香京julia种子在线播放

Most published studies on imitation during the first months of life have focused on typically developing children. It might actually be all studies for imitation of facial gestures during the neonatal period. Researchers seem to avoid including children with any known risk factors for non-optimal development. It follows that reports often state that the children included in studies on neonatal imitation were born full-term and did not have any known medical complications. However, if imitation in the neonatal period is a real phenomenon affecting early social interaction and development, it becomes of uttermost importance to also investigate if children who show developmental deviances, either specifically in the social domain or more generally, differ from typically developing children in this ability. Here we present: (i) a comprehensive literature search focusing on published reports on imitation during the first year of life in children with an Autism Spectrum Disorder (ASD) or Down syndrome (DS); (ii) a home video observation that in our view might be interpreted as showing imitation during the neonatal period in a child later diagnosed with autism; and, (iii) a re-representation of previously published data on near-neonatal imitation in children with Down syndrome, observations that hitherto has gone largely unnoticed by the scientific community (Heimann et al., 1998).

The observation that infants might imitate already as newborns has been with us for a long time, and this ability has been reported by numerous studies since the 1970's (e.g., Meltzoff and Moore, 1977, 1983, 1989; Dunkeld, 1979; Maratos, 1982; Heimann and Schaller, 1985; Heimann et al., 1989; Kugiumutzakis, 1998; Nagy et al., 2013; for a recent meta-analysis, see Davis et al., 2021). Importantly, the existence of this phenomenon has been argued to reflect a rudimentary capacity for primary or innate intersubjectivity (i.e., Trevarthen, 1979, 2011a), that is, an ability to engage in intentional and purposeful exchanges with another human being (e.g., Trevarthen and Aitken, 2003; Rochat and Passos-Ferreira, 2009; Trevarthen, 2011a,b). Although we acknowledge the current debate around how to understand a newborn child's imitative-like responses (e.g., Oostenbroek et al., 2016; Jones, 2017), this paper rests on the assumption that those responses are best described as an act of neonatal imitation (e.g., Simpson et al., 2014; Meltzoff et al., 2018; Heimann and Tjus, 2019). In line with this, we also assume that this early act of imitation reflects a potential expression of primary intersubjectivity and that we need a better understanding of its place in development for children following both typical and atypical developmental pathways. Our hypothesis that newborn imitation is an example of an early social motive that signifies an intersubjective capacity is furthermore anchored in the works by several different theoreticians over the years (e.g., Bråten, 1998; Reddy, 2008; Rochat and Passos-Ferreira, 2009), but foremost on Trevarthen's groundbreaking ideas as exemplified in these two quotes:

If neonatal imitation is one of the first signs of innate or primary intersubjectivity as proposed by Trevarthen (1979, 2011b), Bråten (1998), Reddy (2008), and Kugiumutzakis and Trevarthen (2015), then the question also arises to what degree children following an atypical developmental trajectory would show an early imitative ability. In other words: Is this capacity of the neonate associated primarily with typical development, or is it an ability that also can be observed among infants, following an atypical developmental trajectory?

For autism, imitation has often been highlighted as one of the capacities that develops slowly and possibly represents a core deficit (Sigman et al., 2004; Volkmar et al., 2005; Nadel, 2006, 2014; Rogers, 2006; Vanvuchelen et al., 2011; Gowen, 2012; Vivanti and Hamilton, 2014). Imitation has been viewed as important for children with ASD because it “supports a sense of interpersonal connectedness and mutuality” (Sigman et al., 2004; p. 224), capacities that people with autism often find difficult. This aspect plus the fact that imitation is an important tool for learning through observation have made imitation training an important part of various training and intervention programs for children with autism (see Schreibman et al., 2015; Spjut Jansson et al., 2020). As one central example, Rogers and Pennington (1991) included imitation as one of the early deficits in their theory on autism. According to them, neonatal imitation is an early social competence that would be missing in newborn children that later develop autism. In a subsequent theoretical attempt, Heimann (1998) outlined two possible developmental routes for children with autism. Building on Bråten's (1988, 1998) theoretical formulations that, from the beginning, the mind is both dialogical and intersubjective, two hypothetical models of development were formulated (see Figure 1).

Heimann (1998) reasoned that neonatal imitation, as a marker of primary intersubjectivity, may or may not be observable in infants who later show signs of autism or other developmental disabilities. In one scenario, neonatal imitation is linked to more advanced imitation and intersubjective abilities at later stages in development, which leads to the expectation that neonatal imitation should be impaired in infants with ASD. Thus, a diminished imitation ability in a neonate might be a sign of an unfulfilled capacity to partake in early intersubjective social encounters. A second scenario noted by Heimann (1998) is that neonatal imitation might not have a direct relationship to later imitative ability and intersubjective development. In this scenario, neonatal imitation might be observable in infants with autism or other developmental disorders that are associated with impairments in the social domain. This could mean that mechanisms for primary intersubjectivity are not impaired from birth in these populations and, instead, that atypicality in relating to another person emerges later as other layers of intersubjectivity develop.

One problem in determining whether a lack of the ability for neonatal imitation is an early marker of autism is that the condition is not usually diagnosed before the child is several years old and rarely before 2 years of age (Ozonoff et al., 2015; Goldstein and Ozonoff, 2018; Zwaigenbaum et al., 2020). Even if many parents report retrospectively that they did note problems during the first year of life, it has not been possible to pinpoint an exact cluster of behaviors that makes it possible to reach a definite diagnosis early in life (but see Wetherby et al., 2021 for an early identification protocol). An additional problem is that some children with autism show a typical developmental trajectory from birth to about 12 or 18 months of age whereafter they start to lose abilities (Ozonoff et al., 2008). An example of such a regressive pattern might be a child who, after being able to point and utter his or her first words, suddenly stops both pointing and talking.

Davis and Crompton (2021) highlight the growing insight that the social difficulties associated with autism or other neurodiverse conditions “are at least in part bidirectional” (p. 652). They also argue for researchers to use a difference perspective, in contrast to the more traditional deficit model. This leads to the need for a research framework that charters socio-cognitive abilities in detail and that avoids preconceived expectations of what to expect or not to expect from autistic people or other neurodiverse groups. Within the scope of this paper, this means that we must acquire a better knowledge of the competencies of, for example, infants with the risk of developing autism. If we ever will be able to understand how the different social and communicative abilities of autistic persons evolve, we must differentiate between problems residing within the individual from problems arising from “a mismatch of interpersonal dynamics” (Davis and Crompton, 2021; p 650), as proposed by the dialectical mismatch hypotheses (Bolis et al., 2017).

For children with DS, early development differs from most children with autism (although the two syndromes can also overlap). DS is a chromosomal aberration usually diagnosed at birth or shortly thereafter and almost all children with this syndrome end up having a mild to moderate intellectual disability (Udwin and Dennis, 1995; Di Nuovo and Buono, 2011; Ostermaier, 2019). Since the diagnosis is made early, one would have expected that some studies on imitation at birth or shortly thereafter in this group would have been conducted. But this seems not to be the case. One of the few comprehensive and longitudinal studies of the early psychological development focusing on children with DS is the study by Dunst (1990) that describes sensorimotor development over the first 3 years of life.

The mean age of the nine children constituting the youngest group in Dunst's sample was 2.9 months, and they displayed an almost typical level of imitation according to the Uzgiris-Hunt scale (Uzgiris and Hunt, 1975). However, imitation of facial gestures like tongue protrusion or mouth opening used in studies of imitation in newborn children were not included. Dunst used Piaget's theory when chartering the early development of children with DS and concluded that this group follows a similar developmental trajectory as typical children, although at a slower rate. The development of the youngest group, children younger than 4 months, was almost on par with typical infants (Dunst reported a developmental quotient of 85), but they were clearly below average when they reached their first birthday (DQ = 62). This “slowing down phenomena was most pronounced for vocal imitation” according to Dunst (1990; p. 224).

While studies on neonatal imitation in humans to date have almost exclusively focused on healthy infants (e.g., Meltzoff et al., 2018; Davis et al., 2021), our goal is to provide observations relevant for children developing along atypical trajectories that might affect how the capacity for primary intersubjectivity develops. We will do this through three different paths presented as three different studies:

The psycINFO, PubMed, and Scopus databases were searched for publications in English on the topic of imitation in populations with DS or ASD younger than 2 years of age. Searches of articles were conducted by the second author (E.H.) on December 3, 2020, in psycINFO and PubMed, and on February 26, 2021, in Scopus (for search terms and limiters, see Supplementary Materials). We did not set any limit for publication date.

A flow diagram of the search including the four phases recommended by Liberati et al. (2009) is presented in Figure 2. The search resulted in 85 records in psycINFO, 42 in PubMed, and 232 in Scopus. After review of titles and abstracts, 50 records were kept for full-text review. The most common reason for exclusion was that the publication was not a study of imitation (e.g., mimicry of medical conditions), studied a non-human population (e.g., rodents), or included participants that were older than the targeted populations. Full-text records were reviewed for inclusion independently by both authors. Inclusion criteria were that the publication described an original empirical study with at least one imitation task (e.g., experimental procedure, observational methods, behavioral ratings), and included participants with DS or ASD that were younger than 12 months of age.

After full-text reviews, four articles were deemed eligible for inclusion by both authors, and we agreed to reassess an additional five articles due to uncertainties in study design. Four of these five articles were included after reassessment, and both authors agreed on excluding the fifth. The reasons for exclusion are listed in Figure 2. The eight articles kept for this review are described in Table 1.

Studies included children with DS (n = 2), ASD (n = 4), or children from a high-risk population for ASD (i.e., younger sibling to a child with an ASD, n = 2). The earliest measure of imitation that was reported was from an age of 1 month, and the same participants were also observed at 3 and 4 months of age (Heimann et al., 1998). Age spans in the rest of the studies were in the range of 6–11 months. The largest study included 86 children with a diagnosis, and the smallest was a single-case study. Two studies reported results from an intervention, with baseline measures of spontaneous imitation in mother-child interaction, while the other studies used parental questionnaire (n = 1), notes from medical records (n = 1), or experimental procedures (n = 3). Four of the studies had a control or comparison group, and one of these (Keemink et al., 2021) reported that 6–9-month-old-infants at high-risk for ASD were less likely than a control group to spontaneously imitate facial emotion expressions; in the three other studies, no between-group differences were detected under the age of 12 months.

The following text tells the story of Marcus, a boy with ASD. When he was 3-days old, his parents used a smartphone to take a video of him as parents often do, and Marcus' mother later provided this home video to the first author (M.H.). The very brief home video shows the father modeling tongue protrusion and how Marcus responds. Marcus received an autism diagnosis before his third birthday, and his story is briefly described in this section.

The aim of this endeavor has been to explore what is known from empirical studies on the existence of imitation or imitation-like responses near birth or during the first year of life among children with non-typical trajectories. We have done this through three different venues: by searching published reports via three different databases; by presenting a previously unpublished observation on facial imitation in a 3-day-old infant that later received an ASD diagnosis; and finally by a renewed presentation of previously published observations on imitation in five 1-month-old children with DS. Based on the observations reported here, we tentatively propose that the little empirical evidence that exist implies that children with ASD and DS have a similar capacity for neonatal imitation as do typically developing children and, thus, an innate capacity for primary intersubjectivity. However, our most critical suggestion is that there is a great need of studies investigating neonatal imitation in atypical populations.

Searching Scopus, PubMed, and PsycInfo for papers on neonatal imitation in atypical populations resulted in 50 papers receiving a full-text review, of which 42 were excluded in the end for not fulfilling our criteria (e.g., focused on diseases, non-human populations, or included participants older than 12 months). Of the eight articles included, only two focused on DS, and the only article that described development from birth in a case of a boy with ASD did not comment on imitation before the age of 9 months (Dawson et al., 2000). The one article describing imitation around the first month of life was the one by the first author on this paper (M.H.), described in detail above. Thus, our literature search shows that almost no empirical research on neonatal imitation exists in the target populations of this paper. This is particularly surprising in the case of ASD, since imitation is assumed to be impaired in this population and a possible precursor of later deviant social development (Rogers and Pennington, 1991; Vanvuchelen et al., 2011). The only documented observation that we have been able to identify is the case of Marcus described in this paper and based on that, we do not currently see any support for the hypothesis that imitation is absent in the neonatal period for children on the track to develop ASD. Clearly, more research is highly needed to test this assumption.

For ASD, the home video of Marcus imitating TP when only 3-days-old raises questions about the starting state of a child developing along an autistic path. True, it is only anecdotal evidence based on a very brief video. But even as such, the observation challenges our knowledge of how ASD develops over time. One might argue that Marcus is unique and that the observation says very little about children with ASD overall. Still, to our knowledge this is the first documented observation of its kind. It has, for instance, direct bearing on the two models based on Bråten's theory (Bråten, 1988, 1998) that Heimann (1998) outlined. Based on the video of Marcus, we suggest that Model B should be dismissed in favor of Model A, which, by allowing for an initial state of primary intersubjectivity, probably is closer to the truth. However, even this model is limited since it does not take the heterogeneity of autism into account (see Fountain et al., 2012; Georgiades et al., 2013; Mottron and Bzdok, 2020). We therefore suggest an updated model, Model C, as illustrated in Figure 5. This new model outlines two possible trajectories for children later receiving an ASD diagnosis. Path a illustrates a child developing typically over the first one to one and a half years, whereafter a regressive pattern occurs meaning that some social or communicative skills are lost (Parr et al., 2011; Thompson et al., 2019). The other trajectory, path b, shows an early deviance from typical development, notable well before the child's first birthday. Note that none of the paths deviate at birth. This is not to say that genetic and biological factors might not be different from typical children early on–although imaging studies so far have been unsuccessful in identifying biological indicators of autism in infants below 6 months of age (see Shen and Piven, 2017). Instead, Model C suggests that any differences in social and communicative skills between children with autism and non-autistic children will not be easily detectable this early on a behavioral level. This proposal is in line with what we saw in the literature search, in which few studies reported a difference between children at risk for ASD and typically developing children during the first year of life.

Berger (1990, p. 137) concludes that most “infants with DS are able to enter into reciprocal interactions with their parents soon after birth” even if some delay can be detected for early behaviors such as smiling, vocalizing, and eye contact. In Berger's sample, mutual eye contact displayed a delay of 2.5 weeks in onset. This is also reflected in Model D (Figure 4) that illustrates two main trajectories for how subject-subject (path c) and subject-object (path d) relationships might develop for children with DS. Note that these paths are similar to what we would expect for typical children (Dunst, 1990; Bråten, 1996; Heimann, 1998). The main difference being that children with DS usually show a much slower pace of development and might not reach the expected end state.

According to the findings presented here and in the Heimann et al. (1998) paper, it would, in our view, be wise to add imitation to the list of early social behaviors that children with DS might display. A capacity which signals that most children with DS have the capacity to establish relationships with the quality of primary intersubjectivity at birth or very shortly after. All children in the DS sample discussed here had at least a month of experience before they took part in the experiment. Thus, we cannot disentangle an innate capacity from rapid learning during their first month of life.

From a theoretical point of view, the observations provided suggest the possibility that both ASD and DS children are born with a mind that has an ability for primary intersubjectivity that makes it possible for them to enter into early dialogues as described by, among others, Reddy (2008). In this way, their starting state seems similar to what we expect to observe in typical neonates. This further implies that the difficulties we see later in development most probably are not caused by a lack of ability for rudimentary social interaction but, instead, emerge when the conditions for interaction changes, possibly when typical children start to engage in secondary intersubjectivity. Another possibility is that for children with ASD, the non-social world at some stage becomes more “attractive” than the social. As suggested by Davis and Crompton (2021), the evolving difficulties will, for some autistic children, also be influenced by non-optimal bidirectional processes that repeatedly create a mismatch between interacting partners, in this case within the early parent-infant dialogues. This is not to say that autism is caused by caregivers' responses, only that also children on a path to autism is affected by continuous positive or negative social experiences.

There are some limitations to consider when interpreting the findings reported here. To start, the empirical base for drawing any conclusion varies between our three studies. The literature search rests on a comprehensive scan across three central and relevant databases that allow us to be more definite about the scarcity of studies investigating imitation in newborn and young children with ASD or DS. The empirical support for the existence of an actual capacity to imitate for children with ASD or DS is however much weaker, close to non-existent. For ASD, we have a single home video that is <1 min long and only the parents' reassurance that the situation was spontaneously filmed. According to the information the parents provided they had never tried to elicit imitation before the video was recorded. Furthermore, a further limitation is that the video only contains documentation of imitation of TP and no sequence when the father was passive that could have been used as comparison. However, the fact that no other responses than TP is produced by Marcus during the brief video adds to the quality of the observation.

For DS, we have taken a new look at already published observations on imitation. Although the data consists of a small number of children (n = 5), they represent all published observations on near neonatal imitation for this group to date, as shown by our literature review. It should however be noted that the DS infants responded differently to what we usually observe for typical infants. The participating DS infants produced a higher frequency of TP in comparison with data from studies on typical infants (Heimann et al., 1998). We do not know if this is a difference that is significant when it comes to the capacity to imitate, but it should remind us that the abilities of newborn children with DS might not be identical to typical infants.

Regarding the literature search, although we worked systematically we might have overseen relevant search terms or additional databases. Perhaps even more critical, we did not sweep the field for gray literature (e.g., unpublished dissertations, null findings in file drawers), and we might therefore have failed to include relevant literature. There is reason to believe that a publication bias might exist in the area of neonatal imitation since there is skepticism in the field whether the phenomenon exists (e.g., Oostenbroek et al., 2016). This, in combination with the general issues in conducting studies on atypical populations (e.g., small populations, large heterogeneity, medical complications), might lead to an unwillingness to include clinical groups like ASD or DS in studies on neonatal imitation. Consequently, this leads to an absence of such studies in published literature which, in our view, is very unfortunate since it hinders theoretical advancements. A recent meta-analysis did conclude that there is evidence to suggest that neonatal imitation exists (Davis et al., 2021), and therefore we believe that it is warranted to focus more on individuals from atypical populations as we continue to investigate it.

Our study highlights the lack of empirical support for the notion that neonates with ASD or DS do not have a capacity to imitate. Although resting on limited evidence, we believe that our observations instead tentatively point in the direction of an imitative capacity also for children that follows a developmental trajectory different from neurotypical children. Thus, all newborn children are probably ready for social encounters, and during their first interactions with another human being, they will use their capacity for primary intersubjectivity to establish reciprocity.

It is also striking that our literature search revealed so few studies on neonatal or early imitation in children with ASD or DS. For ASD, this is despite the fact that imitation in general and neonatal imitation specifically have been theoretically in focus for a long time. The lack of studies including infants or neonates with DS is also surprising since this is a group of children identified very early, often before birth. Thus, it ought to be possible for clinicians to gather larger samples systematically over time. This would give us a more solid ground from which to evaluate how children with DS are capable of imitation early in life. For ASD it is more difficult to study neonatal imitation directly, but one possibility could be to include imitation in the neonatal period in future studies on siblings to children with ASD.

The datasets presented in this article are not readily available because the Down syndrome data has been published earlier and the home video is not public material. However, the procedure for the literature review is available in a Supplementary Materials. Any requests regarding the Down syndrome data or the home video should be addressed to the corresponding author.

The studies involving human participants were reviewed and approved by The Ethical Review Board, Göteborg University for the Down Syndrome children. For the ASD home video, the parents gave consent and approved the final text. This study was conducted according to the World Medical Association's ethical principles as outlined in the Helsinki declaration. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin, including consent for the publication of any potentially identifiable images or data included in this article.

The study was conceived by MH. Data were collected and analyzed by MH and EH. The manuscript was written jointly by both authors.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.