Φωνήεντα Chapter 2

Innovations in paradigmatic approaches to IE accent-ablaut

Φωνήεντα·

1 « 2 » 3 · 4 · 5

---

The endingless locative as a link between paradigms

One of the subcategories of athematic nouns which frequently exhibited accent-ablaut alternation were the so-called “heteroclites,” nouns that had a suffix containing *-(e)r- in the strong cases, but *-(e)n- in the weak cases. Schindler (1975a) noted an interesting pattern with some neuter heteroclites: there were some neuter singulars which could be reconstructed along with a related word bearing Schmidt’s (1889) collective *-or- suffix and fitting into a unique variation of the amphikinetic type with a third accent pattern in the locative case, a subtype called “holokinetic.”^❦1 Schindler 1975a p 1-2, 4 For example, Schindler demonstrates that Hittite pa-aḫ-ḫur _NOM/ACC.SG ~ pa-aḫ-ḫu-e-na-aš _GEN.SG ‘fire’ indicates a singular proterokinetic *péh₂·ur _NOM/ACC.SG ~ *ph₂·uén- _OBL, while Gothic fon _NOM/ACC.SG ~ funins _GEN.SG, and Tocharian B pūwar indicates a collective holokinetic *péh₂·uor- _NOM/ACC ~ *ph₂·un-és _GEN ~ *ph₂·uén _LOC.^❦2 Schindler p 13, 1 It is interesting to note that the holokinetic locative *ph₂·uén (with zero-locative ending *-∅) is identical to the proterokinetic weak stem *ph₂·uén-, both with accent in the suffix. This endingless locative *ud·én (~ _NOM/ACC *uód·r^❦3 a proterokinetic noun with *o-grade in the strong stems, cf. *gón·u ‘knee,’ *dór·u ‘wood’) also yields the strong form of Skt udán ‘water,’ which contrasts with genitive udnáḥ, indicating a hysterokinetic paradigm with *ud·én _NOM/ACC ~ *ud·n-és _GEN.^❦4 Schindler 1975a p 5-6 The unique position of the endingless locative could provide insight into the potential relationship between the paradigms of the Erlangen and Leiden models.

In a reevaluation of the hysterodynamic reconstructions of the Leiden model, especially from Beekes (1985), Furlan (2008) argues that the hysterodynamic paradigm can ultimately be derived from an original holokinetic paradigm. Furlan posits that forms such as Hittite keššar ~ kiššeran ‘hand,’ indicating hysterodynamic *gʰés·r _NOM.SG ~ *gʰs·ér-m _ACC.SG ~ *gʰs·r-és _GEN.SG, reflects an older holokinetic *gʰés·or- _STRONG ~ *gʰs·r-és _GEN ~ *gʰs·ér _LOC (Hittite -ar could also reflect *-or-), and that this shift from holokinetic to hysterodynamic was facilitated by a transfer of the endingless locative stem to the accustive: “In this declension the change of acc. *CéC-(R)oR-ṃ into *CeC-(R)éR-ṃ was probably caused by loc.sg. *CC-(R)éR because of the comparable locative function (quo?), which was intrinsic to both cases.”^❦5 Furlan 2008 p 297·^❦6 The hysterodynamic paradigm is needed to explain the short and long a reflexes in Greek ἅλς, Latin sāl ‘salt’ ← *séh₂·l-s _NOM.SG ~ *sh₂·él-m _ACC.SG, cf. Lubotsky 1989 p 60. The strange behavior of the locative and its relationship to strong stems is discussed in Chapter 4.

A correspondence between accent and ablaut grade

An interesting consequence of Furlan’s proposed relationship between the hysterodynamic and holokinetic paradigms is that the unaccented *o-grade of the holokinetic is original, and the one-to-one correspondence between the accent and *e-grade ablaut is an innovation of hysterodynamic nouns.

In contrast, Kloekhorst (2013, 2014) reasserts the correspondence between the accent and full *e-grade of all athematic nouns in the Leiden model by demonstrating that only static athematic nouns with accented *-é- can be confidently reconstructed — there is not enough evidence, according to Kloekhorst, to reconstruct static *é: ~ *-é- or *-ó- or *-é-.^❦7 Kloekhorst 2014 p 140-159 Klimp (2013) goes even further, claiming that “the [static] gradation type probably did not exist, for the evidence adduced in favor of this gradation type does not hold under scrutiny” (p 19). Moreover, Kloekhorst demonstrates that the proterokinetic, hysterodynamic, and amphikinetic (including holokinetic) paradigms postulated of the Erlangen model are valid as reconstructed, but can be derived from the Leiden model’s proterodynamic and hysterodynamic paradigms.^❦8 Kloekhorst 2013 p 10-11 In the exact reversal of Furlan’s hypothesis, Kloekhorst derives the amphikinetic (and holokinetic) paradigm from the hysterodynamic pattern, and the *o-grade of amphikinetic nouns are secondary: the result of de-accented *e.^❦9 Kloekhorst 2013 p 7 Kloekhorst argues for a development following a chronology proposed by Kortlandt and Beekes wherein Proto-Indo-European originally had one vowel phoneme, *e, with full-grade only appearing within an accented syllable; *[o] (under pressure from a lowering older **u) arose as an allophone of *e in unaccented syllables, and later analogy within paradigms caused newly unaccented *e to become the allophone *[o], and original *[o] to become accented in certain positions, rendering it in a new phoneme *o no longer in complementary distribution with *e; in this chronology, phonemic long vowels develop later.⁹ Some examples of the emergence of *o by way of allophony and subsequent analogy include the following:

Table 4 — Hysterodynamic to Holokinetic
‘earth’	1 vowel phoneme */e/ Hysterodynamic > *e [o] >		Analogy >	2 phonemes Holokinetic
NOM.	*dʰégʰ·m	*dʰégʰ·[o]m   é↘	*dʰégʰ·[o]m	*dʰégʰ·om-s
ACC.	*dʰgʰ·ém-m   em↗	*dʰgʰ·ém-m	*dʰégʰ·[o]m-m	*dʰégʰ·om-m
LOC.	*dʰgʰ·ém	*dʰgʰ·ém	*dʰgʰ·ém	*dʰgʰ·ém

In this example, the full-grade from the accusative suffix is analogically extended to the nominative stem, then the full-grade root of the nominative is generalized to the accusative.

Here, the strong full-grade root is generalized (but remains unaccented in the weak stems), then the root-accent is generalized, creating a phonemic distinction between accented *e and *o.

In deriving the hysterokinetic, amphikinetic, and holokinetic paradigms from the hysterodynamic paradigm, Kloekhorst (2013) also observes that “the proterodynamic inflection is mainly found with neuter words, whereas the hysterodynamic inflection occurs almost only with non-neuter words...point[ing] to an original complementary distribution.”⁸ This suggests the possibility of a static paradigm for both animate and neuter (i.e. inanimate) nouns, a mobile proterodynamic pattern for neuter nouns, and a mobile hysterodynamic pattern for nouns:

Table 6 — Kloekhorst’s (2013) noun paradigms
	Static ANIM & INAN			Mobile
				ANIM. (Hysterodynamic)			INAN. (Proterodynamic)
	R	·S	-E	R	·S	-E	R	·S	-E
NOM.	é	∅	∅	é	∅	∅	é	∅	∅
ACC/LOC				∅	é	∅
WEAK				∅	∅	é	∅	é	∅

Such an animacy-based assignment of these paradigms makes sense, as inanimate nouns formally have identical nominative and accusative forms in most daughter languages, and there is more semantic imperative to distinguish between the nominative and accusative singular of animate nouns, since they would be the nouns involved in the action of a clause. I will return to possible semantic relationships between the accent-ablaut paradigms in sections 4 and 5.

Given such a direct relationship between accent and ablaut grade in nouns, with *o being the result of an unaccented allophone of *e — a relationship that is also mirrored in the verbal system ^❦10 Steriade 1988·^❦11 Kortlandt 2004, 2015 — no original vowels other than *e can be reconstructed for the early stages of PIE. Once again, the typological remarkability of a single phoneme *e in lieu of a low vowel phoneme should be stressed.

Φωνήεντα·

1 « 2 » 3 · 4 · 5

---

references

---

bibliography

• Furlan, M 2008. “The origin of Anatolian relations of the type keššar : kiššeran and Balto-Slavic relations of the type akmuo/kamy : akmenį/kamenь.” Linguistica (Ljubljana) 48,1, p 293-307
• Kloekhorst, A 2013. “Indo-European nominal ablaut patterns: the Anatolian evidence.” Museum Tusculanum
• —— 2014. “The Proto-Indo-European acrostatic inflection reconsidered.” Das nomen im Indogermanischen. Morphologie, Substantiv versus Adjektiv, Kollektivum. Akten der Arbeitstagung der Indogermanischen Gesellschaft 14,16, Reichert, p 140-163
• Kortlandt, F 2004. “Accent and ablaut in the Vedic Verb.” Indo-Iranian Journal 41,1, Brill, p 7-15
• —— 2015. “Indo-European o-grade presents and the Anatolian ḫi-conjugation.” MS
• Lubotsky, A 1989. “Against a Proto-Indo-European phoneme *a.” New Sound of Indo European, p 53-66
• Schindler, J 1975a. “The ablaut of the Indo-European r/n-stems.” Indo-European Studies 2, p 210-225
• Steriade, D 1988. “Reduplication and syllable transfer in Sanskrit and elsewhere.” Phonology 5,1, p 73-155

---

philology