the LINC Network

Keystroke Language (and Text) Production: perspectives from cognitive and functional linguistics

23-24 May 2011 - revised dates

Cardiff University, Cardiff, Wales

Supported by the Centre for Language and Communication Research and the LinC Research Network at Cardiff University

The use of keystroke logging as a methodology in language research is not a new field of study since the first Computer Keystroke Logging conference was held at Umeå University in Sweden in 2002. However to date this area of research has primarily focussed on written composition and translation studies. The KEY 2011 workshop and conference intends to broaden this perspective by extending the contributions keystroke logging can make to language production generally, including spontaneous language such as chat messaging. Its theme is to explore functional and cognitive perspectives on the use of keystroke logging in language research where the focus of interest is on the dynamic process of production rather than on the static product of language production.

Keynote Speakers

Professor Kristyan Spelman Miller (University of Winchester)
Dr Mick O’Donnell (Universidad Autónoma de Madrid and Wagsoft Software)

Call for papers

Papers are invited on the general theme of the conference, dealing with the use of keystroke logging in linguistic and language-related research. Presentations will be 20 minutes with 10 minutes discussion time. Papers which present work in progress or that focus on software development and methodology are also welcome. Although we will consider all contributions that relate to the main theme in general, we especially encourage papers that explore:

• Evidence of cognitive processing in electronic language production
• Corpora and the study of electronically produced language
• Language or text as dynamic process (rather than static product)
• Human-Computer Interaction as related to language and keyboard competence
• Linguistic competence (including translation competence)
• Descriptive work that enhances our understanding of electronically produced language
• Functional accounts of language production (including manual and cognitive errors)
• Methodological and/or ethical issues in the use of keystroke logging software

Abstracts

An abstract of approximately 400 words should be submitted electronically at the following webpage: http://linguistlist.org/confcustom/KEY2011. Please state, where appropriate, research questions, approach, method, data and (expected) results. Abstracts will be refereed anonymously by members of the programme committee.
The deadline for submissions is 16 March 2011.
Notification will be sent to authors by 4 April 2011.

KEY2011 website: http://www.cf.ac.uk/encap/newsandevents/events/conferences/key2011/
email: linc-network@cf.ac.uk

Keystroke Language (and Text) Production: perspectives from cognitive and functional linguistics

23-24 May 2011

Cardiff University, Cardiff, Wales

Supported by the Centre for Language and Communication Research and the LinC Research Network at Cardiff University

The use of keystroke logging as a methodology in language research is not a new field of study since the first Computer Keystroke Logging conference was held at Umeå University in Sweden in 2002. However to date this area of research has primarily focussed on written composition and translation studies. The KEY 2011 workshop and conference intends to broaden this perspective by extending the contributions keystroke logging can make to language production generally, including spontaneous language such as chat messaging. Its theme is to explore functional and cognitive perspectives on the use of keystroke logging in language research where the focus of interest is on the dynamic process of production rather than on the static product of language production.

Keynote Speakers

Professor Kristyan Spelman Miller (University of Winchester)
Dr Mick O’Donnell (Universidad Autónoma de Madrid and Wagsoft Software)

Call for papers

Papers are invited on the general theme of the conference, dealing with the use of keystroke logging in linguistic and language-related research. Presentations will be 20 minutes with 10 minutes discussion time. Papers which present work in progress or that focus on software development and methodology are also welcome. Although we will consider all contributions that relate to the main theme in general, we especially encourage papers that explore:

• Evidence of cognitive processing in electronic language production
• Corpora and the study of electronically produced language
• Language or text as dynamic process (rather than static product)
• Human-Computer Interaction as related to language and keyboard competence
• Linguistic competence (including translation competence)
• Descriptive work that enhances our understanding of electronically produced language
• Functional accounts of language production (including manual and cognitive errors)
• Methodological and/or ethical issues in the use of keystroke logging software

Abstracts

An abstract of approximately 400 words should be submitted electronically at the following webpage: http://linguistlist.org/confcustom/KEY2011. Please state, where appropriate, research questions, approach, method, data and (expected) results. Abstracts will be refereed anonymously by members of the programme committee.
The deadline for submissions is 16 March 2011.
Notification will be sent to authors by 4 April 2011.

KEY2011 website: http://www.cf.ac.uk/encap/newsandevents/events/conferences/key2011/
email: linc-network@cf.ac.uk

I have been wanting to do research looking at actual keystrokes in spontaneous language production for several years. I was blocked by IT security issues until last year when I discovered InputLog, a tool developed by researchers for researchers that that logs keyboard and mouse input. Over this academic year (2009 - 2010), I have been collecting data with the very helpful participation of some of our students who have been willing to let me record their keystrokes during chats. I have been collaborating with my colleague Dr Michelle Aldridge on the analysis and we will be presenting our initial results at the 3rd UK Cognitive Linguistics Conference in July, 2010. The paper is entitled Light a tote bag: Insights into electronic language production through keystroke logging (see abstract here).

Recently I discovered that Dr Mick O'Donnell was working on using keystroke logging in his own research and after some discussion we decided to collaborate. I applied for funding through the Cardiff University International Collaboration Award for early stage researchers to fund travel so that we could collaborate face to face. Luckily I was successful so we should be in a position to move this research forward through the next academic year (2010 - 2011). We will be exploring the ways in which we can study dynamic text using keystroke logging software and Mick's UAM CorpusTool.

I'm also working on a larger proposal - something that will continue over an extended period. The long-term project will hopefully include more people who are interested and willing to work on this. Ideally we will get a small team working together on setting up the full version of the project. It'll be great to see my ideas finally getting tested.

It's very exciting - watch this space!

Insights into electronic language production through keystroke logging

This paper reports on one aspect of our Keystroke Project, namely, repairs and pauses in spontaneous, personal electronic language production found, for example, in e-mails, chatrooms and Facebook. The data used in this study were collected through the computer logging software ‘Inputlog’ (Leijten and Van Waes, 2006).

Keystroke logging is a methodological tool which “offers the opportunity to capture details of the activity of writing” (Spelman Miller and Sullivan, 2006:1). For example, Van Waes and Schellens (2003) investigated writing profiles and the effect of the writing mode on pausing and revision patterns of experienced writers and Miller (2005) used keystroke logging to study second language writing processes. To our knowledge, no work, to date, has explored the use of keystroke logging in spontaneous electronic language production and our aim is to fill this research gap.

We will extend Lindgren’s (2005) claim that keystroke logging closely registers traces of cognitive activity during on-line writing because the more spontaneous the language produced, the closer it is to revealing typical human processing. By analysing the spontaneous occurrences of repairs and pauses in data from Facebook and other personal, spontaneous CMC we will offer a description of electronic language production. We identify where pauses and linguistic repairs (lexical and grammatical) occur; we consider whether there is a distinction between the categories of function and content words; and classify spellings in relation to non-standard spellings and ‘ignored’ spellings. The following examples are typically found in our data

1. Your personal bag can be fairly big, light a tote bag. (‘light’ →’ like’)
2. To discuss what we can over (‘over’ → ‘offer’)
3. You had no write to write in his passport (‘write’ → ‘right’)

Through recording and analysing keystrokes we offer innovative insights into the connection between language and cognition.

References

• Leijten, M.; Van Waes, L. (2006) Inputlog: New Perspectives on the Logging of On-Line Writing. In: K. P. H. Sullivan; E. Lindgren (eds.) Computer Key-Stroke Logging and Writing: Methods and Applications. Oxford: Elsevier, 73-94
• Lindgren, E. (2005). Writing and revising: didactic and methodological implications of keystroke logging. Umea: Department of Modern Languages, Umea University.
• Miller, K. S. (2005). ‘Second language writing research and pedagogy: a role for computer logging’. Computers and Composition, 22(3), pp 297-317.
• Spelman Miller and Sullivan, (2006) Keystroke Logging: An Introduction. In: K. P. H. Sullivan; E. Lindgren (eds.) Computer Key-Stroke Logging and Writing: Methods and Applications. Oxford: Elsevier, 1-10.
• Van Waes, L and Schellens, P. J. (2003). ‘Writing profiles: the effect of the writing mode on pausing and revision patterns of experienced writers’. Journal of Pragmatics, 35, 829-853.