The purpose of this paper is two-fold: (1) contribute to improve the representation tools and methodologies for the study of signed languages (SL); (2) provide new evidence on the structure of narrative texts in Italian Sign Language (LIS), focusing on features that are widely found across SL but which are not observed in vocal/written languages (VL), and thus seem to be unique of SL. A novelty of the present study is that we extend the exploration of these features to emergent forms of written LIS. In the first part of the paper we briefly recall and discuss a major theoretical-methodological problem highlighted in several studies (e.g. Brennan, 2001; Cuxac, 2000; Pizzuto et al., 2006; Leeson et al., 2006; Vermeerbergen, 2006): the absence of appropriate means for representing SL in written form, specifically for producing easily readable transcriptions that, aside from specialist linguistic annotation, allow anyone who knows the SL under investigation to reconstruct its forms, and form-meaning correspondences. Drawing on previous work we show how, in transcribing SL productions, the widespread use of VL labels (so-called ‘glosses’) as a primary means for ‘representing’ signs grossly misrepresent the structure of SL. We illustrate the major findings from an ongoing project in which, with the crucial contribution of experienced deaf signers and researchers, we are experimenting the use of Sutton’s (1999) SignWriting (SW) system as a tool for: (a) transcribing LIS face-to-face discourse; (b) composing LIS texts conceived directly in written form. We exemplify data stemming from a comparison of the same LIS productions represented and analyzed using word-labels vs. a SW-encoded form. We show how, using SW-encoded transcripts, LIS signers can easily represent, reconstruct and analyze the form-meaning patterns of their language with an accuracy never experienced with other annotation systems. In contrast, the use of VL word labels induces severe distortions (Di Renzo, 2008). Our results suggest that, even though more research is needed, SW can be a most valuable tool en route towards appropriate transcription for SL. We also find that SW-encoded representations significantly facilitate deaf signers’ metalinguistic reflections on the properties of their language, while this is not the case with annotation systems where the forms of the language are not readily accessible. In the second part of the paper, we describe a study of LIS narrative texts aimed at clarifying lexical and grammatical properties that appear to be unique of the visual-gestural modality. Towards this end, we analyse three narrative texts elicited from three deaf signers (all highly competent in LIS) via the presentation of Chafe’s (1980) “Pear Story” film. The novelty of our study is that, building on previous experience with spontaneous, short written LIS compositions (Di Renzo et al., 2006), the texts analysed include: (1) one signed, face-to-face rendition of the Pear story, subsequently transcribed with the help of the SW system; (2) two texts conceived and expressed directly in written LIS, also encoded via SW ‘orthography’. It must be noted that ‘written LIS’ is still in an “emergent” stage. We do not know whether this form of the language will/will not eventually expand, outside of our research group, in the Italian signing community. Nonetheless, we felt that a comparison between signed/transcribed and written renditions of the same story could provide us special insights for a first exploration of features of SL that may be strongly influenced not only by the visual-gestural channel, but also by the fact that SL, like the vast majority of VL (Ong, 1982) are face-to-face languages, without a written tradition. Our analyses of the signed and written texts focused on the following aspects: (1) The presence and relative distribution of two major types of referential expressions, usually characterized as ‘frozen’ and ‘productive’ lexical and/or grammatical units (see for overviews Brennan, 2001; Vermeerbergen, 2006). Referring to Cuxac’s (2000) theoretical framework, we define these two types of units as respectively: - ‘standard’ signs, i.e. those that are habitually listed in SL dictionaries, and easily translatable in spoken languages words; - Highly Iconic Structures (HIS), i.e. complex sign units that are rarely or never listed in dictionaries, are very frequent in signed discourse, and exhibit not only strong iconic features but also a multi-linear organization that have no parallel in VL. (2) The use of these two major kinds of units, and of different indexical devices (e.g. manual and visual pointing) for deictic and anaphoric reference, i.e. as text cohesion devices (Pizzuto, 2007; Antinoro Pizzuto et al., 2008). (3) How comparable were the forms of ‘standard’ signs found in our texts with those of the ‘same signs’ listed in LIS dictionaries. This analysis concerned a subset of signs for common referents (e.g. for ‘man’, ‘child’, ‘ladder’, ‘tree’ ‘pear’), and aimed at ascertaining context-dependent effects on the stability or variability of standard signs. On the whole, the results of this study evidence major structural similarities, but also interesting differences between the signed and the written texts. Similarities regard especially the presence and distribution of standard signs compared to HIS. Significantly, HIS are well represented not only in the face-to-face, but also in the written texts, and play a major role in carrying out deictic-anaphoric reference operations. Differences concern ‘normalization’ processes detected in the written texts. Context-dependent effects were also found. We conclude discussing the implications of our findings for more comprehensive descriptions of SL narratives and more generally discourse structure. References Antinoro Pizzuto E., Rossini P., Sallandre M-A. & Wilkinson E. (2008). Deixis, anaphora and Highly Iconic Structures: cross-linguistic evidence on American (ASL), French (LSF) and Italian (LIS) Signed Languages. in De Quadros R. (ed.), Sign Languages: spinning and unraveling the past, present and future. TISLR9, Petrópolis RJ (Brazil): Editora Arara Azul. (http://www.editora-arara-azul.com.br/EstudosSurdos.php), 475-495. Brennan M. (2001). Encoding and capturing productive morphology. SL&L, 4 (1/2), 47-62. Chafe W.L. (ed.) (1980). The pear stories: cognitive, cultural and linguistic aspects of narrative production. Nordwood NJ: Ablex. Cuxac C. (2000). La Langue des Signes Française (LSF). Les voies de l’iconicité. Faits de Langues, n. 15-16. Paris: Ophrys. Di Renzo A. (2008). Esperienze e riflessioni su metodi di trascrizione della LIS. in Bagnara C., Corazza S., Fontana S. & Zuccalà A. (eds), I segni parlano. Milano: Franco Angeli, 159-170. Di Renzo A., Lamano L., Lucioli T., Pennacchi B. & Ponzo L. (2006). Italian Sign Language: can we write it and transcribe it with Sign Writing? LREC-2006, Proceedings 2nd Workshop on the Representation and Processing of Sign Languages. Pisa: ILC-CNR, 11-16. Leeson L., Saeed J., Byrne-Dunne D., Macduff A. & Leonard C. (2006). Moving heads and moving hands: developing a digital corpus of Irish Sign Language. The ‘Signs of Ireland’ Corpus Development Project. Information Technology and Telecommunication Conference 2006, Carlow (Ireland), 25-26 October 2006. Ong W.J. (1982). Orality and literacy. The technologizing of the word. London: Methuen. Pizzuto E. (2007). Deixis, anaphora and person reference in signed languages. in Pizzuto E., Pietrandrea P. & Simone R. (eds), Verbal and Signed Languages - Comparing structures, constructs and methodologies. Berlin / New York: Mouton De Gruyter, 275-308. Pizzuto E., Rossini P. & Russo T. (2006). Representing signed languages in written form: questions that need to be posed. LREC-2006, Proceedings 2nd Workshop on the Representation and Processing of Sign Languages. Pisa: ILC-CNR, 1-6. Sutton V. (1999). Lessons in SignWriting. Textbook & workbook. La Jolla CA: Deaf Action Committee for Sign Writing. Vermeerbergen M. (2006). Past and current trends in sign language research. L&C, 26, 168-192.