(1) In this article I discuss the notion stability of language elements. Elements easily tranferred from LI to L2 are stable, elements easily borrowed from L2 by LI are unstable. To a great extend the degree of stability is determined by the degree of consciousness resulting from the process of first language acquisition. Secondary factors may be important during second language acquisition, e.g. the degree of deviation of LI elements from L2 norms. (2) The hypotheses about stability were tested by means of investigations in Borne (Twente) in the period 1985-1987.42 dialect speakers were interviewed in order to investigate to what extent they are aware of Twentisms in Standard Dutch. An acceptability test was an important part of the interview. (3) Later, other groups were also tested by means of the same acceptablility test: non-dialect speakers in Twente, immigrants from the western parts of the Netherlands in Twente, immigrants from Twente in the western parts, speakers born and brought up in the western parts. (4,5) In conformity with our expectations, the dialect speakers of Twente are most tolerant towards Twentisms, the western speakers least tolerant. The non-dialect speakers of Twente are rather tolerant but to a lesser degree than the dialect speakers. The two immigrant groups are characterized by about the same in-between scores. The reactions of the non-dialect speakers and of the immigrant groups show that the language environment in which one lives is of great impoitance. Use of constructions and function words (syntax) turn out to be very stable, content words turn out to be very unstable. Word orders (another syntactic domain) and adverbs of time and place occupy intermediate positions. The results concerning morphology and the lexical phonological aspect are problematic probably because of complications of interview technique. (5.3) In a fïnal section special attention is paid to the stability of syntax not only in contact situations but also in cases of internal change. [ABSTRACT FROM AUTHOR]