You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Eric Ihli a93544a9c3 fixup! Fix bug from refactor 3 years ago
doc Initial commit, syllabification and phonetics. 3 years ago
resources Add resource for spelling variants 3 years ago
src/com/owoga fixup! Fix bug from refactor 3 years ago
test/com/owoga Add function phonify a space-separated phrase 3 years ago
.gitattributes Initial commit, syllabification and phonetics. 3 years ago
.gitignore Initial commit, syllabification and phonetics. 3 years ago Bump version for release, update changelog. 3 years ago
LICENSE Initial commit, syllabification and phonetics. 3 years ago Add link in README to Clojars 3 years ago
deps.edn Add function phonify a space-separated phrase 3 years ago
pom.xml Fix pom.xml repo url 3 years ago


This library:

  • Gives you the phonetic pronunciation of a word.
  • Syllabifies a sequence of phones.

For phonetics, it uses the CMU Pronouncing Dictionary. and CMULexicon from

(require '[com.owoga.phonetics :refer :all])

(get-phones "hello")
;; => [["HH" "AH0" "L" "OW1"] ["HH" "EH0" "L" "OW1"]]

For syllabification, it uses a custom algorithm that is basically:

  • Start from the last phone and work backwards.
  • Take every consonant up through the the first encountered vowel. (This is the "rime" of the syllable).
  • Take every consonant up through the next vowel with the conditions (This is the "onset" of the syllable):

    • Each subsequent consonant is more sonorous than the previous.
    • If the next vowel we'd encounter is lax, it gets followed by at least 1 consonant.
  • Repeat until no phones remain.

That algorithm could be described as the maximal onset principle with an override that lax vowels must be in closed syllables (must end with a consonants).

For example, `pity` would be syllabified as `pi` `ty` given the maximal onset principle but since lax vowels can't end a syllable it actually gets syllabified as `pit` `y`.

(require '[com.owoga.phonetics.syllabify :refer :all])

(let [phones (first (get-phones "alaska"))]
  (syllabify phones))
;; => [["AH0"] ["L" "AE1" "S"] ["K" "AH0"]]

Thanks to CMULexicon, you can also syllabify made-up words.

(let [phones (first (get-phones "foobarbazia"))]
  (syllabify phones))
;; => [["F" "UW1"] ["B" "AA1" "R"] ["B" "AA1"] ["Z" "IY0"] ["AH0"]]

You can also go backwards from a word, however this only works for words that appear in the CMU Dictionary. A future TODO and a request for contribution would be for a function that converts a made-up set of phonemes into a somewhat appropriate spelling of a word.

(let [phones (first (get-phones "diploma"))]
   (syllabify phones)
   (get-word phones)])
;; => [["D" "IH0" "P" "L" "OW1" "M" "AA0"]
;;     [["D" "IH0" "P"] ["L" "OW1"] ["M" "AA0"]]
;;     ["diploma"]]

(let [phones (first (get-phones "foobarbazia"))]
  (get-word phones))
;; => nil


In the typical theory of syllable structure, the general structure of a syllable (σ) consists of three segments. These segments are grouped into two components:

Onset (ω) a consonant or consonant cluster, obligatory in some languages, optional or even restricted in others Rime (ρ) right branch, contrasts with onset, splits into nucleus and coda

Nucleus (ν) a vowel or syllabic consonant, obligatory in most languages Coda (κ) consonant, optional in some languages, highly restricted or prohibited in others

Also, for "ellipsis", ps is not a legal internal coda in English. The s can only occur as an appendix, e.g. the plural -s at the end of a word. So it should be e.lip.sis


Air flow goes through nose.

Examples: "n" in "nose", "m" in "may", "ŋ" in "funk".

"ŋ" is known as the letter "eng" and the technical name of the consonant is the "voiced velar nasal"

"voiced" in the above sentence refers to whether or not your vocal chords are active. Your voice chord doesn't vibrate with voiceless consonants, like "sh" "th" "p" "f". In contrast, notice the vibration in phonemes like "m" "r" "z".


A segment is ambisyllabic if it belongs to two syllables.


The English word hammer cannot be divided into two syllables `ha` and `mer`; the [m] functions both as the final segment of the first syllable and as the initial consonant of the second syllable.

This library doesn't syllabify words based on their letters. It syllabifies words based on their phonemes.

The two `m`'s in "hammer" are represented by a single phoneme, `M`. So, when it gets syllabified, the [m] only functions as an onset to the final rime.

Ambisyllabism TODO

Provide a function that inserts an extra phone where ambisyllabism occurs.


The initial skeleton of this library was generated from

What follows is an unedited part of that skeleton. TODO: Update with syllabify-specific development documentation.

Invoke a library API function from the command-line:

$ clojure -X com.owoga.syllabify/foo :a 1 :b '"two"' {:a 1, :b "two"} "Hello, World!"

Run the project's tests (they'll fail until you edit them):

$ clojure -M:test:runner

Build a deployable jar of this library:

$ clojure -X:jar

This will update the generated `pom.xml` file to keep the dependencies synchronized with your `deps.edn` file. You can update the version (and SCM tag) information in the `pom.xml` using the `:version` argument:

$ clojure -X:jar :version '"1.2.3"'

Install it locally (requires the `pom.xml` file):

$ clojure -X:install

Deploy it to Clojars needs `CLOJARS_USERNAME` and `CLOJARS_PASSWORD` environment variables (requires the `pom.xml` file):

$ clojure -X:deploy

Your library will be deployed to com.owoga/syllabify on by default.

If you don't plan to install/deploy the library, you can remove the `pom.xml` file but you will also need to remove `:sync-pom true` from the `deps.edn` file (in the `:exec-args` for `depstar`).


Copyright © 2021 Eihli

Distributed under the MIT License.