REVIEWS OF THE CONTINUITY OF MIND

Choice Reviews Online 7/16/10 11:16 AM Print Full text 1. 45-4685 BF311 2006-5937 CIP Social & Behavioral Sciences \ Psychology Spivey, Michael. The continuity of mind. Oxford, 2007. 428p bibl index afp (Oxford psychology series, 40) ISBN 0-19-517078-4, $69.95; ISBN 9780195170788, $69.95. Outstanding Title! Reviewed in 2008apr CHOICE. In this fascinating book, Spivey (Cornell Univ.) challenges the long-standing practice of focusing on behavior as a sequence of perception, cognition, and action based on discrete stimulus-response events. Instead, he argues for the continuity of "mental trajectories" and presents a theoretical, methodological, and empirical overview of the flow of thought and action. In wonderful, lucid prose, Spivey presents the conceptual tools for tracking cognitive processing and examines the temporal dynamics of categorization, language comprehension, visual perception, action, and reasoning. The book is scholarly (the bibliography spans 65-page and includes nearly 1,300 references) and demanding, but well worth the effort. An invaluable resource for those interested in psychology, cognitive neuroscience, computer science, philosophy, and related fields, this volume will serve both the present and the next generation of cognitive scientists. Summing Up: Essential. Upper-division undergraduates, graduate students, and researchers. -- G. B. Rollman, University of Western Ontario © American Library Association. Contact permissions@ala-choice.org for permission to reproduce or redistribute. http://www.cro2.org/printcontent.aspx Page 1 of 1 8 of 8 people found the following review helpful: Deep, great, diverse book, September 25, 2008 B y psych reader This review is from: The Continuity of Mind (Oxford Psychology) (Paperback) M. Spivey presents some deep material. The book is just filled with evidence for a very core idea: Our understanding of the mind should be based on continuous functions, and not on the choppy file-copy concepts of a computer system (like the kind I'm typing this on). The book taps into areas that span much of what would satisfy very diverse interests: perception, language, problem solving, consciousness, with neuroscience stuff everywhere. And for any geeks, he supplies programming code at the back of this book for you to run some of the theoretical model systems he uses to demonstrate continuous theories. In short, the book is deep and interesting and will entertain and interest anyone interested in understanding the mind. 4 of 4 people found the following review helpful: 5.0 out of 5 stars A New Psychology, October 3, 2009 B y J. B. - See all my reviews This review is from: The Continuity of Mind (Oxford Psychology) (Paperback) Ever since Jerry Fodor published the widely influential book The Modularity of Mind in the 1980s, continuing from the cognitive revolution in the 1950s and 1960s, the idea that the mind works like a massively modular information processing system (namely, a computer) has been extremely popular. More recently, popular psychologists such as Steven Pinker have endorsed an easily understood and largely popular computational theory of the way that the mind works, which implicitly assumes tricky philosophies such as the internalist view of the mind, and also is implicitly assumed by evolutionary psychology, a largely nativist way of looking at the way the mind has developed. In The Continuity of Mind, Michael Spivey introduces what is probably the most convincing account so far that the mind works an entirely different way. He proposed, as others have before him, that the mind is a continuous dynamical system embodied in its environment, with perception and action being blurred in a continuous cascade of "hungry" mental operations and actions. It's easy to see that this view is much more complicated than the straightforward modular account, but Spivey introduces the mathematics and the nitty-gritty aspects of this fascinating theory easily and without tripping over himself too much. He argues for this embodied dynamical view of cognition with convincing evidence from fields such as visual perception and language, and finishes with hopeful hints that similarly convincing evidence will soon emerge in fields such as reasoning and decision making. The final two chapters on the book lay out the implications that this new and exciting view of the mind has for philosophy of mind at large, touching on concepts of free will, consciousness, and government. Spivey is eloquent and witty, and asserts his arguments with respect for the history that he is building upon. Since this book has come out, research has pointed to this embodied continuous dynamical system view of the mind fitting in quite well with social perception and decision making, and slowly people are beginning to realize that this dynamical systems account of cognition isn't quite as complicated and unintuitive as it has long been believed to be. In fact, one of Spivey's most convincing arguments is that this view of cognition is really much more valid neurophysiologically than its modular counterpart. Many people have argued for this view of cognition before, but Spivey's execution is absolutely perfect and I have no doubt that this book will be the centerpiece for a new study of how the mind works.! 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Nonlinear Dynamics, Psychology, and Life Sciences, Vol. 13, No. 1, pp. 151-154. © 2009 Society for Chaos Theory in Psychology & Life Sciences Book Review The continuity of mind, by Michael Spivey. Oxford: Oxford University Press, 2007. 427pp. ISBN 13: 978-0-19-517078-8. Kudos to Spivey—the self-proclaimed “zombie scientist” -- for helping stimulate the paradigmatic change needed in cognitive science. Spivey forcefully argues: for the field to progress, cognitive researchers need to acknowledge the wide-range of evidence for the continuous nature of mind, and embrace a hybrid approach consisting of Nonlinear Dynamical Systems (NDS), neural network (NN), ecological, and “embodied” perspectives. Within each, numerous models may be used to describe changing behaviors and cognitive states. Spivey narrows the possibilities by first, providing an overview of tools for tracking mental trajectories, and breaks these down (in Chapters 2, 3 and 4) into: conceptual, experimental, and simulation categories. Spivey then spends much of the book focusing on: (a) the NDS state space (i.e., attractor space) description of mental trajectories while carefully stepping readers through various ways to visualize these trajectories, and (b) a simple recurrence net (SRN) model, and how it captures key properties of behavior. Major themes of the book also include strong arguments against cognitive modularity (Chapter 5), and the need for cognitive research focused on temporal dynamics such as the many cited examples including work in: categorization, language comprehension, visual perception, action, and reasoning (Chapter 6). Later chapters include a unifying account of different aspects of cognition, and a final swipe at modularity theories of cognition. In the last chapter, Spivey entertains philosophical implications for consciousness, and his self-identification with “zombie” agents. Spivey”s focus on one version of recurrence analysis (SRN) has both strengths and weaknesses. The strength is in its simplicity and thus accessibility to wide audience. Two weakness: First, its short shrift given to the many other very useful NDS tools that can be used to uncover the dynamical structure of cognition. A future revision of this book might include a description of more NDS tools, or at least provide an informed rationale for why one would choose one approach over another. Pertinent here are considerations of dimensionality, and why some NDS researchers focus on low dimensional, prominent features of behavior (without attempting a link to biology) versus those choosing to study high dimensional behavior through systematic study of subtle patterns of behavioral variability (with the implicit acknowledgement of biological 151 152 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 NDPLS, 13(2), Aks complexity underlying emergence of cognition). Many NN models, for example, are better suited to high dimensional analysis by capturing complex interactions and emergent dynamics resembling those of the brain system. Spivey briefly notes importance of NN models, along with possible generating rules such as predator-prey search examples (p. 321). Further discussion of resulting scaling and power law behaviors would be very helpful along with the role they play in cognitive and other behaviors. A second weakness of SRN is in its key characteristic as a normalized “local” network, having features subject to similar criticism to which Spivey launches against mainstream cognitive science. SRN’s use of local averaging across vectors ignores a key attribute of the human brain: its distributed structure. Similarly, modification of synaptic weights “by hand” is a far cry from (Spivey”s claim for) self-organization. Whether studying key dimensions of overt behavior or the underlying complex system, common goal is to learn whether essential dynamical structure emerges spontaneously from iteration of these simple rules in a network of activated nodes. Spivey might also consider: (a) The relation between SRN and other low dimension descriptions of behavior such as simple differential equation (e.g., logistic equation as in Ward and West (1998), or harmonic oscillator as in Warren (2006). And (b) instead of collapsing to minimal-value functions in SRN approach, the same resulting behavior, along with its pattern of variations, can be described in terms of even simpler equations iterated over time to capture interactive dynamics of brain’s neural populations (e.g., cellular automata or self-organized criticality). The latter approach deliberately takes on the link between brain and behavior, and the fact that it is a complex system. Spivey defers to other sources on some of these issues (i.e., Ward, 2002), and to his credit, covers a vast array of research from the Cognitive literature to show support for continuous nature of mind and how it is “situated” (i.e., mediated) by both body and environment. Making the additional link to NDS is what makes Spivey’s work unique, and arguably its greatest strength aside from the excellent review of cognitive behaviors best described as embodied and continuous. My greatest reservation of the book is that its emphasis on continuity of mind is overstated. The book’s objective may have been better served by placing NDS or embodied cognition at the forefront. One apparent contradiction with the continuity thesis may be found in Spivey’s occasional reference to process of achieving stable mental states. These can easily be regarded as discrete states of perception and cognition. Thus, when describing mind as a trajectory— both continuous and discrete properties exist and can be accounted for in NDS framework. In fact, building a case for NDS theory of mind would be expected to include stable and unstable states as well as description of transitions brought on by changes in control parameters. Shelhamer (2006) is a pertinent source, (an unfortunate omission from Spivey”s book) illustrating excellent examples of transitional states in eye-movement and other behaviors. To his credit, Spivey acknowledges he may have overemphasized “continuous” nature of mind (p. NDPLS, 13(1), Book Review 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 153 322), but does so as “to help pry loose [entrenched] assumptions of stable representation.” In addition to NDS theory of mind not requiring that mind be continuous, it does not eliminate the need for functional-symbolic based understanding of how the mind works. In contrast to conventional stage-based theories of cognition and presumed linear (feed forward) sequence, research now convincingly shows that graded transitions, feedback, and recurrence are essential to high level cognition. Spivey describes this in a section on recurrence and “warping” probabilities where members of population code pass activity back and forth changing percent of activity until one set exceeds a threshold and a decision (response) is made. Similarly, symbol-based theories of cognition rely on improved understanding of such transitional states. Spivey, however, trivializes this point with his insistence on the continuity of mental states, and how discretization of these states is either an illusory or artificial imposition. This focus on continuity (mis)leads him to note that symbolic dynamics (SD) “may sound like an oxymoron” (p. 112), but then counters this by (correctly) pointing out legitimate SD approaches which implement symbols by neuronal assemblies. (Chapter 10 presents SD models from cognitive science such as Anderson’s ACT-R theory, and Complex Systems such as Crutchfield’s deterministic model). Spivey insists though that mental representations are not likely to resemble (or be converted) to rule-based logic. However, this is exactly what might emerge from transitions across different attractor states. Since symbols can be assigned and emitted by different transitional states in a dynamical system this is likely a lucrative path to our understanding of symbolic-rule characteristic of cognition. Noteworthy are the entertaining linguistic gems sprinkled throughout the book, These are quite revealing of Spivey’s expertise in cognitive linguistics and his effort to “lay siege to the ivory tower of consciousness” and assert that “everywhere that free will goes, it scorches the theoretical terrain.” One is left wondering whether the “functionalist view of mind teeters on edge of Cartesian dualism,” and is this in fact “Veja Du-alism all over again?” Many of these perceptual-cognitive “lines in the sand” might be dissolved by seeing that mental trajectories sometimes just “flirt with attractors…,” but do not reach conscious reportability. At other times, “attractors warp into repellors” permitting trajectory to continue on to new concepts.” Overall, in his grand challenge to the “pandemic of homunculitis ideas,” Spivey clearly and boldly expresses his ideas in a manner cognitive psychologists can understand without getting bogged down in mathematics. No doubt, impediments of inaccessibility arise from “modular” and “encapsulated” nature of different disciplines, with their unique jargon and tools. Spivey’s informal and “grounded” literary style—imbued with lots of clever metaphors makes for an enjoyable read even if you disagree with the premise of his ideas. Many are thought provoking and among these I am left pondering: Must one really be a modularist to believe in free will? Or is the true state of our mentalbeing closest to that of a zombie? 154 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 NDPLS, 13(2), Aks REFERENCES Shelhamer, M. (2007). Nonlinear dynamics in physiology: A state-space approach. Hackensack, NJ: World Scientific. Warren, W. (2006). Dynamics of perception and action. Psychological Review, 113, 358-389. Ward, L. M. (2002). Dynamical cognitive science. Cambridge: MIT Press. Ward, L. M., & West, R. L. (1998). Modeling human chaotic behavior: Nonlinear forecasting analysis of logistic iteration. Nonlinear Dynamics, Psychology, and Life Science, 3, 261-282. -- Deborah J. Aks Rutgers University Center for Cognitive Science (RuCCS) Psychology Bld. Annex 152 Frelinghuysen Road Piscataway, NJ 08854-8020 Book Reviews Michael Spivey The Continuity of Mind Oxford University Press. 2006. 448 pp. £43.00 ISBN 978-0195170788 (hbk) Reviewed by Igor Aleksander It used to be said that engineers either think in terms of ‘state spaces’ or they don’t. Those who do make aeroplanes fly and those who don’t, sell them. In The Continuity of Mind psychologist Michael Spivey argues that some such division now applies to those who attempt to understand mind and brain. Either they embrace state spaces to study the structure of mind or they remain bereft of an analytic understanding of their subject. So what is state space and why does it suggest that there is something continuous about the mind? The words ‘mental state’ appear often enough in psychological literature and folk psychology. It gets assumed that a particular ‘thought’ is an experienced state as produced by a state of the machinery of the brain (for example, the collection of firing rates of neurons or the distribution of secretion of neurotransmitters in the brain). For Spivey, such states of mind, clearly packaged in time as thoughts, do not exist. The machinery of the brain just never stands still for long enough for an identifiable experience to emerge. We may also be misled in thinking of pure states of mind (such as the memory of that great sunset last night) because cognitive scientists use the computer metaphor as a model of the relationship of mind to body (as software to hardware). Computers jerk from state to state as controlled by a master clock. So at any point in time the computer can be said to be in a briefly unchanging state both in term of its hardware and its singular point in the execution of a program. Journal of Consciousness Studies, 14, No. 8, 2007, pp. 117–29 Copyright (c) Imprint Academic 2005 For personal use only -- not for reproduction 118 BOOK REVIEWS While nowadays this computer model of mind is pretty well dead (simply due to the non-programmed emergent properties of the brain), Spivey adds another twist to the killing process: there are no singular states due to the continuously changing billions of parameters. Like the flow of particles of lava down the side of a volcano, there is not much point in looking for step-by-step processes, continuous analysis is needed. The unsynchronised nature of the brain does not sustain the notion of pure states of mind: whatever mind is, it has the nature of a continuous flow rather than a sequence of beautifully formed states. As an aside, did William James not make this point with the stream of consciousness? Strangely, James does not feature in this book — but he was not mathematically inclined, whereas Spivey is. To unpack what Spivey suggests, we do need to understand some of the state space analyses of dynamical systems of aeroplane designers. Taking just the neurons in the brain - all 11 billion of them - and forgetting the little problem of how these contribute to conscious experience (a question Spivey accepts but does not address), we think of each of these as having their own state of firing or intensity of firing. In technical terms the neurons are the state variables of the brain and (maybe indirectly) of the mind as the movement of the lava particles are the state variables of the lava flow. Let’s take a toy problem: a system of only three neurons that can fire at a rate of anything between 0 and 10 (arbitrary units). We can visualise a 3-axis plot, one for the firing rate of each neuron. Then whatever is the instantaneous rate of each, the state of the system can be plotted as a single point on this 3-D graph. As the rate of the neurons changes, the point will move about in this space. The space is a state space and the route taken by the point is a state trajectory. One cannot visualise a plot for the brain with its billions of neurons, but the principles are the same. The total instantaneous firing of all the neurons is a point in a space with a vast number of dimensions (unimaginable by ordinary mortals who would want to visualise it, but never a problem for mathematicians who treat the space as being abstract with no need to visualise it). Clearly in this space there are state trajectories, that is, signatures of brain activity. To translate this into mental experience requires the same arguments as used by those who refer to a ‘mental state’ as both experienced thought and its brain machinery support. Often, in dynamic explanations, a ‘concept’ like the memory of Mother’s face is an attractor in the state space. That is, state spaces can have areas which are entered and not exited until a distractor input occurs and some other attractor is entered. In Spivey’s scheme, state trajectories do not hop randomly from attractor to attractor but are continuously Copyright (c) Imprint Academic 2005 For personal use only -- not for reproduction BOOK REVIEWS 119 controlled by input. Instead of attractors he has trajectories going through funnels. So the morning thought to have breakfast, catch a bus and get to work would, for Spivey, be a trajectory through breakfast, bus and work funnels rather than being trapped in attractors related to the same meanings. Both attractors and funnels are a result of physical neuron plasticity that stores experience. While the bulk of the book is aimed at persuading the reader to think in terms of continuous mental state trajectories, the philosophical consequences are considerable. For example, Fodor-like ‘modularity of mind’ is not favoured, as trajectories are continuously wafting in and out of a succession of meanings funnels for the individual. Also, free will is out as trajectories cannot avoid being influenced by the funnels in state space. This may seem like freedom because the state structure of funnels and their effects are not part of one’s consciousness: only the trajectories are. Trajectories can explore alternatives so that it is possible to experience and verbalise a trajectory as ‘if I do A, X will follow’ where X might have some unwanted connotation. Another part of the trajectory might lead to ‘if I do B, Y will follow’. Say that Y is a better outcome then the organism might take the better action B. The astute reader will have rightly anticipated that Spivey has state space models for ‘better’ and ‘unwanted’ to clinch the argument against free will. This is not as theologically traumatic as it sounds — more a case of mathematical ‘non-determinism’ where trajectories can represent that both A or B are allowed within some specific state structure. In many ways this book is an initial shot across the bows of many cognitive scientists to alert them that the language of state trajectories might become a lingua franca in brain and mind studies. It has been this for engineers in the last 50 years with much benefit. So it is as an engineer that I have a slight technical quibble with the ‘continuity’ part of Spivey’s argument. State trajectories need not be continuous. They can have jumps (sometimes, mathematically called catastrophes) when the state variables have non-smooth characteristics. And neurons surely fall into this class, possibly explaining the ‘Aha’ phenomenon. But no matter! All in all I feel that the book should be widely read among mind scientists bringing forward the day that the mind can be spoken of in a unified way. Copyright (c) Imprint Academic 2005 For personal use only -- not for reproduction