Trusting another person's testimony is composed of at least two parts: (1) registering the intentions that distinguish acts of communication and (2) treating those communications as reasons for belief. Jaswal & Akhtar (J&A) discuss alternative explanations for behaviors previously interpreted as indicating a lack of social interest in the autistic population, arguing that social motivation is “more appropriately understood as arising from a dynamic interaction between the individual and how others perceive and react to them” (sect. 1, emphasis added). We agree, and we think that there may be two distinct sources of bias that can affect a observer's evaluation of an autistic person's communication: an experience-based bias for the forms that communication can take and a prejudice against the credibility of certain speakers.

The first bias, an experience-based bias, limits an adult observer's ability to register certain non-conventional or less conventional actions as communicative. J&A discuss various ways in which autistic communications are misidentified or misconstrued. What limits an observer's perceptions of autistic communication? Part of the answer likely relates to increasing exposure to the set of actions, symbols, and routines that mark the conventions of the language we learn. There is a growing body of developmental and cross-cultural research showing that infants begin as quite flexible listeners who are open to a range of communicative actions, and they increasingly develop preferences over time as they encounter more instances of certain recurrent, conventional communications (DeLoache Reference DeLoache2004; Namy & Waxman Reference Namy and Waxman1998; Namy et al. Reference Namy, Campbell and Tomasello2004; Woodward & Hoyne Reference Woodward and Hoyne1999). Also, caretakers from different cultural backgrounds address their infants quite differently. Some seek dyadic interaction using object stimulation and mutual gaze, whereas others predominantly use body contact and physical touch (Kärtner et al. Reference Kärtner, Keller, Lamm, Abels, Yovsi and Chaudhary2008; Keller Reference Keller2007; Richman et al. Reference Richman, Miller and LeVine1992). In response, infants increasingly look to their mothers who seek face-to-face contact; but when caretakers primarily focus on physical touch, infants’ gazing behavior does not change with age, and they produce less vocalizations (Kärtner et al. Reference Kärtner, Keller and Yovsi2010; Keller et al. Reference Keller, Otto, Lamm, Yovsi and Kärtner2008). This inductive process depends upon experience delivering actions that listeners repeatedly encounter and gradually recognize as intentional and communicative for their community, which helps explain both why parents of autistic children are good at interpreting the communications of their children and why others outside of that experience often fail.

This experience-based bias for certain communicative actions also highlights the more general issue of how placing priority upon certain intentional and conventional communicative signals (pointing, eye gaze, words, statements) may thereby neglect the importance and nuance that other forms of dynamic interpersonal communication can take. As J&A suggest, we caution against using a narrow set of communicative actions to evaluate children's communicative abilities and motivations. For example, empirical evidence shows that Mexican infants use significantly fewer hand gestures compared with infants in the Netherlands and China (Salomo & Liszkowski Reference Salomo and Liszkowski2013), but fewer hand gestures does not necessarily mean that Mexican babies are any less socially interested than infants in other countries. Therefore, to better understand communicative development across a diverse group of young speakers and listeners, we think that social cognitive research could benefit from greater consideration of social environments, individual differences, and dynamic interactions.

The second bias, a prejudice against the credibility of certain speakers, can limit a listener's willingness to believe the communication. Underestimating the trustworthiness of certain informants makes it difficult for them to convince us by their testimony, even when we recognize their actions as communicative. We support J&A's recommendation of taking autistic testimony seriously and believe that scientists and clinicians run the risk of committing epistemic injustices to those we work with if we are not aware of prejudices that question the credibility of certain speakers (children, minorities, women, etc.). If we discount the credibility of others’ testimony on the grounds of who they are or what they look like, we may give researchers and clinicians more credibility than warranted and less credibility to certain speakers than they deserve.

Developmental research suggests that even the youngest learners may mistrust speakers from less privileged social groups. Children are more likely to mistrust outgroup members who speak with an accent (Corriveau et al. Reference Corriveau, Kinzler and Harris2013) and speakers of a different race (Chen et al. Reference Chen, Corriveau and Harris2013) or gender (Shutts et al. Reference Shutts, Banaji and Spelke2010). When paired with a more appealing informant, they are also more likely to discredit speakers who demonstrate a non-dominant status (Bernard et al. Reference Bernard, Castelain, Mercier, Kaufmann, Van der Henst and Clément2016), appear to be unconfident (Jaswal & Malone Reference Jaswal and Malone2007), have less prestige (Chudek et al. Reference Chudek, Heller, Birch and Henrich2012), or are less attractive (Bascandziev & Harris Reference Bascandziev and Harris2016), preferences that have also been documented in adults (Anderson & Kilduff Reference Anderson and Kilduff2009; Chaiken Reference Chaiken1979; Henrich & Gil-White Reference Henrich and Gil-White2001). These initial assessments of a speaker's trustworthiness influence the plausibility we give to their communications. If the initial assessment is low, it can lead us to look for signs of incompetence or irrationality, which can serve to confirm or even reduce low initial assessments (Jones Reference Jones and Antony2018). Therefore, without reconsidering and changing our initial assessments of autistic speakers, we may deprive ourselves of the opportunity of learning from them, and we deprive them of opportunities to share their knowledge and experiences with us. To alleviate testimonial injustices such as these, researchers need to have a certain “testimonial sensibility” (Fricker Reference Fricker2003), neutralize the effects of prejudicial stereotypes, and undertake separate assessments of the credibility of the speaker and the plausibility of his or her testimony (Jones Reference Jones and Antony2018).

In conclusion, there are two distinct sources of bias that can affect a listener's evaluation of an autistic person's communication: an experience-based bias and a prejudice against the credibility of certain speakers. To discern the intentions of a speaker's communicative action, developmental researchers need to revise their non-social assumptions about less conventional actions. To treat these actions as reasons for belief, researchers can foster greater testimonial sensibility to diminish the effects of prejudicial stereotypes that would otherwise influence one's credibility judgments.