pursued. Infants and caregivers, that is, share an interest in making sense of and to one
another, and, although only partly and contingently, share interests in the outcome of
their relationship.10 But they cannot interact in symbolic language, since only one of
them is capable of doing so. Symbolic language is an outcome of their communication-
hungry interaction, rather than a resource available to it from the outset.
Other resources are, though, available. These include facial expressions, direction of
gaze, gestures, body-orientation, and prosodic properties of speech, all of which are
powerful media of affective signalling. Caregivers are directly affected and motivated by
displays of infant affect, especially when the infant is their own offspring (e.g.
Wiesenfeld and Klorman 1978). From birth, or very soon after, infants show interest in
faces (e.g. Maurer & Young, 1983), preference for smiling faces (Easterbrook and Barry
2000)11 and evidence of facial imitation (e.g. Meltzoff and Moore 1977). By the time of
birth they attend to, and prefer, the rhythmic properties of the language they heard most
in the muffled world of the womb, and a particular preference for the voice of their
mother, which they reliably identify and prefer to other voices following birth (e.g.
DeCasper and Fifer 1980). Some prosodic features of infant-directed utterances have
been shown to be indicators of approval, disapproval, etc., in their own way just as
universal as facial expressions are indicators of affective state (e.g. Fernald 1992, Ekman
1972).12 Infants across cultures show early preferences for approval vocalisations over
ones whose prosodic character is associated with disapproval.
Neither parent nor infant seem, then, to have to learn how to get started with affective
interaction. In the terms adopted above, we can say that these capacities for affective
response make possible a set of innate indexical associations, or serve as the basis for
their development. They facilitate the setting up of complex patterns of behavioural co-
ordination forming a basis for ongoing development of ever more refined interactive
behaviour. By the middle of the second month of life, infants and caregivers begin to
engage in interactions often described in terms of mutual ‘delight’, in ways showing
evidence of cultural particularity. Trevarthen (1977) refers to such episodes in Britain as
manifesting ‘spontaneity, vivacity and delight’, while Bateson (1979) describes
interactions in Iran as involving ‘delighted, ritualized courtesy’. We might add that our
own data concerning Zulu mothers and infants (see below) includes periods of ‘delighted
musical chorusing’. Around the third month interaction between infants and caregivers
becomes intensely dialogical, involving the production of protoconversation (Bateson
1979) and manifesting what Trevarthen (1979; 1998) called intersubjective
communication. While caregivers respond to infant behaviour, striking phenomena arise
from how they guide and control the infant’ s affectively-based activity. Not only does
this involve the development of joint evaluative behaviour but this outcome influences
how they motivate and rationalise their own behaviour.
For our purposes an especially important feature of this guiding activity is that it is able
to draw on culturally particular expectations concerning appropriate and inappropriate
behaviour. What makes this important is that these expectations are, to varying extents,
culturally specific, and hence that the particular patterns of expectation have, unlike the
responses to smiling, say, to be learned.
It is clear enough that infants occupy what one might call ‘culturally saturated’
environments, in which, for example, the likelihood of an adult allowing an infant’ s
direction of attention to initiate and fix the focus of interactions, is variable. Other areas
of variation include patterns of response to infant distress, where, for example, in some
settings attempts to distract the infant by directing their attention to a visible object are
more likely, whereas in others attempts to comfort or subdue are common. What is not
obvious is when infants themselves begin to show evidence of enculturation, that is, of
behaviour partly shaped by the patterns of interaction prevalent in their own culturally
9 / 23