Two significant language corpora, the Michigan Corpus of Spoken Academic English (MICASE), and the Michigan Corpus of Upper-Level Student Papers (MICUSP) were developed at the English Language Institute. Teachers and researchers across the world use these historically important collections of language for linguistic analysis, classroom activities, and materials development.
Below are links to ELI corpus materials, which are all freely available for study, research and teaching. Please see the MICASE and MICUSP fair use statements for preferred citations for both corpora.
MICASE: The Michigan Corpus of Academic Spoken English
The MICASE corpus is now read-only. If you experience difficulty accessing MICASE, please email micase-help@umich.edu.
Search MICASE for words and phrases in specified contexts, returning concordance results with references to files, full utterances, and speakers
Browse MICASE's collection of transcripts of academic speech events recorded at the University of Michigan
Access MICASE sound files at the TalkBank (Free account required for login access)
MICUSP: The Michigan Corpus of Upper-Level Student Papers
The MICUSP corpus is now read-only. If you experience difficulty accessing MICUSP, please email micusp-help@umich.edu.
Browse papers by discipline.
SUPPORTING MATERIALS FOR TEACHERS, RESEARCHERS, AND LEARNERS
These resources include Kibbitzers*, conference presentations, and other small research projects that that draw on data from MICASE and MICUSP to provide pedagogically-relevant insights into important features of advanced student academic speech and academic writing.
*Note from John Swales: Kibbitzer is a Yiddish term for a person who offers opinions. It was adopted by Tim Johns of the University of Birmingham, UK, who was a pioneer in computer-aided language learning. He used it for short explorations of written academic English. We adapt it here for spoken academic English as well.
Archived Michigan Corpus Website
Additional supporting materials for teachers, learners, and researchers are available at the archived Michigan Corpus Linguistics website. NOTE: This is not an active site and is not maintained by the ELI or University of Michigan, so some pages and resources are not accesible.
