This paper describes work on an ongoing learner corpus in Swedish Sign Language (SSL) as a second language (L2). The purpose of this learner corpus is to provide a solid database for second language research in SSL, as there is a lack of research regarding how adults learn a signed language as a second language, and the availability of such a corpus for research would ultimately lead to new insights in the field. Work on this SSL learner corpus started in 2013 (Schönström & Mesch, 2014), and it now contains longitudinal data collected from 2013 to 2016. The corpus consists of data from two groups of learners. Data collection for the first group was completed in 2014 and contains 9:06 hours of data from a total of 18 learners. Data collection from the second group is ongoing.

The longitudinal data collection consisted of interviews as well as picture and video retellings recorded on four occasions over a period of 1.5 years. The learners consisted of students from a sign language interpreter program at university level. The first collection began one month after course onset, and the second one 1.5 years after onset. The aim was to obtain a wider range of data illustrating the learners’ different developmental stages. The recorded material has been annotated and transcribed in the multimodal annotation tool ELAN using current SSL annotation conventions, especially for annotation of glosses as well as a special annotation schema for L2 analysis according to our particular research objectives.

For those who are learning SSL, we hypothesize that simultaneous and spatial structures in a gestural-visual modality are challenging to learn (cf. Ortega & Morgan, 2015). Earlier we began analyzing the mouth actions of L2 learners (Mesch, Schönström, Riemer-Kankkonen & Wallin, 2016). Data was annotated according to annotation tiers for mouthing categories, such as mouth movements borrowed from Swedish (mouthing without sound), and mouth gestures, as well as L2 tiers. The next step is to analyze a set of complex sign categories (i.e. signs modified according to meaning and space). We are interested in how learners acquire depicting signs as well as other complex sign categories, i.e. modified signs and indicating signs. This overlaps partly with the use of space for meaning and reference, which is a challenge to annotate. In our presentation, we will show our annotation scheme and discuss the challenges of annotating these structures in an L2 context.