1.2. Early and Middle Indo-European

1.2.1. Early and Middle Indo-European evolution

Features of the Middle Proto-Indo-European or Proto-Indo-Anatolian (PIA) parent language can be reconstructed based on Proto-Anatolian (PA) differences with the Common Indo-European (CIE) trunk—defined in turn by differences between Tocharian and other Late PIE dialects—complemented with data informed by internal reconstruction (Kloekhorst 2016, 2017, 2018; Pooth 2018).

Phonology:

·       Laryngeals probably reconstructible as *h and uvular fricatives *χ, *ɣw (Weiss 2016), although possibly uvular stops (Kloekhorst 2018).

·       Vocalic system:

o   Ablauting *e, *o, with *ē, *ō (see below).

o   Dubious existence (or alternatively minimal relevance) of vowel *a.

·       System of stops most likely different from the classically reconstructed *p/*t/*k – *b/*d/*g – *bh/*dh/*gh:

o   Most likely (typologically) *p/*t/*k – *ˀb/*ˀd/*ˀg – *b/*d/*g (Kümmel 2015); also supported by Kortlandt’s glottalic reformulation of Lachmann’s law (Kroonen 2018).

o   Maybe *pː/*tː/*kː – *ˀp/*ˀt/*ˀk – *p/*t/*k, i.e. like Pre-Proto-Anatolian. This could be supported by the divergent evolution of PIE *TT → Late PIE *TsT in compounds with * in Anatolian.

Nominal system:

·       Evolution of the accent-ablaut system (as described by Beekes and Kortlandt):

o   Initially there were apparently only three paradigms: static (inanimate/animate), proterodynamic (inanimate), and hysterodynamic (animate).

o   Sound Law 1: massive vowel reduction, with all accented vowels becoming *e, all unaccented vowels were lost: e.g. nom.-acc. sg. *mén-s, gen. sg. *mn-és-s.

o   Intermediate period A: sometimes spread of vowel *e to unaccented morphemes, e.g. nom. acc. *mén-es. Zero-grade forms may be replaced by its full-grade form in analogy to hysterodynamic paradigms; e.g. gen. *mn-és-es.

o   Sound Law 2: all unaccented *e are weakened to *o. Regular outcome of *mén-es, *mn-és-es is then *mén-os, *mn-és-os.

o   Intermediate period B: new regularisations, e.g. the accented e-grade is generalised throughout the paradigm, yielding *mén-os, *mén-es-os. Vowels *e and *o are now separate phonemes, so *o can spread to accented morphemes.

o   Sound Law 3: In some environments, short *e and *o are lengthened; e.g. *pχ-tḗr ‘father’ is the outcome of an earlier short *e, either because it stood before a word-final resonant, or because it is a compensatory lengthening from **pχ-ters (Szemerényi’s law).

o   Finally, the full reconstructible Middle PIE nominal accent-ablaut system includes also a hysterokinetic (e.g. nom. *pχ-tḗr, acc. *pχ-térm, gen. *pχ-ts) and an amphikinetic one (e.g. *sésor- / *sesr-és ‘sister’).

 

static

proterodynamic

hysterodynamic

 

inanim./anim.

inanim.

anim.

 

R

S

E

R

S

E

R

S

E

nom.

é

-

-

é

-

-

é

-

-

acc.

é

-

-

é

-

-

-

é

-

obl.

é

-

-

-

é

-

-

-

é

loc.

é

-

-

-

é

-

 

é

-

 

Examples of these paradigms are e.g. proterodynamic ‘fire’ nom.-acc. sg. *péχ-ur, gen.sg. *pχ-uén-s; and hysterodynamic ‘hand’ nom. sg. *gés-r, acc. sg. *gs-ér-m, gen. sg. *gs-r-és, although for nom. *gḗs-r see Pooth (2018).

·       The earliest reconstructible PIE gender system showed differences in gender agreement only in the grammatical cases. Different agreement patterns arose primarily in the nominative, with common gender nouns, adjectives, and pronouns showing different case/number endings in contrast to neuter nouns, which did not distinguish the nominative and the accusative (Matasović 2014).

·       Endings:

o   Nom. sg. *-s, *-Ø.

o   Gen. sg. *-(e/o)s, originally probably **-és.

o   Dat.-Loc. in **-i, that develops into an accented *-éi, hence:

§  Dat. unaccented in *-i, accented (hysterodynamic) in *-éi: *CC-éC-i, *CC-C-éi.

§  Loc. unaccented in *-i (proterodynamic, hysterodynamic): *CC-éC-i.

o   Allative in **-é, which developes into *-ó (cf. Hitt. parā ‘forward’, Gk. pró, Skt. prá), possibly with zero-grade *-Ø. Not productive in later stages.

o   Instrumental in *-et, accented *-ét, zero-grade *-t.

o   Development of ablative by adding *-i to the instrumental, cf. PA *-(o)ti. The common ending *-(e)s developed later.

o   Nominal paradigms for Middle PIE:

 

static

proterodynamic

hysterodynamic

 

inanim./anim.

inanim.

anim.

nom.

*CéC-C(-s)

*CéC-C

*CéC-C(-s)

acc.

*CéC-C(-m)

*CéC-C

*CC-éC-m

abl.

*CéC-C-s

*CC-éC-s

*CC-C-és

ins.

*CéC-C-t

*CC-éC-t

dat.

*CéC-C-i

-

*CC-C-éi

i-loc.

*CéC-C-i

*CC-éC-i

all.

*CéC-C

-

*CC-C-é

Ø-loc.

*CéC-C

*CC-éC

 

 

Verbal system:

·       Basic forms were probably injunctive (tenseless) *CéC-t and derivative *CéC-i, with an affix *-i which was either an aspectual (progressive, ongoing at refrence time) or a temporal (hinc et nunc, i.e. ‘here and now’) mark.

·       Endings originally only *-m, *-s, *-t, which added information on person and number.

·       From punctual verbal roots derivatives could be made (by reduplication, n-infix, etc.) with repetitive, durative, causative, etc. meaning; with suffix *-s- a punctual derivative could be made from non-punctual roots.

·       Original distributive-iterative inflectional tipe (*sopé, opposed to *oidé) becomes proto-middle.

 

1.2.2. Schleicher’s fable in Proto-Indo-Anatolian

ɣweu̯is hku̯es-kwe

ɣweu̯is kwoi χelh ne hest

hkums ɣwekwet;

to ˀdenso u̯ogom ugentm̥,

to mgeχ borom,

to ihrom hohku brentm̥.

ēukwt ɣweu̯is hku̯os:

χedgor hme kērˀd,

χnerm̥ u̯iˀdenti hkums χˀgentm̥.”

eukwt hku̯es: “klu ɣwu̯e!

χedgor sme kērˀd u̯iˀdenti,

χnēr, ʔesos, ɣwu̯i̯om χlhen

su̯e gwermom u̯esti kwr̥néuti,

ɣwu̯i̯om-kwe χelh ne hesti.”

To keklu̯us ɣweu̯is pleχnom bēuˀgt.