As explained elsewhere, MICASE has been consciously designed as a research database “open to the world,” and we hope researchers elsewhere will make use of the resources we have made available at micase.umdl.umich.edu.  At present, if quoting MICASE in presentations or publications, the official citation is “R. C. Simpson, S. L. Briggs, J. Ovens, and J. M. Swales. (2002) The Michigan Corpus of Academic Spoken English. Ann Arbor, MI: The Regents of the University of Michigan.”

If you are using MICASE for linguistic or language teaching research please let John know at jmswales@umich.edu, and for testing research Sarah at slbriggs@umich.edu.  More general inquiries or comments should be addressed to micase@umich.edu.

Descriptive Studies 

            Our in-house research and development program continues to evolve, and below we present a snapshot of the presentations and publications carried out up until 2003; after this, we are happy to report that the research projects involving MICASE --both in-house and externally-- became too numerous and difficult for us to keep track of. Most of our early studies consisted of various kinds of “forays” into the database as we searched for both significant findings and for ways of effectively investigating them. One of the earliest, conducted by graduate student research assistant Stephanie Lindemann and external consultant Anna Mauranen, took the much-studied little word “just.” They have been able to show that in the MICASE data available at that juncture only 8% of its occurrences are concerned with time (as in “the mail has just arrived”). Although this temporal use tends to be stressed in ESL pedagogical materials, the primary function of “just” in academic speech (about 75% of occurrences) is to soften comments and suggestions, as in “if we could just move on to the second issue.”

            A comparable study by Deanna Poos and Rita Simpson looked at the uses of “sort of/sorta” and “kind of/kinda.” Traditionally, these and other hedges have been particularly associated with women's speech. However, study of these expressions in academic speech at Michigan show that they are not particularly favored by women; rather, their frequency appears to be a disciplinary effect: less common in science and more common in the humanities. One intriguing explanation for this phenomenon is that in science technical terms tend to be fixed, while in the humanities and social sciences conceptual and technical vocabulary tends to be more fluid and more negotiated. Hence, the higher frequency of these two prototypical hedge words could reflect both a greater tendency toward negotiation of meaning as well as lexical search time on the part of interlocutors in the “softer” areas.

            Other studies include Janine Ovens' investigation of phrasal negation, as in “I have no idea how to do that,” with a more in-depth look at the highest frequency occurrences in the MICASE data; it would appear that in interactive speech events, this type of negation often plays a role in topic transitions. John Swales examined what he called “discussive” uses of point (“the point I wanna make is”) and thing (“The main thing here is”). Both of these are quite common, and have been examined from a number of perspectives, such as whether they are used retrospectively (referring to previous utterances) or prospectively (prefacing propositions about to come). One interesting feature of these two words is that both are overwhelmingly associated with a positive prosody (Stubbs, 1996).

            Somewhat broader studies have been conducted by Swales & Malczewski and by Anna Mauranen. The former investigates the linguistic resources used by participants in a wide variety of university speech events to move from monologue to dialogue (or the reverse), or to change the direction of the lecture or discussion. These they call New Episode Flags. The most common devices are discourse markers, such as “okay,” “so” and “now.” Up until now, these discourse markers have been analyzed as single elements, but the MICASE data clearly shows that they cluster, as in “okay so now let's move on to something else.” Mauranen also has examined discourse practices in the academy, this time focusing on reflexive talk (or “talk about talk”), especially in monologic contexts. She is able to show that these chunks can be highly mitigated and for a myriad of reasons. Here is a fragment (hedges have been italicized):

            "Um, okay let me get into sort of the more serious stuff, and, um, what I'm hoping to do with the remainder of of this first hour, is just give you some uh bit of perspective, show where biology fits into, sort of the rest of your education, and hopefully I can, um begin that framework that we're gonna fill in in the rest of the term."

            John published a chapter in 2002 (see Flowerdew below) relating his experiences with using a corpus to produce advanced EAP materials. He concluded, in the present state of the art, that this is not such an easy or a straightforward process as might appear at first sight. He has been using MICASE data in the chapters of his forthcoming CUP publication on Research Genres that deal with speech, such as dissertation defenses, research group meetings and colloquia.

            In winter 2001, Rita Simpson offered the first U-M course in corpus linguistics through the Program in Linguistics; subsequent to that Rita and Dushy Mendis studied the use and function of idioms in MICASE, presenting their early results at AAAL in March, 2002. Since then, many of our own researchers and visiting scholars have continued to make good use of the corpus (see the Bibliography section).

Teaching and Testing Developments

            Data collected for MICASE has also been used in instructional materials projects. Elizabeth Axelson, ELI Lecturer, used transcripts and sound files to develop training materials for ITAs (International Teaching Assistants), focusing on linguistic aspects of interactive teaching. The materials developed by Axelson, along with other discipline specific instructional materials for ITA training, are available on the “Discourse within the Disciplines” web pages at http://deil.lang.uiuc.edu/TESOL/index.html. Susan Reinhart has also incorporated MICASE data in her forthcoming textbook on oral presentations (U-M Press). We have made some of our in-house teaching materials based on MICASE available at www.lsa.umich.edu/eli/micase/teaching.htm

            The testing division, the major sponsor of MICASE, has been using the database as a resource for test development and validation. Word frequency information based on MICASE has been useful in the development of new items for the listening section of the Examination for the Certificate of Proficiency in English (ECPE). A preliminary analysis of listening test response data found that items containing MICASE phraseology worked particularly well in discriminating high scoring examinees (presumably more proficient users of English) from low scoring examinees (presumably less proficient users of English).

            Item specifications for the development of listening items are currently being modified so that range (the variety and number of speech events a lexical item occurs in) can be added to frequency. Eventually, ELI Testing will be compiling frequency lists of spoken academic American English, particularly lexical bundles, and core and specialized spoken vocabulary. Finally, in addition to using text analysis programs for frequency data, descriptive statistics and concordancing, the sound files provide information about speech rate and prosodic aspects of spoken English in academic settings, the target language use domain.