Why complete a sentence when you can repeat it?: Thoughts about language in ASD

I was reading a paper on methods for researching language development in autism (by Helen Tager-Flusberg1) when I came across the following fascinating tidbit:

"If a child with autism is asked to describe an event enacted by an experimenter or depicted in a photograph or a sequence of pictures that create a story, he or she is just as likely to simply name the objects (e.g., Tager-Flusberg, 1995) as to provide a narrative description..."

And, further:

"Sentence completion tests or sentence formulation tests (which involve asking a child to create or complete a sentence using a word or phrase given by the tester) are often not easily completed by children with autism. ...Autistic children may misunderstand the instructions and imitate what they have heard, rather than ending or formulating their own utterance."

This fascinates me because it seems so characteristically autistic...or at the least, something a neurotypical child wouldn't do.  I also have no idea why they would respond this way.

A researcher would ask: do autistic children have difficulty with the particular tasks they're given, or with completing sentences or constructing narratives in general?  

This problem has been addressed in a variety of ways.  One theory holds that autistic children lack an understanding of "pragmatics," the unspoken contextual and social rules that underlie the use of language.  Pragmatics come into play when the literal words of a sentence could be interpreted in multiple ways.  For example, in the Hobbit, when Bilbo says "good morning" to Gandalf the wizard, Gandalf points out all the meanings the phrase could imply:

"Do you wish me a good morning, or mean that it is a good morning whether I want it or not; or that you feel good this morning; or that it is a morning to be good on?"

Pragmatics involves the ability to use prior knowledge about the use of the phrase, the particular speaker, and the present conversation to determine which meaning a phrase like "good morning" implies.  This account tends to be popular with "theory of mind" researchers, who believe the social disability involved in autism causes difficulties with pragmatics.

Another theory holds that autistic people are bad at semantics--understanding, remembering, and manipulating the meanings of words.  In a neurotypical child, this would be the first explanation given for failure on tasks like these, which are supposed to be based on semantics.  

However, it's possible that autistic children simply don't have the language comprehension ability (or as researchers put it, "receptive language") to understand the instructions.  A number of studies2,3,4,5 suggest that autistic children--or at least, those with language delays--can actually produce more words and more complex sentence structures than they can understand.  

Furthermore, the tendency to repeat language one hears verbatim out of context--echolalia--is related to receptive language ability.  One study with 18 4-12 year old autistic children found that the poorer autistic children's receptive language abilities were, the more likely they were to use echolalia6.  They often used it combined with a "yes" or "no" to answer questions, in the following manner:

Parent: Do you want some juice?
Child: Do you want some juice? Yes.

In a situation where people are likely to mishear each other or miscommunicate, they often cope by repeating what the other person said to check for misunderstandings.  For example, at a drive through, good cashiers will repeat an order, and savvy customers will ask them to do so.  A person with difficulties understanding what other people say might want to follow a similar strategy--and autistic children might be doing the same thing.  In the example above, one can predict the child's response would then be "Do I want some juice? Yes."  

But now suppose the child also has difficulties with grammar, particularly pronouns (which often happens in autism).  Not only might they be unable to construct the sentence "yes I want some juice," they might not even change the pronoun when they repeat the parent's question.  The result: echolalia.

So how can researchers test whether autistic children understand the questions we're asking them?

There's actually a fairly simple experiment one could do.  Quite possibly, it's already been done.

First, you give an autistic child the standard instructions and ask them to complete or create several sentences.  Observe whether they do so or repeat the prompts.
Tell the child they can take a break for a few minutes and play with some toys in the room.  Then bring the child's parent into the room (to prevent the child from getting anxious, it's better to involve a familiar person rather than an unfamiliar "confederate" of the experimenter).  

Tell the child you're going to play the same game with the parent now.  Then give the parent the same task with the same instructions you gave the child.   Make sure the parent knows to focus on the experimenter and not the child--don't allow them to talk to or make hinting gestures toward the child.  This gives the child an opportunity to see how the task is actually supposed to be done.  When the parent has answered a number of questions, tell them they're done and lead them out.  Once the parent leaves the room, tell the child it's time to play the game again.  
If a child repeated the prompts the first time and completes or constructs sentences the second time--however clumsy the attempt--then seeing the parent modeling the task enabled him to understand it.  
Most people get better at a task when they do it more than once.  To make sure participants don't get better just based on doing the task twice, one could assign some of the children to a control condition.  They would do the task twice with a break in the middle, but they would only play with toys during the break, and would receive the same instructions both times.  
The difficulty here would be in interpreting null results--what would it mean if the child continued to repeat the prompts instead of following instructions, even after their parent modeled the appropriate response?  Were they not paying attention to what their parent was saying?  Could they understand the task, but lack the ability to produce the sentences?  
What do you think?  If you're autistic, or have an autistic child, any ideas why children might be responding this way?  If you're a researcher, has this experiment been done?  If not, is it worth doing? 
References1 Tager-Flusberg, Helen (2000). The challenge of studying language development in children with autism. In Methods for Studying Language Production, pp. 313-332.2 K. Hudry, K. Leadbitter, K. Temple, V. Slonims, H. McConachie, C. Aldred, P. Howlin, T. Charman, PACT Consortium, J. Green, A. Pickles, W. MacDonald, L. White, C. Holt, D. Kapadia, K. Bourne, L. Blazey, T. Houghton, C. Taylor, A. Le Couteur, A. Cutress, S. Leach, S. Barron, R. Colmer, S. Randles, K. Beggs, J. Collino, B. Barrett, & S. Byford (2010). Preschoolers with autism show greater impairment in receptive compared with expressive language abilities. Internatl. Journal of Language & Communicaiton Disorders Vol 45 Iss 6, pp 681-690. 3 J. Maljaars, I Noens, E Scholte, & I van Berckelaer-Onnes (2012). Language in low-functioning children with autistic disorder: differences between receptive & expressive skills and concurrent predictors of language.4  J. Volden, I.M. Smith, P. Szatmari, S. Bryson, E. Fombonne, P. Mirenda, W. Roberts, T. Vaillancourt, C. Waddell, L. Zwaigenbaum, S. Georgiades, E. Duku, & A. Thompson (2011). Using the preschool language scale, 4th edition to characterize language in preschoolers with ASDs. American Journal of Speech-language Pathology Vol. 20 Iss. 3 pp 200-208

5 Susan Ellis Weismer, Catherine Lord & Amy Esler (2010). Early language patterns of toddlers on the autism spectrum compared to toddlers with developmental delay. Journal of Autism & Developmental Disorders Vol. 40 Iss. 10, pp. 1259-1273.6 Robin E. McEvoy, Katherine A. Loveland, & Susan H. Landry (1988). The functions of immediate echolalia in autistic children: a developmental perspective. Journal of Autism & Developmental Disorders Vol. 18 Iss. 4 pp. 657-668.