Neologisms: A Case Study

Background: Children with Autism Spectrum Disorder (ASD) use neologisms (i.e., invented words or phrases) more frequently and more persistently than typically developing children (Volden & Lord, 1991). This unusual prevalence suggests a value to these made-up expressions and thus justifies and motivates the clarification of their function. One such neologism was found by Chin (2013) from a child, Audrey, with high-functioning ASD. Audrey said “I’m a (verb),” as in “I’m a watch,” 72 times with 26 different verbs. Chin and colleagues (Chin et al., 2012, 2013) conjectured that the frame referred to the future, demonstrated that it was unattested in her input, and documented that it had mostly disappeared by the end of the study; however, the exact meaning of the frame is still unclear, as was Audrey’s purpose in producing it.

Objectives: We investigated whether Audrey’s “I’m a” frame was an alternative form of an existing expression by examining its privileges of occurrence relative to her future and present frames.

Methods: Audrey was a participant in the Speechome Project, which collected densely-sampled, high-quality audio-video recordings in homes during normal activities over four months (Vosoughi et al., 2012). At the time of the study, Audrey was 2;10-3;1, with a Mean Length of Utterance range of 2.2-3.5. Audrey’s utterances of “I’m a (verb)” were compared with her uses of a conventional future frame (“I’m gonna (verb)”) and a conventional present frame (“I’m (verb)ing”), as follows. First, her uses of “I’m a” and “I’m gonna” were coded as ‘present’ if they overlapped with their referent event or occurred within 30 seconds of it, because of children’s difficulty with the boundaries of ‘present’ (Harner, 1980). Referent events occurring more than 30 seconds later were coded as ‘future’. Implications of this criterion are discussed further in the results. Second, her utterances of “I’m a (verb)” and “I’m (verb)ing” were coded according to the verbs they appeared with. Verbs were categorized as Pro-verbs (i.e., the verbs that children use first with multiple verb inflections: “go,” “do,” “make,” “get,” “eat,” “sit,” “ride,” and “fix”; Bloom et al., 1980), or Other verbs (all others).

Results: (1) All but 1 of Audrey’s uses of “I’m a” were produced in the context of ‘present’ activities, whereas 15 of 73 instances of “I’m gonna” referred to the future (chi2=12.12, p<.001). Moreover, those 15 cases of “I’m gonna” were not boundary cases, but referred to events happening in the distant future (e.g., bed time). (2) Audrey used “I’m a” more often with Pro-verbs, but “I’m (verb)ing” (N=158) more often with Other verbs (Figure 1, chi2=10.93, p<.001).

Conclusions: Audrey’s usage of “I’m a” differed from both her use of the conventional future frame “I’m gonna” as well as of the conventional present frame “I’m (verb)ing”. Thus, it does not appear to be an alternative form of either one, but possibly a wholly distinct construction.  We continue to analyze its semantic/syntactic properties, as well as the contributions from Audrey’s emotional state and cognitive demands, to discover just why it emerged.