Fuzzy theories and daft expressions

Peter Hobson is interested in autistic children. From studies of autism, he hopes to shed light on the "origin of thought". If autism itself were well understood, that might be a sensible project. Otherwise it reminds one of the anecdote about the local's reply when asked for directions: "If I wanted to get where you're going, I wouldn't start from here."

Research on autism highlights the importance of interaction and its emotional dimension. Echoing Blake, Hobson summarises his view of the human child's development: "Engagement with others has taught this soul to fly." But, contrary to what Hobson seems to think, it has long been taken for granted that thought and language cannot develop to maturity without a social matrix, and that expression of emotions is one of the basic functions of human signs, including linguistic signs. The trouble is that "emotional" comes to be a catch-all term that covers everything not covered by "conceptual" or "rational".

Hobson describes a rather rebarbative experiment in which a 45-minute-old baby is taken from its mother to see how it will react to a psychologist in a black shirt pulling faces at it. He clearly thinks this is the right way to go about checking hypotheses concerning innate abilities. Parents gave their permission, we are told. But did the baby? And how could the experimenter know what disturbing image this experience might implant in the infant's memory? The chapter bears the title "Before thought". Whatever we think about the experiment, the question-begging starts here. Unless the unfortunate neonate had come out with some Macaulay-like reaction, such as asking "Why, sir, are you pulling faces at me?", the assumption from the outset is that here we are in the realm of what can be done without thinking. Thinking at the age of 45 minutes? Absurd.

Hobson takes it that in situations forced on them by parents and other people benevolently called "care-givers", babies' first reactions are thoughtless, that is, mindless; but he never puts this notion up for debate. He just assumes that the infant is born without a mind, and that the mindless initial state of the creature can be "observed". This is a modern version of the old tabula rasa thesis: a metaphysical proposition disguised as a simple report because it suits the internal logic of Hobson's own thesis, which requires "thought" to wait upon satisfactory emotional relations with others.

Hobson pays little attention to the fact that his chosen topic, the "origin of thought", has a long ancestry in the western tradition. Locke and Condillac may not have gone around "testing" infants, but at least they produced arguments for their case. They would never have talked about anything as asinine as the child's own "theory of mind". Hobson calls this "a daft expression" but he carries on as if it made sense all the same. The child's theory must adapt to "reality", defined as "the perspective that any reasonable person would agree with if they were in the right position to judge".

Experiments with the very young loom large. Hobson warns against over-interpreting the results and describing infants in terms more applicable to adults. But he does it himself all the time. A crucial moment is when the child begins to "symbolise". In a reckless muddle of metaphors, Hobson announces: "Symbolising in general, and language in particular, are poised at the interface between communication and thought."

He thinks that "grammar" appears in the middle of the second year of life and that words can be used "correctly or incorrectly". Predictably, he has serious problems when it comes to what an infant knows, because he does not want the descriptive terminology to clash with his own account of the origin of thought. So when talking of the postulated pre-thought stage he puts the verb "know" in scare quotes, as if this disposed of the issue.

After a brief comparison with chimpanzees (who, you have already guessed, cannot "think symbolically"), Hobson draws a circular conclusion: only when the human child understands that it has a mind, and that others do too, does it acquire a mind of its own. Hobson has evidently not read enough linguistics to realise that the kind of theoretical framework he needs to explain the emergence of meaning in young children's activities is integrationism. But then he would have to start giving up his own old-fashioned philosophical idées r eçues about the mind, language and communication.

Roy Harris is emeritus professor of general linguistics, University of Oxford.