I critically evaluate the influential new wave account of theory reduction in science developed by Paul Churchland and Clifford Hooker. First, I cast doubt on claims that the new wave account enjoys a number of theoretical virtues over its competitors, such as the ability to represent how false theories are reduced by true theories. Second, I argue that the genuinely novel claim that a corrected theory must be specified entirely by terms from the basic reducing theory is in fact too (...) restrictive for scientific practice and should be rejected. Basic theories co-evolve with nonbasic theories in a mutually interactive way, and thus the basic theories incorporate the concepts and concerns of nonbasic theories. Third, I show that once its ontological consequences are duly noted, the reductive part the new wave account collapses into the classical theory developed within the logical empiricist tradition. As such, it still falls prey to standard anti-reductionist argument based upon multiple realizability and the cross-classification of special science and physical science terms. (shrink)

I resolve an argument over “flat” versus “dimensioned” theories of realization. The theories concern, in part, whether realized and realizing properties are instantiated by the same individual (the flat theory) or different individuals in a part-whole relationship (the dimensioned theory). Carl Gillett has argued that the two views conflict, and that flat theories should be rejected on grounds that they fail to capture scientific cases involving a dimensioned relation between individuals and their constituent parts. I argue on the contrary that (...) the two types of theory complement one another, even on the same range of scientific cases. I illustrate the point with two popular functionalist versions of flat and dimensioned positions – causal-role functionalism and a functional analysis by decomposition – that combine into a larger picture I call “comprehensive functional realization.” I also respond to some possible objections to this synthesis of functionalist views. (shrink)

Thomas Polger and Laurence Shapiro argue that Carl Gillett's much publicized dimensioned theory of realization is incoherent, being subject to a reductio. Their argument turns on the fact that Gillett's definition of realization makes property instances the exclusive relata of the realization relation, while his belief in multiple realization implies its denial, namely, that properties are the relata of the realization relation on occasions of multiple realization. Others like Sydney Shoemaker have also expressed their view of realization in terms of (...) property instances, yet they too have accepted the multiple realizability of properties. Thus I am interested in the more general issue raised by Polger and Shapiro's argument. Specifically, I show how to supplement a theory of realization with a category-inclusive auxiliary assumption, which avoids the stated reductio. I then offer a few reasons to justify the proposed category-inclusive view of realization, making some comparisons to supervenience and causation along the way. (shrink)

Philosophers almost universally believe that concepts of supervenience fail to satisfy the standards for physicalism because they offer mere property correlations that are left unexplained. They are thus compatible with non-physicalist accounts of those relations. Moreover, many philosophers not only prefer some kind of functional-role theory as a physically acceptable account of mind-body and other inter-level relations, but they use it as a form of “superdupervenience” to explain supervenience in a physically acceptable way. But I reject a central part of (...) this common narrative. I argue that functional-role theories fail by the same standards for physicalism because they merely state without explaining how a physical property plays or occupies a functional role. They are thus compatible with non-physicalist accounts of that role-occupying relation. I also argue that one cannot redeploy functional-role theory at a deeper level to explain role occupation, specifically by iterating the role-occupant scheme. Instead, one must use part-whole structural and mechanistic explanations that differ from functional-role theory in important ways. These explanations represent a form of “superduperfunctionalism” that stand to functional-role theory as concepts of superdupervenience stand to concepts of supervenience. (shrink)

I critically evaluate Bickle’s version of scientific theory reduction. I press three main points. First, a small point, Bickle modifies the new wave account of reduction developed by Paul Churchland and Clifford Hooker by treating theories as set-theoretic structures. But that structuralist gloss seems to lose what was distinctive about the Churchland-Hooker account, namely, that a corrected theory must be specified entirely by terms and concepts drawn from the basic reducing theory. Set-theoretic structures are not terms or concepts but the (...) structures that they describe. Second, and more serious, a familiar problem for classical positivist account of reduction resurfaces within this newest wave of thinking, namely, commitment to property identities and inter-theoretic bridge laws (a problem I discussed at more length in "Collapse of the New Wave"). Indeed, this problem is exacerbated by Bickle’s conciliatory treatment of property plasticity, since he is willing to grant that a large number of special science terms denote multiply realized properties, at least if realistically construed. Still, in the end, Bickle sidesteps the reduction of properties by appealing to Hooker’s "function-to-structure token reduction." This is an interesting move with an intriguing concept of reduction. But problems remain. For, third, Bickle and Hooker's function-to-structure token reduction is actually a guised form of eliminative materialism. But that should be unacceptable since the position extends well beyond any modest revisionism for suspect items from a folk theory, say, in folk psychology or folk biology. Instead, it applies to functional terms and concepts employed throughout well-developed and explanatorily successful sciences. (shrink)

Multiple realizability has been at the heart of debates about whether the mind reduces to the brain, or whether the items of a special science reduce to the items of a physical science. I analyze the two central notions implied by the concept of multiple realizability: "multiplicity," otherwise known as property variability, and "realizability." Beginning with the latter, I distinguish three broad conceptual traditions. The Mathematical Tradition equates realization with a form of mapping between objects. Generally speaking, x realizes (or (...) is the realization of) y because elements of y map onto elements of x. The Logico-Semantic Tradition translates realization into a kind of intentional or semantic notion. Generally speaking, x realizes (or is the realization of) a term or concept y because x can be interpreted to meet the conditions for satisfying y. The Metaphysical Tradition views realization as a species of determination between objects. Generally speaking, x realizes (or is the realization of) y because x brings about or determines y. I then turn to the subject of property variability and define it in a formal way. I then conclude by discussing some debates over property identity and scientific theory reduction where the resulting notion of multiple realizability has played a central role, for example, whether the nonreductive consequences of multiple realizability can be circumvented by scientific theories framed in terms of narrow domain-specific properties, or wide disjunctive properties. (shrink)

In light of the phenomenon of multiple realizability, many philosophers wanted to preserve the mind-brain identity theory by resorting to a “narrow reductive strategy” whereby one (a) finds mental properties which are (b) sufficiently narrow to avoid the phenomenon of multiple realization, while being (c) explanatorily adequate to the demands of psychological theorizing. That is, one replaces the conception of a mental property as more general feature of cognitive systems with many less general properties, for example, replacing the conception of (...) pain as a general property across different species with a number of species-specific kinds of pain. Many critics focused on (c), arguing that species-specific properties are not explanatorily adequate because they do not capture what is common in pain across species. In this paper I focused on (b). I argued that if one carries the narrow reductive strategy out in such a way to avoid all multiple realizability that occurs within a species, for example, accommodating individual differences by postulating a more narrow pain-relative-to-an-individual-at-a-time, then the result is for all intents and purposes a claim about tokens and not types or properties at all (the position is now about individuals having properties at times, a structured token event). I also added an important point about explanation, viz., that properties-relative-to-individuals-at-times would not serve the explanatory goals of science by failing to generalize to other individuals. (shrink)

There is currently a significant amount of interest in understanding and developing theories of realization. Naturally arguments have arisen about the adequacy of some theories over others. Many of these arguments have a point. But some can be resolved by seeing that the theories of realization in question are not genuine competitors because they fall under different conceptual traditions with different but compatible goals. I will first describe three different conceptual traditions of realization that are implicated by the arguments under (...) discussion. I will then examine the arguments, from an older complaint by Norman Malcolm against a familiar functional theory to a recent argument by Thomas Polger against an assortment of theories that traffic in inherited causal powers, showing how they can be resolved by situating the theories under their respective conceptual traditions. (shrink)

Functional reduction follows two familiar steps: a definition of a higher-level or special science property in terms of a functional role, then a statement describing a physical property that plays or occupies that role. But Kim (2005) adds a third step, namely, an explanation regarding how the physical property occupies the functional role. I think Kim is correct. But how is the third step satisfied? An examination of the pertinent scientific explanations reveals that the third step is best satisfied by (...) a multiple-subject, part-whole explanation, which is to say, a decomposition of the occupier’s causal capacities or relations. This is true even in cases wherein role and occupant properties are identical, for an occupier’s causal capacities are always underwritten by a part-whole explanation. As a consequence, functional reduction is transformed into a larger picture that at bottom always contains multiple layers of distinct, nonidentical properties that divide between parts and their whole systems. I call it “Part-Based Functional Reduction.” My aim is to develop this larger picture of reduction. (shrink)

This is a largely expository review of Thomas Polger’s and Laurence Shapiro’s The Multiple Realization Book (Oxford Press 2016).

I critically examine some provocative arguments that John Searle presents in his book The Rediscovery of Mind to support the claim that the syntactic states of a classical computational system are "observer relative" or "mind dependent" or otherwise less than fully and objectively real. I begin by explaining how this claim differs from Searle's earlier and more well-known claim that the physical states of a machine, including the syntactic states, are insufficient to determine its semantics. In contrast, his more recent (...) claim concerns the syntax, in particular, whether a machine actually has symbols to underlie its semantics. I then present and respond to a number of arguments that Searle offers to support this claim, including whether machine symbols are observer relative because the assignment of syntax is arbitrary, or linked to universal realizability, or linked to the sub-personal interpretive acts of a homunculus, or linked to a person's consciousness. I conclude that a realist about the computational model need not be troubled by such arguments. Their key premises need further support. (shrink)

I introduce "nomic-role nonreductionism" as an alternative to traditional causal-role functionalism in the philosophy of mind. Rather than identify mental properties by a theory that describes their intra-level causal roles via types of inputs, internal states, and outputs, I suggest that one identify mental properties by a more comprehensive theory that also describes inter-level realization roles via types of lower-level engineering, internal mental states, and still higher-level states generated by them. I defend this position on grounds that mental properties should (...) be understood by our best scientific theories, which at present include informatioin about mental engineering. I further defend this claim by a "parity of reasons" argument. Causal-role functionalists are justified to include sensory stimuli in their theory of mind as opposed to, say, the remote causes of sensory stimuli because the former but not the latter are items of direct mental production. But ditto for the system's physical realizations. They too directly produce mental states, only not by "causing" them but by "realizing" them. Engineering realizations and their input triggering conditions work in tandem. In addition, I tell a related but more general metaphysical story about property identity, namely, that the traditional causal theory should be replaced by a more comprehensive nomic theory that individuates properties by their intra-level causal powers as well as their inter-level realization capacities. (shrink)

I use Carl Gillett’s much heralded dimensioned theory of realization as a platform to develop a plausible part–whole theory. I begin with some basic desiderata for a theory of realization that its key terms should be defined and that it should be explanatory. I then argue that Gillett’s original theory violates these conditions because its explanatory force rests upon an unspecified “in virtue of” relation. I then examine Gillett’s later version that appeals instead to theoretical terms tied to “mechanisms.” Yet (...) I argue that it too violates the desiderata, since it defines realization for mechanisms in terms of two undefined ideas whose explanatory credentials have not been established—“implementation” and “grounds.” Thus I drop those ideas in favor of an explicit constraint that the parts and properties provide a mechanistic explanation. I also distinguish a special mechanistic theory from a preferred general theory that incorporates other kinds of part–whole explanations that target causal powers or capacities. The result is a theory that has the explanatory virtues of mechanistic theories as well as a broader scope desired by Gillett. I also compare the result to a similar idea from Robert Cummins that has been neglected in recent discussions of realization, namely, his general property analysis rather than his functional analysis. Finally, I defend the preferred general theory against possible objections that attempt to show a conflict between metaphysical demands on a theory of realization versus facts about good scientific explanation. (shrink)

Some scientists and philosophers claimed that there is a converse to multiple realizability. While a given higher-level property can be realized by different lower-level properties (multiple realizability), a given lower-level property can in turn serve to realize different higher-level properties (this converse I dubbed the unfortunately obscure "constructival plasticity" to emphasize the constructive metaphysics involved in this converse to multiple realizability). I began by defining multiple realizabilty in a formal way. (Looking back, one point of interest is that I defined (...) modally weak and strong versions of realization, which is relevant to later views about contingent versus necessitarian grounding.) I then turned to the relation in question. By my analysis, for the converse claim to be true, a lower-level property G1 that realizes a higher-level F must be taken in conjunction with some other base condition G2 so that a difference in G2 allows G1 to determine some other higher-level property E but not F (otherwise there would be violations of supervenience). The realization law thus has the form: (G1 & G2) => F. As such, the base property G1 is insufficient by itself to produce F. It is an insufficient but necessary part of a sufficient condition. I also point out that this makes the realization base property an INUS condition, if combined with multiple realizability. Specifically, if F is multiply realized by properties other than the pair G1 and G2, then G1 is an insufficient but necessary part of an unnecessary but sufficient condition. This is just the first clear and explicit INUS condition view of realization in the literature. Looking back, this converse is also relevant to later views, such as Robert Wilson's view about metaphysically contextual realization. (shrink)

Philosophers of science have offered different accounts of what it means for one scientific theory to reduce to another. I propose a more or less friendly amendment to Kenneth Schaffner’s “General Reduction-Replacement” model of scientific unification. Schaffner interprets scientific unification broadly in terms of a continuum from theory reduction to theory replacement. As such, his account leaves no place on its continuum for type irreducible and irreplaceable theories. The same is true for other accounts that incorporate Schaffner's continuum, for example, (...) those developed by Paul Churchland, Clifford Hooker, and John Bickle. Yet I believe a more general account of scientific unification should include type irreducible and irreplaceable theories in an account of their partial reduction, specifically, when there is a reduction of their tokens. Thus I propose a “Reduction-Reception-Replacement” model wherein type irreducible and irreplaceable theories are accepted or received for the purpose of unifying domains of particulars. I also suggest a link between this kind of token reduction and mechanistic explanation. (shrink)

I focus on and criticize John Searle's argument against a classical computational view of the mind according to which the attribution of syntax is observer relative (in Searle, Rediscovery of Mind, MIT Press, 1992). Searle's argument is interesting inasmuch as it differs from his previous and more well-known argument that syntax is not sufficient for semantics. This argument aims to undercut even the syntax as something that exists only in the eye of the beholder. I show that Searle's argument rests (...) on premises that go unsupported. (shrink)