Jayeless.net

Proto-Indo-European

Proto-Indo-European is the theorised, reconstructed ancestor of all the Indo-European languages. It is thought to have been spoken on the Pontic-Cas­pian steppe (spanning from the lands north of the Black Sea to the lands north of the Caspian Sea, so largely within modern Ukraine and Russia) between about 4500 BCE and 2500 BCE. Its speakers seem to have been nomadic pastoralists who had domesticated horses, and are widely believed to have been part of the Yam­naya culture.

While no direct link can be reconstructed, the theorised homeland of PIE speakers is basically right next to the theorised homelands of the Uralic, Northwest Caucasian and Kartvelian language families, and there are certainly some borrowed vocabulary and other linguistic similarities.

Phonology

The phonology of Proto-Indo-European has been reconstructed using the comparative method; basically, looking at all the reflexes of PIE words in the descendant languages to determine what was the most likely original word (and what sound changes would be required to get from there to the modern reflexes). As such, the reconstruction is subject to debate in some ways, and might be revised if further evidence and better theories arise.

PIE is traditionally analysed as having four main vowels, two short and two long, conventionally written *e, *o, *ē, *ō. (Asterisks are used to show these are reconstructions, not actual observed forms.) Some others, like *a, are reconstructed as occurring very rarely – so rarely that there is doubt about whether they really were PIE phonemes, or an artifact of imperfect reconstruction. (For example, *a itself almost always occurs adjacent to the laryngeal *h₂ or reconstructed *k, which may actually have been pronounced [q]. So perhaps *a was actually the result of other vowels being “a-coloured” by the presence of one of those consonants, and the allophone became phonemic in at least some of the daughter languages, such that its existence in PIE is sometimes reconstructed.)

Conversely, PIE was far more abundant in consonants. It’s traditionally reconstructed as having:

Labial Coronal Dorsal (palatal) Dorsal (plain) Dorsal (labial) Laryngeal
Nasals *m *n
Voiceless stops *p *t *ḱ *k *kʷ
Voiced stops (*b) *d *g *gʷ
Aspirated stops *bʰ *dʰ *ǵʰ *gʰ *gʷʰ
Fricatives *s *h₁, *h₂, *h₃
Liquids *r, *l
Semi-vowels *y *w

Some notes:

  • As mentioned, the velar series (all the variants of *k and *g) may not have been pronounced as traditionally reconstructed. Uvular theory describes the idea that actually, the “palatovelars” may have been plain velars *[k], *[g], *[gʰ], and the “plain velars” might have been uvulars *[q], *[ɢ], *[ɢʰ]. See centum-satem split for more, because the split concerns how the velar (or maybe uvular) consonants evolved in PIE’s daughter languages.
  • The laryngeal consonants are not directly reconstructable based on the living descendant languages, but are thought to exist based on the effect they seem to have had on phonemes that appeared around them. That is, if you didn’t hypothesise their existence, you’d have to hypothesise a number of sound changes that occurred “just randomly, lol” and that’s not really a thing that happens. Some of the laryngeals are preserved in the extinct Anatolian languages. Some linguists also think there are a small number of direct descendants of these laryngeals in Western Iranic languages and Albanian. Otherwise, they are reconstructed mostly based correspondences between different vowels in the daughter languages, and particularly on ablaut in Ancient Greek.

PIE had variable stress, including within a paradigm (e.g. a singular form of a noun being stressed on a different syllable to the plural form of the noun). It’s been reconstructed as having a pitch accent.

Grammar

PIE was a highly inflected language. Typologically it’s thought to have been a fusional language, with affixes being added to word roots in order to show the relationship of words within the sentence (rather than word order and grammar particles being used, as in an analytic language like modern English). Some types of inflection that PIE had include:

  • Grammatical gender: late PIE had three genders (masculine, feminine, neuter). Anatolian only shows an animate/inanimate split, so it’s thought that early PIE might only have had this two-gender system, with the feminine gender breaking off from masculine later.
  • Case declinations: it’s thought that PIE had eight or nine noun cases:
    • nominative: marks the subject of a sentence, and also any “renaming” of the subject in the sentence (e.g. after the verb “to be”)
    • accusative: marks the direct object of a transitive verb
    • genitive: marks a noun that modifies another noun (in English our clitic ’s is a relic of this)
    • dative: marks the indirect object of a transitive verb
    • instrumental: marks the means by which, or with which, the subject accomplishes an action (can be a concrete object or abstract)
    • locative: marks the location of something (roughly parallel to the English prepositions at, in, on and some senses of by)
    • ablative: marks movement away from something (parallel to English from or out of)
    • vocative: used to mark a direct addressee, like “John” in, “What do you think, John?”
    • allative: this is the ninth “maybe” case, observed in Anatolian. Replaced in other branches by the locative case, it expresses movement towards something.
  • Grammatical number: PIE had a dual number separate from singular or plural, for three numbers total.
  • Pronouns: Inflected for noun case, and singular/plural, but apparently not dual. PIE only had first- and second-person pronouns; for the third person it instead used demonstratives. Some pronouns had two entirely separate stems, still preserved in the English words I and me.
  • Verb conjugation: oh boy, did PIE verbs ever conjugate. Verb forms varied depending on:
    • Aspect: verbs could be stative (representing a state), perfective (representing a completed action), or imperfective (representing an ongoing or repeated action)
    • Mood: verbs could be indicative (that’s the standard, declarative mood), imperative (used for issuing instructions), subjunctive (used for indicating various kinds of “unreal” states) or optative (used specifically for wishes or hopes)
    • Voice: verbs could be active (where the subject is the verb’s agent) or passive (where it isn’t)
    • Person: first, second or third.
    • Number: singular, dual or plural.

PIE words generally consisted of a root morpheme carrying the core lexical meaning that couldn’t stand alone, a suffix added to create a word stem, then another affix to form a complete word. Many morphemes are believed to have had *e as their inherent vowel, and PIE employed ablaut extensively, changing those vowels into other vowels as part of the word-building process.

PIE seems to have particles used both as adverbs and postpositions. The postpositions became prepositions in most of the descendant languages.

Word order in PIE was flexible because of the way inflections were used to show what roles different words played in the sentence. There is disagreement about what the “default, unmarked” word order would’ve been; some argue for SVO, while others argue for SOV.