Montclair Map Task Corpus


College of Humanities and Social Sciences



Document Type


Publication Date



Montclair State University Digital Commons


The Montclair Map Task Corpus is a collection of files associated with a study of conversational interaction conducted in the Speech Communication Laboratory at Montclair State University. The corpus is useful for anyone interested in investigating naturally-produced, unscripted, task-oriented conversational speech among pairs of English-speaking individuals. The talkers completed an iconic map-matching task with 6 pairs of maps in 16 same-sex female, 16 same-sex male, and 16 mixed-sex pairings. Conversations ranged from 16 to 62 minutes in duration, averaging 32 minutes. This corpus was collected in March-May of 2013 (first 24 pairs) and in May of 2014 (last 24 pairs) as part of a large-scale study of phonetic convergence in spoken communication funded by the National Science Foundation PAC program (BCS #1229033, Jennifer Pardo, Principal Investigator).

The corpus comprises 48 .wav files containing 2-channel audio recordings, 48 Praat .TextGrid files with segmentation and annotation data for each audio recording (viewable with audio files in Praat), and 48 .csv files containing data extracted from the .TextGrid annotations. Additional supporting documents include a .pdf file with 12 Map Task images (6 pairs), a .pdf file containing a listing of all landmark label phrases, and a .csv file with demographic information for all talkers.

A more detailed description of the corpus can be viewed in the MMTC Description document listed below. For a reprint of a journal article detailing the corpus collection methods and preliminary analyses, see the MMTC Publication document below.

The corpus files may only be used for research purposes. Therefore, access to the files is limited to individuals at institutions with IRB oversight of research projects. To request access to the corpus files, contact the corpus administrator, Dr. Jennifer Pardo at

File Format