John Bickle

We introduce a new model of reduction inspired by Kemeny and Oppenheim’s model [Kemeny & Oppenheim 1956] and argue that this model is operative in a “ruthlessly reductive” part of current neuroscience. Kemeny and Oppenheim’s model was quickly rejected in mid-20th-century philosophy of science and replaced by models developed by Ernest Nagel and Kenneth Schaffner [Nagel 1961], [Schaffner 1967]. We think that Kemeny and Oppenheim’s model was correctly rejected, given what a “theory of reduction” was supposed to account for at that time. But their guiding insights about what constitutes scientific reduction—increases in explanatory scope and systematization—reflect actual practices of current reductionistic neuroscience. The key rehabilitative step to make their insights fit current scientific details is to restate them using resources from recent work on causal-mechanistic explanation. We begin with a scientific case study, drawn from the relatively new field of “molecular and cellular cognition”. It provides an explanation of the well-known Ebbinghaus spacing effect on learning and memory in terms of interactions between a transcriptional enhancer protein and its inhibiting phosphatase in neurons recruited into the memory trace. Next we briefly describe some popular models of reduction from mid-20th-century philosophy of science. We point out how these models fail to illuminate key features of our scientific case study. Finally we present our causal-mechanistically updated Kemeny and Oppenheim-inspired model and argue that it nicely accounts for the details of our scientific case study. We close with a remark that will hopefully undercut the surprise many may feel to learn that a long-rejected philosophical account of reduction actually is at work in one of the most prominent reductionistic endeavors in current science.

The Oxford Handbook of Philosophy and Neuroscience is a state-of-the-art collection of interdisciplinary research spanning philosophy (of science, mind, and ethics) and current neuroscience. Containing chapters written by some of the most prominent philosophers working in this area, and in some cases co-authored with neuroscientists, this volume reflects both the breadth and depth of current work in this exciting field. Topics include the nature of explanation in neuroscience; whether and how current neuroscience is reductionistic; consequences of current research on the neurobiology of learning and memory, perception and sensation, neurocomputational modeling, and neuroanatomy; the burgeoning field of neuroethics and the neurobiology of motivation that increasingly informs it; implications from neurology and clinical neuropsychology, especially in light of some bizarre symptoms involving misrepresentations of self; the extent and consequences of multiple realization in actual neuroscience; the new field of neuroeudamonia; and the neurophilosophy of subjectivity. This volume will interest philosophers working in numerous fields who wish to see how current neuroscience is being brought to bear directly on philosophical issues. It will also be of interest to neuroscientists who wish to learn how the research programs of some of their colleagues are being enriched by interaction with philosophers, and finally to those working in any interdisciplinary field who wish to see how two seemingly disparate disciplines--one traditional and humanistic, the other new and scientific--are being brought together to both disciplines' mutual benefit.

The need for representations and computations over their contents in psychological explanations is often cited as both the mark of the genuinely cognitive and a source of skepticism about the reducibility of cognitive theories to neuroscience. A generic version of this anti-reductionist argument is rejected in this paper as unsound, since (i) current thinking about associative learning emphasizes the need for cognitivist resources in theories adequate to explain even the simplest form of this phenomena (Pavlovian conditioning), and yet (ii) the most widely accepted recent theory of associative learning, which utilizes cognitivist resources, has already been reduced to a purely neurophysiological account. Psychoneural reduction of genuinely cognitivist theories is thus already an accomplished scientific fact, despite pronouncements by anti-reductionists about its conceptual impossibility or empirical implausibility. In addition, the specific form of reduction involved in this case (“combinatorial” reduction) provides a promising model for further cognitivist-to-neuroscience theory reductions.

David Marr's three-level method for completely understanding a cognitive system and the importance he attaches to the computational level are so familiar as to scarcely need repeating. Fewer seem to recognize that Marr defends his famous method by criticizing the “reductionistic approach.” This sets up a more interesting relationship between Marr and reductionism than is usually acknowledged. I argue that Marr was correct in his criticism of the reductionists of his time—they were only describing, not explaining. But a careful metascientific account of reductionistic neuroscience over the past two decades reveals that Marr's criticisms no longer have force. Contemporary neuroscience now explains cognition directly, although in a fashion—causal-mechanistically—quite different than Marr recommended. So while Marr was correct to reject the reductionism of his day and offer an alternative method for genuinely explaining cognition, contemporary cognitive scientists now owe us a new defense of Marr's famous method and the advantages of its explanations over the type now pursued successfully in current reductionist neuroscience. There are familiar reasons for thinking that this debt will not be paid easily.

Over the past three decades, philosophy of science has grown increasingly “local.” Concerns have switched from general features of scientific practice to concepts, issues, and puzzles specific to particular disciplines. Philosophy of neuroscience is a natural result. This emerging area was also spurred by remarkable recent growth in the neurosciences. Cognitive and computational neuroscience continues to encroach upon issues traditionally addressed within the humanities, including the nature of consciousness, action, knowledge, and normativity. Empirical discoveries about brain structure and function suggest ways that “naturalistic” programs might develop in detail, beyond the abstract philosophical considerations in their favor. The literature distinguishes “philosophy of neuroscience” and “neurophilosophy.” The former concerns foundational issues within the neurosciences. The latter concerns application of neuroscientific concepts to traditional philosophical questions. Exploring various concepts of representation employed in neuroscientific theories is an example of the former. Examining implications of neurological syndromes for the concept of a unified self is an example of the latter. In this entry, we will assume this distinction and discuss examples of both.

A number of influences have combined to make reductionism an unpopular position in recent philosophy of mind and psychology. Davidson’s Principle of the Anomalousness of the Mental, the multiple realizability arguments of Putnam, Fodor, and others, and attempts to characterize supervenience or dependency as the appropriate nonreductive relation to seek between psychological and physical kinds are the most well-known objections. And these have found their mark. Being a psychophysical reductionist nowadays, as Jaegwon Kim aptly puts it, “is a bit like being a logical positivist or a member of the Old Left—an aura of doctrinaire naiveté hangs over such a person”.

Philosophers and psychologists seeking an accessible introduction to current neuroscience will find much value in this volume. Befitting the neuroscientific focus on sensory processes, many essays address explicitly the binding problem. Theoretical and experimental work pertaining to the “temporal synchronicity” solution is prominent. But there are also some surprising implications for current philosophical concerns, such as the intemalism/extemalism debate about representational content, epistemological realism, a “bottom-up” approach to naturalizing intentionality, Humean concerns about the self, and implications from phantom-limb phenomena. Higher-level theorists about the mind ignore results like these from current neuroscience at their own peril, at least from the point of view of discourse worthy of serious attention as the sciences of the mind/brain push forward into the 21st century.

Social cognition, cognitive neuroscience, and neuroethics have reached a synthesis of late, but some troubling features are present. The neuroscience that currently dominates the study of social cognition is exclusively cognitive neuroscience, as contrasted with the cellular and increasingly molecular emphasis that has gripped mainstream neuroscience over the past three decades. Furthermore, the recent field of molecular and cellular cognition has begun to unravel some molecular mechanisms involved in social cognition, especially pertaining to the consolidation of memories of particular conspecific organisms. Some new experimental techniques for positive interventions into these hypothesized mechanisms offer opportunities for establishing direct causal linkages between intra-neuronal molecular events and the behaviors used to measure social cognitive phenomena. Predicted results from an experiment described below also cast doubt on the application of the “extended mind” approach from recent cognitive science to ground the neuroscience of social cognition. Since neuroethics relies heavily on our best neuroscience of social cognition, that field may soon need to extend its attention beyond cognitive neuroscience, and into neuroscience’s cellular and molecular mainstream

This chapter argues that much discussion between philosophers and neuroscientists is infected by philosophical assumptions about the nature of reduction. Instead we should pursue an unbiased examination of the methods used throughout relevant areas of neuroscience. The chapter focuses on reductionist work in the neurobiological discipline of molecular and cellular cognition. It is argued that reduction is a matter of causal intervention into low level mechanisms, and tracking of the effects of these interventions through levels. When interventions provide evidence that activity in the proposed reductive mechanism co-varies reliably with activity in the target property, such that appealing to higher-level mechanisms will not add any extra explanatory power, reduction can succeed. This is explicitly contrasted with functionalisation and a posteriori approaches to reduction.

This book precis describes the motives behind my recent attempt to bring to bear “ruthlessly reductive” results from cellular and molecular neuroscience onto issues in the philosophy of mind. Since readers of this journal will probably be most interested in results addressing features of conscious experience, I highlight these most prominently. My main challenge is that philosophers (even scientifically-inspired ones) are missing the nature and scope of reductionism in contemporary neuroscience by focusing exclusively on higher-level cognitive neuroscience, and ignoring the discipline's cell-physiological and molecular-biological core.

In their review essay (published in this issue), Looren de Jong and Schouten take my 2003 book to task for (among other things) neglecting to keep up with the latest developments in my favorite scientific case study (memory consolidation). They claim that these developments have been guided by psychological theorizing and have replaced neurobiology's traditional 'static' view of consolidation with a 'dynamic' alternative. This shows that my 'essential but entirely heuristic' treatment of higher-level cognitive theorizing is a mistaken view of actual scientific practice. In response I contend that, on the contrary, a closer look at the memory reconsolidation following reactivation experiments and data suggests (1) a less revolutionary judgment about the proposed alternative, and (2) a now-complete reliance on ruthlessly reductive experimental methods from cellular and molecular neuroscience. These conclusions save the heuristic status I propose for higher-level investigations of behavior and brain. I close with a brief comment on their further charge that I 'sell out' philosophy of science to factual developments in science itself.