have these faults, it is a critical contribution to the contemporary debate on mind-body problem. It scours a variety of fields and its strength is in exploring the question as an empirical matter. What is surely remarkable is Shapiro's effort to link body and mind in a much closer way than classical cognitive science has done so far.

 

© 2007 Elisabetta Sirgiovanni

Elisabetta Sirgiovanni, Ph.D. in Cognitive Science, University of Siena, Siena, Italy.

Categories: Philosophical