Φωνήεντα Chapter 5

Revisiting the PIE vowel phoneme inventory

Φωνήεντα·

1 · 2 · 3 · 4 « 5

---

The semantic origins of *o

Schindler (1972) noted that nouns with *e-grade in the strong stem tended to be agent nouns, whereas nouns with *o-grade in the strong stem more often had a resultative meaning. This closely parallels Pooth’s distribution of agentive nouns oon the one hand, and detransitive nouns with suprasegmental *o) on the other. Sandell (2014) claims that this potential semantic patterning behaves independently from accentuation (the latter of which Sandell explains compositionally with Kiparsky’s Basic Accentuation Principle and a revised version of Kiparsky’s Oxytone Rule). However, Schindler also claimed that nominal accent-ablaut patterns had “ursprünglich gewiß funktionell unterschiedene” [originally funcitonal differences, certainly].^❦1 Schindler 1975b p 260 Additionally, if Klimp (2013) is correct about an absence of sufficient evidence for a static paradigm, and Kloekhorst’s (2013) complementary distribution between proterodynamic inanimates and hysterodynamic animates is accurate, then there was almost certainly a morphologically productive accentual system operating in PIE nouns. On the other hand, if Kloekhorst (2013) is indeed correct about the Kortlandt-Beekes chronology, and accent alone predicts later opposition in vowel quality, then an agentive-detransitive distinction cannot be ascribed to vowel quality (as Pooth suggests). Rather, the surface accent in Kloekhorst’s paradigms can be explained using Frazier’s (2006, 2007) OT constraints, with the only contrastive compositional element between the two paradigms being inherently accented derivational suffixes within inanimate stems.^❦2 Postaccentuation was therefore probably an inherent feature of all roots, assuming no static roots The accentuation of the suffix could be semantically motivated, i.e. the morphology assigns accent to the derivational suffix of inanimate nouns. Such an analysis provides an explanation for the surface accent of both the animate and inanimate nouns while also maintaining the semantic morphological component to the differentiated paradigms.

Alternative vowel reconstructions — the return of *a

Regardless of whether Proto-Indo-European had *e and *o as allophones of each other, or if they were separate phonemes, the typologically extraordinary reconstruction of a language with only mid-vowels and no phonemic low vowel remains an issue. Pulleyblank (1965) notes the strange asymmetry of the Indo-European reflexes of the mid-vowels: *e exhibits raising to i in many Indo-European languages, whereas *o rarely raises, but rather surfaces as a in most Indo-European language families.^❦3 All branches except for Italic, Celtic, Armenian, and Hellenic undergo a change of *o → a Pulleyblank argues that, rather than the usually reconstructed front/back constrast, the reflexes of *o and *e instead suggest a vertical contrast of *a and *ə, respectively.^❦4 Pulleyblank 1965 p 92 Pulleyblank compares the semantic distribution of *o and *e with the “introvert” and “extrovert” semantic qualities posited in Kuipers’ analysis of Kabardian’s two vowel phonemes /a/ and /ə/, respectively^❦5 Pulleyblank 1965 p 93-95 :

The Kabardian distribution of /a/ and /ə/ mirrors the PIE respective distribution of *o and *e very closely; nouns with *o exhibit inanimate/detransitive qualities, while nouns in *e exhibit animate/agentive qualities. Additionally, verbs reconstructed with *o usually exhibit the “perfect” (stative) and the middle (detransitive) voice, signify intransitive motion, or else are clear derivatives from root verbs with *e-grade. However, if *[o] is to be taken as the unaccented allophone of *e, transferring such an analysis to the new values *[a] ~ *ə́ becomes problematic, as it still leaves the low vowel [a] as an allophone alongside a phonemic accented schwa.

Kümmel (2012) rejects the unaccented *[o] allophone hyphothesis, stating that, althoug back vowels tend to have a lowe pitch, “the inherent pitch level of segments is not usually confused with prosodic pitch (tone), and parallels for a development of an unaccented front vowel to a back vowel are not easily found.”^❦7 Kümmel 2012 p 306 Kümmel points to a regular development in Anatolian languages of accented *ó to ā in most positions (all positions in Hittite), as well as Brugmann’s law, whereby *o which paradigmatically contrasted with *e becomes ā in Indo-Iranian open syllables.^❦8 Kümmel 2012 p 308-309 Kümmel thus justifies reconstructing *ā rather than *o, and *a for *e with back allophones adjacent to *h₂ and *h₃ and fronted allophones elsewhere; the latter reconstruction “is similar to the allophony we find in Semitic systems with just one non-high short vowel: e.g. Arabic /a/ and /ā/ are very often fronted in many varieties, except adjacent to uvulars, pharyngeals, and/or pharyngealized consonants.”^❦9 Kümmel 2012 p 313

If Pooth’s templatic hypothesis is to be accepted, a suprasegmental detransitive feature [+round] yielding *o can be replaced with a feature [+long], or perhaps another underlying *a, in which case length would merely be the phonetic realization of */aa/, and only one vowel *a need be reconstructed for PIE. However, if Kloekhorst’s model is preferred, then both *e and *o of later Pre-Indo-European would result from a single phoneme *a(with no length or quality contrasts), with the postulation that *o arose from unaccented *a (= traditional *e) at a stage when full-grade no longer directly corresponded with accent placement. While Kümmel finds an evolution of unaccented *e → *o unlikely, a change from *a to both *e and *o^❦10 potentially through intermediary *ɛ and *ɔ, respectively is less suspect; Pulleyblank (1965) notes that Kabardian has “six [allophones] ranging from [e] to [o] for /a/.”^❦11 Pulleyblank 1965 p 93-94 Deriving both vowel qualities from a single, typologically paralleled vowel *a allows for a purely accent-based explanation for the observed vowel quality in Indo-European athematic nouns (such as that proposed by Kloekhorst) while also resolving the typological issue of an absent low vowel. The consequent vowel inventory would consist of the syllabic variants of semivowel phonemes *i and *u along with one true phonemic vowel *a which developed both front and back allophones, paralleling the phonemic quality distribution of languages like Arabic.

If such an analysis is accurate, then the work of Lubotsky (1989) to eliminate a phoneme *a in favor of *e ~ *o is effectively reversed; those places where traditional *e adjacent to *h₂ can be reanalyzed as the retention of low vowel quality yielding Hellenic, Italic, Celtic, and Armenian a.

Concluding remarks

The articles in this series have surveyed paradigmatic and compositional approaches to PIE nominal accent-ablaut and evaluated their potential for offering insight into the different vowel qualities (*e and *o) traditionally reconstructed for PIE. Some approaches take the vowel apophony observed in athematic nouns as an inherent feature of the paradigms, with no further attempt to account for the placement of *e and *o with relation to accent, but more often, paradigmatic approaches emphasize patterns between the accent placement and the vowel quality, and explain *[o] as an allophone of *e; the assumption of allophony is implicit (or left unexplained) in most compositional approaches to nominal stem allomorphy as well. An alternative to an accent-based explanation for vowel quality, in the form of a highly productive templatic system with a suprasegmental morpheme which manifests itself with the vowel quality *[o], has also been examined against the other approaches, and the semantic and phonological implications of all approaches have been considered. Finally, the possibility that both *e and *o in Indo-European derived from allophones of a single vowel phoneme *a (with a typological parallel in Kabardian) has been considered in the context of paradigmatic allophony, and has been evaluated to fit well with an accent-based semantic distinction between animate and inanimate nouns.

Some issues which require further examination in light of the ideas presented in this study include the observed vowel apophony (*e ~ *o) in verbs, the Caland system of adjectives (and adjectives in general), and the relationship between the accentual system and nasal infixation.

Φωνήεντα·

1 · 2 · 3 · 4 « 5

---

references

---

bibliography

• Frazier, M 2006. “Accent in Proto-Indo-European athematic nouns: Antifaithfulness in Inflectional Paradigms [Thesis].” U of NC Chapel Hill
• —— 2007. “Accent in athematic nouns in Vedic Sanskrit and its development from PIE.” Proceedings of the 18th Annual UCLA Indo-European Conference
• Kiparsky, P 1973. “The inflectional accent in Indo-European.” Language 49, p 794-849
• —— 2010. “Compositional vs. paradigmatic approaches to accent and ablaut.” Proceedings of the 21st Annual UCLA Indo-European Conference
• Kiparsky, P & Halle, M 1997. “Towards a reconstruction of the Indo-European accent.” Studies in Stress and Accent. Southern California Occasional Papers in Linguistics 4, p 206-238
• Klimp, J 2013. “Remnants of *r/n-stem heteroclite inflection in Germanic.” U of Groningen
• Kloekhorst, A 2013. “Indo-European nominal ablaut patterns: the Anatolian evidence.” Museum Tusculanum
• Kümmel, MJ 2007. Konsonantenwandel. Bausteine zu einer Typologie des Lautwandels und ihre Konsequenzen für die vergleichende Rekonstruktion. Reichert
• Pooth, RA 2015a. A typological overview of Proto-Indo-European
• —— 2015b. “Proto-Indo-European nominal morphology part 1. the noun.” Language Arts 1, FIU Cologne
• Pulleyblank, EG 1965. “The Indo-European vowel system and the qualitative ablaut.” WORD 21,1, p 86-101
• Sandell, R 2014. “Perspectives on the reconstruction of the Proto-Indo-European accentual system.” UCLA Indo-European Workshop
• Schindler, J 1972. “L’apophonie des noms racines indo-européens.” Bulletin de la Société de Linguistique de Paris 68, p 31-38
• —— 1975b. “Zum ablaut der neutralen s-Stämme des Indogermanischen.” Flexion und Wortbildung. Akten der V. Fachtagung der Indogermanischen Gesellschaft. Regensburg, Reichert

---

philology