BLICK - a phonotactic probability calculator
Version 1.0 (beta), July 22, 2012
Function and purpose
Input: any string of English phonemes, expressed with a symbol set provided. It can be a real word or one you made up.
Output: a numerical value, inversely reflecting the phonotactic "goodness" (probability, well-formedness, word-likeness ) of the word. For more on the meaning of this score, see the manual.
Examples: the current version of BLICK predicts that ket [K EH1 T] should a completely perfect word of English (penalty score zero), that doit [D OY1 T] should be a somewhat peculiar word of English (score 3.094), and that nguhyee [NG AH H Y IY0] should be a pretty horrible word of English (score 12.295). (Transcriptions in CMU notation; see program interface for translations.)
Purpose: providing a reference point for phonotactic probability in experimentation. This is the same purpose as the pioneering phonotactic model proposed by Vitevitch and Luce (2004); BLICK is a by-product of my efforts, with colleagues, to explore alternative approaches to phonotactic well-formedness based on phonological theory.
Quick start guide
1. Find a Windows machine
2. Download the program and unzip it.
3. Go into the folder thus created and click on the program, BLICK.exe.
4. Left side window is for individual words, right side is for a file name with words in it.
5. Valid transcription format is displayed on the interface.
6. If you need more help read the manual.
- The manual for BLICK
- The grammar used by BLICK (Excel format)
- The phonetic dictionary used in setting the weights of the grammar. This is all the words in the CMU pronouncing dictionary that have a CELEX frequency of at least one. Numerous corrections and annotations.
- A comparison of models on randomly-generated words; this will make sense if you read the talk slides below.
- Slides for a talk I'm giving about this at LabPhon 2012.
Bug reports and feedback
To Bruce Hayes at email@example.com.
Last revised 22 July 2012