Infant Intersubjectivity: Research, Theory, and
Clinical Applications
Article in Journal of Child Psychology and Psychiatry · February 2001 DOI: 10.1017/S0021963001006552 · Source: PubMed CITATIONS
752
READS2,255
2 authors: Some of the authors of this publication are also working on these related projects: Sharing in infancyView project
Sharing in Imagination (in Life and Art)View project
Colwyn Trevarthen The University of Edinburgh 119 PUBLICATIONS 6,915 CITATIONS SEE PROFILE Ken Aitken University of Glasgow 48 PUBLICATIONS 1,293 CITATIONS SEE PROFILEAll content following this page was uploaded by Ken Aitken on 29 September 2014.
' 2001 Association for Child Psychology and Psychiatry Printed in Great Britain. All rights reserved 0021–9630\01 $15n00j0n00
Infant Intersubjectivity : Research, Theory, and Clinical Applications
Colwyn Trevarthen and Kenneth J. AitkenThe University of Edinburgh, U.K.
We review research evidence on the emergence and development of active ‘‘ self-and-other ’’ awareness in infancy, and examine the importance of its motives and emotions to mental health practice with children. This relates to how communication begins and develops in infancy, how it influences the individual subject’s movement, perception, and learning, and how the infant’s biologically grounded self-regulation of internal state and self-conscious purposefulness is sustained through active engagement with sympathetic others. Mutual self-other-consciousness is found to play the lead role in developing a child’s cooperative intelligence for cultural learning and language. A variety of preconceptions have animated rival research traditions investigating infant communication and cognition. We distinguish the concept of ‘‘ intersubjectivity ’’, and outline the history of its use in developmental research.
The transforming body and brain of a human individual grows in active engagement with an environment of human factors—organic at first, then psychological or inter-mental. Adaptive, human-responsive processes are generated first by interneuronal activity within the developing brain as formation of the human embryo is regulated in a support-system of maternal tissues. Neural structures are further elaborated with the benefit of intra-uterine stimuli in the foetus, then supported in the rapidly growing forebrain and cerebellum of the young child by experience of the intuitive responses of parents and other human companions. We focus particularly on intrinsic patterns and processes in pre-natal and post-natal brain maturation that anticipate psychosocial support in infancy. The operation of an intrinsic motive formation (IMF) that developed in the core of the brain before birth is evident in the tightly integrated intermodal sensory-motor coordination of a newborn infant’s orienting to stimuli and preferential learning of human signals, by the temporal coherence and intrinsic rhythms of infant behaviour, especially in communication, and neonates’ extraordinary capacities for reactive and evocative imitation. The correct functioning of this integrated neural motivating system is found to be essential to the development of both the infant’s purposeful consciousness and his or her ability to cooperate with other persons’ actions and interests, and to learn from them.
The relevance of infants’ inherent intersubjectivity to major child mental health issues is highlighted by examining selected areas of clinical concern. We review recent findings on postnatal depression, prematurity, autism, ADHD, specific language impairments, and central auditory processing deficits, and comment on the efficacy of interventions that aim to support intrinsic motives for intersubjective communication when these are not developing normally.
Keywords :Infant intersubjectivity, parent-infant communication, developmental disorders, pathologies of empathy, therapies.
Abbreviations : ADS : adult-directed speech ; DTV : double video link ; F0 : fundamental frequency ; IDS : infant-directed speech ; IMF : intrinsic motive formation ; IMP : intrinsic motive pulse ; PDD : Pervasive Disintegrative Disorder ; PRC : period of rapid change.
Introduction
This is the first Annual Research Review on infancy research for the JCPP. The time is ripe for an examination of changing conceptions of first steps in human psycho-social growth.
The idea that normal human sensitivity for psycho-logical impulses in other persons may have a basis in inherent cognitive and emotional systems of the brain specialised for this function has received attention in psychology recently, much of it sceptical. Given the predominance of individualist, constructivist, and
cog-Requests for reprints to : Professor Colwyn Trevarthen, De-partment of Psychology, University of Edinburgh, 7 George Square, Edinburgh EH8 9JZ, U.K.
nitive theory in empirical psychology, this is hardly surprising. The central problem in early development of the mind has been taken to be object awareness, not person awareness. Nevertheless, there is evidence that even newborn infants, with their very immature though elaborate brains, limited cognitions, and weak bodies, are specifically motivated, beyond instinctive behaviours that attract parental care for immediate biological needs, to communicate intricately with the expressive forms and rhythms of interest and feeling displayed by other humans. This evidence of purposeful intersubjectivity, or an initial psychosocial state, must be fundamental for our understanding of human mental development. It will also be crucial for accurate interpretations of the influences of nature and nurture in the baffling spectrum of psycho-social pathologies in children, as for the development of 3
effective treatment strategies, be they therapeutic or educational.
Observation and experimentation in the motor, sen-sory, and cognitive developments in early childhood has grown spectacularly since the 1960s, stimulating models of developmental change in perceptual discrimination and representation, operational thinking, skills and mem-ory, and attempts to relate these achievements to brain development, as well as to the design of artificial neural nets that simulate cognition and learning. With the success of precise experimental methods developed to measure infant object-perception and cognition in con-trolled conditions, the everyday social-interpersonal fac-tors of development and the intrinsic motives that normally regulate them, for which new evidence was obtained in the 1970s by micro-descriptive studies, have been less regarded. Or they have been explained in reduced physicalistic terms as secondary effects of social contingencies, or as the outcome of some kind of acquired intellectual process that can ‘‘ read ’’ others’ minds. Now research on infants’ special awareness of persons, and their active influence over caregivers’ behaviour, is having a comeback, along with a renewed interest in the motives and emotions that animate consciousness and self-awareness in humans and animals.
We believe that the existence of specialised innate ‘‘ human-environment-expectant ’’ social regulatory and intersubjective functions in the infant mind has been firmly established, and argue that the corresponding anticipatory motives constitute an essential framework for the regulation of all human cognitive development ; guiding, limiting, extending, and evaluating what the individual can discover inside and outside his or her body. Related, though psychologically simpler, processes of intersubjective regulation appear in all animal species that are both highly social and at first dependent on intelligent parental care. The human case is unique in its adaptations, which guide children through dialogic ex-change of emotive and referential narratives in body-mimesis to language and learning of a cultural accumu-lation of well-reasoned knowledge and strategic technical skills. The emotional investment of the child in this ‘‘ learning how to mean ’’ is of primary importance in clinical work with children, as well as in their education. A cognitive description of psychological development in the individual human baby, focusing on the processes that take in perceptual information about objects and the physical situation, is certainly possible. However, it would appear to be a logical category error to infer that interaction between subjects can be explained by decom-posing their behaviours and perceptual discriminations into cognitive components that are adapted to guide one agent in engagement with things that have no psycho-logical anticipation and no adaptive behaviour.
The social intelligence of the infant is evidently a specific human talent—an inherent, intrinsic, psycho-biological capacity that integrates perceptual information from many modalities to serve motive states. Moreover, this capacity is a necessary prerequisite, although not in itself a sufficient cause, for a child to go through psychological development of the kind that leads to and depends on cultural learning. Such a premise leads to a different research agenda in clinical psychology from one that views the cerebral mechanisms of social behaviour, and the emotions that regulate it, as a product of emerging, or constructed, modular components of general representation, or of processing in cognition
(Karmiloff-Smith, 1992 ; Piaget, 1954 ; Rutkowska, 1993, 1997 ; Spelke, 1991), or uni-modal perceptual pattern-recognising mechanisms (Bremner, Slater, & Butter-worth, 1997 ; Johnson & Morton, 1991). We believe that the prevailing logic needs to be reversed ; that object cognition and rational intelligence in infants, and their perceptual preferences, should be viewed as the outcomes of a process that seeks guidance by person-perception and through communication with equivalent processes, of cognition-with-intention-and-emotion, in other persons.
Evidence will be given from a number of clinical areas for the effects of early difficulties in interpersonal func-tions that degrade subsequent developments, including those rational, experience-dependent skilful, and more utilitarian aspects of conscious life subsumed under the titles of ‘‘ social cognition ’’, ‘‘ theory of mind ’’, and ‘‘ pragmatics ’’ of speech and language.
Normal Intersubjectivity in Early Infancy The Discovery of Innate Intersubjectivity in Proto -conversations and Games with Young Infants
The theory of innate intersubjectivity—that the infant is born with awareness specifically receptive to subjective states in other persons—was put forward 25 years ago to account for observations made descriptively or etho-logically from films of the behaviours of infants in natural interaction with their mothers, who were attempting to engage the infants in face-to-face chat, or playing games with them (Trevarthen, 1974, 1979, 1998 a). In the 1970s, researchers in different fields reported the findings of film studies of live interactions between adults and infants a few months old (Bateson, 1971, 1979 ; Brazelton, Kos-lowski, & Main, 1974 ; Stern, 1971, 1974, 1977 ; Tronick, Als, & Adamson, 1979). They were impressed with the similarities of timing and expression between these simplest, intuitive human encounters and informal con-versations between adults. The techniques of conver-sational analysis, with accurate measurement of the timing of the contributions by adult and infant, brought statistical confirmation of this similarity (Beebe, 1982 ; Beebe, Jaffe, Feldstein, Mays, & Alson, 1985 ; Beebe, Stern, & Jaffe, 1979 ; Feldstein et al., 1993 ; Fogel, 1977, 1985 a ; Stern, 1971). It was M. C. Bateson (1971, 1975, 1979) who termed the mother-infant interactions proto-conversations.
Further study revealed that this natural sociability of infants, engaging the interest, purposes, and feelings of willing and affectionate parents, serves to intrinsically motivate companionship, or cooperative awareness, lead-ing the infant towards development of ‘‘ confidence, confiding and acts of meaning ’’, and, eventually, to language (Trevarthen, 1980, 1982, 1987, 1988, 1990 a ; Trevarthen & Hubley, 1978 ; Trevarthen, Murray, & Hubley, 1981). The infant’s communicative motivation, and the intuitive parenting that fosters it, have been identified with the special human aptitude for cultural learning, including language learning (Adamson & Bake-man, 1991 ; Bakeman & Adamson, 1984 ; Bruner, 1976, 1983 ; Butterworth & Grover, 1988 ; Eckerman, Whatley, & McGee, 1979 ; Halliday, 1975 ; Locke, 1993 ; H. Papousek & Bornstein, 1992 ; H. Papousek & Papousek, 1977, 1987 ; Rommetveit, 1979, 1998 ; Ryan, 1974 ; Tomasello, 1988 ; Tomasello & Farrar, 1986 ; Tomasello, Kruger, & Ratner, 1993 ; Vygotsky, 1978). The infant’s
need for communication animates the initial self-other awareness and reception of motives and emotions in the intersubjective messages that underlie all language—a ‘‘ human sense ’’, as Donaldson (1978) called it, that emerges in progressively more powerful forms through the course of infancy (Bra/ ten, 1998; Reddy, Hay, Murray, & Trevarthen, 1997 ; Rommetveit, 1998 ; Ryan, 1974). The earliest meanings are conveyed to the infant or toddler nonverbally or paralinguistically by vocal and gestural expression in natural social situations, at the same time as language is being used by older interactants to convey referential information and to specify purposes, experiences, thoughts, and recollections. Regulation of this primary human communication depends on an innate ‘‘ virtual other ’’ process in the infant’s mind (Bra/ ten, 1988 a, b, 1998).
Researchers found that as early as 2 months, infants and mothers, while they were looking at and listening to each other, were mutually regulating one another’s interests and feelings in intricate, rhythmic patterns, exchanging multimodal signals and imitations of vocal, facial, and gestural expression (M. C. Bateson, 1975, 1979 ; Beebe et al., 1979, 1985 ; Brazelton, Tronick, Adamson, Als, & Wise, 1975 ; Fogel, 1977, 1985 a, b, 1993 a, b ; Fogel & Hannan, 1985 ; Fogel & Thelen, 1987 ; Mayer & Tronick, 1985 ; Stern, Beebe, Jaffe, & Bennett, 1977 ; Stern, Jaffe, Beebe, & Bennett, 1975 ; Tronick, Als, & Brazelton, 1980 ; Weinberg & Tronick, 1994). Mothers and fathers were behaving in an intensely sympathetic and highly expressive way that absorbed the attention of the infants and led to intricate, mutually regulated interchanges with turns of displaying and attending. The infant was thus proved to possess an active and im-mediately responsive conscious appreciation of the adult’s communicative intentions. This is what was called
primary intersubjectivity (Trevarthen, 1979). The
dis-tinction between subjectivity and intersubjectivity in early infancy was defined as follows :
Subjectivity and intersubjectivity : a definition of terms
Human beings understand one another intimately and at many levels. To analyse this ability of persons to act together and to share experience in harmony, we have first to view communication in relation to the private activities of conscious, purposeful action. All voluntary actions are performed in such a way that their effects can be anticipated by the actor and then adjusted within the perceived situation to meet criteria set in advance. Interpersonal communication is controlled by feedback of information, as in all voluntary behaviour. But there is an essential differ-ence between a person doing things in relation to the physical world and the control of communication between persons. Two persons can share control, each can predict what the other will know and do. Physical objects cannot predict intentions and they have no social relationships.
For infants to share mental control with other persons they must have two skills. First, they must be able to exhibit to others at least the rudiments of individual consciousness and intentionality. This attribute of acting agents I call subjectivity. In order to communicate, infants must also be able to adapt or fit this subjective control to the subjectivity of others : they must also demonstrate intersubjectivity. By subjectivity I mean the ability to show by coordinated acts that purposes are being consciously
regulated. Subjectivity implies that infants master the difficulties of relating objects and situations to themselves and predict consequences, not merely in hidden cognitive processes but in manifest, intel-ligible actions (Trevarthen, 1979, pp. 321–322). Perturbation tests, by the still or blank face or double television replay procedures, discussed below, further demonstrated that a 2- to 3-month-old infant was emo-tionally aware of a mother’s contingent and emotion-ally appropriate behaviour, and actively engaging with it (Murray & Trevarthen, 1985, 1986 ; Trevarthen, 1993 a, b ; Trevarthen et al., 1981 ; Tronick, 1989 ; Tronick, Als, Adamson, Wise, & Brazelton, 1978), and these findings are confirmed by more recent investigations (Nadel, Carchon, Kervella, Marcelli, & Re! serbat-Plantey, 1999).
Longitudinal film study of behaviours recorded in semicontrolled lab\studio conditions that favoured close observation revealed an orderly age-related transform-ation of the infant’s motives through the middle of the first year, toward increasingly intricate, precise, and selective coordination with the mother’s richly inflected, rhythmically patterned, and repetitive expressions of communication and dramatised actions of play (Beebe et al., 1979, 1985 ; Bruner & Sherwood, 1975 ; Fogel, 1977 ; Jasnow & Feldstein, 1986 ; Mayer & Tronick, 1985 ; Ratner & Bruner, 1978 ; Stern, 1971 ; Stern et al., 1977 ; Stern & Gibbon, 1980). The baby’s increasing interest in objects was observed to grow in some competition with the earlier developed motives for protoconversational play, and led, around the middle of the first year, to the elaboration of more lively games with objects. Just before the end of the first year, there was a rather sudden development of joint interest of mother and infant in their surroundings, triggered by the infant’s emerging curiosity about the timing and direction and focus of attentions and intentions of the mother (Hubley & Trevarthen, 1979 ; Pecheux, Ruel, & Findji, in press). This change in infants’ experience, and acceptance of joint attention to the world, clearly has momentous consequences in subsequent learning, and profound effects on the ways mothers act with and speak to their infants.
Parallel study of the development of younger infants’ orientations—activities aimed to engage objects and physical events (tracking and reaching, grasping, and manipulating)—clarified the differences between
sub-jective motivesthat led them to experience, for themselves,
the sensations and affordances of their own bodies and of things, and the intersubjective motives that were drawing them into games and self-other regulations of a strictly interpersonal kind, in which the babies had to react alertly to the expressions of purpose and emotion in their partners (Trevarthen et al., 1981). It was confirmed that the differing motives for these two kinds of objective—for object awareness or doing with things, and for person-awareness and communicating with persons (see Trevarthen, 1998 a)—were, indeed, undergoing divergent and periodically competing development during the first year, leading, at around 9 months after the infant’s birth, to integration in the new form of cooperative inter-subjectivity (person-person-object awareness), which was named secondary intersubjectivity (Trevarthen & Hubley, 1978).
It is significant that the evidence for ‘‘ person aware-ness ’’ and a capacity for intersubjectivity came from description in detail, from frame-by-frame analysis with
accurate measurement of the timing, of how infants moved their bodies, especially their expressive organs, in responses, both contingent and provocative, to the expressions of another person. Importantly, the be-haviours selected to define the infant’s intersubjec-tivity—the ways the infants look, express their feelings in face and voice, how they gesture and move in rhythmic cycles to accept or reject contact—were homologous with behaviours that are essential to the elaborate inter-subjectivity of all collaborative intentional activity in adult society, including live conversational language. They are regulated and negotiated purposefully and emotionally, by expressive and receptive processes en-gaging many modalities simultaneously (Bra/ ten, 1998; Dore, 1983 ; Fernald, 1989 ; Jaffe, Stern, & Peery, 1973 ; Kuhl & Meltzoff, 1982 ; Murray & Trevarthen, 1985 ; Stern, 1974 ; Stern et al., 1977 ; Stern, Hofer, Haft, & Dore, 1985 ; Trevarthen, 1978, 1984 a, 1993 a, b ; Trevarthen, Kokkinaki, & Fiamenghi, 1999 ; Weinberg & Tronick, 1994).
By 1 year a baby can not only communicate directly with human expression without language, but can also energetically share complex arbitrary experiences, boldly displaying to familiar persons an individual, socially adapted personality. The baby attends to and imitates conventional vocalisations and gestures, as well as mak-ing orientations to and handlmak-ing objects that other persons use, imitating their actions (Adamson & Bake-man, 1991 ; Bakeman & Adamson, 1984 ; Bates, 1979 ; Bretherton, McNew, & Beeghly-Smith, 1981 ; Bruner, 1976, 1983 ; Butterworth, 1999 ; Butterworth & Grover, 1988 ; Eckerman et al., 1979 ; Halliday, 1975, 1979 ; Hubley & Trevarthen, 1979 ; Locke, 1993 ; Ryan, 1974 ; Tomasello, 1986 ; Tomasello & Farrar, 1986 ; Trevarthen, 1987, 1988, 1990 a, 1992 ; Trevarthen & Hubley, 1978 ; Trevarthen & Logotheti, 1987 ; Trevarthen et al., 1981 ; Uzgiris, 1981, 1999). Motivation to regulate fluent ‘‘ person-person-object ’’ awareness, joint attention, and mutually adjusted intentionality, all at once, is coming to the fore at this age (Trevarthen & Hubley, 1978).
Details of the expression of the developing motives that drive the earliest communications of humans are sum-marised as follows. In the gentle, intimate, affectionate, and rhythmically regulated playful exchanges of proto-conversation, 2-month-old infants look at the eyes and mouth of the person addressing them while listening to the voice. In measured and predictable cycles of response to regular time patterns in the adult’s behaviour, the infant moves its face, which it cannot see or hear, and reacts with movements of face, hands, or vocal system to modified patterns of adult vocal expression that it is incapable of mimicking, and that have not been available in that form in utero. The communicatively active hands of young infants may make expressive movements in rhythmic coordination with a person’s speech (as was first noted by Condon & Sander in 1974), and this can occur when the baby has been blind from birth, and thus never seen its hands, or anyone else’s hands (Tønsberg & Hauge, 1996 ; Trevarthen, 1999 a). Thus we may conclude that the infant has a coherent psychoneural organisation that specifies the timing and form of body movements. This organisation can react with appropriate dynamic changes to another person’s dynamic expressions, match-ing their rhythms and accents. Evidently the responses of the infant are made expressive by internally generated motives and emotions that resemble those carried in the adult expressions. Infant and adult can, for a time,
sympathise closely and apparently equally with one another’s motive states, using similar melodic or prosodic forms of utterance and similar rhythms of gesture. This entails an absorption of the adult’s motivations into an affectionate intuitive parenting mode that tends to mimic the infant and that releases in the adult a specialised, emotionally coordinated ‘‘ musicality ’’ of voicing, with animated but sympathetic and joyful facial expressions and dance like postural, gestural movements that match vocal expressions, and affectionate and playful touching and moving of the infant’s hands, face, or body (H. Papousek & Papousek, 1977, 1987 ; Stern, 1974, 1985, 1993).
Intersubjectivity of Neonates and Foetuses
It has been assumed in medical science and psychology that a human newborn, lacking coherence of psycho-logical representation even of itself as a subject, cannot ‘‘ distance ’’ itself perceptually or conceptually from the adult who cares for it. That, consequently, the relation-ship with the mother is one of ‘‘ symbiotic fusion ’’, to use the term employed by Mahler, Pine, and Bergman (1975). In the same way, the British Object Relations School of psychoanalysis, while developing a framework for ap-preciating the emotional needs of infants, took it as evident that newborn is confluent emotionally with the mother, and emerging to self-awareness within her rational consciousness in a growing attachment to her person (Stern, 1985). With the exception of Fairbairn (Grotstein & Rinsley, 1994), object relations theory (Bion, 1962 ; Guntrip, 1971 ; Klein, 1952 ; Winnicott, 1965) holds that the young baby has no consciousness, no separate ego, no representation of self distinct from the other. These ideas recall ancient philosophical inferences about the primacy of reason, the role of learning by imitation, and the opposition of reason and emotion (Kugiumutzakis, 1998, p. 88).
In fact, while neonates are undoubtedly endowed with reflex ‘‘ panic ’’ responses that serve physiological main-tenance and survival (Panksepp, 1998 a), if a newborn is alert, rested, free of stress, and responded to sympath-etically, voluntary behaviours appear that are well-coordinated, perceptive, and specifically adapted to excite and regulate an engagement with the autonomous expres-sions of interest and emotion of another person, all of which makes the behaviours intensely rewarding for a new mother or father, who feel they are interacting with a human person (Murray & Andrews, 2000 ; Van Rees, Limburg, Smulders, & Kloosterman, 1992). The express-ive behaviours in affectionate chat and play have no immediate role in the regulation of the neonate’s physio-logical state, comfort, or survival. They are distinct from maternal breast-feeding, stroking, holding, rocking, vocal comforting, and the like. The caregiver responds to neonatal signals that are very different from appetitive movements, distress cries, or gestural signs of fear, anger, or fatigue. The interactions are calm, enjoyable, and dependent upon sustained mutual attention and rhythmic synchrony of short ‘‘ utterances ’’ which include, beside vocalisations, touching and showing the face and hands, all these expressions being performed with regulated reciprocity and turn-taking. Newborn and adult spon-taneously display a mutually satisfying intersubjectivity (Kugiumutzakis, 1998 ; Trevarthen et al., 1999).
seen in interspecies communication between a human adult and a newborn chimpanzee (Bard, 1998 ; Bard & Russell, 1999), which proves that face-to-face trans-mission of the basic intersubjective motives is not restricted to humans. The state-regulating communi-cative behaviours and reactions of infants resemble the needs for parental attention shown by the helpless young of many other mammals (Blass, 1994, 1996 ; Carter, Lederhendler, & Kirkpatrick, 1997 ; Hofer, 1990 ; Mc-Kenna & Mosko, 1994 ; Panksepp, 1998 a ; Panksepp, Nelson, & Bekkedal, 1997 ; Panksepp, Nelson, & Siviy, 1994 ; Rosenblatt, 1994 ; Schore, 1994 ; Suomi, 1997).
By recording changes in their heartbeat with attention to novel events, where they chose to look, or by causing their head rotations, leg movements, or sucking to trigger stimuli, it has been possible to show in experimental situations that newborns are sensitive to expressions of emotion in body movements and touching, voice, or facial movements (Bower, 1982 ; DeCasper & Carstens, 1981 ; Eisenberg, 1976 ; Jusczyk, 1985 ; Lipsett, 1967 ; H. Papousek, 1967). However, the most striking evidence for an innate protoconversational readiness comes from intended imitations and provocations of newborns in close reciprocal interaction with adults who are seeking to make their behaviours interesting for, and contingent with, the infant’s signs of attending. Infants only a few hours old are capable of expressing communicative capacities adapted for psychological self-other regulation (Aitken & Trevarthen, 1997 ; Als, 1995 ; Blass, 1999 ; Brazelton, 1984 ; Brazelton et al., 1974, 1975 ; DeCasper & Carstens, 1981 ; DeCasper & Fifer, 1980 ; Heimann, 1998 ; Kugiumutzakis, 1998, 1999 ; Meltzoff, 1985 a ; Meltzoff & Moore, 1977, 1994, 1997, 1998 ; Nagy & Molna! r, 1994; Reissland, 1988; Trevarthen, 1979, 1997a; Trevarthen et al., 1999 ; Zeifman, Delaney, & Blass, 1996). The infants act assertively or apprehensively in appropriate coordination with the assertive phases or watchful apprehensive states of a sympathetic partner (Trevarthen et al., 1999). This active involvement in communication of rudimentary intentions and feelings confirms that the human mind is, from the start, motivated not only to elicit, guide, and learn from maternal physical care to benefit regulation of the infant’s internal biological states, but also for cooperative psycho-logical learning—the mastery of socially or inter-personally contrived meaning specified in intelligent reciprocal social engagements (M. C. Bateson, 1979 ; Bra/ ten, 1998; Dore, 1983; Halliday, 1975, 1979; Hubley & Trevarthen, 1979 ; Newson, 1979 ; H. Papousek & Papousek, 1977 ; Ratner & Bruner, 1978 ; Ryan, 1974 ; Tomasello et al., 1993 ; Trevarthen, 1980, 1987, 1988, 1994 ; Trevarthen & Hubley, 1978 ; Trevarthen & Logotheti, 1987 ; Trevarthen & Marwick, 1986).
In short, infant survival and development depends on communication with a caregiver to service the baby’s needs for an emotional attachment, but also to maintain and develop an intimate emotionally expressed
com-panionship in changing purposes and conscious
ex-periences (Trevarthen, 1998 d, in press). The infant’s mind and body has special functions adapted to anticipate this development of the imagination of meaning (see Fig. 2).
Even an infant born more than 2 months before term can begin to share dynamic protoconversational motives in precisely regulated rhythms of purposeful movement and investigative awareness by exchanging facial ex-pressions, vocalisations, and gestures of the hands with a
sympathetic partner (Trevarthen et al., 1999 ; Van Rees & De Leeuw, 1993). Learning tests that examine preferential orienting or autonomic regulations of newborns prove that perceiving the mother’s rhythmic vocal expressions of motive state from her speech can begin in utero, many weeks before birth (DeCasper & Spence, 1986 ; Fifer & Moon, 1995 ; Hepper, 1995 ; Lecanuet, 1996). Her charac-teristic patterns of speech can be identified by her newborn immediately. Recognition of the visible ap-pearance of the mother’s face is acquired within hours of full-term birth, aided by the newborn’s coherent or integrated capacity for interest in interaction with the feelings behind the other person’s facial expressions and vocalisations (Bushnell, Sai, & Mullin, 1989 ; Field, Cohen, Garcia, & Greenberg, 1984 ; Field et al., 1983 ; Field, Woodson, Greenberg, & Cohen, 1982 ; Goren, Sarty, & Wu, 1975 ; Heimann, 1989, 1998 ; Heimann, Nelson, & Schaller, 1989 ; Heimann & Schaller, 1985 ; Kugiumutzakis, 1993, 1998, 1999 ; Maratos, 1982 ; Meltzoff, 1985 a ; Meltzoff & Moore, 1977, 1994 ; Nagy & Molna! r, 1994; Reissland, 1988; Zeifman et al., 1996).
The complex adaptive structure of the foetal human brain, and notably the peripheral organs and neural systems of social or interpersonal perception and ex-pression, determine directions and limits to future ac-quisition of skills or knowledge by a child (Als, 1995). Although it has been shown that prematurely born infants can imitate facial expressions (Field et al., 1983 ; Kugiu-mutzakis, 1985, 1998), it appears likely that the auditory learning of foetuses and the vocal imitations of premature newborns may be related to the precocity of the auditory sense and its special reception of other persons’ ex-pressions (Mehler et al., 1988). Development of audition may be inhibited after birth by the sudden acceleration of development of the visual system that occurs in early months (Lecours, 1982). Longitudinal data (Kugiu-mutzakis, 1999, pp. 42, 44, 48) indicate that vocal imitations may be declining immediately after birth, as imitations of seen mouth movements increase, and then vocal imitations pick up after 2 months.
Imitation by infants is not mere reproduction or repetition of movements made by another individual, and it serves interpersonal functions, not just acquisition of motor skills and expression (Kugiumutzakis, 1993, 1998, 1999 ; Uzgiris, 1981, 1984). It is, even for newborns, an emotionally charged mutual influence of motive states in which certain salient expressive actions of the other are identified and repeated to further an ongoing communi-cation (Nagy & Molna! r, 1994, 1997). Imitative responses occur at a moment in the stream of interaction where they can act as affirmations, acceptances, or commentaries with respect to accentuated displays of the other person (Trevarthen et al., 1999). Older infants and toddlers imitate to display and reinforce friendship or affiliation, showing great sensitivity to pleasure and praise shown by familiar companions. But even in young infants, imi-tations serve to qualify an attachment relationship (Meltzoff & Moore, 1994), possibly to identify an individual person as an object of heightened affect—i.e. of love or admiration.
The manner of imitating proves the natural complexity and specificity of the infant’s motives for human contact and communication. The imitative reactions recognise communication, as a hand-shake or a head-nod does between adults in conversation, and they have a further peculiar feature. An inventory of the actions that neo-nates can imitate reveals a rather bizarre set : e.g. large
tongue protrusions, exaggerated opening of the mouth or eyes, looking back over the head, holding up a hand, extension of one finger or two fingers, single vocal sounds (vowels) emitted in a rhythmic burst. All these appear to be forced or emphatically marked innovations, not normal spontaneous currency in mutual affective regu-lation, such as ordinary smiles, staring, crying, or frowning. The imitations are emitted after prolonged attention and with effort, and they show improvement in accuracy over a series of repeated attempts. Where it has been found that neonates imitate expressions of emotion (Field et al., 1982), these have been responses to the exaggerated performances of an actress, not normal contingent responses in reciprocal communication. This appears to indicate that, even at this age, imitation is part of a motivation specialised for purposeful negotiation and learning of new or arbitrary social habits or con-ventions, in the form of behaviours that have been isolated and given emphasis in the stream of engagement. Further research on the timing of neonatal imitations is needed to establish this curious point.
Intuitive Parenting and Maternal Speech : Sympathetic Emotions Evoked by the Infant’s Expressions of Feeling, Initiative, and Curiosity
The very distinctive manner of an affectionate adult’s vocalisations to a young baby has now been analysed in detail. ‘‘ Motherese ’’ or infant directed speech (IDS) has defined rhythmic and melodic features as well as voicing qualities. It is organised in repeated phrases, and tends to create slowly changing, cyclic narratives of emotion. The similarity of this speech and vocal play to music and to the rhythmic and rhyming forms, phrases, and verses of poetry has drawn researchers’ attention, and led to a concept of preverbal or subverbal ‘‘ musicality ’’ as a fundamental basis for communication of motives and feelings (Malloch, 1999 ; H. Papousek, 1996). Infants have been found to have astonishing powers of discrimin-ation for subtle features of musical sounds and melodic forms, especially as these are represented in the inflections of a mother’s voice (see below). These features are evi-dently manifestations of a fundamentally innate pro-cess of emotional physiology, by expression of which a primary level of intermental communication is estab-lished between human subjects.
The dynamic narrative envelopes of a mother’s utter-ances, their pitch contours, and other dynamic qualities, have been identified as necessary alimentation for the infant’s developing self-awareness and consciousness of agency (Beebe & Lachmann, 1988 ; H. Papousek & Papousek, 1977, 1987, 1989 ; Stern, 1974, 1985, 1993). On the other hand, the extraordinary precision of the infant’s mirroring, even with a restricted vocal repertoire, has been taken as further evidence for an innate capacity for such ‘‘ on-line ’’ communication (Beebe et al., 1985 ; Stern & Gibbon, 1980 ; Stern et al., 1975, 1977, 1985 ; Trevarthen et al., 1999). Infants, even newborns, can exactly synchronise with certain salient moments in the adult’s message by gesture or utterance, and the vocal emissions can be matched in pitch and quality (timbre) (Malloch, 1999). There clearly is a sensitive two-way mirroring of the emotional values of expression in spite of the great difference in maturity between the participants. Speech directed to infants with concern for their
interest, like speech addressed sympathetically to pets (Burnham, 1998), or to very old people who often think slowly and are hard of hearing, has exaggerated, but modulated, expressivity. It clarifies the feelings, interests, and intentions of the speaker, and it minimises the remembering of meanings of words. This talk is under-standable as effective communication only if it is accepted that even young infants are as sensitive to the feelings behind consciously regulated well-motivated utterances as an old person or a cat. As Bateson, an anthropologist and linguist, pointed out, it is a form of human com-munication that is related not only to education in the forms and meanings of language, but also to the rhythms and melodies of religious ritual and communion, and traditional healing practices (M. C. Bateson, 1979, pp. 74–76).
Comparison of parents’ speech to young infants in different languages confirms that there are universal rhythmic and prosodic features in the expression of human feelings and sympathetic interest (Fernald, 1992 a ; Fernald & Simon, 1984 ; Fernald et al., 1989 ; Grieser & Kuhl, 1988 ; M. Papousek, Papousek, & Symmes, 1991 ; Stern, MacKain, & Spieker, 1982). Thus, for example, motherese or IDS in both a tonal language (Mandarin) and in a nontonal language (English or German), com-pared to adult-directed speech (ADS) in either language, has higher pitch (fundamental frequency, F0), larger F0 range, shorter utterances and longer pauses, fewer syl-lables per phrase, and less phrase time\sample time (Grieser & Kuhl, 1988 ; M. Papousek et al., 1991). Thanavisnuth and Luksaneeyanawin (1998) report simi-lar features in Thai mothers’ speech to their infants. Rising contours, used by parents to elicit infant attention (Stern, Spieker, Barnett, & MacKain, 1983), are similar in English and Mandarin (M. Papousek et al., 1991), and this form of utterance is more effective in eliciting and maintaining infant attention than falling pitch (Sullivan & Horowitz, 1983). Infants prefer approving rather than disapproving intonation (M. Papousek, Bornstein, Nuzzo, Papousek, & Symmes, 1990) ; they show more positive affect to this way of speaking (Fernald, 1993) and are more interactive, interested, and emotionally positive to IDS (Werker & McLeod, 1989). Adults, too, judge role-play better from speech in the IDS register.
Maternal speech has often been studied as if it were just an instructive register of language, an aid for the infant to pick up words and sentence grammar. But cats pre-sumably do not understand words, and nor do 2-month-olds. With toddlers or the aged, linguistic communication may also be part of the effective function of this way of uttering, but its obvious attractiveness and regulatory effects with the youngest infants can have little to do with the grammatical or semantic purposes of language. Vocal communication addressed to infants is, for the infant and largely for the adult too, nonreferential, in the sense that it does not matter that it may specify any reality or object outside the human contact itself. It is intersubjective at a fundamental level. It serves to respond to or affirm the infant’s eagerness to become involved in proto-conversation, which is a nonverbal discourse regulated by dynamic relational affects, and a ‘‘ narrative ’’ sense of transforming feelings.
Research on IDS, at first reacting to the theory of an innate language acquisition device, which argued that language input to infants is so linguistically impoverished that it couldn’t possibly teach grammar (Chomsky, 1965), sought to demonstrate that mothers do provide a graded
instruction in features of language (Snow & Ferguson, 1977). Then acoustic speech analysis led to the dem-onstration that speech to infants has higher pitch, wider pitch excursions, slower tempo, shorter utterances, and longer pauses (Cruttenden, 1994 ; Fernald & Simon, 1984 ; Stern et al., 1982), and it was claimed that the training was perceptual rather than linguistic. But infants clearly already have perceptual biases and attentional preferences that favour awareness of parental speech (R. P. Cooper & Aslin, 1990 ; Fernald, 1985 ; Gleitman, Gleitman, Landau, & Wanner, 1988 ; Pegg, Werker, & McLeod, 1992 ; Werker & McLeod, 1989). Thus three different functions have been attributed to the way adults talk to infants : this speech engages attention (M. Papousek et al., 1991 ; Stern et al., 1982 ; Sullivan & Horowitz, 1983), communicates affect, facilitating social interaction (Fernald, 1989, 1992 a ; Kitamura & Burnham, 1996, 1998 a ; M. Papousek et al., 1990 ; Werker & McLeod, 1989), and facilitates language acqui-sition (Fernald & Mazzie, 1991 ; Hirsh-Pasek et al., 1987).
Experimental Tests of Infants’ Emotions in Protoconversation
The motives and emotions of protoconversation with infants under 3 months of age have been tested by
perturbation experiments, situations that have been
con-trived to obstruct or distort the rhythmic traffic of expressive signals and contingent and sympathetic responses between infant and adult. Two procedures have given valuable evidence on what infants expect from the behaviour of a conversational partner, and how they act when the adult fails to meet these requirements. They confirm that the interaction is generated by coregulation, coconsciousness, and contingent and reafferent mutual regulation in a complex dynamic system wherein the exact course of events is emergent or not defined in advance (Fogel, 1993 a, b ; Fogel & Thelen, 1987 ; Tronick & Weinberg, 1997). They also show that the young infant has expectations of the emotional quality of the en-gagement and the normal contingencies of a sympath-etic adult response, and that these emotions change in ways that affect the adult, regulating positively towards a happy encounter, and defending against fail-ure of contact, by appealing with negative emotional expressions for appropriate remedial action to repair communication.
The still or blank face test (Murray, 1980 ; Murray & Trevarthen, 1985 ; Trevarthen et al., 1981 ; Tronick et al., 1978) requires a mother who has established a proto-conversational interaction with her infant to arrest her movements on a signal from the experimenter, and simply look at the infant without any reaction to what the infant does. This commonly results in the infant showing a succession of appeals for communication by smiling, vocalising, and gesturing, punctuated by increasingly sober staring at the mother, then emission of signs of avoidance of eye-contact and distress. The behaviour has been charted by micro-description and proved by stat-istical analysis and to be a coherent emotional reaction that shows the infant is disturbed or made unhappy by the mother’s unresponsiveness. Indeed, the infant’s be-haviour assumes the configuration and interpersonal timing of an expression of sad avoidance, an expression which, in an older person, we would not hesitate to call
distressed embarrassment or shame (see below for a discussion of the ‘‘ nonbasic emotions ’’ of infants).
A second experiment was designed to deal with the objection that the infant was simply affected by the mother’s sober face and inactivity (Murray, 1980 ; Murray & Trevarthen, 1985 ; Trevarthen et al., 1981). A double video (DTV) link was set up so that infant subjects a few weeks old (less than 3 months) and their mothers could see each other and communicate by seeing one another’s face expressions and hearing vocalisations live. Once good, happy communication was obtained, a portion of the recording of the mother approximately 1 minute in length from an animated and playful period of the encounter was rewound and replayed to the infant. The projection of the mother’s behaviour to the infant was exactly as before, but the physical recording was not, in any reliable way, reacting contingently to what the baby was expressing at any moment. Here infants showed occasional accidental interaction with the taped behav-iour of the mother, confusion when she failed to respond in time and appropriately, then prolonged distress and avoidance as in the still face experiment. It takes time for the infant to recover from this perturbation when the mother resumes normal sympathetic communication, or is on-line again, as was the case in the still face experiment (Weinberg & Tronick, 1996). Replay of the infant’s behaviour to the mother in the DTV apparatus causes her to feel something is wrong, and different mothers ex-perience different emotions and make different verbal evaluations, all uncomfortable, when the infant appears not to connect.
Replication of this DTV replay experiment confirms that 2-month-olds are highly sensitive to the timing and emotion of a mother’s expressions in communication (Nadel, Carchon, et al., 1999). Evidently the infant, at 6 to 12 weeks of age, is able to anticipate and join a sympathetic ‘‘ conversation ’’, and is distressed by mis-timed maternal expressions, no matter how joyful and playful they may be. As we shall explain, this finding is resolutely contested by proponents of the view that infants under 3 or 4 months (or even much older) lack (have not yet constructed) a coherent, intentional ‘‘ self ’’, and therefore do not perceive agency in another person, and cannot be sensitive to the purposeful contingency of another person’s communicative responses (Rochat, Neisser, & Marian, 1998). According to this theory, the self-awareness required for awareness of the other as an agent is a product of acquired social cognition (Lewis, 1999)
Research on the effects of maternal postnatal de-pression, which causes the mother to express herself without pleasure, with flat affect, and with erratic timing of behaviours that do not engage with the infant’s behaviours, leads to the same conclusion as the per-turbation experiments, as is discussed further below. Young infants seeking communication from a depressed mother are affected by unsympathetic and inappro-priately timed maternal behaviour (Breznitz & Sherman, 1987 ; Field, 1992 ; Lundy et al., 1996 ; H. Papousek & Papousek, 1997 ; Tronick & Weinberg, 1997), the mother’s self-referred, unresponsive state (Murray, Kempton, Woolgar, & Hooper, 1993) and the quality of speaking that lacks ‘‘ musicality ’’ (Robb, 1999). They become distressed and avoidant and may develop a lasting depressed state that affects communication with persons other than the mother (Field, 1992 ; Lundy et al., 1996).
Changes in older infants’ reactions to the two per-turbation tests show that a capacity to withstand dis-engagement of a conversational game without distress increases with the infant’s increasing alertness and curi-osity for the environment at large. Infants over 4 months easily engage in agile visual investigation of surroundings, and they can use this new interest to escape an unres-ponsive mother’s gaze. Whereas the 2-month-olds appear to be trapped in the stressful encounter and usually become seriously disturbed, the older infants are much less concerned by brief unresponsiveness of the mother (Biglow, MacLean, & MacDonald, 1996 ; Hains & Muir, 1996) ; they simply avoid looking at her (Muir & Hains, 1999 ; Trevarthen, 1984 a, 1990 a). Further age-related changes in infants’ resistance to distress and separation will affect how they behave in any situation where the mother’s responses are unusual, and when they are older than 6 months infants often regard the still face test as an entertaining game (Trevarthen, 1998 a, p. 40). The tests with 2-month-olds engage the protoconversational motives that are active in ‘‘ primary intersubjectivity ’’, before motives for investigative looking and manipu-lating have become strong, and before the infant has developed a robust self-confidence in game playing, teasing, and showing off.
Developments of Intersubjectivity in the First Year : Age-related Events and Changing Parental
Responses
Perceptuomotor Maturation in Infancy : Increasing Body Awareness and Transitions in Motive and Emotion
Infants consistently show conspicuous age-related changes, not only in their physical size, acuity of perception, and motor strength, but in the coordination of their movements for different purposes, their per-ceptual discrimination and attentional preferences, and their affectionate and cooperative engagements with caregivers. These changes are reflections of trans-formations in brain function that have intrinsic causes. They are clearly also consequences of experience and learning, about the body and of the world and objects, and therefore dependent on the kind and quantity of experiential input. Infants learn recognition of human signals from before birth, and at all ages they react strongly to the emotional support they receive from other persons, and the attentions given to their interests and actions. But they also have their own powerful internal impulses and motivations, and these are always important factors in development of the infants’ awareness and motor coordination, and in their responses to self-generated experience, as well as to care or teaching.
Figure 1 summarises the evidence we have concerning age-related changes in infants’ behaviour. These appear to express periods of rapid change (PRCs) in infants’ psychological motives and capacities for action, cog-nition, and communication (Trevarthen, Aitken, & Plooij, 2000). Four main epochs can be defined, and these appear to reflect changing balance in three principle kinds of intrinsic motive : (a) ergotropic, for attending to the external world with the aid of motor adjustments of the body and selective use of the senses ; (b) trophotropic, for regulation of the internal autonomic or visceral state ; and (c) communicative, this last being effective in regulating
both attachment, serving the developing infant’s tropho-tropic needs, and companionship, by which experiences and skilful actions directed to the environment are shared and learned socially (Trevarthen, 2000 ; and see Fig. 2). The ergotropic\trophotropic distinction in animal motiv-ations was first made by Hess (1954) on the basis of physiological effects of brain stimulation. For modern evidence on motive systems of the mammalian brain see Panksepp (1998 a).
Experimental cognitive psychology has focused prin-cipally on the ergotropic, environment-assimilating func-tions of the infant as an individual perceiver and actor, exploring and using objects, observing events, and ac-quiring skills of perceiving and acting. Consequently primary importance has been given, first, to the very conspicuous changes in visual focus at 4 to 6 weeks, then in attention and manipulation with advances of postural control, visual orienting, and discrimination, and reach-ing and graspreach-ing around 3 and 4 months (e.g. Rochat & Striano, 1999). A second major change in cognition comes toward the end of the first year as the infant, on the threshold of independent locomotion, shows more de-liberate interest in pursuing the kind of purposes which adults see as intentional and becomes more capable of solving problems of the ways objects interact and can be used together.
These changes, at these three ages, have been taken by developmental authorities as the beginnings of various functions of consciousness, object concepts, volition, and self-awareness. The program of developments in object cognition has been assumed to set the pace and strategy for ‘‘ social cognition ’’, a product that must incorporate the capacity for relational emotions, empathy, and intersubjectivity (e.g. Bahrick & Watson, 1985 ; Baron-Cohen, 1994 ; Gergeley & Watson, 1999 ; Izard, 1978, 1994 ; Lewis & Brooks-Gunn, 1979 ; Mahler et al., 1975 ; Piaget, 1954, 1962 ; Rochat & Striano, 1999 ; Rothbart, 1994 ; Schore, 1991 ; Sroufe, 1996 ; Stern, 1985 ; Toma-sello, 1993 ; Yarrow et al., 1984 ; Zahn-Waxler, Radke-Yarrow, Wagner, & Chapman, 1992). An alternative view holds that self-awareness and emotional maturity comes principally with language and social training, after infancy (Dunn, 1994 ; Lewis, 1987, 1992, 1993). Such conclusions can only be supported if the ways in which younger infants, including newborns, react purposefully and emotionally to the environment are disregarded, and their reactions to the communicative signals from other persons are explained as reflexive or ‘‘ mindless ’’. We believe that the changes through the first and second year or childhood are more accurately seen as developmental transformations in prenatally drafted motives that are adapted for intelligent life in the company of other subjects, not the first appearance of the adaptive be-haviours.
Research on the social development of infants focused on the supposed dependence of communication on cognitive development, or on the dependence of the child for emotional development on parental regulation, and, on imitation, takes the child’s self-awareness to be a construct built of experiences acquired concerning how other persons react to what the child does, inferring purposes where they did not exist (Kaye, 1982). The increasing self-consciousness of the infant in the second 6 months of life has been taken as evidence for the beginning of a representation of other individuals’ inten-tions, or intersubjectivity. Advances in the dynamics of interaction and in social signalling are supposed to reflect
Regulation of sleep, feeding and breathing. Innate “pre-reaching”. Imitation of expressions. Smiles to voice.
Fixates eyes with smiling. Protoconversations. Mouth and tongue imitations give way to vocal and gestural imitations. Distressed by “still face” test. “Person-Person” games, mirror recognition.
Smooth visual tracking, strong head support. Reaching and catching.
Imitation of clapping and pointing. “ Person-Person-Object ” games. Accurate reach and grasp. Binocular stereopsis. Manipulative play with objects. Interest in surroundings increases. Playful, self-aware imitating. Showing off. Stranger fear. Persistent manipulation. Babbling and rhythmic banging of objects. Crawling and sitting, pulling up to stand. Cooperation in tasks; follows pointing. Declarations with joint attention. Proto-language. Clowning. Combines objects, executive thinking. Categorises experiences. Walking. Self-feeding with hand. Beginning of mimesis of puposeful actions,
uses of tools and cultural learning. May use first words.
A
B
C
D
E
F
G
Figure 1. Top: In protoconversation, a two-month-old infant and a mother communicate by many modalities of perception and expression, transmitting information about intrinsic motive rhythms and emotions, principally by eye-to-eye contact, voice, facial expression and gesture. Middle : In the first 18 months of life there are marked changes in the infant’s consciousness of other persons and in their motives for communication, without language. Several major transitions can be observed in self-and-other awareness at particular ages. These lead the child toward cooperative interest in actions and objects, and cultural learning. Below : Research studies that have made detailed longitudinal observations at sufficiently frequent intervals have found evidence for major periods of rapid change (PRCs) in motor coordination, perceptual abilities, and communication. All may be described as elaborations of the means by which the initial purposeful, consciously regulated, and intersubjective motives of the newborn infant may be employed to further
change to the social ‘‘ scaffolding ’’ offered to the infant as parents are more playful, presenting game routines, with the infant’s initiative in these changes of communication being regarded as consequences of perceptual and mo-toric developments. The infant’s realisation that other persons are ‘‘ like me ’’ (Mead, 1934) is taken to come as a consequence of acquired contemplative or metacog-nitive functions (Leslie, 1987), developing expectations regarding the dynamic vitality features of other persons, and, especially, an emerging sense of variable contingency in their responses (Bahrick & Watson, 1985 ; Gergeley & Watson, 1999 ; Watson, 1984).
While not denying the importance of learned regu-lations in the management and recollection of experi-ences, and the influence parents may play through their interpretation of infant’s expressions of emotion and ‘‘ behavioural responses ’’, we see the conspicuous de-velopments in infant’s self-consciousness and sociability in this ‘‘ period of games ’’ as continuous with, and de-veloping from, the rhythmically patterned intersubjective motives that were present and active in the newborn. We believe that development is fostered best when parents respond with perceptive sympathy to the motives and feelings infants express to them.
As the infant gains in alertness, discriminative aware-ness, and power of movement in the early weeks, turning more to explore the environment and manipulate objects, exchanges with parents become more lively (Trevarthen, 1990 a, 1998 a ; Trevarthen & Hubley, 1978). After 3 months protoconversations give way to body games, nonsense rhymes, nursery chants, and songs and to ritual play routines involving body bouncing, hand-clapping, tickling, ‘‘ peek-a-boo ’’, and the like (Bruner & Sher-wood, 1975). These are always strongly rhythmic, with regular phrasing and highly predictable repetitions and resolutions of emotional energy or excitement (Tre-varthen, 1999 a). They involve the easy substitution and matching of forms of expression in different modalities (hand gestures, voice sounds, face expressions, looks, and head orientations), which Daniel Stern called ‘‘ inter-modal fluency ’’, in the emotional complementation that he describes as ‘‘ affect attunement ’’ (Stern et al., 1985). By 4 months or so, infants are clearly interested in and responsive to the adults’ changing mood and expressions of excitement, surprise, pleasure, or displeasure, and, with familiar persons, they can appreciate complex teasing games (Nakano & Kanaya, 1993). Imitation games between infants and their mothers and fathers, or between infants, are emotional, and usually very pleasur-able (Fiamenghi, 1997 ; Reddy et al., 1997 ; Trevarthen et al., 1999). Older infants, when they are caught in unfamiliar circumstances, orient purposefully to check their mothers’ emotions (Klinnert, Campos, Sorce, Emde, & Svejda, 1983). It follows that any implication that the parent is giving organisation to the infant by ‘‘ scaf-folding ’’ an erection of immature moves should be qualified by the observation that the adult is often assiduously tracking the infant’s varying mood with imitations, and that the infant can take the role of provocateur or teaser (Reddy, 1991). The infant can be the one who attunes to and accompanies the parent with nicely synchronous gestures or vocalisation, even show-ing anticipation of salient events, such as the prolonged vowel at the end of a phrase or stanza (Malloch, 1999 ; Trevarthen et al., 1999). As in protoconversation, there are frequent occasions in games or songs when the infant takes the role of leader, while the mother responds.
Communicative Musicality in The First Year The subtleties of early preverbal interaction are being studied in detail by methods for objectively assessing the prosody and melody of infant-directed vocalisations and musical sounds, and recording infants’ orientations and preferences to such sounds. The evidence is that infants are selectively attracted to the emotional narratives carried in the human voice, and that they are excited to participate in a shared performance that respects a common pulse, phrasing, and expressive development. Infants respond with synchronous rhythmic patterns of vocalisations, body movements, and gestures to match or complement the musical\poetic feelings expressed by the mother. As infants become more energetic and alert, mothers’ songs and games become more lively. They develop ritual forms which are often repeated, to the great satisfaction of infant and parent. The mood of the mother’s performance changes with the state of alertness and humour of the baby, and reacts with a soothing, calming mode when the infant is tired or distressed. The songs can modulate the emotional state of the infant and the extent to which he or she engages in communication. By 6 months of age, in laboratory discrimination tests, infants respond differently to play songs and lullabies, types of song that are easily recognised by adults. Play songs are associated with increased alertness to the external world and joint attention, whereas lullabies result in more self-focused infant behaviours (Rock, Trainor, & Addison, 1999). These developments parallel, or accompany, changes in the ways parents talk to older infants, conspicuous changes occurring in both musical forms of play and in the affective and directive forms of speech in different languages, first towards 3 months, and then between 9 and 12 months (Kitamura & Burnham, 2000 ; Thanavisnuth & Luksaneeyanawin, 1998 ; Tre-varthen & Marwick, 1986). There are interesting sex differences in these developments, indicating not only that females may be developing slightly ahead of boys in communication, but that they are more responsive of and stimulating to both affective and directive functions of mother’s speech after 9 months (Kitamura & Burnham, 2000 ; Masur, 1987 ; Papaeliou, 1998 ; Thanavisnuth & Luksaneeyanawin, 1998).
Just as infant-directed speech and singing tends to be higher-pitched, slower in tempo, and more repetitious in content than talk addressed to older children or adults (Trainor, 1996 ; Trehub et al., 1997), infants’ responses to female singers confirm that parents’ propensity to interact with infants using a higher vocal range is paralleled by the infants’ preference for higher-pitched singing (Trainor & Zacharias, 1998). However, further investigation has shown that infants are not interested so much in the pitch of singing per se. They are perceiving and mirroring the narratives of emotion in the voice, as we explain below.
The intuitive time-patterns and motive contours of protoconversation and mother-infant games can be car-ried by any means of sensory-motor contact. A recent study focused on the importance of variations in maternal touch and hand gestures during interaction with infants (Stack & Arnold, 1998). Sixty mothers were videotaped with their 5"#-month-old infants during four phases of interaction. At certain times they were instructed to use only touch and gesture, at others to attend to the infant’s face, and at others to engage in normal interaction with their infants using vocalisation as well. Mothers were able to engage successfully with their infants using touch and
gesture alone. This is in accord with studies of the rhythmic tactile forms of communication that familiar and experienced partners use to make contact with profoundly mentally handicapped children or young adults (Burford, 1988, 1993 ; Burford & Trevarthen, 1997 ; Trevarthen & Burford, in press), as well as the findings of research into the most effective ways of supporting self-regulation, communication, and cognitive development in infants and children with sensory loss, including those both deaf and blind (Tønsberg & Hauge, 1996).
Research on infant’s attentions and preferences brings evidence that the essential features of what we may call the intrinsic motive pulse (IMP) of musicality (Trevar-then, 1999 a) are possessed by infants—they are shown in play with adults, or when infants are responding to artificial fragments of musical sound in laboratory tests. Infants listen with perceptive preferences to the melodies of speech, singing, and music, and songs and music make them move in rhythm and register interest and happiness (Baruch & Drake, 1997 ; Demany, 1979 ; Fassbender, 1996 ; Fridman, 1980 ; Lynch, Short, & Chua, 1995 ; H. Papousek, 1996 ; M. Papousek, 1994, 1996 ; M. Papousek & Papousek, 1981 ; Stern, 1971, 1974, 1993 ; Stern et al., 1977 ; Trehub, 1987, 1990 ; Trehub, Trainor, & Unyk, 1993 ; Trehub, Schellenberg, & Hill, 1997 ; Trevarthen, 1986 a, 1987 ; Trevarthen et al., 1999 ; Zentner & Kagan, 1996).
Twenty-five years ago, Condon and Sander (1974) reported ‘‘ entrainment ’’ of newborn arm movements to the syllabic rhythms of adult speech in any language. Since the infant can generate the same arm rhythms without any external guide (Trevarthen, 1974, 1984 b, c ; Von Hofsten, 1983), this coordination is evidently not a passive locking-on of the infant to the adult ‘‘ pace-maker ’’, but a sympathetic and flexible cross-modal, or amodal (aud6itory to proprioceptive, and possibly to visual) monitoring of actively generated impulses in infant and adult (Trevarthen, 1986 a ; Trevarthen et al., 1999). In early protoconversations, when the infant is 6 weeks old, alternation or turn taking on a slow adagio (1 beat in 900 milliseconds or 70\minute) is set up. Within a month or two, in animated games, the beat of shared vocal play with an infant accelerates to andante (1\700 milliseconds; 90\minute) or moderato (1\500 milliseconds ; 120\minute). Different qualities of engage-ment are determined by shared emotions organised mutually in the communications. Homologous feelings and changes in affect of infant and caretaker generate harmony, sympathy, support, comfort, restraint, or antagonism.
In the first 6 months, the emotions become strung together in increasingly impassioned ‘‘ plots ’’, in which protagonists play expressive parts to each other. Stern describes feeling qualities transmitted with distinctive activation contours, which are ‘‘ captured by such kinetic terms as ‘ crescendo, ’ ‘ decrescendo, ’ ‘ fading, ’ ‘ explod-ing, ’ ‘ bursting, ’ ‘ elongated, ’ ‘ fleeting, ’ ‘ pulsing, ’ ‘ wavering, ’ ‘ effortful, ’ ‘ easy, ’ and so on ’’ (Stern, 1993, p. 206). In Stern’s terms, these give ‘‘ vitality forms ’’ to the emotions (vitality affects), which would seem to be homologous with the ‘‘ sentic forms ’’ described in musical expression of feelings by Manfred Clynes (Clynes, 1980, 1983 ; Clynes & Nettheim, 1982). The intuitive parenting behaviour of a father or mother at play with an infant shows that the adult is sensitive to the infant’s emotion and unconsciously skilled in giving the right level of emotionally coloured contingent responses (H. Papousek
& Bornstein, 1992 ; H. Papousek & Papousek, 1987 ; M. Papousek, 1996 ; Stern, 1971, 1974, 1992, 1999).
Laboratory tests have proved that, by the middle of the first year, infants hear musical parameters amazingly well (Demany, 1982 ; Trehub, 1987 ; Trehub et al., 1993, 1997 ; Zentner & Kagan, 1996), and they show preferences for these same parameters in the vocal productions of mothers (DeCasper & Fifer, 1980 ; Trehub, 1990 ; Fass-bender, 1996 ; M. Papousek, 1994, 1996). It seems that the infants’ acute ability to hear musical elements in a mother’s voice is important for state regulation by the mother’s sympathetic response to the infant’s expressions of arousal, fretfulness, tiredness, playfulness, joy, etc. (M. Papousek & Papousek, 1981). But the infant is not just responding to the mother’s signals with a reflex state change, and the adult is responding to the rhythms and emotional quality of the infant’s expressions in a joint, two-way ‘‘ performance ’’. They are ‘‘ musicking ’’ per-formance and listening together (Small, 1998). The mother is attuning to musicality in the infant’s ex-pressions and communicating with the infant (M. Pap-ousek, 1994, 1996 ; H. Papousek & PapPap-ousek, 1987 ; M. Papousek & Papousek, 1981, 1989 ; Stern, 1993, 1999 ; Stern et al., 1985).
Sustained orienting of the baby’s head towards a loudspeaker that is presenting preferred sounds has been used to show that infants 4 to 8 months old can discriminate melodic patterns independent of pitch, and melodic contours with variation of intervals (Chang & Trehub, 1977 a ; Trehub, Bull, & Thorpe, 1984 ; Trehub, Thorpe, & Morrongiello, 1985, 1987 ; for a synthesis, see Trehub et al., 1995). Trehub (1990) concludes that ‘‘ infants’ representation of melodies is abstract and adult-like ’’ (p. 437). It has been shown that infants can distinguish pairs of notes separated by one semi-tone, and they can recall a melody based on the tones of the major triad better than one that is atonal. Other tests dem-onstrate that infants are sensitive to tempo and to rhythmic sequences independent of tempo (Trehub & Thorpe, 1989), and that they experience Gestalt ‘‘ group-ing ’’ effects like adults (Chang & Trehub, 1977 b ; Demany, 1982 ; Demany, McKenzie, & Vurpillot, 1977 ; Fassbender, 1993 ; Me! len, 1999a, b; Thorpe & Trehub, 1989 ; Thorpe, Trehub, Morrongiello, & Bull, 1988). They respond to fundamental pitch independent of tonal composition (Clarkson & Clifton, 1985), perceiving and categorising differences in timbre of nonspeech tones (Clarkson & Clifton, 1985 ; Clarkson, Clifton, & Perris, 1988 ; Trehub, Endman, & Thorpe, 1990). They are sensitive to differences in timbre between vowels [a] and [i], in spite of variations in fundamental frequency, duration, and intensity (Kuhl, 1985).
Trehub argues from her data and observations on human voices that ‘‘ the design features of infant music should embody pitch levels in the vicinity of the octave beginning with middle C (262 Hz), simple contours that are unidirectional or that have few changes in pitch direction (e.g., rise-fall), slow tempos (approximately 2n5 notes\sec), and simple rhythms’’ (Trehub, 1990, p. 443). These predictions match well the vocal patterns infants produce in song-like play, the prosodic patterns parents use to excite or calm their infants, and the songs that adults sing to infants (Fernald, 1992 a ; H. Papousek, 1996 ; M. Papousek, 1996 ; M. Papousek & Papousek, 1981 ; Stern, 1999 ; Stern et al., 1977, 1983, 1985 ; Trainor, 1996 ; Trehub, Schellenberg, & Hill, 1997 ; Trehub, Unyk, et al., 1997).
