explanation of how this organisation is replicated is that there are innate
programs controlling the functional development of cognition.
So the case for nativism and the case for modularity are interconnected,
and evidence which primarily conWrms the one view may also indirectly
fortify the case for the other. Recent research has supplied a great deal of
relevant empirical evidence. This evidence is drawn from a number of
sources, including early developmental studies, case histories of cognitive
dissociations, and brain-scanning data.
3.1 Developmental evidenceMuch of the developmental evidence has been gained by studies reacting
against the seminal work of Piaget (1936, 1937, 1959; Piaget and Inhelder,
1941, 1948, 1966), although Piaget himself was not an extreme empiricist.
In spite of Locke’s striking metaphor of thetabula rasa, or initialclean
slate, the idea of a cognitive subject as an entirely passive receptor holds
out no hope at all of accounting for development. On any view the child
must contribute a good deal to the developmental process. Developmen-
tal empiricisms can then be more or less extreme, depending upon the
number and generality of the mechanisms they postulate for the process-
ing of experiential input. The most extreme form of all is associationistic
behaviourism, according to which learning is just a form of operant
conditioning conforming to the general ‘law of eVect’. Piaget explicitly
rejected this austere version of empiricism (1927,Wnal ch.; 1936). Instead,
he portrayed the child as an active learner who relies on general learning
principles.
However, Piaget’s methods for testing children’s capacities were insuf-
Wciently sensitive, consistently overestimating the age at which a particular
developmental stage is reached. Developmental psychologists have been
able to lower – sometimes very signiWcantly – the ages at which abilities can
be demonstrated to emerge, by adopting more child-adequate techniques
of investigation. (Early examples of the genre are Gelman, 1968; Bryant
and Trabasso, 1971.) Moreover, development does not proceed on an even
front across all domains, as Piaget believed, but follows diVerent trajec-
tories in diVerent domains. (See for example Carey, 1985; Wellman, 1990;
KarmiloV-Smith, 1992.)
As far as young infants are concerned, the techniques required for
investigating their interests and expectations were simply not available in
Piaget’s lifetime. These techniques rely on the few things babies can do –
suck, look, and listen – and on the fact that babies will look longer at
what is new to them, and will suck with greater frequency on a dummy
when interested in a stimulus. The experimenters who have developed
Developmental rigidity and modularity 57