This directory contains procedures for Praat (Boersma & Weenink, 2009) to assist in the measurement of Voice Onset Time (VOT; Lisker & Abramson, 1964; Abramson & Whalen, submitted). It also contains example files for the user to explore.
One procedure, "prepopulate.praat", takes a Praat TextGrid with segments labeled (interval labels) and prepopulates a new interval tier ("Phone") with labels appropriate for measuring VOT:
- VDCLO: the voiced portion of a closure
- VLCLO: the voiceless portion of a closure
- REL: the release burst
- ASP: the aspiration after the burst
For later analysis, only REL is required, and it is required even if the release burst is absent (an arbitrarily short interval should be used). The other intervals may or may not be present. See Abramson & Whalen (submitted) for details.
To use "prepopulate.praat", create labels for a speech file with, at a minimum, an interval tier (default name = "Segment") that labels the segments of interest with IPA symbols. The default set of symbols expected is "b,d,g,p,t,k", but others (e.g., "s,sʰ,pʰ,tʰ,kʰ) can be added. Then, open Praat and run "prepopulate.praat". It will allow the choice of directories and files, as well as segments to be analyzed. For each segment named, the four intervals are inserted in a new tier, "Phone". The duration of that segment is divided arbitrarily into four equal lengths for the initiation of the labels. The user can then modify them in Praat, opening the new TextGrid with the Sound files, and adjusting the boundaries to the true locations; any unneeded labels can be removed.
These labels can also be entered entirely by hand. We have found it quicker to adjust the boundaries than to have to type the label names many times.
Once the "Phone" tier is complete, the second procedure, "get_vot.praat", can be used. It will allow the choice of directories and files, as well as segments to be analyzed. A new tier, "VOT", is created and the VOT measurements are stored there. Positive VOTs will have the label "VOT", while negative VOTs (prevoicing) will have the label "mVOT" (for "minus VOT"). A .csv file can also be created with these values.
Note: Before creating a .csv file, go to Praat -> Preferences -> Text writing preferences. There, choose UTF-8 so that the file will open properly in Excel. If still some fonts are not appearing properly, use Import option on Excel and import the .csv file with UTF-8 setting.
Additional information can be found at the beginning of each procedure.
- get_vot.praat is the main VOT procedure script to return VOT, given Praat tiers with appropriate labels
- prepopulate.praat adds 'Phone' tier if not exist and prepopulates 'Phone' tier with the equidistant labels (i.e. VDCLO VLCLO REL ASP)
- example_getvot includes an example TextGrid file for get_vot.praat
- example_prepopulate includes an example TextGrid file for prepopulate.praat
- example_result has result files from get_vot.praat and example wav file
- Download all files (or git clone)
- Open prepopulate.praat
- Set parameters (e.g. new tier number for Phone tier)
- Set directory of TextGrid files
- Segment tier should be created in TextGrid files
- For more information, see VOT_examples_noPhone.TextGrid under example_prepopulate
- Run
- Download all files (or git clone)
- Open get_vot.praat
- Set parameters (e.g. tier numbers, labels)
- Set directory of TextGrid files
- Run