Dear BilingBank,
There is very little traffic on this list, and perhaps that is not altogether a bad thing. However, it is time now for an update on some of the resources in SLABank, BilingBank and TalkBank more generally.
1. We now have seven funded TalkBank projects, each building corpora and tools for the analysis of spoken language interactions. Four are funded by NIH: CHILDES for child language, AphasiaBank for language in aphasia, PhonBank for phonological development, and FluencyBank for the clinical study of disfluencies. Two are funded by NSF: HomeBank for daylong audio recordings in the home and a grant for database creation for FluencyBank. In addition, we have NEH/DFG funding for web-based learning and study of Latin and Historical German. For access to the data from each of these funded projects, as well as additional not-yet-funded databases, please go to the homepage at
http://talkbank.org .
2. Note that there are three separate “Multilingualism Banks” on the TalkBank home page: SLABank is for learner corpora, BilingBank is for corpora from advanced bilinguals often including code-switching, and Second Language Tutors is for various methods to support second language learning.
3. Although we do not currently have funding for construction of SLABank and BilingBank, we are in fact working actively in these areas. In the Second Language Tutors area, we are creating resources for learning English, Chinese, German, and (believe-it-or-not) Latin.
4. For people interested in linguistic corpus analysis, we have created an importer from TalkBank corpora to the ANNIS corpus analysis system. The first import to this system is Roger Brown’s child language from Eve, but we will soon have many more corpora in the ANNIS system. The URL is
http://gandalf.talkbank.org:8080/annis-gui-3.4.4/ If there are any SLABank or BilingBank corpora that you would like to see available through ANNIS, please tell us.
5. We have developed and refined a tutorial for the learning CLAN. You can download this tutorial from
http://childes.talkbank.org/clan/tutorial.zip . Included in the tutorial is a description of a new method for automatically adding captions from a CHAT file to a linked video.
6. Julia Ruser at SDU and Davida Fromm at CMU are creating YouTube-type tutorials for some of the more difficult aspects of CA coding in CLAN. An earlier set of these videos from Sarah Kress is available from
http://talkbank.org/videos
7. We are also working on the implementation of a system for switching between alternative camera streams while working with a single transcript.
8. Finally, we should note that, because Apple dropped its support for QuickTime on Windows, we had to reprogram CLAN for Windows to work instead with Windows Media Player. That work is now completed. If you use CLAN on Windows, you should download a new version.
We have added a few new corpora, but we are currently mostly focusing on “cleaning up” current corpora by linking them directly to the audio and making sure that they can be analyzed grammatically using the MOR tagging systems we have for English, French, German, Hebrew, Italian, Spanish, Japanese, Mandarin, Cantonese, and Dutch. The goal here is to develop corpora that fully represent code-switching for both learners and bilinguals, while still supporting audio linkage and automatic grammatical analysis. So far, we have been focusing this work primarily on the FLLOC and SPLLOC corpora from classroom learners of French and Spanish.
We are also interested in developing a set of classroom recordings that illustrate some of the most popular methods for foreign language instruction. If yo are wiling to contribute video corpora of this type, or any other corpora on multilingualism, please contact us.
Best regards,
- Brian MacWhinney, Professor of Psychology, CMU
P.S. If you want to stop receiving mail from
bilin...@googlegroups.com , you can send an email to
ma...@cmu.edu.