Referent selection in children with Autism Spectrum Condition and intellectual disabilities: Do social cues affect word-to-object or word-to-location mappings?
Kids with autism can use social cues to learn words, but add intellectual disability and the cues stop working—so teach the word-object link directly and test after any change.
01Research in Context
What this study did
Field et al. (2019) watched kids learn new words while an adult looked or pointed at toys. They tested three groups: children with autism, children with autism plus intellectual disability, and typically developing peers.
After each demo, the team either kept the toy in the same spot or moved it. Then they asked, 'Where is the blick?' to see who still picked the right item.
What they found
Children with autism only used the adult's eye-gaze and pointing to pick the new word. Their scores stayed high even when the toy moved.
Children with autism plus intellectual disability did not use the social cues. When the toy changed places, they lost track of the word.
How this fits with other research
Hempkin et al. (2025) ran a similar task and found that all preschoolers, autism or not, learned words faster when gaze, pointing, and excited voice were added. The 2025 study used an A-B-A-B design, so the cue came and went; kids' correct naming rose and fell with it. Together, the two papers show that social cues help typical kids and autism-only kids, but the benefit shrinks once intellectual disability is added.
Matson et al. (2009) took a different path. They taught first words to non-verbal children with autism by pairing custom sounds or lights with the object. Every child in that study began to speak. Charlotte's work says, 'Don't trust gaze alone for ASC plus ID.' L's work answers, 'Then use a concrete orienting cue and fade it in.'
Congiu et al. (2016) looked at mental-state gaze: kids with autism could follow an eye shift but missed its hidden meaning. Charlotte's task only needed kids to follow the shift, not read minds, explaining why the autism-only group succeeded.
Why it matters
If you teach a child with both autism and ID, do not assume eye-gaze or pointing will lock in the word. Show the item, name it, let the child touch it, then test again after you slide the item to a new spot. Build the word-object link with extra trials, not extra social hints.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After you name an object, move it six inches and ask again—if the child fails, drop the social cue and use errorless teaching to rebuild the word-object link.
02At a glance
03Original abstract
BACKGROUND: There is conflicting evidence regarding whether children with Autism Spectrum Condition (ASC) and intellectual disabilities (ID) follow social pragmatic cues such as a speaker's eye gaze or pointing towards a novel object to assist mapping a new word onto a new object (e.g. fast mapping). AIMS: We test fast mapping from a speaker's gaze and pointing towards objects in children with ASC and ID with varying chronological and receptive language ages compared with receptive language matched groups of typically developing (TD) children. METHODS AND PROCEDURE: Across eight trials, a speaker gazed and/or pointed towards one out of two objects while saying a new word. Pointing was either 'referential' (with intention), or 'incidental' (without obvious intention). To investigate whether children formed more robust word-to-object links rather than associative word-to-location ones, we reversed the original location of the objects in half of the test trials. OUTCOMES AND RESULTS: Children with ASC were as successful as TD children using social cues to form word-to-object mappings. Surprisingly, children with ID did not fast map from referential pointing, or when objects changed location. CONCLUSIONS AND IMPLICATIONS: Children with ID may use different processes to facilitate word learning compared to TD children and even children with ASC.
Research in developmental disabilities, 2019 · doi:10.1016/j.ridd.2019.05.004