SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES

Internet usage continues to increase and people spend more time chatting and forming friendships online. Out of this phenomenon, a new way of speaking is emerging. This poster reports on a pilot study that examines non-standard English features in World of Warcraft chat.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES

The Pew Research Center's (2005) report on Internet usage indicates a 37% increase since their 2000 survey. Increased Internet use has led to new mediums for communication as well as a new way to converse. In this paper, I will refer to the language arising out of such communication as “cyberlanguage.” Cyberlanguage (CL) is the conversational language arising out of the use of online media such as chat, instant messaging, text messaging, games, forums, and the like. It is characterized by the refashioning of standard English into abbreviated and often pictographic representations of existing concepts where layers of meaning are packed into a few simple keystrokes. It is, in part, a result of user adaptation to the constraints and affordances imposed by these new media (e.g. character length restrictions, small screen/window size). Information tools must become as facile with CL as they are with standard language.

Researchers (e.g. Baron, 2008; Cherny, 1999; Crystal, 2008; Ling, 2005; and Werry, 1996) have analyzed language used in electronic media and have uncovered a variety of linguistic features which tend to be higher in prominence than in standard English text, These features include abbreviations of all kinds and surrogate face-to-face (FTF) cues; because such cues are missing in online communication. I have collated and merged all features noted by these and other scholars into a single list and, for this pilot study, am investigating some of these features' existence in World of Warcraft (WoW) gaming chat.

WoW is the most popular massively multiplayer online game (MMOG) at this time, with approximately 11.5 million game players according to Blizzard.com. Because of its strong user community and it being relatively uninvestigated, WoW chat is fertile ground for study.

While my long-term research questions include determining the frequency and proportion of all previously-noted CL features across media, in this pilot study, I focus on three orthographic features, asterisks, slashes, and letter duplication, with the goal of investigating their communicative purposes.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES

World of Warcraft chat logs were collected over 73 days between 12/1/08 and 4/5/09. This corpus includes 100,083 lines of text and 24,324 word types. The chat logs for this study were taken from two U.S. servers at various times of day; however, interlocutors may be from any part of the world. Information about interlocutors, such as real name, sex, ethnicity, etc., are not known. Corpus content is primarily concerned with game play. This study was approved by the UNC-CH IRB.

The corpus was cleaned of timestamps and non-conversational text such as game feedback (e.g. Received item: Hellfire Skiver). Then, a script counted all words within the text, excluding chat channel information and interlocutor pseudonym. For example, in this line of chat:

[Guild] Superdude: so i'm pretty much on autopilot

[Guild] indicates the chat channel and Superdude is the interlocutor pseudonym. In this example, only 6 words were counted: so, i'm, pretty, much, on, and autopilot.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES

Out of the 24,324 total word types in the corpus, 805 were found to contain an asterisk (175, 22% of the 805 types), slash (355, 44%), and/or letter duplication (275, 34%). I then examined how the 805 types were used (i.e. what purpose they served in the conversation). Four major purposes were found: surrogate prosodic cues, surrogate proxemic cues, attempts to conserve effort and abbreviate, and discourse management strategies.

Table 1. Frequency/proportion of communicative purposes across three orthographic features. Some types exhibit multiple purposes, yielding 890 total purposes (e.g. =DDDD is both an emoticon and letter duplication)
 ORTHOGRAPHIC FEATURES EXAMPLES
 Letter DuplicationAsterisksSlashes  
N Types (% of 805)275 (34%)175 (22%)355 (44%)805 
PURPOSE OF USEn (% of 275)n (% of 175)n (% of 355)TOTAL PURPOSES (% of 890) 
Surrogate Prosodic Cues   422 (47%) 
Elongation of sounds275 (100%)2 (1%)18 (5%)295 (33%)EAAASSSSSY
Onomatopoeic expression100 (36%)7 (4%)6 (2%)113 (13%)*bloooooooooop*
Offsetting punctuation for emphasis014 (8%)014 (2%)thats just *my* opinion
Surrogate Proxemic Cues   119 (13%) 
Emotes035 (20%)68 (19%)103 (12%)*tickle tickle* or /tapdances
Emoticons4 (1%)3 (2%)9 (3%)16 (2%)=DDDD (emoticon emphasized)
Economy of Effort   188 (21%) 
Conjunctions00109 (31%)109 (12%)geared/skilled (suggesting that you must be both)
Disjunctions0046 (13%)46 (5%)25vault/25os (asking to do one of two dungeons)
Symbolic substitution01 (1%)32 (9%)33 (4%)h/strat (heroic Stratholme, / signifies modification)
Discourse Management   161 (18%) 
Repairs0115 (66%)2 (1%)117 (13%)Line 1: im stuffing my sace atm (typo sace)
Typo/misspelling1 (0%)7 (4%)36 (10%)44 (5%)Line 2: face* (its repair)
TOTAL FEATURES380184326890 

Table 1 shows the frequency and proportion of communicative-purpose features across the three orthographic features. For example, 35 word types containing asterisks (20%) also exhibited emotes.

In all three orthographic features, there were 422 surrogate prosodic cues, 119 surrogate proxemic cues, 188 effort-economizing types, and 161 discourse management features. Based on these results, asterisks, slashes, and letter duplication are most often used as surrogate prosodic cues. Overall, surrogate FTF cues (proxemic and prosodic) account for the majority of 890 communicative purposes (61%). Elongation of sounds is used the most to convey prosody, emotes to convey proxemics, and conjunctions to economize effort.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES

Resulting from creative adaptation to an evolving landscape of communication venues, cyberlanguage exemplifies rapid language change, making its incorporation into language infrastructures difficult. The study of this phenomenon can inform the design of processes such as search, document clustering, intelligence surveillance, and summarization to better match user needs and practices. Limitations of this pilot study are largely concerned with representativeness and generalizability. Future work will expand this investigation to instant messaging, texting, forums, and other chat, and to consider issues of validation.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. CONCLUSION
  7. REFERENCES
  • Baron, N. S. (2008). Always On: Language in an Online and Mobile World. Oxford: Oxford University Press.
  • Cherny, L. (1999). Conversation and Community: Chat in a Virtual World. Stanford, California: CSLI Publications.
  • Crystal, D. (2008). Txting: the gr8 db8. New York: Oxford University Press.
  • Ling, R. (2005). The sociolinguistics of SMS: An analysis of SMS use by a random sample of Norwegians. In R.Ling & P. E.Pedersen (Eds.), Mobile communications: Renegotiation of the social sphere (pp. 335349). London: Springer-Verlag London Limited.
  • Pew Research Center. (2005). Internet: The Mainstreaming of Online Life. In Trends 2005 (pp. 5669). Washington D.C.: Pew Research Center. Retrieved from http://pewresearch.org/pubs/206/trends-2005.
  • Werry, C. (1996). Linguistic and interactional features of Internet Relay Chat. In S. C.Herring (Ed.), Computer-mediated communication: Linguistic, social and cross-cultural perspectives (pp. 4763). Amsterdam: J. Benjamins.