First, a grounding understanding that race is an ideology, constructed for economic and social power, often culturally engrained and psychologically salient (Salter et al., 2018; Smedley and Smedley, 2005). Race also plays a crucial role in forming group dynamics, and informs practices and policies while holding its hegemonic presence (Sekimoto, 2018). Race provides categories for general public as well as researchers to inform their decisions. In that regard, both scholars and general public essentialize these categories assuming that they provide robust [scientific] benchmarks (Hochman, 2021). But, even the labels used for racial categories are not representative of the population (Martinez, 2023). These labels are found to create ambiguous interpretation for scientists (Hochman, 2021).

Importantly, classifying race is also not an easy answer especially in Multiracial communities. This is especially evident when race categories intersect with ethnicity categories. Ethnicity is driven by shared customs, language(s), and may be racially homogenous or diverse (Smedley and Smedley, 2005). Through years of intersectional work, we know that there is erasure of distinctive experiences of marginalized communities. This is evident in cases of Multiracial and Multiethnic individuals where one dimension is ignored or receives more attention than the other one with the assumptions that these choices are power-neutral (Tripp and Munson, 2023; Umaña-Taylor et al., 2014).

The answer to this question greatly depends on from which perspective, culture, or country this topic is being considered (Song, 2021). The increasing body of literature highlights that racial diversity in one’s environment shapes social-group cognition, blurring or softening binary representations to more continuous category representations. One group which experiences this diversity is Multiracial individuals who have interracial experiences. In the U.S. alone Multiracial self-identified population increased by 276% in the past decade, with the Multiracial youth population increasing from 5.6% in 2010 to 15.1% in 2020 (Bureau, 2021). These individuals maintain more flexibility with one’s racial identity(s), and display less biased thinking (Gaither et al., 2015; Pauker et al., 2018; Pauker et al., 2016). However, perceptual development research still could obtain a richer characterization of variance within Multiracial individuals and experiences. Below, we situate some of the challenges faced thus far and then opportunities for shifting our research narratives to more inclusive and integrative ones.

For many decades, developmental psychology and cognitive psychology (along with many other social sciences) fields as well as publication practices (Roberts et al., 2020) pursued a more static approach in which Monoracially and Monolingually White and Western individuals with no disabilities are perceived as the baseline or modal exemplars of the whole population (Amir and McAuliffe, 2020; Henrich et al., 2010; McMurray et al., 2023; Muthukrishna et al., 2020; Nielsen et al., 2017). This practice is particularly visible through the types of research questions asked, stimuli used, demographic surveys, and participant recruitment. Despite cross-cultural evidence of similarly occurring perceptual narrowing during infancy, our working frameworks of when, why, and how such shifts appear in development are primarily understood from a Monoracial and Monolingual baseline. This results in many opportunities to think more inclusively and incorporate multi-input experiences as part of the baseline understanding of perceptual expertise.

Being Multiracial and/or Multilingual has historically been perceived as exclusion criteria. This practice leaves developmental science with a narrowed benchmark of what being a “representative” or generalizable participant is. This research practice is not exclusively driven by limited numbers of available participants. The exclusion of Multiracial and/or Multilingual individuals is supported by the fact that there still exist racialized ideologies in scientific places that suggest that being White and/or Monoracial/Monolingual is generalizable (Cheng et al., 2021; Nielsen et al., 2017) with minimal flexibility of category boundaries. The critical emphasis we raise below, is to focus on opportunities to shift away from excluding those with multi-category experiences to including them into developmental science.

One of the most used data collection methods in psychological sciences is the use of demographic surveys. These surveys are used to report participant demographics for three reasons: (i) publication, (ii) reporting for agencies, and (iii) statistical analysis control variables. While questioning the validity of demographic surveys is a whole different venue to be explored, the lack of specificity in traditional surveys makes it challenging to fully characterize Multiracial and Multilingual developmental experiences (Surrain and Luk, 2019). However, in recent years some researchers are making the shift to include more open-ended participant-identity focused questionnaires offering richer representations of individual variability of experience (e.g., Atkin et al., 2022; Byers-Heinlein et al., 2020; Ellis et al., 2017; Roberts et al., 2020).

A racially diverse out-of-home context can provide increased opportunities for a multiplicity of input across race and language. Several studies have broadly examined the way racially diverse societies shape developmental perceptual specialization for familiar-versus unfamiliar-race faces (Bar-Haim et al., 2006; Ellis et al., 2017; Hwang et al., 2020; Tham et al., 2019). For example, U.S. 8-month-old White infants with out-of-home exposure to Individuals of Color displayed distinctive eye movement strategies for unfamiliar-race faces not observed in age-matched White infants without such experiences (Ellis et al., 2017). Greater neural responses thought to reflect top-down attentional control and approach systems toward unfamiliar-race faces was found in U.S. White infants from neighborhoods with larger racial diversity (Hwang et al., 2020). These findings may indicate racial diversity, with even minimal contact, improves perceptual engagement with unfamiliar-race individuals. Furthermore, out-of-home racial diversity may improve Monoracial individuals’ face discrimination of out-group members. For example, 3- to 4-month-old and 8- to 9-month-old Chinese infants living in the racially diverse and integrated city of Kuala Lumpur, Malaysia were tested for face differentiation among familiar-race Chinese faces, more-experienced other-race Malay faces and less-experienced other-race White faces (Tham et al., 2019). Younger infants displayed face discrimination capabilities only for familiar-race female Chinese faces. These findings contradict the straw-man conclusion that 3-month-old infants’ face discrimination capabilities are perceptually most flexible and un-biased (e.g., Kelly et al., 2007, 2009). However, this bias is consistent with 3-month-old Monoracial infants living in a Multiracial U.S. city (Gaither et al., 2012). The 8-to 9-month-old Chinese infants displayed face discrimination capabilities for own-race Chinese and the more-experienced other-race Malay female faces, but not for less-experienced White faces (Tham et al., 2019). These findings indicate a broadening or facilitation pattern of perceptual discrimination may result from out-of-home Multiracial experiences. This broadening pattern is consistent with infants’ face-gender discrimination capabilities (Rennels et al., 2017; Righi et al., 2014). When receiving multiple inputs, infants’ face perception system may first tune to the exemplar most salient and then develop expertise for the next most experienced input. This possible pattern is opposite to homogenous Monoracial experiences that begin broad and then quickly narrow.

Despite the dramatic population increase of Multiracial youth, those raised with caregivers not of the same race, there are very few reports on perceptual specialization for infants raised within a racially mixed insular family. Gaither et al. (2012) tested United States 3-month-old Asian-White Multiracial infants, who all had Asian mothers and White fathers, on a face discrimination task. Asian-White Multiracial infants did display greater proportion of looking to the novel relative to familiarized White face, however, did not show this discrimination for familiar-race, Asian, face comparisons. These findings leave us with many questions about the trajectory of perceptual specialization throughout the first years of life and across other Multiracial familial contexts.

Despite possible broadening or other patterns of perceptual specialization throughout infancy, narrowing for familiar-race(s) faces may eventually develop in childhood. For example, a small sample of Black-White US Biracial participants ranging from 5 years to 23 years of age displayed no differences between face memory scores relative to Monoracial participants (Goodman et al., 2007). Starting at 8 years of age participants displayed the classic, familiar-race bias, showing better memory for familiar-race White as well as Black faces relative to unfamiliar-race faces. These findings suggest that for these Black-White Biracial children their face perception expertise formed for the two race groups they are most exposed to. More data is needed to understand if a pattern of narrowing to more than one race group is consistently found across all Multiracial children and/or contexts.

Societal or individual racial diversity alone may not fully characterize developmental perceptual expertise in childhood. Children develop their own sense of identity along with learning about social categories such as race and gender through explicit labels and implicit behavioral messages (Bigler et al., 2006; Quintana, 1998). For example, 3- to-5-year old child with more experience to racial-labels also have neural responses indicative of implicit category formation based on face race (Timeo et al., 2019). The specificity of familiar-group biases may become even more specialized for some children raised in diverse contexts, especially those who are in marginalized groups. An example of the way ethnicity also plays a role in perceptual expertise in childhood is found in 5- to 16-year-old Asian children of different ethnic backgrounds (e.g., Chinese, Filipino, and Korean). These children failed to display a familiar-race face bias when tested with faces rated as racially Asian, but critically not of the same Asian ethnicity as the children who were tested (Gross, 2013). In contrast, White children succeeded in displaying a familiar-race bias for racially White faces of varying ethnicity. Gross (2013) speculated that because many neighborhoods in Southern California are segregated by race and ethnicity, many of the Asian children in the study had more own-ethnicity exposure whereas, White children likely had own-race exposure without ethnic diversity. This is evidence for when ethnic diversity supersedes race in salience for in-group face specialization, despite being situated in a Multiracial society (See Boxes 1, 2, for more). Work by Gross (2013) highlights that experience with environmental racial and ethnic diversity can be both interconnected and distinct in the influence of developmental perceptual specialization. More work is needed to evaluate the way these different categories of faces may influence face perception biases which begin during infancy. Likely, social systems of hierarchy and experiences of marginalization play a macrolevel role for determining when ethnicity or race are more salient for an infant’s or child’s experiences with face diversity and development of face processing biases.

Overall, the reviewed findings indicate there may be several varying pathways to developing specialization for faces of familiar-race group(s) compared to unfamiliar groups. Throughout the next sections brief summaries of the reported similarities and differences between Monolingual and simultaneous Multilingual speech perception are provided as a guide to shaping future face perception research with Multiracial populations. Then the overlap between developmental Multilingual and Multiracial individuals regarding perceptual specialization is considered and highlights two overarching themes that shape current and future research related to these populations. There are also several open questions proposed (see Box 1) that will push our research toward a more inclusive and robust characterization of perceptual specialization.

The majority of the world’s population is Multilingual or Multidialectal (Grosjean, 1989; Kirk, 2023) and quite possibly Multiracial and/or Multiethnic (Aspinall, 2018). However, this is not well reflected in our research practices, much of it is driven by differences in how these social identity categories are defined as well as recognized across the world. These also lead to the (mis)conceptualization that Multilingual or Multidialectal individuals are two Monolinguals or Monodialectals in one mind (Grosjean, 1989). Thus, many theoretical frameworks are situated to understanding individuals of these populations’ experiences, perceptual, and cognitive development from a deviation from the “mono” context.

As speech is perceived as continuous sound, so too may the social construct of racial categories. The perception, representation, and navigation of racial continuum may be particularly salient for Multiracial individuals or those raised in racially diverse and integrated contexts (Nishina and Witkow, 2020). Research indicates similarities and differences between Multilingual and Monolingual infants when it comes to familiar language perceptual specialization. Many of the differences seem to fall within two overarching themes of perceptual specialization (A) variability of input and (B) the pattern of specialization. It is beyond the scope of this paper to offer an in-depth review of developmental Multilingualism literature, however in the following paragraphs a variety of reported patterns and exemplary evidence are highlighted to lay a foundation for the type of questions and content we are missing for familiar-race face specialization for Multiracial experiences. While we overview these studies, we also need to highlight the limitations of the Multilingualism research overviewed below. The majority of the developmental Multilingualism research has had emphasis on simultaneous Multilingualism from specific regions of the world (Rocha-Hidalgo and Barr, 2022; Singh et al., 2022) and without the careful consideration of many issues that minoritized Multiracial and Multiethnic Multilinguals face in everyday life. Furthermore, Multilinguals’ experiences of minoritization through their language practices has a direct effect on racialization of these individuals. Therefore, the intersection of race and language becomes the crucial point in understanding Multiracial and Multilingual individuals’ cognitive processes (see Kutlu et al., 2022a, Kutlu et al., 2022b for examples from adult populations).

Timing, duration, context, and modality of input varies both within and between individuals and societies. These variations of input, particularly perceptual, are all critically influential to developmental neuroplasticity that defines much of the first years of life (for review, see Oakes, 2017). Variability of input has cascading and multidirectional influence on perceptual specialization output during infancy, which in turn shifts future input experiences for the developing child (Oakes, 2017). We see this input variability is both salient and a shared characteristic for those raised in Multilingual/dialectic and Multiracial/ethnic families as well as societies.

Often all Multilinguals are considered to have less input due to the input being split in different languages. Whereas the overall amount of time hearing speech sounds may not vary between Multi-and Monolingual infants (Kalashnikova and Carreiras, 2022), Multilingual infants are hearing more variation within this time. Time spent with a single home language may be lower because of Multilingual infants negotiating between languages (Byers-Heinlein and Fennell, 2014; Hoehle et al., 2020; Orena et al., 2020) and home languages being marginalized in many contexts (Rocha-Hidalgo and Barr, 2022). The source of linguistic exposure and amount of time exposed to each language are two key qualities for describing Multilingualism (Rocha-Hidalgo and Barr, 2022; Sebastian-Galles and Santolin, 2020). There are similar complexities and some nuances when considering the variability of input for Multiracial experiences.

From simultaneous Multilingual literature, authors report a correlation between reported minimum time spent with multiple languages and significant perceptual discrimination findings. For example, infants with more unbalanced dual-language exposure, seem to show greater differences in perceptual discrimination relative to Monolinguals compared to those infants with more balanced dual-language exposure (Bijeljac-Babic et al., 2012; Sebastián-Gallés and Bosch, 2002). This is also reported for Multidialectal infants, with greater variability of accent input broadens older infants’ capability of deciphering within-category linguistic variance (Potter and Saffran, 2017).² Additionally, bilingual infants who hear more mixing of two languages from their mothers appear to be better at segmenting across their dominant and non-dominant languages compared to those infants with low experiences of language mixing (Orena and Polka, 2019).

However, the majority of these studies did not consider the effects of societal barriers on these infants’ environment. For instance, do these infants have access to multiple languages in and outside of the home? Are there any societal pressure on parents regarding maintaining one language over the other one? Crucially, there is now advancements in Multilingualism research that suggest that when we use continuous measures to capture fundamental processes in spoken language such as phoneme categorization, it becomes evident that what was interpreted as deficiency or lack of discrimination ability in Multilinguals become a gradient categorization pattern (Kutlu et al., 2022a). We argue that developmental psychology research should adopt these approaches instead of pushing for an understanding that language and faces are non-variant. When we start with the assumption that input is varied, then we observe that Multilingualism experience offers insights to understand how we cognitively adapt to the variation that surround us.

The second overarching theme is the way infants from racially diverse contexts may achieve specialized perceptual capabilities for familiar-race relative to unfamiliar-race faces. Evidence from the handful of previously reviewed investigations examining face-race perception offer initial evidence for both similarities and differences between Multiracial individuals and racially homogenous Monoracial individuals (Gaither et al., 2012; Iankilevitch et al., 2020; Tham et al., 2019; Tham et al., 2017; Woo et al., 2020). Two starting points are proposed to tackle this large theme: Timeline for specialization and what (perceptual) “narrowing” may look like.