List of Sound Changes

From FrathWiki
Jump to: navigation, search

Below is a large list of attested sound changes in natural languages.


One sound becomes more like another, or (much more rarely) two sounds become more like each other. (See also: )

Place Assimilation

One sound changes its place of articulation to match that of a nearby sound.

  • In consonant clusters:
    • Anticipatory: C > [α place] / _C[α place]
      • This is a very common sound change to affect nasals, and also fairly frequent for obstruent consonants.
      • Example: 半角 /haNkakɯ/ [hãŋkakɯ̥] "half-width" (Japanese)
      • Example: octō /oktoː/ "eight" (Latin) > otto /ɔtto/ "eight" (Italian)
    • Lagging: C > [α place] / C[α place]_
      • Lagging place assimilation is much rarer, and tends to affect word-final consonant clusters only.[citation needed]
      • Example: Regen /ʁeːɡən/ [ʁeːɡŋ̩] "rain" (German)
  • In a CV or VC sequence:
    • Anticipatory consonant-to-vowel assimilation: C > [α place] / _V[α place]
      • A very common change of this type is palatalisation of velar or coronal consonants before front vowels. Other reasonably frequent subtypes are labialisation of velars before rounded vowels, and uvularisation of velars before (open) back vowels.
      • Example: decem /dekem/ "ten" (Latin) > dieci /diɛt͡ʃi/ "ten" (Italian)
      • TO DO: example for labialisation *k > p / _{u,o}
      • TO DO: example for uvularisation *k > q / _ɑ
    • Anticipatory vowel-to-consonant assimilation: V > [α place] / _C[α place]
      • As with anticipatory consonant-to-vowel assimilation, anticipatory vowel-to-consonant assimilation can happen in several dimensions: Fronting of central or back vowels before palatal and/or coronal consonants, backing of front vowels before dorsal and/or retroflex consonants, rounding of unrounded vowels before labial and/or labialised consonants, and lowering of high vowels before uvular and/or pharyngeal consonants.
      • TO DO: examples
    • Lagging consonant-to-vowel assimilation: C > [α place] / V[α place]_
      • A relatively rare change, but attested in all the same subtypes as anticipatory consonant-to-vowel assimilation.
      • TO DO: example: Slavic third palatalisation
      • TO DO: example for labialisation *k > kʷ or p / {u,o}_
      • TO DO: example for uvularisation *k > q / ɑ_
    • Lagging vowel-to-consonant assimilation: V > [α place] / C[α place]_
      • TO DO: examples
  • At a distance:
    • Consonant Harmony
      • TO DO: example: sibilant harmony (Chumash, Ts'amakko)
      • TO DO: example: dorsal harmony (Totonacan)
    • Vowel umlaut (anticipatory): V > [α place] / _C(C)V[α place]
      • Usually affects vowels in stressed syllables, anticipating the quality of the vowel in a following unstressed syllable. Most common is i-umlaut (fronting and/or raising), but u-umlaut (rounding and/or backing) and a-umlaut (lowering) are also attested.
      • Example: *fōts /foːts/, PL *fōtiz /foːtiz/ (Proto-Germanic) > *fōt /foːt/, PL *fø̄ti /føːti/ (Early Old English) > foot /fʊt/, PL feet /fiːt/ (Modern English)
      • Example: *nistaz /nistaz/ (Proto-Germanic) > Nest /nɛst/ "nest" (German)
    • Vowel harmony
      • Usually affects vowels in unstressed syllables, which become similar to the quality of the word's stressed syllable. Both anticipatory and lagging vowel harmony is attested. Common dimensions of vowel harmony are frontness/backness, rounding, and [±ATR] (advanced tongue root).
      • Example: tie /tie/ "road" + -llA "LOC" > tiellä /tiellæ/ "on the road"; talo /tɑlo/ "house" + -llA "LOC" > talolla /tɑlollɑ/ "at the house" (Finnish)
      • Example: O- "3SG" + bE- "FUT" + hor "wash" > obehor /obehor/ "s/he will wash"; O- "3SG" + bE- "FUT" + "go" > ɔbɛkɔ /ɔbɛkɔ/ "s/he will go" (Akan)

Manner Assimilation

One sound changes its manner of articulation to match that of a nearby sound.

  • In consonant clusters:
    • Nasality Assimilation: C > [+nasal] / C[+nasal]_, _C[+nasal]
      • Fairly common (mostly affecting plosives and/or liquids), especially in anticipatory direction.
      • Example: 학년 /haknjʌn/ [háŋnjʌ̀n] "school year" (Korean)
    • Liquid Assimilation: C > [+liquid] / C[+liquid]_, _C[+liquid]
      • Fairly common (usually affecting only coronal plosives, nasals, and liquids), especially in anticipatory direction.
      • TO DO: example (Finnish)
  • In a CV or VC sequence:
    • TO DO: vowel nasalisation
  • At a distance:
    • TO DO: nasal consonant harmony (Guaraní & other South American languages)
    • TO DO: liquid harmony (Bukusu)

A good resource:

Laryngeal Assimilation

One sound changes its laryngeal features (most commonly voicing, but also aspiration or glottalisation) to match those of a nearby sound. Strictly speaking a subtype of manner assimilation, but it's useful to treat it separately.

  • In consonant clusters:
    • Devoicing: C[+voice] > [-voice] / C[-voice]_, _C[-voice]
      • This type of sound change very commonly affects voiced obstruents, and rarely also nasals and/or liquids. The direction of assimilatory devoicing is usually (but not always) anticipatory.
      • TO DO: example of a plosive-plosive cluster and/or fricative-plosive cluster
      • TO DO: example of voiceless resonants (Icelandic)
    • Voicing: C[-voice] > [+voice] / C[+voice]_, _C[+voice]
      • Affects voiceless obstruents. Less frequent than devoicing, but attested in both directions. The most common subtype of assimilatory voicing is probably that of voiceless plosives preceded by a nasal and/or liquid.
      • TO DO: example of a fricative-plosive cluster (Slavic languages)
      • TO DO: example of NT > ND, maybe also LT > LD
    • (De-)Aspiration: C > Cʰ / Cʰ_, _Cʰ // Cʰ > C / C_, _C
      • TO DO: examples in both directions (Ancient Greek, Indo-Iranian)
    • (De-)Glottalisation: C > Cˀ / Cˀ_, _Cˀ // Cˀ > C_, _C
      • TO DO: examples in both directions (Pacific NW? Caucasian?)
  • In a CV or VC sequence:
    • Vowel devoicing: V > [-voice] / C[-voice]_, _C[-voice]
      • Example: /watasi/ [wataɕi̥] "I" (Japanese)
      • Example: 拡張 /kakɯtɕoo/ [kakɯ̹̥tɕoː] "expansion" (Japanese)[1]
      • Notes: In Japanese, this ordinarily only affects high vowels, and successive potential devoicing areas can block devoicing where it might ordinarily occur. Devoicing and tone patterns interact in complex ways.
    • TO DO: Vowel phonation (breathy, creaky...)
    • TO DO: Consonant changes from an adjacent vowel with special phonation
  • At a distance:
    • TO DO: Vowel phonation harmony (does this exist somewhere?)
    • TO DO: Consonant aspiration or glottalisation harmony (Chadic, might also be attested in the Pacific NW?)



  • {C/V} > Ø / #_
    • Example: μνημονικός /mnemonikos/ (Greek) > mnemonic /nǝmɑnɪk/ (English)
    • Example: *cneo /kneo/ (Old English) > knee /niː/ (English)
    • Notes: This sound change is also spelled "apheresis". Often this is a way to resolve complex onset consonant clusters. As a counterpart to this sound change, see prothesis.




  • Ø > V / _#
    • Example: தஞ்சாவூர் /t̪antɕaːʋuɾ/ [t̪andʑaːʋuɾu] "Thanjavur (name of a city)" (Tamil)[2]
    • Notes: The vowel inserted at the end is usually either a copy vowel of the previous vowel, or some neutral vowel in the system. In the Tamil example, it looks like the former, but it's actually the latter: [u] is always inserted.


  • Ø > V / #_
    • Example: *scola /skola/ (Latin) > escuela /eskuela/ "school" (Spanish)[3]
    • Example: *scola /skola/ (Latin) > école /ekol/ "school" (French)[3]
    • Notes: Often the vowel inserted has the status of the most neutral or basic vowel in the system (with the terms "neutral" and "basic" being defined internal to the system). Also, this sound change often applies to words that begin with an impermissible consonant cluster. As a counterpart to this sound change, see aphaeresis.


A "stronger", less sonorous sound becomes a "weaker", more sonorous one.

Intervocalic Voicing

  • C > [+voice] / V_V
    • Example: *metus /metus/ (Latin) > miedo /miedo/ "fear" (Spanish)
    • Notes: This is a very common sound change. Sometimes it can be triggered when one (or both) of the elements on either side of the consonant is a sonorant, not simply a vowel (e.g. Latin patrem > Spanish padre).


A "weaker", more sonorous sound becomes a "stronger", less sonorous one.

Word Final Devoicing

  • {C/V} > [-voice] / _#
    • Example: друг /drug/ [dɾuk] “friend” (Russian)
    • Example: /xaːj/ [xaːj̥] "house" (Mayan)[3]
    • Notes: Often these are near mergers, where the vowel length may differ between forms that end with a phonologically voiced vs. voiceless consonant, or there may be a difference in how the voiceless stop is released if it’s phonologically voiced vs. voiceless.

Offglide Hardening

  • V[-syllabic] > [+consonantal] / V_C
    • Example: αὐτός /a͡utos/ "self" (Ancient Greek) > αυτός /aftos/ "this" (Modern Greek)
    • Notes: In Greek this change happens only to the u-offglide of earlier diphthongs, which becomes /v/ before voiced obstruents, /f/ before voiceless obstruents, and /m/ before nasals. Other languages might have something similar for an i-offglide becoming /ʝ~ʑ~ʒ~z/, /ç~ɕ~ʃ~s/, and /ɲ~n/ in the same environments.[citation needed?]


Consonant Opposition Loss

  • C[phonation]V > CV[tone]
    • Example: 개다 /keta/ [kʰèdá] "fold", 캐다 /kʰeta/ [kʰédà] "dig" (Korean)
    • Notes: Typically "more voiced" consonants become lower tone, and "more aspirated" consonants become higher tone.

Final Consonant Loss

  • C > [tone] / _#
    • Example: *pah > pâ (Vietnamese)
    • Notes: Typically lower-sonority consonants (e.g. voiceless stops) become higher tone, while higher-sonority consonants (voiced stops, fricatives) become lower tones.


  1. From
  2. From A Reference Grammar of Spoken Tamil by Harold F. Schiffman (2006).
  3. 3.0 3.1 3.2 From Historical Linguistics: An Introduction by Lyle Campbell (1999).