The Crossroads project was a collaborative and interdisciplinary research project investigating small scale multilingualism in a shared cultural space in Southern Senegal. The Crossroads area forms a triangle of three villages tied by dense social networks, family relationships and shared languages, yet every local setting has its own distinctive language ecology.

Rather than switching to only one or two regional or national languages of wider communication, people at the Crossroads are maintaining multiple languages in their repertoires, and have done so for centuries, if not millennia. It is not unusual for an individual to speak anything upwards of five languages, many of which are small, village-based languages that are not used as linguae francae. We approached this type of multilingualism from two angles: the social and the lexical, structural and cognitive.

Governed by the assumption that the maintenance of this type of multilingualism must be motivated by social, political and religious factors and serve to express particular aspects of identity in different contexts, the project investigated the ideologies that nurture it.

In addition, we looked at how the social networks and communities of practice of the very mobile speakers of these languages shape and change their repertoires. Our third research focus explored the impact of social practice on the structural side of language – where convergences between languages occur (or don’t), to what extent the languages are fully reified and to what extent they keep re-emerging in and from multilingual discourse.

Our research has crucially advanced the empirical, theoretical and methodological basis for the study of small-scale multilingual settings. We pioneered new epistemological approaches for the study of settings that, because of their great fluidity and variability, cannot be described in terms of the interaction of language pairs but require in-depth knowledge of their full historically grown language ecologies. We developed a number of new sociolinguistic concepts such as patrimonial language and patrimonial deixis (Lüpke 2018b). Our research connecting sociolinguistic and descriptive and theoretical linguistic perspectives also resulted in a nuanced picture of language contact (Cobbinah 2017, forthcoming d), including gesture (Krajcik 2019), and a novel approach to language categorisation, based on individuals’ repertoires (Watson 2019), and the Because of the relational identities expressed through claiming and speaking languages, we built perspective and scale into our research paradigm (Goodchild 2016, Goodchild forthcoming 2019, Weidl 2019). We have developed the innovative literacy programme LILIEMA (Lüpke et al forthcoming) and plan to develop LILIEMA into an innovative a tool for sociolinguistic data collection that allows to access repertoires beyond local language ideologies. We have created training sessions and an elaborate workflow to collaborate on site and remotely (without internet access in Senegal) in the collection, annotation and analysis of highly multilingual data and intend to publish on these methodologies. Our corpus constitutes a unique data set that we only started to mine, and we plan to seek collaborations with computational linguists and language technologists to visualise convergences and divergences in linguistic forms and to indicate their distance from the empirically determined language prototypes.

The physical focal point of our research in Casamance was our field base in Brin, where our local transcriber team worked with us on making sense of the multilingual recordings we collected. There, we came together as a team with members of the village communities for shared meals, workshops, training sessions, and, not to forget, parties.