The study of language acquisition has enjoyed a robust history in recent years, with the advent of developmental psycholinguistics as a separate field, and with much attention to bilingualism and the acquisition of second languages by both children and adults. The loss of language skills by individual speakers has, by contrast, been a little researched area, with the exception of the field of aphasiology, which has developed roughly parallel with modern psycholinguistics. Typical situations in which language skills may be lost occur when an individual speaker of a language moves to an area where another language is dominant; when an ethnolinguistic minority child enters school and adopts the societal language; when a second language is no longer studied or needed; when a local language drops out of use and its speakers must adopt a more dominant language. As they grow older, young children appear to lose some language-related skills, such as the ability to make fine phonetic discriminations (see Burnham, this issue) and aging adults lose some language skills they once had.

Young infants perceive a variety of speech contrasts, even those that are not relevant in their own language environment. Evidence regarding the loss of perceptual ability with phonologically irrelevant contrasts is reviewed and a theoretical distinction is advanced. It is proposed that the perception of “fragile” contrasts is lost in infancy due to lack of exposure to particular sounds, while the perception of “robust” contrasts is lost around the onset of formal language training due to lack of experience with phonologically irrelevant contrasts. This distinction should provide a heuristic framework for future research on the development of speech perception.

In the present paper a linguistic description is given of the process of first language acquisition of Turkish children aged 4 to 8 in a Dutch submersion environment in the Netherlands. On the basis of the assumption that language development involves the acquisition of distinct subskills in differential patternings, the development of lexical, morphosyntactic, and pragmatic abilities have been investigated separately. Furthermore, these longitudinal data are compared with cross-sectional language data of five- and seven-year-old children in Turkey. This comparison was made in order to be able to interpret whether the Turkish language skills of native Turkish-speaking children in the Netherlands were showing delay, stagnation, or attrition of skills. The overall results suggest that in the age range of 4 to 8 years the acquisition of first language skills by Turkish children in the Netherlands can best be characterized as stagnated.

Relative proficiency in Gaelic varies by age in a Highland Scottish region where Gaelic is spoken by a decreasing number of bilinguals in a steadily aging population segment. Although there is a Gaelic proficiency continuum ranging from full fluency (and Gaelic dominance) to minimal generative skills (and English dominance), there are also deviations at both extremes of skill: Fully fluent speakers show self-conscious difficulty with one morphological and four syntactic structures, while minimally proficient speakers show less difficulty than might be expected in the use of conjunctions. An exceptionally keen though low-proficiency speaker can devise unusual means of compensation in using the language. Neither full fluency nor marked deficiency produces altogether predictable levels of structural conservatism or communicative success.

Elderly adults (70 to 89 years) and young adults (30 to 49 years) were asked to imitate complex sentences involving embedded gerunds, wh-clauses, that-clauses, and relative clauses. The young adults were able to imitate accurately or correctly paraphrase the sentences regardless of the length, position, or type of embedded clause. The elderly adults could accurately imitate or paraphrase short constructions. The elderly adults were unable to imitate or paraphrase correctly long constructions, especially those in which the embedded clause was sentence-initial. The pattern of results demonstrates an age-related decline in syntactic processing abilities due, perhaps, to the increased processing demands of the long or sentence-initial constructions.