Infants explore their vocal possibilities from birth, producing vocalisations that are on a continuum from cries (rough sounds with a high amplitude and fundamental frequency) to speech-like vocalisations, or protophones (sounds whose morphological and spectro-temporal features resemble speech sounds) (Kent et al., 1987; Nathani et al., 2006; Oller et al., 2019). This pre-linguistic phase of vocal exploration is thought to be crucial for the emergence of speech: it could serve as a base for a selection mechanism whereby caregivers’ differential responses to their infants’ vocal outputs (e.g. different responses to cries versus protophones) progressively lead to the prioritisation of speech signals for communication, both at developmental and evolutionary scales (Ghazanfar and Zhang, 2016; Locke, 2006; Oller and Griebel, 2020). Temporal contingencies (Yoo et al., 2018) could be especially important for infants, allowing them to realise through repeated interactions that some sounds are privileged communicative signals that are particularly efficient to engage their social partners in conversation.

But what determines when infant vocalisations occur initially, and what their acoustic characteristics are? Are infants’ early vocal explorations constrained, and if so, how? One possibility is that vocal explorations follow a stochastic regime early on, and that infants’ explorations of their vocal tract possibilities produce a wide and unconstrained repertoire of outputs that is then narrowed down through the parental selection mechanism described above. Consistent with this idea, Oller and colleagues have proposed that a fundamental ability that supports the emergence of speech is functional flexibility (Oller and Griebel, 2020; Oller et al., 2013). An individual has functional flexibility when at least some of their vocalisations can occur alongside variable underlying affective states, and are not tied to specific communicative functions (e.g. expressing distress). This ability is necessary for the establishment of a language system: it is because we can produce specific sounds to convey different meanings that arbitrary, symbolic systems can emerge (Oller et al., 2013). In short, functional flexibility is a necessary condition for arbitrariness, a key feature of words that supports the emergence of conventional symbolic systems. By contrast, non-human primate vocalisations remain largely inflexible with respect to arousal even in adulthood (Borjon et al., 2016) (although see Taylor et al., 2022).

Infants would be said to have functional flexibility if specific vocalisations that they produce (e.g. protophones) were not tied to specific communicative functions, instead occuring alonside variable affective states. Consistent with this idea, by 3 months, infants can produce speech-like vocalisations in conjunction with both positive and negative facial displays, which suggests that their vocal explorations are functionally flexible in terms of valence (Oller et al., 2013). It remains possible, however, that their vocalisations remain tied to other affective dimensions, in particular autonomic arousal, the fast-acting neural substrate of the body’s stress response mediated by the Autonomic Nervous System (ANS) (Cacioppo et al., 2001; Wass, 2020; Pfaff, 2018; Porges, 2007). Arousal and valence vary in an orthogonal fashion (Kreibig, 2010), so it remains possible that infants vocalisations are tied to arousal, while remaining relatively flexible with respect to valence (e.g., vocalisations produced with both positive and negative affect could be monotonically linked to arousal).

Consistent with this hypothesis, one factor that does appear to influence vocalisation likelihood early on in development is the presence of an interactive social partner (Baumwell et al., 1997; Gros-Louis et al., 2006; Goldstein et al., 2003): although infants also vocalise when they are alone (Oller et al., 2019; Long et al., 2020), from the first few days of life, most infants’ vocalisations cluster with parental speech in time when infants are awake and actively engaged with a partner (Caskey et al., 2011; Dominguez et al., 2016). This might suggest that infants mostly vocalise when they are aroused, in the context of social interactions in particular, and thus, that their vocalisations might remain relatively inflexible – at least with respects to states of arousal, early on in development. This assumption has not been formally tested in humans, but research with marmoset monkeys has shown that vocalisation likelihood and the acoustic properties of vocalisations are both driven by rhythmic fluctuations in the autonomic nervous system (ANS) across multiple temporal scales, in infants and in adults (Ghazanfar and Zhang, 2016; Zhang and Ghazanfar, 2020; McFarland et al., 2020). These influences span across temporal scales, from the temporally fine-grained (spectral features of vocalisations, in the kHz range), through to context-dependent vocalisations likelihood on the scale of minutes to hours (Zhang and Ghazanfar, 2020).

Relatively little research has investigated whether vocal behaviours in human infants are influenced in a similar way by fluctuations in autonomic arousal. Although a number of authors have discussed the relationship between physiological arousal and vocal behaviour, particularly in the context of infant cries (Wolff, 1967; Zeskind et al., 1985; Wilder, 1974), no research to our knowledge (other than work from McFarland and colleagues, discussed below) have directly measured it. Studying this is important for two reasons.

First, as mentioned above, so far, research on functional flexibility has focused on valence and correspondences between visible facial affects and vocalisations only, which limits our understanding of how and when functional flexibility emerges across development (Oller and Griebel, 2020). Thus, here, we investigate whether each infant’s and adult’s vocalisation likelihood is overall more contingent on arousal (either their own arousal level or their partner’s). We also subdivide infant vocalisations into several types: cries (which our supplementary analyses show tend to be negative affective valence) and speech-like vocalisations (which our analyses show tend to be neutral affective valence). The functional flexibility hypothesis predicts that speech-like vocalisations should be relatively independent from arousal, while cries (alarm signals) should largely co-vary with arousal (Altenmüller et al., 2013). Another important aspect is that functionally inflexiblility of vocalisations with regard to arousal in early infancy could be inidicative of a specific learning mechanism that underlies the development of speech over time. That is, it might be that instead of a stochastic regime, what determines the acoustic features of infants vocalizations early on is precisely fluctuations in arousal, that impact on the tension of the vocal folds, and thus, on the roughness and loudness of vocalizations (Fitch et al., 2002). The parental selection mechanism would then operate on a repertoire of vocalisations that is not stochastic, but grounded in physiology (Ghazanfar and Zhang, 2016).

Second, studying how vocalisations relate to arousal changes and autonomic arousal coupling across the dyad would deepen our understanding of how caregivers identify and respond to various types of vocalisations, leading to selective reinforcement (Locke, 2006; Oller and Griebel, 2020; Zhang and Ghazanfar, 2016; Goldstein and Schwade, 2008; Albert et al., 2018). Many authors have described how mimicry and vocal turn-taking behaviours play roles in socio-communicative development (Condon and Sander, 1974; Schneirla, 1946; Lester et al., 1985; Wilson and Wilson, 2005; Fogel, 2017), and how empathy and physiological synchrony play roles in the development of self-regulation and caregiver-child affiliative bonding (Feldman, 2007; Feldman, 2006; Fogel, 1993; Ham and Tronick, 2009); but the relationship between these two areas remains relatively under-explored. McFarland and colleagues have shown that contingent vocalisations (mother to infant and infant to mother) are more common during periods of respiratory-marked synchrony (McFarland et al., 2020; McFarland, 2001). And our own previous research has shown transient increases in caregiver-child physiological synchrony following negative affect vocalisations ^4223,38. Traditionally, the co-regulation of arousal (i.e. management of arousal across the caregiver-child dyad) has been considered important for the early development of self-regulation (Fogel, 2017; Feldman, 2007; Kopp, 1982; Beebe et al., 2016), but it has rarely been linked to the development of communicative skills. The limited previous research in this area suggests that negative affect vocalisations are more common at high arousal states, and are more likely to elicit contingent caregiver responding (Wass et al., 2019; Tronick, 2007). But, if this is the case, and cries are more likely to elicit parental responses, then how might speech-like vocalisations (i.e. non-cries) become progressively prioritised? To answer this question, it seems crucial to examine whether non-cry vocalisations also elicit changes in arousal, arousal stability and arousal coupling across the caregiver-child dyad, and whether this is related to changes in caregivers’ responses to these vocalisations; but to our knowledge no previous research has examined this.

To investigate these questions, we designed new miniaturised wearable autonomic monitors (electrocardiograms and actigraphs) and miniaturised microphones and video cameras that could be worn by infant and caregivers to obtain day-long recordings in home settings. For technical reasons our microphones recorded a 5-s sample every minute (i.e. 8% of each minute) and therefore our analyses examined only large-scale arousal changes during the 20 min before and after each vocalisation (see Methods for further discussion).

We had two main research questions. First, are caregiver and infant vocalisations as inflexible with regard to arousal as those documented in non-human primates? That is, do different types of vocalisation, such as cries and speech-like sounds, show different patterns of association with arousal across the infant-caregiver dyad? Our hypothesis was that even speech-like vocalisations remain relatively tied to fluctuations in arousal during infancy, in contrast with adulthood. Second, do spontaneously occurring vocalisations during the day co-occur with specific patterns of arousal synchrony and co-regulation? Here, our hypothesis, in line with the parental selection mechanism, was that caregivers would track infants’ arousal fluctuations, and that as a consequence their vocalisations and arousal would be largely tied to their infants rather than their own.

Our results section is in two parts. In part 1, we analyse individual and cross-dyadic arousal changes relative to all vocalisations obtained in our data. In part 2, we subdivide vocalisations into cries and speech-like vocalisations; and, in Appendix 1, by additionally subdividing vocalisations based on vocal intensity and affective valence, manually rated by trained coders (Appendix 1 sections 2.5–2.6).

Part 1 – All vocalisations

Our first research question was: are caregiver and infant vocalisations as inflexible with regard to arousal as those documented in non-human primates? To examine this, we conducted three analyses. First, as a preliminary analysis, we examine how vocalisations are clustered together in time. Second, we examine caregivers’ and infants’ arousal levels around vocalisations, using three approaches: (1) average arousal levels around vocalisations; (2) vocalisation likelihood around arousal peaks; and (3) Receiver Operator Characteristic (ROC) curves. Third, we examine arousal around vocalisations subdivided by the partner’s arousal at the time of the vocalisation.

Temporal clustering

To examine whether infants and caregivers produce clusters of vocalisations simultaneously, we performed the following analysis. For each vocalisation, we estimated the likelihood of another vocalisation occurring both before and after that vocalisation. See Figure 1a, which shows as an example the likelihood of a subsequent infant vocalisation occurring 1–3 min after an initial infant vocalisation. To compare the observed probabilities with chance we performed a control analysis in which we inserted random ‘non-vocalisation’ events into the data and repeated the analysis relative to these ‘non-vocalisations’, and compared the ‘real’ and ‘control’ datasets using a Mann-Whitney U test. We then repeated this analysis across multiple time windows from 20 min before the vocalisation to 20 min after. We also repeated it across multiple contrasts, looking both within an individual (e.g. infant vocalisations relative to infant vocalisations) and across a dyad (e.g. infant vocalisations relative to adult vocalisations). Multiple comparisons were corrected for using permutation-based clustering analysis (described Appendix 1 section 1.9).

Figure 1

Download asset Open asset

After correcting for multiple comparisons, infant vocalisations were more likely to occur relative to a caregiver vocalisation across all time windows examined from 12 min prior to a vocalisation to 8 min after (all ps <0.05). This indicates that, when a caregiver vocalisation had occurred, there was an elevated likelihood of a infant vocalisation occurring for all time windows from 12 min prior to that caregiver vocalisation to 8 min after it. Similarly, caregiver vocalisations were significantly more likely to occur relative to an infant vocalisation from 20 min before to 20 min after; infant vocalisations were significantly more likely to occur relative to another infant vocalisation from 16 min prior to a vocalisation to 20 min after (all ps <.05); and caregiver vocalisations were significantly more likely to occur relative to another caregiver vocalisation from from 16 min prior to a vocalisation to 16 min after. Overall, these findings indicate that vocalisations occurred in clusters both within an individual and across the dyad, which confirms previous reports that infant vocalisations tend to cluster with those of their social partners (; Slone et al., 2023).

Arousal around vocalisations. We conducted three analyses to examine the relation between arousal levels and vocalisations. First, we examined how average arousal levels change before and after vocalisations. Second, we examined how the likelihood of vocalisation changes around peak moments in arousal. Third, we calculated ROC curves to examine whether arousal levels alone can predict vocalisation likelihood.

Analysis 1 - average arousal levels around vocalisations. Figure 2a shows histograms of arousal levels at the time of a vocalisation, including both within-individual (e.g. infant arousal to infant vocalisations) and across the dyad (e.g. infant arousal to adult vocalisations). Figure 2b shows the same information, but also examines change in arousal during the period from 20 min before and 20 min after each vocalisation. Figure 2a is identical to the Time 0 values in Figure 2b. Significance testing was performed by comparing the observed arousal levels around vocalisations with a chance value of 0 (which, for z-scored data, is that individual’s average arousal level across the entire day). Multiple comparisons were corrected for using a permutation-based cluster test (described Appendix 1 section 1.9). For infant arousal to infant vocalisations, significant increases in arousal were observed from 16 min before to 16 min after the vocalisation (all ps <0.05, after correction). Note that these findings are not affected by autocorrelation in the arousal data as this was removed (see Appendix 1 section 1.6). For infant arousal to caregiver vocalisations, significant increases were observed from 16 min before to 18 min after each vocalisation; for caregiver arousal to infant vocalisations, significant increases were observed from 4 min before to 6 min after; for caregiver arousal to caregiver vocalisations, no significant difference from 0 was observed at Time 0, but a significant difference was observed for the period between 4 minutes to 2 minutes before each vocalisation. Overall, these results suggest that infant arousal levels are elevated around both infant and caregiver vocalisations, and that caregivers also show smaller but significant increases in arousal around infant vocalisations. By contrast caregivers’ vocalisations show little association with their own arousal, congruent with the hypothesis of heightened vocal flexibility in human adults as compared to infants.

Figure 2

Download asset Open asset

Analysis 2 - vocalisation likelihood around arousal peaks. Conversely, we also examined the likelihood of vocalisations occurring around peaks in arousal. We identified the moments when the infants’ and the caregivers’ z-scored arousal levels exceeded the top 10% most elevated values observed for that participant that day (Figure 2c and e). Appendix 1 section 2.4 shows the same analysis repeated with different threshold levels (5% and 20%). We then examined the likelihood of vocalisations occurring during the time windows around arousal peaks, and compared this with control data generated in the same way as described in section 1.1. Significance was calculated by performing Mann Whitney U tests and correcting for multiple comparisons using a permutation-based clustering analysis (see Appendix 1 section 1.9). Significant increases in vocalisation likelihood were observed only when we examined the likelihood of infant vocalisations around infant arousal peaks, and when we examined the likelihood of adult vocalisations around infant arousal peaks. Note that this latter finding was not significant when other threshold values were used instead (see Appendix 1 section 2.4). No significant increases in vocalisation likelihood were observed relative to adult arousal peaks. Overall, these results confirm that infant arousal peaks are associated with an increased vocalisation likelihood in both infants and (to a lesser extent) adults, but that peaks in adult arousal are not associated with increased vocalisation likelihood (a marker of greater vocal functional flexibility).

Analysis 3

As a further test of whether arousal levels predict vocalisation likelihood differently in infants and adults, we employed a signal detection framework based on the ROC (see Figure 2d). Each dataset was systematically thresholded at all possible values from its minimum to maximum value. At each threshold, each epoch was individually classified either as a True Positive (above-threshold arousal, vocalisation present) or a False Positive (above-threshold arousal, vocalisation absent). If the systematic thresholding produced as many false alarms as hits, then the feature dimension could not be said to aid in predicting vocalisation likelihood. Following calculation of the ROC curves, the Area Under the Curve (AUC) was calculated: a higher AUC indicates that the feature dimension is more predictive. AUC values were calculated per participant and compared with a chance value of 0.5 using the non-parametric Mann-Whitney U test. Results indicated that the infant arousal was significantly predictive of infant vocalisation likelihood (p<0.001), but that other relationships were not. This is again consistent with the idea that infants’ vocalisations are inflexibly related to their arousal.

Arousal around vocalisations subdivided by partner arousal at the time of vocalisation

Our final method for examining how contingent infant and caregivers’ vocalisations are on arousal levels across the dyad was to subdivide all vocalisations by the partner’s arousal at the time of the vocalisation. Figure 3a shows caregiver arousal relative to infant vocalisations (i.e. the same as the purple line from Figure 2b), but subdivided using a quartile split by infant arousal at the time of the vocalisation. Figure 3b shows infant arousal relative to caregiver vocalisation.

Figure 3

Download asset Open asset

To estimate whether caregivers showed larger arousal changes to high arousal infant vocalisations, we performed a one-way ANOVA repeatedly for each time bin and used a permutation-based temporal clustering analysis to correct for multiple comparisons (see section 1.9). Significant effects were found (P<0.01) such that increased caregiver arousal was observed during the time periods 2–6 min and 10–14 min after high arousal infant vocalisations (Figure 3a). For infant arousal, the opposite finding was observed: high arousal caregiver vocalisations were accompanied by increased infant arousal during the period 10–6 min before the caregiver vocalisation (Figure 3b). Overall, these results suggest that high arousal infant vocalisations are followed by subsequent increases in caregiver arousal, and that high arousal caregiver vocalisations are preceded by increases in infant arousal.

Control analyses

Overall, results thus far suggest that infants’ vocalisations are contingent on their arousal state, whereas adults’ vocalisations are independent of arousal. However, we also considered two possible alternative explanations for this finding. The first is that it may be because vocalisations are more likely to occur while the participants are in physical positions associated with increased arousal. To examine this possibility, we conducted an additional analysis in which we performed video coding to examine infants’ physical position while vocalising (Appendix 1 section 2.2). In brief, this analysis suggested that 49% of infant vocalisations occurred while the infant was freely moving; 33% occurred while they were free but stationary; 7% while strapped sitting; 11% while carried. For adult vocalisations, 44% occurred while the infant was freely moving; 33% while free stationary; 10% while strapped sitting; 12% while carried. Overall when we examined how arousal levels differed by physical position we found no evidence that arousal increases around vocalisations are attributable to changes in physical position.

The second possibility is that arousal increases around vocalisations may be attributable to the physical act of vocalising itself. This may seem unlikely given that we also observed increases in infant arousal relative to caregiver vocalisations (Figure 2b). Yet, because we also observed that caregiver and infant vocalisations occur in clusters (Figure 1b), it remained possible that vocalising itself increased infant arousal in these periods. To address this, we conducted a more fine-grained analysis on a different dataset in which we continuously recorded vocalisations and arousal in 11-month-old infants and their caregivers during two 5-min tabletop interactions (see Appendix 1 section 2.3). The timings and durations of vocalisations were coded to an accuracy of 20 Hz (i.e. 50 ms), and our findings examine heart rate changes on a much finer time-scale (1 sample per second compared with 1 sample per minute for the main analyses). Overall our results suggested that, in a seated tabletop interaction, caregivers showed no change in arousal relative either to vocalisations either from themselves or their partner (the infant). Infants showed non-significant increases in arousal relative to their own vocalisations, which started to increase 5 s before a vocalisation and returned to baseline 20 s after. No changes in infant arousal were observed relative to caregiver vocalisations. The fact that arousal levels start to increase before a vocalisation suggests, consistent with animal research (Borjon et al., 2016), that it is unlikely that arousal changes around vocalisations are purely attributable to the physical act of vocalising itself. The fact that no changes were observed in caregiver arousal around caregiver vocalisations is also consistent with this conclusion.

Arousal stability and arousal coupling around vocalisations

Our second research question was: do spontaneously occurring vocalisations during the day co-occur with specific patterns of arousal, arousal synchrony and arousal co-regulation? To address this we performed the calculation described in the Methods and illustrated in Figure 6.

Arousal stability

Arousal stability was indexed by calculating the auto-correlation in infant and caregiver arousal. No significant changes in infant and caregiver arousal stability were observed relative to adult vocalisations (Figure 4a and e). By contrast, infant vocalisations were associated with decreased arousal stability in infants (Figure 4b), and increased arousal stability in adults (Figure 4f), in the time windows prior to the event. These findings differ markedly, however, when we subdivide infant vocalisations into cries and speech-like vocalisations, as shown in part 2.

Figure 4

Download asset Open asset

Arousal coupling

To measure arousal coupling we calculated the cross-correlation in infant-caregiver arousal, as described in the Methods and illustrated in Figure 8. Results suggested that significantly increased infant-caregiver arousal coupling was observed in the time windows following an adult vocalisation. For infant vocalisations, the same directional effect was observed but results were not significant. These findings again differ markedly when we subdivide infant vocalisations into cries and speech-like vocalisations, as shown in part 2.

Part 2 – Infant vocalisations subdivided by vocalisation type

The findings described in part 1 indicate that 12-month-old infants’ vocalisations are contingent on their arousal state, whereas adults’ vocalisations are independent of arousal. However, there may be important differences between cries and speech-like vocalisations or protophones, which have been argued to already be used flexibly by infants during infancy (Oller et al., 2013). To further test our first research question, therefore, we examined whether different types of vocalisation, such as cries and speech-like sounds, show different patterns of association with arousal. To examine this, we recorded arousal changes relative to vocalisations subdivided by infant vocalisation type, differentiating between cries and speech-like vocalisations (see Methods).

Temporal clustering

To examine whether infants and caregivers produce clusters of vocalisations differently as a function of vocal type, we performed the same analysis as described for part 1, this time splitting cries and speech-like vocalisations. A significant increase in the likelihood of another infant vocalisation occurring was observed from –20 min to +20 min after each infant speech-like vocalisation. For cries, a significantly increased likelihood of a subsequent vocalisation was observed for all time intervals from –20 min to +16 min. For caregiver vocalisations following infant speech-like vocalisations, significant differences from the control were observed from –4 min to +12; for caregiver vocalisations following infant cries, from –4 min to +8 min. Overall, these results suggest that in naturalistic data, vocalisations occur in clusters around both speech-like vocalisations and cries, which is inconsistent with previous reports based on laboratory recordings which suggested that infants often produce speech-like vocalisations that are not directed to social partners (Long et al., 2020).

Arousal around vocalisations

To examine how arousal levels changed relative to cries and speech-like vocalisations, we performed the same three analyses as described for part 1.

Analysis 1 - average arousal levels around vocalisations

The peak arousal (Time 0) at the time of the vocalisation was z-score.56 for cries and.41 for speech-like vocalisations, which for both categories was significantly higher than chance (both ps <0.001) (Figure 5c). A separate Mann-Whitney U test indicated arousal at the time of the vocalisation was significantly higher for cries than for speech-like vocalisations (p<0.01). However, arousal levels after the vocalisation regress to baseline levels more rapidly following cries than following speech-like vocalisations (Figure 5d). For speech-like vocalisations, significant increases in infant arousal were observed from –20 min to +20 min after (Figure 5c); for cries, significant increases in infant arousal were observed from –20 min to 10 min after. Significant increases in caregiver arousal were observed around infant cries (from –1 to +2 min) but not infant speech-like vocalisations.

Figure 5

Download asset Open asset

Analysis 2 - vocalisation likelihood around arousal peaks

Conversely, we also found that both cries and speech-like vocalisations are significantly more likely to occur during the time periods around infant arousal peaks, defined as the top 10% most elevated values observed for that participant that day (Figure 5b). Speech-like vocalisations were significantly more likely to occur from 3 min before to 5 min are infant arousal peaks. Cries were more likely to occur up to 3 min following an infant arousal peak.

Analysis 3 – ROC curves

Results indicated that the infant arousal was significantly predictive of infant cries and speech-like vocalisations (both ps <0.001), but that caregiver arousal was not significantly predictive of either vocalisation type (Figure 5e).

Overall, these results suggest that both cries and speech-like vocalisations are associated with increases in infant arousal, but that infant arousal at the time of the vocalisation is higher for cries than speech-like vocalisations. However, speech-like vocalisations lead to more long-lasting increases in arousal. Adults show arousal changes to cries but not infant speech-like vocalisations.

Using day-long home recordings obtained using miniaturised wearable autonomic monitors and microphones, we examined how autonomic arousal cofluctuates with vocal behaviours in caregiver-infant dyads recorded in a naturalistic home setting. We examined within-individual relationships (e.g. how infant arousal relates to infant vocalisation likelihood) and cross-dyad relationships (e.g. caregiver arousal to infant vocalisation likelihood). We also examined how caregiver-child arousal coupling cofluctuates with vocalisation likelihood. From our results the following conclusions can be drawn:

First, the within-individual relationship between arousal and vocalisation likelihood is strong in infants, and weaker in adults (Figure 2a and b). Symmetrically, for infants, there is an increase in the likelihood of vocalising around arousal peaks, which is not the case for adults (see Figure 2). The same result was confirmed by our ROC analyses, suggesting that vocalisation timings can be predicted based on arousal levels (Figure 2). Our findings are based on analyses conducted at the minute-level temporal scale. Findings from a more temporally fine-grained supplementary analysis (see Appendix 1 section 2.2) suggest, consistent with findings from animal research, that it is unlikely that these changes are purely attributable to the physical act of vocalising itself. These findings are also unlikely to be attributable to changes in physical positioning around vocalisations, as our supplementary analyses suggest that most vocalisations occur while free roaming, and that differences in arousal contingent on physical positioning are limited (see Appendix 1 section Supplementary Analysis 1 – video analyses of physical position while vocalising). This suggests that infants vocalisations are still relatively inflexible with respects to states of arousal at 12 months.

Our finding that the association between arousal and vocalisation likelihood is strong in infants, but weaker in adults, can be contextualised by previous research suggesting that the association between arousal and vocalisation likelihood is present in both infant and adult marmoset monkeys (Ghazanfar and Zhang, 2016; Borjon et al., 2016). Whereas previous research has suggested that human infants already show vocal flexibility with respect to affective valence (i.e. they can use similar vocalisations in conjunction with various facial affects) (Oller and Griebel, 2020; Oller et al., 2013), our findings suggest that infant vocalisations are relatively inflexible with regard to arousal during early development. This was true both for cries and for speech-like vocalisations, which were also more likely to occur around arousal peaks (Figure 5b), as confirmed by our ROC analyses (Figure 5e). This discrepancy with previous findings might be due to a genuine difference in functional flexibility across arousal and valence, which is possible given the orthogonality of these two constructs (one can be happy, highly aroused and positive, and angry, highly aroused and negative). However, at this stage it remains equally possible that the discrepancy between our study and previous studies focusing on facial affects stem from methodological differences. Physiological measures might be more sensitive than relying on overt displays, and in future we could use other measures (e.g. acoustic analyses of vocalisations Fitch et al., 2002) to try to understand whether and how functional flexibility develops at the same rate for arousal and valence. Nonetheless, our findings suggest that, in early infancy, vocal production is directly related to the arousal state of the infant, with increases in arousal necessary for infants to vocalise, whatever vocalisation type.

By contrast, our results show that caregiver vocalisation likelihood is more influenced by the infant’s arousal than by the caregiver’s own arousal. The likelihood of an adult vocalising increases around infant arousal peaks (Figure 2e – purple line), more than around the caregiver’s own arousal peaks (Figure 2e – blue line). No relationship was observed between caregiver vocal intensity and caregiver arousal (Fig S9c), but high-intensity caregiver vocalisations are preceded by high infant arousal (Fig S9a). Similarly, no associations were noted between caregiver vocalisation type and caregiver arousal (Fig S9d), whereas positive caregiver vocalisations are more likely to be preceded by high infant arousal (Fig S9b). Finally, caregivers are more likely to produce high arousal vocalisations following an increase in infant arousal (Figure 3b).

Taken together, these findings confirm that caregivers’ speech is flexible with respect to their own levels of arousal (in contrast with what has been documented in other primate species Zhang and Ghazanfar, 2016) but reveal that it is finely attuned to their child’s levels of arousal. This suggests that speech is an under-appreciated mechanism for arousal co-regulation during early life. Previous research has pointed to caregiver-child arousal synchrony as a mechanism for arousal co-regulation (Feldman, 2007; Tronick, 2007). One mechanism for this might be that parents dynamically modulate their own arousal level to match their child (Wass et al., 2019). The present findings reinforce this by suggesting, for example, that high arousal infant vocalisations tend to be followed by subsequent increases in caregiver arousal (Figure 3a). Vocalisations thus appear as an important medium through which arousal is made manifest between caregivers and their infants, supporting coregulation at the level of the dyad.

Our vocal type findings also point to a clear role for vocalisations in arousal co-regulation (Wolff, 1967; Zeskind et al., 1985). Infant arousal is higher when they produce cries (Figure 5c), and they show decreased arousal stability around cries (Figure 4c). Infant cries also lead to changes across the dyad: increased caregiver arousal stability is observed in the time following the vocalisation (Figure 4g), as well as increased caregiver-child arousal coupling (Figure 4k). Comparing arousal levels before and after a vocalisation suggests that cries tend to be followed by decreases in arousal (Figure 5c) and increases in arousal stability (Figure 4c). This may be because cries are less likely to occur in clusters (i.e. to be followed by another vocalisation) than speech-like vocalisations are (Figure 5a). This suggests that cries are events that serve to down-regulate the infant’s arousal, and that the likely mechanism for this is through caregiver-infant arousal coupling.

Speech-like vocalisations show a strikingly different profile. Although arousal levels at the time of the vocalisation are also elevated relative to baseline (Figure 5c), the profile of change after the vocalisation is markedly different as compared to cries. Speech-like vocalisations tend to lead to sustained increases in infant arousal, in contrast to cries which lead to decreases in arousal (Figure 5c and b). Infant speech-like vocalisations are also more likely to be followed by other speech-like vocalisations from both the infant and the caregiver (Figure 5a). This potentially indicates increased attentional processes in the time-period after a speech-like vocalisation that could support the processing of caregiver’s verbal responses, information encoding, and influence later vocal production (Zhang and Ghazanfar, 2020; Goldstein and Schwade, 2008).

Most important to selective reinforcement accounts of early speech development, whereas cries vocalisations lead to changes in arousal stability and coupling across the dyad, speech-like vocalisations do not (Figure 4h and l). Thus, despite the fact that they are both associated with fluctuations in the infant’s own arousal, there are functional differences between cries and speech-like vocalisations at the level of the dyad: compared to cries, speech-like vocalisations do not seem to relate to arousal co-regulation, and they induce less changes in arousal in caregivers. This may be an important mechanism for selective learning in early speech development, with caregivers’ unimpaired arousal allowing for more flexible, timelier, and, potentially, more semantically attuned responses to these vocalisations. Supporting this suggestion, parents overlap more frequently in their responses to infant cry-sounds, compared to protophones (Yoo et al., 2018), and, in our data, speech-like vocalisations were more likely to be followed by caregiver vocalisations over a longer time-period, compared to cries (Figure 5a). These data are consistent with the idea that a parental selection mechanisms grounded in stress physiology is the ‘engine for vocal development’ (Ghazanfar and Zhang, 2016).

Although technical factors meant that we were confined to random sampling during the day rather than continuous recordings, our analyses suggest that this sparse sampling preserves the temporal structure of the data (see Appendix 2 section Supplementary analysis – simulation to examine the effects of sparse sampling on the data); furthermore, we examine event-related changes relative to vocalisations, and so the presence of undetected vocalisations can only have weakened the patterns of event-related change that we have documented here. Nevertheless, future research based on continuous recordings would allow us to examine in more detail the role of turn-taking behaviours in communicative exchanges – examining, for example, whether arousal facilitates effective communication between caregiver and child by making children more likely to respond to verbal initiations by the caregiver. Furthermore, it would also be interesting to examine whether the link between arousal and vocalisations remains unchanged even in the absence of the caregiver, or where the caregiver is unresponsive.

In future, it would also be interesting to explore the role of vocalisations in developmental psychopathology. For example, our present results suggested that, in typical dyads, high arousal infant vocalisations tend to be followed by increases in caregiver arousal. But other work from our group has shown that, in caregivers with elevated anxiety, moments of high infant arousal were more likely to be accompanied by high caregiver arousal; that anxious caregivers were more likely to vocalise intensely at high arousal, and to produce intense vocalisations that occurred in clusters; and that high intensity vocalisations were associated with more sustained increases in autonomic arousal for both anxious caregivers and their infants (Smith et al., 2021a; Smith et al., 2021b). Understanding how arousal relates to vocalisation likelihood across typical and atypical development is an important goal for future research. Relatedly, longitudinal studies are needed to examine the association between infant arousal and vocalisations across development; when, for example, do speech-like vocalisations become functionally flexible from arousal, and how is this affected by early infant-caregiver coregulation?

Overall, our data show that there is a functional dissociation between speech-like vocalisations and cries: cries are more likely to lead to changes in the caregiver’s arousal, while speech-like vocalisations are more likely to associate with sustained increases in infant arousal, as well as an increase in vocalisations in both infants and caregivers. These results are consistent with the idea that caregivers’ differential responses to specific types of vocalisations (i.e. speech-like vocalisations), which are not yet produced flexibly by infants, may be an important factor driving speech development, and they suggest that this bidirectional physiological process supports progressively specialised vocalisations through parental selection.

Due to the personally identifiable nature of this data (home voice recordings from infants) the raw data is not publicly accessible. Researchers who wish to access the raw data should email the lead author s.v.wass@uel.ac.uk. Permission to access the raw data will be granted as long as the applicant can guarantee that certain privacy guidelines (e.g. storing the data only on secure, encrypted servers, and a guarantee not to share it with anyone else) can be provided. In order to allow access to the raw data the name of the applicant will also need to be added to our current ethics approval from the University of East London. This is expected to be routine, as long as the applicant is able to provide these guarantees. De-identified versions of the data - i.e. the processed Autonomic Nervous System data, and the raw coding showing the timings of when vocalisations were recorded, is available here: https://doi.org/10.5061/dryad.612jm6473. The code used to conduct the analyses is available here: https://doi.org/10.5281/zenodo.7409281. A readme file in the Data folder explains how to run the analyses to reproduce the results that we report in the paper. For any queries, please contact the lead author s.v.wass@uel.ac.uk.

1. 1. Wass S

   (2022) Dryad Digital Repository

   Vocal communication is tied to interpersonal arousal coupling in caregiver-infant dyads.

   https://doi.org/10.5061/dryad.612jm6473
2. 1. Wass SV

   (2022) Zenodo

   Vocal communication is tied to interpersonal arousal coupling in caregiver-infant dyads.

   https://doi.org/10.5281/zenodo.7409281

1. 1. Abney DH
   2. Dale R
   3. Louwerse MM
   4. Kello CT

   (2018) The bursts and lulls of multimodal interaction: temporal distributions of behavior reveal differences between verbal and non-verbal communication

   Cognitive Science 42:1297–1316.

   https://doi.org/10.1111/cogs.12612

   • PubMed
   • Google Scholar
2. 1. Albert RR
   2. Schwade JA
   3. Goldstein MH

   (2018) The social functions of babbling: acoustic and contextual characteristics that facilitate maternal responsiveness

   Developmental Science 21:e12641.

   https://doi.org/10.1111/desc.12641

   • PubMed
   • Google Scholar
3. Book

   1. Altenmüller E
   2. Schmidt S
   3. Zimmermann E

   (2013) Evolution of Emotional Communication

   Oxford Academic.

   https://doi.org/10.1093/acprof:oso/9780199583560.001.0001

   • Google Scholar
4. Conference

   1. Aurobinda A
   2. Mohanty BP
   3. Mohanty MN

   (2016) R-peak detection of ECG using adaptive thresholding

   2016 International Conference on Communication and Signal Processing (ICCSP) IEEE.

   https://doi.org/10.1109/ICCSP.2016.7754140

   • Google Scholar
5. 1. Baumwell L
   2. Tamis-LeMonda CS
   3. Bornstein MH

   (1997) Maternal verbal sensitivity and child language comprehension

   Infant Behavior and Development 20:247–258.

   https://doi.org/10.1016/S0163-6383(97)90026-6

   • Google Scholar
6. 1. Beebe B
   2. Messinger D
   3. Bahrick LE
   4. Margolis A
   5. Buck KA
   6. Chen H

   (2016) A systems view of mother-infant face-to-face communication

   Developmental Psychology 52:556–571.

   https://doi.org/10.1037/a0040085

   • PubMed
   • Google Scholar
7. 1. Borjon JI
   2. Takahashi DY
   3. Cervantes DC
   4. Ghazanfar AA

   (2016) Arousal dynamics drive vocal production in marmoset monkeys

   Journal of Neurophysiology 116:753–764.

   https://doi.org/10.1152/jn.00136.2016

   • PubMed
   • Google Scholar
8. Book

   1. Cacioppo JT
   2. Tassinary LG
   3. Berntson GG

   (2001) Handbook of Psychophysiology

   Cambridge University Press.

   https://doi.org/10.1017/CBO9780511546396

   • Google Scholar
9. 1. Calderon DP
   2. Kilinc M
   3. Maritan A
   4. Banavar JR
   5. Pfaff D

   (2016) Generalized CNS arousal: an elementary force within the vertebrate nervous system

   Neuroscience and Biobehavioral Reviews 68:167–176.

   https://doi.org/10.1016/j.neubiorev.2016.05.014

   • PubMed
   • Google Scholar
10. 1. Caskey M
    2. Stephens B
    3. Tucker R
    4. Vohr B

    (2011) Importance of parent talk on the development of preterm infant vocalizations

    Pediatrics 128:910–916.

    https://doi.org/10.1542/peds.2011-0609

    • PubMed
    • Google Scholar
11. 1. Condon WS
    2. Sander LW

    (1974) Neonate movement is synchronized with adult speech: interactional participation and language acquisition

    Science 183:99–101.

    https://doi.org/10.1126/science.183.4120.99

    • PubMed
    • Google Scholar
12. 1. Dominguez S
    2. Devouche E
    3. Apter G
    4. Gratier M

    (2016) The roots of turn-taking in the neonatal period

    Infant and Child Development 25:240–255.

    https://doi.org/10.1002/icd.1976

    • Google Scholar
13. 1. Fawcett T

    (2006) An introduction to ROC analysis

    Pattern Recognition Letters 27:861–874.

    https://doi.org/10.1016/j.patrec.2005.10.010

    • Google Scholar
14. 1. Feldman R
    2. Greenbaum CW
    3. Yirmiya N

    (1999) Mother–infant affect synchrony as an antecedent of the emergence of self-control

    Developmental Psychology 35:223.

    https://doi.org/10.1037/0012-1649.35.1.223

    • Google Scholar
15. 1. Feldman R

    (2006) From biological rhythms to social rhythms: physiological precursors of mother-infant synchrony

    Developmental Psychology 42:175–188.

    https://doi.org/10.1037/0012-1649.42.1.175

    • PubMed
    • Google Scholar
16. 1. Feldman R

    (2007) Parent-Infant synchrony and the construction of shared timing; physiological precursors, developmental outcomes, and risk conditions

    Journal of Child Psychology and Psychiatry, and Allied Disciplines 48:329–354.

    https://doi.org/10.1111/j.1469-7610.2006.01701.x

    • PubMed
    • Google Scholar
17. 1. Feldman R
    2. Magori-Cohen R
    3. Galili G
    4. Singer M
    5. Louzoun Y

    (2011) Mother and infant coordinate heart rhythms through episodes of interaction synchrony

    Infant Behavior & Development 34:569–577.

    https://doi.org/10.1016/j.infbeh.2011.06.008

    • PubMed
    • Google Scholar
18. 1. Fitch WT
    2. Neubauer J
    3. Herzel H

    (2002) Calls out of chaos: the adaptive significance of nonlinear phenomena in mammalian vocal production

    Animal Behaviour 63:407–418.

    https://doi.org/10.1006/anbe.2001.1912

    • Google Scholar
19. Book

    1. Fogel A

    (1993)

    Developing through Relationships

    University of Chicago Press.

    • Google Scholar
20. Conference

    1. Fogel A

    (2017)

    Two principles of communication: Co-regulation and framing

    New perspectives in early communicative development. pp. 9–22.

    • Google Scholar
21. 1. Ghazanfar AA
    2. Zhang YS

    (2016) The autonomic nervous system is the engine for vocal development through social feedback

    Current Opinion in Neurobiology 40:155–160.

    https://doi.org/10.1016/j.conb.2016.07.016

    • PubMed
    • Google Scholar
22. 1. Goldstein MH
    2. King AP
    3. West MJ

    (2003) Social interaction shapes babbling: testing parallels between birdsong and speech

    PNAS 100:8030–8035.

    https://doi.org/10.1073/pnas.1332441100

    • PubMed
    • Google Scholar
23. 1. Goldstein MH
    2. Schwade JA

    (2008) Social feedback to infants’ babbling facilitates rapid phonological learning

    Psychological Science 19:515–523.

    https://doi.org/10.1111/j.1467-9280.2008.02117.x

    • PubMed
    • Google Scholar
24. 1. Graham FK
    2. Jackson JC

    (1970) Arousal systems and infant heart rate responses

    Advances in Child Development and Behavior 5:59–117.

    https://doi.org/10.1016/s0065-2407(08)60465-6

    • PubMed
    • Google Scholar
25. 1. Gros-Louis J
    2. West MJ
    3. Goldstein MH
    4. King AP

    (2006) Mothers provide differential feedback to infants’ prelinguistic sounds

    International Journal of Behavioral Development 30:509–516.

    https://doi.org/10.1177/0165025406071914

    • Google Scholar
26. 1. Ham J
    2. Tronick E

    (2009) Relational psychophysiology: lessons from mother-infant physiology research on dyadically expanded states of consciousness

    Psychotherapy Research 19:619–632.

    https://doi.org/10.1080/10503300802609672

    • PubMed
    • Google Scholar
27. Book

    1. Jaffe J
    2. Beebe B
    3. Feldstein S
    4. Crown CL
    5. Jasnow MD
    6. Rochat P
    7. Stern DN

    (2001)

    Rhythms of dialogue in infancy: coordinated timing in development

    In: Moore G, editors. Monographs of the Society for Research in Child Development. Wiley-Blackwell. pp. 1–149.

    • Google Scholar
28. 1. Jänig W
    2. Häbler HJ

    (2000) Specificity in the organization of the autonomic nervous system: a basis for precise neural regulation of homeostatic and protective body functions

    Progress in Brain Research 122:351–367.

    https://doi.org/10.1016/s0079-6123(08)62150-0

    • PubMed
    • Google Scholar
29. 1. Kent RD
    2. Osberger MJ
    3. Netsell R
    4. Hustedde CG

    (1987) Phonetic development in identical twins differing in auditory function

    The Journal of Speech and Hearing Disorders 52:64–75.

    https://doi.org/10.1044/jshd.5201.64

    • PubMed
    • Google Scholar
30. 1. Kopp CB

    (1982) Antecedents of self-regulation: a developmental perspective

    Developmental Psychology 18:199–214.

    https://doi.org/10.1037/0012-1649.18.2.199

    • Google Scholar
31. 1. Kreibig SD

    (2010) Autonomic nervous system activity in emotion: a review

    Biological Psychology 84:394–421.

    https://doi.org/10.1016/j.biopsycho.2010.03.010

    • PubMed
    • Google Scholar
32. 1. Lacey JI

    (1967)

    Somatic response patterning and stress: some revisions of activation theory

    Psychological Stress: Issues in Research 1:14–37.

    • Google Scholar
33. 1. Lester BM
    2. Hoffman J
    3. Brazelton TB

    (1985)

    The rhythmic structure of mother-infant interaction in term and preterm infants

    Child Development 56:15–27.

    • PubMed
    • Google Scholar
34. 1. Levenson RW

    (2014) The autonomic nervous system and emotion

    Emotion Review 6:100–112.

    https://doi.org/10.1177/1754073913512003

    • Google Scholar
35. 1. Locke JL

    (2006) Parental selection of vocal behavior

    Human Nature 17:155–168.

    https://doi.org/10.1007/s12110-006-1015-x

    • PubMed
    • Google Scholar
36. 1. Long HL
    2. Bowman DD
    3. Yoo H
    4. Burkhardt-Reed MM
    5. Bene ER
    6. Oller DK

    (2020) Social and endogenous infant vocalizations

    PLOS ONE 15:e0224956.

    https://doi.org/10.1371/journal.pone.0224956

    • PubMed
    • Google Scholar
37. 1. Maris E
    2. Oostenveld R

    (2007) Nonparametric statistical testing of EEG- and MEG-data

    Journal of Neuroscience Methods 164:177–190.

    https://doi.org/10.1016/j.jneumeth.2007.03.024

    • PubMed
    • Google Scholar
38. 1. Maris E

    (2012) Statistical testing in electrophysiological studies

    Psychophysiology 49:549–565.

    https://doi.org/10.1111/j.1469-8986.2011.01320.x

    • PubMed
    • Google Scholar
39. 1. McFarland DH

    (2001) Respiratory markers of conversational interaction

    Journal of Speech, Language, and Hearing Research 44:128–143.

    https://doi.org/10.1044/1092-4388(2001/012)

    • PubMed
    • Google Scholar
40. 1. McFarland DH
    2. Fortin AJ
    3. Polka L

    (2020) Physiological measures of mother-infant interactional synchrony

    Developmental Psychobiology 62:50–61.

    https://doi.org/10.1002/dev.21913

    • PubMed
    • Google Scholar
41. 1. McHugh ML

    (2012) Interrater reliability: the kappa statistic

    Biochemia Medica 22:276–282.

    https://doi.org/10.11613/BM.2012.031

    • Google Scholar
42. 1. Nathani S
    2. Ertmer DJ
    3. Stark RE

    (2006) Assessing vocal development in infants and toddlers

    Clinical Linguistics & Phonetics 20:351–369.

    https://doi.org/10.1080/02699200500211451

    • PubMed
    • Google Scholar
43. 1. Neumaier A
    2. Schneider T

    (2001) Estimation of parameters and eigenmodes of multivariate autoregressive models

    ACM Transactions on Mathematical Software 27:27–57.

    https://doi.org/10.1145/382043.382304

    • Google Scholar
44. 1. Oakes LM
    2. Baumgartner HA
    3. Barrett FS
    4. Messenger IM
    5. Luck SJ

    (2013) Developmental changes in visual short-term memory in infancy: evidence from eye-tracking

    Frontiers in Psychology 4:697.

    https://doi.org/10.3389/fpsyg.2013.00697

    • PubMed
    • Google Scholar
45. 1. Oller DK
    2. Buder EH
    3. Ramsdell HL
    4. Warlaumont AS
    5. Chorna L
    6. Bakeman R

    (2013) Functional flexibility of infant vocalization and the emergence of language

    PNAS 110:6318–6323.

    https://doi.org/10.1073/pnas.1300337110

    • PubMed
    • Google Scholar
46. 1. Oller DK
    2. Caskey M
    3. Yoo H
    4. Bene ER
    5. Jhang Y
    6. Lee CC
    7. Bowman DD
    8. Long HL
    9. Buder EH
    10. Vohr B

    (2019) Preterm and full term infant vocalization and the origin of language

    Scientific Reports 9:14734.

    https://doi.org/10.1038/s41598-019-51352-0

    • PubMed
    • Google Scholar
47. 1. Oller DK
    2. Griebel U

    (2020) Functionally flexible signaling and the origin of language

    Frontiers in Psychology 11:626138.

    https://doi.org/10.3389/fpsyg.2020.626138

    • PubMed
    • Google Scholar
48. Book

    1. Pfaff D

    (2018)

    How Brain Arousal Mechanisms Work: Paths Toward Consciousness

    Cambridge University Press.

    • Google Scholar
49. 1. Porges SW

    (2007) The polyvagal perspective

    Biological Psychology 74:116–143.

    https://doi.org/10.1016/j.biopsycho.2006.06.009

    • PubMed
    • Google Scholar
50. 1. Quas JA
    2. Yim IS
    3. Oberlander TF
    4. Nordstokke D
    5. Essex MJ
    6. Armstrong JM
    7. Bush N
    8. Obradović J
    9. Boyce WT

    (2014) The symphonic structure of childhood stress reactivity: patterns of sympathetic, parasympathetic, and adrenocortical responses to psychological challenge

    Development and Psychopathology 26:963–982.

    https://doi.org/10.1017/S0954579414000480

    • PubMed
    • Google Scholar
51. 1. Samuels ER
    2. Szabadi E

    (2008) Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function Part I: principles of functional organisation

    Current Neuropharmacology 6:235–253.

    https://doi.org/10.2174/157015908785777229

    • PubMed
    • Google Scholar
52. 1. Schneirla TC

    (1946) Problems in the bio-psychology of social organization

    The Journal of Abnormal and Social Psychology 41:385–402.

    https://doi.org/10.1037/h0055210

    • Google Scholar
53. 1. Slone LK
    2. Abney DH
    3. Smith LB
    4. Yu C

    (2023) The temporal structure of parent talk to toddlers about objects

    Cognition 230:105266.

    https://doi.org/10.1016/j.cognition.2022.105266

    • PubMed
    • Google Scholar
54. 1. Smith CG
    2. Jones EJH
    3. Charman T
    4. Clackson K
    5. Mirza FU
    6. Wass SV

    (2021a) Anxious parents show higher physiological synchrony with their infants

    Psychological Medicine 52:1–11.

    https://doi.org/10.1017/S0033291720005085

    • PubMed
    • Google Scholar
55. Preprint

    1. Smith C
    2. Jones EJH
    3. Charman T
    4. Clackson K
    5. Mirza F
    6. Wass S

    (2021b) Vocalisation and Physiological Hyperarousal in Infant-Caregiver Dyads Where the Caregiver Has Elevated Anxiety

    PsyArXiv.

    https://doi.org/10.31234/osf.io/wse8t

    • Google Scholar
56. 1. Suveg C
    2. Shaffer A
    3. Davis M

    (2016) Family stress moderates relations between physiological and behavioral synchrony and child self‐regulation in mother–preschooler dyads

    Developmental Psychobiology 58:83–97.

    https://doi.org/10.1002/dev.21358

    • Google Scholar
57. 1. Taylor D
    2. Clay Z
    3. Dahl CD
    4. Zuberbühler K
    5. Davila-Ross M
    6. Dezecache G

    (2022) Vocal functional flexibility: what it is and why it matters

    Animal Behaviour 186:93–100.

    https://doi.org/10.1016/j.anbehav.2022.01.015

    • Google Scholar
58. 1. Thomas KA
    2. Burr RL

    (2008) Circadian research in mothers and infants: how many days of actigraphy data are needed to fit cosinor parameters?

    Journal of Nursing Measurement 16:201–206.

    https://doi.org/10.1891/1061-3749.16.3.201

    • PubMed
    • Google Scholar
59. Book

    1. Tronick E

    (2007)

    The Neurobehavioral and Social-Emotional Development of Infants and Children

    WW Norton & Company.

    • Google Scholar
60. 1. Vest AN
    2. Da Poian G
    3. Li Q
    4. Liu C
    5. Nemati S
    6. Shah AJ
    7. Clifford GD

    (2018) An open source benchmarked toolbox for cardiovascular waveform and interval analysis

    Physiological Measurement 39:105004.

    https://doi.org/10.1088/1361-6579/aae021

    • PubMed
    • Google Scholar
61. 1. Wass SV
    2. de Barbaro K
    3. Clackson K

    (2015) Tonic and phasic co-variation of peripheral arousal indices in infants

    Biological Psychology 111:26–39.

    https://doi.org/10.1016/j.biopsycho.2015.08.006

    • PubMed
    • Google Scholar
62. 1. Wass SV
    2. Clackson K
    3. de Barbaro K

    (2016) Temporal dynamics of arousal and attention in infants

    Developmental Psychobiology 64:1–17.

    https://doi.org/10.1002/dev.21406

    • Google Scholar
63. 1. Wass SV
    2. Smith CG
    3. Clackson K
    4. Gibb C
    5. Eitzenberger J
    6. Mirza FU

    (2019) Parents mimic and influence their infant’s autonomic state through dynamic affective state matching

    Current Biology 29:2415–2422.

    https://doi.org/10.1016/j.cub.2019.06.016

    • PubMed
    • Google Scholar
64. Preprint

    1. Wass S v

    (2020) The Origins of Effortful Control: How Early Development within Arousal/Regulatory Systems Influences Cognitive and Affective Control

    PsyArXiv.

    https://doi.org/10.31234/osf.io/n6fjm

    • Google Scholar
65. 1. Wilder CN

    (1974)

    Respiratory patterns in infant cry

    Developmental Psychology 6:293–301.

    • Google Scholar
66. 1. Wilson M
    2. Wilson TP

    (2005) An oscillator model of the timing of turn-taking

    Psychonomic Bulletin & Review 12:957–968.

    https://doi.org/10.3758/bf03206432

    • PubMed
    • Google Scholar
67. 1. Wolff PH

    (1967)

    The role of biological rhythms in early psychological development

    Bulletin of the Menninger Clinic 31:197–218.

    • PubMed
    • Google Scholar
68. 1. Yoo H
    2. Bowman DA
    3. Oller DK

    (2018) The origin of protoconversation: an examination of caregiver responses to cry and speech-like vocalizations

    Frontiers in Psychology 9:1510.

    https://doi.org/10.3389/fpsyg.2018.01510

    • PubMed
    • Google Scholar
69. 1. Zeskind PS
    2. Sale J
    3. Maio ML
    4. Huntington L
    5. Weiseman JR

    (1985)

    Adult perceptions of pain and hunger cries: a synchrony of arousal

    Child Development 56:549–554.

    • PubMed
    • Google Scholar
70. 1. Zhang YS
    2. Ghazanfar AA

    (2016) Perinatally influenced autonomic system fluctuations drive infant vocal sequences

    Current Biology 26:1249–1260.

    https://doi.org/10.1016/j.cub.2016.03.023

    • PubMed
    • Google Scholar
71. 1. Zhang YS
    2. Ghazanfar AA

    (2020) A hierarchy of autonomous systems for vocal production

    Trends in Neurosciences 43:115–126.

    https://doi.org/10.1016/j.tins.2019.12.006

    • PubMed
    • Google Scholar

Author details

1. Sam Wass

   Department of Psychology, University of East London, London, United Kingdom

   Contribution

   Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing

   For correspondence

   s.v.wass@uel.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7421-3493

2. Emily Phillips

   Department of Psychology, University of East London, London, United Kingdom

   Contribution

   Formal analysis, Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

3. Celia Smith

   Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, United Kingdom

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

4. Elizabeth OOB Fatimehin

   Department of Psychology, University of East London, London, United Kingdom

   Contribution

   Formal analysis

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5179-9583

5. Louise Goupil

   Université Grenoble Alpes, Grenoble, France

   Contribution

   Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

• 641

  views

• 123

  downloads

• 4

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.