Elena Pizzuto (Institute of Psychology, National Research Council (C.N.R.), Rome, Italy)
Brita Bergman (University of Stockholm, Institute ofLinguistics, Department of Sign Language, Stockholm, Sweden)
Research conducted in the last thirty years on individual sign languages, along with crosslinguistic investigations of different sign languages, have provided important information on the morphological and morphosyntactic structure of signed as compared to spoken languages, and much has been learned on the extent to which the structure of a linguistic system is dependent upon the specific modality through which language is perceived and produced. For example, it has been clarified that the visual-gestural modality proper of sign languages significantly affects the organization of morphology and syntax according to primarily spatial patterns that are not found in spoken languages, where a primarily temporal organization prevails. The temporal organization of signs within sentences and connected discourse also significantly differs from that found in speech. Due to the fact that in sign languages the two hands (and/or also other parts of the body such as the shoulders, the face, the eyes) can be used as distinct articulators, sign languages allow the simultaneous production (or co-articulation) of distinct lexical or morphological elements to a degree that is not approached in speech.
Despite the advancement of our knowledge of sign language structure, current research still faces a major methodological problem. No sign language to date investigated has autonomously developed an appropriate system of written notation for the analysis and coding of signed texts. The written notation systems currently available are usually considered appropriate for the transcription and analysis of individual signs, but cannot be easily used for transcribing, coding and analysing longer sequences of signs as they occur in spontaneous conversation and discourse. Most research on sign language morphology and syntax is still carried out using what are called "sign glosses", or written labels for the meanings of the signs in, for example, written Italian, English or French. Although enriched with particular notation devices that attempt to capture the particular spatial-temporal features that characterize the morphology and syntax of signs in discourse, these glosses are very often inadequate for appropriate analyses of the morphological alterations of signs, and of the spatial-temporal arrangement of syntactic and discourse patterns. In some cases, the use of such glosses can indeed obscure, rather than clarify, important features of the linguistic structures under investigation.
Recent advancement in multimedia technologies have raised the possibility that new notation systems be devised for coding and analysing signs in discourse. For example, digital video-recording can be, and has been fruitfully extended to recording signs in discourse. Computerized video archives can be constructed, and subsequent coding and analysis can be performed on video images of the signs: the visible morphological and syntactic structure of the signs can be directly inspected, overcoming the constraints formerly imposed by the almost exclusive use of sign glosses.
The Italian workshop on morphosyntax and discourse will be the third in a series of five workshops planned within the Intersign network. Whereas the two previous workshops have focussed on lexical and phonological databases respectively, the present workshop will focus on the above mentioned problems in the creation of databases for, and in the analysis and description of signed texts. The workshop will aim to provide a survey and a critical analysis of text corpora, transcription and notation systems currently developed for the analysis of sign language morphology, syntax and discourse (including multimedia coding systems). Since one of the goals of the network is to develop standards and techniques for storing, coding and analysing sign language, the workshop will also address the question of how to provide data for comparative analyses and discuss future collaboration in this area.
Participants are expected to present/demonstrate and discuss the systems and techniques they are using and to point to different problems they are facing in terms of storing, notation and analysis. Participants are also encouraged to include considerations of possible specifications for a common European database. Participants of the workshop should also be prepared to participate in the work following the workshop such as contributing in the preparation of the final publication of the workshop proceedings.
What follows is a by no means exhaustive list of general as well as more specific issues and questions we propose to address and, hopefully, clarify during the workshop.
A major methodological issue we wish to consider concerns the possible reasons which render so difficult to transcribe signed texts. It seems somewhat paradoxical that current research has developed tools for the notation of individual signs, and yet these notation tools are rarely used, or are very difficult to use, for transcribing sequences of signs as they occur in ordinary discourse. This situation is hard to compare with the one observable in spoken language research. The orthographic or phonetic systems that are used for notating and transcribing spoken languages normally allow the transcription of both the individual words of a language (as they occur in their citation forms out of context), and sequences of words in discourse, with all the morphological alterations they can undergo. Is it possible that the difficulty we face in our notation of signed languages texts is, at least in part, due to inappropriate segmentation and/or analysis of the linguistically relevant elements that compose signs and sign discourse?
Some specific questions that need to be addressed concern the frequent use of sign glosses in the transcription and coding of signed texts. Even though researchers are obviously very cautious when attributing glosses to signs, the extent to which glosses can obscure the morphological features of the signs is often underestimated, and deserves special attention. The methodological problems connected with the use of sign glosses may also have a different weight in different language communities, due to the different morphological structure of the language in which the sign glosses are provided. For example, providing sign glosses in written English, a language in which the morphological distinction between nouns and verbs is for the most left unmarked in the surface forms of the words, does not force a sign language researcher to immediately categorize (via the gloss used) a given sign as a noun or a verb. However, a different situation holds when a researcher is glossing signs in written Italian, a language in which the morphological distinction between nouns and verbs is almost always overtly marked. The possibility of inappropriately attributing the category of verb (or noun) to a sign, even in the lack of independent evidence coming from a more accurate analysis of the sign structure, is much greater when using written Italian glosses. Similar observations could also be made with respect to sign glosses given in languages that have morphological features similar to those found in Italian.
Concerning signed text corpora, a wealth of related questions arise: