skip menus and go right to content



LifeWatch Employee Assistance Program




Topic Home  Related:  
Book Review - Surprise, Uncertainty, and Mental Structures
Surprise, Uncertainty, and Mental Structures
by Jerome Kagan
Harvard University Press, 2002
Review by Max Hocutt, Ph.D.
Oct 25th 2002

In the Epilogue to this compact book, Jerome Kagan, professor emeritus of psychology at Harvard University, makes a confession:  He did not fully understand the “plot” of his book until he had finished writing it.  I have to make a similar confession: If by a “plot” is meant a single thesis which determines the book’s organization, or which its various arguments undertake to support or illustrate, I doubt that this book has a plot.  It has what might be called themes, and I will state these below, but it was apparently not written to advance these themes. As the author’s confession suggests, they appear to have emerged in an attempt to survey the experimental work of a lifetime, say something about the beliefs that guided it, and contrast it with the work of others in the field.

One gathers that the effort must have threatened to get out of hand.  In the Introduction, Professor Kagan observes that his book could easily have been a thousand pages long, instead of a mere two hundred.  Although his notes contain little but bibliographical data, they alone total over thirty pages.  A complete and thorough discussion of the literature cited in them would have made for a very large book indeed.  It would also have made for some very tough reading; for the subject is difficult and the findings reported are contentious.  Commendably, Professor Kagan chose to give us a more compendious treatment, which must have required large cuts from earlier drafts.  Unfortunately, the results of Kagan’s effort at condensation are sometimes a little too condensed.  The professor’s diction is always careful, and his syntax is always firm, but his accounts are sometimes so brief as to be nearly cryptic. As we shall see later, he is also prone to make obiter dicta that are not motivated by the context or supported by the evidence.

Developed in its first three chapters, the book’s most important theme is the need for a distinction between what the author calls schemata and what he calls semantic networks, meaning in both cases hypothetical structures, or representations, in the brain.   Schemata are structures and representations that result from sensory/motor contact with physical objects and events —the mental pictures, as it were, that you get from seeing an elephant, smelling a skunk, hearing a train pass by, tasting an onion, lifting a board, and so on.  Once established, the resulting schemata become the patterns that you use to recognize, make sense of, and react to, or deal with, similar objects and events.  Sharing some of the features of what they represent, schemata have verisimilitude; they enjoy the truth of correspondence.  By contrast, semantic networks are interlocking systems of words and concepts having only an artificial relation to what they represent.  Some semantic forms may be tied to schemata, but others are not.  They represent not discernible realities but logical possibilities enjoying only the truth of coherence.  The reader familiar with Immanuel Kant, whom Kagan cites frequently, will recognize here the influence of his famous distinction between empirical content and logical form. There are also vestiges of Plato in this talk of forms.

Kagan says that both kinds of form are resident in the brain; they are, in fact, brain structures.  Kagan emphatically denies, however, that the representational, or mental, powers and traits of these structures can be described in physiological terms.  Psychology can make use of brain physiology.  In fact, psychology cannot do without physiology; but it also cannot be reduced to it.  Cognitive science must have its own distinctive language.  Unfortunately, Kagan is not as forthcoming as might be desired about what distinguishes this language from others. Grant that neuro-physiology cannot displace psychology.  One would like to know precisely why not.  Grant that the cognitive scientist is not talking about the physiological features of brain states.  One would still like to know what she is talking about.  No doubt, we can safely suppose that the defining feature of mental representations is their representational character, but Kagan does not venture to tell us where that is to be found or in what it consists.

What he does tell us, in Chapter 1, is that it is discrepancies between our schemata and our observations and, in Chapter 2, that it is inconsistencies in our networks that often have the most significance for us. We are sometimes motivated more by the unfamiliar than by the familiar.  Novelty, change, surprise, puzzlement, incongruity, uncertainty, and the like can make a greater difference than repetition, reinforcement, etc.  Thus, small children sometimes show more interest in objects that differ from those they encountered earlier, and adults sometimes show more interest in facts that belie their preconceptions. Hence, we are told in Chapter 3, a subject’s response to a new stimulus cannot be predicted without knowledge of her history.  All of this seems correct, even indisputable, but a behaviorist will wonder what is gained by phrasing it in terms of as yet unidentified brain structures.  Grant that where there is learning there must be alterations in the nervous system.  Grant too that we now have the technology to detect brain responses to various stimuli.  One may still want to know how postulating otherwise unspecified “schemata” explains an infant’s act of focusing his attention on a new or different object—especially given Kagan’s well-advised warnings about the uncertainties of interpreting behavior. 

Chapter 4 on “Implications for Development” begins by criticizing the hypothesis that infants have an innate understanding of numbers and solid objects   The chapter also includes a critique of Piaget’s ascriptions of conceptual knowledge to infants, but Kagan wants us to know that he does not go as far as Locke in supposing that the infant mind is a tabula rasa.   He looks favorably, for example, on Chomsky’s belief in an innate capacity for language, but he also wants us to know that he does not think all human capacities are innate.  In particular, he understands that adverse circumstances and unfavorable environments can hinder the cognitive development of those less fortunate than we are.  In short, he believes nothing that might constitute a threat to egalitarianism.

Chapters 1 through 4 having focused on the experiments, Chapter 5 on Creativity and Personality casts a wider net.  It begins by speculating on the reasons for the popularity of T.S. Elliot, Jane Austen, Frank McCourt and other literary figures.  According to Kagan, we count these authors as creative because of the novelties in their work.  Kagan extends the same theory to Sigmund Freud, who became popular when he overturned established views about sexuality.  Resistance to the theory of continental drift in geology shows, however, that the novelty of an idea cannot be too great.  After making this observation, Kagan treats us to broadsides against reinforcement theory and sociobiology, which are said to have been popular, like Freud, for reasons unrelated to their merits—which Kagan’s undocumented caricatures do nothing to reveal.  Chapter 5 ends with a critique of personality inventories and with apt warnings about how a choice of terminology can beg questions and mislead.

The author appears sometimes to criticize other thinkers on the basis of rumor rather than reading.  Thus, he quotes Wittgenstein as wrongly denying that we have a concept of a game, when what Wittgenstein denied was only that there is anything that all games have in common.  Kagan also attributes to behaviorists belief that animals are motivated by a desire for pleasure and to sociobiologists belief that human beings are innately selfish, although real behaviorists don’t talk about mental states, and sociobiology began with the theory that that there is a genetic basis for altruism. The only explanation I can think of for these lapses is that Kagan has not always read, or not always read carefully, the objects of his criticism.  Instead, he has accepted second hand the caricatures of their critics.

That said, however, it must also be said that Professor Kagan’s remarkably compact summary of his life’s work and thought contains much learning and more than a little wisdom.  Many cognitive scientists will want it in their libraries.


© 2002 Max Hocutt


Max Hocutt, Ph.D., Professor of Philosophy Emeritus, The University of Alabama