Auslautgesetze

This subject is usually called 'Auslautgesetze' in Germanic linguistics.

The topic of Germanic final endings is one fraught with much debate and is made even more unclear by the fact that the Germanic languages inherited a strong inclination towards regularization in their early stages, causing lots of restructuring in paradigms, both verbal and nominal, resulting in a murky picture of some of the finer aspects of Germanic morphological reconstruction.

A general rule here on the Þiudisk wiki, at least at the moment, we respectfully reject the quantitative theory, or bimoraic vs trimoraic vowel length, of Germanic Auslautgesetze, as is widely accepted and followed by some Germanicists. This isn't a permanent policy and would be subject to change following more study and discussion. For now however, it is easier to accetp that in the intermediate period between the advent of Proto-Germanic ca. 500 BCE and the start of a written record of the Germanic language family beginning some 700 years after the fact, that the individual Germanic dialects and families saw widespread leveling, restructuring and general change as subject to the will of its speakers.

Here we will dissect and interpret the Germanic situation from both before and after its period of existence, by reconstructing and transposing the attested forms in older Germanic languages, as well as what can be projected from Proto-Indo-European and other Indo-European languages.

Kortlandt, Boutkan and Ringe will be referenced haphazardly here. Proper references will be added later.

Adverbial *-ō and *-ē
These two adverbial endings, represented by Gothic -leikō and hidrē, contain long vowels that by all accounts should have been shortened on their way into Gothic, giving expected **-leika and **hidra. However, by retaining their length, one must assume they were not simply long vowels, but either overlong (quantitative theory) or protected by an intervening sound since lost, either *-t or *-ą (see the genitive plural discussion below).

A point to keep in mind, in almost all Germanic languages, and certainly to be found in all three branches, the word-final vowel reflex of Germanic *-līkō is similar to the reflex of the feminine ōn-stem nominative singular *-ǭ, compare: OE -līċe and bēċe (< *bōkijǭ), ON -liga and ganga (< *gangǭ), and Gothic -leikō and fullō (< *fullǭ), exceptions are found in Old Saxon and Old High German. Thus, it is far more likely that some quality of sound protected the long vowel from shortening, just as nasalization protected ōn-stems from the same changes that affected ō-stems.

Word-final -t
A problem is that word-final -t's were already lost by the time of Proto-Germanic, compare the third person plural preterite ending *-un from earlier *-unt. A possible solution to this problem, is that word-final -t didn't simply disappear, but rather left a glottal stop *-Ɂ in it's place, lost after consonants but retained following vowels, therefore *-un, not **-unɁ, and *-ōɁ, not **-ō.

Main a- and ō-thematic stems
The proto-Germanic a-stem masculine class is reconstructed in the following by Ringe (2006):

... Although more accurately the only real difference being that some linguists reconstruct a PIE genitive plural *-oH-om, however only the ending *-om is reconstructible, compare archaic Hittite walkuwan vs. Ancient Greek λῠ́κων (lúkōn) vs. Sanskrit वृकाणाम् (vṛ́kāṇām). We can assume that the very simple genitive plural *-om, was less desirable as contrasted against the accusative singular *-om in the same paradigm, and therefore liable to be subject to change. Such is obviously the case in Latin where the o-stem genitive plural -ōrum was likely created by attaching the plural ending to the accusative plural, as if **-ons-om > *-ōs-om, compare the same development in Sanskrit: accusative plural vṛ́kān, genitive plural vṛ́kāṇām.

In the sections following, we will judge each of the relevant daughter languages vs parent language, first we will show the expected development vs the actual attestations, with notes and discussions.

Gothic

Expected development from Proto-Germanic to Gothic:

Attested forms of Gothic vs Proto-Germanic:

Disregarding the loss of the instrumental case, the biggest differences are:
 * 1) The genitive singular -is vs the loss of -a- in the a-stem declension, wulfs < *wulfaz and the loss of -i- in the i-stem declension, alds < *aldiz
 * 2) genitive plural -ē vs *-Ø, as in the nominative/accusative neuter singular, leik < *līką
 * 3) third syllable apocope of dative plural *-am-az

As to the third:
 * It is likely that the genitive singular -is was borrowed from the light ja-stem genitive singular which was -is < -js from earlier *-jas. A semivowel -j- was not reinserted in this paradigm because there was no basis for which to do so: it wasn't part of this stem.
 * Gothic -ē- should be assumed to be only the result of Germanic -ē₂-, reflecting earlier Germanic *-ija- > *-eia- > *-ej- or *-ē-, with a-mutation. This is shown in Gothic hēr, "here", which can only be assumed to reflect earlier Germanic *hijar, *hi-, "this" + -ar, "place; locative suffix", cf. þar, hwar. Therefore this must have spread from the heavy ja-stems and the i-stem genitive plural *-iją, as argued by Kortlandt.
 * Assumption of weakly stressed third syllable deletion, however this is in clear contrast to n-stem plural -ans < *-aniz. Therefore we would have to assume only *-amaz would be affected.
 * Deletion of remnant -s after second syllable bilabial nasals only.