American Scientist

The Symbolic Species: The Co-evolution of Language and the Brain. Terrence W. Deacon. 527 pp. W. W. Norton & Co., 1997. $29.95.

The Symbolic Species is an outstanding work of synthesis that provides an enviably coherent theory of how the human brain and language have coevolved through time. Its chapters span almost all aspects of the neurosciences (including "paleoneurology"), along with a goodly dose of linguistics, philosophy, embryology, cognitive and developmental psychology, ethology, and, well, you name it. From my own perspective as one who studies the evolution of the brain, this book might well become a bible of sorts for those within and on the borders of paleoanthropology, and in particular anyone interested in the evolution of language. (Who knows, perhaps even sociocultural anthropologists might gain from it.) The author makes no bones about it: Language is unique to humankind—is completely dependent on the particular confluence of hard and soft wiring that comes about through human development and its evolution. Other higher primates, at least from Macaca through Pan (chimpanzee), have most if not all of the basic neural centers and interconnecting fiber systems that we have, but our brains are specialized in that the social-structural environments in which we develop (both in the ontogenetic and phylogenetic sense) select for language. Chapter after chapter abounds in the details underlying this coevolution, and the reader will have to attend well to the arguments because this is not a book for skimming. As I will argue later in this review, however, there are some details, such as neural Darwinism and prefrontalization, that are well worth closer examination.

The book is made up of three parts: The first, "Language," is divided into four chapters that attack the real problem of what represents reference—that is, what symbols mean and do. This part was hard sledding to my aging mind but is the essential background for what follows. Part two, "The Brain," includes six very technical chapters that deal with the neuroanatomy, neurophysiology, neurochemistry and neuropsychology of the human and other primate brains. It is imperative that these chapters as well as the footnotes be read carefully, because it is here that some assumptions become reified as the building blocks for later chapters. It is also here that I find criticism most necessary. Part three, "Co-Evolution," provides four chapters that explain how exactly the brain and language coevolved and why we are the way we are. They range from some wonderful speculations on sexual selection made from a modern sociobiological viewpoint, to the necessity for ritual, and how this combination is woven together into symbolic reference. The last two chapters deal with the problems of what constitutes mind, consciousness, computers and artificial intelligence—none of it glossed over, incidentally. There is some beautiful writing within the twelfth chapter, in particular. The text is followed by chapter footnotes, an annotated selection of suggested additional readings, a large bibliography and an index. This book is just leaps and leaps above other evolution-of-language books published during the past five years. It is a nonpareil.

Well, with this enthusiasm, let’s see if I can work up a little negativity, which I promise to keep "little."

First of all, the book is needlessly overwritten in the beginning, taking about 144 pages of tough going through the minefields of defining icons, indexing and, finally, symbolic referencing in comparative perspective before we ever get into the meat of the nervous system. Despite its brilliance, I imagine that this will be the part picked at by linguists.

My main disagreement comes because the author places an extraordinary degree of dependence on the role of the prefrontal lobe of the human brain in the coevolution of language, whereas he all but ignores the parietal cortex, particularly the inferior parietal lobule. Deacon believes that the primary visual cortex somehow became reduced as it overhung the underlying cerebellar lobes, whereas I believe the posterior parietal cortex showed a relative increase. Just to warn anyone reading this review: A battle of sorts looms on this question.

His premise starts with very dubious data from the areal extent of surfaces, which are supposedly delimitations of particular cytoarchitectonic regions. These data come from works that have never been truly replicated, and this speculation then appears throughout the remainder of the book as if no further evidence of the prefrontalization process were necessary. The data are from The Human Brain in Figures and Tables (S. M. Blinkov and I. I. Glezer, Plenum Press, 1968) and include K. Brodmann's 1912 figures for surface area of the frontal lobe and for the prefrontal cortex, once the premotor cortex is subtracted. They purport to show the human prefrontal cortex as disproportionately large when compared to that of the chimpanzee, our closest primate cousin. The scaling, however, is not log-log, so it is not possible to be certain how far from an allometrically scaled version of prefrontal lobe the human point might appear.

If one uses just the primate data from Brodmann ("Neuer Ergebnisse uber die Vergleichende histologische Localisation der Grossrinde mit besonderer Berucksichtigung des Stimhims," Anatomischer Anzeiger. Ergänzungsheft, 1912, 41:157– 216) and calculates the log-log equation without the human data, when one inserts the total surface area for the human back into the equation, the prefrontal cortex is 50 percent greater than would be expected for a primate with a neocortex of such a size. How significant is 50 percent? The reduction in primary visual striate cortex is on the order of 121 percent for the human brain. There is a crying need for replication here, and as Semerdeferi et al. ("The evolution of the frontal lobes: A volumetric analysis based on three-dimensional reconstructions of magnetic resonance scans of human and ape brains," Journal of Human Evolution, 1997, 32:375–388) point out, the many tables in the Blinkov and Gleser book are particularly difficult to interpret.

Deacon is well aware that there is some problem here, for he explicitly discusses it (but in footnote 13, p. 476) when he mentions briefly the work of Semerdeferi et al. and Uylings et al. ("Qualitative and quantitative comparison of the prefrontal cortex in rat and primates, including humans," Progress in Brain Research, 1990, 85:31–62). Semerdeferi et al. found from their MRI studies that the chimpanzee frontal lobe was about the same proportion as that for the human brain, something I harped on back in 1968 ("The evolution of the primate brain: Some aspects of quantitative relations," Brain Research, 1968, 7:121–172) and Gerhard von Bonin raised 20 years earlier ("The frontal lobe of primates: cytoarchitectural studies." Research Publications Association for Research in Nervous and Mental Disease, 1948, 27:67–83). Neither von Bonin's nor Semerdeferi's work involved delineation of prefrontal cortex, however. Uylings's work did although it was based on the orangutan rather than the chimpanzee. When plotted as a log-log scaled regression, prefrontal cortex versus total cerebral cortex shows the human point just about exactly on the line, and the slope is just slightly greater than 1.0 (1.06), which means, essentially, that the amount of prefrontal cortex in a human being is exactly what would be expected for a primate with its volume of cerebral cortex.

Something is very strange here, as Brodmann's lateral drawing of the chimpanzee brain shows a frontal lobe that appears to be at least one-third the total volume, yet the figure he provides is a paltry 16.9 percent, whereas the human volume is about 30 percent. These studies should have been discussed up front in the main text, for I fear only the most dedicated people will bother to examine the footnotes.

A final little criticism is that in the beginning of the book the author explains that he will not be citing all of his colleagues' works, and he hopes they understand that he wants to present his ideas freshly. Well, that sort of stings, particularly when one's pet theories are not discussed or even cited as a part of the literature. To be clearer, although his book does include a curious discussion of my controversy with Dean Falk regarding the fossil hominid brain endocasts and the position of the lunate sulcus that represents the anterior boundary of Brodmann’s area 17, primary visual striate cortex, in which I come out looking pretty good, Deacon omits three other papers of mine. Specifically, there is a 1967 paper ("The evolution of the human brain: Some notes toward a synthesis between neural structure and the evolution of complex behavior," General Systems, XII:3–19) that offers a veritable theory of how the human brain evolved, including an actual "initial-kick" scenario to the interactive positive feedback between behavioral and neural complexity during human evolution—what the author now labels "co-evolution." In 1969, I suggested that stone tool making, the only evidence from the fossil record for hominid cognition, suggested grammatical structure akin to language ("Culture: A human domain," Current Anthropology, 10:395–412). And in 1995, I offered what I thought was a major synthesis between the fossil record of endocasts and comparative neuroanatomy ("Toward a synthetic theory of human brain evolution," in Origins of the Human Brain, J. P. Changeux and J. Chavaillon, eds., Oxford University Press, pp. 42–54) that was similarly left out. Nor am I not the only person ignored here. The ideas might well be wrong, but I would have hoped they were worthy of notice and discussion!

Having carped thus so slightly about the shortcomings of this book, I'd like to end this review by quoting directly two key (at least in my mind) passages in this book that are the major kernels in his theory:

Establishing such social-sexual relationships cannot be accomplished by indexical communication alone, that is, by systems of animal calls, postures, and display behaviors, no matter how sophisticated and complex. And yet, even extremely crude symbolic communication can serve this need. Only a few types of symbols and only a few classes of combinatorial relationships between them are necessary. But without symbols that refer publicly and unambiguously to certain abstract social relationships and their future extension, including reciprocal obligations and prohibitions, hominids could not have taken advantage of the critical resource available to habitual hunters. The need to mark these reciprocally altruistic (and reciprocally selfish) relationships arose as an adaptation to the extreme evolutionary instability of the combination of group hunting/scavenging and male provisioning of mates and offspring. This was the question for which symbolization was the only viable answer. Symbolic culture was a response to a reproductive problem that only symbols could solve: the imperative of representing a social contract.

Finally (but there is much else worthy of citation herein):

In conclusion, then, the theory of symbolic origins I have outlined is not just a new twist on Rousseau's "social contract" theory. It is not a theory of the origins of social behavior, but of the translation of social behavior into symbolic form. More important, it is not a scenario for how our intelligence triumphed over our reproductive competition, but rather how unique demands of reproductive competition and cooperation created the conditions that led to our unique form of intelligence. By answering the evolutionary question of how to take advantage of a new foraging trick, our ancestors unwittingly turned the tables of natural selection so that social evolution could reshape the brain in its own image. We reflect on this from the other end of an extensive co-evolutionary process, where the indispensable uses of symbolic communication as a social organizing tool were long ago relegated to being only one among a multitude of selection pressures mutually converging on making this communication more and more efficient. Two and a half million years of sustained selection in an unprecedented socioecological niche, maintained by unprecedented communicational and cognitive tricks have taken us far from these beginnings in both the physical changes in the brain that resulted and in the mental and cultural world that coevolved with them.

These two quotes include the essential kernels of how the synthesis proceeds, with a surplus of other insights that I found both profound and exciting. I wish a review such as this could mention all of them. Deacon is to be congratulated for this superb synthesis!