Reductionism

Reductionism is the position that holds that theories or things of one sort can exhaustively account for theories or things of another sort. So, for example, reductionism within the cognitive sciences holds that neuroscientific theories will explain the success of psychological theories and, therefore, will reveal that psychological states and processes are nothing but bodily states and processes.

Traditional reductionism within the philosophy of science (Nagel 1961/1979) especially emphasizes some theories' abilities to explain others. It focuses on theories' formal linguistic structures and endorses a model of EXPLANATION that requires that the reduced theory's laws follow deductively from the reducing theory's laws in combination with bridge laws that connect the two theories' predicates.

According to more ambitious reductionism, successful reductive explanation in any science allegedly reveals the in-principle dispensability of the reduced theory and its ontology. PHYSICALISM in the philosophy of mind anticipates a comprehensive reduction of just this kind for our best psychological theories. Both reductionism generally and psychology's reducibility in particular remain controversial.

Reductionistic analyses presuppose that a hierarchy of analytical (or explanatory) levels in science -- encompassing the psychological and sociocultural sciences at its highest levels and the physical sciences at its lowest -- duplicates a hierarchy of organizational levels in nature (Wimsatt 1976). Associated with each organizational level is a distinctive ontology and with each analytical level a distinctive set of theoretical concepts, explanatory principles, and investigative techniques. Traditional reductionism promotes a theoretical and ontological UNITY OF SCIENCE based on a series of reductive explanations of the theories at each level by the theories at the next lower level so that all theories in science, finally, are reducible to, which is to say, explained by, the theories of physics.

Bridge laws establish systematic connections between the predicates of the reduced and reducing theories. The more stringent the reductionism the more stringent the envisioned connections are. Minimally, bridge laws specify lower-level conditions sufficient for upper-level patterns. Such minimal bridge laws would be conditional in form; for example, if humans have a deficiency of noradrenaline in parts of their limbic systems, then they will experience depression. Under such circumstances reduction proves limited in scope, inasmuch as such bridge laws are domain specific. (Other conditions might suffice for depression as well.)

Reductionists are often more ambitious, seeking grounds not just for explaining the upper-level theory but also for claiming that its entities are "nothing but" configurations of lower-level entities. The reducing theory should replace the upper-level theory without explanatory or ontological loss. Because conditions sufficient for a reductive explanation do not guarantee these ontological consequences (Richardson 1979), reductionists seeking ontological economies (Causey 1977) in addition to explanatory consolidation argue that only identities between the theories' predicates will suffice as bridge principles. Such bridge laws amount to empirical hypotheses that initiate new lines of research. Consider, for example, the wealth of new hypotheses concerning the neural mechanisms constituting consciousness.

The identity theory of mind , which resolves the MIND-BODY PROBLEM by holding that mental states simply are brain states, is harmonious with ambitious reductionists' goals. The identity theory avoids the explanatory and metaphysical complexities that dualism introduces. It preserves MENTAL CAUSATION by identifying mental operations with the neural processes that are causally implicated in behavior.

The identity theory presumes many of the conditions necessary for reducing psychology to neuroscience; however, two controversies surround the presumption that all psychological states can be identified with neural ones. First, proponents of non-reductive theories of consciousness insist that no sort of physical information about the brain can explain how conscious mental states feel. The identity theory's defenders respond by showing how neurocomputational findings explain some intuitions about the similarity and relative intensity of QUALIA (Churchland 1989).

The second controversy concerns whether most types of states to which commonsense or FOLK PSYCHOLOGY appeals are readily identifiable with types of brain states. Most philosophers think not, citing numerous considerations to defend that conclusion. The critical question concerns what they make of that failure of systematic intertheoretic mapping.

Those who hold to ELIMINATIVE MATERIALISM exceed reductionists in their confidence in the explanatory sufficiency of neuroscience. They contend that this mapping failure impugns folk psychology, which -- in the face of neuroscience's superior merits and promise -- deserves eradication (Churchland 1989). Like traditional reductionists, eliminativists assume that one account of intertheoretic relations suffices to model not just interlevel relations but changes in theoretical commitments within particular levels too. Some question that assumption and whether theory eradication in science on the scale that eliminativists anticipate ever actually arises in interlevel contexts (McCauley 1986).

Accepting token identities of the mental with the physical, others interpret this failure of systematic mapping as support not for adopting a more radical eliminativism, but for repudiating psychology's reducibility. ANOMALOUS MONISM contends that psycho-physical laws are impossible, because evolving ideals of rationality partially control our use of intentional concepts. Fodor (1981) argues that PSYCHOLOGICAL LAWS are not exceptionless and that the multiple realizability of psychological types is widespread, resulting in uninformative bridge principles with lengthy disjunctions of neural types that might realize some psychological type. Both considerations impede reduction and favor the AUTONOMY OF PSYCHOLOGY. Because psychological explanations rely on intentional contents, reductionism's requirements fail to guarantee their subsumption by neuroscience. Fodor claims that reducing psychology requires demonstrating that neural states instantiate a code but that this yields a neuroscience unamenable to further reduction.

Noting reductionism's restrictiveness in accounting for cross-scientific relations, opponents of the autonomy of psychology are untroubled by mapping failures. Paul Churchland stresses approximate reduction, where lower-level theories preserve an equipotent image of upper-level theories without comprehensive mapping. Patricia Churchland (1986) emphasizes how coevolving theories will often prove mutually instructive -- yielding progressively better mappings.

Others have adopted even more pragmatic approaches, advocating integrative models of cross-scientific relations (Bechtel 1986). Examining issues of discovery, evidence, method, and more, they foresee many illuminating relationships (besides possible reductions) between psychological, neurocomputational, and neuroscientific models (McCauley 1996). Within some suitably restricted domains reductions of some psychological principles are possible (Bickle 1995), but Bechtel and Richardson (1993) argue that the chief goal of reductionistic research is the discovery and explication of the mechanisms underlying the functioning of complex systems. Integrative modelers pursue analyses showing how upper-level research often plays central roles in justifying lower-level proposals (McCauley 1996), motivating innovative research at intermediate levels (such as connectionist modeling), and stimulating research in what Hardcastle (1996) calls "bridge sciences" (such as event-related potential studies). Compared to standard reductionism, these integrative models examine a wider range of the cross-scientific relations that arise in the sort of interdisciplinary research characteristic of cognitive science.

Additional links

-- Robert McCauley

References

Bechtel, W., Ed. (1986). Integrating Scientific Disciplines. The Hague: Martinus Nijhoff.

Bechtel, W., and R. C. Richardson. (1993). Discovering Complexity. Princeton: Princeton University Press.

Bickle, J. (1995). Psychoneural reduction of the genuinely cognitive: Some accomplished results.Philosophical Psychology 8:265-285.

Causey, R. (1977). Unity of Science. Dordrecht: Reidel.

Churchland, P. M. (1989). A Neurocomputational Perspective. Cambridge, MA: MIT Press.

Churchland, P. S. (1986). Neurophilosophy. Cambridge, MA: MIT Press.

Fodor, J. A. (1981). Representations. Cambridge, MA: MIT Press.

H ardcastle, V. G. (1996). How to Build a Theory in Cognitive Science. Albany: SUNY Press.

McCauley, R. N. (1986). Intertheoretic relations and the future of psychology. Philosophy of Science 53:179-199.

McCauley, R. N. (1996). Explanatory pluralism and the coevolution of theories in science. In R. N. McCauley, Ed., The Churchlands and Their Critics. Oxford: Blackwell.

Nagel, E. (1961/1979). The Structure of Science. Indianapolis: Hackett.

Richardson, R. (1979). Functionalism and reductionism. Philosophy of Science 46:533-558.

Wimsatt, W. (1976). Reductionism, levels of organization, and the mind-body problem. In G. Globus, G. Maxwell, and I. Savodnik, Eds., Consciousness and the Brain. New York: Plenum Press.

Reductionism

Additional links

References

Further Readings