Date: Fri, 02 Aug 2002 14:49:21 -0400 From: Katharine Beals Subject: Loritz (2002) How the Brain Evolved Language
Loritz, Donald (2002) How the Brain Evolved Language. Oxford University Press, 227pp, paperback ISBN 0-19-515124-0, $19.95. (Hardback ed. published 1999)
SUMMARY In "How the Brain Evolved Language" Donald Loritz argues why human language requires no special linguistic module, and how its underlying neurology can, and indeed must, be explained without recourse to the abstract mechanisms of generative grammar.
A complete explanation of language, Loritz believes, must start with the one-celled life forms whose communication with the outside world prefigured multicelled life. Loritz, thus, begins by describing single-celled organisms, and how their membranes admit food and block toxins; the Cambrian explosion of multicelled organisms, a consequence of the invention of the neuron (an "excellent prototype" for which was the single-celled organism called the flagellate, with its axon-like tail); and the formula for life: f(x) = Ax + t, where x is a form, A is a geometric transform of that form, and f(x), therefore, is the general equation for a fractal.
Indeed, the fractal principle of self-similarity-- the principle that "patterns in nature repeat themselves on different scales" (p. 29)-- pervades Loritz's sketches of biology in general and the brain in particular. The cell, with its membranes within membranes, is self-similar both in its design and function; by the same principle, so is each bodily organ, and so is the body as a whole.
The recapitulation of phylogeny by ontogeny in fetal development is a case of "procedural self-similarity" (p. 30). Within an organism that itself communicates, so is communication among neurons, the complicated mechanics of which is Loritz' next topic. Following this, and next in the fractal hierarchy of human communication, is "the society of the brain." Here we visit the limbic system, the cerebellum, and the cerebrum, with all its substructures and their different functions, and discover, in Loritz' overview of the occipital lobe and vision, the principle of "on-center off-surround" anatomies that turns out to be key to language learning and processing. In such an anatomy, some "center cells", reacting to a stimulus, relay their excitement up to other cells, which echo this back to the center cells, keeping them excited, while also inhibiting those at the periphery (p. 63). Reviewing evidence about brain function in deaf signers, left-handers, and children who recover from aphasia, Loritz concludes that the cerebrum is initially plastic, and that there is no particular location into which language is hard-wired.
The subject of the next chapter, Adaptive Resonance, is the theory of neural communication which underlies how brain cells learn language without hard-wiring. Here, with myriad schematic diagrams, mathematical equations, and quantitative models, Loritz addresses how Adaptive Resonance, via on-center off-surround anatomies, enables such functions as contrast enhancement, noise suppression, resonance, self-similarity, and neural rebounds-functions which, though they turn out to be key to language, also underlie sensory processing, especially vision.
In the following chapters, Loritz discusses the extent to which Adaptive Resonance Theory underpins language learning and processing, from sound and speech perception, to serial processing, to morphology, syntax, semantics and pragmatics. Key to morphology and syntax is a specific Adaptive Resonance mechanism: that of the dipole rhythm generator, which enables the hierarchical, binary, parsing of language into phrases, words, feet and syllables. Emerging from these dipoles, therefore, are metrical phonology, spoonerisms, and syntactic structure. Syntax, in particular word order, also results from a "topic gradient" of old to new information, which includes verbal as well as nominal elements, and which orders older before newer. Combine the topic gradient with the principles of Adaptive Resonance, and we get cross-linguistic tendencies towards subject first, as well as topicalization, passivization, dative movement, particle movement, the ordering of arguments within the VP, the distribution of pronouns, the scope of negation, and the syntax of questions. Generative grammar, which had its time, and its virtues, but whose tree diagrams were starting to look "more and more like stimulus-response chains and crayfish brains" (p. 147), is no longer needed.
In the closing chapter, "What if Language is Learned by Brain Cells," Loritz revisits, from the vantage point of language acquisition, how language can exist without a linguistic module. He questions the so-called "critical period" for language learning, arguing that, to the extent that it exists, it is merely part and parcel of a critical period for learning in general: over time the cerebrum in particular becomes less plastic. He then shows how various language disorders, some of which (e.g. Specific Language Impairment) purportedly evince different linguistic modules, may instead result from more general dysfunctions, e.g. in Adaptive Resonance mechanisms. Finally, he questions two key claims of generative grammarians: that children learn language effortlessly, and that parents don't teach it to them. In particular, he argues, when a child refers to a cow as a "koo," the typical parental response-e.g. "That's right! It's a cow"- is an expansion or recasting that yields implicit negative evidence of the sort that generative grammarians assume is completely absent, but which in fact, through the principles of Adaptive Resonance, causes automatic, subconscious corrections.
DISCUSSION Maybe Loritz is on to something. In particular, his general observation about implicit negative evidence in language acquisition rings true. But, for the most part, it's awfully hard to tell. His explanation of Adaptive Resonance Theory is at once too technical and too sketchy for the non-mathematical, non-computational lay person to glean even a general gist; his disposal of the major problems of phonology, morphology and syntax is way too sweeping and superficial: morphology in three pages; syntax in 13. Farfetched claims-- whether about flagellae and axons, or syntactic trees and crayfish brains-- undermine his credibility. So do some of his specific linguistic judgments, e.g. the purported acceptability of "Irv poured the glass with ice water," which to me, in comparison with "Irv filled the glass with ice water," sounds unequivocally odd.
The general exposition is also confusing. Key terms and concepts-"self-similar," "on-center off-surround" and "dipole" appear pages in advance of anything approaching a definition. The topics ramble along with many apparently unmotivated digressions: it's never really clear, for example, what the point is of recapitulating phylogeny and exposing self-similarity in Chapter 2, or what the meditation in Chapter 11 on truth vs. meaning (which tracks the rise and fall of Linguistics from the Holy Roman Empire and the Reformation through Hitler and oral contraceptives) has to do with the rest of the book.
"How the Brain Evolved Language" glimmers with interesting and important ideas. If only Loritz had used more words to describe and connect them!
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Katharine Beals received her Ph.D. in linguistics from the University of Chicago in 1995. From 1995 to 2000 she worked as a Senior Software Engineer with the Natural Language Group at Unisys. She is currently at home with her baby daughter and at work on a book about her deaf, autistic son, which explores such issues as language modality, cochlear implants, and language and consciousness in autistic people.