SELIM03.pdf 27 Juan de la Cruz, Selim 3 (1993): 26-42 OLD ENGLISH NOUN DECLENSIONS ABSTRACT The Old English declensions have traditionally bewildered students for various reasons, but perhaps two of these merit special attention. (1) The ap- parent incoherence between the ‘historic’ labels attached to them in the light of historical reconstruction and their actual forms as they appear in the lan- guage. (2) The fact that many students of Old English have had some training in Latin -a language with neat and clearly marked paradigms. This obviously makes it even more puzzling for them to understand the ‘historic’ labels of the Old English declensions let alone the actual ‘Indo-European’ links between one declension system and the other. This paper aims at clarifying the situa- tion and purports to justify both the labellings and the historic links in a comparatively simple way. O. INTRODUCTION The ‘historical’ labels attached to the OE. noun declensions have tradi- tionally bewildered beginners to the subject. They appear to them opaque or, worse still, confusing. So, the newcomer to the field faces, for example, the great ‘A declension’ of masculines and neuters (including the stems in -JA and -WA) and the also great ‘O˜ declension’ of feminine nouns (including the stems in -JO˜ and -WO˜), when the fact is that the N. s. of the great ‘A declen- sion’ appears without any -a ending at all and the N. s. of the great ‘O declen- 28 sion’ appears in some cases with a final -u and in other cases without any fi- nal vowel whatsoever. It is not necessary to mention here the long list of ‘A masculines and neuters’ (se cyning, se song, thaet bearn, that huµs, etc.) nor the equally long list of ‘O feminines’ (seµo giefu, seµo laµr, etc.). The puzzle pre sented by such labels (A declension, O˜ declension) is aggravated by the observation made in many handbooks that the OE. ‘A declension’ corresponds to the ‘second Latin declension’ (masculines in -US and neuters in -UM whereas the OE. ‘O˜ declenslon’ corresponds to the first Latin declenslon (feminines in -A). The situation is similar with regard to the other old Germanic languages. Schematically then we have: GERMANIC (OE. included) LATIN masculines in -A = masculines in -US (cf. Gk. OS) neuters in -A = neuters in -UM (cf. Gk. -ON) feminines in - O˜ = feminines ln -A (cf. Gk. -A) Clever students must surely be intrigued both by the ‘labels’ attached to the OE. declensions and even more so by the correspondence posed if they are not accompanied by some explanation. A knowledge of Latin does not solve the problem at all. On the contrary, it may frustrate the interested stu- dent even more. Note that Latin is a language with neat and clearly marked paradigms (first declension in -A: rosa. porta, puella …, second dclension in -US, UN: hortus, lupus, templum, donum …). as against the OE. state of af- fairs. On the other hand, the rather limited school acquaintance with the common Indo-European origin of Latin and Germanic, even though revised in most University syllabuses, is not enough to solve the riddle of the corre - spondences. It is these considerations that have moved me to make known some personal reflections on the subject of the OE. declensions, always bearing in mind the needs and frequent perplexity of our pupils. The historical data that we need have not completely disappeared from the OE. paradigms. Although scarce, they are a perennial testimony to the past. I shall 29 concentrate on the the two great declensions I have already referred to. This will confirm the validity of the correspondences and provide a justification for the ‘historical labels’. 1. A BRIEF SYNOPSIS OF THE OE. NOUN DECLENSIONS The two ‘great’ OE. declensions we have referred to are also known as the two great ‘vocalic declensions’ because of their vowel stem. They are the most important of the so-called strong declensions (the term ‘strong’ is not a particularly happy one and refers to their relative degree of inflectional com- plexity). These two ‘great’ OE. declensions and the so-called weak declension of nouns constitute the great bulk of OE. nouns. The ‘weak’ declension has an ‘N stem’ and corresponds to one of the various types of consonant stems that we have in the third Latin declension. Cf. N. s. nomen, G. s. nominis, which corresponds exactly to OE. N. s. nama , G. s. naman. From the point of view of the beginner, the OE. weak declension of nouns is easy and neat, since the N,s. is always -a in masculine nouns and -e in feminine and neuter nouns (there are only three neuter nouns in this declension). Cf., for instance, se nama , already mentioned, se moµna, se wita, seµo sunne, seµo heorte, thaet eµare, that eµage. In addition to the three great declensions considered so far, the two great strong vocalic declensions and the great weak declension (or ‘consonant de- clension’), there are other strong vocalic declensions with fewer elements. Two of these, although with comparatively few elements, are nevertheless traditionally considered ‘major declensions’ together with the three ‘great ones’. The two declensions concerned are the so-called ‘I declension’, and the so-called ‘U declension’, which correspond to the third and fourth Latin declensions respectively, both with vowel stem. (Remember that the OE weak declension in ‘N’ also corresponded to the third Latin declension but, of course, to one of its consonant varieties). As I have said, there are few 30 recorded OE. nouns that belong to these declensions. Only nine nouns are known to belong to the ‘U declension’. This is due to the fact that many of the nouns that originally belonged to this declension have joined the great ‘A declension’ and ‘O declension’ and follow their inflectional patterns.1 The same applies to the ‘I declension’. Examples of elements still remaining in the ‘I declension’ are se wine, se cyre, se lyre …, with -e as a relic of a stem in -I, seµo dæµµd, seµo speµd, seµo tȵd …, without any visible testimony of stem vowel -I. Examples of the ‘U declension’ are se sunu, se sidu, seµo duru … with maintenance of the stem vowel -U, and seµo hand with no vis ible testi- mony of stem vowel -U. The remainder of the OE. declensions, known as ‘minor declensions’, are out- side our scope. THE OLD ENGLISH NOUN DECLENSIONS (EXCEPTING THE ‘MINOR DECLENSIONS’)2 The great declensions by reason of the number of elements: ‘A declen- sion’ (masc. cyning, biscop …, neut. bearn , scip -N. Ac. pl. scipu-);’O declen- sion’ (fem. giefu, laµr, weard ); ‘N declension’ or ‘consonant declension’ or ‘weak declension’ (masc. moµna, G. s. moµnan; fem. sunne, G. s. sunnan; neut. eµare, G. s. eµaran. Other so-called ‘major declensions’: ‘I declension (masc. wine, fem. dæµµd , neut. gedrync); ‘U declension (masc. sunu. fem. duru, hand). 1 de la Cruz, J., 1986: Iniciacioån Praåctica al Ingles Antiguo, Madrid, p. 23. 2 de la Cruz J, 1983: La Prosa de los Anglosajones, Maålaga and Salamanca, p. 298. 31 2. THE OLD ENGLISH NOUN DECLENSIONS WITHIN THE INDO-EUROPEAN FRAMEWORK Needless to say, the correspondence between the ‘great’ OE. declensions and the Latin declensions is not an exclusive phenomenon of these lan- guages. It can also be found if we compare any two or more of the old Indo-European languages. In point of fact, a certain amount of ‘reconstruction’ is possible thanks to the comparative method. Such recon- struction is what enables us to say that the basic OE. noun declensions with the historical lables ‘A declension’, ‘O declension’, ‘I declension’ and ‘U de- clension’ are the OE. representatives of the Indo-European nominal flexion with vowel stem, which of course has other representatives in other Indo-European languages, for instance, Latln. The same applies to the OE. weak noun declension with consonant stem (‘N declension’). Cf. the follow- ing distribution: (I) DECLENSIONS WITH PURE STEM VOWEL Indo-European -A˜ = Germanic and OE. O˜. Note that the Indo-European vowel is long and does not correspond to Germanic short A. IDE. A˜ corre - sponds to Germanic O˜! The declension concerned corresponds to the first Latin declension. The quality changes but the quantitity is maintained (at least originally). Indo-European -O = Germanic and OE. A. Note that the Indo-European vowel is short and does not correspond to Germanic O˜. IDE. O corresponds to Germanic A! The declension concerned corresponds to the second Latin declension. The quality changes but the quantity is maintained. Indo-European -I (there is also stem vowel -I ˜ = Germanic and OE. I. The declension concerned corresponds to the third Latin declension. Indo-European -U (there is also stem vowel -U˜). The declension con- cerned corresponds to the fourth Latin declension. 32 (II) DECLENSIONS WITH CONSONANT STEM 1/ IDE. plosive, 2/ IDE. -S, 3/ IDE. nasal = Germanic and OE. N,1 4/ IDE. liquid, 5/ IDE. mixed group of 3/ and 4/. 3. CORRESPONDENCE AND CHANGE It is self-evident that a considerable percentage of the words of the Indo-European languages share one and the same root. One significant case is the Indo-European root ULQUO-/LUQUO-2 ‘wolf’ with IDE. ‘O stem’: (Singular) N. *ulquos/luquos, OE. wulf,3 L. lupus;4 Ac. *ulquom/luquom, OE. wulf, L. lupum; G. *ulquosio/luquosio, OE. wulfes, L. lupÈ µ;5 D. *ulquoµi/luquoµi, OE. wulfe, L. lupoµ; Abl. *ulquoµd/luquoµd, L. lupoµ. (Plural) 1 This is the stem of the so -called 'weak declension'. There are other consonant stemmed declensions. See reference in note 2. 2 Sanskrit vrkah, Lithuanian vilkas, and Goth. wulfs, etc., represent IDE. *ULQUO. L. lupus and Gk. lykos represent IDE. *LUQUO. The only difference between IDE. *ULQUO and IDE. *LUQUO is the metathesis 'UL-/LU-'. 3 The does not represent the regular development of IDE. *QU- as illustrated in cases llke IDE. *QUO- = Goth hwas 'who', OE. hwaµ. The of wulf is due to labial assimilation by reason of a preceding [w]. 4 This is not the regular development of IDE. *QU- in Latin. Cf. L. quis, quinque, etc., whlch lllustrate the ordinary development of IDE. *QU- in Latin as against Osco-Umbrian

: Osco-Umbrian pis = L. quis 'who'. L. lupus 'wolf' must be a loan-word from Osco-Umbrian. 5 Note the the elements in square brackets are all innovations. See explanation in the sections dealing with the first and second Latin declenslons. 33 N. *ulquôs/luquôs, OE. wulfas, L. lupi; Ac. *ulquoµns/luquoµns, OE. wulfas, L. lupoµs; G. *ulquôm/luquôm, ingl. ant. wulfa, L. lupoµrum, D. /Abl.1 OE. wulfum, L. lupÈ µs. We can see a perfect coincidence of stem and paradigm: OE. wulf and L. lupus. They do not only share the same root and stem but also follow the same declension pattern. Nevertheless, given the vicissitudes of linguistic evolution, we should not be surprised if elements which share one and the same Indo-European root do in fact follow one inflectional type in one lan- guage and quite a different one in another. An interesting case is that of OE. wylf and L. lupa ‘she-wolf’. They certainly share the same root which they in turn share with wulf and lupus. The only difference between both sets is the stem vowel. It so happens, however, that OE. wylf and L. lupa do not exactly represent the same stem. L. lupa is a comparatively recent formation with a pure ‘A stem’, whereas OE. wylf represents a stem in ‘JA˜’ = Germanic ‘JO˜’. Note also that wylf ‘she-wolf’ shows the effect of i-mutation (hence the vowel lupae, viai > viae), is an ana- logical formation which replaces the form in -as. The original Abl. s. was identical to the G. s. in all stems, the only excep- tion being the ‘O stems’, that is, the second declension. Therefore here we have another innovation. See below, in archaic Latin, the influence from the second declension. The G. pl. was not the original form. The ending -aµru m comes from the pronominal inflection IDE * -aµsoµm). The D. pl. and Abl. pl. are not the original forms either, since the majority of the Indo-European languages show either an ending with -bh or an ending with -m-. The -bh- ending is the one we see in L. -bus in the 3rd and 4th de- clensions and even in the old forms in -aµbus attested in the 1st declension, such as deaµbus and filiaµbus from dea and filia respectively, instead of 36 diÈ µs, filiÈ µs. These forms in -is are analogical and come from the 2nd declension. 37 THE SECOND LATIN DECLENSION. The ending of G. s. in Indo-European was *-sio. This is precisely the origin of the Germanic Genitive. Cf. Goth. wulfis, OE. wulfes, OSax. wulfas, etc. This is therefore the origin of the so-called ‘genitivo sajoån’ in Spanish grammars of English and which only recently is intruding into Spanish shop names in our modern cities, as the ubiquitous Paco’s and company bear wit - ness to. We leave for the moment the disputed question as to the possible adverblal origin of the old ending * -sio. The Indo-European Abl. s. was in * -eµd/-oµd. The result of -oµd was -oµ in Latin and this explalns L. lupoµ. Archaic Latin shows relics of an ending with -d- in some adverbs and also in analogical formations of the 1st and 4th declensions. Cf. the case of Abl. s. sententiaµd = Classical Latin sententiaµ. The ending of N. pl. comes from the pronominal inflection and the G. pl. is analogical of the 1st declension. For the same reason as in the 1st declension, the D. pl. and Abl. pl. are not the original forms either. The endlng -È µs may come from the Indo-European locative and/or instrumental. Note that the D. pl. endings of the Germanlc declensions point to an Indo-European instru - mental which, as in the case of the Indo-European D. pl. and Abl. pl., also possessed one variety with -bh and another with -m-. Hence the Germanic forms of the type OE. cyningum. Linguistic evolution may have even more devastating effects, to the ex- tent of completely obscuring the paradigms themselves. This is particularly the case of the two great stem vowel declensions in Old English, as we shall see in more detail in our next section. 38 4. THE CASE OF THE TWO FIRST STEM VOWEL DECLENSIONS. As pointed out in section 2., the stem vowels of these two great declen- sions appear qualitatively reversed in the Germanic languages. Moreover, they are already very deteriorated in Old English, since the OE. declensions have lost a considerable amount of phonic substance in the process from the original Indo-European inflectional pattern to the historical data of Old En - glish. In order to facilitate the visibility of this phenomenon it is necessary to reconstruct the missing steps and this we will do in the remainder of this pa- per. So, to justify the traditional nomenclature of the two great declensions as well as their cross-linguistic correspondence, the discusslon of each de- clenslon is preceded by a reconstruction synopsis with the elements which will serve as our points of reference. We shall exemplify the case of IDE. *GHEBHA ‘gift’. We illustrate first the IDE. endings, then the hypothetical IDE. forms and finally the attested Gothic and Old English forms: (INDO-EUROPEAN RECONSTRUCTION OF VOWEL STEMS IN -A˜ = GERMANIC -O˜) (Singular) N. IDE. -ø (ghebhaµ), Goth. giba, OE. giefu; Ac. IDE. -m (ghebhaµm), Goth. giba, OE. giefe; G. /Abl. IDE. -es/-os (ghebhâs), Goth. giboµs, OE. giefe; D. IDE. ei (ghebhaµi), Goth. gibai, OE. giefe. (Plural) N. IDE. -es (ghebhâs), Goth. giboµs, OE. giefa; Ac. IDE. -ns (ghebhaµns), Goth. giboµs, OE. giefa; G. IDE. - ôm (ghebhâm/ ôm), Goth. giboµ, OE. giefa; D. /Abl. IDE. (-bh-/-mr), Goth. giboµm, OE. giefum. The important correspondence ‘Indo-European A˜ = Germanlc O˜’ is absolutely clear in cases like L. maµter, OE. moµdor, or like L. fraµter, OE. broµthor. Nevertheless, such an important correspondence may appear obscure or even invislble when the vowel ln question is part of an inflectional 39 ending and as such in final position. This is due to the fact that the strong word initial stress which characterizes the Germanic family of languages tends to weaken the quality of the vowels of the inflectional endings. In some cases such strong word initial stress is responsible for thelr complete disappearance. Even so, the correspondence ‘Indo-European A˜ = Germanic O˜’ can be clearly seen in Gothlc. Cf. the Indo-European endings marked with a circum- flex accent in the reconstruction synopsis. With this accent we represent a stressing which, it is assumed, fell on some long vowels and diphthongs and prevented them from succumbing to the general process of Germanic word-end weakening. For this reason the ‘A’ which possessed such stressing (Â) survives as O˜ in final position in Germanic. So we have Goth. G. s. and N. pl. giboµs since it comes from IDE. * ghebhâs. The Ac. pl. is also giboµs from IDE *ghebhas for *ghebhaµns. However, the N. s. and Ac. s. is giba since, coming from IDE. *ghebhaµ, *ghebham, without the aforementioned stressing, does not survive as ‘O’ in Germanic. It undergoes shortening, which in the East domain of Germanic (Gothic) follows the general tendency towards opening and hence the change from ‘O’ to ‘A’. As far as Old English is concerned, N. s. giefu is due to the opposite ten- dency in West and North Germanic, whereby the shortening materializes in subsequent closeness. Hence the fact that we have ‘U’ instead of ‘A’. The contrast between giefu and laµr, both feminine nouns belonging to the same declension, is easily explained in Old English, where the ending -u is kept only in short stem elements, never in long stem elements. Remember that a stem is long when its vowel is long or when the vowel, although short, is fol- lowed by two consonants. OE. giefu is therefore an interesting testimony of the correspondence ‘Indo-European A˜ = Germanic O˜’ as well as of the correspondence ‘lndo-European feminine declension in -A˜ = Germanic feminine declension in -O˜’. We are not so lucky with the rest of the declension. Ac. s. giefe is the regular and normal result of OE. * -om: -a > -ae > -e. G. s. giefe is an analogical formation; the regular development of IDE. 40 *ghebhâs in Old English, after the common Germanic period, would have been giefa with -a. D. s. giefe is the result of regular development: -ai > -ae > -e. The development of N. pl. is parallel to that of G. s. So we have giefa . The Ac. pl. adopts the ending of the N. pl. (the regular development would have been giefe ). The development of the G. pl. is perfectly regular and it is also giefa . Lastly, in the D. pl. we have the endin g -um as in the whole of West and North Germanic. We can now appreciate the tremendous importance of Gothic and the tes - timony of the Old English N. s. of the feminine ‘O˜ declension’, when the noun is short stemmed. We refer, of course, to the invaluable witness of the -u ending. In the synopsis that follows we shall exemplify IDE. *ULQUO ‘wolf’ and IDE. *IUGO ‘yoke’. We illustrate first the IDE. endings, then the hypothetical IDE. forms and finally the attested Gothic and Old English forms: (INDO-EUROPEAN RECONSTRUCTION OF VOWEL STEMS IN -O = GERMANIC -A) (IDE. *ULQUO) (Singular) N. IDE. -s (ulquos), Goth. wulfs, OE. wulf; AC. IDE. -m (ulquom), Goth. wulf, OE. wulf; G. IDE. -sio (ulquosio), Goth. wulfis, OE. wulfes; D. IDE. -ei (ulquoµi), Goth. wulfa, OE. wulfe; Abl. IDE. -oµd/-eµd (ulquoµd). (Plural) N. IDE. -s (ulquôs), Goth. wulfoµs, OE. wulfas; Ac. IDE. -ns (ulquoµns), Goth. wulfans, OE. wulfas; G. IDE. -ôm (ulquôm), Goth. wulfeµ, OE. wulfa; D. /Abl. IDE. (-bh-/-m-), Goth. wulfam, OE. wulfum. (IDE. *IUGO) (Singular) N. IDE. -m (iugom), Goth. juk, OE. geoc; Ac. IDE. -m (iugom), Goth. juk, OE. geoc; G. IDE. -sio (iugosio), Goth. jukis, OE. geoces, D. IDE. -ei (iugoi), Goth. juka, OE. geoce, Abl. IDE. -oµd/-eµd (iugoµd). 41 (Plural) N. IDE. -a (iugaµ), Goth juka, OE. geocu; Ac. IDE. -a (iugaµ), Goth. juka, OE. geocu; G. IDE. -om (iugôm), Goth juke, OE. geoca; D. /Abl. IDE. (-bh-/-m-), Goth. jukam, OE. geocum. The correspondence ‘Indo-European O = Germanic A’ is absolutely clear in a case like L. octo ‘eight’, Goth. ahtau. (The initial diphthong in OE. eahta ‘eight’ is due, as is well known to students of Old English, to a later develop- ment). However, the situation is very different in the realm of declension endings. Here the short vowels in final position tend to disappear by reason of the strong stressing of Germanic initial syllables in general, as we have al- ready said. So, the a that we see, for example, in OE. N. pl. and Ac. pl. wulfas, G. pl. wulfa , is no witness of the correspondence ‘Indo-European O = Germanic A’. Such an a cannot be the original -a of the stem. This, being short, would have disappeared at the end of a word, exactly the same as in Goth. N. s. wulfs or OE. N. s. wulf. Cf. the second u of L. lupus, ultimately from IDE. *LUQUO for *ULQUO. Therefore, what we have in OE. N. pl. and Ac. pl wulfas, G. pl. wulfa, is the result of the contraction of the stem vowel + the ending. Cf. our previous synopses. I must conclude this section with an important observation. The N. pl. and Ac. pl. of the Indo-European neuter declension in ‘O’ coincides with the N. s. of the Indo-European feminine declension in ‘A˜’. Also the N. pl. and Ac. pl. of the Germanic neuter declension in ‘A’ coincides with the N. s. of the Ger- manic feminine declension in ‘ O˜’ if the noun in question is short -stemmed. The reason is simple. The N. Ac. pl. neuter has the IDE. ending -aµ, which was originally a collective form with singular meaning, exactly identical with the IDE. ending -a of the N. s. of the Indo-European feminine ‘A˜ declension’. So, the explanation we have given concerning giefu