Thread:User talk:CodeCat/Proto-Indo-European/reply (7)

Firstly, I wonder where you are finding words like 'paw', 'sem' and 'mehl'. If they are pronouns (which they are used as here) then they ought to have case endings, they can't just be used as bare roots. (If a PIE term ends with - then it is a root, not a full word!)

The ablaut system allows most instances of e to be replaced by o, or to be removed entirely. But the conditions under which that applies are strictly dictated by the grammar, so you can't just say 'let's put o there instead'. One example of how ablaut still survives in modern English is in verbs like sing-sang-sung. From a PIE perspective, the verb has the e-grade in the present aspect (not tense!), the o-grade in the perfect singular, and the zero grade in the perfect dual/plural and perfect participle. However, that is just the rule that was generalised in Germanic verbs, ablaut in PIE was more complex than that.

The system for forming words is like this. Every fully inflected verb or nominal (a nominal is any word that has cases, so a noun, adjective or pronoun) is formed with three pieces: the root (which gives the basic meaning), the suffix, and the ending (for case/number in nominals, or for person/number in verbs). R+S+E for short. R+S together form the stem of the word. Some nominals, but especially a lot of verbs, have no S, so they are just R+E. There are also words with several suffixes, so they are R+S+S+E and so on. Most later IE languages have different declensions like o-stems, ā-stems, n-stems, s-stems and so on. And verbs come in a wide variety like -ā-, -i- etc in Latin. In PIE, verbs and nominals come in just two basic types: athematic and thematic. Athematic stems have no thematic vowel at the end of the stem, thematic stems do. From the perspective of the later languages, o-stems and (most) ā-stems are thematic, all others (including i- and u-stems!) are athematic.

Thematic stems are easier to understand. They have no ablaut and accent variations: the accent is the same in every form, and the ablaut grades of the syllables don't change. Thematic stems all end in the so-called thematic vowel, which is -o- (sometimes -e- in verbs, but only depending on the ending). An example of a thematic nominal is sénos "old", which consists of three pieces, sen-o-s: sen- is the root, -o- is the suffix (which consists of just the thematic vowel in this case), and -s is the ending (for the nominative singular masculine). A thematic verb is bʰéreti "carries", which is bʰer-e-ti.

Athematic stems are much more complex, because they have ablaut and accent variations. So the accent and ablaut vowels may change in different forms. But it doesn't change arbitrarily, there are fixed patterns that the accent and ablaut may follow. In nominals, two types of cases are distinguished: direct (nominative, accusative, vocative) and oblique (the others). In verbs, there are also two types of form: singular and nonsingular. Ablaut and accent changes always happen between direct/singular and oblique/nonsingular forms. For example, the noun dn̥ǵʰ-wéh₂-s "tongue" is athematic, and its genitive form is dn̥ǵʰ-uh₂-és (notice how the suffix went from e-grade to zero grade, and the ending got an e-grade). The verb h₁es- "to be" is h₁és-ti "is" in the 3rd person singular, but h₁s-énti in the third person plural (here, the root went from e-grade to zero grade). One important thing is that if the accent changes between direct/singular and oblique/nonsingular, then the accent is always further to the right in the oblique/nonsingular forms.

Athematic nouns come in four different types, based on the patterns that the ablaut and accent follow. Verbs also have similar patterns. You'll notice that in athematic stems, the syllable with the accent tends to have the e-grade (sometimes o-grade), and the syllable without accent has zero grade, but that's not always true (it is a good rule of thumb though).
 * Acrostatic nominals only have ablaut changes, but no accent changes. The accent is always fixed on the root. Examples: (gen. nébʰesos),  (gen. nékʷts).
 * Amphikinetic nominals have the accent on the leftmost syllable of the stem in the direct cases, and on the ending in the oblique cases. Examples: (gen. (dʰ)ǵʰmés)
 * Hysterokinetic nominals have the accent on the rightmost syllable of the stem in the direct cases, and on the ending in the oblique cases. Examples: (gen. dn̥ǵʰuh₂és),  (gen. ph₂trés)
 * Proterokinetic nominals have the accent on the second-rightmost (penultimate) syllable of the stem in the direct cases, and on the rightmost syllable of the stem in the oblique cases. Examples: (gen. ǵustéws),  (gen. sh₁m̥éns)

Now, concerning verbs, things get very... different I suppose. We're used to verbs having tenses, but PIE verbs didn't have tenses as such. Rather, PIE verbs are based on aspect which is "the flow of time through the event". That sounds like tense but it is not. The present aspect describes events that have some kind of internal structure in time, such as being continuous, habitual, ongoing, unfinished. The aorist aspect describes events that have no internal structure in time, so they just 'happened' and we don't care about the details (they generally have past tense meanings). The perfect or stative aspects describes not events, but certain states of being (which themselves might be the result of something happening earlier). From a historical point of view, the present tense in Germanic derives from the present aspect in PIE (or sometimes the aorist), while the past tense derives from the perfect aspect.

What is confusing at first though is that there is strong evidence that the three aspects in PIE were originally separate verbs altogether. So that means that at one point in time, there was one present verb, one aorist verb, and one perfect verb, and they eventually merged after PIE to form part of a single verb's inflection. In PIE itself, a single verb root may have only one or two aspects instead of all three. The root for example only has the present aspect. There are even some verb roots that have the same aspect twice, formed in two different ways! For example, the root apparently had two present stems, thematic ǵn̥h₃ské- and athematic ǵn̥néh₃-/ǵn̥nh₃-, and also two aorist stems, ǵnéh₃-/ǵn̥h₃- and ǵnḗh₃s-/ǵnéh₃s- (both athematic). One thing you may notice here is that the inflection of each aspect is separate: one may have a suffix, the other may not, and one may be thematic while the other is athematic. So you really need to treat each verb aspect as if it were a verb by itself, and reconstruct each aspect separately.