1.2 Peculiarities of the East Germanic subgroup

As far as East Germanic dialect were the first to separate from the rest of Germanic-speaking community (North and West Germanic dialects), there is a number of East Germanic innovations that result from the early splitting of the Gothic branch and distinguish it from West and North Germanic branches. To the innovations belong mostly features that were preserved in East but lost in West and North Germanic as well as those developed by East Germanic languages autonomously. Other features differentiating East from two other Germanic groups developed due to phonetic, morphological and lexical processes that took place in West and North Germanic languages whereas East ones remained inert to them.

So, East Germanic innovations include:

1. at the end of the word voiced spirants are devoiced;

2. ē>ei, ō>ū

3. no rhotacism

4. ending –a in proper names of the weak declension (Wulfila, Dagila), like the Nominative Case of the weak declension.

There are also features which are common to Gothic and Scandianvian languages:

1. peculiarities of Holtzmann’s Law (Gmc *ii, *uu developed consonantal clusters);

2. Preterite of Strong verbs, 2^nd sg, is formed according to the Indo-European perfect (ablaut change in the second form + ending t);

3. the first person Subjunctive in the present tense is formed by –au, -a from au with u as the imperative particle.

4. A group of verbs with a stem-suffix *ne/ *no (Gothic weak verbs of Class IV, ending in –nan);

5. Interrogative pronouns are formed by means of an additional suffix: *hwa, *hwe + suffix.

6. Participle I is declined according to the weak declension.

7. Some common lexical units.

Common West and East Germanic features include:

1. common suffix of abstract nouns;

2. Genetive and Dative Cases of numerals 14-19 which are declined according to the i- stem declension;

3. some common lexical units.

9.3 Gothic and Germanic

Our chief sources for ascertaining the approximate quality and quantity of the pronunciation of Wulfilian Gothic are:

1. the pronunciation of the Greek and Latin alphabets as they were obtained in the fourth century, the former taken by Wulfila as the basis for representing his own native sound-system;

2. comparison of the Gothic spelling (Wulfilian orthography) of Greek loan words and proper names occurring in Wulfila’s translation with the original Greek words. Cf.:

Pawlus for Gk. Paúlos, but Pauntius for Gk. Póntius;

3. the spelling of Gothic proper names found in Greek and Latin records of the fourth-eighth centuries;

4. the occasional fluctuating orthography of one and the same word in the translation of the Bible;

5. special sound laws within the Gothic language itself. Gf.:

e → i everywhere (OHG weg, Go. wigs “way”; also, whereas the change [x] → [h], [x^w] → [h^w] in North and West Germanic seems to have occurred only initially and before a vowel, in Gothic the corresponding change was everywhere;

6. the comparison of Gothic with the other Old Germanic languages. Cf.:

Go. wulfs, Far. ulvur, Norw. ulv, Fris. ulw, etc.

Traditionally, the Gothic vowel-system is represented by 5 elementary letters a, e, i, o, u and 4 digraphs ei, iu, ai, au. The vowels /e, o/ were always long (/e:/ was a long close e-sound like in English “she”; /o:/ was a long close vowel, strongly tinctured with the vowel like in German “gut”). /a/ and /u/ had both a short and long quantity. /a/ had the same sound as the a in German Mann; /a:/ occurred only in the combination āh in native Gothic words: jah “and”, brahta “took”. /u/ had the same sound as the vowel in English “put”, /u:/ had the same sound as the vowel in German “gut”. /ı/ was probably a short open vowel like the i in English “bit”; the corresponding long vowel /ı:/ was regularly expressed by the digraph ei.

/iu/, after the analogy of the Greek pronunciation in the fourth century, was a falling diphthong.

Highly arguable is a traditional approach to the pronunciation of /ai/ and /au/: some scholars think they have single value, others differentiate 3 positions: 1) ai – a short open e – sound, almost like the a in English “hat” in front of r, h and h : Go. airþa “earth”; ái – a diphthong, the same sound as in German “mein” and nearly the same sound as the i in English “mine”: hláifs “bread”; ai – probably the same pronunciation as Old English æ; it occurs only in very few words, like saian “sow”; 2) aú – a short open o-sound, like in English “hot” before r, h, h ; áu – a pure diphthong with approximately the same sound as in English “house” – áugo “eye”; au – a long open o-sound like in English “ought” which occurs in a few words before a following vowel.

T he contemporary consensus (Bennett, Voyles, Marchard and others) tends to the view that ai and au represent /ε / and /ɔ/. Though absolute certainty is impossible to attain in such matters, it has been shown to be highly improbable that ai and au signaled more than one phoneme each. For example, Marchand defines the position of the sound represented by ai as, relatively, higher and more front and less lip-rounded than the sound represented by the sign e.

Marchand considers ai and au together, since they are parallel in all aspects. The method of correlation shows that Wulfila distinguished degrees of aperture in vowels. He used au to read o in the names from the Bible. Besides, if Wulfila had two phonemes, spelt au, he would have signalled the phonemic difference if means were available, and those means were available. So, au was a more open sound than o. On the whole, ai and au were opposed as vowels of front vs back articulation.

The fact that Old Norse and Old High German have diphthongs in corresponding words tells us only that Proto-Germanic had diphthongs in these words. But the course of monophthongization in Romance (Proto-Romance *ai and *au) shows us that various sister languages may exhibit monophthongization at various dates. So, the fact that diphthongs in Old High German correspond to ai and au in Gothic cannot be construed to mean that Go. ai and au must be interpreted as diphthongs in the corresponding etyma.

According to the transcription the Gothic consonant system is represented by b, d, f, g, h, h , j, k, l, m, p, q, r, s, t, þ, w, and z. Cf.:

b: had a twofold pronunciation: initially, medially and finally after consonsnts, it was a voiced plosive like English b; medially after vowels it was a voiced bilabial spirant, like v in English “live”: Go. liban “live”, haban “have”.

d: initially, medially and finally after consonants was a voiced plosive like English do; medially after vowels it was a voiced interdental spirant nearly like th: Go. fadar.

g: initially the same sound as English “go”; medially between vowels it was a voiced spirant; final postvocalic g and g in the final combination gs (Go. dags “day”) was probably a voiceless spirant like the ch in German “ich”; before another guttural it was used to express a guttural nasal ŋ: briggan “bring”.

h: initially before vowels and probably also medially between vowels it was strong aspirate as haban “have”, hairto “heart”; in other positions it was a voiceless spirant like in German “nacht”.

h: was either a labialized h or a voiceless w; it may be pronounced like the wh in the Scotch pronunciation of “when”, as heila /xwı:la/ “time”.

j: like in yet: juggs “young”.

r: was a trilled linqual r, like the r in Lowland Scotch.

Besides, b, d, z alternate with f, þ, s in word final position:

Go. giban “give”

gaf “gave”

B and d alternate with f and þ before s. This is usually considered as proof that b, d, z in Wulfilian pronunciation were spirants in the position after a vowel.

The pronunciation of h and q as clusters is argued. Moulton regards h and q as representing the phonemic sequences /h/ + /w/ and /k/ + /w/. Unlike Moulton, Bennett considers h and q unit phonemes. His principal reasons for such interpretation are three. First, in reduplication (Strong verbs of Class 7/h is treated as a single consonant. Secondly, the verb saih an “see” belongs to Class 5 of strong verbs whose stems end with single consonants. Thirdly, the operation of Thurneysen’s “law of dissimilation” would indicate that w is voiced, whereas h and q remain voiceless even between sonorous and operate as single consonants. Voyles accepts only the latter two of these reasons, because some other phonemic sequences - /st/, /sk/, and /sp/ - can reduplicate in their entity: skaisaiþ from skaidan “cut”.

Among disputable items on behalf of Wulfilian orthography and its interpretation is the sequence ggw. In Bennett’s opinion, written ggw represents in all cases [ŋgw].He maintains that early Go. /ggw/ (from Gmc *ww) merged with /ngw/, so such words as triggws “true”, bliggwan “beat” and siggwan “sing” must have been pronounced similarly. Bennett gives three reasons in support of his point of view. First, both siggwan and bliggwan belong to Strong Verbs of Class 3 which are formed by means of the ablauting vowel followed by a nasal plus a consonant. Second, a change of /ggw/ to /ngw/ is an instance of contact dissimilation. Third, new /ngw/ was much more frequent than /ggw/.

A s for Voyles, he resolutely opposes to two of Bennett’s reasons. First, he calls to mind that not all strong verbs of Class 3 have the canonical form ablauting vowel + nasal + consonant (for example, þriskan “thresh”). Then, original /dd/ and /bb/ were used in few words, whereas /nd/ and /mb/ were much more frequent. Finally, scribal variations of the type bringiþ/briggiþ are attested, whereas not once were they used in case of triggws or its derivative forms. To sum up, Voyles fails to consider ggw’s in all cases as /ngw/.

While considering the position of East Germanic within the Germanic group of languages, the comparativists came to the conclusion that North and West Germanic share more features than North and East Germanic (as represented by Gothic) do. The scholars assume that Gothic had already split off from North-West Germanic, while the sound changes common to North and West Germanic were still going on. This assumption is based on the order of phonological changes ascribed to the earliest stages in the development of Germanic: Grimm’s and Verner’s Laws (stage I); [x^w] → [x], [ k^w] → [k] in certain environments (stage II); o → a in certain environments and e → i in unstressed position (stage III); later vocalic changes: e → i before nasal + consonant; and before zero to three consonants + i or j; [i] → [e], [u] → [o] before one to three consonants (excluding nasals) + a non-diffuse vowel, like /a(:), e(:), o(:)/ (stage 4).

The above 4 stages are usually ascribed to Proto-Germanic times. But, according to Voyles, “it would seem that even as early as stage 3 the first dialect split in Proto-Germanic had begun.” He is of the opinion that “the most economical derivation of Gothic requires a departure at stage 3 from the changes evinced in North and West Germanic with similar, but not idential, rules” (Voyles 1968: 739).

First, the change of [o] → [a] (stage 3) seems to have occurred everywhere in Gothic, whereas in North and West Germanic [o] → [a] would seem to have been with exceptions (cf. the infectional ending of Go. dat. Pl. dagam “days” and 1pl.pres. bairam “we carry” and the corresponding inflectional endings of Old High German tagum and berumēs).

S econd, whereas the change of [e] → [i] in unstressed position (stage 3) is the first step in the derivation of North and West Germanic (cf. IE * seghés – “seizure” → Gmc * segés- → * ségis- → * sígis-), in Gothic [e] → [i] is found everywhere.

Third, while in Gothic [x^w] → [h^w], [x] → [h] everywhere, in North-West Germanic [x^w] → [h^w], [x] → [h] only initially and before a vowel.

Finally, in Gothic [ı] → [ε], [u] → [ɔ] before r, h, h^w, North and West Germanic had other conditions, as well.

It was suggested that North and West Germanic provide instances of at least thirteen sound changes shared by these two groups to the exclusion of Gothic. Moreover, they outweigh the one phonological change which is traditionally known to be shared by North and East Germanic – Holtzmann’s Law. But it is quite possible that the rules subsumed under the name “ Holtzmann’s Law” (Gmc. jj and ww appear as ddj in Gothic, but as ggw in Old Norse) represented two separate sound changes for Norse and Gothic, i.e. somewhat similar types of sound change.

To it sum up, the ethnological differences in Proto-Germanic would seem to have arisen before the various Germanic peoples left Scandinavia and before the time of our historical records. At least the split of Proto-Germanic into North-West and East Germanic would seem to have taken place at about the same time.