Viewed retrospectively, shyness appears to be a remarkably stable concept: many people who describe themselves as shy claim that they have been shy since childhood. Viewed prospectively, however, the stability appears less compelling. Even Kagan (1997: 35), whose typological orientation implies a high level of stability in the generalized tendency to approach or avoid novel stimuli, warns that ‘further environmental conditions can modulate the behavioural profile; levels of motor attention and crying are not constant from day to day and daily experiences permit some children to learn to control their irritability and, later, their fear. It is even possible that experiences that reduce levels of uncertainty can alter the excitability of the limbic systems or change the density of receptors on neurons.’