Sunday, April 01, 2001

Imaging: High Tech Cannons, Phantom Targets?

The New Phrenology: The Limits of Localizing Cognitive Processes in the Brain

By: William R. Uttal Ph.D.

Both the local newspaper and the international research journal hail the latest triumph: neuroimaging shows the spot in the brain where yet another mental process takes place—say, perceiving a single word or experiencing fear. But wait, says psychologist and engineer Uttal: We may be misusing this powerful technology. In this excerpt from The New Phrenology, he argues that the quest for increasingly detailed localization may be doomed to failure if we cannot first ask the right questions.

 exc_v3n2uttal_1

Do you want to know where that idea (or emotion, or talent, or impulse) arises in the brain? Just wait for the next report from the imaging laboratory. As fast as brain scientists and their postdocs can get a turn on the PET scanner or MRI machine, a new taxonomy of the mind is being drawn on our skulls.

Semantic processing takes place there, romantic love is here, murderous impulses are over there.

But wait, wasn’t there once something called phrenology? Actually, there still is, argues William R. Uttal, Ph.D., an industrial engineer at Arizona State University, who is also professor emeritus of psychology at the University of Michigan. He calls it the new phrenology in a book to be published this summer by The MIT Press.

The New Phrenology: The Limits of Localizing Cognitive Processes in the Brain examines a premise with deep roots in psychological science: the localization hypothesis. Uttal asks: “Can psychological processes be localized, that is, can they be adequately defined and isolated in a way that permits them to be associated with particular brain regions?” If they cannot, a few brain journals are out of business. With noninvasive techniques such as PET, MRI, and other more exotic technologies widely available, brain scientists (and psychologists, anthropologists, criminologists, and art historians, as well) are looking for the home addresses in our brains of visual imagery, short-term memory, spatial thinking, mathematical thinking, color detection, attention, verbal memory, and dozens of other mental processes.

The technology, for all its potential, is not without its problems, Uttal suggests, but the big unknown is on the other side of the equation. How solid are our definitions of the dozens of mental processes we would like to localize in the brain?

In this excerpt from The New Phrenology, Uttal lays out the problem and takes us on a guided tour of 2000 or more years of changing taxonomies of human mental processes, asking if our era has somehow hit on the right taxonomy. If PET had existed 30 years ago, would imagers be seeking the location of the Id, the Ego, and the Superego? And can our emerging brain map be any better than our map of the mental landscape?

EXCERPT

From The New Phrenology: The Limits of Localizing Cognitive Processes in the Brain by William R. Uttal. To be published by The MIT Press, July 2001. Used by permission of The MIT Press.

In recent years, noninvasive tomographic techniques using radioactivity, X rays, or magnetic fields, and even more exotic methods have allowed us to peer into the human brain while it is actively engaged in mental activities. As computational procedures have improved our ability to extract three-dimensional images of brain anatomy and, some have argued, perhaps even of mental functions, more and more psychologists have applied these powerful procedures to the localization question...

Much of this work on the localization of psychological functions in the brain seems to have plunged off the rock of scientific certainty into a lake of unknowns with an exuberance typical of a science suddenly provided a powerful new tool—or, perhaps, of a child given a new toy. Unfortunately, all too often inadequate attention has been paid to the conceptual, logical, and technical considerations that should accompany any scientific undertaking...

Indeed, so abundant, provocative, and perplexing have the findings of the new technologies and procedures been that it is time to step back and ask fundamental questions. What does it all mean? What can be accomplished? What are the capabilities of this particular device or procedure? What are its limitations?...

We must also consider what our definitions of psychological functions and processes themselves signify. It is here that the greatest impediments to understanding the relation between psychological constructs and brain mechanisms actually lie...

I approach this work with considerable trepidation. The scientific community at large has become so thoroughly enamored of the localization hypothesis that any single critical voice is likely to be drowned in the sea of current enthusiasm...

One has only to examine the contents of one of the world’s leading scientific journals— Science—to appreciate the impact and popularity of the current effort to use imaging techniques to locate psychological processes. Most of the articles from 1998 to mid-2000 that deal with mental processing in any way include some fMRI or imaging data. Conventional psychophysical studies based on purely behavioral observations are almost completely absent. Moreover, some of our most prestigious universities have replaced traditional cognitive (previously known as “experimental”) psychology programs with cognitive neuroscience programs that completely accept both the assumptions and the findings of the search for the cerebral localization of hypothetical cognitive modules with imaging techniques...

LOCALIZING COGNITIVE PROCESSES: THE PROBLEM

Can psychological processes be adequately defined and isolated in a way that permits them to be associated with particular brain locales? Embedded within this question are three subquestions:

  1. Can the mind be subdivided into components, modules, or parts?
  2. Does the brain operate as an equipotential mass or is it also divisible into interacting but separable functional units?
  3. Can the components, modules, or parts of the mind, if they exist in some valid psychological sense, be assigned to localized portions of the brain? 

None of these questions is new. Rather they have been at the heart of much of scientific psychology’s research throughout its history. Indeed, historically, the respective answers to subquestion 1 are the bases of many of psychology’s great schisms. For example, “mentalisms” assume cognitive analyzability and “behaviorisms” typically eschew it. Given the fundamental conceptual importance of these issues, it is surprising that questions of this kind have seldom been studied in detail in recent years. In particular, modern cognitive neuropsychology bases a substantial amount of current research on the a priori assumption that affirmative answers to all three of these subquestions are justified....

The localization of cognitive functions in the brain is not just an esoteric or arcane exercise; it expresses a profound and deeply important point of view. 

The localization of cognitive functions in the brain is not just an esoteric or arcane exercise; it expresses a profound and deeply important point of view. It should be understood, however, that by championing psychobiology or cognitive neuroscience in this manner, I am in no way reneging on my promise to provide a critical analysis of its findings and conclusions, as well as of its strengths and weaknesses. The practical concerns revolving around neuroscientific research make it essential that we understand its limitations and misdirections as well as its progress and successes. Localization studies, for example, have significant bearing on both neurosurgical procedures and popular notions about child development and rearing. Inferences drawn from these studies all too easily go far beyond the legitimate implications of what has been discovered in the laboratory. How many of us have heard about the putative differences between the “left: and “right” brain and the pedagogic techniques that have “followed” from the work on split-brain patients? On the other hand, how few of us are familiar with critiques challenging the entire concept of hemispheric specialization? Efron..., for example, suggests that the “thriving ‘low tech’ cottage industry” of psychological research purportedly into the specialized functions of the cerebral hemispheres is based on “performance asymmetries” that are, in fact, incapable of distinguishing between the separate functions of the hemispheres. Thus, left ear-right ear behavioral differences and left hemiretina-right hemiretina phenomenal differences, uncritically interpreted as differences between left and right cerebral hemisphere functions, cannot be scientifically correlated. To assume that they can leads to erroneous scientific theories and practical applications. In the classroom, similar and perhaps even more dangerously silly notions of “training a particular part of the brain” reflect a profound ignorance of the way in which most parts of the brain interact to produce unified mental outcomes.

The treatment of childhood hyperactivity with drugs of unknown mechanism and questionable efficacy...also suggests that we should step back and examine contemporary views of brain-mind relationships. Whereas the dangers of incorrectly localizing a cognitive function on the neurosurgeon’s operating table are self-evident, the dangers in the classroom are not so obvious.

It should be understood that the unity of subjective experience therefore contradicts the hypothesis of cognitive localization in a fundamental way. 

There is thus an enormous chasm between what can and cannot be done to solve the localization problem, on the one hand, and what has actually been concluded from well-intentioned, but inadequately reasoned research, on the other. Whatever specialization and localization may occur in the brain, and however isolated the dimensions of a sensory code may be in the peripheral transmission pathways, it seems far more likely that the mind, consciousness, or self-awareness represent the merging or binding of many different underlying processes and mechanisms into an integrated singular experience. It should be understood that the unity of subjective experience therefore contradicts the hypothesis of cognitive localization in a fundamental way...

“ASSOCIATED” WITH WHAT?

It became clear in the twentieth century that large regions of the brain, called “association” or “intrinsic areas,” seemed not to be associated with an particular sensory or motor function. The immediate speculation was that these were the areas where cognitive activities such as thinking, perceiving, decision making, problem solving took place—where the raw sensory information was evaluated, transformed, and linked to the motor regions. It is here that much of the crucial and most controversial work on brain localization is being done. It is also here that proving localization of mental functions becomes most difficult. For one of many reasons, the anchors to the outside world are least defined or measurable in these areas. By contrast, the sensory regions are well anchored to the physical energies used as stimuli; the motor regions are equally well anchored to the physical mechanisms of motor responses or even measurable glandular secretions. The mysterious association areas, however, are associated with constructs, metaphors, mental events, and perceptual phenomena that are much less defined or measurable. Nevertheless, much has been suggested about the role of these areas. Indeed, evaluating those suggestions is the central focus of this book.

The idea that the brain consisted of a very large number of modules with distinguishable functions and influences has not always been uncritically accepted. The major opponent to the idea of localized modules in the brain in recent times was Lashley..., whose studies of learning in rats led him to conclude there was no specific brain locus for the neural changes (the “engram”) that must have occurred during behavioral changes produced by experience. Rather, Lashley championed a “mass action” or “equipotentiality” theory: it was simply the gross amount of cortical tissue remaining after an ablation [or surgical removal of a section] that correlated with the presence or absence of learned behavior. It is now believed that it was the complexity of the process Lashley used to study the learning behavior of his rats and the recruitment by that process of so many areas of the rat’s brain that made the brain’s activity with regard to such behavior seem equipotential.

Although many psychobiologists no longer subscribe to Lashley’s idea of mass action, the notion of widely distributed representation of complex cognitive processes still has significant support. For example, the definitive and now classic study reported by Olds, Disterhoft, Segal, Kornblith, and Hirsch (1972) argued strongly for the idea that even such an apparently simple learning paradigm as classical conditioning is capable of activating widely distributed regions of the brain.

The controversy between mass action and localized function is not an active arena of debate these days. There is little current opinion or theory arguing that the brain is homogeneous or equipotential or that it operates with any semblance of mass action. Rather, contemporary theory and research is generally characterized by the assumption of discrete and specialized functions in various regions...

The more complex the psychological process, the less likely it is that a narrowly circumscribed region uniquely associated with that process will be found. 

There is thus no question that specialized sensory and motor regions exist and that other regions are involved in, if not dedicated to, particular cognitive processes in some yet-to-bediscovered ways. On the other hand, however, it seems clear that the more complex the psychological process, the less likely it is that a narrowly circumscribed region uniquely associated with that process will be found. Much of the rest of this book is aimed at providing support for this assertion...

Although it may seem at first that this extreme difficulty arises solely from the limitations of recording techniques or from the brain’s inherent complexity, I argue that, on the contrary, the preeminent problem in achieving a general solution to the localization issue lies in defining the psychological processes and mechanisms for which loci are being sought. The ease and precision of defining a spatial location of a region in the brain contrasts starkly with the difficulty of defining even the simplest of psychological constructs. Just consider for a moment what “looking at something” may mean. It involves not just the sensory encoding of visual signals but, at the least, memory, perceptions, recognition, and semantic interpretation in diverse and unknown ways. If our psychological science and technology allowed us to parse an act as complex as, say, reading into components, perhaps solving the localization problem could move ahead more rapidly than it currently does. However, defining and parsing even the simplest mental process is fraught with technical and conceptual difficulties of enormous proportions. What are the dimensions of consciousness, of imagination, or even of some of the operationally best defined and superficially simplest constructs that emerge as experimental outcomes? For example, do the processes often proposed....as the underlying components accounting for measured reaction times—stimulus identification, responses selection, and response evocation—have any independent psychobiological reality? Questions like these are not easy to answer despite the prevailing assumption that such components are real entities and not just hypothetical, post hoc constructions. As we shall see repeatedly...and as evidenced throughout the history of psychology, some of the putative mental and cognitive processes suggested as candidates for localization experiments are ephemeral, indeed...

If what one seeks to localize is itself nebulous and ill defined, finding a particular brain locus in which it is uniquely or even partially represented is going to pose a considerable challenge. 

IS A TAXONOMY OR EVEN A LEXICON OF COGNITIVE PROCESSES POSSIBLE?

The great interpreter and historian of experimental psychological thought E. G. Boring is reputed to have said, “Intelligence is what intelligence tests measure.” Herein lies the kernel of the greatest problem faced by psychology in general and the essence of the difficulty one faces when one attempts to localize mental processes or components in particular parts of the brain. The difficulty is that, beyond the physically anchored sensory and motor processes, it has always been extremely difficult to define exactly what is meant by a psychological, cognitive, or mental term. Yet, in the search for the brain loci responsible for psychological processes, we glibly go about trying to locate some such nebulously defined entity in the three-dimensional space inhabited by the brain. Clearly, if what one seeks to localize is itself nebulous and ill defined, finding a particular brain locus in which it is uniquely or even partially represented is going to pose a considerable challenge—a challenge fraught with its own conceptual, technical, and logical difficulties. This criticism holds true no matter how precisely the coordinates of the brain regions themselves may be defined.

How do we deal with this formidable obstacle—ill-defined mental modules, faculties, traits, or components? One way is to examine the various psychological processes and mechanisms that have been proposed over the years to see if there is any convergence toward a consensus on what these mental components might be. In other words, is there any scientific basis for a reductive classification of mental components that permits us to develop an orderly taxonomy based on sound cladistic logic?...

There is probably nothing that divides psychologists of all stripes more than the inadequacies and ambiguities of our efforts to define mind, consciousness, and the enormous variety of mental events and phenomena. All of us have been involved in disputes that are ultimately seen to founder on the definition of mentalist terminologies...

Unlike lepidopterists, who have the relatively simple task of gathering and classifying butterflies, psychologists have few such convenient physical anchors. Organizing the myriad proposed psychological components— “butterflies”—of our minds has been and is one of the great unfulfilled challenges of our science. Indeed, it is not only unfulfilled; it has not, in my opinion, been adequately engaged. Rather, hypothetical psychological constructs are invented ad lib and ad hoc without adequate consideration of the fundamental issue of the very plausibility of precise definition.

Where entomologists have objective measures such as the number of legs or wings and the physical types of antennae (as well as the new measures offered by DNA analysis), psychologists have far less objective metrics on which to base our definitions. Indeed, the hugely disorganized body of processes, ideas, phenomena, and other ephemeral entities that psychology confronts is probably far more bewildering than the 1.5 million species of insects currently thought to exist. Psychology does not even have simple dimensions such as “number of” or “shape of” along which to classify and organize our specimens...

The centerpiece of a point of view about the nature of the human mind in one generation is forgotten or ridiculed in the next. Where today, indeed, would one find even the most enthusiastic fMRI researcher searching for Freud’s “Id” or Gall’s “acquisitiveness?”  Nevertheless, psychologists are seeking to localize some equally tentative and perhaps equally transitory new hypothetical constructs among the many centers and nuclei of the brain. Indeed, vaguely defined concepts such as language, emotion, or consciousness, and other mental processes of uncertain specificity and existence are all too easily deemed both specific and “real” enough to be localized in a particular portion of the brain. 

The effort to organize mental processes into some kind of classificatory order is not new. Each generation’s psychologists or protopsychologists seem to have suggested one based on their own contemporary worldview. 

PREVIOUS TAXONOMIES

The effort to organize mental processes into some kind of classificatory order is not new. Each generation’s psychologists or protopsychologists seem to have suggested one based on their own contemporary worldview. Certainly, from the dawn of language, and perhaps even from earlier methods of interpersonal communication, a rich vocabulary or set of nonverbal symbols of mental events existed that described experiences, feelings, memories, intents, desires, emotions, and other kinds of thought...

In what is still one of the best histories of scientific psychology, Klein (1970) points out that Maimonides (1135-1204) had developed a surprisingly complete and modern-sounding list of mental faculties...

Later, Thomas Hobbes (1588-1679) and Rene Descartes (1596-1650) offered their mechanistic theories of the nature of humanity and its relation to other parts of our world. In particular, it is often overlooked that the father of the reductivist method, Descartes, championed the idea that the mind was more or less indivisible.

From these earlier ideas emerged the early associationist theories of the British empiricists including John Locke (1632-1704), David Hartley (1705-1757)...among many others...

By the time of the introspective structuralists—Wilhelm Wundt (1832-1920), Edward Titchener (1867-1927) and the philosopher James Ward (1843-1925)—and for much of this century, the prevailing taxonomy included a long list of basic sensations that had to be combined by associationist processes to produce more complex perceptions, thoughts, feelings, and emotions. Even Watson, the great behaviorist (who, as a behaviorist, should have been more committed to a holistic point of view) adhered to the idea that the higher levels of cognition were constructed in some way from the components of stimulus and response.

An avalanche of new vocabulary and ideas championing the separateness of the components of mental activity, however, occurred with the emergence of modern “cognitive” psychology, the newest version of traditional mentalism. The main difference between behaviorism and cognitive mentalism is their respective attitudes toward the question of the accessibility of mental processes. The behaviorists asserted that mind is not directly accessible, whereas the cognitivists asserted that it is.

A second issue distinguishing these two great schools of thought revolves around the analyzability of mental processes. Behaviorists argue that mental processes are unitary and that externally measured responses do not therefore allow us to analyze mind into its components, even if they do exist (an issue on which behaviorists of different stripes continue to disagree). The core of much cognitive thinking in recent years is just the opposite—cognitivists very strongly argue that mind is made up of a system of more or less insolatable and independently assayable components. Unlike behaviorists, therefore, cognitivists accept that it is plausible—indeed, sensible—to search for the internal structure and architecture of the mind as well as that of the brain...

The mentalist theory that mind is made up of separable and identifiable components is clearly stated in a recent book by Pinker (1997):

The mind is a system of organs of computation, designed by natural selection to solve the kinds of problems our ancestors faced in their foraging way of life, in particular, understanding and outmaneuvering objects, animals, plants, and other people... The mind is organized into modules or mental organs, each with a specialized design that makes it expert in one arena of interaction with the world. The modules’ basic logic is specified by our genetic program...

Pinker is crystal clear here on his commitment to the “modularity and separability of mental processes” even if he does not provide us with the empirical justification for his version of contemporary cognitive psychology. On the basis of that commitment, he vigorously supports the analytical approach of modern cognitivism— and presumably also the localization effort. In doing so, however, he also inadvertently highlights the great difficulty faced by localization theorists—the still unsatisfied need to sharply demarcate that which is to be localized.

What we see exemplified in Pinker...is the a priori assumption that mental modules and “organs” exist and can therefore provide a conceptual basis for the entire localization enterprise. This is done in the face of a surprisingly large amount of empirical evidence to the contrary. I argue here that, by accepting this a priori assumption, psychology abdicates its responsibility as a profession to evaluate the logic of its primary fundamental premises—the accessibility and the analyzability of mental activity...

Most psychologists at work today laugh at the “bumps on the skill” idea that Gall and Spurzheim (1808) offered as the main tenet of phrenology or, at it was sometimes called, “craniology.” This evolutionary outgrowth of the preexisting faculty psychology tradition seems, at first, a distant and antique idea that should best be classified among the charlatanisms of the nineteenth century. However, there are two things that suggest that such an offhand rejection of their teaching is not in order. First, few psychologists today appreciate the enormous popularity of the movement. At its time, phrenology was ranked with Darwin’s theories, at least in the popular mind. Although many scientists of the time severely criticized it, they provided at least a modicum of scientific respectability to phrenology. We must acknowledge Gall and Spurzheim’s proposal of one of the first specific theories of brain localization. Right or wrong (and they were certainly wrong with regard to almost every technical issue), they did associate particular brain (and to their eventual and ultimate ridicule) skull regions to a very specific set of psychological properties or, as they had come to be known by that time—“faculties.” Although one may challenge this fundamental assumption (as I do in this book), this idea still has wide, if cryptic and implicit, acceptance in modern cognitive neuroscience...

INTELLIGENCE AND THE FACTOR ANALYSIS APPROACH

The analysis of mental activity into components reached what many consider to be its apex with the development of the powerful mathematical tool called “factor analysis” by workers such as Spearman (1904)... Each of these psychologists had their repertoire of research tools broadened by their unusual (for psychologists) mathematical training. The main idea behind the factor analysis approach is to carry out a series of “ability” tests and then to develop a matrix of the correlations that exist between all pairwise combinations of these tests. Correlations among the various tests are supposed to indicate the presence of common mental abilities that account for their common scores. The correlation matrix is then transformed to reduce the variance among these correlations into a much smaller set of “factors” that presumably map onto the underlying mental abilities...

Although there are many special problems associated with the application of factor analysis to intelligence testing and the study of individual differences that I do not consider in detail here, the method represents a persistent and popular theme of psychological research. Like many of the other taxonomies discussed, the factor analysis approach also seeks to uncover a set of real mental components. Perhaps better than any other, however, it represents a rigorous and formal method for suggesting what these mental components might be...

Another contemporary psychologist who has developed a specific taxonomy of mental process is [Howard] Gardner (1983), whose taxonomy includes the following “kinds” of intelligences:

Object-free intelligence

Linguistic, Musical

Object-related intelligence

Logical-mathematical, Spatial, Bodily-Kinesthetic, Person-related, Interpersonal, Intrapersonal

To which an eighth category has been added (Gardner, 1999): Naturalist

As specific as Gardner is in classifying the various kinds of intelligences, he hesitates to apply such a scheme to other aspects of human mentation. Indeed, his argument is that the categories in his taxonomy are only valuable when one is concerned with how we process objects in the world, and that to extend this classification system to mental processes such as “personality, motivation, will, attention, character, creativity, and other important and significant human capabilities...may be ill advised.” Again, even though it is limited to only one domain of psychological processes—“intelligences”—Gardner’s taxonomy clearly provides another example of a modern “modular” system.

Psychology abdicates its responsibility as a profession to evaluate the logic of its primary fundamental premises—the accessibility and the analyzability of mental activity.

THE EMERGING “MODERN” IMAGING TAXONOMY

Finally, there may be no better way to end this discussion of classic and modern taxonomies of mental components than to consider the various processes that have been the object of the localization quest by means of imaging techniques. The database of cognitive neuroscience is itself an implicit taxonomy of mental modules and components. The variety of mental activity invoked in this quest is voluminous and the following list is far from exhaustive, but it does give an idea of the enormous span of current research attention...

In providing a very limited sample of the mental processes being investigated with PET or MRI imaging techniques, my purpose is to show that a new vocabulary and thus, implicitly, a new taxonomy is emerging, one unfortunately based on ad hoc and methodological criteria rather than on the logical structure required of a solid cladistic system:

  1. Kosslyn et al. (1999) have studied visual imagery.
  2. Lumer, Friston, and Rees (1998) have studied the brain loci of perceptual rivalry.
  3. Spitzer et al. (1996) have attacked the problem of semantic information processing.
  4. Smith and Jonides (1999) have explored the role of frontal cortex in short-term memory storage and the executive processes that act on that stored information.
  5. Buchel, Coull, and Friston (1999) have studied associative learning.
  6. McIntosh, Rajah, and Lobaugh (1999), on the other hand, have explored sensory learning.
  7. Kopelman, Stahopoe, and Kingsley (1997) have dissected temporal and spatial context memory into different processes represented by different regions.
  8. Dehane, Spelke, Pinel, Stanescu, and Tsivkin (1999) have studied the basis of mathematical thinking, implicating both linguistic competence and visuospatial representations.
  9. Gulyas and Roland (1994...) have studied binocular disparity detection, form, and color.
  10. Roland and Gulyas (1992) have studied visual pattern discrimination and the formation of visual memories.
  11. Kanwisher, McDermott, and Chun (1997), among others, have designated one or another locus in the brain as a face perception area.
  12. Elfgren and Risberg (1998) have studied verbal fluency and “design” fluency.
  13. Petersen, Fox, Posner, Mintum, and Raichle (1988) have studied the brain correlates of linguistic single word processing.
  14. Blakemore, Rees, and Firth (1998) have studied the differential brain effects of expected and unexpected events.
  15. Kastner, De Weerd, Desimone, and Ungerleider (1998) have studied directed attention.
  16. Tranel, Damasio, and Damasio (1997) have sought to determine which brain regions are involved in the retrieval of conceptual knowledge, that is, the categories of concrete things.
  17. Barch et al. (1997) have assigned working memory and task difficulty to separate brain regions.
  18. Cohen et al. (1997) have also explored working memory.
  19. Smith, Jonides, Marshuetz, and Koeppe (1998) have studied verbal working memory.
  20. Kapur et al. (1996) have studied intentional learning of verbal materials.
  21. Lane et al. (1997) have assigned pleasant and unpleasant emotions to different combinations of parts of the limbic system.
  22. Ploghaus et al. (1999) have used tomographic images to distinguish between pain and the anticipation of pain

It is important to keep in mind that what we are concerned with here is the relationship between the mental or cognitive processes and brain regions. PET, MRI, and other imaging techniques have been used for many useful and valuable medical, surgical, and research purposes. It is the questionable application of these techniques to the localization of psychological functions that is the principal focus of this book’s discussion.

Nevertheless, and with specific regard to the psychobiological issue at hand, this list, partial and incomplete as it is, does make one strong point. There are a large number of more or less inadequately defined mental components and activities that have been associated with particular regions of the brain by PET and MRI techniques and the list is growing rapidly. I contend that the recent wave of research using imaging techniques has led to a proliferation of hypothetical psychological, cognitive, or mental components without a proper foundation of clear-cut answers to the questions of accessibility and analyzability.

Implicit in this and all other lists and taxonomies presented in this chapter is the central assumption of analyzability. That is, it is increasingly clear that the a priori assumption that the mind can be finely divided into any number of components or modules is alive and well. The current approach has deep intellectual roots and has been at the core of thinking about mental processes since philosophers and scientists began to consider these issues.

Compartmentalization of the mind into a set of components is a convenient concept to help us organize what would otherwise have to be considered a composite and integrated as well as unified “mind.” Indeed, it may be that it is extremely difficult to study mentation in any other way. Yet exactly because of this pragmatic convenience and ease, it is possible that we are being led astray from a truer, more valid, and more realistic conceptualization of the unified nature of mental processes based on widely distributed brain mechanisms. If our search is for truth and an actual rather than a convenient or pragmatic understanding of mind and brain, then it may be necessary to consider what the alternative— an indivisible, composite, aggregated, and unanalyzable mind—might be like. This is the main alternative raised by those who argue against the uncritical acceptance of mental analyzability and the cerebral localization of high-level cognitive (i.e., nonsensory and nonmotor) processes...

If our search is for truth and an actual rather than a convenient or pragmatic understanding of mind and brain, then it may be necessary to consider what the alternative—an indivisible, composite, aggregated, and unanalyzable mind—might be like.

The main alternative to the analyzability hypothesis is that the proposed “components” of mental activity are actually properties of some unified mind—rather than distinct entities. By implication, the modules or components of mind that have been invoked for centuries are simply manifestations of the procedures and methods used to measure behavioral functions, that is, any given mental module is a manifestation of just one of the infinitely many ways of measuring the complex, but unitary process traditionally called “mind.” For example, the widespread acceptance of a family of different types of memory (e.g., short-term, working, and long-term memory) depends in large part on the methods used to study the retention of information...

I argue that, no matter how useful it may be or how entrenched in scientific psychology, the idea of separate mental components or modules is deeply flawed. At the very least, the whole question should be reconsidered as a research issue rather than simply accepted as an a priori fundamental premise of our science.

CONCLUSION

Although it may be that it is far easier just to assume faculties and components than it is to search for a conclusive answer to the question of their existence, there is one body of evidence that suggests an indirect answer to this question. As I look over the studies reviewed in this chapter, one thing becomes completely clear. There is no historical trend, no evolutionary sequence, no convergence toward a taxonomy of mental components. We seem to have no community memory; the components suggested in one generation are largely forgotten in the next...

If there is no impending—indeed, no possible—agreement on what mental components are being measured or what a valid taxonomy of mental components should look like, what credence can be given to the quest to localize such phantoms? If mental components are only a semantic convenience, but have neither scientific validity nor the promise of being scientifically validated, what can be made of the ultraprecise maps showing sharply demarcated “hot spots” of neural activity when “pure” forms of these mental components are supposedly activated? Obviously, much needs to be done to answer these questions and to properly interpret the often fascinating findings from the imaging laboratories. It is likely that what these findings will turn out to mean will be far different from what we now believe they mean.



About Cerebrum

Bill Glovin, editor
Carolyn Asbury, Ph.D., consultant

Scientific Advisory Board
Joseph T. Coyle, M.D., Harvard Medical School
Kay Redfield Jamison, Ph.D., The Johns Hopkins University School of Medicine
Pierre J. Magistretti, M.D., Ph.D., University of Lausanne Medical School and Hospital
Robert Malenka, M.D., Ph.D., Stanford University School of Medicine
Bruce S. McEwen, Ph.D., The Rockefeller University
Donald Price, M.D., The Johns Hopkins University School of Medicine

Do you have a comment or question about something you've read in CerebrumContact Cerebrum Now.