On the Origin of LARSPecies

David Crystal

The origins of LARSP lie in a serendipitous encounter with a language-delayed child, at a time when a linguistically informed analysis of language was not a routine weapon in the speech pathologist's armoury. In Britain, it was not until 1972 that a government report on speech therapy services (the Quirk Report) made a statement that now seems blindingly obvious, but which at the time indicated the distance still to be travelled in clinical training and practice:

the would-be practitioner of therapy ... must in future regard language as the central core of his basic discipline.

The consequence was an immediate raising of the profile for linguistics, as the relevant science of language. A new climate formed and a clinical linguistic perspective became routine. LARSP, expounded in 1976 in the book The grammatical analysis of language disability, was part of that climate. But its origins lie a decade before, in that chance encounter.

The Department of Linguistic Science at the University of Reading was established in 1965. One of my roles was to teach courses in the structure of English and child language acquisition, and so, when a phone call came through from the Royal Berkshire Hospital, just down the road, it was put through to me. It was from the audiologist, Dr Kevin Murphy, who was hoping that there was someone who could visit his department to advise on the assessment of a 3-year-old child whose speech was puzzling them. There were no hearing problems, it seems, and intelligence and social skills were normal, but she wasn't saying very much, and when she did speak her utterances evidently sounded immature.

I spent an afternoon observing a session with her therapist. It was the first language-delayed child I had ever seen, but her language output was immediately recognizable to anyone familiar with the limited but growing literature on language development in children. Asked for my opinion about what was going on, I gave a report which (if I were giving it today) would say that this child was making slow progress at Stage II, with several isolated phrases, verb-related gaps in clause structure, no clause element expansion in the direction of Stage III and a worryingly high proportion of minor sentences at Stage I. But this was not today: it was 1968. And as I continued with my description of the child's grammatical difficulties and her developmental level, it was evident that my listeners had no idea what I was talking about.

We took time out to discuss the problem. Grammatical analysis, it appeared, had not been part of the training of any of the professionals in the room. They had a vague memory of 'doing grammar' in their school days, but that had not been a particularly pleasant experience; they had forgotten most of it, and in any case it didn't seem to relate to the practical demands of the clinic. Nor had they ever been given a course on child language development. That was hardly surprising. Child language studies were in their infancy in the 1960s. There was no journal of child language (that did not begin until 1975), no child language association (that was 1970), and the major books that would one day define the subject were yet to be written (Roger Brown's A First Language, for example, wasn't published until 1973). I well recall the difficulty of finding material for students to read, in those days, and – in a pre-CHILDES era – getting hold of audio examples of real children to demonstrate developmental reality. I spent many hours recording my young children, as a consequence, and I know several other linguists who were doing the same.

LARSP, I can see now, was born on that afternoon. I was asked to write up my observations in a report, and present it to a group at the hospital. I did so, giving them a handout which had primitive levels of development (what would later be Stages I to V) illustrated by a few types of construction. The child was evidently about two years grammatically delayed. My session included a basic grammar tutorial about clause and phrase structure. The data from the child was analysed and located at the appropriate points. It was possible to see a pattern: where she was. It was also possible to see the deficit: where she ought to be (for her age). And because the material was organized into developmental stages, it was possible to address the question that was foremost in everyone's mind: how do we get her from where she was to where she ought to be? The principle of following normal development was already well established in paediatric circles, so it was possible to suggest to the speech therapist that the next session should focus on structures which would first consolidate the stage where the child was (introducing verbs, in particular), and then integrating phrase structure within clause elements. A remedial programme was agreed, and therapy began. It is no surprise today, several thousand interventions later, to know that it worked. But at the time, the fact that the therapy was principled, that the child made progress, and that the procedure was generalizable to other patients with language disability (adults as well as children), produced reactions of genuine surprise, delight and relief.

It was grammar that had to be the primary focus of attention, and rightly so. Although the child I had seen evidently had an immature vocabulary and pronunciation, it was her grammatical difficulties which needed the most urgent treatment. This is because grammar is the key to understanding language disability. Vocabulary, because of its huge scope (tens of thousands of words to be learned) is sometimes said to be the primary goal of therapy, and it has certainly taken...