Heine, Bernd. 2002. On the role of context in grammaticalization. In New Reflections on Grammaticalization. Edited by Ilse Wischer & Gabriele Diewald. Typological Studies in Language 49. Amsterdam; Philadelphia, Pa.: Benjamins. 83-101.
Hopper, Paul J. 1987. Emergent Grammar. In Berkeley Linguistic Society. Proceedings of the Thirteenth Annual Meeting, February 14-16, 1987: General Session and Parasession on Grammar and Cognition. Edited by Jon Aske, Natasha Beery, Laura Michaelis, Hana Filip. 13. Berkeley: Berkeley Linguistic Society. 139-57.
Kuhle, Anneliese. 2019. Tool Intelligence as an Explanation of Cross-Linguistic Variation and Family Resemblance: An Evolutionary and Typological Investigation. Lanham, My.: Lexington.
von Mengden, Ferdinand & Anneliese Kuhle. 2020. Recontextualization and language change. Folia Linguistica Historica 41: 253-81.
Traugott, Elizabeth C. & Robert B. Dasher. 2002. Regularity in Semantic Change. Cambridge Studies in Linguistics 96. Cambridge: CUP.

In this paper, I argue for an emergent view on language and language change as sketched by Hopper 1987. In contrast to structuralist tenets, which see language as a pre-established system that exists prior to usage, Emergent Grammar implies that the linguistic system “is always deferred, always in a process but never arriving, and therefore emergent” (Hopper 1987: 141).

While recent usage-based approaches to language change have taken the variability and the dynamic character of language into consideration, they have remained structuralist in spirit by still seeing language change as a transition between default stages (‘A becomes B, with a transitory period in which A and B coexist’). Concepts like ‘bridging contexts’ (Heine 2002) or ‘invited inferences’ (Traugott/Dasher 2002) suggest that, when a linguistic form changes its function or meaning this requires contexts in which both old and new function constitute part of the interpretation of an utterance. Accordingly, English since, usually encodes causality on the basis of a temporal relation on the propositional level.

Although an advantage over the structuralist approach, in which change is simply seen as a difference between an earlier and a later “stage”, this view takes contextual contributions as something exceptional that accounts for a transition between two defaults. It doesn’t account for the fact (among other things) that attestations of since which are unambiguously either exclusively temporal or exclusively causal are extremely rare. I would therefore like to go a step further and argue that the linguistic sign is inherently negotiable, underspecified and variable. Rather than striving for logical clarity, interlocutors generally handle ambiguities through clues provided by the respective communicative setting. Language change, then, doesn’t require innovation but ‘recontextualization’ – that is, the use of an existing sign / construction in a different context (rather than the use of a new or altered sign). I will demonstrate that canonical types of language change (e.g. the grammaticalization / reanalysis in I’m going to Vienna > I’m gonna like Vienna) don’t require any innovative behaviour on part of a speaker, but reflect the use of one and the same construction being constantly recontextualized. A beneficial theoretical side effect of this claim is that the notion of ‘recontextualization’ is well-compatible with other systems that have been described as ‘emergent’ in various fields outside linguistics, e.g. primate behavior studies (‘flexible transfer’) or cultural anthropology (‘bricolage’) (Kuhle 2019; von Mengden & Kuhle 2020) and is also compatible with more recent approaches in sociolinguistics that focus on small-scale practices rather than on ‘languages’.

The theoretical assumption that language doesn’t exist outside the specific context of a conversation entails that the contextual and interactional conditions of each conversation are essential for (rather than external to) the linguistic sign. Rather than speaking of an impact of context on language (change), ‘recontextualization’ suggests a symbiotic relationship between the sign and the environment of usage. Accordingly context is a necessary ingredient of language which allows for communication with inherently vague, variable and ambiguous signs.