Zelotes and elnvnges: The Extension of Genitive Singular -es in the Gloss to the Durham Collectar

Nieves Rodríguez Ledesma* Zelotes and elnvnges: The Extension of Genitive Singular ‑es in the Gloss to the Durham Collectar https://doi.org/10.1515/ang-2018-0061 Abstract: The aim of this article is to study the extension of genitive singular ‑es from the a-stems to other noun classes in the gloss to the Durham Collectar (Durham, Cathedral Library, A.iv.19). To this end a quantitative analysis of sixtyfive nouns has been carried out in contexts where they gloss a Latin genitive form. The nouns have been selected on the basis that their etymological inflection for the genitive singular is other than ‑es, and they consist of feminine nouns (nouns ending in ‑ung, nouns ending in ‑ness,ō‑stems, i‑stems), kinship r‑stems and weak nouns or n‑stems. The results of the analysis of the Durham Collectar are compared with those found in the Lindisfarne Gospels (London, British Library, Nero D.iv) with three main purposes: first, to determine the degree of similarity between the two glosses; second, to establish whether Durham is more conservative or not with regard to this feature, and finally, to contribute to the discussion on the question of authorship. 1 Introduction 1.1 Glosses to the Durham Collectar and to the Lindisfarne Gospels: Authorship and Relationship Some scholars have argued that the Aldred that glossed the Lindisfarne Gospels is not the same as the one who glossed the Durham Collectar. Skeat (1877–1879), for example, does not think that all the glosses in the latter are in one hand and remarks that “there is nothing to connect this Aldred with the Aldred, son of Tilwine, who glossed the Lindisfarne MS”(55). 1 Similarly, Lindelöf (1927) notes *Corresponding author: Nieves Rodríguez Ledesma, University of Sevilla E-Mail: [email protected] 1With regard to the date of the Durham Collectar, Skeat notes that it cannot be later than the close of the tenth century. In a footnote he adds that Dr. Murray and Mr. Maunde Thompson Anglia 2018; 136(4): 611–642 that there are differences in the writing of some letters, such as <ð>, as well as sufficient divergences in spelling, use of abbreviations, grammar and vocabulary “to render it highly improbable that they could be the work of the same man”(lii). Ker (1943), however, remarks that “the similarities are obvious”(8), the linguistic differences “slight”(10), and that the differences in appearance may be attributed to the formality of the gloss to the Gospels (9). Based on palaeographical evidence, he concludes that the gloss to the Gospels and the gloss to the Ritual [Durham Collectar] are the work of the same scribe, and he identifies Aldred the son of Alfred with Aldred the Provost (11–12). 2 Other scholars who identify Aldred as the glossator of both Lindisfarne and Durham include Brown et al. (1969), Jones (1970), Ross (1970, 1978, 1982) and, more recently, Jolly (2016) and Roberts (2016). Although written later, the gloss to the Durham Collectar has been shown to be more conservative than that to the Lindisfarne Gospels. Ross describes five points in which Durham is more conservative: the retention of the dative singular in ‑ein what he refers to as the ‘Type α’declension (Ross 1960: 39); the maintenance of the distinction between the third person singular present indicative and the plural present indicative (gefeð,‑es vs. geafað,‑as), and between the infinitive and the present subjunctive as regards the vowels of the inflections (geafa vs. gefe); the absence of ðy for the feminine form of the demonstrative; the retention of ‑ð(vs. ‑s) in the third singular and plural present indicative of the verb, 3 and the virtual absence of ‑e‑in the preterite of verbs of the second weak class, ‑a‑being the normal form (Ross 1970: 363–364). On the other hand, he mentions two features in which Durham is less conservative than Lindisfarne: in the n‑stems, ‑efor ‑ais much more frequent than in Lindisfarne, and in the present subjunctive ‑afor ‑eis also more frequent (365). As a possible explanation, he argues that some of the conservatisms may be due to West Saxon influence, since the colophon to Durham was written at Oakley, between Salisbury and Blandford, and others to the fact that, when glossing Durham, Aldred had “settled down”and decided which variants to favour among those he had at his disposal (365–366). conclude that the glosses are not by the same hand and that the Ritual gloss is “somewhat earlier than that of the Gospels”(1877–1879: 55). 2For a more detailed account of the controversy concerning the identity of the glossator, cf. Brown et al. (1969: 23–29). 3Although s‑forms are also found in Durham, they are proportionately much rarer than in Lindisfarne. Ross (1970: 364) gives the following figures, based on Blakeley (1949–1950): 1519 ð‑forms and 1441 s‑forms in Lindisfarne, as against 331 ð‑forms and 48 s‑forms in Durham. See also Cole (2014) for the distribution of these forms in Lindisfarne. 612 Nieves Rodríguez Ledesma 1.2 Previous Studies on the Genitive Previous studies on the gloss include general remarks on the genitive. In his study on the development of the grammatical category of gender in the glosses to the Lindisfarne Gospels and to the Durham Collectar, Jones (1967) focuses on modifiers, rather than nouns. He remarks that the theory of ‘neutralization’put forward by Ross (1936) to account for the unhistorical use of the ꝥ form in these glosses is “not sufficiently powerful”(1967: 101), and observes a tendency for the inflections ‑ne,‑(e)s and ‑(u)m to spread in the accusative, genitive and dative cases respectively, regardless of the etymological gender of the noun (102). His hypothesis is that “at this period there appears to be a developing tendency for the language to generate sentences the relation of whose parts is characterized by the use of certain inflexional types of pre and post head modifiers regardless of the gender of the head word itself in each case”(102). With regard to the genitive, Jones (1967: 105) observes an overwhelming use of the‑(e)s form of both pre- and post-head modifiers with historically femininenouns in Lindisfarne, “a spread which was probably analogous to a similar one in the genitive inflexion of the head word in the nominal group itself”(105–106). Feminine forms of modifiers, however, are also attested in this gloss with historically masculine and neuter nouns so that “at least for a time, both ‑(e)s and ‑re forms were being used as markers of possessive relationships between nominal groups” (106). 4 The situation in Durham with regard to the genitive, however, is different in two respects (Jones 1967: 108): on the one hand, the spread of the ‑(e)s form of modifiers to indicate the possessive relationship between nominal groups of all genders is found to a greater extent than in Lindisfarne. On the other, there is no spread in this gloss of genitive ‑re forms with historically masculine and neuter nouns. 5 These two factors seem to indicate that the generalization of ‑(e)s as possessive marker has advanced one step further in Durham. 4The figures for the genitive singular of the definite article in Lindisfarne are the following: a) with etymologically masculine nouns: 89 tokens of ðæs and 5 tokens of ðære; b) with etymologically feminine nouns: 14 tokens of ðæs and 7 tokens of ðære; c) with etymologically neuter nouns: 34 tokens of ðæs and 5 tokens of ðære (Jones 1967: 111). 5The figures for the genitive singular of all modifiers in Durham are the following: a) with etymologically masculine nouns: 138 tokens of ‑(e)s forms and 0 tokens of ‑re; b) with etymologically feminine nouns: 73 tokens of ‑(e)s forms and 40 tokens of ‑re; c) with etymologically neuter nouns: 53 tokens of ‑(e)s forms and 0 tokens of ‑re (Jones 1967: 111). Gen. sg. -es in the Durham Collectar Gloss 613 Focusing more specifically on nouns, Ross (1971: 57) remarks with regard to the s‑genitive that “analogical extension is wide-spread in Ritual as in L[indisfarne]”and gives examples of nouns belonging to different declensions, such as: gefes,frofres,ricsunges,sibbes,dedes,lichomes,earðes, instead of the etymologically expected forms gefe,frofre,ricsunga,sibbe,dede,lichoma(n),earða(n) (57). 6 However, he points out a difference between the two glosses regarding the nouns ending in ‑ness: in Durham the genitive singular in ‑es coexists with that in ‑e, and he remarks that “here, in contradistinction to his practice in L[indisfarne], Aldred has no feeling against ‑es”(1971: 56). 7 1.3 Aims and Methodology This article offers a quantitative study of the genitive singular inflection in the gloss to the Durham Collectar. In a previous work on the gloss to the Lindisfarne Gospels I focused on zero or uninflected genitives and on the word order of adnominal genitives, and showed that the gloss is innovative at the morphological level, as shown in the extension of genitive singular ‑es from the a‑stems to other noun classes, including proper nouns, which tend to add ‑es regardless of the gender of the name and of the ending of the Latin original (Rodríguez Ledesma 2016). The aim of this article is to establish the actual scale on which this phenomenon –the extension of genitive singular ‑es from the a‑stems to other noun classes –is found in the gloss to the Durham Collectar. It is in line with Jones’ work on modifiers (1967), but focuses on nouns and, although mainly descriptive, it suggests some factors which may have conditioned the choice of a particular inflection (etymological vs. innovative ‑es) in some cases. To this end a quantitative analysis of sixty-five nouns has been carried out in contexts where they gloss a Latin genitive form. The nouns have been selected on the basis that their etymological ending for the genitive singular is other than ‑es, and they consist of feminine nouns (nouns ending in ‑ung, nouns ending in ‑ness, ō‑stems, i‑stems), kinship r‑stems and weak nouns or n‑stems. A reading of some sections of the text (Lindelöf 1927) revealed examples of the extension of genitive 6With regard to Lindisfarne, Ross (1937: 99) remarks that the ending of the genitive singular ‑es “was extended to practically all classes”and gives the following examples: lufes,saules,sibbes, brydes,oxes,widues. He adds that “in all these classes this must be considered as the normal form of the gen. sg. though older forms are often preserved also”(1937: 99). 7In her study of the reshaping of the nominal morphology in early Northern West Germanic, Adamczyk (2018) also deals with the extension of the genitive singular ending ‑es to other nouns in Old English. 614 Nieves Rodríguez Ledesma singular ‑es to feminine and weak nouns, and these were supplemented by other nouns included in Lindelöf’s glossary (1901) which did not originally belong to the a‑stems and were attested in the genitive singular in the gloss. The Dictionary of Old English Web Corpus (hereafter DOEC), based on Lindelöf’s edition, has been used to collect the data. 8 The original idea was to collate all the tokens with the facsimile copy of the manuscript (Brown et al. 1969), but the quality of the facsimile, although good for the Latin text, is not so for the gloss and many sections are not readable. 9 All the tokens have nevertheless been checked with Lindelöf’s edition (1927), which is quite faithful to the manuscript. 10 Lindelöf (1923) and Squires (1971) have also been used. With regard to the methodology and because of the wealth of spelling and morphological variants attested in the gloss, the DOEC was searched for the Latin genitive form of the nouns which form the basis of the present study. Thus, for the kinship r‑stems, the terms of search were L. patris,matris,fratris,sororis and filiae. In those instances in which the Latin noun has the same form for several cases (nominative and genitive natalis, genitive and dative misercordiae), a careful examination of the syntax and the context of all the examples obtained was necessary in order to identify the genitive forms. Following the DOEC, the examples given throughout the article offer both the Northumbrian gloss and the Latin text, to which a word-for-word translation has been added. A problem for this study is posed by the numerous abbreviations and contractions found in Durham, 11 and by the fact that “the glossator does not seem to have followed any clear principle with regard to the PLACE he gives the waved stroke when used as a general mark of abbreviation or contraction”(Lindelöf 1927: 204). In the case of nouns ending in ‑nisse, for example, the abbreviation mark is sometimes placed over the <s>, sometimes over the <i>, sometimes over the <n>, and often it covers two or more letters (Lindelöf 1927: 204). There are also numerous instances in which the glossator used an abbreviation mark without any apparent reason. Among the examples Lindelöf (1927: 205) 8The title abbreviations and editions of all the Old English texts mentioned in this article are those employed by the DOEC. 9As Campbell (1973: 259) remarks: “On the plates the Latin text is usually clear, but the gloss, which is to Anglo-Saxon scholars the most interesting item in the manuscript, is often obscure, sometimes totally illegible”. 10 Cf. Squires (1971: 362): “In preparing this collation I have made a transcript of the manuscript [...] and this has shown how very reliable is the standard edition by U. Lindelöf, Rituale Ecclesiae Dunelmensis”. 11 Cf. Brown et al.’s (1969: 26) remark: “In the ‘Ritual’gloss, as one would expect in a liturgical text in which formulas are often repeated, Aldred makes more use of capricious abbreviation”. Gen. sg. -es in the Durham Collectar Gloss 615 gives to illustrate this “irrational use of the waved stroke”, he mentions the placing of a curl over one of the last letters of some genitive singular forms ending in ‑es, such as gidvoles (34.14), hæles (35.4), wætmes (105.9) or ymbstondennisses (182.9). Squires (1973: 405) remarks that “certain forms and certain cases seem particularly prone to be misabbreviated”, the genitive singular being among these, with about twenty per cent of the instances found. According to her, “the fact that the glossator seems to be selective in his errors would suggest [...] that there is some reason for his tendency to make them”(405–406), and she puts forward the hypothesis that an awareness of the variation existing in the genitive singular inflection may have caused the glossator “either to make mistakes or to express the termination as vaguely as possible, because he was uncertain of the correct form”(407). The presentation and analysis of the data is organized as follows: feminine nouns are studied first (2), and they are subdivided into four categories: nouns ending in ‑ung (2.1), nouns ending in ‑ness (2.2), ō‑stems (2.3) and feminine i‑stems (2.4). Kinship r‑stems are analysed in (3) and weak nouns or n‑stems in (4). The genitive singular of some adjectives is studied in (5) and finally some conclusions are provided in (6). In each section the results of the analysis of Durham are compared with those found in Lindisfarne with three main purposes: to determine the degree of similarity between the two glosses, to establish whether Durham is more conservative or not with regard to the genitive singular inflection and, finally, to contribute to the discussion on the question of authorship. 2 Feminine Nouns 2.1 Nouns Ending in ‑ung These nouns belong to the historical ō‑stems. Hogg and Fulk (2011: 28) note that there are two important processes of derivation associated with ō‑stems. The first, which is very frequent, forms deverbal nouns by adding the suffixes ‑ung/‑ing. The suffix ‑ung dominates in derivatives of weak verbs of class II, and the suffix ‑ing in derivatives of strong verbs and of weak verbs of class I (Hogg and Fulk 2011: 29; see also Campbell 1959: 158). In North Northumbrian these nouns often take ‑es in the genitive singular, with these forms outnumbering instances with ‑a, but the same is not true of South Northumbrian (i. e. Rushworth 2 ) (Hogg and Fulk 2011: 111). In Lindisfarne there are also examples of nominative and accusative plural forms ending in ‑as, such as ebolsungas ‘blasphemies’(2011: 112, n. 3). Ross (1971: 54) also notes the form bloedsungas in Lindisfarne. 616 Nieves Rodríguez Ledesma Table 1 12 gives the list of nouns belonging to this group which are attested in the genitive singular in Durham, together with the Latin form they gloss, and the number of occurrences of the different inflections. Table 1: Genitive singular of nouns ending in ‑ung n‑sn‑a/e(Etymological Inflection) n Abbreviation n Total bloedsvng (L. benedictio,‑onis) 3–14 eftniwvng 13 (L. reparatio,‑onis) 2––2 forescea(w)vng (L. providentia,‑ae) 1––1 hvoerflvnges 14 (L. vicissitudinis) 1––1 ricsvng (L. dominatio,‑onis) 2––3 15 somnvng (L. congregatio,‑onis) 1––1 ðingvng (L. intercessio,‑onis) 1–12 ðrovng (L. martyrium,‑ii, L. passio,‑onis) 16 13 ––13 12 In the Tables the forms are given in accordance with the relevant headword in Lindelöf’s glossary (1901). All the examples throughout the article are actual attestations that reflect the special graphematic and spelling conventions of Durham. 13 The only examples of the spelling <eft‑> given in Bosworth and Toller (BT) and the Dictionary of Old English (DOE s.v. eftniwung) are those from Durham. The form edniwung is attested in other texts, including Lindisfarne (gloss to Matthew) (DOE s.v. edniwung). 14 In Lindelöf’s glossary (1901), based on Stevenson (1840), the headword is echvoerflvnges, but in his edition it is amended to hvoerflvnges (DurRitGl 1 [Thomp-Lind] 28.6). The genitive form is the only attestation of this noun in Durham. 15 Besides the two inflected forms, there is also one token in which the genitive is uninflected: in ælcvm stove ricsvng ðæs (DurRitGl 2 [Thomp-Lind] 171.9). It is found in the same context as the other two examples, glossing L. in omni loco dominationis ipsius ‘in every place of your kingdom’. Lindelöf (1901: 179) wonders whether it is the nominative singular (“ns. (?) –171, 9”), but I think it could be considered an example of a zero genitive, this being a feature of Northern dialects and attested, for example, in Lindisfarne (cf. Rodríguez Ledesma 2016). 16 Seven tokens of OE ðrovng gloss L. gen. passionis, and one glosses L. ablat. passione (DurRitGl 1 [Thomp-Lind] 23.3). Four tokens gloss L. gen. martyrii, and one glosses ablat. martyrio (DurRitGl Gen. sg. -es in the Durham Collectar Gloss 617 n‑sn‑a/e(Etymological Inflection) n Abbreviation n Total worðvng (L. honor,‑oris) 3––3 wvldrvng (L. glorificatio,‑onis) 1––1 Total 28 (90.32 %) –2 (6.45 %) 31 As can be seen in the table, no example with the etymological inflection is attested for the genitive of these nouns in Durham, and very few instances are found with an abbreviation mark. On the contrary, they tend to take the analogical ending ‑es whenever they are found in the genitive singular: 28 out of a total of 31 occurrences (90.32 %). Most of these nouns are found glossing Latin nouns belonging to the third declension, whose genitive singular ends in ‑is; so there is the possibility of influence of the Latin inflection (the <s> in Latin triggering an <s> in Old English). However, ‑es is also found when rendering Latin nouns ending in a vowel, as is the case with fʹesceavnges, glossing L. providentiae, and ðrovnges, which is found four times glossing L. martyrii. In a few examples the noun is modified by a possessive or a demonstrative, following Latin. In some of these instances, the analogical genitive form ending in ‑es is modified by the feminine form of the possessive, with no internal gender agreement within the noun phrase. This is the case with somnvng: (1) gimyndig voes ðv somnvnges ðinræ þ ðv gisceope fro' frvma Memor esto congregationis tuæ quam creasti ab initio ‘Be mindful of your congregation, whom you created from the beginning’ DurRitGl 2 [Thomp-Lind] 173.2 In other instances, however, the possessive also takes the ending ‑es. This is the case with bloedsvng: there are three tokens of the noun phrase bloetsvnges ðines gefe, twice glossing L. benedictionis tue gratiam (DurRitGl 1 [Thomp-Lind] 81.20, 91.19), and once glossing L. benedictionis tuae dono (DurRitGl 1 [Thomp-Lind] 95.6). 17 The Table 1: (continued) 1 [Thomp-Lind] 80.8). In the two cases in which it glosses a Latin ablative, there may have been some confusion because it is preceded by nouns in the genitive. 17 The example in which the abbreviated form is found is also modified by the possessive ðines: ðines bloedsʹ(DurRitGl 1 [Thomp-Lind] 121.14). In other contexts, when glossing an ablative, this 618 Nieves Rodríguez Ledesma noun forescea(w)vng is also modified by the historically masculine/neuter form of thepossessiveinthegenitivesingular:fʹesceavngesðines,glossingL.prouidentietuę (DurRitGl1[Thomp-Lind]110.8). 18 In Durham, therefore, ‑es has been generalized as the genitive singular inflection of feminine nouns ending in ‑ung, to the extent that no example of the etymological ending is attested in the gloss. This inflection, which originally marked jointly case and gender, is now being used as a grammatical function marker (genitive case), and can be added to both the noun and the possessive in a noun phrase (as in the last examples) or just to the noun (as in (1)). 19 The generalization of ‑es as a marker of genitive singular can clearly be seen with the example which gives the title to this article: elnvnges, glossing L. zelotes (DurRitGlAbbrev 192.4). This form is somewhat curious, since the base form elnvng is attested rendering L. zelus (DurRitGlAbbrev 192.3). There seems to have been a mistake by the scribe, since zelus and zelotes are not forms of the same noun in Latin, but are glossed as though they were, probably because in the dialect of the glossator the ending ‑es has become a marker of genitive singular, regardless of gender and declension class, and that has been applied to Latin as well. 20 In Lindisfarne two of these nouns are attested in the genitive: ðrouung (3x) and somnung (2x), 21 and both take the analogical ending in all instances: ðrounges,ðrouunges,ðrowunges,somnunges. 2.2 Nouns Ending in ‑ness These nouns were historically jō‑stems, the abstract suffix ‑ness being the most frequent jō‑stem suffix (Hogg and Fulk 2011: 33). This is the usual form in West noun is also modified by the historically masculine/neuter form of the adjective: ælcvm bloedsvnge gastlicvʹ(DurRitGl 1 [Thomp-Lind] 110.13). 18 When glossing an ablative, this noun is also modified by the historically masculine/neuter form of the possessive: fʹesceavvnge ðinvʹ, glossing L. prouidentia tua (DurRitGl 1 [Thomp-Lind] 108.12). 19 This is in line with Jones’s results concerning determiners and modifiers in Lindisfarne and Durham (1967 and 1987). 20 Cf. DOE s.v. elnung 3.: “in an anomalous gloss: glossing zelotes ‘one that loves with jealousy, one that is jealous’apparently as if a form of zelus ‘zeal, fervour’–DurRitGlAbbrev 192.4b: zelotes elnvnges (the glossator seems to have taken the Lat. as gen. of the preceding word zelus [...]; perhaps the sin elnunges shows the influence of the Lat. ending)”. 21 The Old English forms in Lindisfarne are given in accordance with the relevant headword in Ross and Stanley’s glossary (1960). The inflected forms reflect the exact spellings attested in the gloss. Gen. sg. -es in the Durham Collectar Gloss 619 As with the previous group, many of these nouns are found glossing Latin forms ending in ‑is; so there is the possibility of influence of the Latin inflection (the <s> in Latin triggering an <s> in Old English). This hypothesis seems to be confirmed by the following example: (15) no in daerstv' aldv' ne æc in dærstv' yfelgiornisse ⁊vnwisnise ah on dærstv' ł on ðearfv' bilvitnises ⁊soðfæstnises non in fermento ueteri neque in fermento malitiæ et nequitiæ sed in azymis sinceritatis et ueritatis ‘not with the old yeast, the yeast of malice and evil, but with the unleavened bread of sincerity and truth’ DurRitGl 1 [Thomp-Lind] 25.9–11 There are four nouns in the genitive: those glossing a Latin form ending in a vowel end in a vowel as well (‑e, the etymological inflection): yfelgiornisse and vnwisnise, glossing L. malitiæ and nequitiæ. By contrast, those glossing a Latin noun whose genitive ends in ‑is take ‑es:bilvitnises and soðfæstnises, glossing L. sinceritatis and ueritatis. Although influence of the Latin original is clear in (15), the inflection ‑es is also found when glossing Latin nouns whose genitive does not end in ‑is. This is the case with bihaldennisses (L. observantiae,continentiae), gicyðnisses (L. testamenti), hernisses (L. ministerii), miltheartnisses (L. misericordiae) and soðfæstnises (L. iustitiae). Some of the nouns included in Table 2 above are attested in the genitive in Lindisfarne: accenisse (2x), cyðnisse (3x), hernisse (3x), miltheartnisse (1x), selenisse (3x), soðfæstnisse (3x), unrehtuisnisse (3x). Genitive singular forms in ‑es are attested for the following: cyðnisse (1x), hernisse (3x), selenisse (1x), soðfæstnisse (2x), unrehtuisnisse (3x). The main difference is that the nouns accenisse and miltheartnisse do not take ‑es in Lindisfarne, whereas they do in Durham (cf. example (4)), which confirms Ross’statement (1971: 56) quoted above, in Section 1.2. The extension of ‑es to these nouns, therefore, has advanced further in Durham. 2.3 ō-Stem Nouns This group consists of the historical ō‑stem nouns, including the jō‑and wō‑stems, and is the chief declension for feminine nouns in Old English. In Late West Saxon they have ‑ethroughout the singular, except for the nominative (‑u/‑Ø), and ‑athroughout the plural, except for the dative (‑um). Light-stemmed nouns also show examples of extension of ‑uthroughout the singular: gifu,lufu,talu. 626 Nieves Rodríguez Ledesma Hogg and Fulk (2011: 110, fn. 2) note that all the instances found are accusative singular, except for lufu, which is dative singular, and remark that they have not found any examples of ‑ufor the genitive singular in Late West Saxon, “and this absence is probably not accidental”. There is, however, one token of lvfv glossing L. dilectionis in Durham (DurRitGl 1 [Thomp-Lind] 95.13), which may be considered an example of a zero genitive, this being a feature of Northern dialects in Middle English and attested in late Northumbrian as well (cf. Rodríguez Ledesma 2016). Hogg and Fulk (2011: 109–210) also remark that in Late West Saxon nouns of this declension may occur with the genitive singular ending of the a‑stems in a small number of idiomatic constructions, and give examples such as helpes biddende,helpes tid,sybbes lufu. In Durham, however, instances of ‑es genitives for these nouns are not restricted to idiomatic constructions, but are generalized, as can be seen in Table 3 below. This confirms Hogg and Fulk’s observation that “a particular characteristic of NNbr is the widespread use of ‑es as the marker of the gen.sg., such forms greatly outnumbering instances with ‑e”(2011: 111). However, they argue that “this is not to be taken as a shift in gender, but rather the generalization of the ‑es inflexion outside the original as-declension”(ibid.). Table 3 gives the list of ō-stem nouns which are attested in the genitive singular in Durham (excluding nouns with the suffixes ‑ung and ‑ness), together with the Latin form they gloss, and the number of occurrences of the different inflections: Table 3: Genitive singular of the historical ō‑stem class, including the jō‑and wō‑stems n‑sn‑e(Etymological Inflection) n Abbreviation n Total ældo (L. aetas,‑atis) 2–13 gefe (L. gratia,‑ae) 4155 35 24 35 Although some abbreviations could stand for either the etymological inflection ‑eor ‑es, others seem to stand for ‑es. This is the case with gefeʹ, which is attested five times glossing a genitive form. In two instances the noun is modified by the historically masculine/neuter form of the modifier (ðines, DurRitGl 1 [Thomp-Lind] 16.15; ecelices, DurRitGl 1 [Thomp-Lind] 35.9), and in another two by the feminine form of the possessive, ðinrae (DurRitGl 1 [Thomp-Lind] 98.5, 123.3). There is one token with no modifier. Gen. sg. -es in the Durham Collectar Gloss 627 n‑sn‑e(Etymological Inflection) n Abbreviation n Total hælo (L. salus,‑utis, L. salvatio,‑onis) 12 –4 36 17 37 lar (L. consilium,‑ii) 1––1 lvfv 38 (L. fides,fidei, L. dilectio,‑onis) 6–3 39 10 rod (L. crux,crucis) 81413 savel (L. anima,‑ae) 9––9 sibb (L. pax,pacis) 12 ––12 snytro (L. sapientia,‑ae) 4–15 strengo (L. fortitudo,‑inis) 1––1 Table 3: (continued) 36 The abbreviated forms are hælʹ(DurRitGl 1 [Thomp-Lind] 28.18), hælesʹ(DurRitGl 1 [Thomp- Lind] 35.4), hæloʹ(DurRitGl 1 [Thomp-Lind] 33.15) and hæleʹ(DurRitGl 1 [Thomp-Lind] 99.8). The second and the last forms probably stand for hæles, with the extension of ‑es. According to Squires, the form hæloʹ“might well seem convenient shorthand to the glossator for hælo uel hæles or hæle”, the abbreviation mark indicating not variant spellings, but alternative grammatical forms (Squires 1973: 407). 37 There is also one token of hælo. Hogg & Fulk (2011: 111) notice that in the case of light stems, “the nom.acc.dat.sg. all most frequently show ‑o, which varies with ‑u,‑a”and argue that “it seems possible that these case forms have fallen together and that the observable vowel variation is due to confusion amongst unstressed back vowels”. They do not mention, however, that this vowel is also found in the genitive singular of these nouns in Northumbrian. 38 Originally OE lufu was a weak noun, but it was also declined according to the strong feminine declension (cf. Campbell 1959: 251). According to Hogg and Fulk (2011: 110, fn. 1), “in WS, lufu generally declines as an a-declension noun, but there is a minority of forms from the an-declen- sion, most notably in the dat.sg. In Angl it declines as a member of the an‑declension”. In Durham, however, it does not decline as an an‑declension noun, but shows extension of ‑uthroughout the singular. 39 The abbreviated forms are lvfeʹ(DurRitGl 1 [Thomp-Lind] 49.12) and lvfvʹ(DurRitGl 1 [Thomp- Lind] 93.2, 93.3). As in the case of hæloʹdiscussed above (cf. fn. 36), the form lvfvʹmay indicate variation in the genitive singular of this noun: lvfv or lvfes. 628 Nieves Rodríguez Ledesma n‑sn‑e(Etymological Inflection) n Abbreviation n Total synn (L. peccatum,‑i) 2–13 ðiostre (L. tenebra,‑ae) 40 1––1 wræððo (L. iracundia,‑ae, L. indignatio,‑onis) 11–7 41 Total 63 (59.43 %) 17 (16.03 %) 19 (17.92 %) 106 As can be seen, forms with ‑es for the genitive of these nouns are dominant in Durham (63x, 59.43 %), followed by abbreviated forms (19x, 17.92 %), and finally by forms with the etymological inflection (17x, 16.03 %). Looking at the results more closely, however, there is one noun which stands out and behaves differently from the others in favouring the etymological inflection over innovative ‑es: gefe. So the results seem to be somewhat skewed. As with the previous class, in some instances the grammatical context seems to have conditioned the choice of a particular inflection: ‑evs. ‑es. This is the case with gefe, the noun which stands out in this group: all the examples which take etymological ‑e(15x) gloss a Latin noun modified by a possessive or an adjective. There are eight tokens glossing L. gratiæ tuæ, four of which are modified by the feminine form of the possessive: gefe ðinrae (DurRitGl 1 [Thomp-Lind] 37.1, 37.20), gefe ðinrę(DurRitGl 1 [Thomp-Lind] 71.20, 100.1). As illustrated in Section 2.2 above, the feminine form of the possessive (ðinræ) triggers the etymological inflection ‑eas against innovative ‑es. In the remaining four examples, the noun is modified by the historically masculine/neuter form of the possessive, ðines, with no internal gender agreement within the noun phrase: geafæ ðines (DurRitGl 1 [Thomp-Lind] 4.11), gefe ðines (DurRitGl 1 [Thomp-Lind] 7.10, 8.4, 23.10). There are also six instances in which the noun takes the etymological inflection ‑eand Table 3: (continued) 40 Campbell (1959) considers þeostru as an ō‑stem noun, beside strong neuter þeostre (1959: 237). Lindelöf (1901: 203) does not give the gender of the noun. 41 There are also five forms ending in ‑o:wræððo (3x), vræððo (2x), which may be considered examples of zero or uninflected genitives (cf. Rodríguez Ledesma 2016). These forms have been extended throughout the singular and are found glossing Latin accusative, genitive and ablative forms, both in Durham and Lindisfarne (cf. Ross 1971: 55). Gen. sg. -es in the Durham Collectar Gloss 629 is modified by an adjective ending in ‑es, as in: gefe gastlices, glossing L. gratię spiritalis (DurRitGl 1 [Thomp-Lind] 99.7–8) or godcvndes gefe, glossing L. diuinę gratiæ (DurRitGl 1 [Thomp-Lind] 121.16). In all these examples it is the modifier, either with etymological ‑re or with innovative ‑es, that unambiguously marks the grammatical function of the noun phrase. As against the 15 tokens of gefe with the etymological inflection in the genitive singular, there are four examples with ‑es:gefes. Two of them are modified by an adjective: gefes heofonlices, glossing L. gratiæ celestis (DurRitGl 1 [Thomp-Lind] 17.4), heofonlic gefes, glossing L. caelestis gratiæ (DurRitGl 1 [Thomp-Lind] 69.7), and the other two have no modifier: giselenise gefes, glossing L. donatione gratie (DurRitGl 1 [Thomp-Lind] 52.5), and ðing gefes, glossing L. munus gratie (DurRitGl 1 [Thomp-Lind] 63.11). In the case of wræððo, the six examples which take a vowel (as against ‑es)in the genitive singular gloss a Latin noun modified by the possessive tuae. 42 Four gloss L. iram tuæ indignationis and are modified by the feminine form of the possessive, ðinræ (DurRitGl 1 [Thomp-Lind] 40.18, 41.8, 42.17, 43.12). The remaining two gloss L. flagella tuae iracundiæ and are modified by the historically masculine/neuter form of the possessive, ðines, with no gender agreement within the noun phrase (DurRitGl 1 [Thomp-Lind] 8.18, 15.13). As in the case of gefe,in these examples it is the possessive, either with etymological ‑re or with innovative ‑es, that unambiguously marks genitive singular. Interestingly, the only instance of this noun which takes the genitive singular inflection ‑es glosses a Latin noun not modified by the possessive tuae: (16) in tide wræððes in tempore iracundię ‘in time of wrath’ DurRitGl 1 [Thomp-Lind] 88.4 As with gefe and wræððo, in the case of rod the only token with ‑eoccurs when glossing a Latin noun modified by tuae, the possessive ðinræ triggering etymological ‑e(as against innovative ‑es): (17) tacne ðære hælga rode ðinræ signo sanctę crucis tuae ‘with the sign of thy holy cross’ DurRitGl 1 [Thomp-Lind] 120.5 42 The only token of genitive singular lvfv (ending in a vowel) is also found glossing a Latin noun modified by the possessive tuae:lvfv ðinre, glossing L. dilectionis tuae (DurRitGl 1 [Thomp-Lind] 95.13). 630 Nieves Rodríguez Ledesma As with the previous classes, the Latin original may have conditioned the choice of ‑es with some of these nouns, the <s> in Latin triggering an <s> in Old English. That could be the case with hælo,rod or sibb, which gloss Latin nouns whose genitive ends in ‑is:salutis,crucis and pacis respectively. In this line, it could be claimed that the high number of occurrences of etymological ‑ewith gefe (15x) may be due to the fact that it does not gloss a Latin form ending in ‑is, but in a vowel (gratiae). However, as explained above, the grammatical context does play a role, since all these tokens are modified by a possessive or an adjective ending in ‑re or ‑es which unambiguously mark genitive case. Another reason for the high number of instances of etymological ‑ewith gefe could be its frequency of occurrence, since frequent nouns tend to preserve historical forms and resist analogical processes. Although ‑es has been generalized as the genitive singular marker of these nouns, they retain the etymological inflection ending in a vowel in the nominative and accusative plural, as illustrated by lar, which takes the ending ‑o;savel, with the forms savlo,savelo;sibb, with the form sibba; and synn, with the forms synna, synno (cf. Lindelöf 1901, s.vv.). Seven of the nouns included in Table 3 above are attested in the genitive in Lindisfarne: hælo (2x), lar (1x), lufu (3x), rod (1x), sauel (3x), sibb (3x), synn (2x). Except for lar, genitive singular forms in ‑es are attested for all of them: hælo (hæles 1 x, hælo 1x), lufu (lufes 2 x, lufæ 1x), rod (rodes 1x), sauel (saules 3x), sibb (sibbes 2 x, sibb 1x), synn (synnes 2x) (cf. Ross and Stanley 1960). Although the number of tokens is much lower than in Durham (15x vs. 65x), there is also a preference for ‑es in Lindisfarne: 11 instances, as against one of the etymological inflection, and three of uninflected genitive forms. Unfortunately, gefe, the noun which behaved differently from the others in favouring the etymological inflection over ‑es in Durham, is not attested in the genitive singular in Lindisfarne. 2.4 Feminine i-Stems Although historically belonging to a different stem class, at a very early date the masculine and neuter i‑stems adopted the inflections of the a‑stems, and the feminines adopted the inflections of the ō‑stems so that “even in early Old English, their membership in that stem class is primarily an historical fact rather than an indication of their structure in OE”(Hogg and Fulk 2011: 37). Table 4 gives the feminine i‑stems which are attested in the genitive singular in Durham, together with the Latin form they gloss, and the number of occurrences of the different inflections. Gen. sg. -es in the Durham Collectar Gloss 631 Table 4: Genitive singular of feminine i‑stems n‑sn‑e(Etymological Inflection) n Abbreviation n Total æ (L. lex,legis) 2––2 ded (L. actio,‑onis) 1––1 mæht (L. potentia,‑ae, L. virtus,‑utis) 3–14 portcvoen(e) (L. meretrix,‑icis) 1––1 tid (L. tempus,‑oris) 1––1 worvld (L. saeculum,‑i) 2–14 43 Total 10 (76.92 %) –2 (15.38 %) 13 Although few tokens of these nouns in the genitive singular are attested in Durham (13x), most of them take ‑es (10x, 76.92 %), no example having been found with the etymological inflection. Forms with abbreviation mark are also very scarce (2x, 15.38 %). With regard to æ, Campbell notices that it has an indeclinable singular and nominative and accusative plural, genitive plural æa, and also accusative, genitive and dative singular æwe. Late Northumbrian has genitive singular æs, and in late West Saxon the form æys is once found (Campbell 1959: 244). Two forms are attested in Durham for the genitive singular of this noun: æes (DurRitGl 1 [Thomp- Lind] 6.19) and aes (DurRitGl 1 [Thomp-Lind] 100.7). As with the previous classes, influence of the Latin original may be adduced to account for the ‑es tokens of some of these nouns, since most of them gloss Latin forms ending in ‑is. However, ‑es is also found glossing Latin nouns whose genitive ends in a vowel, as in the case of vorvldes (L. seculi)ormæhtes (L. potentiae). 43 There is also one token of a zero genitive: in worvlde vorld, glossing L. in sæculum sæculi (DurRitGl 2 [Thomp-Lind] 169.3). 632 Nieves Rodríguez Ledesma Four of these nouns are attested in the genitive in Lindisfarne: ae (19x), maeht (1x), tid (6x) and uorold (11x), and all of them take genitive singular forms in ‑es. In fact, all the tokens are with ‑es (36x), except for one instance of uninflected ae. In Lindisfarne the number of tokens is much higher than in Durham (37x vs. 13x), but in both glosses there is widespread extension of the genitive marker ‑es with these nouns, no example being attested with the etymological inflection. 3 Kinship r-Stems In Durham only two nouns belonging to this declension are attested in the genitive singular: moder and fæder. The first is found only once, glossing L. matris, with the abbreviated form moderʹ. There are 17 occurrences of fæder glossing L. patris. In six instances the noun takes the ending ‑es:fadores. One example of a zero genitive, the etymological inflection for these nouns, is attested in Durham: fador (163.15) (Lindelöf 1923: 278). The remaining 10 tokens are abbreviated forms: fadorʹ(7x), fædorʹ(1x), fadoʹ(1x), fęderʹ(1x). Different variants alternate in the same contexts, as illustrated by the following: (18) on nome fador' ⁊svn' ⁊gast hal' in nomine patris et filii et spiritus sancti ‘in the name of the Father, and of the Son, and of the Holy Spirit’ DurRitGl 1 [Thomp-Lind] 116.6 (19) on nome fadores ⁊svnv' ⁊gastes halges in nomine patris et filii et spiritus sancti ‘in the name of the Father, and of the Son, and of the Holy Spirit’ DurRitGl 1 [Thomp-Lind] 117.18 In Lindisfarne four nouns belonging to this declension are attested in the genitive: fæder (36x), moder (5x), broðer (10x), sweoster (1x). Genitive singular forms in ‑es are dominant for all except for sweoster:fæder (30x), moder (5x), broðer (8x) (cf. Rodríguez Ledesma 2016: 217). Although with fewer occurrences and many more examples of abbreviated forms, there is also extension of the inflection ‑es to these nouns in Durham. With regard to the vowel of the unstressed syllable in the genitive singular forms of fæder, <o> is the norm in Durham (16x, as against 1 xof <e>), as is the case in Lindisfarne as well (23x, as against 4 x of <e>) (cf. Rodríguez Ledesma 2016: 217). This confirms Hogg and Fulk’s claim that in northern Northumbrian the usual form of the genitive singular of fæder is fadores,“with occasional syncopated forms and a few forms without inflexion”(2011: § 3.68, n. 3). Gen. sg. -es in the Durham Collectar Gloss 633 4 Weak Nouns or n-Stems The genitive singular inflection of weak nouns was ‑an, but in late Northumbrian the final consonant was lost, with the result that forms ending in a vowel are found throughout the singular and in the nominative and accusative plural of these nouns. There is great fluctuation with regard to the vowel, and all <a, e, u, o, æ> occur. Hogg and Fulk also note that in Northumbrian weak nouns may acquire endings of the a‑stems, such as ‑as in the nominative/accusative plural and ‑es in the genitive singular, the latter being “the commonest ending by far for the gen.sg. in both masc. and fem. nouns”(2011: 126). Table 5 gives the weak nouns which are attested in the genitive singular in Durham, together with the Latin form they gloss and the number of occurrences of the different inflections. The first three nouns are feminine and the other three masculine. Table 5: Genitive singular of weak nouns n‑sn Vowel (Etymological Inflection) n Abbreviation n Total cirica (f.) (L. ecclesia,‑ae) 2417 earðe (f.) (L. terra,‑ae) 45110 nedre (f.) (L. serpens,‑tis) 1––1 gidvola (m.) (L. error,‑is, L. ignorantia,‑ae) 3–1 44 4 erendvreca (m.) (L. apostolus,‑i) 1––1 45 gileafa (m.) (L. fides,‑ei) 2––2 44 The attested form is gidvoles' (DurRitGl 1 [Thomp-Lind] 34.13), with a “puzzling”abbreviation mark at the end which seems to be a mistake of the glossator; cf. remarks by Lindelöf (1927) and Squires (1973) above, in Section 1.3. 45 There are 18 tokens of the Latin genitive singular apostoli in Durham: 11 x are not glossed, 6 x are glossed with an abbreviated form of the borrowing (apostol' (2x), apl' (3x), apost' (1x)), and only one is glossed with the native word: erendvrecæs (DurRitGl 1 [Thomp-Lind] 79.8). 634 Nieves Rodríguez Ledesma n‑sn Vowel (Etymological Inflection) n Abbreviation n Total lichoma (m.) (L. corpus,‑oris, L. caro,carnis) 16 1 46 320 noma (m.) (L. nomen,‑inis) 16815 Total 30 (50 %) 16 (26.66 %) 14 (23.33 %) 60 As can be seen in the table, forms with ‑es for the genitive of these nouns are the most frequent in Durham (30x, 50 %), followed by forms with the etymological ending (16x, 26.66 %), and finally by abbreviated forms (14x, 23.33 %). As in previous classes, some abbreviated forms seem to stand for the ‑es inflection: gidvoles' (1x), earðe' (1x) and nome' (6x), whereas others could stand for either the etymological or the ‑es ending: nom' (2x), lichom' (1x). With regard to forms like nome', Squires (1973: 406–407) remarks that, although normally interpreted as nomes, genitive singular nome is also frequent in Durham, and therefore she considers the abbreviated mark as having the same effect as nome s , signalling alternation, nome uel nomes, and producing variant inflectional forms (407). Looking at the results more closely, not all nouns behave in the same way: those which are more frequent in the gloss favour the etymological inflection over ‑es (cirica,earðe and noma), the only exception being lichoma, with an overwhelming number of ‑es forms in the genitive (16x out of 20x). Another factor conditioning the choice of inflection may be the gender of the noun: the extension of ‑es may have taken place in masculine nouns earlier than in feminine ones. That could explain the behaviour of cirica and earðe (feminine), but not that of noma (masculine), with only one token of ‑es. 47 As in previous classes, in some instances the grammatical context seems to have conditioned the choice of a particular inflection: etymological vs. innovative Table 5: (continued) 46 The only token of the etymological form is found rendering a Latin noun modified by a possessive: mið vser lichome bisene, glossing L. cum nostrę carnis substantia (DurRitGl 1 [Thomp- Lind] 4.13–14). In the same context the inflection ‑es is also found: in vnstondennisse vser lichomes, glossing L. in substantia nostre carnis (DurRitGl 1 [Thomp-Lind] 2.9). 47 However, if the six tokens of nome' are interpreted as nomes, the instances of the etymological inflection and innovative ‑es would be roughly the same (6x vs. 7x). Gen. sg. -es in the Durham Collectar Gloss 635 Jones, Charles. 1970. “Some Features of Determiner Usage in the Old English Glosses to the Lindisfarne Gospels and the Durham Ritual”.Indogermanische Forschungen 75: 198–219. Jones, Charles. 1987. Grammatical Gender in English, 950–1250. London: Croon Helm. Kendrick, T.D. et al. (eds.). 1956–1960. Evangeliorum quattuor Codex Lindisfarnensis, Musei Britannici Codex Cottonianus Nero D. IV. 3 vols. in 2. Olten/Lausanne: Graf. Ker, N. R. 1943. “Aldred the Scribe”.Essays and Studies by Members of the English Association 28: 7–12. Lindelöf, Uno. 1901. Wörterbuch zur Interlinearglosse des Rituale Ecclesiae Dunelmensis. Bonn: Hanstein. 105–220. Lindelöf, Uno. 1923. “A New Collation of the Gloss of the Durham Ritual”.The Modern Language Review 18: 273–280. Lindelöf, Uno (ed.). 1927. Rituale Ecclesiae Dunelmensis: The Durham Collectar. London: Surtees Society. Roberts, Jane. 2016. “Aldred: Glossator and Book Historian”. In: Julia Fernández Cuesta and Sara M. Pons-Sanz (eds.). The Old English Gloss to the Lindisfarne Gospels: Language, Author and Context. Berlin: De Gruyter. 37–59. Rodríguez Ledesma, Mª Nieves. 2016. “Dauides sunu vs. filii david: The Genitive in the Gloss to the Lindisfarne Gospels”. In: Julia Fernández Cuesta and Sara M. Pons-Sanz (eds.). The Old English Gloss to the Lindisfarne Gospels: Language, Author and Context. Berlin: De Gruyter. 213–238. Ross, Alan S. C. 1936. “Sex and Gender in the Lindisfarne Gospels”.Journal of English and Germanic Philology 35: 321–330. Ross, Alan S. C. 1937. Studies in the Accidence of the Lindisfarne Gospels. Leeds: University of Leeds. Ross, Alan S. C. 1960. “Standard Paradigms”. In: Kendrick et al. (eds.). Evangeliorum quattuor Codex Lindisfarnensis, Musei Britannici Codex Cottonianus Nero D. IV. Book II. Olten/ Lausanne: Graf. 37–42. Ross, Alan S. C. 1970. “Conservatism in the Anglo-Saxon Gloss to the Durham Ritual”.Notes & Queries 215: 363–366. Ross, Alan S. C. 1971. “Aldrediana XXIII: Notes on the Accidence of the Durham Ritual”.Leeds Studies in English 5: 53–67. Ross, Alan S. C. 1978. “A Point of Comparison between Aldred’s Two Glosses”.Notes & Queries 223: 197–199. Ross, Alan S. C. 1982. “Rare Words in Old Northumbrian”.Notes & Queries 227: 196–198. Ross, Alan S. C. and Eric Stanley. 1960. “Index Verborum Glossematicus”. In: Kendrick et al. (eds.). Evangeliorum quattuor Codex Lindisfarnensis, Musei Britannici Codex Cottonianus Nero D. IV. Book II. Olten/Lausanne: Graf. 45–176. Skeat, Walter W. 1877–1879. “Collation of the Durham Ritual; with Notes”.Transactions of the Philological Society 1 (Issue 3. Appendix II): 49–72. Squires, Ann. 1971. “Collation of the Anglo-Saxon Gloss to the Durham Ritual”.Notes & Queries 216: 362–366. Squires, Ann. 1973. “Some Curious Abbreviations in the Durham Ritual”.Notes & Queries 218: 403–409. Stevenson, Joseph (ed.). 1840. Rituale Ecclesiae Dunelmensis. London: Surtees Society. Tamoto, Kanichi (ed.). 2013. The MacRegol Gospels or The Rushworth Gospels. Amsterdam/ Philadelphia, PA: Benjamins. 642 Nieves Rodríguez Ledesma