Southeast Asian Linguistics Research Unit
Research Spotlight

To me, or not to me: How ASD children produce and comprehend personal reference terms

Speakers of Thai may acknowledge the fact that the language boasts a generous repertoire of personal reference terms, comprising not only over 50 different personal pronouns, but also kinship terms, occupational titles, and personal names (Bandhumedha, 2011; Cooke, 1968; Iwasaki & Ingkapirom, 2009). In a novel experimental semantics and pragmatics study presented at the 22nd Sinn und Bedeutung Conference, Dr. Nattanun Chanchaochai inquires into the production and comprehension of semantically and pragmatically distinct Thai personal reference terms among children with autism spectrum disorders (ASD) and typically developing children (TD).

Turning first to personal pronouns, one crucial concept to distinguish between them has to do with presuppositions. Presuppositions are pieces of information that are implicitly assumed. They can be triggered by the use of certain lexical items, as well as exist in the common ground between discourse participants.

First and second-person pronouns trigger lexical presuppositions, in that the pronominal forms refer to the addresser and the addressee, respectively. The third-person pronoun, however, refers to neither (Lyons, 1977), resulting in no lexical presuppositions being triggered. This unmarked quality of the third person brings about a different type of presupposition: an implicated presupposition (Heim, 1991; Sauerland, 2003, 2008a, 2008b), the inference of which can only be derived through pragmatics — in this case, that of being “anti-participant” or not referring to either of the participants involved.

In non-person dimensions, similar cases can also be observed. Masculine and feminine distinctions exist among first and second-person pronouns in Thai; nevertheless, the restricted epistemic status of male pronoun in (1a) and (1c), as compared to their more general counterparts in (1b) and (1d), suggests that male pronouns are marked in first and second persons. Meanwhile, following Sauerland’s (2008b) proposal, female third-person pronouns are marked.


(1a) ผมกำลังจะไปแล้วนะ

‘I am leaving’ (I = first-person, male)


(1b) ฉันกำลังจะไปแล้วนะ

‘I am leaving’ (I = first-person)


(1c) นายกำลังจะไปแล้วนะ

‘‘You are leaving’ (You = second-person, male)


(1d) เธอกำลังจะไปแล้วนะ

‘You are leaving’ (You = second-person)


Similarly, human pronouns in Thai are also marked, in contrast with the unmarked non-human pronouns, indicating implicated presuppositions at work.

Personal reference terms are of much interest in the autism literature. Numerous studies have shown difficulties with personal pronouns (e.g. Kanner, 1943; Loveland, 1984; Mizuno et al., 2011). Even more attractive are languages with diversified options for personal reference terms, a notable candidate of which is Thai. Apart from Chanchaochai’s (2013) previous work where preferences for personal reference terms with lower deictic levels such as personal names were observed in Thai-speaking children with ASD, not much is known about the bigger picture of personal reference terms and their relevance to the study of autism among speakers of the Thai language, which could bear theoretical implications for the order of acquisition among different types of presuppositions.

In this study, 29 children with ASD and 67 age-, gender-, and non-verbal-IQ-matched TD control participants underwent a production and comprehension task. With each child situated among the experimenter and small cardboard cutouts of a boy, a girl, and a monkey, picture cards were given out to each participant. During the test phase, the participant (in the production task) or the term (in the comprehension task) was randomized as the target at least twice. Table 1 outlines the personal reference terms tested in the experiment, while the instructions are provided in (2a)-(2b).


Table 1 Tested personal reference terms


(2a) Production Task

Experimenter: “Who is holding X?” (twice for each target)

Child: “___ (is holding X).”


(2b) Comprehension Task

Experimenter: “What is Y [tested pronoun] holding?”

Child: “(Y is holding) ___.”


Overall, the results showed that both groups answered accurately in the production task (94.6% for ASD; 90.6% for TD), yet their performances dropped sharply in the comprehension task, with a significantly greater decrease in children with ASD (60.4%; 82.3% for TD). 


Figure 1 Overall accuracy across tasks


The production task produced intriguing results, indicating that the most common first-person personal reference terms for both groups were personal names and personal pronouns — albeit in a reverse order: Children with ASD used personal names more often than personal pronouns, and vice versa for TD children. 


Figure 2 Choices of terms the children used to refer to themselves


While such obverse patterns also arose for second-person terms referring to the experimenter, the proportion of choices was not significantly different across groups. The occupational title ครู appeared more frequently than the kinship term พี่ among children with ASD, while the contrary was the case for TD children. Moreover, for third-person references to the cardboard cutouts, both groups mostly used common nouns.

Additionally, in the comprehension task, the ASD group performed most poorly for third person, while only the male third person yielded poorer performance for the TD group. One instance where children with ASD gave more accurate results than the control group was the formal second-person pronoun คุณ (when the referent is not ambiguous).


Figure 3 Accuracy in comprehension task by item


Based on the aforementioned results and the error analyses provided in the full paper, several pieces of the puzzle start to fit together. Lexical presuppositions appeared to be easier than implicated presuppositions across all groups. Furthermore, recall the different dimensions used to distinguish pronouns: there seems to be a hierarchy of prominence among each cues within a pronoun. Outperforming on the comprehension of the formal second-person pronoun คุณ suggests that the person, not the social-deictic aspect, serves as the primary cue for the ASD group’s identification of the referent (as the TD group’s performance might have been held back by the awkwardness of formality). This is supported by how children with ASD made significantly more mistakes with the pronoun เธอ when referring to the girl figure, despite the female second-person pronoun being supposedly marked, as the gender and social descriptive features were ignored.

In the context of production, in accordance with the results in Chanchaochai (2013), children with ASD still preferred personal names to pronouns. Additionally, both groups frequently used kinship terms and occupational titles. This can also be accounted for by the notion that the usage of pronouns presupposes salience (Roberts, 2004), which means that pronouns might not be as salient as other available referents in the utterance: usually fixed referential terms.

Finally, returning to implicated presuppositions, person unmarkedness is shown to single-handedly decrease the ASD group’s performance. Moreover, when the gender is also unmarked, the accuracy gets dragged down even further. As the TD group’s performance only became poorer after the gender unmarkedness was added, there are two possible interpretations: 1) two implicated presuppositions together are difficult, and 2) gender unmarkedness is especially difficult.

Drawing a clearer picture onto previous literature, this study supports the hypothesis that lexical presuppositions are acquired earlier than implicated presuppositions (Yatsushiro, 2008; Legendre et al., 2001) and that their type and quantity affect acquisition. As for clinical implications, children with ASD tend to have the most trouble with the pragmatic aspects of personal reference terms related to implicated presuppositions. Therefore, children who, instead, struggle with pronouns with marked features and produce pronoun reversals might be revealing signs of autism with language impairment (ALI).



Bandhumedha, N. (2011). Waiyakorn Thai, [Thai grammar] (6th ed.). Bangkok: Academic Publications Project, Faculty of Arts, Chulalongkorn University.

Chanchaochai, N. (2013). “Well, there are no pronoun reversals here”: Case studies of person-referring terms used by Thai-speaking children with autism spectrum disorders. Paper presented at Thai Student Academic Conference – Thai Student Interdisciplinary Symposium 2013 (TSAC-TSIS 2013), March 29-31.

Cooke, J. R. (1968). Pronominal reference in Thai, Burmese, and Vietnamese. Berkeley and Los Angeles: University of California Press.

Heim, I. (1991). Artikel und Definitheit, pp. 487–535. Berlin: De Gruyter.

Iwasaki, S. and P. Ingkapirom (2009). A reference grammar of Thai. Cambridge: Cambridge University Press. 

Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child 2, 217–250.

Legendre, G., I. Barriere, L. Goyet, and T. Nazzi (2011). Quantifier acquisition: presuppositions of ‘every’. In M. Pirvulescu et al. (Eds.), Selected Proceedings of the 4th Conference on Generative Approaches to Language Acquisition North America (GALANA 2010), Somerville, MA, pp. 150–162. Cascadilla Proceedings Project. 

Loveland, K. A. (1984). Learning about points of view: Spatial perspective and the acquisition of I/you. Journal of Child Language 11, 535–556. 

Lyons, J. (1977). Semantics, Vol. 2. Cambridge: Cambridge University Press. 

Mizuno, A., Y. Liu, D. L. Williams, T. A. Keller, N. J. Minshew, and M. A. Just (2011). The neural basis of deictic shifting in linguistic perspective-taking in high-functioning autism. Brain 134, 2422–2435.

Roberts, C. (2004). Pronouns as Definites. Oxford: Oxford University Press.

Sauerland, U. (2003). A new semantics for number. In Proceedings of SALT13, pp. 258–275.

Sauerland, U. (2008a). Implicated presuppositions. Mouton de Gruyter. 

Sauerland, U. (2008b). On the semantic markedness of phi-features, pp. 57–82. Oxford University Press. 

Yatsushiro, K. (2008). Quantifier acquisition: presuppositions of ‘every’. In Proceedings of Sinn und Bedeutung 12, Oslo, pp. 663–677. University of Oslo.

Related posts

Probabilistic Measure for Diffusion of Linguistic Innovation: As Seen in the Usage of Verbal “Nok” in Thai Twitter

How to (more precisely) analyze vowel variation from acoustic data in Thai


Linguistic motivations behind tonal contour changes: A case of Bangkok Thai tones


Kathoey’s reinterpretation of Thai feminine pronouns and the construction of gender identity