Very long‐term retention of a first language without rehearsal: A case study

Abstract

A case study is presented of an American sign language-using deaf woman whose life circumstances were such that she had no opportunities to use her sign language over nearly 40 years, but whose recall of its lexicon and syntax were almost unimpaired when she resumed its use after that period. Implications for the role of memory in first language learning are compared with results for second languages, and the role of rehearsal and interference in remembering over very long periods is considered.In recent years memory researchers have turned increasingly to the study of memory in 'natural' or 'everyday' contexts to supplement traditional restricted and artificial laboratory studies (Neisser, 1982;Bahrick and Karis, 1982;Harris and Morris, 1984; Johnston and Hasher, 1987). Recent studies have been not only of remembering and forgetting in natural contexts, but have also been of these phenomena over very long periods (up to, for example, 50 years in Bahrick's (1984a) study of recall of Spanish learned in school and university and Bahrick and Hall's (1991) of school and university mathematics). Bahrick (1984a) has shown that recall of Spanish learned in school and university as a second language shows a discontinuous function; with different degrees of forgetting taking place in the first 3-6 years after learning, and then virtually no loss of material for at least 25 years thereafter. He therefore posits the existence of a 'permastore' in memory, particularly for 'semantic' materials (of which language is a prime example). The extent to which learned material is entered into permastore varies with level of original training, the original grade achieved, and the nature of the response required at later testing, with those trained to higher levels and obtaining higher original grades retaining more. Bahrick was also able to show that amount of retention was unrelated to opportunities for rehearsal and he concluded that 'very significant portions of semantic knowledge remain perfectly accessible for decades without being used at all' (p. 23). He is also of the view that interference from later learning is not particularly important in determining level of forgetting, but that the major factor at work is the level of original learning, particularly for semantic material which typically is learned over long periods so that 'extension of the re-exposure to information over long time periods produces a cumulative 0888-4080/93/030229-09/$09.50