Compensatory lengthening
Compensatory lengthening in phonology and historical linguistics is the lengthening of a vowel sound that happens upon the loss of a following consonant, usually in the syllable coda, or of a vowel in an adjacent syllable. Lengthening triggered by consonant loss may be considered an extreme form of fusion (Crowley 1997:46). Both types may arise from speakers' attempts to preserve a word's moraic count.[1]
| Sound change and alternation |
|---|
| Fortition |
| Dissimilation |
Examples
English
An example from the history of English is the lengthening of vowels that happened when the voiceless velar fricative /x/ and its palatal allophone [ç][2] were lost from the language. For example, in the Middle English of Chaucer's time the word night was phonemically /nixt/; later the /x/ was lost, but the /i/ was lengthened to /iː/ to compensate, causing the word to be pronounced /niːt/. (Later the /iː/ became /aɪ/ by the Great Vowel Shift.)
Both the Germanic spirant law and the Ingvaeonic nasal spirant law show vowel lengthening compensating for the loss of a nasal.
Non-rhotic forms of English have a lengthened vowel before a historical post-vocalic */r/: in Scottish English, girl has a short /ɪ/ followed by a light alveolar /r/, as presumably it did in Middle English; in Southern British English, the */r/ has dropped out of the spoken form and the vowel has become a "long schwa" [əː].
Classical Hebrew and Aramaic
Compensatory lengthening in Classical Hebrew and Aramaic is dependent on the class of consonant which follows the prefix (definite article in Hebrew and prefix waw-hahipukh in both languages).
E.g. (using the Hebrew definite article [hey with pataḥ plus dagesh in following consonant]):[3]
- Before ע and א it is usually [hey with qametz].
- Before ח and ה it is usually [hey with pataḥ]. If it is pretonic it may be [hey with qametz].
- But when it is propretonic, whatever the guttural, it will usually be [hey with segol].
Ancient Greek
Compensatory lengthening is very common in Ancient Greek. It is particularly notable in forms where n or nt comes together with s, y (= ι̯), or i. The development of nt + y was perhaps thus:
- *mont-yă → montsa (palatalization ty → ts) → mõtsa (nasalization and vowel lengthening) → mõssa → mõsa (shortening ss → s) → mōsa (denasalization, retention of long vowel) = μοῦσα "muse"
Forms with this type of compensatory lengthening include the nominative singular and dative plural of many participles, adjectives, and nouns, the 3rd person plural ending for present and future active of all verbs, and the 3rd person singular present of athematic verbs:
- *πάντ-ς → πᾶς "every, whole" (masculine nominative singular)[4]
- *πάντ-ι̯ᾰ → *πάντσα → πᾶσα (feminine)
- *πάντ-σι → πᾶσι (masculine/neuter dative plural)
- compare παντ-ός (m./n. genitive singular)
- *όντ-ι̯ᾰ → *όντσα → οὖσα participle "being" (feminine nominative singular)[5]
- *οντ-ίᾱ → *ονσία → οὐσία "property, essence"
- compare ὀντ-ός (m./n. genitive singular, from participle ὤν "being",)
- Doric ἄγ-ο-ντι → ἄγοντσι → Attic/Ionic ἄγουσι "they drive"
- Doric φα-ντί → *φαντσί → Attic/Ionic φᾱσί "they say"[6]
Indo-Aryan Languages
In the evolution of the modern Indo-Aryan languages, there is a first stage in which consonant clusters with dissimilar consonants preceded by a short vowel undergo assimilation resulting in consonant clusters with similar consonants. In the second stage, the first consonant of the cluster or geminate was lost, which was accompanied by the lengthening of that vowel and sometimes additional nasalization. In Punjabi, only the first stage occurred, while most of the other modern Indo-Aryan languages underwent the second stage as well.
| Sanskrit | Punjabi | Hindi | Translation |
|---|---|---|---|
| हस्तः (hastaḥ) | ਹੱਥ (hatth) | हाथ (hāth) | hand |
| सप्त (sapta) | ਸੱਤ (satt) | सात (sāt) | seven |
| अष्ट (aṣṭa) | ਅੱਠ (aṭṭh) | आठ (āṭh) | eight |
| कर्तनम् (kartanaṃ) | ਕੱਟਨਾ (kaṭṭanā) | काटना (kāṭanā) | cutting |
| कर्म (karma) | ਕੰਮ (kamm) | काम (kām) | work |
| अर्धम् (ardhaṃ) | ਅੱਧਾ (addhā) | आधा (ādhā) | half |
| अद्य (adya) | ਅੱਜ (ajj) | आज (āj) | today |
| सर्पः (sarpaḥ) | ਸੱਪ (sapp) | साँप (sā(n)p) | snake |
| अक्षि (akṣi) | ਅੱਖ (akkh) | आँख (ā(n)kh) | eye |
| दुग्धम् (dugdhaṃ) | ਦੁੱਧ (duddh) | दूध (dūdh) | milk |
| पुत्रः (putraḥ) | ਪੁੱਤ (putt) | पूत (pūt) | son |
See also
- Mora
- Phonological history of the English language
- Ancient Greek nouns
Notes
- Hayes, Bruce (1989). "Compensatory Lengthening in Moraic Phonology". Linguistic Inquiry. The Massachusetts Institute of Technology. 20 (2): 253–306.
- Millward, C. M. (1996). A Biography of the English Language. Boston: Wadsworth. p. 84.
- Hoffer, Victoria. Biblical Hebrew: Supplement for Enhanced Comprehension. New Haven, Conn.: Yale University Press, 2005. Pg. 58. See also Garrett, Duane A., and Jason S. DeRouchie, A Modern Grammar for Biblical Hebrew. Nashville: Broadman & Holman Academic, 2009. Chapter 8.
- Smyth, par. 299: adjs. in nt.
- Smyth, par. 301 a and d: participles in nt.
- Smyth, par. 462 note: Doric athematic verb endings.