All the changes described above affected accented vowels. The development of vowels in unstressed syllables, final syllables in particular, was basically different. Whereas in stressed position the number of vowels had grown (as compared with the PG system), due to the appearance of new qualitative differences, the number of vowels distinguished in unstressed position had been reduced. In unaccented syllables, especially final, long vowels were shortened, and thus the opposition of vowels – long to short – was neutralised. Cf. OE nama (NE name) to the earlier *namon. It must also be mentioned that some short vowels in final unaccented syllables were dropped. After long syllables, that is syllables containing a long vowel, or a short vowel followed by more than one consonant, the vowels [i] and [u] were lost. Cf. the following pairs, which illustrate the retention of [u] and [i] after a short syllable, and their loss after a long one: OE scipu and sceap (NE ships, sheep, pl from *skeapu); OE werian—demon (NE wear, deem; cf. Gt domjan).
Old English Vowel System (9th-10th c.)
The vowels shown in parentheses were unstable and soon fused with resembling sounds: [a] with [a] or [o], [ie, ie:] with [y, y:].
The vowels are arranged in two lines in accordance with the chief phonemic opposition: they were contrasted through quantity as long to short and were further distinguished within these sets through qualitative differences as monophthongs and diphthongs, open and close, front and back, labialised and non-labialised. Cf. some minimal pairs showing the phonemic opposition of short and long vowels:
OE dæl — dæl (NE dale, 'part') is — īs (NE is, ice))col — cōl (NE coal, cool).
The following examples confirm the phonemic relevance of some qualitative differences:
OE ræd — rād — rēad (NE 'advice', road, red), sē — sēo 'that' Masc. and Fern. mā — mē (NE
more, me)
The OE vowel system displayed an obvious tendency towards a symmetrical, balanced arrangement since almost every long vowel had a corresponding short counterpart. However, it was not quite symmetrical: the existence of the nasalised [a] in the set of short vowels and the debatable phonemic status of short diphthongs appear to break the balance.
All the vowels listed in the table could occur in stressed position. In unstressed syllables we find only five monophthongs, and even these five vowels could not be used for phonemic contrast:
i – ænig (NE any)
e – stāne, Dat. sg of stān as opposed to
a – stānaGen. pl of the same noun (NE stone)
o – bæron — Past pl Ind (of beran as opposed to bæren. Subj. (NE bear)
u — talu (NE tale), Nom. sg as opposed to tale in other cases
The examples show that [e] was not contrasted to [i], and [o] was not contrasted to [u]. The system of phonemes appearing in unstressed syllables consists of three units: e/i a o/u
On the whole, consonants were historically more stable than vowels, though certain changes took place in all historical periods.
It may seem hat being a typical OG language OE ought to contain all the consonants that arose in PG under Grimm's and Verner's Law. Yet it appears that very few noise consonants in OE correspond to the same sounds in PG; for in the intervening period most consonants underwent diverse changes: qualitative and quantitative, independent and positional.
Some of the consonant changes dated in pre-written periods are referred to as "West Germanic" (WG) as they are shared by all the languages of the WG subgroup; WG changes may have taken place at the transitional stage from PG to Early OE prior to the Germanic settlement of Britain.
After the changes under Grimm's Law and Verner's Law PG had the following two sets of fricative consonants-voiceless [f, 0, x, s] and voiced [v, ð, y, z].
In WG and in Early OE the difference between the two groups was supported by new features. PG voiced fricatives tended to be hardened to corresponding plosives while voiceless fricatives, being contrasted to them primarily as fricatives to plosives, developed new voiced allophones.
The PG voiced [ð] (due to Verner's Law or to the third act of the shift) was always hardened to [d] in OE and other WG languages, cf., for instance, Gt goþs, godai [ð],OIcel goðr and OE god (NE good), The two other fricatives, [v] and [y]were hardened to [b] and [g] initially and after nasals, otherwise they remained fricatives.
PG [z] underwent a phonetic modification through the stage of [ж] into [r] and thus became a sonorant, which ultimately merged with the older IE [r]. Cf. Gt. wasjan, 0 Icel verja and OE werian (NE wear). This process, termed rhotacism, is characteristic not only of WG but also of NG.
In the meantime or somewhat later the PG set of voiceless fricatives [f, 0, x, s] and also those of the voiced fricatives which had not turned into plosives, that is, [v] and [y], were subjected to a new process of voicing and devoicing. In Early OE they became or remained voiced mtervocally and between vowels, sonorants and voiced consonants; they remained or became voiceless in other environments, namely, initially, finally and next to other voiceless consonants Cf. Gt qiþian, qaþi with [0] in both forms, and OE cweðan [ð] between vowels and cwæð[0] at the end of the word (NE arch, quoth 'say').
The mutually exclusive phonetic conditions for voiced and voiceless fricatives prove that in OE they were not phonemes, but allophones.
In all WG languages, at an early stage of their independent history, most consonants were lengthened after a short vowel before [j]. This process is known as WG "gemination" or "doubling" of consonants, as the resulting long consonants are indicated by means of double letters, e.g.: *fuljan > OE fyllan (NE fill); * sætjanOE > settan (NE set), cf. Gt satjan.
During the process, or some time later, [j] was lost, so that the long consonants ceased to be phonetically conditioned. When the long and short consonants began to occur in identical phonetic conditions, namely between vowels, their distinction became phonemic.
The change did not affect the sonorant [r], e.g. OE werian (NE wear); nor did it operate if the consonant was preceded by a long vowel, e. g. OE demon, metan (NE deem, meet) — the earlier forms of these words contained [j], which had caused palatal mutation but had not led to the lengthening of consonants (the reconstruction of pre-written forms *motjan and *domjan is confirmed by OS motion and Gt domjan).
In Early OE velar consonants split into two distinct sets of sounds, which eventually led to the growth of new phonemes.
The velar consonants [k, g, x, y] were palatalised before a front vowel, and sometimes also after a front vowel, unless followed by a back vowel. Thus in OE cild (NE child) the velar consonant [k] was softened to [k'] as it stood before the front vowel [i]: [*kild]>[k'ild]; similarly [k] became [k'] in OE spræc(NE speech) after a front vowel but not in OE sprecan ("NE speak) where [k] was followed by the back vowel [a]. In the absence of these phonetic conditions the consonants did not change, with the result that lingual consonants split into two sets, palatal and velar. The difference between them became phonemic when, a short time later, velar and palatal consonants began to occur in similar phonetic conditions; cf. OE cild [k'ild], ciest [k'iest] (NE child, chest) with palatal [k'] and ceald, cepan (NE cold, keep) with hard, velar [k] — both before front vowels.
Though the difference between velar and palatal consonants was not shown in the spellings of the OE period, the two sets were undoubtedly differentiated since a very early date. In the course of time the phonetic difference between them grew and towards the end of the period the palatal consonants developed into sibilants and affricates: [k']>[tſ], [g']>[dz]; in ME texts they were indicated by means of special digraphs and letter sequences.
The date of the palatalisation can be fixed with considerable precision in relation to other Early OE sound changes. It must have taken place after the appearance of [æ, æ:] (referred to the 5th c.) but prior to palatal mutation (late 6th or 7th c.); for [æ, æ:] could bring about the palatalisation of consonants (recall OE spræc, NE speech), while the front vowels which arose by palatal mutation could not. In OE cepan. (from *kopjan) and OE cyning (with [e:] and [y] through palatal mutation) the consonant [k] was not softened, which is confirmed by their modern descendants, keep and king. The front vowels [y] and [e:] in these and similar words must have appeared only when the splitting of velar consonants was well under way. Yet it is their appearance that transformed the two sets of positional allophones into phonemes, for a velar and a palatal consonant could now occur before a front vowel, that is, in identical phonetic conditions: cf. OE cyning and cyse (NE king, cheese).
Comparison with other OG languages, especially Gothic and O Icel, has revealed certain instances of the loss of consonants in WG and Early OE.
Nasal sonorants were regularly lost before fricative consonants; in the process the preceding vowel was probably nasalised and lengthened. Cf.:
Gt fimf, 0 Icel fim, OHG fimf — OE fif (NE five)
Gt uns, OHG uns — OE ūs (NE us)
Fricative consonants could be dropped between vowels and before some plosive consonants; these losses were accompanied by a compensatory lengthening of the preceding vowel or the fusion of the preceding and succeeding vowel into a diphthong, cf. OE sēon, which corresponds to Gt saihwan, OE slēan (NE slay), Gt slahan, G. schlagen, OE sægdeand sæde(NE said).
We should also mention the loss of semi-vowels and consonants in unstressed final syllables, [j] was regularly dropped in suffixes after producing various changes in the root: palatal mutation of vowels, lengthening of consonants after short vowels. The loss of [w] is seen in some case forms of nouns: Norn, treo, Dat. treowe (NE tree);
Nom. sæ, Dat. sæwe(NE sea), cf. Gt triwa, saiws.
Lecture 3. OLD ENGLISH GRAMMAR
Preliminary Remarks
OE was a synthetic, or inflected type of language; it showed the relations between words and expressed other grammatical meanings mainly with the help of simple (synthetic) grammatical forms. In building grammatical forms OE employed grammatical endings, sound interchanges in the root, grammatical prefixes, and suppletive formation.
Grammatical endings, or inflections, were certainly the principal form-building means used: they were found in all the parts of speech that could change their form; they were usually used alone but could also occur in combination with other means.
Sound interchanges were employed on a more limited scale and were often combined with other form-building means, especially endings. Vowel interchanges were more common than interchanges of consonants.
The use of prefixes in grammatical forms was rare and was confined to verbs. Suppletive forms were restricted to several pronouns, a few adjectives and a couple of verbs.
The parts of speech to be distinguished in OE are as follows: the noun, the adjective, the pronoun, the numeral (all referred to as nominal parts of speech or nominal, the verb, the adverb, the preposition, the conjunction, and the interjection. Inflected parts of speech possessed certain grammatical categories displayed in formal and semantic correlations and oppositions of grammatical forms. Grammatical categories are usually subdivided into nominal categories, found in nominal parts of speech and verbal categories found chiefly in the finite verb.
We shall assume that there were five nominal grammatical categories in OE: number, case, gender, degrees of comparison, and the category of definiteness / indefiniteness. Each part of speech had its own peculiarities in the inventory of categories and the number of members within the category (categorial forms). The noun had only two grammatical categories proper: number and case. The adjective had the maximum number of categories — five. The number of members in the same grammatical categories in different parts of speech did not necessarily coincide: thus the noun had four cases. Nominative, Genitive, Dative, and Accusative, whereas the adjective had five (the same four cases plus the Instrumental case). The personal pronouns of the 1st and 2nd p., unlike other parts of speech, distinguished three numbers — Singular, Plural and Dual. Cf.
sg OE ic (NE I), dual wit 'we two', pl we (NE we)
OE stān (NE stone) — stānas (NE stones).
Verbal grammatical categories were not numerous: tense and mood — verbal categories proper — and number and person, showing agreement between the verb-predicate and the subject of the sentence.
The distinction of categorial forms by the noun and the verb was to a large extent determined by their division into morphological classes: declensions and conjugations.
In OE there were with the following parts of speech: the noun, the adjective, the pronoun, and the verb.
The OE grammatical system is described synchronically as appearing in the texts of the 9th and 10th c. (mainly WS); facts of earlier, prewritten, history will sometimes be mentioned to account for the features of written OE and to explain their origin.
The OE noun had two grammatical or morphological categories: number and case. In addition, nouns distinguished three genders, but this distinction was not a grammatical category; it was merely a classifying feature accounting, alongside other features, for the division of nouns into morphological classes.
The category of number consisted of two members, singular and plural. As will be seen below, they were well distinguished formally in all the declensions, there being very few homonymous forms.
The noun had four cases: Nominative, Genitive, Dative and Accusative. In most declensions two, or even three, forms were homonymous, so that the formal distinction of cases was less consistent than that of numbers.
Before considering the declension of nouns, we shall briefly touch upon the meaning and use of cases. The functions of cases in OE require little explanation for the Russian student, since they are those, which ought to be expected in a language with a well-developed case system.
The Nom. can be loosely defined as the case of the active agent, for it was the case of the subject mainly used with verbs denoting activity; the Nom. could also indicate the subject characterised by a certain quality or state; could serve as a predicative and as the case of address, there being no special Vocative case, e. g.: