Presenting at Conferences
Attending and presenting at conferences are an essential part of academic careers and the dissimination of our results. Below, we've summarized some thoughts on what role conferences can or should play in the planning of an academic career. Further down, you'll also find a list of regular conferences that are of interest to the lab.
1) Conferences vs. Journal papers: At the end, it's (journal) papers that count. For linguistics, I'd say about three conferences with proceedings are like a journal paper. For computational linguistics, prestigious conference proceedings (ACL, EMNLP, NIPS, and, to a lesser extent, COLING, NAACL) are basically the same as a journal paper. But for cognitive science, conference proceedings (a) are rarely cited and (b) too many of them might raise questions as to why there is no follow up article.
Conferences are a great opportunity to get feedback from outside of your bee hive, to meet new collaborators, and to mingle with folks. They are also a great way to get to know and introduce yourself to a new field of research. But conferences are also distractions - you need to write abstracts and prepare the presentation and that always takes more time than you think. Conferences can easily become a cheap buzz, due to the immediacy of the feedback (which, however, is also shortlived). Also, think about whether your project is better suited for direct journal publication. Any thing your write consumes time. Even if you look at short paper abstracts (CogSci, NIPS), and think that they are a great way to start your paper, keep in mind that a good paper is self-contained. So, there will be extra work involved in making the short paper a good stand-alone paper. Whether it makes sense to submit such a paper rather than to go straight for a journal paper depends on (a) the proportion of time that will be put into making that short-paper stand alone (i.e. time that will not contribute to your overall project write-up), (b) how important this conference is to you -- including whether it would be a good idea to get some "visibility" on the job market, (c) maybe whether you fear that someone else may talk about this before you do, (d) how awesome the conference location is ;).
2) Plan ahead: Be aware of the major annual conferences in your subfield (see below). Be part of the relevant email lists, announcing deadlines (CUNY's sentproc, the AMLaP list, and Linguist List are a good start; for some of you funknet, lingtype, or the international corpus list might be a good idea, too). Most regular meetings have their deadlines around the same time every year. Put a reminder into your calendar a few weeks before that time to check whether the deadline has been posted (see also the CLS calendar, which has many of these deadlines).
3) Remind your collaborators and advisors and start early: Please keep in mind that your collaborators and advisors may need time to look over your abstract presentations. As a rule of thumb, remind them at least a month ahead. Keep in mind that they may have other commitments. For myself, for short 1-2 page abstracts, having a complete draft about 5 days in advance of the deadline is usually ok. But if many people in the lab attend the conference (CUNY, AMLaP, LSA), then please give me 7-10 days at least. For short paper abstracts of 6-9 pages, a complete draft should typically be ready about 2 weeks in advance, preferably more. See also our notes on abstract writing.
4) How many conferences should I plan to present at? As a rule of thumb, I'd say present at least one conference a year during your graduate years, perhaps two, but not more than three (with the possible exception of year 4 or 5). In planning ahead, keep in mind that there are workshops in addition to the regular meetings. These workshops may fit your interests very well. So don't book your calendar full of conferences.
Regular Meetings and Conferences of Interest to the Lab
Psychonomics (Annual meeting)
- Deadline: December
- Conference: late March/beginning of April
- Summary: This has been our main conference so far, but that may change. Accepts mostly work on sentence processing and is sometimes somewhat narrowly defined. For example, it's hard to get talks with phonological or speech topics. We had reasonable success of computational topics, though occasionally the reviews suggest lack of familiarity with computational approaches. That being said, I love going to CUNY: great people, usually well-organized, not too big, and so it always ends up being a productive and fun time.
Architectures and Mechanisms for Language Processing (AMLaP) (AMLaP)
- Deadline: May
- Conference: Early September, Barcelona
- Presentation types: same as CUNY
Summary: Same audience as CUNY, but more folks from Europe (since it takes place in Europe). Maybe traditionally more accepting of modeling work. Also less limited to sentence processing.
International Workshop on Language Production (IWLP)
- Conference: 2010 (probably on a bi-annual schedule now)
- Presentation types: invited talks, posters
- Summary: Relatively small gathering (~60 people?) focused on language production. Talks are by invitation only, with a restriction that no speaker will be invited more than once. Talks are ~45 minutes, with ~15 minutes for questions. Discussion is heavily encouraged, and there are generally several panel-type sessions.
- Publications: Work presented here is usually collect for a special issue of Language and Cognitive Processes.
Experimental Pragmatics (XPrag 2009)
- Conference: 23-25 April, Lyon, France (biannual schedule)
- Presentation types: oral (25+5 min, ca. 20% chance), posters (ca. 40% chance), invited talks, commentaries on sessions (by invitation only)
- Summary: Fairly small (~70) but growing group of experimental pragmaticists who gather to talk about the processing of scalar implicatures, presuppositions, perspective-taking, communicative intentions, information structure, metaphor, ad hoc concepts,... Methods are behavioral, psycho-/neurolinguistic. No modeling (yet).
- Publications: Talks get a proceedings paper.
2. Cognitive Science
Neural Information Processing Systems (NIPS)
- Deadline: Early May
- Conference: Early December, Vancouver
Presentation types: oral (< 2% chance), spot-light poster (2min; < 5%), poster
- Summary: Annual conference on machine learning on topics related to neural or cognitive themes, but actually pretty much hard core machine learning. About 1,000+ people from CS, engineering, cogsci, computational linguistics, neuroscience. Both theoretical and method related work is accepted as long as it meets the (high!) formal standards.
Society for Mathematical Psychology annual meeting (SMP)
- Deadline: April 3, 2009.
Conference: Workshop on Bayesian Models for Cognitive Science on morning of August 1 2009. Conference August 1-4 2009 at the University of Amsterdam. Designed to be compatible with CogSci.
Annual Interdisciplinary Conference (AIC)
- Conference: February 8-13 2009.
- Summary: Best kept secret in Cognitive Science. Great talks, skiing.
3. Language and Cognitive Development
Boston University Conference on Language Development (BUCLD) (http://www.bu.edu/linguistics/APPLIED/BUCLD/)
- Deadline: May 15, 2009.
- Conference: November 6-8, 2009 at Boston University.
- Presentation types: Primarily talks (20 min. each), some posters. 2008 stats: 450 submissions, 30% acceptance rate (87 talks, 66 posters).
- Summary: Major conference for research on first- and second-language acquisition. Primarily parallel sessions. Roughly 500 attendees.
Sixth Biennial Meeting of the Cognitive Development Society (http://cds2009.spc.uchicago.edu/index.php)
- Deadline: June 1, 2009.
Conference: October 16-17, 2009 in San Antonio, Texas. 2009 Plenary speakers: Michael Tomasello (MPI) & Josh Tenenbaum (MIT).
- Presentation types: Symposia/workshops (1 hr, 45 min each) and posters.
- Summary: Conference for research on change and continuity in development (including but not exclusively pertaining to language development). Submitted symposia/workshops/poster sessions presented in parallel sessions. Roughly 400 attendees.
International Conference on Infant Studies (ICIS) 2010 (http://www.isisweb.org/)
- Deadline: TBA.
- Conference: March 10-14, 2010 in Baltimore, Maryland.
- Presentation types: Talks and posters. Submission format TBA.
- Summary: Conference for research on the development of sensation, perception, cognition, emotion, language, and social behavior. Conference includes invited and investigator-initiated symposia, lectures, and poster presentations.
4. Computational Linguistics
ACL: yearly conference on natural language processing; mostly not directly related to work on human language processing, but occasionally formal approaches (e.g. machine learning inspired work) on human language processing are presented at ACL. This is a good venue for presentations of collaborations with CS. It's a pretty prestigious conference.
NAACL: North American version of ACL, held yearly, unless ACL located in North America.
- EACL: same as NAACL, but for Europe; held yearly, unless ACL located in Europe.
- EMNLP: yearly
- COLING: every other year
ACLWeb (external CL resource page)
LREC: every other year. Presentations are mostly concerned with new language resources. From our lab, Carlos Gomez Gallo has submitted. Several people in CS occasionally submit.
Joel Tetreault maintains a calendar with call for papers deadlines and conferences dates for a lot of the computational linguistic conferences. He does not include some of the dialogue specific ones such as:
- SIGDIAL : Every year. It is collocated with either ACL or Interspeech.
- SEMDIAL : Every year. Dedicated to dialogue and semantic work. Its name changes depending on the location where it is being held. In 2008 it was LONDIAL, since it took place in London. (cute).
Linguistics Society of America (LSA) (LSA)
- Deadline: Beginning September (need to be member for submission; can become member along with submission)
- Conference: January 7th-10th 2010 in Baltimore, MD.
- Presentation types: oral (20+10), poster
- Summary: Annual linguistics meeting. About 1,000+ linguists from all over the world. Many (6ish) parallel sessions about all types of linguistic work. Sometimes hard to attract enough people with processing or computational interest, but it's getting better. Good for overview talks, and to get to know people (if you're up for the social stress). We usually have some presentations there, and I (Florian) try to attend every other year.
Conference on Quantitative Social Science Research Using R (QR 2009)
- Deadline: rolling, May 15 final submission date. Acceptance notification within 2 weeks of submission.
- Conference: June 18-19, 2009 in Manhattan (hosted by Fordham University).
- Presentation types: Platform session talks, parallel session talks, posters
- Summary: This conference is about applications of the R software and graphics system to important policy and research problems. It provides an excellent opportunity to bring together researchers from various disciplines using R in their work.
- Publications: All accepted papers have to be accompanied by R code and data. Code and data will be made available on CD-ROM. Full papers will be published by Springer in the UseR! series.
Joint Statistical Meetings (JSM)