John Benjamins Publishing Company

This is a contribution from Interaction Studies 6:3 © 2005. John Benjamins Publishing Company This electronic file may not be altered in any way. The author(s) of this article is/are permitted to use this PDF file to generate printed copies to be used by way of offprints, for their personal use only. Permission is granted by the publishers to post this file on a closed server which is accessible to members (students and staff) only of the author’s/s’ institute. For any other use of this material prior written permission should be obtained from the publishers or through the Copyright Clearance Center (for USA: www.copyright.com). Please contact [email protected] or consult our website: www.benjamins.com Tables of Contents, abstracts and guidelines are available at www.benjamins.com

Experiencing contingency and agency First step toward self-understanding in making a mind? Jacqueline Nadel1, Ken Prepin1 and Mako Okanda2 1UMR

CNRS 7593 / 2Kyoto University

Precursors of inferential capacities concerning self- and other- understanding may be found in the basic experience of social contingency and emotional sharing. The emergence of a sense of self- and other-agency receives special attention here, as a foundation for self-understanding. We propose that synchrony, an amodal parameter of contingent self-other relationships, should be especially involved in the development of a sense of agency. To explore this framework, we have manipulated synchrony in various ways, either by delaying mother’s response to infant’s behaviour, disorganizing mother’s internal synchrony between face and voice, freezing the partner in a still attitude, or on the contrary maximizing synchrony through imitation. We report results obtained with healthy and clinical populations that are supposed to be at the beginning of basic experiences concerning the ownership of their actions: infants of 2 months and 6 months, low-functioning children with autism and MA matched young children with Down Syndrome. Our results support the idea of a two-step process linking understanding of self to understanding of other and leading on to form the concept of human beings as universally contingent entities. Keywords: infant, imitation, agency, self-understanding

Numerous studies using classical conditioning have shown that neonates do associate and anticipate events (Gewirtz, 1969). A few hours after birth, they already discriminate regularities such as: “just after A (= a finger pressing my forehead left/right), comes B (= a drop of water on lips left/right side)”. They do not only turn their head in the appropriate direction, they also express distress when B no longer occurs (Blass et al., 1994). Current research on social

© 2005. John Benjamins Company Interaction StudiesPublishing 6:3 (2005), 447–462. issn 1572–0373 / e-issn 1572–0381 © John Benjamins Publishing Company All rights reserved

448 Jacqueline Nadel, Ken Prepin and Mako Okanda

perception shows that the early capacity to perceive regularities, and to anticipate B from A, rapidly includes social events. In this case, the anticipated event is not a physiological gain, but rather a psychological benefit: what is gained is not a drop of water, but a glance, a smile, a word. How can we explain such an early detection of social contingency? Gergely and Watson (1996; 1999) have proposed a theoretical framework that may provide a fruitful background for a complete exploration of infants’ subtle capacity to associate and anticipate social events that are related to their current behaviour. Starting from Watson’s demonstration that 2 month-olds not only can relate external events to their own behaviour, but also are keen with such events (Watson, 1972), Gergely and Watson (1996; 1999) postulate an innate module of contingency detection (DCM) that concerns social as well as physical events. DCM enables the infant to perceive causal relationships, to establish expectancies for contingency, thus acting as a determinant of social responsiveness. DCM is seen as composed of two independent mechanisms, one anticipating the probability of a future event as regard to present behaviour, and the second retrospectively searching for a link between a present event and past behaviour. Such properties of CDM are hypothesized to lead the infants to distinguish between the sensorial consequences of their motor behaviour, and external sources of perception. This clearly fits the important distinction proposed by Russell (1996) between external perception and perceptions that are at will, since they are caused by one’s own actions: from this distinction will emerge a sense of ownership of one’s own action, as opposed to and complemented by a sense of other-agency. Given that the unique actions producing perfectly synchronous perceptions are our own ones, it follows that the detection of synchrony may be considered as a primitive basis for an early distinction between self and external world, a first milestone in making a mind. In particular, Rochat (2002) has underlined that self-imitation is a primary source of knowledge about the self and a basic process by which infants gain self-reflective abilities (p. 86). Results by Watson (1985), Rochat and Morgan (1995) and Schmuckler (1996) support the hypothesis of a developmental switch from an initial attention bias in favour of perfect synchrony between action and perception to a later attraction toward imperfect synchrony. In other words, the preference for perception of own actions soon gives place to a preference for behaviours that reflect, though imperfectly, the infant’s behaviour. Computing temporal, spatial and intensity information, CDM takes account of the global degree of relationship between events. According to the prediction attached to the model, imitation of her

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 449

behaviour should be the most attractive social response among all those that can be offered to a young infant, and a tight temporal contingency accompanied by an attuned comment of the infant’s behaviour (Stern, 1985) will be preferred to simple temporal contingency. Looking back to twenty years of research of our group, it is now obvious to us that we have been turning around the question processed by Gergely and Watson’ model, without a clear awareness of such a convergence. In this paper we will stress the convergence and differences between the above mentioned theoretical model and our findings with young infants. As we will see, infants are able to detect a non-contingent behaviour of their mother earlier than predicted by Gergely and Watson’s model. They do so more efficiently if the history of their relationship with their mother is a contingently stable one. We will next suggest that the model may need further specification of what is synchrony for older infants and propose an additional dynamical perspective. We will then refer to the second part of Gergely and Watson’s model (1996; 1999), that emphasizes the role of parents’ mirroring of their infant’s action. We will show the relevance of this model to explain our findings with healthy young infants and low-functioning children with autism, and stress its convergence with a theoretical formalism designed by Gaussier and colleagues (Gaussier, Baccon, Prepin, Nadel & Hafemeister, 2004). Finally, we examine the relevance of the hypothesis of an impairment of CDM in autism. Watson (1994) and Gergely and Watson (1999) postulate that the shift of orientation from self-based perfect contingencies to environment-based contingencies occurring around two- to -three months in typically developing infants, does not take place strongly enough to supplant the search for perfect contingency. According to the authors this may explain stereotypies, that can be seen as a kind of self-imitation, and social avoidance. We will see that even low-functioning children with autism are highly sensitive to almost perfect contingency in imitative responses.

I.

Early detection of non-contingent responses of mother

One of the first studies demonstrating the early sensitivity to social non-contingency was conducted by Tronick, Als, Adamson, Wise and Brazelton (1978). This study showed that 3-month-olds display negative reactions and finally disengage when their mothers pause with a still face, like infants of depressed mothers tend to do. Gusella, Muir and Tronick (1988) replicated the results at

© 2005. John Benjamins Publishing Company All rights reserved

450 Jacqueline Nadel, Ken Prepin and Mako Okanda

the same age via a televised face to face situation. However, several concerns were raised about the all-or-none nature of the still face paradigm. To test more convincingly infants’ sensitivity to social non-contingency, Murray and Trevarthen (1985) organized a TV face-to-face interaction between mothers and 6-to-12 week-olds. After a pause, they replayed the episode which therefore was no more contingent with the current behaviour of the infant. Comparing the live and replayed episode, they found a strong negative effect of the replay on the infant’s behavioural state, and concluded that young infants expect contingency and detect non-contingent behaviour. Rochat, Neisser and Marian (1998) did not replicate the findings and concluded that Murray and Trevarthen’s results could be explained by a natural decline of interest for the TV interaction. In a series of experiments with an updated design and procedure, we have tested the capacity of 2 month-old infants to detect and expect contingency via experimental violations of social contingency. In an initial experiment with 10 infants aged 2 months, (Nadel, Carchon, Kervella et al., 1999), we have designed a live1-replay-live2 procedure, instead of the live-replay procedure used by Murray and Trevarthen (1985). We have used a double teleprompter device that allowed us to offer to mothers and infants a continuous image and voice of their partner and to present to the infants alternately 30 seconds of live and replay episodes of their mother’s communication with a seamless shift. Results replicated Murray and Trevarthen’s previous ones. In addition, they showed a recovery of infant’s positive state during mother’s second contingent communication for the seven infants who did not cry during replay. A recent experiment with fourteen 2-month-old infants of healthy mothers gave similar results (Nadel, Soussignan, Canet, Libert & Gérardin, in press). When the 14 infants of healthy mothers were compared with 14 same-age infants of depressed mothers, results showed a linear decrease of smile in infants of depressed mothers contrasting with a curvilinear curve in infants of healthy mothers. They also demonstrate lower negative reactions to non-contingent maternal behaviour. These findings suggest the importance of a stable contingent relationship with the mother for the development of expectancies for contingency and extended use of the DCM. A third experiment was conducted with 50 infants aged 2 months, that were randomly assigned to two conditions (Soussignan, Nadel, Canet & Gérardin, submitted): One group was presented the live-replay-live condition and the other was presented a continuous live interaction (live1-live2-live3) with the mother. Results demonstrate that the second episode of 30 seconds is not processed similarly in the case of live

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 45

interaction and in case of non-contingent interaction: only non-contingent interaction induces a change of infants’ affect. Taken together, the findings of these experiments clearly establish that infants as young as 2 months detect and expect social contingency.

II. Six-month-olds’ processing of partially contingent responses Gergely and Watson’s model of contingency predicts young infants’ preference for high level of synchrony compared to lower levels. This prediction suggests that visual plus auditory contingent responses will be more attractive than a synchronic response in one modality only. This is what research demonstrates (Walker-Andrews, 1997; see Muir & Nadel, 1998, for a review). However, no prediction can be derived from the CDM model with regard to how young infants process the information coming from two sensory modalities, one contingent to her behaviour and the other non-contingent. Does the infant react differentially (and thus distinguish) when two distinct sources produce the two sensory messages and when there is only one source? Does she expect two modalities coming from the same source to be intrinsically synchronic? One of us is currently developing a general model of coupling between dynamic systems exchanging energy (Prepin, 2003). Agents engaged in a TV face-to-face interaction can be considered as dynamic systems exchanging energy. Energy can be exchanged using two different flows through two different channels (see Figure 1): the visual channel where facial and bodily motor outputs of one system are visual inputs for the other(s), and the auditory channel, where vocal and verbal outputs of one system are auditory inputs for the other(s). The two flows of energy produced by a dynamic system can be related according to various combinations. Generally when different flows of energy co-occur, they are produced by a unique source, but they may co-occur although produced independently by two different systems. Notice that the receptor can process and combine energy flows in a number of ways, independently or not from their real state of connection: the receptor can process each flow separately, select one flow and neglect the other, sum up the two flows, consider the two flows as competing, or extract their common properties, thus capturing their intrinsic coherence. This way of processing depends on the receptor’s ability to determine if the energy flows it receives come from one source or two independent sources.

© 2005. John Benjamins Publishing Company All rights reserved

452 Jacqueline Nadel, Ken Prepin and Mako Okanda

Our model predicts that the receptor will distinguish co-occurring flows of information from co-varying ones if it detects shared a-modal parameters, such as speed, rhythm, periodicity, and intensity of flows that reflect their inherent tight connections. When common properties are extracted from two flows of energy, the guess is that the two flows are expected to come from the same system. Conversely, two flows coming from the same system are expected to be tightly connected, not only co-occurring but also coherently responding. Is the young infant able to extract these common properties so as to distinguish between incoherent co-occurring information that comes from two independent sources and incoherent co-occurring information that comes from a unique source? To explore this question and evaluate the role of multimodal synchrony in contingency detection, our experimental design was updated so as to disconnect visual and auditory modalities. We can at will present to the infant a coherently contingent mother, a mother whose response to the infant is contingent for one modality (voice contingent, for instance) but non- contingent for the other (face replayed, for instance) or two modalities coming from two distinct sources (contingent mother’s voice coupled with the non-contingent pre-recorded face of a stranger). To test this model, two TV experiments were led with 6 month-old infants. In the first experiment (Prepin, Simon, Canet, Mahé, Soussignan & Nadel, submitted), 19 infants were presented three 30-second uninterrupted episodes of maternal interaction in the following order: mother’s contingent face and voice (live 1), mother’s contingent voice with non-contingent (replayed) face, and Figure . A model of coupling between dynamic systems Given a system Syst.1 which receives simultaneously information from different channels, how will this system process the two flows? We present here a schema of the situations that Syst.1 can encounter: Syst.1, Syst.2 and Syst.3 are dynamic systems exchanging energy (as shown by arrows) through different channels: the visual channel where facial and bodily motor outputs of one system are visual inputs for the other(s), and the auditory channel, where vocal and verbal outputs of one system are auditory inputs for the other(s). Those two channels are symbolised by red and blue colours. Syst.1 receive flows of information from both channels. Two conditions can be distinguished: a) The two information flows come from one source (Syst.2), which receives Syst.1’s outputs and is thus contingent with Syst.1 (i.e. the Live conditions in our experiments). b) The two information flows can come from two independent sources: Syst.2 contingent with Syst.1, and Syst.3 which does not receive any ouputs from Syst.1 or Syst.2 and which is thus non-contingent with Syst.1 (i.e. the conditions Contingent voice of the mother presented with non-contingent face of mother or of stranger).

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 453

again mother’s contingent face and voice (live2). The three episodes were presented with a seamless shift. The presence/absence of Gaze to the screen, Smile, Grimace and self-centred movements were coded each 40/100th of a second for the three episodes. In the second experiment, 10 infants were presented three 30-second uninterrupted episodes of maternal interaction in the following order: mother’s contingent face and voice (live 1), mother’s contingent voice coupled with the non-contingent (pre-recorded) face of another mother responding to her infant, and again mother’s contingent face and voice (live 2). Comparing the results of the two experiments for the four indices, we found no significant difference during the first live interaction. The infants’ response to the perturbation episode (episode 2) however, was significantly

�� ������� ������� ��������������������������� ������������������������

�� ������ �������

������� Figure . The visual-auditory coupling between dynamic systems.

© 2005. John Benjamins Publishing Company All rights reserved

454 Jacqueline Nadel, Ken Prepin and Mako Okanda

different for gaze: while infants withdraw from the image of the dysfunctioning mother, they maintained gaze at the screen in experiment 2, where the mother’s voice was coupled with a non-contingent stranger’s face [t (27) = 2.13, p < .04]. Infants showed also a significantly higher amount of self-centred movements in experiment 1 [t (27) = 2.07, p < .05] and a marginally significant higher level of grimacing for experiment 1. Taken together, these findings are in agreement with the hypothesis that 6month-old infants have formed the concept of mother as an intermodal entity whose sensorial outputs should cohere in a contingent bimodal response to the infant’s behaviour, and clearly distinguish a dysfunctioning intermodal entity from co-occurring sensory messages originating from two different sources. To the CDM model, we propose to add the early distinction between contingency detection and detection of co-occurring messages that are not necessarily coherent and may be partially contingent only.

III. Social biofeedback and the effect of being imitated in young children and low-functioning children with autism According to Gergely and Watson’s model (1996; 1999), parents’ mirroring of their infants behaviour acts as a social biofeedback. Considering that the internal changes of emotional states are beyond perceptual awareness, the model proposes that those states cannot be understood as such without an external feedback of what is felt. In other words, parents’ mirroring of infants’ facial expressions will play a role similar to the effect of a biofeedback documenting upon what is related to our internal experience: getting a perceptual knowledge of our blood pressure allows us to relate various internal states to various levels of blood pressure. Here the biofeedback is social and is provided through imitation. Recently, a theoretical formalism designed by Gaussier and colleagues in an epigenetic robotic perspective (Gaussier, Baccon, Prepin, Nadel & Hafemeister, 2004) resulted in similar proposals. At the start, the conceived architecture can couple perception and action and is driven by a homeostatic principle according to which perception and action should be in accordance. It follows that any perception that perfectly fits the kinesthetic feedback of action reinforces the link between perception and action, and leads to a repetitive behaviour. Reversely, any deviation between perception and action is registered as a signal of error which leads the architecture to act next so as to correct the error via a

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 455

mirroring of the perceived action: the architecture will thus act according to a process similar to that of social biofeedback. This formalism can describe very simply the distinction between perfect synchrony resulting in self-imitation (Rochat, 2002) and imitation of another’s action. It can also describe the process by which facial expressions of internal states that are felt, but not seen, can gain an intermodal representation via a re-enactment of what is externally perceived and resemblance to what was felt: this clearly deals with the principle of social biofeedback. To illustrate the relevance of the formalism designed by Gaussier and colleagues (2004) and of the social biofeedback model, we will report previous findings with 2-year-old infants and recent research with 2-month-olds.

Young infants prefer imitative contingency A preference for imitative contingency compared to other contingent responses is already to be found very early in development. Three month-olds react more to an imitative mother (Field, Guy & Umbel, 1985). Two-month-olds imitate more imitative than non-imitative mothers (Nadel, Revel, Andry & Gaussier, 2004). Infants gaze and smile more to an imitative adult than to an adult who is timely contingent only (Meltzoff, 1990), and this finding was replicated with 9 to 18 month-old infants (Agnetta & Rochat, 2004). In ecologically valid settings, we have demonstrated the preference of 2year-olds for a mirroring of their behaviour, compared to other contingent responses of peer-age partners (Nadel, 1986; Nadel & Fontaine, 1989). The same dyads of 30 month-olds were presented with two different settings with the same partner within 24 hours. Once they met in a setting that afforded synchronic imitation insofar as it was furnished with two identical sets of 10 objects, and another time they were presented with 20 different objects. When we compared the social indices in the two conditions, we found that attention to partner was significantly higher in the setting with two identical sets of objects, that imitation was the main social behaviour in this setting (70% of total time), and that verbalizations were equally rare in the two settings. Initiations of contact were more frequently successful (i.e. answered) when they took place during imitative actions; social signals such as smiles, and non verbal behaviours such as offering, occurred more frequently during the imitative sequences. There were seven times more laughs in the double setting than in the setting with single objects, although the objects proposed were equally attractive in the two settings. Another interesting finding confirmed the preference of young children for a combination of temporal, spatial and intensity signals

© 2005. John Benjamins Publishing Company All rights reserved

456 Jacqueline Nadel, Ken Prepin and Mako Okanda

of contingency. Children in the double setting monitored their temporal synchrony during the imitative sequence, the model waiting for the imitator, and the imitator rushing to imitate on time (Nadel-Brulfert & Baudonnière, 1982). As a consequence, actions with identical objects were in tight temporal connection, thus producing a high level of contingency between partners.

Verbal children switch to a preference for low contingency A few months later, when children master verbal language, the preference for synchronic among other non-verbal means of interaction declines abruptly in dyads of 42–46 months meeting in similar conditions than 30 month-olds (Nadel, 1986; Nadel & Fontaine, 1989). After 4 years, hostility grows toward synchronic imitation now considered as mockery. This hostility is still present in adulthood and was expressed fully in the Platonic tradition, when synchronic imitation was considered as holding danger for individual identity and for self-consciousness. A way to complete the CDM model is to take account of this later switch toward low contingency and propose a three-step model of contingency.

IV. Autism: An impairment of the Contingency Detection Module? Low-functioning children with autism may be so deeply involved in repetitive behaviour that they do not even notice their social environment. Except for these cases of extreme focus on self-produced behaviour, the majority of non verbal children with autism are attracted toward highly contingent responses to their behaviour. They show various levels of detection of being imitated, that parallel the levels described elsewhere for typical development (Nadel, 2002). A high level will consist in testing the intentionality of the imitator through strategies like changing the rhythm of activity, changing the object used, or stopping action, while gazing at the experimenter. Those who do not test the experimenter but reciprocate imitation, smile, approach the experimenter, or gaze fixedly at her, show an awareness of being imitated but probably do not understand the intentionality of the imitator. However, this basic response to being imitated is a first step toward an access to the attribution of intentionality to the imitator. This was shown in a collaborative study with Field and colleagues (Field, Field, Sanders & Nadel, 2002). In this study, repeated sessions of being imitated during one week proved to significantly generate awareness

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 457

of being imitated, increase imitation, and enhance close proximity with the imitator. These findings were recently replicated by Heimann and colleagues who proposed three sessions of imitative interactions in a day instead of in a week (Heimann, Laberg, & Nordoen, in press). The burst of social awareness that these findings suggest is to be linked to the discovery of self-agency during imitative sessions. While acting, children discover the external consequence of their action mirrored by another person (a clear example of social biofeedback). They perceive not only that they are the author of their action but also that they are at the origin of the other’s action, that they have a power over another person. Some of them will perceive next that this power is restricted to the other’s will, since the person is an agent. This scenario was the bet of a pilot study led with the still face paradigm that we revisited for this purpose (Nadel, Croué, Mattlinger et al., 2000). The Still Face paradigm pioneered by Tronick and colleagues (1978 ) has generated a large amount of robust data. Interestingly, it appears to also be a powerful tool to test other questions such as: When do children start forming a general concept of persons as contingent agents intending to interact? Do lowfunctioning children with autism form such a general concept? Do high levels of contingent behaviour favor the building of such concept? To address these questions, we substituted the ‘interaction-still face-interaction’ procedure with a ‘still face-interaction-still face’ procedure. And, instead of a familiar partner, we “froze” a stranger. If the children show negative reactions in front of a still adult that they have never met before, we will conclude that they expected him/ her to share with them, thus that they had formed the concept of a person as a contingent agent. If they show concern only during the second still face, after a highly contingent interaction with the adult, we will conclude that they need prior imitative experience with a person, before they will form expectancies about her social behaviour and willingness to share. Results showed a significant difference between behaviour displayed by the children during Still face 1 and Still face 2. During Still Face 1, they explored the new arrangement of their sport room and looked at or manipulated the objects displayed, all in two exemplars. During Still Face 2, after three minutes of having been imitated by the stranger, they appeared to have a unique focus: have the stranger be socially responsive. Their disappointment, their embarrassment, their surprise, their attempts to attract the stranger’s attention show that they understand the still behaviour as being at will. They now react to the still face like people typically react to ostracism (see Williams and Gerber, 2005; this volume): it is an insult to their being there.

© 2005. John Benjamins Publishing Company All rights reserved

458 Jacqueline Nadel, Ken Prepin and Mako Okanda

Infants with Down Syndrome matched on developmental age do not need a previous interaction with the still stranger to react as if ostracized: they immediately withdraw, take a distance and look wary. More precisely, when comparing the two clinical matched groups during still face 1, we found that children with Down Syndrome presented significantly more negative facial expressions (U = 5; p < .001), complained more (U = 28; p < .05), and had a lower amount of object manipulation (U = 27; p < .05.) than children with autism. During the second still face however, the only difference between groups concerned the higher amount of gaze to stranger in children with Down Syndrome (U = 14,5; p < .01), a result in conformity with established criterion of social gaze avoidance in children with autism. The equivalence of other social indices in the two groups document a clear-cut change in the social behaviour of children with autism after the imitative session. This change in children with autism reveals the importance of being imitated in the understanding of other’s agency. In another ‘still face’ experiment, we have tested the prediction that higher contingent behaviours are more efficient to awaken social awareness in children with autism. In this experiment, two matched groups of children with autism were presented with two different interactive conditions: a non imitative interaction versus an imitative interaction. Imitative interaction was proved to be a more powerful way of initiating positive emotional behaviours towards the stranger during the second still face than non imitative contingent interaction (Escalona, Field, Nadel & Lundy, 2002).

Concluding comments How to make a mind is certainly a question that cannot be answered without the help of a developmental model of perception-action coupling (Prinz, 1990). Data concerning the early role of action on perception and of perception on action show that the precedence of one on the other is far from being demonstrated (Hauf, Elsner & Aschersleben, 2004; Hauf & Prinz, 2005; this volume; Sommerville & Woodward, 2005). All data however converge to underline the developmental role of action. A basic coupling of perception and action concerns imitation. Synchronic imitation of others and its reverse facet, sensitivity to being imitated, both contribute to exert the distinction between perceptions caused by self and perceptions caused by the external world, in a context of high contingency. Russell (1990) has highlighted the basic role of this distinction in understanding intentionality. The model of a Contingency

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 459

Detection Module proposed by Gergely and Watson (1996; 1999), as well as the perception-action-based formalism developed by Gaussier and colleagues (2004) provide inspiring frameworks for a systematic approach to studying the role of imitation in the development of agency and self-understanding in young infants and non verbal children with autism. This paper was aimed at providing developmental and psychopathological data that support and complete these frameworks, in an attempt to approach the basic conditions that make an imitative infant a developing mind.

Acknowledgements More recent research presented in this paper was funded by ADAPT, UE contract, Vth Framework, IST- 2001–37173.

References Agnetta, B., & Rochat, P. (2004). Imitative games by 9-, 14-, and 18-month-old infants. Infancy, 6, 1, 1–36. Blass, E.M., Ganchrow, J.R., & Steiner, J.E. (1984). Classical conditioning in newborn humans 2- 48 hours of age. Infant Behavior and Development, 7, 223–35. Escalona, A., Field, T., Nadel, J., & Lundy, B. (2002). Imitation effects on children with autism. Journal of Autism and Developmental Disorders, 32, 2, 141–144. Field, T., Field, T., Sanders, C., & Nadel, J. (2001). Children display more social behaviors after repeated imitation sessions; Autism, 5, 3, 317–323. Field, T., Guy, L., & Umbel, V. (1985). Infants’ responses to mothers’ imitative behaviors. Infant Mental Health Journal, 6, 40–44. Gaussier, p., Baccon, J.C., Prepin, K., Nadel, J., & Hafemeister, L. (2004). Formalization of recognition, affordances and learning in isolated or interacting animats. The Society for Adaptive Behavior SAB’04: MIT Press. Gergely, G., & Watson, J. (1996). The social biofeedback model of parental affect-mirroring. International Journal of Psycho-Analysis, 77, 1181, 1212. Gergely, G., & Watson, J. (1999). Early social-emotional development: Contingency perception and the social biofeedback model. In P. Rochat (Ed.), Early social cognition (pp. 101–136). Hillsdale, NJ: Erlbaum. Gewirtz, J. (1969). Mechanisms of social learning: some roles of stimulation and behaviour in early development. In D. Goslin (Ed.), Handbook of socialization theory and research (pp. 57–212). Chicago: Rand McNally. Heimann, M., Laberg, K., & Nordoen, B. (in press). Imitative interaction increases social interest and elicited imitation in nonverbal children with autism. Infant and Child Development.

© 2005. John Benjamins Publishing Company All rights reserved

460 Jacqueline Nadel, Ken Prepin and Mako Okanda

Hauf, P., & Prinz, W. (2005). The understanding of own and others’ actions during infancy. Interaction Studies: Social and Behaviour Communication in Biological and Artificial Systems, 6:3, same issue. Hauf, P., Elsner, B., & Aschersleben, G., (2004). The role of action effects in infants’ action control. Psychological Research, 6, 115–125. Meltzoff, A. (1990). Foundations for developing a concept of self: the role of imitation in relating self to other and the value of social mirroring, social modeling, and self practice in infancy. In D. Cicchetti & M. Beeghly (Eds.), The self in transition: Infancy to childhood (pp. 139–164). Chicago: University of Chicago Press. Meltzoff, A., & Gopnik, A. (1993). The role of imitation in understanding persons and developing a theory of mind. In S. Baron-Cohen, H. Flusberg & D. Cohen (Eds.), Understanding other minds (pp. 335–366). Oxford: Oxford University Press. Muir, D., & Nadel, J. (1998). Infant social perception. In A. Slater (Ed.), Perceptual development (pp.247–285). Hove, UK: Psychology Press Ltd. Murray, L., & Trevarthen, C. (1985). Emotional regulation of interaction between twomonth-olds and their mothers. In T.M. Field & N.A. Fox (Eds.), Social perception in infants (pp. 101–125). Norwood, NJ: Ablex. Nadel, J. (1986). Imitation et communication entre jeunes enfants. Paris: Presses Universitaires de France. Nadel-Brulfert, J., & Baudonnière, P. M. (1982). The social function of reciprocal imitation in 2-year-old peers. International Journal of Behavioral Development, 5, 95–109. Nadel, J., Carchon, I., Kervella, C., Marcelli, D., & Réserbat-Plantey, D. (1999). Expectancies for social contingency in 2-month-olds. Developmental Science, 2, 2, 164–173. Nadel, J., Croué, S., Mattlinger, M.-J., Canet, P., Hudelot, C., Lecuyer, C., & Martini, M. (2000). Do autistic children have expectancies about the social behaviour of unfamiliar people? A pilot study with the still face paradigm. Autism, 2, 133–145. Nadel, J., & Fontaine, A.M. (1989). Communicating by imitation: a developmental and comparative approach to transitory social competence. In B. Schneider et al. (Eds.), Social competence in developmental perspective (pp.131–144). Dordrecht: Kluwer. Nadel, J., Guérini, C., Pezé, A., & Rivet, C. (1999). The evolving nature of imitation as a format for communication. In J. Nadel & G. Butterworth (Eds.), Imitation in infancy (pp. 209–234). Cambridge: Cambridge University Press. Nadel, J., & Muir, D. (Eds.) (2005). Emotional Development. Oxford, NY: Oxford University Press. Nadel, J., Revel, A., Andry, P., & Gaussier, P. (2004). Toward communication: first imitations in infants, low-functioning children with autism and robots. Interaction Studies: Social Behaviour and Communication in Biological and Artificial Systems, 5, 1, 45–74. Nadel, J., Soussignan, R., Canet, P., Libert, G., & Gérardin, P. (2005). Two –month-old infants of depressed mothers show mild, delayed and persistent change in emotional state after non-contingent interaction. Infant Behavior and Development, 4 (in press). Prepin, K. (2003). Development of tools for the study and modelling of imitation games. Postmaster report, University of Cergy, 40 p. Prepin, K., Simon, M., Canet, P., Mahé, A.S., Soussignan, R., & Nadel, J. (submitted). The effect of a mismatch between mother’s voice and face on 6-month-old-infants interaction.

© 2005. John Benjamins Publishing Company All rights reserved

Experiencing contingency and agency 46

Prinz, W. (1990). A common coding approach to perception and action. In O. Neumann & W. Prinz (Eds.), Relationships between perception and action (pp.167–201). Berlin: Springer Verlag. Rochat, P. (2002). Ego function of early imitation. In A. Meltzoff & W. Prinz (Eds.), The imitative mind (pp. 85–97). Cambridge, MA: Cambridge University Press. Rochat, P., & Morgan, R. (1995). Spatial determinants in the perception of self-produced leg-movements in 3- to 5- month-old infants. Developmental Psychology, 31, 626–636. Rochat, P., Neisser, U., & Marian, V. (1998). Are young infants sensitive to interpersonal contingency? Infant Behavior and Development, 21 (2), 355–366. Russell, J. (1996). Agency. Hove: Erlbaum (UK) Taylor and Francis Ltd. Schmuckler, M.A. (1996). Visual-proprioceptive intermodal perception in infancy. Infant Behavior and Development, 19, 221–232. Sommerville, J., & Woodward, A. (2005). Pulling out the intentional structure of action: the relation between action processing and action production in infancy. Cognition, 95, 1, 1–30. Soussignan, R., Nadel, J., Canet, P., & Gérardin, P. (submitted). Sensitivity, not tiredness, accounts for 2-month-olds emotional changes during non-contingent maternal behavior. Stern, D. (1985). The interpersonal world of the infant. New York: BasicBooks. Tronick, E., Als, H., Adamson, L., Wise, S., & Brazelton, T. (1978). The infant’s response to entrapment between contradictory messages in face-to-face interaction. Journal of the American Academy of Child Psychiatry, 17, 1–13. Walker-Andrews, A. (1997). Infants’ perception of expressive behaviors: differentiation of multimodal information. Psychological Bulletin, 121, 437–456. Watson, J. (1972). Smiling, cooing and the “game”. Merrill-Palmer Quarterly, 18, 323–339. Watson, J. (1985). Contingency perception in early social development. In T. Field & N. Fox (Eds.), Social perception in infants (pp. 157–176 ). Norwood, NJ: Ablex. Watson, J. (1994). Detection of self: the perfect algorithm. In S. Parker, R. Mitchell & M. Boccia (Eds.), Self-awareness in animals and humans: Developmental perspectives (p.131–148). Cambridge, MA: Cambridge University Press. Williams, K., & Gerber, J. (2005). The making of the ignored and excluded mind. Interaction Studies: Social and Behaviour Communication in Biological and Artificial Systems, 6:3, same issue.

Authors’ addresses Jacqueline Nadel CNRS Unit UMR7593 Pavillon Clérambault, Hôpital de La Salpêtrière 47, Bd de l’Hôpital F-75013 Paris France E-mail: [email protected] Web: www.ccr.jussieu.fr/nadel/overview.htm

© 2005. John Benjamins Publishing Company All rights reserved

462 Jacqueline Nadel, Ken Prepin and Mako Okanda Ken Prepin CNRS 8051 & CNRS 7593 Pavillon Clérambault, Hôpital de La Salpêtrière 47, Bd de l’Hôpital F-75013 Paris France E-mail: [email protected] Mako Okanda Kyoto University Department of Psychology Yoshida-Honmachi, Sakyo-ku Kyoto 606–8501 Japan E-mail: [email protected]

About the authors Jacqueline Nadel is a research director at the French National Centre of Scientific Research. She leads the group “Development and Psychopathology”, with a focus on the origin of social cognition without language. Her areas of expertise concern the development of imitation, emotion and precursors of intentionality in infants and children with autism. She is especially involved in interdisciplinary programs interfacing epigenetic robotics and human development. Ken Prepin received a Master degree of mathematics and a M.S. degree in Cognitive Science from the Paris- 6 University. He is currently running a PhD program in an interdisciplinary perspective, modelling social interaction via imitation between humans and robots. Mako Okanda received a Master degree in psychology in 2004, and is now a PhD student at the University of Kyoto, studying mother–infant interaction with Dr. Shoji Itakura. She has been trained to use the double video design and related coding system in Dr. Nadel’s lab.

© 2005. John Benjamins Publishing Company All rights reserved