📱 Emoji in a 🇫 🇷 SMS Corpus 😀 💓 😉 😜 😘 😳 😍 ☺ 😔 ❤ - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

📱 Emoji in a 🇫 🇷 SMS Corpus 😀 💓 😉 😜 😘 😳 😍 ☺ 😔 ❤

Rachel Panckhurst
  • Fonction : Auteur
  • PersonId : 846670
Francesca Frontini

Résumé

A pluridisciplinary team of linguists and computer scientists collected more than 88,000 French authentic SMS in Montpellier (http://www.sud4science.org/), within the context of a vast international project (http://www.sms4science.org/), coordinated by the CENTAL (UCL, Belgium). After semi-automatic treatment using natural language processing (NLP) techniques, the anonymized text messages were made available to the research community and the general public as a corpus, entitled 88milSMS (http://88milsms.huma-num.fr, TEI/XML version: https://hdl.handle.net/11403/comere/cmr-88milsms) (Panckhurst et al. 2014, 2016a, 2016b). In this article, we examine the usage of emoji (e 絵, “picture”; moji 文字, “character”) in 88milSMS. We have observed three main situations (partly inspired by Marcoccia & Gauducheau, 2007, Wilson, 1993): 1.redundant addition: an emoji is used as well as written text, but it is not required in order to understand the SMS (within the context of the exchange); 2.necessary addition: an emoji is used as well as written text, and its inclusion is necessary in order to avoid misinterpretation; 3.lexical replacement: an emoji is used instead of a word. Emoji are more often used in addition (redundantly or necessarily, sometimes as ‘softeners’ (adoucisseurs, Détrie & Verine, 2014)) than for lexical replacement, denoting a reference/referential function (Referenzfunktion, Dürscheid & Siever, 2017). Emoji positioning appears in descending order: at the end, somewhere within the SMS and finally at the onset. Part-of-speech tagging (POS) and NLP software are used to annotate SMS containing emojis in order to investigate the grammatical environment. This allows more in-depth analysis of linguistic functions and to check sociolinguistic factors. In future research, it will be essential to diachronically compare our results with more recent data, emanating from: Swiss Whatsup (http://www.whatsup-switzerland.ch/index.php/fr/) (Ueberwasser & Stark, 2017), Belgian Vos Pouces (http://www.vospouces.org/) (Cougnon et al., 2016) projects. Intercultural variation, crossplatform differences in graphical rendering (Miller et al., 2016) and ‘emoji semiotics’ (Danesi, 2017) will also be explored. This will enable one to visualize evolving mediated digital discourse several years down the track. Keywords: French SMS, emoji, 88milSMS, mediated digital discourse (MDD).
Fichier non déposé

Dates et versions

hal-01651897 , version 1 (29-11-2017)

Identifiants

  • HAL Id : hal-01651897 , version 1

Citer

Rachel Panckhurst, Francesca Frontini. 📱 Emoji in a 🇫 🇷 SMS Corpus 😀 💓 😉 😜 😘 😳 😍 ☺ 😔 ❤. Visualizing in the New Media (VINM), Nov 2017, Neuchâtel, Switzerland. ⟨hal-01651897⟩
314 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More