16 What enables infants to give care? Prosociality and learning by alter-centric participation

In my lecture last fall here in The Academy, marking the beginning of a new academic year in The Centre of Advanced Study, I tried to give you an idea of what we are trying to accomplish here at the Centre in our Theory Forum group, as a part of a more comprehensive researchers' network. Some of our work will be published in a collective source book (for which Cambridge University Press has just offered me a contract). In this evening's talk the focus will be on my own studies of intra- and interpersonal processes in dyadic systems in which care-giving and reciprocity are exhibited. To this question, 'What enables infants to give and reciprocate care?', I shall offer a reply derived from a systems-theoretical postulate which I proposed 1986 (actually at a Gordon Research conference on Cybernetics of Cognition). Four of those scholars who were the first to realize the significance of the postulate and refer to it in their writings are present here tonight -- Sophie Freud, Karsten Hundeide, Ragnar Rommetveit, and Colwyn Trevarthen -- making this a special occasion for me since I shall make a further step derived from that postulate.
It has already yielded a collateral prediction, which now appears to be partly confirmed. I predicted in 1994 that an ego-centric kind of imitation errors would be found in autism in face-to-face situations, reflecting an impairment in the capacity which is the keynote concern of this talk: namely, the capacity for what I term alter-centric participation, i.e. the capacity to virtually move with the facing other; the very reverse of seeing others from an ego-centric perspective. I shall try to indicate how infants can learn from caregivers and afford care to others in virtue of bodily intersubjective communion enabled by alter-centric participation. without any presuppositions about toddlers having to "de-center" in virtue of higher-order
In the conclusion I shall venture a new prediction that may come to invite this new label: "neurosociological", and which some of you will find too reductionistic for your taste. Please, bear in mind, though, my systems-concern with the interplay between nature, nurture, and culture, and that the horizon for this talk is this:"Intersubjective Communication and Emotion in Early Ontogeny" -- the title of the forthcoming source book:. Here we purport, with a comparative eye also on non-human primate communication, to re-write, as it were, some of the story of child development, replacing earlier theoretical views of infants as ego-centric and a-social with a new understanding of infant sociality from the outset, for example, of
- how the newborn can communicate by imitation and soon affectively engage in dialogue-like interplay with adults;
- how babies attempt pre-speech and show perceptual attunement to speech patterns in the community into which they are born; and
- how toddlers sometimes help and comfort other children, i.e. exhibit early prosociality.

Occurrences of early prosociality
So, how come? What enables infants and toddlers to give care? That is the question for this lecture.2 It presupposes that an empirically based positive reply be given to the question of whether they are capable of prosociality. In some previous papers (Bråten 1992, 1996ab) I have referred to case studies and cross-cultural studies of children, including toddlers, even infants, helping or comforting others in nursery, family, and cultural settings, as reported inter alia by Anna Freud (1951, 1973), by Zahn-Waxler and Radke-Yarrow (1979, 1982), and by Elisabeth Whiting and Carolyn Edwards (1988), who works in our group here.
The question about the nature and nurture of pro-sociality cannot be quite dissociated from questions about anti-sociality, even abuse. We know today that the quality of child rearing and socializing background appears to have a profound impact on the manner in which toddlers react towards other children in need or distress (Berk 1994:405; George and Main 1979; Main and George 1985; Trickett and Kuczynski 1986).
Background variables alone do not suffice, however, in an account of the aetiology of early prosociality. There is a nature-nurture interplay here. Aside from individual differences, we have to search for possible mechanisms at play when companions afford models for and experiences of socio-emotional behaviour. I invite you to consider the possibility that the complex nature-nurture equation of prosociality entails some sort of endowed sociocapacity in infants to tune into the feelings of caregivers and peers, and to virtually participate in the other's activity, moving with the others' movements even in an en face situation. In contrast to viewing the adult-infant caregiver situation as one in which the infant merely receives care and observes the adult from an ego-centric position, I shall draw your attention to the infant's apparent capacity to feel to be moving with the caregiver's movements from the caregiver's position. For this I propose a new term: alter-centric participation. This is the reverse of perception of facing subjects from an ego-centric perspective, and complementary to allo-centric perception by the disengaged observer -- for example viewing the table in this room independently of the observer's position.
Let me exemplify by three anecdotal reports, the first two from a wartime nursery in Great Britain, reported by Anna Freud and Burlingham, the last one from my own recording in a home in Oslo. Consider first this incident:
(C 1) "Edith (21 months) had taken off her shoe and sock and tried hard to put them on again. Paul (23 months) watched from a distance, then rushed over to her, sat down on the floor, and took the sock out of her hand. He tried...to put it on Edith's foot, his mouth open, his tongue far out, breathing heavily. Edith watched his face and immediately imitated his expression. For two or three minutes both children were absorbed...an expression of utmost strain on their faces." (Freud & Burlingham 1973)

Here we see demonstrated two pertinent aspects: First, Paul exhibits care-giving before his second birthday, re-enacting, as it were, the kind of helping which other caregivers have afforded him. Second, even though Paul is the author of the efforts of trying to put the sock on Edith's foot, Edith gives evidence of participating in Paul's activity and feelings, attested by the way in which they both are absorbed in his effort, with the same expression of utmost strain on their faces.
Even more telling, demonstrating how a one-and-half-year old is capable of transcending her own bodily ego-centric position and egoistical needs, is this occurrence, again reported by Anna Freud, albeit not with my theoretical spectacles:

(C 2) "Rose (19 months) sat at the table and drank her cocoa. Edith (17 months) climbed up and tried to take the mug from Rose's mouth. Rose looked at her in surprise, then turns the mug and holds it for Edith so that she could drink the cocoa."
(Freud and Burlingham, 1973)

Rose's reaction to Edith's clumsy attempt to get cocoa is most telling, not only because of her nurturant act, helping Edith to drink cocoa at the cost of stopping her own drinking, but because of the way in which she reverses the mug in view of Edith's bodily position. By turning the mug so that Edith can drink from Edith's position, reverse to that of her own, Rose here demonstrates her capacity to transcend her own bodily ego-centric perspective. She gives evidence of what I term alter-centricity.
Here is another example, from my own recordings:
(C 3) Thomas (11 3/4 month) is being spoon-fed by his sister (11 years). When Thomas from time to time is allowed to take the spoon in his own hand, he feeds his sister in return (Bråten 1996b).
What Rose does in the middle of her second year, and what Thomas does before his first birthday anniversary, may look simple to the untrained observer, and puzzling to researchers who in line with Piaget's theory attributes only ego-centricity to toddlers, that is, their being able to see the world and others in it only from their own perspective.
My postulate, in contrast, permits the expectation that what Rose and Thomas do comes quite natural to them. Long before normal children develop generalized means for perspective-taking in G.H. Mead's mediating sense, and for simulating processes in others (like adults do in symbolic interaction, previously proposed by Grice (1974) and Bråten (1974)), infants should be capable of such alter-centric participation in the movements of facing others. This follows from my attributing to the infant a virtual other perspective that complements the infant's bodily self perspective in a asymmetric chiral-like manner (from Greek for 'handedness'), and hence, which permits a virtual mirror reversal in the experience of others' movements in such en face situations in which Rose and Thomas previously have taken part. The postulate thus invites attention to be paid to the face-to-face situational aspect of feeding situations (C3). Reciprocating the caregiving, having learnt to do so by being spoon-fed in previous care-giving situations in the same face-to-face format, this boy -- before his first birthday anniversary -- gives evidence of having participated in the caregiver's movements from her position, which is the reverse of the boy's own bodily ego-centric position. That is what I term learning by alter-centric participation.
Thomas demonstrates his ability to reciprocate the care-giving, but not necessarily from altruistic motives. To him and his sister it is also fun, even funny. But not such great a fun that he shares all the food with her. The sweet dessert that follows the main dish, he keeps to himself. There is the limit to his reciprocating the care-giving. Thomas re-enacts the spoon-feeding from his previous experiences of having felt to be co-enacting such spoon-feeding.
What we see manifested here, however, is of course much more than learning by participation in the caregiver. Being quintessential of caregiving situations, such feeding situations are precursory of the joint meal, the very prototypical Gemeinschaft situation in society and culture. The Latin word companion catches this beautifully: its root meaning is the joint breaking of bread. Thomas and Victoria, his sister (C3), are jointly and reciprocally sharing a meal, as we say, including each other in a shared companion space in the mode of felt immediacy. In the underlying photos and video-recording we see manifestations of bodily intersubjective communion, attested by the way in which they both participate in the other's eating. Unwittingly the sister opens her mouth as she offers the spoon to the infant's mouth, and as he reciprocates he opens his mouth -- now sharing in his sister's mouth opening. They complement and fulfil each other's movements as companion processes in a self-organizing and re-creative dynamic whole, which I have termed "dialogic closure" (Bråten 1988) and which Stern (1995) captures with his term 'proto-narrative envelope'.

Moving with the (m)other's movements
In our culture much of infant-caretaker interplay occurs in face-to-face situations, even when the baby is lying at the nursing table. That is pertinent to my theme of postures and perspectives. Such face-to-face interplay my be contrasted to other cultures, for example the Gusii culture, in which infants are carried on the back of their older siblings (Levine & Levine 1988).

On offspring postures and prosociality in nonhuman primates
It may also be contrasted to other species, carrying their offsprings on the back, such as shown by the lowland gorilla rescuing a boy who fell down in the gorilla pit. This prosocial act occurred in the Brookfield Zoo outside of Chicago on 16 August 1996. A three-year-old boy fell into the gorilla pit, plunging 18 feet to the hard floor. The gorilla mother, Binty, picked up the unconscious boy, carried him across the pit, and laid him gently at the exit door, while zoo keepers came to and kept the other gorillas at a distance with a water hose. (The rescue was captured on video and broadcast on television around the world, and Binty was named The Hero of the Year in the special Winter 1997 issue of Newsweek).
In this occurrence, we see demonstrated a clear case of prosociality in one of our closest species relatives. It serves to remind us of the double sided nature-nurture equation of the aetiology of prosociality. But it also serves to illustrate these three pertinent aspects: First, the way in which the offspring bodily moves with the mother's movements. Second, the offspring experiences to be moving with the mother from the mother's perspective; there can be joint attention from the offspring's ego-centric perspective, without transcending or reversing the mother's perspective. Third, from this perspective the offspring is even afforded the bodily experience of participating in the mother's rescue of a third person.
When old enough to ride on their mother's back, offsprings of gorillas and chimpanzee, in wildlife as well as in captivity, ride in this manner. This means that the offsprings on their backs of the mothers are afforded to move with the mother's movements and to see the world and other subjects and things in it, from the mother's position. From such a position they may learn to see and manoeuvre in the world from the ego-centric perspective which they are afforded in virtue or riding on their mother's back.3

Being hand-guided by the parent
The same kind of opportunity to move with the parent's movement, and in addition being hand-guided, we may see in Gregory Bateson's photos of the Balinese boy Karba (11-12 months) being taught by his parents. Karba is being taught to pick food by his mother and to play an instrument by his father (appendix in Darwin 1955). Facing the same direction as his parent, and in virtue of being hand-guided, Karba bodily feels to be moving with his parent's movements (cf. figure 15.2 in the previous essay, this volume)..

Virtually moving with the model's movements as if the learner were hand-guided
Now, Karba's posture is quite the opposite of the face-to-face situations in Meltzoff's (1988) deferred imitation experiments, in which nine-month-old face the model's performance on an object out of their reach (replicated also by Heimann and Meltzoff 1996). Infants are able to re-enact that performance on the next day, in spite of their being face-to-face with the model's enacting his performance. This face-to-face issue has not been paid attention to by infancy researchers. Now, it was the systems-theoretical postulate that drew my attention to this aspect, opening for a way in which both the pair of Karba and his parent, and the pair of the nine-month-old and the model, could be specified in the same operationally closed format of a dynamic dyadic system, entailing bodily engagement in felt immediacy: In the Balinese case it is manifested in actual bodily movements in concert; in the deferred imitation experiments the subsequent re-enactment by the nine-month-olds suggests that they have virtually been moving with the performing model during the previous model exposure (cf. figure 15.3).
The latter, however, presupposes a reversal of the model's movements as felt by the learner. And in a pre-conference talk on imitation in Paris 1.June 1994, I posed this question, which to my knowledge hitherto had never been asked: What underlies the perceptual inversion entailed by such re-enactment of the other's enactment experienced in a face-to-face situation? Trevarthen (1986) distinguishes alteroception as the motivated perception of others that depends on the specific cerebral response to the other's body movements, and I define the virtual other as a non-specific companion perspective that complements the bodily self perspective with the operational efficiency (virtus) of an actual companion perspective (Bråten 1988). Body movements of actual others included in the companion space of the virtual other afford crude alteroception in an appropriating sense, which in face-to-face situations entails perceptual inversion. Hence, the model and the learner need not face the same direction or use a mirror, unless the novel movements are very complex, or the child has (autistic) learning problems (Bråten 1994).

Subjects with autism
Thus, to be or not to be in a face-to-face situation, makes no difference if you have the capacity to feel to be moving with the actual other's movements, reversed in virtue of the putative bodily-self/virtual-other asymmetry, while it makes a crucial difference if you are deprived of or impaired in the capacity to engage in such felt immediacy with facing others.
Now, in a Behavioral and Brain Science commentary the year before, I suggested that in subjects with autism, internal auto-enclosed cycles are operative without the social-emotional nurture of other's perspectives as felt from the inside (Bråten 1993:515). Unlike Rose who reverts the mug so that Edith can drink (C2), and Thomas (C3) who re-enacts the movements from the caregiver's position in a complementary and reciprocal format in virtue of the mechanism of alter-centric participation, subjects with autism should at best be expected to re-enact the model's enactment as seen and related to their own ego-centric position:

"Difficulties in imitating body movements in autism are here predicted to relate to impairment of the above mechanism and implied to be most acute in face-to-face situations" (Bråten 1994:16)

The same summer that this systems-theoretical prediction was voiced in Paris and repeated in Oslo, an empirical review paper on imitation in autism appeared. Here Isabel Smith & Susan Bryson (1994) point to Ohta's (1987) experimental findings of partial imitation errors in subject with autism trying to re-enact gestural and bi-manual tasks, and which Smith (1995) replicated, yielding the same result: children with autism, as compared to control children, manifested errors in imitating gestures, including rotation of the hand.4 This appears to support the above prediction.
Smith and Bryson refer to a suggestion about "the intriguing possibility that these errors represent attempts of the children with autism to reproduce the gesture so that their view of their own hands matches the view of the model" (Smith & Bryson 1994:264). This suggestion is made and followed up by Barresi and Moore (1996:120). They suspect that the problem in autism is a failure of integrated representation of matched first and third person information. I see us pointing in the same direction, albeit with different accounts. Subjects with autism appear incapable of observing others from anything but an auto-enclosed ego-centric perspective, expected to be revealed by comparisons with a normal child when both are placed in a face-to-face situation. While normal children can do what the other is doing, when invited to do so, children with autism who understand the invitation, are expected to have problems in virtue of seeing the other only from an ego-centric position: Observing the model from the outside, the subject is seeing, but not feeling from the inside, what the model is doing. Seeing the resemblance between the inside of the model's hands with the insides of his own, and being incapable of a virtual reversal of the model's movements as felt, the child with autism can only do what is being seen from own position, and will raise hands with inside inwards.
This may reflect the more general problem in autism of relating to others. Engagement with others in complementary and reciprocal format appears to be impaired. They may echo and follow others -- subjects and objects -- in line, as it were, but the reciprocal and complementary engagement which Thomas (C3) demonstrates may be beyond their capacity.
Early learning from prototypical caregiving situations
When Thomas reciprocates the care-giving afforded by his sister and other previous care-givers, he give evidence of having learnt by what I term alter-centric participation. Such alter-centric participation is inferred to come about in virtue of the posited complementarity between the bodily self and the virtual companion perspective. When the actual other fills the companion space (in the place of the baby's virtual other), perceptual inversion and supra-modal binding of somaesthetic and visual qualities are afforded by the actual companion movements in virtue of being included in that space in felt immediacy. Through my theoretical spectacles, that is what I see Thomas to have been doing. The previous spoon-feeding situations which he has experienced, is of course much more complicated than situations involving gestural imitation. Quite a feat, actually, if we stop to consider that it reflects learning from a virtual mirror inversion of the caregiver's movements. In virtue of complementing the baby's bodily self (in the same asymmetric manner as the baby's virtual other is complementing the baby's bodily self) the actual other's movements are felt to be co-enacted by the baby (as if being hand-guided by the facing feeder).
Attribute to infants such innate sociocapacity for alter-centric participation in others in the mode of felt immediacy. If others in need or distress reactivate in the child feelings semblant of the form of bodily self-feelings evoked in situations in which the infant has experienced caregiving, then they invite caring efforts resembling the caring afforded by others earlier in infancy by virtue of having partaken in such caring in the reciprocal mode of felt immediacy.
Thus, for example in the feeding situation, the infant is not just a recipient of food from the care-giver, but feels to take a part in the food-giving in virtue of including the care-giver in the infant's companion space in felt immediacy. Alter-centric participation in the feeder's bodily movements leaves the infant with not just the experience of having been cared for, but of having taken part in the caregiving movements in a reciprocal and complementary manner. The caregiver included in the infant's companion space permits the infant not just the bodily self-experience of being cared for, but to engage in the caregiving in the reciprocal mode of felt immediacy. The body movements and position of the feeder included in the infant's companion space afford the feeling of virtually moving with the caregiver in an appropriating sense even though the face-to-face situation requires mirror-like reversion. More specifically, the proprioceptively felt reciprocal feeding on the part of the infant is in-formed by the alter-centric feelings of feeding movements afforded by the caregiver's feeding act.
Thus, the feeding situation in which the adult feeds the infant may be regarded as prototypical of caregiving situations more generally, for example, when being comforted or cared for when in need or distress. So then, when another in need or distress is included in the toddler's companion space, affected by such previous experiences in the reciprocal mode of felt immediacy, there is a phenomenological and experiential basis for re-enactment in an appropriating manner.
Whenever the caregiver has been included in the infant's companion space, the caregiving has afforded felt experience of co-enacting the caregiving. The infant's alter-centric participation complements the bodily self-feeling of being afforded care.
Thus, when peers in need or distress are included in the child's companion space and reactivate in the child feelings semblant of the form of bodily self-feelings evoked in situations in which the infant has experienced caregiving or comforting affording such learning by participation, then such peer situations should invite semblant feeling and efforts.
These are the propositions:

(1) Caregiving situations, which may appear to be unilateral activities, are re-defined to be reciprocal activities in virtue of which the infant takes part in what the caregiver does, and thereby learns from alter-centric participation in that very caregiving.

This fits with studies revealing how the quality of the caregiving background appear to play a role in children's reaction towards others in need: Those from a nurturant and caring background are most likely to help and offer comfort to other children in need or distress (Berk 1994; Zahn-Waxler et al. 1979). The second proposition, however, is more rich in its implied consequences in the form of potentially vicious circles:

(2) The kind of caretaking frequently experienced by the infant in virtue of alter-centric participation provides a basis for circular re-enactment of that kind of caretaking towards other children in need or distress.

Implication: circular re-enactment of abuse
Thus, sensitive caretaking frequently experienced by the infant in the reciprocal mode of felt immediacy provides a basis for circular re-enactment of semblant kinds of caretaking towards other children in need or distress. But caretaking experiences need of course not only be experiences of caring, comfort and holding (in Winnicott's sense). Parent, caretaker and others may provide various severe forms of abuse.
Thus, if the above applies, then we should also expect that experiences of abuse and rejection in caretaking may come to invite vicious circles of re-enactment. Prior to the onset of defence mechanisms, experience of indifference, rejection or abuse may come to be initially experienced in felt immediacy. In the case of abuse, if the perpetrator manages to be included in the victim's companion space, the victim may be compelled to feel to participate in the abusive movements. In virtue of alter-centric participation the victim may come to experience engagement in the bodily motions and feelings of the abuser, not just the suffering. That will leave the victim with a compelling bodily and emotional remembrance that increase the likelihood of circular re-enactment of abuse in peer relations or towards younger children later in ontogeny.
The above implies that children who have experienced caretaking or parenting in a harsh, punitive, neglecting or abusive manner should be more likely to respond with fear, anger, or even attack peers or younger children in distress, as compared to responses by children with a different experiential background. Empirical studies point in this direction. For example, George and Main (1985) have observed abused toddlers to abuse other infants. Severely abused toddlers have been observed at a day-care centre to react fearfully or aggressively towards other children in distress, and by the second year of their life to re-enact the abusive behaviour of their parents, while non-abused children in similar situations reacted with clear signs of concern (Berk 1994:405; Klimes-Dogan & Kistner 1990).
There is thus a potentially double tragedy of child victims of abuse. Not only are they deprived of full emotional holding quality in their own life. By virtue of circular re-enactment from e-motional memory of abuse, some of the victims may later in ontogeny be driven to deprive others of that same quality of life -- if other paths of intersubjective communion and transitional endeavours are not opened up. But fortunately, even vicious circles are sometimes transformed into self-other creative spirals transcending the tragic experiences.
We may thus speak of prosociality and abuse in the same vein: human nature and social nurture are at play, albeit with this radical difference: abuse nurtures abuse, while prosociality may emerge in spite of -- and even be evoked in -- the most horrible and insane circumstances. An example is Anna Freud's (1951) report on the six orphans rescued at about 3 years of age from Nazi extinction camps, and who helped each other and showed care in close companionship, for example, when eating: giving food to one's companions appeared to be more important than eating oneself.

Conclusion -- with a neurosociological prediction
In this talk the meal has indeed been used as quintessential of companionship and caregiving. I have offered an explanation of how infants may be able to give and reciprocate care in virtue of an innate capacity for participation in felt immediacy in caregivers and in other children, and contrasted such alter-centric participation by the ego-centric mode of perception, predictably exhibited in autism. From my postulate of an inborn companion space with a virtual other (Bråten 1988ab), this twofold reply has been inferred about the nature and nurture of prosociality:
First, as for the part played by nature, infants have the innate capacity to engage in others in the mode of felt immediacy -- whether it be in other children felt to be in need, or in the caregivers affording care, and, hence, do not only take a part in the receiving end of care-giving. Second, as for the part played by nurture, in virtue of such sociocapacity for participating in the care-giving movements and accompanying feelings, infants are afforded an experiential social learning ground for potential re-enactment -- when actual circumstances and motor resources permit.

Predicting a neural basis for alter-centric perception
The keynote here is what I term infant learning by alter-centric participation. And now for the prediction promised in the introduction. In a King's College workshop on perception of subjects and objects, John O'Keefe (1992) raised the issue of the relation between self-consciousness and allocentric maps, i.e. maps of the landscape in front of you independent of your position in relation to that landscape, distinct from seeing the landscape from an ego-centric perspective. For example, in a study of a monkey moved to different places in a spatial environment, O'Keefe found evidence of place cells, dependent upon the place where the monkey was, and different from view cells, defined primarily by the view of the environment, and not by the place where the monkey was (cf. Rolls 1995). Here is a ground for assuming that differentiated neural systems may be sensitized to allocentric aspects, different from aspects viewed from an ego-centric perspective.
Now, if by way of experimental procedures, the neural basis sensitized to ego-centric and allo-centric perception is uncovered in humans, then I predict that neural systems, perhaps even neurons, sensitized to alter-centric perception will be uncovered in experiments designed to test this prediction. Weaker, and less reductionistic, is this hypothesis: If neurons sensitized to ego-centric and allo-centric perception are found in humans, then I predict that capacity for alter-centric participation will be found to be realized by more comprehensive neural systems and pathways, modifying ego-centric inputs, reversed in face-to-face situations.5 Should any of these predictions be experimentally confirmed by neural scientists sometimes in the future, and without being divorced from the dynamic interpersonal companion systems level examined in this talk, then neurosociology would be born.
I thank you for your attention!

1 Centre for Advanced Study lecture in the Norwegian Academy of Science and Letters, Oslo, 4 March 1997. Figures have been omitted in this edition. For drawings and diagrams, see other essays in this volume, and chapter 5 in Bråten (ed.) 1998. Scandinavian readers may find pertinent photo illustrations and drawings in chapters 1,2 and 13 in Bråten (1998) and in Bråten (1999; Danish edition).
2 This question has rarely been posed, perhaps for these two reasons: First, it is somewhat at odds with folk psychology. Although parents and caretakers cannot have failed to notice that infants sometimes do exhibit care and concern, caregiving situations are usually considered a unilateral give-and-take situation: the adult is the giver and the infant is the recipient of care. Second, expectations of caregiving by infants conflicts with traditional theories of child development which viewed the infant as awaiting a long period of development before becoming social and, indeed, prosocial. Even so, there have been deviating voices in theory; in Object Relations Theory, for example, Winnicott pointed to the early capacity for concern, and in Attachment Theory, Bowlby spoke of the reversed caregiver relation that sometimes could occur in mother-infant dyads, albeit regarded as pathological.
3 Later, they probably acquire also an allo-centric perspective, becoming able to see things, for example, as localized in the corner of this room, irrespective of one's own position in this room. See O'Keefe (1985; 1992) for a discussion of the relations between self-consciousness and allocentric maps.
4 I am grateful to Mikael Heimann for providing me with these references during his stay here at the Centre for Advanced Study 1996-97. After my first talk in Paris, Julie Brown told me about her data for a doctoral thesis in preparation (Brown 1996), documenting that subjects with autism failed in grasp-a-thumb imitation. That was the first indication of potential support of my prediction.
5 By way of connectionist modelling I am presently in the process of trying to explore here at the Centre some fragments of such different architectures for perceptual inversion. Such 'neural net' simulations may not tell us much, however, -- except, perhaps, serving to spell out some of the implications.