Visually compare the phonetic inventories of two languages.
When learning a new language, you need to know the new sounds: the ones that aren't a part of your native sound system. I wanted a tool to quickly compare the different sounds of two languages. pho-diff presents two languages in the form of a diff of their IPA charts.
For two languages a and b, if a letter (which represents a distinctive sound) is absent in b's chart but present in a's, it is colored green. Likewise, if a letter is absent in a but present in b, the letter is colored red (see Examples). pho-diff also outputs a Clojure map with additional information (other sounds not on the chart, links to the source URLs, etc.).
pho-diff can be useful for second language learners trying to learn pronunciation. While IPA resources are biased for English speakers, pho-diff can assist with developing resources for ESL: it can identify sounds from someone's first language that may not exist in English or identify sounds in English that do not exist in someone's native language.
pho-diff can diff between two arbitrary languages as long as they're in the Speech Language Archive; paired with other tools, this can be helpful for languages that have fewer common resources or speakers.
Use pho-diff to find the new sounds in the target language (colored green) and use the Interactive IPA Chart to hear roughly the sounds that the symbols represent.
lein run "english" "spanish"
lein run "spanish" "english"
lein run "korean" "esperanto"
lein run "german" "tagalog"
- The Speech Accent Archive provides the set of native language phonetic inventories (you can read more about the project here).
- Phonetics, sound inventory (Babylon) — an explanation of phonetics and phonology.
- Interactive IPA Chart — hear the approximant sounds that the symbols represent
- Fluent Forever — a language learning method that emphasizes learning the new sounds of a language and getting the pronunciation right.
- American Speech-Language-Hearing Association — a curated list of resources for learning the phonemic inventories and cultural information across languages.
- UCLA Language Materials Project — teaching resources for less commonly taught languages.
Download from https://github.com/sgepigon/pho-diff.
git clone https://github.com/sgepigon/pho-diff.gitlein run "a" "b"a and b should be languages from the Speech Accent Archive. See Bugs for caveats.
lein run "english" "tagalog"{:keys [:a "english" :b "tagalog"],
:charts
{:cons "resources/output/english-tagalog-cons.gif",
:vowels "resources/output/english-tagalog-vowels.gif"},
:other-sounds
{:a #{"labio-velar voiced central approximant [w]" "5 diphthongs"},
:b #{"labio-velar central approximant [w]"}},
:sources
{:a
"http://accent.gmu.edu/browse_native.php?function=detail&languageid=18",
:b
"http://accent.gmu.edu/browse_native.php?function=detail&languageid=64"}}
If either languages a or b lack an IPA chart, pho-diff will return nil.
Not all languages listed on the Speech Accent Archive have an inventory chart. Instead, the pages say "Coming soon…" e.g. "malagasy", "yapese", "sotho", "hmong daw", and "tamajeq".
"newari" actually does not say "Coming soon…" but it is missing the charts. "newari" points to "newar" which does say "Coming soon…"
There are some languages that do have IPA charts, but are slightly off, e.g. "yupik", "mandinka", and "swiss german". This misalignment results in ugly diffs:
The diff is usable, but still an eyesore.
- Clojure – Written in Clojure
- ImageMagick – Uses the convert program to create the diffs
Copyright © 2017 Santiago Gepigon III
Distributed under the Eclipse Public License either version 1.0 or (at your option) any later version.