The generalisation of linguistic constructions is performed through analogical reasoning. Children with developmental language disorders (DLD) are impaired in analogical reasoning and in generalisation. However, these processes are improved by an input involving variability and similarity. Here we investigated the performance of children with or without DLD in a construction generalisation task. We also compared their performance following training with an input involving progressive alignment (combining similarity and variability) or high variability. Progressive alignment improves construction generalisation in children with or without DLD, which could have implications for our understanding of language development and for interventions conducted with children with DLD.

The Picture Exchange Communication System (PECS) is a widely used intervention strategy designed to teach communication skills to children with developmental delays, including autism. The Picture Exchange Communication System incorporates the teaching of mand initiations that are thought to be pivotal response behaviours, and have been demonstrated to lead to generalised improvements in other nontargeted behaviours. The aim of the present study was to assess the acquisition of PECS with a 3-year-old boy with autism using the established PECS training program, and to evaluate concomitant changes in spoken language, social–communicative behaviours, and functional play. Results indicated that the participant rapidly acquired the criterion behaviours for Phases 1 to 3 of the PECS program. Although PECS exchanges were rarely observed in the generalisation settings, clear increases were evident in verbal mands and other initiations in both home and kindergarten generalisation settings. Increases in spoken vocabulary and in the length of comprehensible spoken utterances in free-play were observed as were gains in time spent in developmentally appropriate play. Implications of these results and directions for future research are discussed.

The limited communicative abilities of young children with autism and developmental disabilities may be enhanced by augmentative communication systems such as the Picture Exchange Communication System (PECS). Both children and adults can learn to use PECS, but research is inconsistent in establishing the degree to which PECS training transfers reliably from the training setting to other settings, for example, regular classrooms, and home. Three 5–6 year-old children, diagnosed as autistic and/or developmentally disabled, were given 1:1 PECS training at school, structured to enhance generalisation and transfer. Concurrent observations were made in their regular classroom and at home to probe the extent of generalisation. All three acquired the ability to request a preferred object or activity by exchanging a picture symbol with a communicative partner. In tests of transfer and generalisation, a multiple-baseline across subjects design showed no use of PECS symbols either in the classroom or at home prior to training, with spontaneous initiations of picture exchange (transfer) occurring to a slight to moderate degree in the classroom and to a slight degree at home. Even with PECS training structured to enhance transfer, functionally significant degrees of generalisation of training may be difficult to achieve, and cannot be assumed to occur.

This article is concerned with the status of generalisation in interpretive sociology. The case made is that generalisation is inevitable, desirable and possible. It is held that interpretivism must employ a special kind of generalisation, characterised here as moderatum. However, an acknowledgement that such generalisations can be made must bring us to specify the limits of generalisation in interpretive research. These limits are the limits of interpretivism itself and the paper concludes that this implies the adoption of methodological pluralism in order to realise the full potential of the method.