Why is it so hard to reach agreement on terminology? The case of developmental language disorder (DLD).

A recent project entitled CATALISE used the Delphi method to reach a consensus on terminology for unexplained language problems in children. 'Developmental language disorder' (DLD) was the term agreed by a panel of 57 experts. Here I reflect on points of difficulty that arose when attempting to reach a consensus, using qualitative information from comments made by panel members to illustrate the kinds of argument used. One issue of debate was the use of labels, in particular the term 'disorder', which was seen as having both pros and cons. The potential for labels to stigmatize or create low expectations was a particular concern. However, labels could also ensure language problems were not trivialized and could help avoid stigma by providing an explanation for behaviours that might otherwise meet with disapproval. Further debate surrounded issues of how best to identify cases of disorder. Although it was agreed there should be a focus on cases with a poor prognosis, it was recognized that our knowledge of factors related to prognosis was still incomplete. Furthermore, there was a tension between use of standardized tests, which allow for a relatively objective and reliable assessment of language, and more qualitative observations, which can capture functional aspects of communication that are not always picked up on formal assessment. Debate also surrounded the issue of the relationship between DLD and other conditions. Some favoured drawing a distinction between DLD and language disorders associated with other conditions, and others regarded such distinctions as unnecessary. We concluded that it was misleading to assume co-occurring conditions were causes of language disorder, but it was helpful to distinguish DLD from cases of language disorder associated with 'differentiating conditions' that had a known or likely biomedical origin, including brain injury, sensorineural hearing loss, genetic syndromes, intellectual disability and autism spectrum disorder. Furthermore, DLD could co-occur with milder neurodevelopmental disorders that did not have a clear biomedical aetiology. Normal-range non-verbal IQ has traditionally been incorporated in the diagnosis of DLD, but this was rejected as unsupported by evidence. DLD is a category that has utility in identifying children who would benefit from speech-language therapy services, but it should not be thought of as a well-defined condition. DLD has a multifactorial aetiology, is heterogeneous in terms of language features and overlaps with other neurodevelopmental disorders. Our notions of DLD are likely to be refined by further research into aetiology, associated characteristics and intervention effectiveness.