Merlin Donald
Origins of the Modern Mind

Harvard University Press. 1991


page 2
Keywords: Human cognition - aspects of behavior, including reflexes, instincts, perceptual capacities conditioned behavior, problem solving, insight, curiosity, communication, language, social intelligence, and memory - revolution in cognitive science, a new style of conceptualizing the mind has developed, one that is best described as structural - structural models of mind - my hypothesis is that the modern human mind evolved from the primate mind through a series of major adaptations, each of which led to the emergence of a new representational system - modern representational structure of the human mind - our representational apparatus somehow perceived the utility of symbols and invented them from whole cloth; no symbolic environment preceded them - symbolic reference - to understand or use a symbol appropriately in context you must first understand what it represents, and this referential understanding is inherently nonsymbolic -


pg. 2
Human cognition is complex and multidimensional. How should its evolution be approached? There is a large literature comparing humans with other species on particular aspects of behavior, including reflexes, instincts, perceptual capacities conditioned behavior, problem solving, insight, curiosity, communication, language, social intelligence, and memory.

But unfortunately the continuum from reptile to mammal to primate to human has not always proven to be a smooth one; some animals appear better at one thing, some at another. For this reason, comparative studies have painted a somewhat confusing picture. However, there is another possible approach, based on
the underlying structure of mind. During the recent revolution in cognitive science, a new style of conceptualizing the mind has developed, one that is best described as structural.

One major objective of structural models of mind is to describe modular architecture, that is, the overall configuration of components making up the totality of mind. Clinical neuropsychological evidence is central to such models, because it addresses the question of cognitive breakdown or dysfunction; components of mind sometimes break down independently of one another, providing clues as to their place in an overall structural scheme.

The essence of my hypothesis is that the modern human mind evolved from the primate mind through a series of major adaptations, each of which led to the emergence of a new representational system. Each successive new representational system has remained intact within our current mental architecture, so that the modern mind is a mosaic structure of cognitive vestiges from earlier stages of human emergence. Cognitive vestiges invoke the evolutionary principle of conservation of previous gains and are similar in principle to the many other vestigial behaviors we possess - for instance, baring the teeth in anger, or wailing in grief.

The modern representational structure of the human mind, I will argue, thus encompasses the gains of all our hominid ancestors, as well as those of certain apes. Far from being a diffuse tabula rasa, modern human cognitive architecture is highly differentiated and specialized. And despite our close genetic relationship to apes, the cognitive distance from apes to humans is extraordinarily great, much greater than might be imagined from comparative anatomy.

The key word here is representation. Humans did not simply evolve a larger brain, an expanded memory, a lexicon, or a special speech apparatus;
we evolved new systems for representing reality. During this process, our representational apparatus somehow perceived the utility of symbols and invented them from whole cloth; no symbolic environment preceded them.

The problem of symbolic reference has always been the Achilles heel of computational approaches to language. The difficulty is this: to understand or use a symbol appropriately in context you must first understand what it represents, and this referential understanding is inherently nonsymbolic. For instance, to understand why it is amusing to name a reclusive hound dog "Raskolnikov," one must know something of Dostoevsky's novel, something of the brooding, asocial intelligence of hound dogs, and something of the style of contemporary humor. None of this knowledge can be obtained just by looking up other symbols. But that is all most computational algorithms can do. The programmer, or the user, must eventually provide meaning, since the computer has no knowledge of its own.

Moving to a nonrepresentational computer model (Brooks, 1989) won't solve the problem either; such systems must ultimately face the same limitations as an animal without symbolic intelligence. Just as traditional Al programs use nothing but symbols, most animals cannot use symbols at all, and no animal except humans has ever invented a symbolic device in its natural environment.

The question is, how did humans, given their nonsymbolic mammalian heritage, come to represent their knowledge in symbolic form? -Through what stages must this development have passed? How did humans bridge the tremendous gap between symbolic thought and the nonsymbolic forms of intelligence that still dominate the rest of the animal kingdom?

During the relatively short time of human emergence, the structure of the primate mind was radically altered; or rather, it was gradually surrounded by new representational systems and absorbed into a larger cognitive apparatus.


Language-Evolution

HOME      BOE     SAL     TEXTE