In historical linguistics, vowel breaking, vowel fracture,[1] or diphthongization is the sound change of a monophthong into a diphthong or triphthong.
Vowel breaking may be unconditioned or conditioned. It may be triggered by the presence of another sound, by stress, or in no particular way.
Vowel breaking is sometimes defined as a subtype of diphthongization, when it refers to harmonic (assimilatory) process that involves diphthongization triggered by a following vowel or consonant.
The original pure vowel typically breaks into two segments. The first segment matches the original vowel, and the second segment is harmonic with the nature of the triggering vowel or consonant. For example, the second segment may be pronounced as //u// (a back vowel) if the following vowel or consonant is back (such as velar or pharyngeal), and the second segment may be pronounced as //i// (a front vowel) if the following vowel or consonant is front (such as palatal).
Thus, vowel breaking, in the restricted sense, can be viewed as an example of assimilation of a vowel to a following vowel or consonant.
Vowel breaking is sometimes not assimilatory and is then not triggered by a neighboring sound. That was the case with the Great Vowel Shift in English in which all cases of pronounced as //iː// and pronounced as //uː// changed to diphthongs.
Vowel breaking sometimes occurs only in stressed syllables. For instance, Vulgar Latin open-mid pronounced as //ɛ// and pronounced as //ɔ// changed to diphthongs only when they were stressed.
Vowel breaking is a very common sound change in the history of the English language, occurring at least three times (with some varieties adding a fourth) listed here in reverse chronological order:
See main article: Southern American English. Vowel breaking is characteristic of the "Southern drawl" of Southern American English, where the short front vowels have developed a glide up to [j], and then in some areas back down to schwa: pat pronounced as /[pæjət]/, pet pronounced as /[pɛjət]/, pit pronounced as /[pɪjət]/.[2]
The Great Vowel Shift changed the long vowels pronounced as //iː uː// to diphthongs, which became Modern English pronounced as //aɪ aʊ//.
In early Middle English, a vowel pronounced as //i// was inserted between a front vowel and a following pronounced as //h// (pronounced pronounced as /[ç]/ in this context), and a vowel pronounced as //u// was inserted between a back vowel and a following pronounced as //h// (pronounced pronounced as /[x]/ in this context).
That is a prototypical example of the narrow sense of "vowel breaking" as described above: the original vowel breaks into a diphthong that assimilates to the following consonant, gaining a front pronounced as //i// before a palatal consonant and pronounced as //u// before a velar consonant.
In Old English, two forms of harmonic vowel breaking occurred: breaking and retraction and back mutation.
In prehistoric Old English, breaking and retraction changed stressed short and long front vowels i, e, æ to short and long diphthongs spelled io, eo, ea when followed by h or by r, l + another consonant (short vowels only), and sometimes w (only for certain short vowels):[3]
In late prehistoric Old English, back mutation changed short front i, e, æ to short diphthongs spelled io, eo, ea before a back vowel in the next syllable if the intervening consonant was of a certain nature. The specific nature of the consonants that trigger back umlaut or block it varied from dialect to dialect.
Proto-Germanic stressed short e becomes ja or (before u) jǫ regularly in Old Norse except after w, r, l. Examples are:
According to some scholars,[4] the diphthongisation of e is an unconditioned sound change, whereas other scholars speak about epenthesis[5] or umlaut.[6]
The long high vowels of Middle High German underwent breaking during the transition to Early New High German: pronounced as //iː yː uː// → pronounced as //aɪ̯ ɔʏ̯ aʊ̯//. In Yiddish, the diphthongization affected the long mid vowels as well: pronounced as //ɛː oː øː iː yː uː// → pronounced as //ɛɪ̯ ɔɪ̯ ɛɪ̯ aɪ̯ aɪ̯ ɔɪ̯//
This change started as early as the 12th century in Upper Bavarian and reached Moselle Franconian only in the 16th century. It did not affect Alemannic or Ripuarian dialects, which still retain the original long vowels.
In Yiddish, the diphthongization applied not only to MHG long vowels but also to pronounced as //ɛː oː// in words of Hebrew (in stressed open syllables) or Slavic origin:
Vowel breaking is present in Scottish Gaelic with the following changes occurring often but variably between dialects: Archaic Irish eː → Scottish Gaelic iə and Archaic Irish oː → Scottish Gaelic uə [7] Specifically, central dialects have more vowel breaking than others.
Many Romance languages underwent vowel breaking. The Vulgar Latin open vowels e pronounced as //ɛ// and o pronounced as //ɔ// in stressed position underwent breaking only in open syllables in French and Italian, but in both open and closed syllables in Spanish. Vowel breaking was mostly absent in Catalan, in which pronounced as //ɛ// and pronounced as //ɔ// became diphthongs only before a palatal consonant: Latin coxa 'thigh', octō 'eight', lectum 'bed' > Old Catalan pronounced as /
In the table below, words with breaking are bolded.
| Syllable shape | Latin | Spanish | French | Italian | Portuguese | Catalan | |
|---|---|---|---|---|---|---|---|
| Open | petram, focum | piedra, fuego | pierre, feu | pietra, fuoco | pedra, fogo | pedra, foc | |
| Closed | festam, portam | fiesta, puerta | fête, porte | festa, porta | festa, porta | festa, porta |
Romanian underwent the general Romance breaking only with pronounced as //ɛ//, as it did not have pronounced as //ɔ//:
It underwent a later breaking of stressed e and o to ea and oa before a mid or open vowel:
Sometimes a word underwent both forms of breaking in succession:
The diphthongs that resulted from the Romance and the Romanian breakings were modified when they occurred after palatalized consonants.
See main article: Quebec French phonology. In Quebec French, long vowels are generally diphthongized when followed by a consonant in the same syllable (even when a final [ʁ] is optionally made silent).
Some scholars[8] believe that Proto-Indo-European (PIE) i, u had vowel-breaking before an original laryngeal in Greek, Armenian and Tocharian but that the other Indo-European languages kept the monophthongs:
However, the hypothesis has not been widely adopted.
Some languages in Sumatra have vowel breaking processes, almost exclusively in syllable-final position. In Minangkabau, the Proto-Malayic vowels *i and *u are broken to ia and ua before word-final *h, *k, *l, *ŋ, *r (*təlur > *təluar > talua "egg").[9] In Rejang, the Proto-Malayo-Polynesian vowels *ə, i, and u are broken to êa, ea, and oa before any of word-final consonants above except *k and *ŋ (*tənur > *tənoar > tênoa "egg").[10] This process has been transphonologized by loss of *l and *r and merging of several word-final consonants into a glottal stop (*p, *t, *k in Minangkabau, or *k, *h in most dialects of Rejang except Kebanagung).
Word-final Proto-Malayo-Polynesian *-i and *-u were also broken in Sumatra. In Rejang, these vowels are broken into -ai and -au in Pesisir dialect, or into -êi and -êu elsewhere.[10]