This item is only available as the following downloads:
THE HARD REDUCTION: PSYCHOLOGY AND NEUROSCIENCE BY CALEB KAUFMAN A Thesis Submitted to the Division of Social Sciences New College of Florida In partial fulfillment of the requirement of the requirements for the degree Bachelor of Arts Under the Sponsorship of Dr.Gordon Bauer Sarasota, Florida February, 2009
ii Acknowledgments I acknowledge Gordon Bauer, for guiding me th rough a conceptually broad investigation with patience and insight; Al Beulig, for hi s fostering in me an interest in mind-body science; and Heidi Harley, for her humor and support. I would also like to thank Aron Edidin for his willingness to help me fill in for the philosophy educat ion I never actually got, and my roommates: Ian and Lauren; to them: fish sauce.
iii Table of Contents Acknowledgments .....ii Table of Contents ..iii Abstract .........iv Chapter I: What is Reductionism? .............................................................................1 Chapter II: Mind-Body Th eories and Reduction........................................................18 Chapter III: Reducing the Personal Reality................................................................46 Chapter IV: Psychoneuroimmunology as Interactionist Exemplar............................82 Chapter V: Conclusion ..............................................................................................118 References..................................................................................................................123
iv THE HARD REDUCTION: PSYCHOLOGY AND NEUROSCIENCE Caleb Kaufman New College of Florida, 2010 ABSTRACT As theory from cognitive psychology and clinical psychology are being formally included under the auspice of neuroscience, the questi on of integration arises. Reductionism and the mind-body problem provide a framework from which to examine this question. Reductionism highlights partic ular problems in attempting to explain the mental, experiential content of psychology in terms of neurobiology, such that identity claims cannot yet be made. Nonethele ss, attempts to explain, and explain away the mental experiential content of psychology ma ke themselves evident in contemporary neuroscience. Interactionism as a method, is a viable option which respects the irreducibility of the mental, while a llowing productive research to continue. Dr. Gordon Bauer Division of Social Sciences
CHAPTER 1: WHAT IS REDUCTIONI SM, AND HOW DOES IT RELATE TO PSYCHOLOGY? It is not absurd to suppose that psychologica l laws may eventually be explained in terms of the behavior of individual neurons in the brain; that the behavio r of individual cells including neuronsmay eventually be explained in terms of their biochemical constitution; and that the behavior of mol eculesincluding the macro-molecules that make up living cellsmay eventually be explained in terms of atomic ph ysics. If this is achieved, then psychological laws will have, in principle, been reduced to laws of atomic physics. -Paul Oppenheim How can optics be explained by electromagnetism, or thermodynamics by statistical mechanics? The relation of scientific concepts, theories, and disciplines to each other has been a topic of inte rest for perhaps as long as th e classical mechanics of the 17th century (Sarkar, 1992). Galileo developed the Method of Resolution and Composition, which says that to understand a phenomenon you n eed to break it into parts. According to Galileo, you then study these parts and attemp t to derive the phenomenon from what you learn from the parts (Horst, 2007). This idea contributed greatly to classical mechanics, but it was not until 1949, according to Sahotra Sark ar (1992), that a rigorous analysis into the logic of scientific expl anation was commissioned, by the philosopher Ernest Nagel. Nagel wanted to know the way in which Galileos parts relate to each other. Nagel came out of the positivist-empiricist tradition, predominant in the first half of the 20th century (Hooker, 1981). For the positi vists, deductive reason and the human
2 senses were the primary means of doing scien ce; thus, when Nagel introduced his idea of inter-theoretic reduction (Sar kar, 1991), it was based prim arily upon establishing logical connections between concepts and laws (Hooker, 1981). Reduction for Nagel was a form of explanation that evidenced, base d upon logical deduction and empirical data collected in science, the relation be tween two things (Nagel, 1961). Reductive explanation says that when one concept, theory, or discipline can be explained by another, it has been reduced by it (Horst, 2007). In the philosophy of scien ce, reductionism is a term that connotes a method for thinking about how science can be used to e xplain science. The word reductionism has been accused of vagueness, and is frequently us ed pejoratively to mean all the things one does not like about science (Dennett, 1995), but reductionism has sp arked many different ideas about scientific inquiry, and the plausibility of reduc tive explanation (Horst, 2007). Nagel was not the first to use the word; it can be traced back in academic literature to W.M. Urban, in 1925. Urban uses it in a different way however, in analyzing the concept of evolutionism, he says that evolutionism is reductionism (p.93). He means that by showing that the complexities su ch as mind and life emerge from different things then they are shown not to evolve at all (Urban, 1925). Urban does not mean the word here but in the general sense of explanation-by-components, akin to Galileos method. If taken in this broad sense, reduc tionism has existed since the time of the ancient Greeks, who thought for a time that all things were composed of the primal elements. It was not until Nagel that the word took on its specific and operational meaning, which is currently employed in philosophy (Hooker, 1981). The Oxford English Dictionary (2009, p.1) defines reductionism as:
3 The practice of describing or explaining a complex (especially mental, social, or biological) phenomenon in terms of relatively simple or fundamental concepts, especially when this is said to provide a sufficient description or explanation; belief in or advocacy of such an approach. Analogously, the Stanford Encyclopedia (2008) of philosophy says the following about reductionism (p.1): The basic question of reduction is whethe r the properties, concepts, explanations, or methods from one scientific domain (typi cally at higher levels of organization) can be deduced from or explained by th e properties, concepts, explanations, or methods from another domain of scienc e (typically one about lower levels of organization). For this thesis I draw from both of these descriptions, as well as those who have written on reductionism, in the wake of Nagel. Following his description of the logical contingencies of reduction, ph ilosophers attempted to apply reductionism to extant sciences. Initial debates on reductionism were concerned with the basal disciplines: physics and chemistry (Sarkar, 1992). But si nce its inception in the 1950s, reductionism has extended beyond the bounds of physics and chemistry, into the special sciences (those sciences that, as of yet, have not been explained by physics). It has also branched out into the philosophy of mind (Horst, 2007). Steven Horst says in his book, Beyond Reductionism, that currently the aprioristic, mathematical approach espoused by the logica l empiricists has been abandoned. He says that philosophy of science is more concer ned with studying the methods of specific scientific disciplines (Horst, 2007), and that the treatment of science in terms of reductive explanation has passed on. But while philos ophy of science may be beyond reductionism,
4 Horst goes on to say that it continues on in the philosophy of mind (Horst, 2007); particularly, he sees reductionism in philo sophy of mind as concerned primarily with the necessity of being able to explain the special sc iences in terms of the physical sciences (Horst, 2007). For this thesis, I am ex amining the special science of psychology. Karl Popper and John Eccles (1974) provide a diagram of the canonical reductionist hierarchy (Table 1) Rather than including specific disciplines, they describe levels of description, which may encompass a num ber of distinct fiel ds of inquiry. It should be noted that at the time this hi erarchy was formulat ed, psychology was not central to reductionist debate. It can, howev er, be assumed to fall under the realm of ecosystems, along with the other social sciences, whic h exist at the level of description of people (Medawar, 1975). Table 1 Summary of Reductionist Hierarchy Level of Description Ecosystems Populations of metazoa and plants Metazoan and multicellular plants Tissues and organs Populations of unicellular organisms Cells and of unicellular organisms Organelles Liquids and solids Molecules Atoms Elementary Particles Sub-elementary particles Unknown: sub-sub-elementary particles Can psychology be explained reductively by the physical sciences? Considering the recent inclusion of psychological concepts wi thin contemporary neuroscientific research
5 (Bear, Connors, Paradiso, 2007), I think this is an important question to ask. Subsumed by this question are two further concerns: if psychology can be reduc tively explained, can it be abandoned as a sc ientific discipline; a nd how might this explanation take place, in light of contemporary neuroscience? To examine this question, I will draw from different leve ls of abstraction, including terms from philosophy of scie nce, philosophy of mind, psychology, and neuroscience. I do not consider these te rms hierarchically; any mention of the hippocampus should not be considered more im portant than memory or mental state. I consider all of these levels of description to be important to the discussion. Because of the range of objects that can be considered in reductionism (theories, concepts, models, etc.) using the terms T for the reduced thing in question, and T for the reducing thing in question is helpful. T and T will be utilized when necessary to help clarify a statement or give an example, where no concrete one is available; T will usually denote the level of explanation of psychology, and T will denote that of neurophysiology. I believe examining how neuroscience may attempt reductions from psychology to the physical sciences is a useful endeavor. I believe that if there is not one already, there may become a palpable threat to th e mental terms that psychology utilizes, one example of which might be mental illness. I also believe that philosophically examining neuroscientific research will make transp arent extant reductionist attitudes. If philosophical inquiries, part icularly those related to reductionism and the mind-body question, are not considered, then the re search produced by neuroscience may be conceptually misguided. Bennett and Hacker (2 009, p. 3-4) clarify this idea, discussing how philosophical questioning is related to science: Conceptual questions antecede matters of truth and falsehood. They are questions
6 concerning our forms of representation not questions con cerning the truth or falsehood of empirical statements whe n empirical problems are addressed without adequate conceptual clarity, mi sconceived notions are bound to be raised, and misdirected research is likely to ensue Science is no more immune to conceptual error and confusion than a ny other form of inte llectual endeavor. According to Bennett and Hack er, science is concerned wi th truth and falsehood, while philosophy is concerned with forms of re presentation (Bennett & Hacker, 2009). How can research be conceptually misguided due to its forms of representation? What is a form of representation? Two examples are time and space. Time and space are fundamental forms used to represent things in the world. They have both constrained and informed our methods for studying it. If time and space did not exist as real concepts, then the conclusions we would draw about th e world would be radically different. If the pre-Socratics had never consid ered the philosophical question, What is the world made of?, and had Democritus not posited that ma tter is composed of smaller parts, for instance, then we may have never developed chemistry. But assuming most of the fundamental questions have been asked, why is it not enough to the scientific met hod, investigating new and interesting phenomena? I believe that, in fact, the question of reductionism is considered implicitly by many nonphilosophers to be resolve d. But not addressing the way in which explanations are provided in scientific resear ch may, in fact, lead to faulty conclusions, on logical grounds. Specifically, I believe th at contingent on assumpti ons about whether psychology may be reduced to the physical sciences, neur oscientists will draw different conclusions from their research.
7 There are other views which I believe ar e even more implicit than reductionism. Though they are not incompatible with reducti onism, they do not necessarily entail it. These two views are naturalism and physicalis m. Naturalism and physicalism themselves are closely related, but are di stinct (Horst, 2007). Physical ism posits that there are no properties of things in the world which are not physical. It is the same as materialism. Basically, it makes the strong metaphysical conj ecture that for even those things which we do not yet understand, we may still assume that those things are physical. Naturalism, for the proposes of discussi ons in philosophy of science, means that for any given domain of inquiry, all features of the domain are provided by the definition of nature, as accepted by the natural sciences (Horst 2007). In other words, naturalistic theories cannot posit the existence of supern atural phenomena, or a metaphysical theory which is not grounded in the natural sciences (Hor st 2007). This does not mean that one must dogmatically adhere to the scientific method. Rather, naturalism says that for any inquiry into any domain of the world, that inquiry ma y be satisfied w ithin the bounds of the concepts employed by the natural sciences. I will argue that an implicit assu mption of physicalism makes certain neuroscientific endeavors simultaneously reduc tive endeavors. This argument may be framed in two ways: as an argument against what is an extant, but unspoken reductionist attitude, or as an argument against hypothe tical reductionist claims. The latter framing may make my thesis susceptible to criticis m, via the Straw Man argument, but I do not believe it falls victim to this logical fallacy. A straw man argument creates the illu sion of having refuted an argument by substituting it with a superficially similar one (Madsen, 2007). In this case the paradigm I am arguing against is hypothetical. I am not pointing at specific individuals and saying
8 there, that is who I am arguing against. Ra ther, I am taking the principles set out by Nagel, and his followers, and applying them again in light of contemporary neuroscience. That being said, my thesis could also be accused of an argument from probability, saying that because something could happen, it will happen (Murphey, 2007); alternatively, ignoratio elenchi which is an irrelevant thesis; presenting a valid argument but not addressing the issue in question (Mill, 1959). At the end of this thesis I will revisit these potential criticisms. How exactly science may draw different conclusions based upon its implicit answer to reductionism requires further b ackground. Reductionism is concerned with how and if things can be reductively explained. It does not assume physicalism, although it is in the repertoire of di fferent reductionist arguments. Also, successful reductions come in different forms. A broad reducti on has two formal attr ibutes (Horst, 2007). The first attribute of a broad reduction is part-whole explanation. This means that explanations must describe feat ures in an entire system or discipline in terms of the relations of its proper parts, or components th at are no more complex than its parts. For example, a computer may be considered in terms of its hardware components: hard drives, central processing unit, battery, et c., which individually, are no more complex than the computer as a whole. The second formal attribute is explanations without remainder or explanations that are comprehensive in the sense that they can describe the full range of observations in the reduced system (Horst, 2007). The explanation, in most cases, need only be bottom-up. If a discipline can explain anot her in terms that are no more complex than the ones employed in that discipline, and if it explains it fully, then the latter discipline can be considered reduced (Horst 2007); though, broad reductions are not limited to
9 disciplines. Any broad question relating some T to some T is open to the requirements of broad reductionism. If someone were to say, all ecology can be explained by Darwinism, then to evaluate that claim, th e principles of broad reduction may be used. Austen Clark, in his book Psychological Models and Neural Mechanisms, expounds upon derivative, or derivation reduction, and eliminative reduction. Derivation reduction considers the reducing explanations, given by T, as a second set of explanations for T (Clark, 1980). That is, T gives an alternative set of explanations. Derivational relations are usually put in term s of logical sentences conjoining laws, and connecting principles. They w ould play out like syllogisms. Such a sentence might take the following form: For every x, if x is P, then x is Q. In derivation reduction, one derives the validity of a claim based upon the logical continuity between its premises. One builds the explanation for T in terms of T. For example, for every apple, if an apple is r ound, and for every sphere, if a sphere is round, then apples are spheres. Deri vative reduction has the requirem ent that each theory or law be unambiguous, and operationalized in terms of observable, testable hypotheses. If any single theory in T describes observations radically different than in T then explanations in T cannot be used to predict, or be cons idered a corollary to explanations in T Clark considers this problem directly by admitting th at there are obvious discrepancies between theories in different disciplines (Clark, 1980). For instance, in the above example, appl e is simply a nomin al description of some object that has properties, while sphere is a concept from geometry, which has its own properties, put in terms of mathematics. In order to say that apples are spheres, you would have to connect both, in terms of these properties. But there seems to be no innate commonality between an apple and the mathema tical axioms dictating the dimensions of
10 a sphere. Because of this difficulty, Clark offe rs that derivation reduc tion can take into account connecting principles. Connecting principles will often attempt to find a conceptual middle-ground. For example, in relating thermodynami cs to statistical mechanics, temperature is used to also mean mean kinetic energy of mol ecules (Clark, 1980). In the case above, roundness is the connecting principle. Conn ecting principles need not be ontological claims (claims about what exists) but need only to make the sentences relating T to T understandable. This might take the form of integrating terms as operators, e.g. If x is Q then x is W or If x is Q then x is not W or, Either x is Q or x is W. (Clark, 1980). There is some ambiguity as to how extensive these connective principles may be, but deriviative reduction is conc erned with general laws in T being proven through logical deduction to be derived from laws in T As long as propositional content is logical within the frameworks of both theori es, containing both necessary and sufficient conditions, then it is open to deri vation reduction (Clark 1980). The other form of reducti on worth consideration is eliminative reduction. Eliminative reduction states that, assuming T will provide a more fundamental explanation, it should subsume T This is what Clark terms the replacement assumption: If T reduces to T then any explanation using a term from T can be replaced by some explanation using a term or terms from T (Clark, 1980). That is to say, one theory should be reduced to the other, and no longer used in an explanatory way (Clark, 1980). If a term from psychology is reduced to a brain term, then in any instance where you might use that psychological term, you would instead use the brain term. The key difference between eliminative and broad reductionism is that while broad reductionism states that the reducing theory T explains all of the features of the
11 reduced theory T it does not have to abide by the replacement assumption. Broad reduction provides a framework, and fundament al principles by which other types of reduction must abide to be considered su ccessful. In the philos ophy of mind, there are two specific stances which employ eliminativ e reduction. These are eliminativism and reductive materialism. It is from these that I perceive an extant threat to psychology. One example of eliminative reduction is that of light to electromagnetism. Before optics, vision was understood intuitivel y. Many philosophers believed that the phenomenon of seeing played out through mean s of the eyes alone (Churchland, 2005; Gundert, 2000). For them, there existed a di rect relation between gaze and percept. When you gaze at an object you send out a sort of laser from your eye which connects you to the object and allows you to see it (Churchland, 2005). There existed no concept of light being reflected by objects. Eventually it was discovered that light is the mechanism by which we see, and that it consists of sub-atomic particles os cillating in space. Depending on when it is observed, light sometimes behaves as a partic le (a photon) and sometimes as a wave. Whatever the case, over time the conception of seeing as being centere d fully in the eyes was reduced to the theory of light, which wa s even further reduced to the operations of sub-atomic particles (Churchland, 2005). Eliminative reduction also runs into th e problem of conceptu al incompatibility, but sometimes a term from a reduced th eory describes common enough phenomenon that it can be brought over to the eliminating theor y. The word itself will continue to be used, but its definition changes. For example, before thermodynamics was understood, phlogiston was assumed to be what gave mate rial objects the genera l property of heat. Phlogiston as a concept was eliminated, but we still speak of heat. The concept of heat is
12 common enough that it gets carried on, but we also understand that what heat is includes thermodynamic forces, and the collisions of pa rticles causing electrons to move into a lower quantum orbital. Within the reducti on, an identity claim is embedded, wherein heat is shown to be identical to a thing wh ich previously it was not (Churchland, 2005). Summary of Reductionism Reductionism is concerned with how and if things can be reductively explained. It does not assume naturalism, nor physicalism. Although its concerns are philosophical, to prove a successful reduction empirical da ta are often relied upon. At some point, reductionism was applied to the special sciences, and branched out into the philosophy of mind, T becomes the body and T becomes the mind. Because psychology employs mental terms such as mental illness, I believe asking how the mind and the body relate is analogous to asking how psychol ogy and neuroscience relate. Derivative reduction attempts to an swer this question by building logical sentences, using terms and connecting principles from neuroscience to derive psychology. Eliminative reduction attempts to show that explanations in neuroscience prove psychological explanations to be inaccura te. It uses the same methods as derivative reduction, but makes the replacement assu mption, which says that after reducing psychology to neuroscience, you eliminate it. Bo th derivative and eliminative-reduction must adhere to the principles of broad re duction, which are part-w hole explanations, and explanations without remainder. Neuroscience and Psychology Psychologys inclusion into neuroscience I believe is often predicated on the assumption that the constructs employed by ps ychology are, in some sense, reducible to the neurobiological. Indeed, neuroscience ha s been accused of utilizing a reductionist
13 approach (Bear, Connors, Paradiso, 2007). Th is general claim is difficult to falsify, considering the broadnes s of the field of neuroscience, which may be better described as a meta-discipline, in that it integrates fi ndings from a variety of different fields. Specifically, it groups its constituen t researchers into computational neuroscientists, developmental neurobi ologists, molecula r neurobiologists, neuroanatomists, neurochemists, neuroeth ologists, neurophysiologists, physiological psychologists, and psychophysicists (Bear, C onnors, Paradiso, 2007). It has evolved in relation to several technological advances. These include, but are not limited to, gene sequencing, proteomics, structural mode ling, and brain imaging (Bear, Connors, Paradiso, 2007). For a phenomenon in question, these techniques can be used concurrently to attempt to provide an integrated explanation. For ex ample, after a geneticist has identified a gene related to some phylogenetic trait, gene seque ncing can elucidate the specific sequence of nucleotides, which encodes for amino acids. A molecular biologist, working off this knowledge, can attempt to isolate the proteins that this gene encodes. Structural mechanics can assist in examining the struct ure of those encoded proteins, and how they interact with each other. Th e other disciplines may use their methods of investigation; developmental neurobiology might look at the way in which molecules organize into tissues, which develop into the nervous system; this eventually might lead to the level of neurophysiology, wherein brain im aging could investigate the neural activity associated with the given phylogenetic trait. But how is psychology, a special science, been related to neuroscience? Freud, an influential figure in clinical psychology, was himself a neurobiologist. He at least nominally identified his theory about the origin of the Id to sub-thalamic brain structures
14 (Freud, 1940). Even before the brain was well understood, Freud was attempting to relate a mental concept with a physical one. But since the days of psychoanalysis, psychology has gone though several paradigm sh ifts, leading to more explicit overlap between its interests and ne urosciences interests. With the rise of behavioral research in the early 20th century there was increased focus on directly observable phenomena, which allowed for sp ecific laws to be generated in order to predict behavior. Behaviorism di d not think that intern al, mental processes, were worth study (Milner, Squire, Kandel, 1998). After this, there was a cognitivist turn during the mid-20th century, which oriented the fiel d towards an increased focus on models of perceptive and sens ory phenomena, rather than behavior (Milner, Squire, Kandel, 1998). That is not to say that the method of behaviorism was abandoned. Cognitive psychology began with Francis D onders investigation of decision making, in 1868, and then Broadbent and Gibsons work in the 1940s on human interactions with machines (Goldstein, 2007). By 1967, Ul ric Neisser published the first cognitive psychology textbook, simply named Cognitive Psychology which helped to concretize it as a field of inquiry. In Cognitive Psychology Neisser advocated the use of experimental approaches to understanding and solving problems related to external sensory stimuli, perception, and the mental processing that mediates them (Neisser, 1967). Where previously a sort of hierarchical division existed between the processes of stim ulation, sensation, perception, cognition, and behavior, cogni tive psychology assumed that perception entailed both external sensory information and intern al mental processing. Hence, cognitive psychology attempted to bridge the gap be tween sensation and perception (Milner & Squire, 1998). Unlike behaviorism, it acknowle dged the importance of mental states.
15 Making the claim that cognitive psyc hology acknowledged the mind is more difficult. Cognitive psychology took an approach which did not necessarily attempt to make metaphysical conjectures, was more c oncerned with establishing black box models that could be used to explain phenomena such as memory and attention. Instead of saying thinking, the term cognition is used (Gol dstein, 2007; Neisser, 1967). Neisser defined cognition as the study of all processes by which sensory i nput is transformed, reduced, elaborated, stored, recovered and used (Nei sser, 1967). He included in cognition things that occur independent ly, without external input, includi ng internal representations of images and hallucinations (Neisser, 1967). This functional, information processing approach had signifi cant overlap with systems neurobiology, which was interested in understanding the st ructure and function of neural circuits (Milner, Squire, Kandel, 1998). The two were integrated, and called cognitive neuroscience which is a term denoting all of the pursuits in neuroscience which attempt to find neural networks associated with cognitive processes (Milner, Squire, Kandel, 1998). Acknowledging that cognitive neuroscience is just one division within neuroscience and the sheer amount of research that has, and is being conducted (last years Annual Meeting garnered over 30,000 attendees), is there any room for philosophical inquiry? Is the mind body question irrelevant at this point? To get a better understanding of this questi on, a meta-search of annual re view volumes for psychology, medicine, physiology, biochemistry, geneti cs and neuroscience was conducted. The criteria included being publishe d from 1980 2009, and titles containing both the words mind and brain. The sear ch yielded 344 separate publications. This would seem to indicate that there exists at le ast a sub-section of scientists interested in
16 publishing on the mind-body problem. I believe that the mind-body problem is still relevant, and is worth consideration. As stated above, I believe neuroscientific experimental methodology is in fact predicated on implicit assumptions about the mindbody problem and reductionism; hence, a rec onsideration of the mind-body problem in light of neuroscience is worthwhile. Statement of Goals The primary concerns of this thes is include the reduction of psychology, particularly cognitive theori es, to neurosbiology, and how a philosophical examination of neuroscience, in the context of the mi nd-body problem, may affect neuroscientific inquiry. The possibility of reducing psychol ogy to biology will be examined. The type of reductions espoused by certain mind-body theori es will be shown to be in violation of one of the broad reduction requirements, which is explanation without remainder. That is, a broad reduction of psychology to ne uroscience would leave something out. I understand this to be subjective experience. This is not a critique of neuroscience, a nd this thesis is explorative, rather than destructive. I believe the dom ains of neuroscience are apt in revealing facts and findings about the human nervous system, which will lead to great advances in medicine. However, without a philosophical bedrock, sorting through a large body of data and trying to draw conclusions, especially as regards the cause of any given behavioral phenomena is, will be difficult. A rigorous ph ilosophical introspection will perhaps not answer this type of question, but it may pr ovide a more conceptually guided means by which to try and answer it (Bennett & Hacker, 2009; Devlin, Poldrack, 2007).
17 The goals of this thesis are as follows: 1. Provide an overview of mind-body theories within the context of reductionism. 2. Examine the problems with an eliminativist and reductive materialist approach. 3. Examine a reconciliatory way of going a bout neuroscientific research that does not require strong redu ctionist claims.
18 CHAPTER II: MINDBODY THEORIES AND REDUCTION All things are water. -Thales of Miletus Context Predominant theories in philosophy of mind may be considered under the domain of reductionism. By asking the mind-body question, one is implicitly asking: how do the mind and body relate? This question is analogous to asking how two scientific disciplines may relate. If one replaces m ind with the domain and theories of psychology, and body with neur obiology, the contrast is cl earer. I acknowledge that psychology, as a discipline, does not limit itself to mental terms, but these terms are what I am concerned with. Further examples might include mental illness, emotion, volition, and personality; essen tially, those terms which attempt to describe phenomena that occur inside the mind of a person. Besides mind and brain, I will also employ the terms mental and physical, which should be considered synonymous, resp ectively. This chapter will continue to introduce important concepts relevant to psyc ho-biological reduction, as well as consider some predominant mind-body theori es, in the context of reductionism. It is my belief that any given theorys answer to the mind-body que stion will reveal it s attitude towards reduction. One useful way to examine mind-body theori es is by considering their approach. The question can be framed in three ways: the ontological, epistemological, and methodological. The ontological question dates as far back as philosophy itself. It is
19 primarily concerned with what exists, and what properties those existi ng things have. It asks what the mind and body are, what their constituent properties are, and how they relate, if at all. The epistemological question is concerne d with our ability to understand reality; more strictly, the limits of human understanding as regards the facets of mind-body relation. The methodological is re lated to the epistemological, in that it offers the correct way in which the mind-body question should be addressed. For instance, logical positivism said that any statement is only meaningful if it can be evaluated in terms of being true or false (Hempel, 1950). I am con cerned primarily with the ontological and the methodological. Within the mind-body question, there ar e two broad categories of theories: monistic ones and dualistic ones. Monism is the belief that the mental and the physical are the same. Dualism is the belief that the mental and the physical are different (Churchland, 1988). Often, monism is conf lated with physicalism, but as will be demonstrated, this is not necessary. Th ere are subtle versio ns which, though not incompatible with physicalism, take different reductionist stances. The Mind-Brain Problem The mind-body question now tends to connote the mindbrain problem (Bennett & Hacker, 2009). This shift was probably in fluenced by the lesion studies of Broca, in the 1860s. Broca found that there are speci alized areas of the brain which, when damaged, result in explicit and predictable defi cits in behavior (Kandel, 2000). Perhaps the most canonical of these is Brocas area, a small section of neural tissue in the posterior inferior frontal gyrus Damage to Brocas area re sults in Brocas aphasia, a condition which entails the loss of th e ability to vocalize language.
20 Later, Korbinian Brodmann, through his observations of the brain using Nissl stains, posited a cytoarchitectural map (Fi gure 1), which organizes the neocortex into different regions (Kandel, 2000). These regions have been correlated with a diversity of cognitive functions. For example, the primar y somatosensory cortex in the postcentral gyrus (Brodmanns areas 1,2,3) has been correl ated to the sensation of touch (Kandel 2000). It can be seen in the figure below co lored orange, on the dorsal side of the brain (Figure 2). Brodmanns areas have been further sp ecified through brain imaging, and an increased understanding of the way neurons are connected to each other, via cortical columns and neural networks (Kandel, 2000). These discoveries of functionally localized areas have contextualized what I believe is the reductive approach, neuroscience has been accused of utilizing.
F Fi g u r F i g ure 2. T h r e 1 Brod m h ree Dimens i m anns Cyto a i onal Interp r a rchitectural r etation of B Map of the B rodmanns A Brain (Bro d A reas d mann, 1913 21)
22 A change in cognitive functioning due to brain damage, such as in the discovery of Brocas area, is seen by mindbody theorists as an example of neural dependence (Churchland, 1988). Neural dependence says that because faculties of the mind can change from brain damage, they cannot exist se parately from it. The concept of neural dependence has been used to provide argument for and against certain mind-body theories. This consideration might be consider ed as a bottom-up approach, or the bodymind question, instead of a top-down a pproach, or the mind-body question (Shanon, 2009). It seems that the approach will likely reflect underp inning ontological assumptions, whether consciously considered or not. If a neurosci entist operates under the assumption that the mind is purely material then an attempt to use brain states to explain or reduce mind states will follow. In contrast, if a neuroscientist is a dua list, then perhaps a modular approach would be taken, wherein she or he begins with a mental state, and attempts to explore the relation it has with brain state: if it co-occurs, if it arises from it, if it interacts with it, etc. What specific approach might be taken will depend on the type of dualism a neuroscientist believes in. Substance Dualism Rene Descartes is typically associat ed with substance dualism. In fact, Cartesian dualism is synonymous with substance dualism (Gundert, 2000). Descartes was one of the first to offer a specific brain based mechanics for mindbody effects. Observing the now classic patellar, or kn ee-jerk reflex, he offered an explanation that nerves, conceptualized as networks of hollow tubes r unning from the heart to the brain, transmute animal spirits which cause movement. Animal spirits are produced in the pineal gland,
23 which would send them towards muscles. The pi neal gland was Descartes answer to the question of how mind stuff a nd physical stuff were supposed to interact. The pineal gland translates the volition of the so ul into bodily moti on (Gundert, 2000). Substance dualism says that each mi nd is a distinct, sp atially constrained, nonphysical substance. Its identity is i ndependent of any body to which it might be attached to (Churchland, 1988). It is s upposed that mental phenomena like language, reason, and volition arise from th is mental substance. Proba bly the best argument against substance dualism is the problem of causation. If the mind is supposed to affect the body, and the mind is non-physical, how can it ever be proven in naturalis tic terms that the mind causes motion? Whereas physical substances are gover ned by natural laws and properties like mass, position, color, etc., mental substan ces presumably do not have such rigid dimensions. Rather, Descartes believed that the essential characteristic of mental substance was thinking (Gundert, 2000). How can the tw o be connected? Descartes did attempt to find a connecting principle for the mental substance thr ough animal spirits. Even he knew that momentum cannot origin ate from nothing; but unfortunately the theory of animal spirits cannot be tested on naturalistic grounds, and the problem remains (Gundert, 2000). Because substance dualis m considers the mind as fundamentally separated from the body, it does not make an atte mpt to explain reduce it in terms of the body. Neural dependence is provided as an ar gument here. If mental states can be influenced by changes in brain states, then thinking cannot occur independently from the activity in the brain. For exampl e, if one considers negative thinking to be a symptom of depression, then anti-depressants, which work to try and alter brain chemistry, would
24 seem to be a good description of how the mental is affected by the physical, and not vice versa. As mentioned earlier, this is the type of bottom-up approach commonly employed by materialists (Shanon, 2009). Property Dualism Property dualism says that while the brain is a physical substance, its properties are non-physical. This may be confusing at first, because when scientists talk about properties they typically mean physical pr operties of matter (C hurchland, 1988). A nonphysical property must be explained on its own terms, typically in a negative sense -to be proven as not a physical property. Property dualis m attempts to bridge the gap of rigorous commitment to physicalism, with reco gnition that mental states exist, but are properties of brain states. Epiphenomenalism Epiphenomenalism says that mental propert ies are the intuitive constructs that we think of as having a causal effect on physical states, but that they do not in fact. These include things like vo lition and emotion. Epiphenomena are supervenient upon physical states in that physical states are assumed to constitute them, but cannot be used to explain them from the ground up. One formal defi nition of supervenience says the following (Bennett, McLaughlin, 2005, p.1): A set of properties A supervenes upon another set B just in case no two things can differ with respect to A -properties without also diffe ring with respect to their B properties. In slogan form, there cannot be an A -difference without a B difference. An example of supervenience might be indi viduals and society. There can be no change in society, without a change in individuals, but you could not necessa rily explain society
25 from the ground up, therefore society superv enes upon individuals. Supervenience is important to discussions of the mind and br ain, especially in the context of neural dependence, which seems to support the idea that the brain supervenes on the mind. Besides the lesion studies of Broca, th ere are other historical examples which seem to evidence that mental states supervene upon brain states. In the mid 19th century a railroad worker named Phineas Gage suffered an injury to the frontal lobe of his brain, due to an explosion that shot a tamping ir on through his skull. He survived, but the damage caused dramatic changes in his personality. He became irresponsible, untrustworthy, and impatient of suggestion when it conflicted with his desires (Damasio, 2004). This was one of the first instances in neuropsychology that suggested a functional localization of emotion and personality in the frontal lobes (Kandel, 2000). Seemingly this would allow one to make the claim that decision making exists in the frontal lobes; however, this example did not prove that this change could only be brought about from damage to this specific ar ea. In other words, it did not achieve a derivation reduction which might say something like, If personality is neural firing in the frontal lobe, then a change in neural firing in the frontal lo be is a change in personality. It merely shows that there is an association. Important to note is that th ese epiphenomena are impotent in the sense that they have no backwards causal effect on physical brain states. Physical brai n states occur, and the things we call mental st ates result. One useful me taphor for epiphenomena are as little sparkles of shimmeri ng light that occur on the wr inkled surface of the brain, sparkles which are caused to occur by physic al activity in the brain, but which have no causal effects on the brain in return.(Churchland, 1988, p.12).
26 While epiphenomalism says that our comm on understanding of volition is false, it does not attempt to eliminate it. It merely says that our experience of having brought about some action is merely the byproduct of the physical brai n state that has actually brought the action about. The ex act latency from brain stat e to mental state is not discussed. They occur close enough together th at we tend to conflate them as the same, but are not necessarily on e qual footing ontologically. Interactionism Interactionist property dualism, or interactionism says that there does exist a causal relation between mental properties and the brain (Eccles, 1974). One has a cognition, which causes a bodily action, and th en the bodily action in turn acts on the environment. The environment acts on the body through sensation, which affects cognition in turn. Interactionism still says that mental st ates are properties of brain states. Interactionists would say that mental states are materially constrained by the existence of a complex nervous system (Churchland, 1988), rather than the body, as such. An analogy might be how the property of red is constrained by the existence of an object, to call red. There is nothing necessarily that is red in the object, but redness arises from the object. This arising is an important concept in mind-body debates; it is called emergence In emergence, one thing arises out of another, seemingly without cause. The emerging thing is more complex and novel than its constituent can produce, within the understanding of its functional c onstraints. In the case of in teractionism, mental states emerge out of brain states, and also superv ene on them. In other words, there is no change in mental state that does not entail a ch ange in brain state, but that mental states
27 cannot be explained in terms of the change in brain state, because there is nothing about brain states that should account for th e emergence of the mental state. Because interactionism accepts the emerge nce of mental states, it does not try to reduce them. It also does not consider mi nd states supervenient upon brain states. A change in brain states does not necessitate a change in mental state, because mental states exert a causal effect, themselves, upon brain st ates. This is why interactionism is a dualist position, and hence falls under the same general criticisms. If a mental phenomenon is emergent, yet cannot be said to be explainable in terms of that which it emerges from, i.e. the brain, how can it exert influence over it? In 1974, John Eccles and Karl Popper, wrote The Self and Its Brain a controversial discussion of dua lism, reduction, and interactionism. They attempted to reconceptualize the problem from that of brain and mind by splitting all extant phenomena into three worlds. These worlds are considered emergent, and are called cosmic evolutionary stages (Table 2). In their book, Poppe r and Eccles try to explain the way the content of the three worlds interact with each other, and address problems arising from causation and supervenience (Eccles, Popper, 1974).
28 Table 2 Summary of Eccles and Poppers Three Worlds (Eccles, Popper, 1974) Name of World Constituent Entities The Products of the Human Mind (III) Works of Art and of Science, Technology Human Language. Theories of Self and Death. Subjective Experiences (II) Cons ciousness of Self and Death Sentience (Animal Consciousness) Physical Objects ( I) Living Organisms The Heavier Elements; Liquids, Crystals Hydrogen and Helium I find it difficult to argue that physical objects, and those thin gs listed under the products of the human mind do not exist. It w ould also take a stark solipsism to say that language does not exist. Wh ere the confusion arises is about World II: subjective experiences. Why they are they categorized th e way they are, and how exactly are they supposed to emerge from World I? Interactive Emergence There are those who believe that the me ntal is irreducible, but not emergent (Churchland, 1988). They believe that mental properties exist in terms of their own independent operating principles, analogous to laws like electromagnetism. Mental properties adhere to them, but cannot be reduced to them. Rather, the laws used as a way of explaining them. For example, light is not electromagnetism, but electromagnetism can explain what light is, and how it works. To come up with an under pinning law of the mental would allow for mental phenomena to be described, on their own terms. But when it comes to making predictions, there must be a way of linking th e mental law to th e physical laws. There
29 would need to be a means of interaction. Wh ether this linkage must adhere to purely naturalistic terms is unclear, because no such psycho-physical inte raction has yet been described. But assuming it was possible to come up with some fundamental psycho-physical law, accounting for phenomena would still be difficult, because in interactionism the interaction may take place in a bi-directional way. For example, Jane may feel a cascade of hormones and neurotransmitters and think, I am in love, or, she may think I am in love, and feel the same cascade. How is one to know which comes first, in the causal chain? While it may not be a result of assume d interactionism, this type of problem in causal directionality has been en countered before in psychology. Within debates about emotion, there were several predominant theories. William James and Carl Lange challenged the layman s understanding of em otion by saying that the environmental stimulus causes a bodily re sponse, and as a resu lt, the bodily response causes emotions. This theory was later challenged in the 1920s by the Cannon-Bard theory, which reversed the second step in th e causal chain, saying that we first encounter an environmental stimulus, have an emotion, and then act upon that emotion, resulting in physiological states. The problem with the Cannon-Bard theory was that it did not provide a means of interaction. Schachter a nd Singers two factor theory of em otions attempted to describe a mechanism. They said that the way we act upon an emotion is through cognitive evaluation. When I say, I am afraid what I mean is that my body has reacted to a stimulus -a large animal, for instance. I pe rceive this physiologi cal reaction, and then cognitively evaluate it, leading to me evaluating the reaction as fear. Where the CannonBard theory posits a one-to-one relations hip between an emotion and a physiological
30 reaction, the two factor theo ry leaves room for things like memory and context. All of these theories have explanator y power, but they cannot solve the problem of directionality. If one considers the idea that organisms act upon their environment through their physiological and emotional states then it becomes difficult to see where the causal chain begins. Although the addition of cognitive evaluation provides an extra step that, hypothetically, could fill the place of World II subjective experience, it cannot pinpoint how exactly emotion occu rs. It could be the case that cognition occurs first, and that the physiological reaction follows; or, th e emotion could occur first, and then the cognition. Because saying there are psycho-phys ical laws does not allow for emergence, it cannot reduce one part of the cau sal chain in terms of another. Emotion cannot be reduced to either cognition or physiology. Argument from Knowledge While dualism faces several de facto problems in explanation, such as whether the mind is a property or substance, or whether it arises or s upervenes upon brain states, or whether the mind and the brain interact, there are several compelling arguments for the irreducibility of mental st ates that, while controversial, are still considered in contemporary mind-body debates. The argument from knowledge says that it is possible to know some aspect of reality which is not grounded in physical prope rties. The typical example runs as such: Mary is a scientist. Mary knows every physical fact about the process of seeing: sensory transduction, color recognition, motion and feat ure detection, etc. She can describe, in every detail, the physiological process of th e photon hitting the corn ea, traveling through the retina, being transduced and coded, as we ll as how this information travels along the visual pathways to the visual co rtex, where it gets integrated.
31 For all intents and proposes, Mary is an expert in vision. The problem is that Mary has lived her entire life in a black and white room. One day Mary decides to leave her room; she sees a bed of red roses. After seeing these roses, would Mary have learned anything new about vision? The argument from knowledge says that she does, and that it is the experience of seeing a rose. (Churchland, 1988) Another name for the knowledge Mary gains is quale Quale is the singular form of quaila, wh ich is philosophical jargon for subjective experiences (Churchland, 1986). Qualia are the certain unmistakable phenomenological character the qualitative character of experiences, sensations, feelings, and so forth, to which we have introspective access. (Churchland, 1986, p.327). Most people agree that qualia exist (Churchland, 1988), but there are several problems that arise. For example, which mental states have qualia attr ibuted to them? It seems possible that those in altered states of consciousness, such as the comatose, might not. Do animals have qualia? Are qualia universal, in that they adhere to some predictive law, relating them to the physical world? Finally, ar e qualia intrinsic characteri stics of those having them? (Tye, 2007). It is hard to deny that there exists a phenomenal character to the experience of seeing red. It intuitively makes sense. But is that experience of seei ng red explainable in terms of the physical facts that underlie th e sensation of seeing red, or is it truly emergent? If qualia are supervenient upon physical facts; that is, if qualia cannot be accounted for by physical facts alone, then it would seem that there is something about qualia that are separate from physical facts. If there are things that are separate from physical facts, then it seems that dualism is true. Whether one believes that qualia are supe rvenient properties of physical states, or
32 that they are emergent, or interactive, will dictate what type of dualism is appropriate. The problems about the characteristics of qua lia aforementioned may never be resolved, but whatever the case, the Mary the Scient ist example is a comp elling thought question which brings up a concern a bout the thoroughness of a purel y physicalist explanation. Reductive Materialism: The Identity Theory Monism encompasses a variety of theories that assume simply that the world is composed of only one substance. This doe s not necessarily entail physicalism. For instance, Berkeleys Idealism might also be call ed monist, in that it says that all things are mental. But the identity theo ry is physicalist, and it says th at all mental states are just physical states of the brain (Churchland, 1988). It says that for any type of mental state, there is a brain state type which is identical to it. While identity th eorists admit that as of yet, there does not exist a one-to-one taxonomy of mental to physical states, they are committed to the idea that there will be eventu ally. This is the sense in which identity theorists are also reductive. At the level of any given phenomena general anxiety, fo r instance if a neuroscientist could prove that certain physical brain states corresponded to the behavioral, mental, and all me ntally significant aspects of the symptoms, then you could say they are identical. In this sense, re ductive materialism acknowledges that as of yet that mental states supervene on brain states. If a brain state was show n to be identical to the mental state of anxiety, then there can be no change in anxiety (within the degrees of freedom of observation for the term) wit hout a change in the brain state. How such an identity claim would come about would require a broad reduction. There could be no in remainder explana tion. Methodologically, a derivative reduction would be required. Statements that allow you to build an explanation, from the level of T
33 to the level of T would need to be established. For in stance, if trying to reduce anxiety to a brain state, you might have a statement th at says, if for all anxious creatures, irritability is present, then irritability is a part of anxiety. You would need several such statements, connected by common terms, in order to make an identity statement understandable. But for anxiety, how could a successfully broad reduction occur in the face of such complexly defined psychological phenom ena? DSM-IV criteria for Generalized Anxiety Disorder include fatigue irritability, muscle tensi on, and disturbed sleep (DSMIV, 2000, p. 472). These are all complicated, in themselves, and would likely require many connecting principles to be explained in terms of th e physical. DSM criteria also includes subjective distress due to constant worry and difficulty controlling the worry as criteria for anxiety. While it seems feasible that after extensive controlled research, neuroscience could identify the brain states that corres pond to the observational characteristics of anxiety, it would be leaving a remainder of subjective distress. If an identity theorist chooses not to believe in the existence of qualia, then they would probably just remove subjective distress from a description of anxiety, because it would be assumed that a physical explanation is sufficient. Otherwise, an identity theorist must wait patiently, assuming that subjective experience has a physical explanation, and that it will be found some day. There is a view called non-reductive physicalism which says that all things are indeed physical, but that physical explanations will never be able to successfully reduce the mental. As opposed to being reduc tionist or anti-reductionist, non-reductive physicalism seems that for the sake of e quivocation, to aband on the problem. Non-
34 reductive physicalists are similar to identity th eorists, in that they would say that qualia are physical, but they would not attemp t any sort of derivative reductions. At face value, identity theory wit hout successful reductions seems to be no different from an assertion of materialism. It is a daunting task to reduce all mental phenomena, but for arguments sake, let us sa y that it may occur. For anxiety, say that someday all of the criteria, including some connecting principle to explain subjective distress, become explainable in terms of th e physical. At that point you could say anxiety is this set of physical states. With a derivative reduction, from this poi nt on it would just be understood that anxiety is both those things desc ribed in the DSM, as well as th is set of physical states. It could be considered an alternate set of explan ations, especially if in some cases it is linguistically more convenient to use the ps ychological explanati on. A psychiatrist may know that anxiety is th is set of physical states, but for the convenience of his patient, he calls it fatigue, irritability, muscle tension, and disturbed sleep, etc. Although I have my doubts as to how feas ible such a reduction is, especially as regards mental illness, it is not my largest concern. My concern is going a step further and making a derivative reduction into an elimin ative one. If a psychiatrist decided that the physical explanation of a nxiety could account for the beha vior of his patients better than the psychological explanation, then he might decide to eliminate the psychological explanation altogether. That would be after the fact. Eliminative-materialists make a consorted effort to eliminate T in terms of T The reductionist agenda is not just to show that T is explainable in terms of T but that the explanation in terms of T is a better explanation in the first place. There are othe rs who have taken a more extreme stance,
35 who have advocated the elimination of ment al explanations before any reduction could even take place. Eliminativism Eliminativism, it should be noted, is not the predominant stance in philosophy of mind. Even internally, it is considered radical (Ramsey, 2007). Eliminativism shares the claim of the identity theorists, saying that a ll mental states are brai n states, but precludes that an attempted reduction of th e mental to the physical will fail. This is because of the conceptual vagueness of the mental, or ps ychological, terms. This vagueness and ineptitude render eliminativis ts to label certain mental terms as folk psychology (Churchland, 1988; 1989). According to eliminativists like Paul and Patricia Churchland and Daniel Dennett, folk psychology makes false predictions, and ge nerally misleads the masses in regards to the nature of human behavior and cognition (Churchland, 1988). If scientists attempted to reduce folk psychology, they would have extreme difficulty because the gap in explanation would be due to th e inaccuracy of the terms in T rather than some innate metaphysical division. (Churchland, 1988). Elimin ativists also include qualia as folk psychology, and they say that any ineffability attr ibuted to it is merely because it is such a vague term, with little explanatory value (Churchland, 1988; Dennett, 1995). Eliminativists feel that the vocabular y of folk psychology should be eliminated, prior to any type of attempte d reduction -that a completed neuroscience would reveal them in their failure. Some examples of fa iled folk psychology include the ideas of sleep, memory, mental illness, and desire (Church land, 1989). Where previously the folk description of sleep might have said that sleep was equal, in that each hour contributes to a recovery of function; neuroscience has prove n that it is primarily REM sleep and slow
36 wave sleep which are necessary for higher cognitive functioning (Born, Rasch & Gais, 2006). In this example, the common sense understanding of sleep is inaccurate. Some eliminativists claim that our folk psychol ogical explanations have not changed since antiquity, in that they are equally us eless in explanation (Churchland, 1989). Eliminativists advocate a purely bottom-up approach, ut ilizing the most reliable and fundamental levels of explanation, which typically include physics and chemistry. Because eliminativism advocates throwing out psychological terms, something like anxiety would not be studied. Instead, rese arch would be conducted via something like neurocomputation, which examines the way in which brains process information, by means of neural networks (Churchland, 1989). The relevant data would involve action potentials, chemical cascades, transduction patterns, and the lik e (Churchland, 1989). At some point if commonalities were shown, you could say, groups of people show decreased activation in the locu s coeruleus, bu t you would not call it anxiety. Eliminativists feel that even attempting a reduction is giving too much credence to folk psychology -that the hypothetical T already retains more explanatory power and accuracy than folk psychology could hope for. It assumes that a purely physiological account is already more apt to explain behavior than folk psychology is. Richard Rorty and Paul Fe yerabend (both canonical propon ents of eliminativism) speak on the deprecation of psychological terms. They both agree that if it were not a linguistic inconvenience, as it stands, psychol ogical terms could be dropped with no real harm (Rorty, 1965; Feyerabend, 1963). In fact getting rid of them now would lessen the ambiguity inherent to the folk psychologi cal, in favor of th e more objective explanations; namely, the neurobiological ones They recognize that this is unlikely to happen, but advocate it all the same:
37 [Rorty] If the Identity theo rist is taken to be predic ting that some day sensation, pain, mental image, and the like will drop out of our voca bulary, he is almost certainly wrong. But if he is saying that, at no greater cost than an inconvenient linguistic reform, we could drop such terms, he is entire ly justified If we knew enough physiology, we could forget about commonsense ways of accounting for the behavior of our friends and rela tions common sense psychological theory is a dispensable crutch [Fe yerabend] There are enough independent predictions available, many more predictions in fact than the mentalist could ever provide for physiology to provide its own explanatory system (Rorty, 1965, p.198; Feyerabend, 1963, p.140) Eliminativism would accept an eliminative -reduction, should such a completed neuroscience ever arise, but does not requir e it for the deprecation of folk psychology. Its contentions with reduction make some assu mptions, most of them linguistically based. Primarily, it makes the natura listic assumption: that anyt hing worth describing should adhere to the inventory employed by the physic al sciences. It says that in order for psychological phenomena to be scientifically respectable, or ontologically legitimate, they must be reducible to fact s of the natural scie nces (Horst, 2007). This is ironic, considering that eliminativists also believe such a reduction, as it stands, to be impossible. In effect, th at is, until such reductions have occurred, eliminativism may just be a fo rm of identity theory that is pessimistic about psychology. As both the identity theory and eliminativism ar e central concerns of this thesis, they will be discussed in more detail in the following chapter. But I would like to review what some consider an answer to reductionism, in general.
38 Functionalism Sometime in the 1960s identity theory was usurped by functionalism, as the predominant and popular mind-body stan ce (Churchland, 2004) Functionalism approaches things not in terms of properti es, but in terms of functions: input, output relations. For example, functionalism does not see a heart as a collection of smooth muscle, or the center of emotion. It is a he art by means because of its function, which is to pump blood. One of the interesting things that functionalism allowed for is called multiple realizability. Multiple realiz ability says that for any given function there may be an infinite number of physical substrates in wh ich to instantiate it. For instance, a river might be considered in terms of corrosion, vi scocity, volume, and velocity, and a function could be derived which could predict values of one, in terms of another. This function could be instantiated in just about any liquid medium; also, it could be programmed into a computer. According to functionalism the central goal of Artificial Intelligence is to do this: to create novel realizations of the abstract functions that exist in the world, but primarily those that we, as humans, are computing (Churchland, 2004). Functionalism treats folk psychology as already embodying a crude or ba sic investigation in to these functions. They provide good additive knowledge. For instance, within cognitive psychology there is a model for working memory, put forth by Baddely. It onl y employs other psychological terms, with the assumption that there exists some underlying mechanism th at makes the model function, but that it is up to the realm of neurobiology to discover it. A schematic of Baddeleys model is
39 represented below Figure 3 Baddeleys Model of Working Memory. Each of the items in the above model fu lfills general tasks. For instance, the central executives job is to govern planni ng or decision making, trouble shooting in situations in which the auto matic processes appear to be running into difficulty, novel situations, dangerous or tec hnically difficult situations, and situations where strong habitual responses or temptations are invol ved (Goldstein, 2007). Though this is an extremely broad model, it is testable. Because it attempts to explain genera l functions, cognitive psychology is given a special status. Functionalism admits it is a worthy endeavor to investigate the idiosyncratic ways in which the cognitive functions are represented in our brain architecture, but that to reduce psychologica l kind-terms to brain terms would be to exclude the multiply realizable ways that they can be represented. Cognitive psychology should be left alone, because it can contribute to the global function which can only be revealed in the situated molar-level beha vior of an entire cr eature (Churchland, 2005). Since its inception, functionalism has remained a popular stance in the mind-body
40 question, but its goal of inst antiating the human global func tion in a computer has run into problems. For instance, the discovery of parallel distributed processing evidenced limitations in computer modeling of cognitive functions. Parallel distributed pr ocessing, or PDP, was first introduced by James Mclleland in 1986. It was the forerunner of what is called connectionism, which sees mental processes as being composed of interconnected neural networks, composed of simple units (neur ons) (Kandel, 2000; McCl elland & Rumelhart, 1986). It was discovered that de spite the relative slowness of neural communication (no greater than 120m/sec), compared with electroni c communication (around 108 m/sec), computing things such as language recognition required tremendous computational power (Churchland, 2005). Whereas computer s process functions serially, the brain works in parallel, breaking up sense data in to constituent aspects, and then rebuilding them at the level of the percept (McClelland & Rumelhart, 1986). For this reason, the modeling of the gl obal organismic function (the goal of Strong AI) has been presumed to fail. It woul d also seem to indicate that functionalisms presumption that functions are multiply realiz able is false, but Paul Churchland argues that the cognitive terms themselves are the cause of these problems, not the computers. If considered neurocomputationally, all syna ptic connections can be thought of as components of the massive system of parallel distributed processing. It is at the level of synaptic connections that functions should be derived, according to Churchland, not from cognitive psychology. He considers the claim that cognitive phenomena are multiply realizable true, but he does not think that this detr acts from its ability to be reduced. He admits that such a claim seems counter-intuitive, because this would imply that cognitive phenomena can be
41 reduced to levels which adhere to different physical laws, but he thinks this is not so problematic (Churchland, 2005). He give s the example of temperature. Temperature is a molar-level phenomenon, but it can be realized in multiple ways. It can be seen in the Earths atmosphere, or in a gas of any molecular constitution. This is because temperature is identical with, or reducible to mean kinetic energy of molecules that make up a gas (Churchland, 2005). By this definition, temperature adheres to the underlying laws of Newtonian physics. Laws of motion are the same for every type of molecule that could hypotheti cally be involved. Ev en the idiosyncratic details such as the velocity of dispersion of a gas will be reducible to laws of classical thermodynamics, e. g. The Ideal Gas Law, wh ich in turn are reduc ible to classical mechanics (Churchland, 2005). What, at face value, seems to necessitate an appeal to different laws at each level of explanation turns out to be untrue. Churchland says that high-level regularities across diverse material substrates will typically find a successful reduc tion to underlying and general physical laws laws that are simply blind to the id iosyncratic and irrelevant differences that happen to distinguish the several substrates. (Churchland, 2005, p. 15). But the problem with this statement is that in our particular substrate, there exists a significant idiosyncrasy, which is that of qualia A reduction of this type would need to exclude qualia as inherent to any given function, and would inherently leave it as a remainder. Summary In this review I have introduced severa l concepts important to the discussion of reduction. These include supervenience, em ergence, qualia, elimination, neural dependence, and causal interaction. At the be ginning of this chap ter, I posited that
42 different mind-body assumptions can affect the way neuroscience goes about research. Having reviewed, generally, the ontologi cal components of predominant mind-body theories, how might the adoption of any one of them affect a given experiment? Take decision making as an example. A substance dualist might pursue the explanation under the assumption that th e phenomenon of making a decision is constrained by neural activit y, but that neural activ ity and decision making are fundamentally separate. In this case, there would be no reductionist agenda; the neurophysiology could be considered on its own terms, perhaps in trying to examine the way in which some type of decision making occurs, through game theory or psychology. Once understood on the neurophysiological level, there would be no attempt to eliminate decision making because it is supposed to be non-physical. An epiphenomenalist would approach th e method of research similarly, also without a reductionist agenda. Because it would consider de cisions as a property, or an emergence from neural activity, there is not much use in trying to correlate the specific types of mental states which co-occur with br ain states. It is simply assumed that when some neural activity occurs, a decision is made Statements about correlations to brain states would be unnecessary, because decision making is considered a facet of a brain state, occurring at the same time, but not linked to it in any causal way. An interactionist would do the opposite. He or she would attempt to establish the specific contingencies in which a change in me ntal state entails a change in brain state, and vice versa. Experimentally, this would probably take place on the instance, or token, level. For instance, when faced with a choi ce to flee or fight a large animal, the brain receives sense information. This sense info rmation in some way interacts with the mind, which makes a decision. This decision affects, and effects (brings about) another brain
43 state. While there may be perhaps an infinite number of ways describing the sequence of interaction, both the levels of description at the mind level and brain level are retained. For the sake of produci ng a holistic picture of what it means to make a decision to fight or flee, as well as the assumption that mental states cannot be reduced to brain states (because they are metaphysically separate), both the physical and the mental are given importance. At the theoretical level, a mode l of interaction could be formulated, and its generalizability could be tested in other token circumstances. A reductive materialist makes the assumption, somewhat akin to the epiphenomenalist, that mind stat es and brain states are a pa rt of the same phenomenon. The difference is that experimentally, a re ductive materialist woul d try to explain the mental state in terms of the brain states. It accepts th at at this point, nominally, the two are considered separate. In other words, reductive materialism accepts the prima facie supervenience of psychology on neuroscien ce. A phenomenon like decision making would be approached in tryi ng to show which brain states, and which brain states only constitute the observable phenomenon of a decision. Connecting principles may be used, but at the end of the day, a reductive materialist s goal would be a logical sentence that has enough predictive value to reduce decision making to brain states. A non-reduc tive materialist would believe, because of technological or epistemological reasons, that this is impossibl e. He or she would believe that the phenomenon of decision making is, in a sense, a brain stat e (believing that the world is composed only of physical substan ces), but would not at tempt a derivative or eliminative reduction. An eliminativist would accept an elim inative reduction, and say that after
44 decision making has been effectively redu ced, it should not be used as a concept, because the reducing explanation is a better ex planation than the psychological theory of decision making itself. Because the psychol ogical construct of d ecision making may be predicated on the idea that there is some intention or volition, both of which eliminativists consider myths (Churchland, 1986; 1988; 1989; Dennet, 1987), the effort to reduce would be considered wasted time. Eliminativis m says that there is nothing substantive in folk psychology worth reducing, but it would not refute a sound eliminative-reduction. Finally, a functionalist w ould consider the function of making a decision. Because of multiple realizability, a reduction of the psychological description of decision making to the neurobiological one would be unnecessary. Functionalism says that the primary function of: <
45 Table 3 Mind-Body Theories and Th eir Reductionist Stands Name of Theory Attitude Towards Relation Attitude towards Reduction Substance Dualism Mind and Body are separate substances; the mind is constrained by the body, and causes physical states. Mental states are irreducible to physical explanation. Epiphenomenalism Mind states are properties of brain states occurring with, and above them. Mental properties are supervenient upon physical states. There is no causal relation between them. Interactionism Mind states are properties of brain states; they are ontologically on equal footing. Mental properties are irreducible to physical states. Mind states have effects on brain states, and vice versa. Reductive Materialism/Identity Theory Mind states are brain states. Eventually all mind states will be shown to be the same as brain states, utilizing a derivational and broad reduction. Eliminativism Mind states do not exist. An attempt to reduce is unnecessary, because folk psychology is de facto misleading. The next chapter will descri be some of the subtleties of eliminative reduction, as well as argue for a non-reductive stance. To relate eliminativism to psychology, questions about mental illness will be use d. Arguments will be framed from both the methodological and epistemological perspective.
46 CHAPTER III: REDUCING THE PERSONAL REALITY From 1.5 kilograms of flaccid matter, c onvoluted folds, about 100 billion neuronal components, hundreds of trillions of inte rconnections, many thousand kilometers of cabling, and a short cultural history emerge d calculus, Swan Lake, Kind of Blue, the Macintosh, and The Master and Margarita -Kristof Koch Outline In this chapter I will discuss some of the problems with eliminative-reduction, focusing on two examples: mental illness, from the lens of psychiat ry, and brain imaging, from the lens of clinical neuroscience. I wi ll discuss concerns about identity theorists reduction of psychological constructs, and address methodological and epistemological criticisms to eliminativism and eliminative-reduction. I am not making the assumption that every neuroscientist is a reductive materialist or an eliminativist. I raised a c oncern in the first chapter about the prevalence of individuals attempting to reduce psychol ogy to neurobiology, but any neuroscientist, any psychiatrist, any psychologi st may maintain a reductive materialist agenda. They also may be nonreductive materialists, or dualist s. In any case, I believe that scientists method to treatment, or research, will be affected by their philo sophical orientation. One approach to explaining depression will be examined, in the context of identity theory. After describing this approach, I will bring up several arguments, on the token level (that is, against this specific example), which I believe refute an identity claim. I will next move on to clinical ne uroscience, which exemplifies an eliminative-
47 approach to mental illness and the brain. In th is case I will bring up an argument on the methodological level, which limits the ability of clinical neuroscience to make identity claims, or pursue true eliminative reductions. Finally, I will address e liminativism. I will introduce a concept which cuts to the heart of reductionism in neuroscience, and address the most salient remainder, which eliminativist criticism cannot account for. Biological Psychiatry Psychiatry poses an interesting example from which to discuss identity theory. Being trained in both psychol ogy and medicine, psychiatrist s field seems to entail a conflict between the psychological and the biol ogical, in explaining mental illness. To address this conflict, the biopsychosocial m odel of mental illness was posited by George Engel, in 1977. It was offered as an integra tive approach to mental illness, quelling the problem of conflict between psychological and biological ex planations (DeGrury, 2003). It has been called a more comprehensive way to think about clinical conditions (DeGrury, 2003 p.84): When one considers the biological, psyc hological, and social dimensions of a medical condition, mental problems become connected to physical problems etiologically, consequentially, or by some less direct association like overlapping areas of a Venn diagram. (DeGrury, 2003, p.85). Do the dimensions of the Venn diagram overlap neatly? For those who claim to adhere to the biopsychosocial model, is it possible that one area has become more emphasized? There is no conclusive answer. There has l ong been a divide within psychiatry between those who emphasize the psychological, a nd those who emphasize the biological (Ghaemi, 2003; Luhrmann, 2000); specifically, in the time of Engel, the divide was between the medical model and psychoanalysi s (Ghaemi, 2003). The medical model, or
48 objective-descriptive mode l, involves the following: observing symptoms, relating them to syndromes, diagnosing a disease, and finally treating that di sease with medicine (Ghaemi, 2003). The psychological empha sizes the use of psychotherapy, and emphasizes psycho-social factors in tr eating mental illness (Ghaemi, 2003). While the biopyschosocial model exists t oday as a means of integrating all the components thought important to mental illne ss, Nassir Ghaemi, a psychiatrist from Harvard, says in his book The Philosophical Founda tions of Psychiatry that the biopsychosocial model is deceptive. He believes th at it is used to mask a lack of critical, philosophical introspection. Particularly, he says, psychiatrists us e the biopsychosocial model as a means to avoid discussions of the mind-brain problem (Ghaemi, 2003). He believes the biopsychosocial mode l is professed in place of a veiled dogmatism. Psychiatrists are dogmatic either to the psychological, or the medical model. Ghaemi also believes that the biopsychosoc ial model is also made synonymous with eclecticism. By eclecticism, Ghaemi means the view that psychi atrists choose to be agnostic to the mind-body problem, or deny that they have one single view about how to explain mental illness (Ghaemi, 2003). Eclectic s eschew any particular model of mental illness, but, in practice, do in fact have a veiled dogmatism towards the biological or the psychological (Ghaemi, 2003). Dogmatists buy into a mistaken monism (Ghaemi, 2003), in that they approach their patients with the understanding that one approach is sufficient to explain the patients mental health problems. Because Ghaemi believes that most psychiatrists who espouse the biopsychosocial model are in fact dogm atists of one sort or another, he says that the biopsychosocial mode l is flawed (Ghaemi, 2003). Though they might claim to not be dogma tists, Ghaemi thinks that they are not
49 being honest with themselves about their unde rlying views. Worse, they may just be doing what intuitively seems right for any given patient: making diagnoses, increasing drug doses, and offering etiol ogies that may not be base d in anything except personal experience (Ghaemi, 2003). Ghaemi particularly notes the biopsychos ocial model for its vagueness. It merely says that when treating mental illness, you acknowledge biological, cognitive, and interpersonal variables (Engel, 1977). It makes no conjecture as to how these variables interrelate. This leaves mu ch room for interpretation (G haemi, 2003). As an eclectic, a psychiatrist may condemn a reductive, monis tic form of explanation, but in practice will be enacting implicit mind-body assumptions. Because of its vagueness, the biopsychosocial model allows this mistaken monism to occur, under the attractive auspice of holism. Considering the fact that many psychiatrists only meet with their patients for 15 or 20 minutes to conduct med-checks, or receive updates about major life events (Ghaemi, 2003), this is worrying. If a psychi atrist is making a faulty causal attribution, then a patient runs the risk of not being trea ted effectively, or possi bly even being harmed by the long term side-effects of medications. Patients concerned about the effects of medications, particularly new ones that may not have comp letely understood mechanisms may decide to see a psychotherapist, or a ps ychiatrist less oriented towards the medical model. That is not to say that patients have any better understandi ng of the nosology of mental illness than psychiatrists. Psychiatrists are trained in both medicine and psychology for the purposes of providing mo re effective treatment (Ghaemi, 2003); therefore, their professional opinion has more sway than a patients. But patients do
50 retain some understanding of mental illness, which I believe psychiatrists have limited access to. They have their own subjective experience of being mentally ill. For those psychiatrists who adhere to the medical model, the idea of subjective experience may seem irrelevant. They are likel y to be reductive materialists (Insel & Quirion, 2005, Glannon, 2003; Ghaemi, 2003; Kandel, 2000), and because reductive materialists believe that mental states are not hing more than brain states, an appeal to subjective experience ha s little explanatory value. But what is the matter? Why should a psychiatrists philosophical position exert any effect on their success as healers? There was a time when subjective experi ence was acknowledged as important for treatment. Originally the DS M-II incorporated subjective ac counts of mental illness as a part of the diagnostic process. At the publishing of the DS M-III though, subjective accounts were removed and replaced with obj ective signs and symptoms, which could be verified through inter-rater reliability and generalized to populations (Flanagan, Davidson & Strauss, 2007). The DSM-IV-TR now bases its diagnostic cr iteria along five axes: major clinical disorder, personality disorder, medical illn ess, environmental contributors, and global assessment, ranging from 0-100 (as rated by the psychiatrist) (DSM-IV-TR, 2000). Treatments are guided by these criteria (DSM-IV-TR). Assumedly, the objective accounts of mental illnesses would account fo r the subjective experi ences of those who are mentally ill, but there have been studies that have shown that the DSM criteria for certain mental illnesses do not coin cide with individuals experience. For instance, a criterion for borderline personality disorder (BPD) is identity disturbance: markedly and persistently unsta ble self-image or sense of self (DSM-IVTR, p.710). Miller (1994) conducted a study in which he interviewed those diagnosed
51 with BPD, rather than reporting a loss of identity, could cogently and consistently describe who they were, and could speak about times when they felt as though they could not be themselves (Miller, 1994). Patients reported feeling estranged and inadequate, when they compared themselves to others (Miller, 1994). Miller clar ifies, saying, rather than having an impaired sense of self, they seem to have a sense of themselves as impaired (p. 1216). It would seem that integrating subjecti ve experiences into diagnostic criteria would be a prudent goal for the next issue of the DSM. Despite the belief that patients do not have insight into their own illnesses (Flanagan, Davidson & Strauss, 2007), those insights may be necessary, in order to formul ate a truly holistic understanding of mental illness (Flanagan, Davidson & Strauss, 2007). Particularly, if patients subjective experiences are in conflict w ith the objective account, th en psychiatrists treatment strategy may run the risk of be ing unnecessarily deleterious. Whether the inclusion of subjective experience necessitate s a dualistic mind-body stance is an open question. But the manner in which reductive materialism instantiates itself in the practices of psyc hiatry is worth exploration. Psychiatrists who are reductive materialists would say that mental illness is an imbalance of neurotransmitters in the body (Insel & Quirion, 2005, Glannon, 2003; Gh aemi, 2003; Kandel, 2000). Because mental illness is a disease, and diseases are treated with medicine, it seems to make sense to say that mental illness is a dysfunction of the brain. But the extent to which this position is feasible should be considered. To do this, I will examine a commonly utilized explanation for mental illness, and use it as a token example from which to argue against reductive materialism, e .g. the identity theory.
52 Depression and the Serotonin Hypothesis The phrase mental illness carries with it a certain linguistic suggestion (Kendel, 2005). An illness is something that someone catches. It is caused by some pathogen, and is an abstract term to refer to the eff ects that pathogen has on the body. When speaking about mental illness, is the substrate of illness actually mental? Is mental illness literally a disease of the mind? (Kendel, 2005). How someone answers this question is central to mind-body assumptions. And when these assumptions are not made cl ear by those who are commissioned to treat mental illness, I believe laymen may make the conclusion that mental illness is isomorphic to any other type physical disease. Because mood disorders are some of most common forms of mental illn ess (DSM-IV-TR, 2000), one type will be used as an example, wherein confusions about physic al causes may lead to identity claims. Major Depressive Disorder (MDD) has been associated as a deficiency of serotonin levels. This is referred to as the serotonin hypothesis. (Lacasse & Leo, 2005). The serotonin hypothesis is often relied on h eavily in direct-to-consumer advertising campaigns, which purport that selective serotonin reuptake i nhibitors (SSRIs) can correct a chemical imbalance due to a lack of serotonin in the brain. Though these advertisements imply that this claim is a proven fact, identifying depression with serotonin is incongruent with scientific eviden ce, which has not, as of yet, to make such determinations (Lacasse & Leo, 2005). I will review some reasons why an identity claim between serotonin dysfunction and depr ession is, as of yet, impossible. The term imbalance implies that an e quilibrium of serotonin levels for nondepressive functioning has been established. This is a false claim, both because of methodological limitations in assaying brain ch emistry, and the fact that no such baseline
53 has yet been conceived (Lacasse & Leo, 2005). As regards neurochemical assays, direct analysis of neurotransmitters in the brain is possible but requires a sample of brain tissue. This is unrealistic for live populations (Kellermen, 2004). Other resources for transmitter collection include taking cerebrospinal fluid (CSF) samples and urinalysis. CSF runs throughout th e ventricles in the brain, as well as the spinal cord. To take a sample, the lumbar spin e must be incised what is called a spinal tap. Spinal taps run the risk of permanent dama ge to the spine or sp inal cord, (Kellermen, 2004); though they may be necessary in some cases for the detection of certain biomarkers, lie after acute injuries, they are not preferable for the detection of MDD. Urinalysis is an indirect way of meas uring neurotransmitter levels, and yields counts which include both neurotransmitters in the central nervous system, and the peripheral nervous system (Kellermen, 2004). In the central nervous system, along with its hypothesized effects on mood, serotonin is also associated with the regulation of appetite, sleep, and muscle contraction (K andel, 2000). Also, roughly 80% of the bodys serotonin is located in the gastrointestinal tract. A urinalysis for serotonin would not be able to differentiate between amounts fr om the CNS or the PNS. Although decreased appetite and insomnia are common in M DD (DSM-IV, 2000), it is not clear whether depression is caused by a decrease of serotonin levels in the periphery or in the brain. SSRIs work to block reuptake of serot onin by selectively blocking receptors that collect quanta of serotonin from the post-synapt ic cleft, back into the presynaptic cell. Increasing the extracellular level of serotoni n in the extracellular space between the presynaptic and postsynaptic cell increases the likelihood of chemodynamic binding to postsynaptic receptors (Kandel, 2000). This is th ought to exacerbate excitatory effects, in serotonergic pathways. And hypothetically, beca use of serotonins role as an agonist,
54 more activation in serotinergic pa thways leads to improved mood. Experimental data have shown that the degree of dosage used for many SSRIs may not be high enough to produce the effect s they claim (Mendels, Stinnett, Burns, Frazer, 1975; Beulig, 2008). Furthermore, the multi-purpose nature of serotonin as a neurotransmitter makes it unlikely to be causally responsible for MDD (Lacasse & Leo, 2005). Some argue that the fact that some have experienced reduced depressive symptoms from taking SSRIs provides eviden ce for the serotonin hypothesis, but such reasoning is invalid, and akin to saying that because aspirin relieves headaches, headaches are caused by a deficit of aspi rin in the brain (L acasse & Leo, 2005). It may actually be the ca se that SSRIs themselves induce a chemical imbalance of serotonin, due to the desensitization of autor eceptors in post-synaptic cells. This causes a down-regulation of the amount of serotonin which the post-synaptic cell would normally be producing, on its own. The existence of discontinuation syndrome also seems to support this idea. Although SSRIs are not thought of as addic ting, or as entailing nearly as high of a dependence-risk as benzodiazepin es (tranquilizers), some who attempt to stop taking SSRIs experience severe, withdr awal like symptoms (Broekhoven, Kan & Zitman, 2002). These are frequently thought by patients to be a recurrence of their original symptomology (Broekhoven, Kan & Zi tman, 2002), and may necessitate the use of SSRIs again, just to relieve the symp toms due to discontin uation (Broekhoven, Kan &, Zitman, 2002). Efficacy trials have also shown SSRIs to be equivocally effective compared to other means of relieving depressive sympto ms. Kirsch, Moore, Scoboria and Nicholls (2000), using the Freedom of Information Ac t, reviewed all cl inical trials of antidepressants submitted to the FDA for approval, and found that for both published and
55 unpublished studies, the placebo was able to dup licate 80% of the antidepressant effect (Kirsch, Moore, Scoboria, & Nicholls, 2000). There are also studies measuring the comparison of SSRIs with other medications, such as tricyclic antidepressants (TCAs), buproprion, and reboxetine. These drugs were f ound to be just as effective in treating depressive symptoms (Karvoussi, Segraves, Hughes, Ascher, 1995). TCAs block the reuptake of biogenic amine neurotransmitters, both serotonin and norepinephrine (noradrenalin). Norepinephr ine has its own, complicated pathways throughout the central nervous system, which i nnervate broad areas of the brain and act on an array of receptor type s (Kandel, 2000). The mechanism of SSRIs also cannot account for the relative success of electro-c onvulsive therapy (ECT), which has also shown to relieve depressive symptoms (Fi nk, 2001). ECT is a procedure which induces a controlled brain seizure (Fink, 2001). A ra ndomized, controlled study examined the effects of SSRIs combined with cognitive be havioral therapy to SSRIs alone, and found no significant difference between the two in relieving majorly depressed patients symptoms (Goodyer et al., 2008). Another issue is the latency between drug administration and symptom relief. Average time for effects to be shown with SSRIs is six weeks (Kirsch et al., 2000) while for those who experience relief from TCAs, onset is typically immediate (Kandel, 2000; Kirsch et al., 2000). Considering TCAs differe nt mechanism of action, as well as the fact that they act upon norepinephrine, in addition to serotonin, sh ared relief of depressive symptoms between TCAs and SSRIs seem to be due to different physiological processes in the brain. Although the data presented above are limited, they do demonstrate that the serotonin hypothesis is still a hypothesis for a reason. The efficacy trials that show SSRIs
56 to be equivocally effective at relieving depressive symptoms compared to other antidepressants and treatments seem to indicate that, though serotonin may be implicated in MDD, it is not the cause of it. If it were possible to esta blish a baseline normal le vel of serotonin, and it could be shown experimentally that all of these methods work towa rds restoring this baseline, then you could make a stronger claim. But, as of now, if a lack of serotonin cannot explain MDD without remainder, then one cannot make the claim that depression is a lack of serotonin in the brain. Because any eliminative reducti on is also a broad reduction, it is susceptible to the same principles. That being said, the different treatments for depression might be suitable for a potential derivation reduction. If connecting principles c ould be found, then a reduction could be accomplished; however, no conclu sive connecting prin ciples have been established. Perhaps, after investigation, form ulas will be derived that take into account all contingencies that go into MDD: things like gene tic predisposition, brain physiology, environmental factors, etc., but within th e bounds of broad reducti on, an identity claim between depression and serotonin dysfuncti on does not work. You could say that depression is supervenient on some sort of neurochemical imbalance, but as of yet, it cannot be proven that it is serotonin. Y ou cannot explain depression, in terms of serotonin. Many psychiatrists use anti-depressants in combination with psychotherapy to provide multiple approaches to treatment. Whether this eclectic approach is used due to a non-reductive mind-brain stance, or merely becau se of the fact that as of yet, antidepressants alone cannot treat depression, is difficult to as certain. But the patients who show improvements probably do not care about such philosophical meanderings.
57 Pragmatically speaking, if a medication shows positive effects, why should a psychiatrist hesitate to prescribe it? I have no contention with those who garn er positive effects from an eclectic approach; however, as outlined in the beginning of this section, the biopsychosocial creates significant space for am biguity. Within this ambiguit y, the possibility that there are psychiatrists operating under a veiled reductive materialis m brings up concerns about their prescription patterns and etiolo gical suggestions. I have attempted to show in this sect ion that the serotoni n hypothesis does not retain enough explanatory power to warra nt making a causal attribution between a serotonin imbalance and depression. This argument was primarily based upon the principles outlined in chapter I: that any broad reduction show that T accounts for all observable phenomena in T The long latency between onset of prescription of SSRIs and relief of depressive e ffects, as well as the evidence that treatments not using serotonin as their primary mechanism (such as ECT ) can achieve the same effects are both examples of why a causal attribution between serotonin and depression cannot be made, because of the large remainder. This calls into question potential overemphasis on psychopharmaceuticals, which a reductive materialist may fall victim to. Clinical Neuroscience and Constrai nts on Functional Localization I will now move in the more explicit di rection of neuroscience, and examine the ways in which neuroscientists could pursue eliminative reductions. In other words, h ow would a neuroscientist actually go about trying to reduce mental states to brain activity, and what are the problems with such a reduc tive endeavor? Besides the data gained from neuropsychology, which examines those with brain damage and makes conclusions about functional localization via negative means (a fter the fact), there is also a positive
58 (additive) endeavor to achiev e functional localization. This takes the form of brain imaging (Roth, Randolph, Koven & Isquith, 2006). We understand how sensory systems operate We can describe the way in which stimuli are transduced into el ectrical signals, and how these signals are in turn conducted along axons, via action potentials. We know that information about the external stimulus is encoded by the frequency of these action potentials (Kandel, 2000), and the mechanism by which these action potent ials are perpetuated. Myelin insulates axons, and allows for i onic gradients within axons to renew the electrical pulse via saltat ory conduction (Kandel, 2000). Fi nally, we understand the role that neurotransmitters play in the excitati on and inhibition of action potentials, which perpetuate diffusely throughout complicated neural networks composed of billions of cells. But is this knowledge -this explanation of diffuse neural activity -sufficient to explain mental illness? Just as there is a s ub-division within neuroscience called cognitive neuroscience, there is another, called clinical neuroscience, which concerns itself in explaining mental illness in terms of neural activation (Milner, Squire, Kandel, 1998). While many of the conclusions in cognitive neuroscience are drawn from brain imaging studies wherein participants will be put into an MRI machine, and given certain, highly specific stimulus-response tasks, previously as sociated with the c ognitive faculties in question (Milner, Squire & Ka ndel, 1998), clinical neurosci ence investigates clinical populations and conditions. These can range from mental illness, to brain tumors, to the perception of pain (Journal of Clinical Neuroscience, 2010). Findings from cognitive neuroscience are drawn upon, when constructing study protocols, but stimulus-response tasks may or may
59 not be utilized. The Journal of Clinical Ne uroscience Research cites its purpose and content (JCNR, 2010, p.1): clinical investigations which furthe r our knowledge of cognition, mood, behavior and motor function of both normal and a bnormal brain function. It will include articles on the cellular, neurophysiologi cal, and molecular functioning of the central nervous system. From the above description, one of the aims of clinical neuroscience includes comparing clinical populations to nonclinical populations, for the purposes of investigating the neural functions associated with particular disorders. While it is concerned with explanations at severa l levels, it is the neurophysiologi cal which I am concerned with. Previously these types of investig ations could only be gained through neuropsychology, which makes conclusions about brain function by looking at deficits, after damage, or after the f act. The advent of functional magnetic resonance imaging (fMRI) has made investigations into the f unctioning of the brain more accurate, because it can collect data about brain activity closer to its occurren ce, and does not require brain damage (Roth et al., 2006). I am taking brain imaging as an exampl e of an already existing endeavor to reduce the mental to the physical As mentioned in the first ch apter, I do not think that this reductionism need instantiate itself explicitly in the individual attitudes of researchers, but rather it shows implicitly in their research. William Uttal (2001, p.4) elaborates on this idea: So seductive and exciting have the new technologies been, and so important the promises of understanding at last how the brain is re lated to the mind, that the bedrock issues of the accessibility and analyzability of cognitive processes and
60 brain loci have largely been ignored The very posi ng of the problem [relating brain activity to cognitive processes] presumes a monistic ontology and a nonmystical, realistic, materia listic, and natura listic philosophy. Functional magnetic resonance imaging (f MRI) relies on the finding that after an action potential, there is an influx of oxygenated blood into neurons. This influx occurs to compensate for the energy required by sodium -potassium pumps in the neuron, as well as to prepare the neuron for further metabolic processes (Roth et al., 2006, Uttal, 2001) To measure this blood flow, the ratio of deoxygenated blood to oxygenated blood is calculated, based upon the res onance patterns of hydrogen, wh en run through a magnetic field. Hydrogen, when present in oxygenated hemoglobin, versus deoxygenated hemoglobin, shows a slightly different res onance pattern; the method of measuring the difference between these res onance patterns is called bl ood oxygenation level dependent (BOLD). It takes 4-6 seconds to detect this difference between oxygenated and deoxygenated hemoglobin. For this reason, fMRI has poor temporal resolution. It cannot determine, for instance, that at this exact moment some neural activity has occurred. An activation threshold is designated, and the areas of the brain where BOLD activity above this threshold are averaged, a nd represented three dimensionally (Roth et al., 2006; Uttal, 2001). Volumetric pixels, or voxels, are colored ba sed upon the intensity of the difference in deoxygenated to oxygenated blood, and displayed on an x/y/z coordinate plane, which itself is superimpos ed onto the actual MRI scan of the brain. Though fMRI has poor temporal resolution, it has comparatively good spatial resolution. It is able to pinpoint BOLD changes to within a cubic centim eter (Uttal, 2001). But what does a measurement about differences in blood flow actually say about
61 neural activity? The precise relation between cerebral blood flow and neural activity is still under question (Roth et al., 2006, Uttal, 2009). Because it is as of yet, impossible to directly measure focal action potentials, in vivo (this would require invasive measures) it would seem to be a misnomer to say that neural activity is be ing measured directly Rather, cerebral blood flow is an indicator, of what might be called neural activity. Many of the imaging studies in clinical neuroscience borrow tasks from cognitive neuroscience, particularly thos e that pertain to executive f unctions (Roth et al., 2006). Executive functions are supervisory cognitive processes involved in the selection, initiation, execution and monito ring of complex domain spec ific cognitive processes and motor responses, as well as aspects of emotional processing (R oth et al., 2006, p.2) Findings are used to predict expected activation areas, and clinical samples are compared to healthy ones to detect differences. An example paradigm is in the investigation of self-awareness, or self -monitoring. Self-monitoring, presumably, would be one aspect of types of mental illness. Indeed, excessive rumination upon self-other interactions is one characteristic ma ny diagnosed with MDD (DSM-IV-TR, 2000). Studies on self-monitoring have involved aski ng subjects to perform tasks or answer questions in the scanner; they are designe d to make the subjects think either about themselves, or others. For example, to induce self-monitoring one study asked non-clinical subjects to read a list of words, which were descrip tions about their own personalities (Kircher, Senior, Phillips, Benson, Bullmore & Bra mmer, 2000). The subjects had provided these descriptions prior to the study, so the la tency between writing and remembering was controlled for. Descriptions were read with subjects lying down in the MRI scanner for about an hour (Kircher et al., 2000). It should be noted that although improvements have
62 been made to MRI scanners in recent days the magnet is typically quite loud (Uttal, 2009). From the self-descriptive study, results s howed activation in the left-hemisphere, left inferior frontal, left ante rior cingulate, left parietal cortex, and bilateral precuneus regions (Kircher et al., 2000). Activation pa tterns garnered from other self-monitoring studies typically showed activa tion in the anterior and posteri or cingulate cortex, as well as bilateral frontal regions (Roth et al., 2006). The generali ty of these findings, however, is disputed (Morin, 2002). Other studies show drastically different activation patterns, which seem to change depending on the task being used to in duce self-monitoring. For example, another study asked particip ants to read brief stories, and then attribute responsibility to themselves, or to others (Vogeley, Bussfield, Newen, Herrmann, Happe & Falkai, 2001). Activation wa s recorded in the bilateral anterior cingulated cortex, right premot or cortex, bilateral precune us, and the right temporalparietal junction (Vogeley et al., 2001). These findings may indicate that self-monitoring is actually composed of several, different processes; or, it may just be that selfmonitoring involves several spat ially separate, but functionally overlapping regions of the brain. In a review paper of a host of the brain imaging studies done on self-monitoring (Roth et al., 2006) in the last ten years, it was posited that not only self-monitoring, but many other executive functions, are involve d in mental illness (Roth et al., 2006). Working off of this idea, other researchers ha ve attempted to correlate self-monitoring--a construct created for the proposes of brain im aging--to mental illness. In one study, self-monitoring was examin ed in a sample of schizophrenics. Researchers attempted to see whether a lack of self-monitoring cont ributes to the core
63 syndromes of schizophrenia, which in this case included either paranoid-hallucinatory syndrome, or formal thought-disorder (K noblich, Stottmeiser, Kircher, 2004). Participants included those diagnosed under DSM-IV-TR criteria for schizophrenia, with either of the syndromes ( n = 27), and a matched, non-clinical control group ( n = 23). For this experiment, participants were asked to draw circles on a PC-t ablet, while observing the way in which their hand movements corre lated with the drawings on the display (Knoblich, et al., 2004). After a few circles were drawn, the mappi ng between the tablet and the displayed drawing was dramatically altere d, so that the pointer gained distance from its point of reference. The purpose of this task was to determine whether those with schizophrenia would be able to compensate for the change (Knoblich, et al., 2004). Measured by the root mean squared error of the spatial distan ce from the tracking signa l, healthy controls and schizophrenics were shown to not be si gnificantly different (Knoblich, et al., 2004). It should be noted, this study took a completely different approach from the fMRI studies previously described. That is, the type of self-monitoring be ing correlated with schizophrenia may have not even been th e type measured in the fMRI studies. Even though both Kircher et al.s pe rsonality-description study (2000) and Knoblich et al.s drawing study (2004) have elucidated wh at are probably important details relating self-monitoring and clinical symptoms, the two cannot be seen as complimentary. Kircher et al.s study certainly added to the unders tanding of functional areas of the brain, but whether those areas are actually involved in self-monitoring, or some other cognitive process, is unclear. To extrapolate investigations into clinical populations based upon constructs that have not been psyc hometrically validated may lead to faulty conclusions or type II errors.
64 This does not mean self-monitoring is not involved with mental illness. The problem is that the functional localization of self-monitoring is still at a preliminary stage. The array of reviewed fMRI studies all used different means of initiating selfmonitoring, and hence instead of localizing its function fo cally, could only localize it diffusely (Roth et al., 2006). Some of the sp ecific areas reported active in self-monitoring are themselves associated independently with a breadth of other functions; one example is the prefrontal cortex (Ka ndel, 2000). Most of Phineus Gages injuries included focal damage to the pre-frontal cortex, and Gage s symptoms were highly complex, including entire personality changes (Damasio, 2004). Other studies, based upon the assumption that there are impairments in executive functions in mental illness, carry on in a si milarly disjointed fashion. Research into executive dysfunction for Atte ntion Deficit/Hyperactivity Di sorder (ADHD) has yielded an array of data which seem to show that for both children and adults with ADHD there are primary deficits in working memory, as well as response inhibition (Roth et al., 2006). These conclusions were drawn based upon a smattering of previous studies thought to have localized these executive f unctions in the brain (Roth et al., 2006). From those previous studies working me mory was associated with activation in the inferior frontal gyrus, dor solateral and ventral frontal cortices (Goldberg, Herman, Randolph, Gold & Weinberger, 1996; Jonides, Smith, Joeppe, Awh, Minoshima & Mintum, 1993). Also, bilateral frontal and pa rietal lobe activati on was noted, though in some cases there was hemispheric asymmetry. Th is was attributed to the different types of tasks used, which included verbal, spatia l, and delayed respons e types (Roth et al., 2006). The posterior parietal cort ex, anterior cingulate gyrus, and basal ganglia were also implicated (Petrides, 1996).
65 From these brief examples, it should be clear that the overlap between cognitive and clinical neuroscience is perhaps not so clean. While data keeps pouring in on functional areas implicated for executive functions, clinical trials use that data to try and see how they are involved in mental illn ess. But the groundwork has not been laid. Presumably, one of the goals of cognitive neuroscience would be to correlate all of the voxel maps for any given construct to extract common factors, in orde r to clarify exactly what self-monitoring is but it seems to rather be occupi ed with collecting as much data from as wide an array of c ognitive tasks as possible. At what point, assuming that fMRI studies continue in this somewhat insulated manner, attempting to study independent executi ve functions while controlling for others, could findings be integrated into a holistic pi cture of mental illnes s? The difficulty of controlling for other stimuli besides those directly under question seems to hinder the progress of being able to make strict localiz ations. After all, ther e is no point when the brain is not receiving stimulation (Kandel, 2000). After a long period of sca nning, thresholds could be ad justed according to what might be considered baseline activation patter ns (from things like vi sual input from the walls of the scanner, auditory input from the magnet, emoti onal reactions, etc.). Eventrelated paradigms in fMRI claim to be se nsitive to fluctuations in mood and other environmental variables, in that they present stimuli in a random or der (Roth et al., 2006) as opposed to in blocks. But event-related desi gns still adhere to th e exact same cognitive tasks, which may, or may not, have anythi ng to do with the cognitive function under question. But my concerns are less about the co rrect way to do brain imaging and more about the potential claims that clinical ne uroscience may make about mental illness and
66 functional localization. While loose localizations have been achieved from cognitive neuroscience, strict (one-to-one) localizatio ns have not. To take these general findings and attempt to map them onto mental illness seems to me premature. To say, analogous to the claim that depression is a malfunction that ADHD is a dysfunction of the working memory areas of the brain, is, in my opinion, infeasible. But assuming that there may be a point when the cognitive tasks utilized ar e narrowed down, and executive functions can be concisely described at the psychological level, gaining explanatory power, there still exist further impediments to eliminative reduction. Neuroplasticity and Di sjunctive Statements For a long time it was thought that there we re critical developmental periods, in which if a domain specific area like Bro cas area -did not deve lop, it never would (Huttenlocher, 2003). Once these critical peri ods passed, it was surmised that the brain lost its capacity to learn certain things (H uttenlocher, 2003). These areas were considered immutable (Nudo, 2006). It was also thought th at the number of neurons in the brain, after early childhood, was fixed (Nudo, 2006). In the 1960s several animal studies began to demonstrate that the brain is not so immutable. Rat studies began to be published that described larger brains and increased cortical volume, for those that were raised in enriched environments (Bennett, Diamond, Krech, Rosenzweig, 1964). Later, it was shown that rats raised in enriched environments also had larger neuronal somas, larger syna ptic contacts and dendri tic spine density, more complicated dendritic branching, and a higher synapse-to-neuron ratio (Mohammed et al., 2002). While it was still believed that once adulthood was reached, no new neurons could be formed, it was found that even in adult rats there exis ts neurogenesis (the
67 creation of new nerve cells) within th e hippocampus (Kempermann, Kuhn & Gage, 1998). These initial studies ign ited a whole new realm of inqu iry. Through further animal studies and refined theories regarding exact ly how the environment (particularly, sense data) shapes cortical change s, it was found that through expe rience, both the structure and physiology of the brain can change (Huttenloc her, 2003). This phenomenon is referred to as neuroplasticity (Nudo, 2006). It has been found to also exist in humans (Nudo, 2006; Huttenlocher, 2003). Human adults have also demonstrated neurogenesis in the hippocampus. There has also been much research into the recove ry of sensory and motor functions post-stroke and acute central nervous system damage. It has been shown that those recovering from stroke can regain some speech function, and th at depending on sensorimotor input, those recovering from debilitating paralysis can even regain some of their previous motor function (Nudo, 2006; 2003; Huttenlocher, 2003). Studies have also demonstrated the existence of extensive corti cal rewiring following traumatic brain injury in monkeys (Dancause et al., 2005), as we ll as cases of spontaneous re covery of motor function in humans, following traumatic brain injury (Nudo, 2006). The discovery that the brain can literally rewire itself, that ax ons can form entirely new functional pathways, and that in certain extreme cases, such as in hemispherectomy, gross structures can take over the functions of others (Hutte nlocher, 2003) must influence the way that we think of the brain, as an or gan to be studied. Though there is still much research being conducted into the mechanisms by which ne uroplasticity works, it still retains implications for discussions on eliminative-reduction. Neuroplasticity puts a strong limitation on the potential norma tivity of strict functional localizations. Because a given brain structure has the potential of changing,
68 based upon extenuating environmental input, a generalizing attri bution of functioning cannot be made. It may be the case that for 99.9% of the human population, the corticospinal pathway projects down through th e thalamus and the cerebral peduncles, crosses over at the pyramids, and extends through the dorsal pa rt of the spinal cord. But because of the possibility of people who are different, and in light of evidence that evidences motor remapping, you cannot sa y that the cor ticospinal tract is the motor control pathway. You could sa y that generally it is, but also must acknowledge the variance. The corticospinal tract is relatively simple, anatomically, compared to other parts of the central nervous system. Perhaps you could say, so and so activity in the amygdala indicates a fear response. Perhaps you could predict that any human being, when forced to encounter a large bear, w ould show this same activati on pattern. But considering the possibility that at some point in the future this structur e of the brain may take on the functioning of another, you cannot make this identity claim, at least not until it has been proven experimentally that the amygdala is immutable. But where does this leave a reductionist? Should scientists stop attempti ng to describe the neurophysiology correlated with fear responses? I do not believe so. Reductive explanations would need to be limited to derivations. In regards to functional localizations, most of the derivations or logical se ntences would have to take the form of disjunctive statements. For example, a fear response is this or this or this [n] . Alternatively, scientists could take a non-reductive stance, accepting that in some cases certain mental states are too complex, and that they supervene upon brain states (Clark, 1980). But for those dedicated to reductive mate rialism, there are still other limitations
69 besides neuroplasticity that will make any reductive endeavor, whether eliminative or derivative, difficult. They are mainly limits on normativity, which, if not considered, would lead to explanations with remainders For instance, compari ng clinical populations to healthy populations is problematic. Consid ering that criteria for clinical diagnoses typically include DSM-IV-TR criteria, wh ich are not widely agreed upon. The DSM-V has been in deliberation for ten years, and is poised to dramatically revise certain criteria for mental illnesses, as well as add new categories (American Psychiatric Association, 2010). Although random sampling is good for dist ributing variance cau sed by individual differences, brain imaging studies typically have few participants, both because of the ethical concerns using clinical samples, and the cost of using an fMRI machine for even an hour (Uttal, 2009). Small sample sizes, as we ll as the previously mentioned difficulty in controlling for baseline activit y are also limits to normativity. Another consideration is that of childr en. While it may someday be possible to make identity statements between a given c ognitive phenomena associated with mental illness and a local burst of neural activity, this may never be true for children. Seeing as the prevalence of diagnosing children with mental diso rders has increased (Dawdy, 2010), the use of psychopharmaceuticals such as SSRIs, making the assumption that what works for adults will work is potentially dangerous. Left-handed people are under-represented in imaging studies, be cause of different hemispheric lateralization (Lacasse & Leo, 2005). There are also cross cultural issues. There have been few comparative studies between brain functioning of those in one country, to those in another. For exampl e, one study did attempt to investigate Americans neural activity in response to domestic and foreign music (Morrison et al.,
70 2003), but its results were inconclusive. Consid ering the role environment is posited to play in neuroplasticity, more cross-cultural research is n eeded to make generalizable claims. Summary of Token Arguments against Iden tity Theory and Eliminative-Reduction The biopsychosocial model of mental illne ss, while superficially attractive, entails the same philosophical problems encountered in the mind-body problem. Because psychiatry seems to be the locus where these problems converge, it was briefly discussed. Making the assumption that some psychiatrists are identity theorists, believing that the mind and the brain are the same, and that this view arises via psychopharmacology, the serotonin hypothesis was examined. The possibil ity of making an identity claim between a serotonin imbalance and depression was refuted. Functional magnetic resonance imagi ng was examined under the lens of eliminative-reduction. The feasibility of strict (one-to-one) locali zations was examined. Methodological limitations were discussed; these included the temporal latency of measurements, the difficulty of controlling for complex stimuli while attempting to isolate others, and the incons istency of findings due to di fferent experimental designs. Borrowing inconclusive findings about execu tive functioning and attempting to explain mental illness in terms of them becomes problematic. The body of research for fMRI is limited at this point. Didactic complaining about its methods is unlikely to lead to any major changes; however, precisely because of the infancy of the domain of research, the types of claims made by researchers need to be carefully examined. For instance, researchers ha ve claimed that the areas of the brain for mate choice have been discovered, based upon one study with only 17 participants
71 (Fisher, Aron & Brown, 2007). This claim need s to be considered critically under the requirements of any eliminative reduction. Th e previous section attempted to describe these requirements, and the problems encount ered in attempting to localize function in the brain. Eliminativism and The Hard Problem As stated earlier, any broad reductionist claim must adhere to the principle of explanations without remainde r. If one assumes that someday we may have the technology to be able to make identity claims about mental illness a nd brain states, is it still possible that some fundamental aspect would be left unexplained? Returning to the formal requirements of elim inative reduction, if psychological variables are being reduced, and there is a remainder, then the reduction is unsuccessful. The reducing theory must have at least as much explanatory power as the reduced theory; it must be able to explain at least the same phenomena as the reduced theory. In chapter two, an argument was described that is employed in favor of dualism; this was the argument from knowledge. The exam ple of Mary the scientist was used to demonstrate that a purely ne urobiological account of the pr ocess of vision still left something unaccounted for. In that case, the un accounted thing was said to be qualia. For Mary, the neurobiological e xplanation cannot account for what it is like for her to see something. It cannot account for th e subjective expe rience of seeing. The phrase, what it is like originated with another thought experiment. In his 1974 essay, What is it like to be a bat? Thomas Nagel proposed that we could describe echolocation, and all of the physiological under pinnings of what it means to be called a bat, but that we cannot know what it is like to be a bat, because we are not bats (Nagel, 1974). He speculates that if any organism has conscious experience at all, then it means
72 that there is something it is like to be that organism. The nature of this being is for the organism, something that only it has access to (Nagel, 1974). He goes on to call it the subjective character of experience (p.392). Qualia, per their character, arise out of a subject, not an object. Therefore, they cannot be studied in an objective way like prope rties or functions might, but rather must be approached in a completely different ma nner. Either that, or a connecting principle bridging the objective with the subjective must be discovered. But as of yet, no general theorem of subjective experience, no psychophysical law, has yet been deduced. There have been proposals, influenced by Hegel s phenomenology (Dennett, 1991) but none of which have caught on in mainstream thought. This separation between the objectiv e aspects and subjective aspects of phenomena is called the explanatory gap (H orst, 2007; Levine, 1983). The idea of the explanatory gap was first posited by Joseph Le vine. He spoke particularly of the mindbody problem, and explanations about consci ousness (Levine, 1983). He elaborates on Nagels initial concerns a bout being able to reduce c onscious experience, or the what it is like to naturalistic, or even functionalist terms. Using the example of an identity claim between pain and c-fiber fi rings, he says (1983, p.357): There is more to our concept of pain th an its causal role, there is its qualitative character, how it feels; and what is le ft unexplained by the discovery of C-fiber firing is why pain should feel the way it does! For there seems to be nothing about C-fiber firing which makes it naturally fit the phenomenal properties of pain, any more than it would fit some other set of phenomenal properties Unlike its functional role, the identification of the qualitative side of pain with C-fiber pairing (or some property of C-fiber fi ring) leaves the connection between it and
73 what we identity it with completely mysterious. Unlike any other reductive explanation, accoun ting for subjective experience seems to pose a problem that cannot be answered in naturalistic or functionalist terms. Whereas a heart can be explained naturalistically as a collection of cardiac muscle, connected by gap junctions, nerves and blood; or functionall y, as a thing that pumps blood, neither suffice to account for what it is like to feel your heart pumping. Both Levine and Nagel agree that successful reductions must not leave unexplained phenomena hence the explanatory gap. The explanatory gap has resurfaced in cont emporary times, partly due to the work of David Chalmers, who has contextualized it within contemporary neuroscientific debates about consciousness (Horst, 2007; Chalmers, 1995; 1996). Chalmers speaks of the hard problem of consciousness, which ex ists because of the explanatory gap, and the soft problems of consciousne ss. In his influential paper, Facing Up Chalmers clarifies (1995, p.3): The easy problems of consciousness are thos e that seem directly susceptible to the standard methods of cognitive science, whereby a phenomenon is explained in terms of computational or neural mechan isms. The hard problems are those that seem to resist those methods. The soft problem of consciousness encompasses some of the contemporary breakdowns in neuroscience. These include: the ability to discriminate, categorize, and react to environmental stimuli; the integration of information by a cognitive system; the reportability of mental states; the ability of a system to acce ss its own internal states; the focus of attention; the de liberate control of behavior ; and the difference between wakefulness and sleep (Chalmers, 1995).
74 Many of these aspects of consciousne ss are assumed to be represented by investigations into the executive functions of the brain. The aforementioned construct of self-monitoring likely entails, to some extent parts of the soft problems. The specific contingencies necessary to make reductions fr om brain imaging studies into the executive functions apply generally to the soft problem s. But again, taking an optimistic stance and saying that neuroscience may someday be able to make derivative reductions and identity claims between certain brain states, what would be reduced are only the soft problems (Chalmers, 1995). The hard problem is hard precisely because neuroscience seems to have no access to it. It is relevant to psychology in that psychologi cal states also retain a what it is like to be in those states. Clinical neuroscien ce may be able to explain all of the neurophysiological and neurochemical facets of MDD, for instance, but it cannot explain what it is like for a person to be depressed Because any derivative or eliminative reduction must also be a broad reduction, the only options are to reje ct the existence of the hard problem, or acknowledge caveats to every reductive explanation acknowledging that there is a yet-un explained part of it. Eliminativism rejects the hard probl em of consciousness (Churchland, 1989; Dennett, 1996; 1991). Because eliminativis m adheres to a strict physicalism and naturalism, which says that the only things in the world are physical, and that all explanations of that world ma y be formulated on the basis of the terms utilized by natural sciences, the idea of conscious experience r uns into obvious conflict. Analogous to its criticisms of dualism, in general, eliminativism says that merely declaring the existence of a thing which cannot be expl ained naturalistically does not entail that it actually exists (Churchland, 1988). Vice versa, if har d consciousness cannot be explained
75 naturalistically, then it does not exist. All of the terms used to substantiate a non-reductive attitude : qualia, experience, even consciousness itself, are considered to be so vague and ephemeral, they could never explain phenomena with the same level of accuracy as the natura l sciences. For this reason, they are considered illegitimate concepts. Recalling the replacement assumption from chapter one, which says, assuming T will provide a more fundamental explanation, it should subsume T (Clark, 1980, p. 27), eliminativis m assumes that explanations put in terms of the physical sciences provide a mo re fundamental description of reality than terms like consciousness, or qualia (Churchland, 1989). Dennetts Heterophenomenology Daniel Dennett, a stark eliminativist, has published widely on his reasoning why qualia and consciousness are illegitimate concepts. In Consciousness Explained (1991), he speaks on why he thinks so much empha sis has been placed on qualia. He believes that because many philosophers begin from what is closest to them; that is, their own intuition, but that they take for gr anted that their own intuition is actually their own (Dennett, 1991). Dennett believes that such an intuition exists, but that it is nothing more than an intuition, or a sort of strivi ng towards (Dennett, 1991). He th inks that any exploration of consciousness, while taking into account subjective accounts will simply reveal that there is some reason for a subjective acc ount to be occurring (Dennett, 1991). Presumably, these reasons can be explained natu ralistically, and so what is understood in the folk sense as intuition, really entails someth ing like the reportability of mental states. To substantiate these claims, Dennett proposes what he calls heterophenomenology, which is method he considers as a scientific, 3rd person
76 approach to the study of consciousness. He begins by noting how t hose authors who write on consciousness fall victim to what mi ght be called the first-person-plural presumption (Dennett, 1991, p. 67). They phras e their statements in terms of we, meaning both I and you. It makes the presumption that the I intuition about consciousness is comparable to the you, in that phenomenologically the you is introspecting in the same way as the I (p.68). Secondly, and perhaps more significantly, he says that people who speak on consciousness overestimate the reliability of introspection. He says (p. 68): What we are fooling ourselves about is the idea that the activity of introspection is ever a matter of just looking and seeing. I suspect that when we claim to be just using our powers of inner observati on, we are always actually engaging in a sort of impromptu theorizing and we are remarkably gullible theorizers. Heterophenomenology involves applying the scientific method by combining an individuals self-report with all other naturalistically observable data in order to determine their mental state. It does not disqua lify what a subject is saying, but takes into account things like physiologica l states, which may render th e subject to be wrong about her own mind (Dennett, 1991). For instance, a person may say she is not hungry, but an observer may see that she in fact is hungry because of an increase of stomach acid. She may just be saying that because there is so me underlying reason why she sees herself as not being hungry (Dennett, 1991). He continues (p.83): The heterophenomenological method neither challenges nor accepts as entirely true the assertions of subjects, but rath er maintains a constructive and sympathetic neutrality, in the hopes of compiling a definitive description of the world a according to the subjects.
77 From practicing the method himself, Dennett conc ludes that the defini tive description is sufficient for a reductive explanation, and that it would render s ubjective experience as nothing more than a cause for reporting e xperience (Dennett, 1991). Dennett concludes that when he looks inside, all he sees is the thing causing him to look inside. That is, he sees attention, self-monitoring, etc. (Dennett, 1991). For any intuitive question a subject asks, there is a corresponding question wh ich asks why does the subject ask that question? The answer is supposed to en compasses the total range of what is called consciousness, reducing it to pr ocesses like reportability. Chalmers Response Chalmers response to Dennetts h eterophenomenology includes pointing out that you cannot discount the Hard Problem through third person accounts, even if those accounts acknowledge the fallibility of self-report. There is no way to discount something without being able to stu dy it, and there is no way to study phenomenal experience without there being something it is like to study phenomenal experience. By taking a stand on personal experience, you are yourse lf experiencing th at taking-a-stand (Chalmers, 1995). Furthermore, he says that while describi ng things in terms of functions works for other forms of explanation, consciousness is unique in that it lies at the center of an individuals epistemic world (Chalmers, 1995) He says that the proposition of turning ones first person perspective into a third person perspective already assumes that all that is needed to explain consciousness is a desc ription of structure a nd function (Chalmers, 1995). It also assumes naturalism, which, as stated previously, is prerequisite of eliminativism. Reasserting the explanatory gap, Chalmers says (1995) that conscious experience
78 is not "postulated" to explain other phenom ena in turn; rather, it is a phenomenon to be explained in its own right. And if it turns out that it cannot be explai ned in terms of more basic entities, then it must be understood as ir reducible. (p.3). He admits that both his and Dennetts arguments rest upon conf licting assumptions. He laments (p.4): we will probably just have to get used to the fact that there is a basic division in the field: that between those who think the "easy" problems are the only problems, and those who think that subjec tive experience needs to be explained as well. We can therefore expect two qu ite distinct sorts of theories of consciousness: those which explain the f unctions and then say "that's all", and those which take on an extra burden. This type of conflict is conceptual, a nd wholly dissatisfying. Both Chalmers and Dennetts arguments seem to rest upon the que stion of whether one accepts the existence of the explanatory gap. While the arguments considered previously about eliminative reduction and identity theory were primarily methodological, the arguments in this case are epistemological. They are concerned with how we construct explanations. Eliminativists say we should construct arguments naturalistically, while those in favor of the explanatory gap say we should construc t them intuitively. How does one go from these conflicting views on describing the world, to conclusions about explanations without remainder? Negative Epistemological to Metaphysical Connection Are conceptual difficulties enough to abandon the acceptance of the explanatory gap? If one accepts the disproving of metaphys ical content by epistemological necessity, or the negative epistemol ogical-to-metaphysical connec tion (Negative EMC), and believes that an explanatory gap still exists then one has reason for believing that an
79 elimination is unsuccessful. In this case, epis temological necessity tends to take the form of possible-world s cenarios (Horst, 2007). If one can prove logically that some A supervenes epistemologically upon B then it makes sense to say that it is metaphysically necessary, for A supervenes on B if there can be no change in A without a change in B. But there are different types of supervenience which do not en tail metaphysical necessity. Although Chalmers stance is non-reductive as regards the hard problem, that is, it believes the explanatory gap is real, rejecti ng Negative EMC, allows one to not have to consider them metaphysically separate. In other words, one can accept the existence of the explanatory gap without being a substance dua list (Horst, 2007). The general description of supervenience, provided above must be differentiated in two ways before one can consider th e plausibility of Negative EMC. These differentiations include natura l supervenience and metaphysi cal supervenience. Natural supervenience means that there can be no tw o possible worlds where the properties of B can change without changes in A properties, assuming its ph ysical laws are identical (Horst, 2007, Chalmers, 1996). Metaphysical su pervenience says that when all the B properties are fixed, then the A properties follow. At first glance, these are confusing st atements. Relying on hypothetical possible worlds seems useless, because possible worl ds, by means of them being hypothetical, do not actually exist. But using possible worl ds scenarios helps clarify the difference between metaphysical (ontological) necessity, and epistemological (logical) necessity. If one believes that there are mental properties, T for which there are no physical properties, T that can give a conceptually ade quate (no remainder) explanation of T then one accepts the explanatory gap (Horst, 2007) You would say that mental properties T
80 do no naturally supervene upon physical properties T. If one believes in Negative EMC then it would follow that it is not metaphys ically necessary that physical properties necessitate mental properties (Horst, 2007) Put another way, you cannot derive mental properties from physical propertie s; mental properties would be irreducible. This is the stance of Chalmers (Chalmers, 1996). If one rejects Negative EMC, and still accepts the explanatory gap, then one might be a non-reductive physicalist, believing that mental properties do metaphysically supervene upon the physical, but th at there can be no explanati ons of the mental in terms of the physical. Non-reductive physicalism does make an identity claim between the physical and the mental, but it does not es pouse derivation or eliminative reduction. Finally, one may reject both the explan atory gap and Negative EMC. This is the stance of eliminativism. It believes that consciousness can be explained by physical things without remainder. It also believes in a positive epistemological to metaphysical connection, in that by showing that consciousne ss can be reduced to physical processes, consciousness is metaphysically supervenient on the physical. Because eliminativism accepts eliminative reductions, although it is pessimistic about the explanatory power of T in the first place, it would advocate elimin ating consciousness from the scientific vocabulary. Summary of Eliminativism and the Hard Problem Seen from the context of Negative EM C, the stark discontinuities between eliminativism and non-reductive stances seems to make more sense. But where does all this leave a neuroscientist, who may acknowle dge the current explanatory gap, but does not believe in dualism? Neuroscience is ce rtainly not going to stop, for the sake of unresolved philosophic debates. As men tioned above, non-reductive materialism is a
81 viable option, although in practice, it is no different from identity theory. Does one adopt a dualist perspective? Certainly those who purport the existence of the explanatory gap may be dualist, like Ch almers himself, but is there perhaps another way of relating the apparent separation of mind and body, which is conducive to productive clinical research ? Having addressed the si gnificant problems with both eliminativism and reductive materialism, I w ill now consider an alternative non-reductive approach that does not offer an explanation for the gap, but provide s a model of relating the mind to the body, addressing the problem of interaction. There ha ve been few cogent theories, in this regard; however, John Eccl es interactionism has a unique perspective on the problem which I believe is worth consideration.
82 CHAPTER IV: PSYCHONEUROIMMUNOLOGY AS INTERACTIONIST EXEMPLAR Yet, even if the metaphysical mindbody problem remains insoluble, it does not follow that we cannot make progress in coming to a better understanding of the medical mind body problem: how mind and body interact in mainta ining health or causing disease. It is the second question raised above that is germ ane to psychiatry in general and depression in particular, especially regarding the idea that the mind can affect the brain and body, and vice versa. -Walter Glannon Interactionism Revisited Interactionism divides reality into th ree worlds (Table 1). These worlds are supposed to represent what Eccles calls cosm ic evolutionary stages (Eccles & Popper, 1974, p.16). Table 1 Summary of Eccles and Poppers Three Worlds Name of World Constituent Entities The Products of the Human Mind (III) Works of Art and of Science, Technology Human Language. Theories of Self and Death. Subjective Experiences (II) Cons ciousness of Self and Death Sentience (Animal Consciousness) Physical Objects ( I) Living Organisms The Heavier Elements; Liquids, Crystals Hydrogen and Helium
83 Each world is a gross representation of the stages of human evolution. But when Eccles and Popper mean evolution, they do not mean it in strictly Darwinist terms (Eccles & Popper, 1974). Each stage emerges from the previous; therefore, they are considered irreducible. Eccles admits that it appears th at all living things, plants and animals, are bound by physical and chemical laws (they emerge from them), but they are not reducible or necessarily supervenient upon them (Eccles & Popper, 1974). Physical processes like forces interact with each ot her, and thus do physical objects interact, within World I. Popper and Eccles consider the constituen ts of World II to exist, just as the objects of World I. They include mental stat es, and consciousness ( both the hard and the soft types) (Eccles & Popper, 1974). World I and World II ar e considered to interact, although Eccles and Popper admit that this id ea runs into the same problems as any mindbrain problem, in that it posits a relation between physic al and non-physical entities (Eccles & Popper, 1974). What makes this type of interactioni sm peculiar is the addition of World III, which includes the products of the human mind. Eccles and Popper include things like language which is intangible. It is an idea. But it exists, according to Eccles and Popper. World III also includes physical objects, like sc ulptures, which are physical. In this sense, there is an interaction betw een World I and World III. Theories of self can instigate creative efforts, but creative efforts can also be instantiated in physical object s (Eccles & Popper, 1974). So in some sense, components of World I are also components of World III. But how is this possible? While emergence occurs in the grand scenario, stage by stage, sculptures do not inde pendently arise from limestone. There needs to be an intervening st ep. This step is the human mind; the means by which World I and World III interact is World II (Eccles & Popper, 1974).
84 Subjective experiences a nd consciousness, mental states, mediate World I World III interactions, as well as World III World I interactions. Eccles and Popper assert that products of the human mind exis t: books, technology, la nguage, art, etc., and that some of these products are physical objects, but how does the human mind go about translating physical obj ects into products? World III emerges from World II; that is, the products of the human mind, such as language, are supervenient upon World II, a nd accepting that both World III and World I actually exist ontologically, the existence of World II becomes a necessity (Eccles & Popper, 1974). In fact, for most phenomena, a valid description necessitates mutually inclusive interactions between all three worlds. For exampl e, a book cannot exist without the physical material that composes it (an in tra World II emergence). But the materials could not be organized into a book on thei r own; they need a human organism, in combination with a human mind, to put them together. A book is also not a book unless it has wo rds written in it. Language is a product of the human mind, and written language requ ires a physical object, thus interactions between World II, III, and I. It seems appare nt that this way of describing phenomena is potentially recursive in an infinite nu mber of ways, but Eccles and Poppers interactionism manages to addr ess some of the pitfalls of reductive materialism, as well as Cartesian dualism. Because subjective experience is ack nowledged to exist, explanations of phenomena in the world will not be overlooking the hard problem; also, because the cosmic stages are emergent but not hierarchically supervenient this interactionism is non-reductive. Although it does no t posit connecting principles, or psycho-physical concepts like the pineal gla nd to account for the World I World II interaction, Eccles
85 and Poppers interactionism s till retains explanatory power, in that the existence of the mind is not simply posited to exist through in trospection, but rather observance of real things in the world, such as language and art. But assuming that a neuroscientist or psyc hiatrist chose to adopt this outlook, how exactly might it allow for productive resear ch to continue? How does one translate a philosophical stance into a model for research? I believe that there already exists a field in neuroscience which does this, and does so successfully. Psychoneuroimmunology Psychoneuroimmunology (PNI) investigates the specific interactions between psychological variables, the immune system and the neuroendocrine system (Tausk, Elenkovt, Moynihan, 2008). It has evolved out of advances in the three disparate fields, but firstly owes allegiance to those philosophe rs who first posited specific theories of mind-body interaction, such as Hippocrates, Des cartes, Eccles, and Popper. I believe that Eccles and Poppers interactioni sm can be extrapolated to ac count for these three levels of description, particularly in the multi-directionality of interactions, and within-World interactions. Relevant background information relevant to the different levels of explanation will be provided, as well as a description of the theoretical basis on which they were integrated. Finally, I will return to M DD, and examine it through the lens of psychoneuroimmunology, pointing out the ways in which an interactionist approach (as opposed to a reductive materialist one) is condu cive to exploring alte rnative explanations that are not necessarily relate d to the serotonin hypothesis.
86 Stress Theory Discoveries by Russian scientis t Ivan Pavlov in the late 19th century predicted an ideological shift in psychology that paved th e way for a series of animal studies, which directly lead to the f ounding of psychoneuroimmunol ogy (Garcia, Kimeldorf, & Koelling, 1955). In Pavlovs basic design, he pa ired a conditioned stim ulus (the ringing of a bell) with an unconditioned stimulus (dog food). He was able, after repeated pairings, to cause the ringing bell to elic it the conditioned response of salivation. This accomplishment is an example of classica l conditioning. Classical conditioning shows that you can pair a neutral stimulus with an eliciting stimulus, and eventually the neutral stimulus will become an e liciting stimulus on its own. Classical conditioning differs from instrumental conditioning, which was formulated and discussed by Edward Thor ndike, and B.F. Skinner (who developed Thorndikes ideas later). Instru mental conditioning concentrat ed less on the ability of an animal to adopt new eliciting stimuli, but rather to examine eliciting stimuli (Sd) and how they initiate a response. It also examines the way in which the response affects the environment, causing feedback (Sr) which will either in crease the likelihood of responding again, or decrease it. Behavior ism in psychology employed instrumental conditioning for many of its experiments. Walter Cannon, in 1915, coined the term homeostasis wherein he stressed the importance of balancing psychological and physio logical processes, in maintaining health (Cannon, 1932). Cannons theories also examined the relationship between stimuli and response, but focused instead on internal, phys iological reactions, as opposed to external ones. This started when he noticed a trend in hi s work with animals. When in a state of
87 arousal, animals digestive functions w ould cease (Cannon, 1932). This led Cannon to investigate the relation between emotional st ates and physiological reactions. He was the originator of the fight or flight response Since this point, other effects have been associated with the fight or flight response; these include the releasing of noradrenaline, increased heart rate, and increa sed blood pressure (Rabin, 2005). Hans Selye was one of the first to de scribe stress, which he defined as the nonspecific result of any demand upon th e body, which can be produced by a wide variety of stimuli. Selye called these stim uli stressors (Selye, 1951). Selye conducted research wherein he subjected animals to st ress for varied length s of time (Selye, 1951). He studied trends in their bodily reacti ons, especially their ability to heal. It was from this research Selye formul ated his GAS model, which conceptualizes a generalized stress reaction as consisting of the stages of alarm resistance and exhaustion The alarm stage entails the initial stress response, or the fight or flight response. Normally the body returns to homeost asis following the alarm stage, but if it stress continues then the body goes into the resistance stage. The body produces a cascade of hormones that raise blood pressure and increase blood sugar, maintaining the stress response (Selye, 1951). If the body does not rest, or go through periods of relaxation to counter-balance the energy expe nditure, then it goes into the exhaustion stage. In the exhaustion stage, the bodys en ergy reserves are depleted; significant decreases in mental and emotional resources occur, and the bodys immune functions are inhibited. Disease and death typically follow (Selye, 1951). The GAS has proven extremely useful, applied to specific questions such as the effects of chronic stress on cardiovascular functions (Selye, 1951). The GAS is conceptually integrated with
88 Cannons findings, particularly in conceptualizin g stress as a disruption of homeostasis, and in seeing the fight or flight response as part of the alarm stage of the GAS. Conditioning the Immune Response Throughout the 20th century, there was an increas e in understanding of both the immune system, and the mechanisms of the autonomic nervous system. The details of such discoveries will be reviewed, but two important studies will be discussed first, because they initiated the future conver gence of stress theory, neurophysiology, and immunology. In 1955 a discovery was made with animal experimentation. Scientists (Garcia, Kimeldorf, and Koelling, 1955) observed th at following exposure to a low-dose of gamma radiation, rats exhibite d depressed food and water cons umption. The severity of this effect increased with the length of exposure. Th ey hypothesized that this might be due to a learning mechanism: reinforced avoidance through repeated pairings with radiation. To investigate this phenomenon furt her, they attempted to elicit the avoidance using classical conditioning. Garcia et al. pu t saccharin in the rats water, which was made available to the rats during subsequent irradiation sessions (G arcia, Kimeldorf, & Koelling, 1955) Rats were split into several groups, varying both in the length of radiation exposure and the condition of wate r-tap or saccharine. The first day after irradiation, those groups who received tap water were give n saccharin water, and vice versa, for 6 hours. Following the second day, both tap and saccharin water was made available to rats on a continuous basis. Consumption rates were measured in median percentages. Those rats who did not receive radiation were shown to maintain their initial preference for saccharin water (w ith a median preference of 86. 1% ), whereas the groups
89 who received radiation showed marked decrease s in preference. With 30rads, the median preference = 33. 4%, and with 47rads = 3. 1% This conditioned ta ste aversion lasted up to 30 days following the initial irradiation. This finding was important in that a very strong preference for the saccharine water was extinguished due to short pairings with aversive stimuli (Garcia et al., 1955). Almost 20 years later, Ader and C ohen (1975) conducted another important experiment with rats, following which, the field of psychoneuroimmunology became formally established. What Ader and Cohe n did was take the original design and go a step further. They discovered that physio logical immune reactions could be conditioned with the same classical conditioning methods that Pavlov used with his dogs. Some previous taste aversion studies, surprisingly had resu lted in the deaths of the rats, and the proposed hypothesis was that along with aversion to the saccharine substance, immunosuppression had been conditione d as well. This lead to Ader and Cohens study, in which they set up a compli cated protocol with a placebo, conditioned, and unconditioned group of rats, along with further subgroups of conditioned rats, which received varied amounts of saccharine, fo llowed by injections of cyclophosphamide (to condition sickness) along with sheep red bl ood cells (SRBCs), to test the effects of immunosuppression. Ader and Cohen found that those rats who received the cyclophasphamide acquired taste aversion to the saccharine water, as found in previous studies; furthermore, they also found significantly attenuated an tibody titres in those rats that had the cyclophasphamide paired with SRBCs (A der & Cohen, 1975). This was the first experiment that provided direct evidence fo r a link between the ne rvous system and the immune system. Ader and Cohen were able to make the unconditioned immune response
90 conditioned to saccharine water. By pairing a naturally occurring immune reaction with cyclophasphamide, Ader and Cohen managed to take what was thought to be a function unrelated to learning (the imm une response), and show that it could be conditioned just like the salivary response. Following this experiment, the term psychoneuroimmunology was conceived. As a new field, it began to further investigate the interactive nature of stress, the nervous system, and the immune system (Task, Elenkov & Moynihan, 2008). To retroactively understand what is meant by immunosuppres sion and to interpret contemporary findings, the relevant tenets of immunology will be discussed. Fundamentals of the Immune System The immune system may be generally di vided into two categories: adaptive and innate (Irwin and Cole, 2005). Innate imm unity includes classes of cells which all mammals retain. It is the f irst line of defense against invading pathogens. Pathogens include any disease causing infectious agents Though the chemical messengers involved in triggering an innate res ponse are complicated, they pale in comparison to the adaptive immune response. Adaptive immunity includes special cells which are able to recognize antigens (literally, an tibody-generators) and adapt to them, giving organisms a way to recognize and fight repeated exposures. Both components of the immune system are characterized by their constituent cell types, signaling molecules, and medium (whe re they originate from and proliferate to) (Herbert & Cohen, 1993). Most classes of i mmune cells originate in the bone marrow, thymus, spleen, lymph nodes, tonsils, appendi x, and peyers patche s, which are clumps of immune tissue in the small intestines (Herbert & Cohen, 1993). These cells tend to circulate in the lymph and pe ripheral blood stream, migrating to sites of inflammation as
91 a result of molecular signaling cycles. As it is difficult to gain access to many of the sites where immune cells arise, blood samp ling is a common extraction method in immunological assays. In the bone marrow, along with re d blood cells, white blood cells begin development in by means of a cascade of hor mones onto pluripotent stem cells. White blood cells are called leukocyt es; a class of agranular white blood cells are called lymphocytes. A sub-class of lymphocytes trav els from bone marrow to the thymus, where they mature to become T lymphocytes. T lymphocytes are involved in the adaptive immune system. They recognize foreign anti gens that have been introduced to the thymus (Rabin, 2005). T cells may be further divided, depending on certain cell surface markers, into helper T cells and cytotoxic T cells. Helper T cells assist other cells by releasing special molecules called cytokines, which act as ha rbingers for phagocytic cells as well as B lymphocytes, whose primary function is to produce antibodies (Irwin and Cole, 2005). Cytotoxic T cells can help to destroy tissue th at is being invaded by viruses. They use a special method called lysis, which involves punc hing holes in the cell membrane to cause its contents to spill out, resulting in ce ll death (Rabin, 2005). Monocytes are another class of non-granulated cells, which engul f foreign materials and expand, becoming macrophages. After engulfing antigens, the m acrophages wait until th ey are activated by CD4 helper T cells. At this point th ey destroy the antigen (Rabin, 2005). During an adaptive reaction, exposure to an antigen results in the production of lymphocytes which become reactant to it; they will, in the future maintain a memory against future introductions, causing a quick er and stronger defense. As a caveat, the difference between the terms pathogen and an tigen is that pathogens are any invasive
92 exogenous invader, including viruses, while an tigens specifically cause antibodies to be produced. The primary adaptive response takes four days after initial pathogen introduction (Rabin, 2005). It involves se veral stages, which involve a host of different cells and processes. First there is the collection of an antigen by den dritic cells, which reside in tissue or mucous. The dendritic cell breaks the antigen into 8-12 chain peptides, which then become bound to major histocompatabi lity complexes (MHC I or MHCII) (Rabin, 2005). The pathogens are next presented to CD4 help er T cells, either in the spleen or the paracortical region of a lymph node. At this point dendritic cells are now called antigenpresenting cells. This pres entation causes naive CD4 cells to become activated CD4 cells, which gain the ability to recognize the antigen when re-i ntroduced at a future point. When this happens, the response is called a seco ndary adaptive response. It can occur in a much more expedient manner than the primar y response -as quickly as 24 hours (Rabin, 2005). B cells mature in the bone marrow excl usively. They produce antibodies through the following process: in lymph nodes, naive B cells come into direct contact with antigens, binding them to antigen-specific an tibody molecules, dige sting them, and then presenting them in MHC molecules to activated T cells. The activated T cells then bind to the antigen and releases cytokines. These cy tokines cause activated B cells to release antibodies (Rabin, 2005). Antibodies come in differe nt classes (IgG, IgA, IgE, IgM, IgD) and serve different protective functions. Structurally speaking, most antibodies are similar. They are glycoproteins with arrangements of light and heavy chains. Antibodies also contain
93 regions called fragment antigen binding (Fab) regions, which can vary by type. It is the Fab region that gives an an tibody its pathogenic specificity (Rabin, 2005). Antibodies work by binding to invading antigens, slowi ng them down, and damaging them. They coat the antigen, initiating its removal by macrophages and other cells which recognize them. Other immune reactions can also be in itiated by the presence of an antigen coated by antibodies (Rabin, 2005). Antibodies work to signal other molecules, calling them to the site of infection. Cytokines are also signaling molecules. They encompass a multitude of cell-tocell communicators, and assist in such f unctions as the activa tion of T cells, the production of antibodies, and the activation of the inflammatory response. The latter function is controlled by so called pro-inflamm atory cytokines, of which interleukin-1( IL-1) is integral (R abin, 2005). IL-1 is produced by ma crophages, monocytes, fibroblasts and dendritic cells. IL-1 works by increas ing endothelial adhesion factors in blood vessels, enabling the transmigrati on of white blood cells to the site of infection. IL-1 also causes the hypothalamus to produce fever. The chronology of both innate and adaptiv e responses independent of each other is well understood. The way these two components relate to each othe r precisely is not; however, there are ongoing inves tigations in molecular biol ogy attempting to elucidate this question. Microbial pattern recogniti on receptors (PRRs) ar e the means by which innate cells recognize microbes, and have b een offered as a potential explanation. PRRs detect pathogen associated molecula r patterns. They come in three different classes; secreted, which bi nd to pathogens and opsoni ze them for phagocytosis; transmembrane, which are able to detect a va riety of features about a pathogen, such as the lipposacharide of gram-negative bacteria; and cytosolic, which are able to recognize
94 viral ribonucleic acid (RNA), fo rms of stress (for instance, ultraviolet irradiation), microbial products, and noninfectious crys tal particles (Iwasak i & Medzhitov, 2010). These PRRs are on, and in, a variety of different cells (Iwasaki & Medzhitov, 2010). PRRs can activate adaptive immune res ponses such as the IgM, IgG, and IgA antibody responses; T helper cell (TH1) and TH17 CD4+ T cell responses, as well as CD8+ T cell responses (Iwasaki & Medzhitov, 20 10). It has been sugge sted that there are different mechanisms by which a cell-extrinsi c or cell-intrinsic PRR might initiate an adaptive response. Associate recognition, in either case, is thought to play a strong role in the magnitude of the response. For example, B cell receptors along with ot her innate signaling pathways, such as complement, results in an enhancement of the antibody response (Iwasaki & Medzhitov, 2010). Similarly, coengagement of B cell recep tors and transmembrane pattern receptors enhances the antibody response (Iwasaki & Medzhitov, 2010). The way in which PRRs mediate the endogenous release of cytokines, and initiate other ad aptive processes such as the activation of T-4 helper cells is still unde r investigation, but PRRs have shown promise as a potential explanation of the tr ansition between innate and adaptive immune responses. Measurement of the Immune System As should be evident from the above re view, the cells and chemical messengers that make up the immune system are numer ous. There are almost as many biological tests and assays, which measure the differe nt processes these cells and chemical messengers perform. These tests can be broken down into those that look for the presence of a signaling molecule or cell t ype (enumerative), and those that test the
95 effectiveness of certain cells in fighting o ff foreign antigens (func tional) (Ader, Cohen, 1975). Some examples of tests include lym phocyte subsets, cytotoxicity assays, lymphocyte proliferations, cytokine receptor counts, and antibody assays. These tests all retain advantages and disadvantages. One general disadvantage is that the amount of circulating molecule does not necessarily correlate to a bett er functioning immune system (Vedhara, Wang, 2005). In the case of anti bodies, an enumerative assay will indicate only that the amount found circulating will respond to a specific type of antigen. A person with low levels of a certain an tibody may maintain perfectly good health (Vedhara, Wang, 2005). Another issue is the sens itivity of certain hormones or signaling molecules to stress. For example, certain hormones are extremely sensitive to acute stressors, and levels may change within minutes of being pricked with a needle. These hormones also often operate under circadian rhythms. For exam ple, cortisol, a glucocortoicoid, runs on a diurnal cycle throughout the day, having a morn ing zenith and an evening nadir. This means that blood tests measuring cortisol need to be taken at similar times and in similar conditions in order to give an accurate estimation of functioning. Many assays are in-vitro or outside of a living orga nism. Because they are not examining immune functions inside the body, th ese tests may not be the most accurate indicator of immunological functio ning. But there do exist several in-vivo tests. One type of in-vivo test involves the administration of a live or weakened virus, and the subsequent monitoring of internal, enumerative changes am ong lymphocyte subsets. There are also a variety of delayed hypersensitivy tests which measure different functional responses. But in-vivo tests also retain disadvantages. Logistical i ssues include finding healthy
96 participants, isolating a pathoge n to administer, and controlli ng for normal responses in a representative sample. A Model of Neuro-Endocrine Interaction Since the mechanics of the immune system have been discussed, the neural and endocrine details should be examined. In the last century, scientists have increased their knowledge about neurophysiology greatly. Le arning, at the neuronal level, has been elucidated. This is due, according to Para diso (2007), to the theory of long-term potentiation (LTP), which was formed by th e collective efforts of Hermann Ebbinghaus, who showed that memory could be studied experimentally, Donald Hebb, who posited the idea of synaptic strengt hening (Hebbian learning), Terj e Lmo, for his work with hippocampal networks in rabbits, and Eric Kandel and his work with the sea snail aplysia Kandel investigated the molecular basi s for habituation and sensitization, which are both important concepts in PNI studies (Par adiso et. al, 2007). LTP says that the more a neuron fires, the stronger its synaptic stre ngth. This causes simila rly encoded signals to be transmitted faster to the structures invol ved in memory (Lmo, 2003). This is relevant to another discovery in neurophysiology, one wh ich has greatly aided PNI investigations: the discovery of Papez circuit. The Papez circuit is part of the limbic sy stem. It is composed of several functional loops, also called pathways. The Papez circui t is one of those loops, and it has been implicated in both memory and emotion (Rab in, 2005). Grossly speak ing, it includes the hippocampus, amygdala, mammillary bodies, ante rior thalamus, and the cingulate gyrus. Papez circuit also has connections from the frontal cortex, which may serve a regulatory purpose. An unpleasant memory, triggered in the amygdala, may cause aggression and a fight response through ac tivation of the hippocampus, and then the hypothalamus
97 (colored purple)(Figure 4). This may be ove rridden by downward projections through the cingulated gyrus to the thalamus and hypot halamus (colored or ange)(Figure 4). Figure 4 Medial Slice of Hippocampal Affe rent and Efferent Projections. The hypothalamus (colored green)(Figure 4) is on e of the starting points in PNI models, as it produces many immunologically relate d hormones. The way it is affected by cognitive functions, especially those in memo ry centers, is relevant to substantiating some of the earlier findings in learned helplessness and c onditioning. Specifically, it can help explain the connection between stressor evaluation and a biol ogical response.
98 Many complex processes involving Papez circuit and a variety of hormonal feedback loops are involved. But despite the difficulty invol ved in the peculiarities of stress recognition and the immune system, certain models or axes have been conceptualized to describe th ese feedback loops. The first such axis is the hypothalamicpituitary-adrenal, or HPA axis. The hypothalamus is situated anterior a nd basal to the thalamus, which is a hub for incoming sensory, and outgoing motor in formation. The hypothalamus influences a variety of human functions, many subsumed by the four Fsfeed ing, fighting, fleeing and fornicating (Vedhara & Wang, 2005). Th ese functions are in fluenced through the secretion of hormones, which, generally sp eaking, migrate to the pituitary gland to mediate the release of further hormones, which have peripheral effects on the body. The pituitary gland sits anterior and ba sal to the hypothalamus, and is split into anterior and posterior sections These sections are by the way hormones migrate to them, and which type of hormone they receive. All hormones travel to the pituitary through the hypophyseal portal system, but the paths towards the anterior and posterior pituitary are separated by blood vessel walls. The full range of hormones that are secreted by the hypothalamus will not be described. Rather, one specifi c hormonal sequence that has pr oven to be relevant in the mediation of immune reactions will be. Th is sequence begins w ith the secretion of corticotrophin releasi ng hormone (CRH), from the parave ntricular nucleus (PVN) of the hypothalamus. CRH is released in response to stress. The stimulatory pathways that cause this to happen are noradrenergic and serotonergic, and the cascades of which are the same, irrelevant of the type of stressor -whether they be emotional or physical (Kaye & Lightman, 2005).
99 The PVN also releases arginine vasopr essin, which has a diuretic effect, along with a vasoconstrictive one, working to in crease blood pressure and cause the kidneys to retain water. CRH travel s to the anterior pituitary gland, which releases adrenocorticotropic hormone (ACTH), as we ll as several types of enkephalins and endorphins. ACTH then travels via the blood stream to the adre nal cortex of the kidneys, causing the release of gluccocorticoids, most notably cortisol. The gluccocorticoids are involved in a negative feedback loop back to the hypothalamus. Cortisol has a variety of effects on th e body. Generally speaking, it functions in two major capacities: regulati on of glucose and fat uptake for quick response to acute stress, and immunosuppression, by preventing T-ce ll proliferation. Cortisol achieves the latter effect by rendering IL -2 producer T-cells unresponsiv e to IL-1, and making them unable to produce the T-cell growth fact or (Kaye & Lightman, 2005; Palacious & Sugawara, 1982). Cortisol also has some effects on the other major model in the neuroendocrine system, the sympatho-adrenomedullary, or SA M. As mentioned previously, CRH and noradrenergic neurons in the brain cause th e release of ACTH. CRH also causes the release of norepinephrine by the peripheral sympathetic nervous system, and the release of epinephrine and norepinephrine from the adrenal medulla. These hormones contribute to the fight response described by Walter Ca nnon in the alarm phase of the GAS. Both the HPA (Figure 5) and SAM axes are m odels for how the neuroendocrine system interacts with the immune system (Figure 6).
100 Figure 5. The HPA-Axis With Cortisol Feedback (Lane, at al., 2009)
101 Figure 6. Brain-Immune System Interface (Lane, at al., 2009) The negative feedback loop involving cort isol (Figure 6) is important in PNI studies. Through general circul ation, cortisol migrates back to the hypothalamus and anterior pituitary, causing CRH and ACTH to be down regulated. This in turn results in the adrenal gland synthesizing less cortisol. When cortisol gets low, CRH and ACTH
102 will again be produced. This can happen from psychological stress, or metabolic demand. Chronic stress may cause an exce ss of ACTH to be produced, resulting in bursts of cortisol and keeping its blood concen tration at a high level. This can have deleterious effects (V edhara & Irwin, 2005). Hypercortisolism, or Cushings syndrome, is the result of an excess of cortisol. Its symptoms can include weight gain, fa tigue, slow healing of superficial wounds, decreased libido, headaches, and in extrem e cases a glucose intolerance leading to diabetes (MFMER, 2008). ACTH and cortisol can be elevated because of bottom-up, non psychological stressors. These include su ch things as receiv ing surgery, or being wounded (Vedhara et. al, 2005). These are phys ical stressors, which may or may not be influenced by the complex emotional pathways of Papez circuit. Despite this, the physical stressors share the same immunosuppressive e ffects as emotional st ressors (Vedhara et al., 2005). Prolonged physiological imbalance from up-regulated immune systems can cause serious problems in higher functioning areas of the brain, such as the hippocampus (Tollenaar, Elzinga, Spinhoven & Everaerd, 2008). As previously described, the hippocampus has been implicated in the evalua tion of stressors. The negative feedback loop involving the hippocampus can be disrupte d, wherein an excess of cortisol will cause the hippocampus to suppress the act ivity of the hypothalamus to produce CRH (Vedhara et al, 2005). There is evidence to suggest that regula tion from the anteri or cingulate may be able to override this inhibition, but extr eme psychological stress not only causes an excess of cortisol, but also decreases the se nsitivity of the negative feedback systems, which may result in habituation (Vedhara et al., 2005). While in some circumstances,
103 habituation to certain stressors may be adva ntageous, in the case of chronic stressors which are not life threatening, the constant up-regulation of cortisol, due to depressed release of CRH, can result in harmful effects on the body. Psychological Stressors In 1994 Ader and Cohen conducted a meta -analysis examining the effects of a broad array of stressors on immune functioni ng. They classified stressors according to negative life events, both object ive discrete events and self -reported checklists. It was recognized that any individual may have a heterogeneous mix of these stressors. Both subjective and objective classes of stressors we re measured over a specified time period. The negative events included in the analysis te nded to be salient, and occurring within a short time of the immune assessment (Ader and Cohen, 1994). Ader and Cohen recognized that the length of the stressor probably had an effect on the magnitude of the immune response; that is, whether the stre ssor was acute or chronic had an effect on the strength of th e immune response. They speculated that certain chronic stressors woul d lead to habituation to co rtisol, where acute stressors would lead to a more pronounced immunosuppr ession. They recognized however that this distinction is not clear cut. For exampl e, there is evidence that bereavement involves elements of chronic stress with burst s of acute stress (Ader & Cohen, 1994). Ader and Cohen concluded that there wa s substantial evidence for the relation between stress and decreases in immunologi cal functioning, as well as measures of lymphocyte cell counts. They further conc luded that objective lif e events tended to correlate with more severe immunosuppression, showing that stressor length was related to magnitude of immune response, and also that interpersonally related stressors were more prone to immunological changes than non-social ones (Ader & Cohen, 1994).
104 Ader and Cohen recognized that the stre ngth of the mean-effects found were not overwhelmingly large. They admit methodologi cal flaws inherent in meta-analyses, as well as the difficulty of obtaining reliable data from immunological assays (Ader & Cohen, 1994). Despite this, they reiterated th eir conclusion that there exists a strong relationship between objective, proximal stressors, and immune response, along with distal, chronic subjective ones (Ader & Cohen, 1994). They note the fact that they could not isolate the difference between immunosuppression elicited by one type of stressor over another, but also describe interesting findings relating interpersonal and nonsocia l events. For the immune outcomes of helper-to-suppressor ratios, as well as percents of suppressor/cytotoxic T cells it was found that interpersonal events resulted in greater immune alteration than nonsocial events (Ade r and Cohen, 1994). They note that although interpersonal stressors are more likely to trigger depressive affect than other types of even ts, it is unclear that on the whole they are stronger immunosuppressive variables than nons ocial events. They conclude by saying, we need a more comprehensive picture of immune effects associated with specific stressor characteristics, a nd, we need to determine the personal and social characteristics of individuals that render th em more or less suscep tible to stress-induced immune alteration. (p. 375). A l ack of social support is a ri sk factor of major depression (MDD) (DSM-IV). Understanding the way in which DSM criteria relate to stress theory and immunosuppression seems a valuable endeavor. Psychoneuroimmunology of Depression MDD is one of the most common mood -related disorders, with a lifetime prevalence estimated at 15% for normal popula tions, and 15-36% for medical in-patients
105 (Irwin & Cole, 2005). The reason for the latter statistics magnitude is that medical inpatients experience pre-disposing conditions, some of which increase the risk of developing MDD. These can occur in isola tion, or co-morbidly, and may include such things as: recovering from surgery, sleep dist urbance, alcohol dependence, hospitalization stress, or tobacco usage (Irwin, Cole, 2005). These risk factors also are independently correlated with immunosuppressive effects. Fo r this reason, controlling history effects can be difficult, when attempting to ascertain the effects of DSM cr iteria depression on the immune system. Sleep deprivation, incr eased consumption of nicotine, and alcohol usage all have been shown to produce certa in immunosuppressive eff ects (Irwin & Cole, 2005). As reviewed previousl y, the serotonin hypothesis of major depression retains certain problems that limit its explanatory pow er. Alternative explanations for MDD are desirable, especially ones that integrate a lternative physiological findings, while also taking into account the qualit ative aspects of depression -particularly the interaction between personality traits and interpersonal stressors. One avenue to find alternative explanations is in the PNI literature. It has been demonstrated through previ ous studies that ther e exists a complex interplay between glucocortico ids, prefrontal areas of th e brain, Papez circuit, the hypothalamus, the pituitary, and the kidneys. While some of the specific chemical mechanisms from each structure are unclear, they all contribute to the evaluation of stressors, neuroendocrine res ponses, and immune regulation. There are two helpful ways of relati ng depression to these processes, conceptually. The first is in the inte rpretation of MDD symptoms. MDD symptoms typically include weight change sleep disturbance, psychomot or retardation or agitation,
106 pervasive fatigue, feelings of worthlessness or irrational guilt, mental concentration difficulties, and recurrent suicidal ideation (Irwin & Cole, 2005). These symptoms can be considered within a behavioral analytic pe rspective; specificall y, as contingencies of operant conditioning. Unlike in rat studies, in this case the mental can be taken into account. These contingencies generally take th e form of: establishing operation (EO) discriminative stimulus (Sd) response (R) reinforcing stimulus (Sr for reinforcers, Save for aversive stimuli) (Pierce & Chene y, 2004). As an example, an adolescent who previously used to have no punishing or re inforcing relationship to standing in large groups of people may get to the point wher ein she/he experiences a large degree of distress, leading to rumination and guilt. The chain of contingencies might take the form of: not being in a group of people (EO) entering into a large group of people (Sd) sweating, feeling anxious, etc. (R) experiencing guilt (Save). This approach integrates symptomol ogy from generalized anxiety, while also taking into account the subjective nature of any given situ ation. This is compatible with the idea of neuroplasticity and individua l differences, which I believe mentalistic psychology is sensitive to. This model allows fo r specific contingencies to be defined at the idiographic level. For exampl e, irrational guilt could be c onsidered in the context of a class of stimuli that have lo st their neutral or reinforc ing value, and have become aversive. Key to this is the personalized nature of operationalization. As regards relating this approach to neuroendocrine and immunological assays, stress theory may again be utilized. If an individual is thought of as an organism constantly attempting to maintain a homeostatic relationship to its stressors, then depression can be considered a failure to cope with some, or all of those stressors. This
107 fits with Selyes GAS model. Hypothesis te sting, based upon previous findings with animals that are in the stages of alarm, resist ance, or exhaustion could be used to examine the physiological correlates of MDD. Indeed, Martin Seligman (1975) offered that depression could be thought of on the basis of learned helplessness, and the physiological correlates experienced in the exhaustion stage. As regards integration with neuroendocrine systems, the two major axes used in other strains of research can again be used when investigating MDD. Within the context of the SAM, depressed individua ls show increased levels of catecholamines (epinephrine, norephinephrine) and neuropeptide Y in res ponse to acute, physio logical, as well as psychological stressors (Brown, Pa tterson, Mascovivch & Grant, 1991). This finding maps nicely on the earlier example of guilt contingencies. The response of sweating, anxiety, etc. can serve as a descri ption of a fight reaction, which includes an increased heart ra te, increased blood pressure, and rapid breathing (Vedhara & Irwin, 2005). SAM activation can be activ ated by less acute stressors, such as the mere anticipation of a stressful situation. It is gene rally implicated in increased risk for metabolic and immune related disorders (Ve dhara & Irwin, 2005). This probably occurs by means of the amygdala, which contributes to emotion, as well as emotional memory. There also exists evidence that activation of the SAM is the m echanism by which acute myocardial infarction occurs after significant physical exer tion, or significant emotional trauma (Vedhara & Irwin, 2005). Within the context of the HPA axis, cortis ol is again hypothesized as relevant for depression. Both hypercortisolism and depres sion also have the shared symptom of fatigue, and depressed individua ls usually show an increase of CRH (Vedhar & Irwin, 2005). Psychiatrists, colloquially at least, se em to corroborate this link, as Cushings is
108 listed in the DSM as an Axis III (medical c ondition) criteria for depression (DSM-IV-TR, 2000). The DSM also offers th at the etiology for depression may be related to an imbalance of neurotransmitters, but it does not list it as specific criteria. Increased cortisol has effects on proinflammatory cytokines, immune functioning, and bodily homeostasis in general. To examine the initial hypothesis th at depression is related to immune functioning, Zorrilla et al., (2001) conducted a meta-analysis of over 180 studies utilizing more than 40 immune measures. They did not speculate as to the mechanism, but were interested in a general pi cture of how those with M DD functioned immunologically. They also attempted to compare the re lation of the results of MDD-related immunosuppression with general, stressor-re lated immunosuppression (comparable to the earlier meta-analysis conducted by Ader and Cohen). It was shown that MDD is associated with leukocytosis, reductions in NK-cell counts and T-cell proportions, increases in CD4/CD8 marker ratios, and decr eases in Tand NKfunctioning (Zorilla et al., 2001). Linking these findings with naturally occurring stressors, there were shared effects of overall leukocytosis, manifesting as a relative neutr ophilia and lymphoenia, increased CD4/CD8 ratios, increased haptogl obin, PGE, and IL-6 levels, reduced NK cytotoxicity, and reduced lymphocyte prolifer ative responses to an introduced mitogen. For stressors alone, Zorilla et al. ( 2001) found leukocytosis, manifesting as lymphocytosis, alterations in cytotoxic ly mphocyte levels, CD4/CD8 ratios, and NK toxicity, a relative reducti on of T cells, reduced lymphoc yte response to mitogen, and proportion of IL-2r bearing cells following s timulation. Considering the shared effects between the independent analyses of depressi on and stressors, it would seem that there
109 are concrete immunological changes that occur with depression. In another meta-analysis, conducted by Burke, Davis, Otte and Mohr (2005), seven studies were analyzed to understand the relation of cortis ol levels in depressed, and non-depressed individuals, in relation to stress. They found that baseline levels of plasma cortisol were not significantly differe nt between depressed and non-depressed individuals. Twenty-five minutes after the onset of laboratory stressors, what Burke et al. called the recovery period, majorly depresse d individuals showed significantly higher cortisol levels than non-depresse d individuals (Burke et al., 2 005). This is consistent with the idea that previous and chronic unregulated cortisol can impair depressed individuals ability to evaluate and recover from stressors. One mechanism by which the deregulati on of the SAM and HPA may contribute to depression is in pro-inflammatory cytokine s. Pro-inflammatory cy tokines such as IL-1 and IL-6 regulate the innate immune res ponse by signaling for leukoc ytes and starting the antibody response; they are ex tremely important for the acute phase response. With depression, it has been shown that there are increases in pro-inflammatory cytokines, such as IL-6, tumor necrosis factor (TNF) and IL-1 (Motivala, Sarfatti, Olmos & Irwin, 2005). They have been theoreti cally linked to depression through their neuroendocrine effects as well as effects in neuronal activity (Capuron et al ., 2007). Apart from roles in the acute phase, pro-inflammatory cytokines can affect the HPA axis directly (Capuron, Miller & Irwin, 2007). In the first instance, pr o-inflammatory cytokines, including IL-6, Il-1, IFN-alpha, and TNF-al pha have been shown to cau se the amount of CRH being released by the hypothalamus to increase. CRH is integral for the regulation of the HPA and SAM, and a hypersecretion of it may lead to the habituating response to
110 glucocorticoids, described earlier. CSF samples in depressed patients relia bly show an increased amount of CRH (Pariante & Miller, 2001). It has also been shown that for those undergoing cytokine therapy for cancer and viral infection, approximately 30-50% will develop MDD (Capuron et al., 2007). CRH leads to the releas e of ACTH, which cont rols the release of gluccocorticoids as well as catecholamines from the adrenal medulla. Immune cells have receptors for cortisol, as well as adrena line and noradrenaline. When CRH is upregulated, there is a resulting down-regul ation of NK-cell functioning, as well as further increases in pro-inflammatory cytokines su ch as IL-6 and TNF (Capuron et al., 2007). This means that at some point when an organi sm is under stress, cytokines used to initiate an acute response lead to a dysfunctioning of the HPAs inhibitory feedback loop, which includes CRH, ACTH, and cortisol. Furthermore, the pro-inflammatory cytoki nes which feed back to the brain cause dramatic changes to the glucocorticoid re ceptors in the hypothalamus and hippocampus (Pariante & Miller, 2001). Thes e changes alter glucocorticoid receptor (GR) functioning by decreasing their expression, phosphorylation st ate, protein-protein interactions, and/or GR binding to DNA (Capuron et al., 2007; Pa riante & Miller, 2001). An increased inflammatory reaction may, in this case, cause a cytokine-induced feedforward cascade leading to decreased effectiveness of the GRs, and the gluccocorticoid mediated feedback loop that they are involved in (Capuron, et al., 2007). Cortisol itself can cau se negative alterations in brain physiology. In 2008, Marieke et al. found a negative correlation between cortisol levels and gross size of hippocampal tissue. This was a worrying fi nding, and several other studies have attempted to pinpoint the relation between cortisol, depression, and the hippocampus;
111 specifically, the relation of depr ession with cortisol levels, the relation of cortisol to limbic functioning, and the relation of depressi on and cortisol levels with memory and executive functioning. These studies have variously found that elevated cortisol is related to poorer performance on memory tasks, and that depr ession and cortisol interact for poorer performance on a color-word indices of res ponse inhibition (Gomez, Posener, Keller, DeBattista, Solvason & Schatzberg, 2009). Depression independently has been correlated with poorer memory performance on a delayed recall test (Gorwood, Corruble, Falissard & Goodwin, 2008), which was asso ciated with hippocampal dysfunction. Supporting this interpretation, studies have s hown that depression is associated with decreased hippocampal volume, neural atr ophy, and loss of neurons (Marieke et al., 2008; Schmidt & Duman, 2006). In a review, Sapolsky (2000) describes th e role that neurodegeneration plays in several types of psychiatric disorders. He sa ys that too many glucocorticoids (GCs) in the hippocampus cause an increase of excitatory amino neurotransmitters, such as glutamate, in the synapses. GCs will also affect the availability and effectiveness of certain neurotrophins: brain-derived ne urotrophic factor in particul ar, which helps to maintain neural connections. These combined effects can cause atrophy in dendritic processes, leading to explicit memory deficits (Sa polsky, 2000). GCs are also shown to be neurotoxic to CA3 neurons in the dentate gy rus, and also to inhibit adult hippocampal neurogenesis (Sapolsky, 2000). Various MRI studies have shown gross hippocampal volume reduction in depressives, ranging from 8% to 15%, depending on whethe r depression was severe or mild, and also whether patients were in remission (Bremmer, Narayan, Anderson, Staib,
112 Miller & Charney, 2000). One study claimed to be able to predict the actual number of days an individual had been experienci ng depressive symptoms, depending on the magnititude of hippocampal de generation (MacQueen, Campbell, McEwen, Macdonald, Amano & Joffe, 2003). Seeing as the hippocampus is involved in the feedback loop for CRH, both in the evaluation of stressors and recognition of ci rculating GCs, these findings are worrying. Damage to the hippocampus inhibits the func tioning of the Papez ci rcuit, in that the evaluation of emotionally relevant stressors will be impaired. It also means that following an alarm response, the body will be unable to downregulate its circulating GCs, because of the effects of pro-inflammatory cytokine s in altering the morphology of GC receptors, which makes the hippocampus unable to inhibit the production of CRH. At both the levels of neurophysiology (World II) and immunology (World I) there exist bi-directional effects that contribute to depressive symptoms. The memory and response inhibition deficits resulting from hippocampal degeneration (Gomez et al., 2009; Gorwood, et al., 2008) also re presents a significant World II World I interaction, in that impaired memory and c ognitive functioning are likely to reinforce negative self-ideations, as well as make potentially neutral stimuli aversive. The elucidation of the relation be tween stressor evaluation, immunology and depression has lead to interest ing revisionist hypothesisizing. It has been suggested that there may exist a mechanism by which anti-depressants can increase hippocampal neurogenesis, effectively treating the memory impair ing, immunological factors of depression at the biological level (Schmid & Duman, 2006). In animal models, anxiety and behavioral impairment due to psychosoc ial stress have been shown to decrease hippocampal neurogenesis (Dranofsky & Hen, 2006). Further, chronic treatment with
113 serotonin reuptake inhibitors such as fluoxetin e have proven effective in reversing stressinduced impairment of neurogenesis, as we ll as behavior irregul arities (Malberg & Duman, 2003). A proposed mechanism by which monoami ne modulators (such as SSRIs) may accomplish these effects involves the fact that chronic stress results in the decreased transcription of brain-derive d neurotrophic factor (BDNF) (Dranofsky & Hen, 2006) As described earlier, BDNF functions in maintaining the su rvival of existing neurons, as well as inciting the growth and differentiation of new neurons and synapses. This happens throughout the body, but in adulthood it is most relevant in the dentate gyrus of the hippocampus, where neurogenesis occurs. It has been shown that monoamine modulators results in an increase of BDNF transcription ( Dranofsky & Hen, 2006) Combined with the findings that proinflammatory cytokines can decrease BDNF, it would seem that while stress impairs hippocampal neurogenesis, an tidepressants promote it. This might be a way of explaining the positive effects SSRIs have shown, but th rough a different mechanism, tied in with stress theory and a psychoneuroimmunologica l approach to depression, rather than the serotonin hypothesis. Summary It seems that there are grounds to as sociate depression with findings of the deleterious effects of cortisol, by means of the HPA axis. Cortisol is involved in a complex array of feedback loops; an upregula tion can cause various effects including the inability to react appropriately to threatening stimuli, a maintained resistance stage (contributing to risk of CVD, as well as ge neral health problems), and different cognitive impairments due to neurodegeration in the hippocampus.
114 Depression research at th e level of psychology has cont ributed much in as far understanding symptoms, risk factors, and trea tments (Kanter, 2008), but insight into the corollary activity in the immune, neural, a nd endocrine systems can provide valuable additive knowledge. Utilizing an interactionist approach, th e conceptual differences that may arise between the two levels of investig ation (the psychological and biological) do not necessitate elimination. The psychologica l, neuroendocrine and immunological levels of description are not used as a means of expl aining each other, but rather as a means of explaining a general phenomenon: in this case, MDD. Returning to my initial claim that Eccles interactionism can be applicable in PNI, I believe that the three worlds can be appropr iated as such: World I includes the smallest level of study, which is the individual immune cells, as well as the hormones involved in neuroendocrine initiations. World II include s the neurophysiological structures and processes, such as those involved in the Pap ez circuit, particularly the hippocampus, the anterior cingulate, the hypothalamus, and p ituitary gland. Finally, World III encompasses stress theory. This includes the environment (life stressors, interpersonal stressors, etc.), and the mind, which attributes meaning to those stressors. As shown in this review, all of these Worlds interact among each other, and with themselves. The neurophysiological is influenced by the mental (World III World II), and the neurophysiological affects itself, e .g. through electrical stimulation in Papez circuit. World II affects the release of CRH and eventual cortisol, along with IL-6 and other pro-inflammatory cytokines (World II World I). World I affects both World II and World III. Cortisol feeds back to the hypothalamus, and has effects on the hippocampus and amygdala (World I World II) which in turn changes the way stressors are evaluated (World II World III).
115 These are just some examples. Causal chains, as mentioned in the previous analysis of interactionism, can be conceptua lized in potentially infinite ways. But rather than being a determent, this conceptual flex ibility provides a useful groundwork to relate somewhat disparate levels of explanation, w ithout trying to reduce them. It also manages to avoid the problem of accounting for ps ycho-physical laws, because the goal of interactionism is not to solve the mind-brain problem, but ra ther to provide a useful and accurate way of describing real ity (Eccles & Popper, 1974). It can be thought of more as a model than a metaphysics (Figure 7).
116 Figure 7 Diagram of Interacti ons in Psychoneuroimmunology.
117 While scientists may consider approach as an implicit separation of the mind, or mental, with the body, or physical, I believe that this is an unnecessary conclusion. My argument in this case is not necessarily fo r an interactionist ontology, but rather an interactionist methodology. Imagining the levels of psychological description as interacting with the level of neurophysiology and imm unology has shown to catalyze discoveries, rather than inhibit them. Alt hough a neuroscientist could choose to be a substance dualist, with no real harm, be lieving that the mind and the body are two separate substances is not required for this t ype of interactionism. It is a framework from which models can be built and tested empirically. There may come a point in the future where every nominal me ntal state can be related identically to a brain state, but at this point this has not been accomplished. Because the psychological level of explanation is helpful in describing facets of mental illness that the level of biology cannot, such as behavioral contingencies, and takes into account the subjective experience of having a mental illness, I believe as of yet, a nonreductive stance is a good choice for ne uroscientists. Rather than discounting psychological variables as having little explan atory value, interactionism attempts to integrate them with the p hysiological. The psychological and the physiological are not seen as independent and competing levels of explanation, which must be reduced, but rather as facets of a larger pi cture: that of human beings.
118 CHAPTER V: CONCLUSION AND FINAL STATEMENTS The Brainis wider than the Sky Forput them side by side The one the other will contain With easeand Youbeside The Brain is deeper than the sea Forhold themBlue to Blue The one the other will absorb As SpongesBucketsdo The Brain is just the weight of God ForHeft themPound for Pound And they will differif they do As Syllable from Sound -Emily Dickinson I have discussed the pragmatism in accepting interactionism. I have conceded, for those still dedicated to physica lism, that one does not have to be an ontological dualist to take into account the import ance of maintaining folk psychology. Folk psychology may in fact retain a level of explanatory power that ne urophysiology does not. At the beginning of this thesis, the goals were set out as the following: 1) Provide an overview of mind-body theori es within the context of reductionism 2) Examine the problems with an eliminativist and reductive materialist approach 3) Examine a non-reductive philosophical stance which allows for productive neuroscientific research. Chapter two set out to describe the major di visions in philosophy of mind. Monistic and
119 dualistic theories were discussed, as well as arguments for and against them. There was a recurrent problem of interacti on. While insisting that they ar e separate, dualistic theories cannot account for how the mind and the body interact. Part of the conflict was shown to be the assumption that any account of interac tion must be naturalis tic -explainable in terms of the natural sciences. I discussed reductive materialism and eliminativism at length. Both have contentions with the psychological level of expl anation; reductive materialism is in favor of eliminative reduction, which would eliminat e psychological terms after they have been reduced to neurobiology, while eliminativis m is in favor of the concatenation of psychological terms before any such reduction could occur. Eliminativism contends that folk psychology is inept, and generally provide s explanations that ar e not only inaccurate, but false. Chapter three examined problems with eliminative reduction and eliminativism. Eliminative reduction was in itially attacked on met hodological grounds, using the examples of the serotonin hypothesis, and neuroimaging. I asked the question, If one assumes that someday we may have the technology to be able to make identity claims about mental illness and brain states, is it still possible that some fundamental aspect would be left unexplained? I concluded that the unexplained remainder would be subjective experience. This was shown to be in violation of one of the requirements of a broad reduction, which is expl anations without remainder. I considered the problem of subjective e xperience, as discusse d within the hard problem of consciousness. In particular, I reviewed David Chalmers differentiation between the soft problems and hard problem. The soft problems, Chalmers considers as following under the realm of potential reduction, through the means of neuroscience. Soft
120 problems include examinations into phenomena like the executive func tions. In contrast the hard problem resists reduction through neuroscientific means. Daniel Dennett contends that thr ough his method of heterophenomenology, it can be shown that the hard problem of consciousne ss is reducible to c ognitive functions like self-report. Dennett condones ex amining these self-reports and contrasting them to physiological recordings in orde r to show that subjective e xperience is no more than a physical event, causing a person to report some experience that they actually are not having. Chalmers responds to those who discount the hard problem by reiterating that any supposed objective examination of personal experience will itself entail subjective experience. For Daniel Dennett to ta ke a stand on looking inside through heterophenomenology, there will be something it is like for Daniel Dennett to do so. This response does not discount the eliminativ ist complaint however, which is that the something it is like does not exist in the first place. The two arguments were reexamined in the light of Negative EMC, which sets up contingencies for making metaphysical claims about the world through epistemological necessity. Chapter four showed how in teractionism, which divides reality into three separate worlds, can be extrapolated for the use of different academic disciplines. I used the example of psychoneuroimmunology, whic h integrates psychology, neuroendocrinology, and immunology. I reviewed so me studies on depression, which were shown to take into account the subjective interp ersonal variables, as well as the objective, physiological ones. Alternative hypotheses of the etiology of depression were discussed; specifically, that from the lens of stress theory, Papez circuit, and the HPA axis. In this thesis I have attempted to synt hesize an array of complex theories, perhaps
121 not spending enough time acknowledging the subtle ties and internal dialectics inherent to any one of them. I made assumptions that there eliminative-materialist and eliminativist neuroscientists exist, and that the presence of such scientists lends a threat to the psychological level of description. As a stude nt with limited exposure to the realms of psychiatry and neuroscience, it is possible th at my notions about the implications of philosophical assumptions are unfounded. I admit that the difficulty in providing evidence to substantiate these claims lends to this criticism. I have encountered no studies attempting to explore the underlying philosophical assumptions of neuroscientists; also, because such assumptions are implicit, rather than explicit, such a study would be diffi cult to conduct. Despite this, in examining several key texts written by those who are psychiatrists, philosophers, or neuroscientists, I still believe that th e question of reducti onism is important. In chapter three I attempted to demo nstrate the interface between reductionist attitudes and scientific inquiries and claims, but secondary to this, I also argued in what might be considered an ethical mode, implyi ng that psychiatrists s hould not overprescribe when the mechanism of a psychopharmaceutical is inconclusive (such as with SSRIs), and that neuroscientists shoul d not begin brain imaging i nvestigations into mental illnesses until the relevant dependent variables have been isolated and operationalized. The latter concern was particularly oriented towards laymen, who may read articles in pop science magazines purporting that the mate choice center of the brain has been discovered (Fischer et al., 2007), and believe that such an identity claim is true. Even if one believes that philosophical questioni ng is irrelevant to research, if it can lead to better adherence to the Hippoc ratic Oath, or elucidate ethica l considerations, then it is worth investigation.
123 References Ader, R., & Cohen, N. (1975). Behaviorally Conditioned Immunosuppression. Psychosomatic Medicine 37 (4). American Psychiatric Association. (2010). DSM-V: The Future Manual Retrieved January 2010, from DSM-V: http://www.psych.org/MainMe nu/Research/DSMIV/DSMV.aspx Bear, M., & Connors, B. P. (2007). Neuroscience: Exploring the Brain (Vol. 3rd). Baltimore: Lippincott Williams & Wilkins. Bennett, E. D. (1964). Chemical and an atomical plasticity of the brain. Science 610-619. Bennett, M., & Hacker, P. (2009). Philosophical Foundations of Neuroscience (9th ed.). MA: Blackwell. Beulig, A., & Fowler, J. (2008). Fish on Prozac: Effect of Serotonin Reuptake Inhibitors on Cognition in Goldfish. Behavioral Neuroscience 122 (2), 426-432. Born, J., Rasch, J., & & Gais, S. (2006). Sleep to remember. Neuroscientist 12 410. Bremner, J., Narayan, M., Anderson, E., St aib, L., Miller, H., & Charney, D. (2000). Hippocampal volume reduction in major depression. Am J Psychiatry, 2000 157 115-127. Brigandt, I., & Love, A. (2008). Reductionism in Biology. Retrieved February 1st, 2010, from Stanford Encyclopedia of Philosophy: http://plato.stanford.edu/en tries/reduction-biology/ Broekhoven, F., Kan, C. C., & Zitman, F. G. (2002). Dependence potential of antidepressants compared to benzodiazepines. Progress in NeuroPsychopharmacology & Biological Psychiatry 26 939 943.
124 Burke, H., Davis, M., Otte, C., & Mohr, D. (2005). Depression and cortisol responses to psychological stress: A meta-analysis. Psychoneuroendocrinology Cannon, W. (1932). The Widsom of the Body. New York: WW Norton & Co. Capuron, L., Miller, A., & Irwin, M. ( 2007). Psychoneuroimmunology of Depressive Disorder: Mechanisms and Clinical Implications. In R. Ader, Psychoneuroimmunology. Elsevier Academic Press. Churchland, P. (2005). Functionalism at Forty: A Critical Retrospective. Department of Philosophy, UCSD. Churchland, P. (1988). Matter and Consciousness: A C ontemporary Introduction to the Philosophy of Mind. The MIT Press. Churchland, P. (1986). Neurophilosophy: Toward a Unified Science of the Mind-Brain. Cambridge: The MIT Press. Clark, A. (1980). Psychological Models and Neural Mechanisms. Dawdy, P. (2010). Antipsychotic Use Doubles In America's Toddlers. Retrieved 2010, from Furious Seasons: http://www.furiousseasons.com/archiv es/2010/01/antipsyc hotic_use_doubles_in_ americas_toddlers.html Dranovsky, A., & Hen, R. (2006). Hippocampal Neurogenesis: Regulation by Stress and Antidepressants. Biological Psychiatry 59 1136-1143. DSM Organizing Members. (2000). The Enigma. In DSM-IV-Casebook (p. 329). Eccles, J., & Popper, K. (1974). The Self and its Brain. Engel, G. L. (1977). The Need for a New Medical Model. Science 196 129-136. Feyerabend, P. K. (1963). Me ntal events and the brain. The Journal of Philosophy 140141.
125 Fisher, H., Aron, A., & Brown, L. (2007). Romantic love: An fMRI study of a neural mechanism for mate choice. Journal of Comparative Neurology 493 (1), 58 62. Flanagan, E., Davidson, L., & Strauss, J. (2007). Issues for DSM-V: Incorporating Patients Subjective Experiences. American Journal of Psychiatry 164 (3). Frank, D. (2003). The Biopsychosocial Perspec tive on Mental Disord ers: Depression in the Primary Care Setting. In The Biopsychosocial Approach: Past, Present and Future (p. 85). Freud, S. (1940). An Outline of Psychoanalysis. Garcia, J., Kimeldorf, D., & Koellino, R. (1955). Conditioned Aversion to Saccarhin Resulting from Exposure to Gamma Radiation. Science Ghaemi, N. (2003). The Concepts of Psychiatry: A Pl uralistic Approach to the Mind and Mental Illness. Boston, MA: The Johns Hopki ns University Press. Glannon, W. (2003). Depressi on as a mind-body problem. PPP Goldstein, B. (2007). Cognitive Psychology: Connecting Mind, Research and Everyday Experience (2nd ed.). Wadsworth. Gomez, R., Posener, J., Keller, J., DeBattist a, C., Solvason, B., & Shatzberg, A. (2009). Effects of Major Depression Diagnosis and Cortisol Levels on Indices of Neurocognitive Function. Psychoneuroendocrinology 1012-1018. Goodyer, I., Dubicka, B., Wilkinson, P., Kelvin, R., Roberts, C., Byford, S., et al. (2008). A randomised controlled trial of cognitive behavior therapy in adolescents with major depression treated by selective serotonin reuptake inhibitors. Health Technology Assessment 12 (14).
126 Gorwood, P., Corruble, E., Falissard, B., & Goodwin, G. (2008). Toxic Effects of Depression on Brain Function: Impairment of Delayed Recall and the Cumulative Length of Depressive Disorder in a La rge Sample of Depressed Outpatients. Am J Psychiatry 731-739. Gundert, B. (2000). Soma and Psyche in Hi ppocratic Medicine. In J. WRight, & P. Potter, Psyche and Soma: Physicians and Me taphysicians on the Mind-Body Problem from Antiquity to Enlightenment (pp. 13-37). Oxford University Press. Hempel, C. (1950). Problems and Changes in the Empiricist Criterion of Meaning. Revue Internationale de Philosophie 41 41-63. Herbert, T., & Cohen, S. (1993). Stress and Immunity in Humans: A meta-analytic review. Psychosomatic Medicine 364-379. Hiroto, D., & Seligman. (1975). General ity of Learned Helplessness in Man. Journal of Personality and Social Psychology Hooker, C. (1981). Toward a General Theory of Reduction. Dialogue 20 (1), 38-59. Horst, S. (2007). Beyond Reductionism. Oxford University Press. Huttenlocher, P. (2003). Neural Plasticity: The effects of the environment on the development of the cerebral cortex. Cambridge University Press. Insel, T., & Quirion, R. (2005). Psychi atry as a clinical neuroscience. Journal of the American Medical Association 294 (17). Irwin, M., & Cole, J. (2005). De pression and Psychoneuroimmunology. In Human Psychoneuroimmunology (p. 251). Iwasaki, A., & Medzhitov, R. (2010). Regula tion of Adaptive Immunity by the Innate Immune System. Science 327 (291). Joseph Devlin, R. P. (2007). In Praise of Tedious Anatomy. Neuroimage 10331041.
127 Journal of Clinical Neuroscience. (2010). Clinical Neuroscience Research home Retrieved January 2nd, 2010, from Elsevier: http://www.elsevier.com/wps/find/jo urnaldescription.cws_home/621012/descripti on#description Kandel, E. (2000). Principles of Neuroscience. McGraw-Hill Medical. Kanter, J., Callaghan, G., Landes, S., Bu sch, A., & Brown, K. (2004). Behavioral Analytic Conceptualization and Treatment of Depression: Traditional Models and Recent Advances. The Behavior Analyst Today 5 (4). Karvoussi, R., Segraves, R., Hughes, A., As cher, J., & Johnston, J. (1997). Double-blind comparison of bupropion sustained release a nd sertraline in depressed outpatients. Journal of Clinical Psychiatry Kaye, J., & Lightman, S. (2005). Psychol ogical Stress and Endocrine Axes. In Human Psychoneuroimmunology. Kellermann, G. (2004). Urinary Measurement of Neurotransmitters. Retrieved January 18th, 2010, from NeuroScience: Improving Clinical Outcomes: https://www.neurorelief.com/ showarticle.php?A rticleID=436 Kempermann, G., Kuhn, H., & Gage, F. (1998). Experience-induced ne urogenesis in the senescent dentate gyrus. Journal of Neuroscience Kendler, K. (2001). A Psychiatric Dialogue on the Mind Brain Problem. American Journal of Psychiatry 989-1000. Knoblich, G., Stottmeister, F., & Kircher, T. (2004). Self-monitoring in patients with schizophrenia. Psychological Medicine, 2004, 34, 15611569 1561-1569. Lacasse, J., & Leo, J. (2005). Serotonin a nd Depression: A Disconnect between the Advertisements and the Scientific Literature. PLoS Med 2 (12).
128 Lomo, T. (2003). The Discovery of Long-Term Potentiation. The Royal Society Luhrmann, T. M. (2000). Of two minds: The growing diso rder in American psychiatry. New York: Knopf. MacQueen, G., Campbell, S., McEwen, B., Macdonald, K., Amano, S., & Joffe, R. (2003). Course of illness, hippocam pal function, and hippocampal volume in major depression. Proc Natl Acad Sci U S A 100 1387-1392. Madsen, P. (2007). How to Win Every Argument: The Use and Abuse of Logic. UK: Continuum International Publishing Group. Malberg, J., & Duman, R. (2003). Cell prolifer ation in adult hippocampus is decreased by inescapable stress: reversal by flouxetine treatment Neuropsychopharmacology 15621571. Marieke, T., Elzinga, B., Spinhoven, P., & Ever aerd, W. (2008). The Effects of Cortisol Increase on Long-Term Memory Retrieva l During and After Acute Psychosocial Stress. Learning, Memory and Psychopathology Mayo Clinic Staff. (2008). Cushing's Syndrome Retrieved 2009, from MayoClinic.com: http://www.mayoclinic.com/health/cushings-syndrome/DS00470 McLaughlin, B., & Bennett, K. (2005). Supervenience. Retrieved 2009, from Stanford Encyclopedia of Philosophy: http://plato.stanford.e du/entries/supervenience/ Mendels, J., Stinnet, J., Burns, D., & Frazer, A. (1975). Amine Precursors and Depression. Arch Gen Psychiatry 32 22-32. Milner, B., Squire, L. R., & Kandel, E. ( 1998). Cognitive Neuroscience and the Study of Memory. Neuron 20 445468. Mohammed, A., Zhu, S., Darmopil, S., Hjerling-Leffler, J., Ernfors, P., Winblad, B., et al. (2002). Environmental enrichment and the brain. Prog Brain Res
129 Motivala, S., Sarfatti, A., Olmos, L., & Irwin, M. (2005). Inflammatory markers and sleep disturbance in major depression. Psychosomatic Medicine 67 187-194. Nagel, E. (1961). The Structure of Science: Proble ms in the Logic of Scientific Explanation. Neisser, U. (1967). Cognitive Psychology. Prentice Hall. Oxford English Dictionary. (2009). Reductionism. Retrieved February 1st, 2010, from Oxford English Dictionary: http://dictionary.oed.com/cgi/ent ry/50200344?single=1&query_type=word&quer yword=reductionism&f irst=1&max_to_show=10 Palacios, R., & Sugawara, I. (1982). Hydrocorti sone abrogates prolifer ation of T cells in autologous mixed lymphocyte reaction by rendering the interleukin-2 producer T cells unresponsive to interleukin-1 and una b le to synthesize the T-Cell growth factor. Scand J. Immunol. 25-31. Pariante, C., & Miller, A. (2001). Clucocortico id receptors in major depression: relevance to pathophysiology and treatment. Biological Psychology 49 391-404. Rabin, B. (2005). Introduction to Immunology a nd immune-endocrine interactions. In K. Vedhara, & M. Irwin (Eds.), Human Psychoneuroimmunology. Oxford University Press. Ramsey, W. (2007). Eliminative Materialism. Retrieved 2009, from Stanford Encyclopedia of Philosophy: http://plato.stanford. edu/entries/materialismeliminative/ Rorty, R. (1965). Mind-body identi ty, privacy, and categories. Journal of Philosophy 198-9.
130 Roth, R. M., Randolph, J. J., & Nancy Koven, P. I. (2006). Neural Substrates of Executive Functions: Insights from Func tional Magnetic Resonance Imaging. In D. Joshua (Ed.), Focus on Neuropsychology Research. Nova Science Publishers. Sapolsky, R. (2000). Glucocorticoids and hippocampal atrophy in neuropsychiatric disorders. Research in Gen Psychiatry 57 (10), 925-35. Sarkar, S. (1992). Models of Reduction and Categories of Reductionism. Synthese 91 167-194. Selye, H. (1951). The Gene ral-Adaptation Syndrome. Annual Review Medicine 327342. Tausk, F., Elenkov, I., & Moyni han. (2008). Psychoneuroimmunology. Dermatologic Therapy 23-32. Tye, M. (2007). Qualia. Retrieved 2009, from Stanford Encyclopedia of Philosophy: http://plato.stanfor d.edu/entries/qualia/ Uttal, W. (2009). Distributed Neural Systems: Beyond the New Phrenology. Sloan Publishing. Uttal, W. (2001). The New Phrenology. Boston: Bradford. Vedhara, K., & Wang, E. (2005). Assessmen t of the Immune System in Human Psychoneuroimmunology. In Human Psychoneuroimmunology. Oxford University Press. Vogeley, L., Bussfeld, P., Newen, A., Herrm ann, S., Happe, F., & Falkai, P. (2001). Mind reading: neural mechanisms of theory of mind and self-perspective. Neurimage 14 Warner-Schmidt, J., & Duman, R. (2006). Hi ppocampal Neurogenesis: Opposing Effects of Stress and Antidepressant Treatmenbt. Hippocampus 239-249.
131 Zorilla, E., Luborsky, L., McKay, J., Rosentha l, R., Houldin, A., Tax, A., et al. (2001). The Relationship of Depression and Stre ssors to Immunological Assays: A MetaAnalytic Review. Brain, Behavior, and Immunity