The Inflection of Latin Proper Names in the Old English Translation of Bede’s Historia Ecclesiastica

Abstract

This paper discusses the inflections of Latin proper names in the Old English translation of Bede’s Historia Ecclesiastica. Whereas most common Latin loans are integrated into the Old English system as far as inflections are concerned, proper names, like scientific loans, can retain Latin inflections in some contexts. The analysis of the more than 700 tokens in this text reveals that the prototypical paradigm of Latin proper names results from a mixture of Latin and Old English patterns. The choice of inflections seems to be chiefly conditioned by grammatical case. While the nominative and accusative are modeled after Latin with very few exceptions, the dative and genitive are influenced by Old English paradigms as well. Both Latin and Old English inflections are evenly distributed in the dative. However, marking on names seems to be secondary and determined primarily by additional morphosyntactic means such as determiners or prepositions. As for the genitive, the predominant inflection, thematic vowel plus -s, results from the fusion of the inflections in both languages grounded in phonetic or spelling similarities.

1. Introduction

In any language with a relatively rich morphological system, proper names tend to follow the same rules as common nouns with regard to inflections in that they are ascribed to one of the available nominal inflectional paradigms. Thus, names of Anglo-Saxon origin follow one of several inflectional paradigms in Old English. For example, Æðelwulf declines as the a-stem wulf ‘wolf’, while Beda follows the n-stem declension, i.e., guma ‘man’ (see

Table 1. Singular paradigms of the a-stem wulf ‘wolf’, ō-stem cwen ‘woman’, or n-stem guma ‘man’.

	a-stem	ō-stems	n-stem
Nominative	wulf	cwen	guma
Accusative	wulf	cwene	guman
Genitive	wulfes	cwene	guman
Dative	wulfe	cwene	guman

below). Colman (2014) and Okasha (2011) offer detailed information on this point. Latin is no exception to this general rule. Therefore, a name such as Marius shares its inflectional endings with common nouns belonging to the second declension, such as dominus ‘master’ or servus ‘slave’. When Latin names are used in Old English texts, however, some difficulties arise.

This paper studies proper names that entered Old English via Latin regardless of whether they are ultimately of Latin origin, such as Claudius or Marcus, or not, as is the case with Greek Theodorus or Irish Colmanus. Given their foreign origin, these names present certain peculiarities concerning their morphology. As Campbell (1959, p. 219) points out, these proper names do not seem to fit in the Old English system as they can retain Latin inflections in some contexts and in some grammatical cases. Consequently, the same Latin name can be attested with a Latin inflection in a given text, and with an Old English one elsewhere. Examples (1) and (2) below illustrate this behavior. In (1) the name Petrus has its corresponding Latin dative inflection -o, whereas in (2), it shows the Old English dative ending -e.

(1)	Ðætte Agustinus þæt mynster þara apostola Petrus & Paulus getimbrade; & be his þam ærestan abbude Petro (dative). ‘That Augustine built the monastery of the apostles Peter and Paul, and about its first abbot Peter’ BedeHead [0032 (1.10.20)]1.

(2)	Ðætte Laurentius se arcebiscop wæs geþread fram þam apostole Sancte Petre (dative). ‘That the archbishop Laurentius was chastised by the apostle St. Peter’ BedeHead [0041 (2.12.6)].

Latin influence is one of the major topics in most works on Old English vocabulary such as Kastovsky (1992, 2006) or Gneuss (1993). However, proper names tend to be overlooked, most likely due to their peculiarities compared to common nouns as Durkin (2014, p. 11 and fn7) mentions. Other works, including Baker (1998) and Rodríguez Ledesma (2016), focused more on the interaction of the Latin and Old English morphological systems, do take them into account, although they do not discuss their idiosyncrasy in detail. Others, for example, Bately (1980, pp. cviii–cix), offer a succinct overview of their features as part of the wider introduction to a specific edited text.

Following the lines established by Ruiz Narbona (2023a, 2023b), this work sets out to analyze the inflectional system of Latin proper names in the Old English Translation of Bede’s Historia Ecclesiastica (OEHE henceforth). This text was chosen because of the large number of prominent characters it features. Thus, it stands out as a major source of Latin proper names in Old English. This paper explores how these names are adapted to the Old English system and tries to shed light on the rules underlying the choice of inflections, the factors that may contribute to it, and the type of possible interactions between Latin and Old English. It ventures as well to offer insights into the dynamics of noun morphology within Old English itself with a special focus on the decline in productive inflections in the language from an early period (Hogg 1992, pp. 124, 137–38).

This paper is further divided into four sections. Section 2 discusses previous works on Latin loanwords and names with a special focus on their morphology. Section 3 concerns the methodology. It provides a brief introduction to the OEHE and explains the corpus design. This is followed by Section 4, where the main results of the present investigation are discussed. Finally, Section 5 provides a summary of the main conclusions and discusses further research.

2. Latin Loanwords in Old English and Their Morphology

The type of loan and the period when they entered Old English are the major concerns of the literature with respect to the discussion of Latin loanwords. Usually, three periods have been distinguished, namely the Continental period, the Anglo-Saxon settlement period, and the Christianization period (see Cole and Laker (2022, pp. 6–7) with references). Other authors such as Campbell (1959, p. 200), Wollmann (1990), or Durkin (2014, p. 105), however, advocate for a twofold division marked by Christianization.2

Regardless of whether a threefold periodization is favored, the introduction of Christianity represents the clearest boundary line as far as the nature of loanwords is concerned (Kastovsky 1992, p. 305). From the 7th century onwards, the church, with its organization and rituals, becomes the richest source of Latin loanwords into Old English. The characteristics of loanwords also changed as, with time, an increasing number of words entered Old English through the written medium, and therefore, from Classical Latin, rather than from Vulgar Latin as in previous periods. This tendency grows stronger as the Christianization period continues, especially after the Benedictine Reform introduced in the late 9th and early 10th centuries, which Kastovsky considers a sub-grouping within the Christianization period (Kastovsky 1992, p. 307; 2006, pp. 221–22). This period is characterized by an influx of scientific and scholarly loans—around 150 according to Strang (1970, p. 314)—that are not adapted into the Old English inflectional system as regularly as previous loans (Funke 1914; Durkin 2014, p. 103).

When loans enter a language with a rich inflectional system such as Old English, they are normally ascribed to one of the available paradigms and take native inflectional endings (Gneuss 1996, pp. 3–4). Such was the case of most Latin loans in Old English (Durkin 2014, pp. 123–24). Although Old English had a wide array of nominal paradigms (see Campbell (1959, pp. 222–60)), most Old English nouns were ascribed to one of three major declensions, namely a-stems, ō-stems, or n-stems, illustrated in Table 1 (Hogg 1992, pp. 137–38). Not only were these classes dominant in terms of the number of nouns ascribed to them, but they also attracted nouns from other minor declensions, while the opposite shift was rare. Additionally, with very few exceptions, such as the names of certain nations (see Campbell (1959, p. 208) and Brunner (1965, p. 215)), loanwords that entered Old English were integrated into one of these three primary declensions, which as Gneuss (1996, p. 4) highlights, is a good indicator in assessing the predominance of certain declensional types in a given language.

A succinct account of how Latin loanwords were integrated into Old English is offered below following the detailed overview by Campbell (1959, pp. 208–9 and 218–19) and Gneuss (1996). Latin first declension nouns, mostly feminine, are adapted following either the ō-stem model, e.g., cylen ‘kiln’ from Latin culīna ‘kitchen’, or n-stems as is the case with mæsse from Latin missa ‘mass’. Adaptations to n-stems are preferred as the Old English period progresses. Second declension nouns tend to follow the a-stem declension, such as disc from Latin discus ‘dish’. So are third declension nouns such as font, Latin fons ‘fountain’, regardless of whether they are consonant stems or i-stems. However, third declension n-stems, e.g., Latin draco ‘dragon’, Old English draca, appear as n-stems.

Although the rules described above apply generally, several authors have emphasized the ability of a-stems to attract all types of Latin loans irrespective of their gender or original declension (see, for instance, Pogatscher (1888, pp. 154–65), Funke (1914, pp. 131–34), and Campbell (1959, pp. 207–10)). As discussed in Welna (1978, 1980), this is not exclusive to Old English, though. Actually, the predominance of a-stems eventually shaped the English nominal system as it became practically the sole available paradigm in later periods as pointed out by Hogg (1992, pp. 137–38).

There is still a group of Latin loanwords that deserve further discussion given their morphological peculiarities. I refer to the scholarly and scientific loans incorporated mostly during and after the Benedictine Reform. Contrary to most Latin loans in Old English, these words are not fully incorporated into the Old English system and tend to retain Latin inflections, at least partially. Funke (1914) analyses them in depth and concludes that as far as their morphology is concerned, they could be divided into two groups: on the one hand, Lehnwörter, that is loanwords, whose inflectional endings are modeled after Old English patterns, and Fremdwörter ‘foreignisms’, which tend to retain Latin inflections. Durkin (2014, pp. 124–27), however, disagrees with this classification and even with the fact that they should be considered Old English words. He suggests they could be better described as examples of code-switching3 in that they seem to behave like Latin words embedded in Old English texts, as is usually the case with Greek words in Latin scientific texts. A strong argument in favor of the code-switching hypothesis is the fact that these foreignisms are frequently accompanied by meta-textual explanations that serve to clarify the meaning of the word, as illustrated by exorcista in (3). It is also interesting to point out that in example (3) exorcista does not have a dative ending as would be expected from context, but a nominative one. This seems to reinforce the idea that the word is being cited or used in its ‘dictionary form’ and is not integrated into the Old English sentence.

(3)	Þa underget se bisceop þæt he mihte hine gibigan gif he him bude læssan had, and bead him þæt he wære gehadod to exorcista, þæt we hatað halsigend þe ðe bebyt deoflum, þæt hi of gedrehtum mannum faran.
	‘Then the bishop perceived that he might persuade him if he offered him a lesser office, and offered him that he should be ordained exorcist, as we call one who adjureth, or one who commandeth devils that they should go out of possessed men’ ÆLS (Martin) [0037 (139)]4.

In addition to this, the classification based on morphological terms may also be problematic as acknowledged by Funke (1914, p. 144), as he considers a third group of so-called mixed loanwords, i.e., loans with Latin nominatives and Old English inflections in oblique cases. Nevertheless, as Durkin (2014, pp. 127–28) shows, the paradigm of mixed loanwords is not as consistent as Funke seems to imply. As a way of illustration, the word cometa ‘comet’ is always attested with a Latin nominative and Old English -an inflections in the rest of the grammatical cases, but only in the singular. The nominative plural inflection, however, can either be Latin -e or Old English -an as exemplified in (4) and (5).

(4)	Comete (nominative plural) sind gehatene þa steorran ðe færlice & ungewunelice æteowiað. ‘The stars that appear suddenly and unusually are called comets’ ÆTemp [0212 (9.13)].

(5)	Her æteowdan twegen cometan (nominative plural). ‘Here appeared two comets’ ChronD (Cubbin) [0089 (729.1)].

These scientific loanwords are relevant for the discussion in this paper because proper names can also retain their Latin inflections in some cases (Campbell 1959, p. 219). Kirkman (1930, pp. 19–21) and Bately (1980, pp. cviii–cix) discuss Latin proper names in Orosius’ Historia Adversus Paganos and conclude that inflectional endings from both languages are regularly used. They also mention the common attestation of endings that do not match their corresponding syntactic context as well as the frequent addition of Old English inflectional endings to Latin nominatives, for example, to Antiouchuse ‘to Antioch’.

Ruiz Narbona (2023a, 2023b) analyzes the inflection of Latin proper names in the Old English Martyrology. Conclusions show that the inflectional system of names is the result of the interaction of the Latin and Old English systems. As far as masculine names are concerned, nominative and dative inflections are dominated by Latin and Old English, respectively. Accusative marking is based on syntactic means, as accusative names are distinguished from their nominative counterparts mainly by means of the determiner þone. The genitive inflection consisting of a thematic vowel plus a sibilant arises from the fusion of Old English genitive -es and other Latin inflectional endings in -s, such as the third declension genitive -is or the second declension nominative -us. Additionally, feminine names are better integrated into the Old English system. Even if they are consistently Latin-like in the nominative, they display weak -an endings in oblique cases in most attestations. The only exception is the contextually and syntactically restricted genitive inflection -e. As will be shown below, the inflection of names in the OEHE shares some similarities with the aforementioned works. Whereas Latin inflections dominate the nominative and accusative cases, the genitive follows the model discussed in the Martyrology. Finally, dative inflections seem to alternate freely between Latin and Old English.

3. Data and Methodology

3.1. The Old English Translation of Bede’s Historia Ecclesiastica

The source of the proper names under study in this paper is the Old English Translation of Bede’s Historia Ecclesiastica. The translation of this major historical work by one of the most prominent scholars of his age has often been associated with the learning program devised by King Alfred the Great (d. 899) (Fulk and Cain 2013, p. 99) or an earlier school of translation in Mercia (Rowley 2011, p. 2)5. It seems to have served the purpose of celebrating England’s glory in terms of its past achievements while at the same time exemplifying it through the work of one of its greatest scholars. Far from being a slavish translation of the original, this version is more focused on Anglo-Saxon matters, while at the same time, it discards material of little interest to a ninth-century audience (Scragg 1997, p. 47). Given the large number of prominent characters featured in this work, it stands out as a major source of Latin proper names. Furthermore, they are attested in a wide variety of syntactic contexts, contrary to what happens in other surviving sources, such as lists of kings, which makes this text a valuable source of information for researchers working on proper names.

Paleographical evidence suggests that the earliest window for the translation must have been between the late ninth century and the early tenth century. However, the five more or less complete surviving manuscripts known as T, C, O, B, and Ca, plus some excerpts edited by Zupitza (1878) (Zu), date from the tenth and eleventh centuries. For an overview of the different manuscripts and editions of the text (see Rowley (2011, pp. 15–28)). MS T is considered the best manuscript, and it conforms the backbone of the most accurate edition up to the present day, Miller’s (1890–1891) The Old English Version of Bede’s Ecclesiastical History of the English People. Even if all surviving manuscripts are West-Saxon, Miller (1890–1891, p. v) determines that the original translation, now lost, is an Anglian text, and therefore, with his edition, he offers a text that represents the Anglian text as far as possible. In order to do so, he discards MS B and follows T, C, O, and Ca in order of preference.

3.2. Corpus Design and Analysis of the Data

The data analyzed in this study were retrieved using the Dictionary of Old English Web Corpus (DOEC henceforth) (Healey et al. 2015). This major Old English corpus includes Miller’s (1890–1891) edition of the text. Therefore, both the Old English text and its present-day English translation referred to in this paper come ultimately from this edition. References to specific fragments, however, are made following the DOEC.

The complete corpus of examples consists of 753 tokens and 120 types that include proper names that have entered Old English via Latin regardless of their ultimate origin, be it Greek or other as mentioned in the introduction. With regard to the analysis of each attestation, seven categories have been considered. Three are referential in nature and include the fragment where the name appears with its reference, taken from the DOEC, as well as the translation of the fragment to present-day English from Miller (1890–1891). Once the attestations were collected, each fragment was analyzed with a focus on four additional categories. The process is explained using (6) below as an example. The first step was to identify the grammatical case of the word form under study, i.e., Daniele. Given its context, Daniele is clearly in the dative case. Next, the inflection of the name, i.e., -e, was analyzed and it was determined whether or not the inflectional ending was compatible with the Latin or Old English systems. In this case, -e is the typical Old English dative strong ending. Following this, information regarding the lemma and the original Latin declension of the name was included. In some less clear cases, it proved more useful to analyze this information before determining whether the inflectional ending in a given occurrence was Latin or Old English. Thus, the four additional non-referential categories taken into account in this study are the following: (1) the lemma of the name; (2) the form of the name as it appears in the attestation in question; (3) the grammatical case of the form under analysis and information regarding the inflectional ending of the token; and (4) the original Latin declension the name belongs to. An example of the complete coding of the data is illustrated in (6).

(6)	Lemma: Daniel
	Word form: Daniele
	Inflectional ending: -e. Old English strong declension
	Latin declension: Third
	Fragment: Sio biscopscir Wihte þæs ealondes belimpeþ to Daniele Wintanceastre biscope
	Fragment translation: The episcopal charge of the Isle of Wight falls to Daniel bishop of Winchester
	References: Bede 5 [0571 (22.478.17)]

4. Results of the Analysis

The focus will be laid firstly on the scarce number of Latin feminine proper names in the OEHE. Their morphological idiosyncrasy differs from that of the majority of masculine names discussed below. Because of this and their low numbers and limited distribution, I will deal with them separately.

4.1. Feminine Names

Latin feminine names in the text under study are limited to four types and nine tokens. Two of these items appear in the nominative case. These are Fara (Bede 3 [0133 (6.172.11)]) and Euae (Bede 1 [0400 (16.86.28)]). The former has the typical Latin nominative first declension inflection -a, while the latter shows the first declension genitive ending -ae. As argued by Baker (1998, p. 195), the presence of genitive endings instead of nominatives is not particularly uncommon, especially in martyrologies or calendric material (see Ruiz Narbona (2023b, p. 45) for more examples).

The rest of the feminine attestations, two types and seven tokens, are in the genitive case. They are Marian (x5), Mariam (7), and Cecilian (Bede 5 [0231 (12.422.1)]). The inflectional ending -an is typical of the Old English weak declension. As for Mariam, -am might be taken as a Latin first declension accusative. However, the context shows that this is clearly a genitive, as it is in the original Latin text: et per intercessionem beatae suae genetrices semperque uirginis Mariae (Colgrave and Mynors 1969, pp. 526, 528)6.

(7)	Forgeaf ðe Drihten lif þurh gebedo […] & þurh þingunge his þære eadigan meder ðære unmælan fæmnan Sancta Mariam (genitive).
	‘The Lord has given you life through the prayers […] and the intercession of his blessed mother the immaculate virgin St. Mary’ Bede 5 [0482 (17.462.26)].

In conclusion, the few Latin feminine names in my corpus seem to conform to what was described by Ruiz Narbona (2023b), namely that these names tend to have a Latin inflection in the nominative and the ending -an in oblique cases, thus conforming rather closely to the Old English weak declension. However, the scarcity of the data in the OEHE makes this conclusion only tentative at best.

4.2. Nominative

The nominative case is the one with the highest number of attestations in type, 75, and second to the genitive in number of tokens with 264 examples. It is the most uniform case as far as its inflectional endings are concerned since most names follow Latin models.

First declension names constitute the smallest sub-group with only three types, namely Columba (x2), Esaias, and Tobias. Both Esaias and Tobias have the -as ending typical of first declension masculine names that have entered Latin via Greek. As for the -a ending in Columba, it could either be accounted for as the Latin first declension nominative ending, for example, poeta, ‘poet’, or as the nominative inflection of some Old English weak names such as Beda. The fact that its two genitive attestations in the OEHE have the Old English weak inflection -an, Columban (Bede 5 B9.6.7 [0170 (9.410.17)] and [0177 (9.412.1)]) seems to point to the latter alternative as a more suitable explanation.

Most attested names in the nominative are original Latin second declension names. This class is made up of 58 types and 234 tokens including names such as Theodor (x19), Paulinus (x18), Iustus (x9), Laurentius (x7), Aidan (x6), Adrianus (x5), or Constantinus (x2) to name but a few. The names in this group present two alternative inflections: either Latin -us as Adrianus, or zero inflection, e.g., Aidan. The great majority of names, 52 types and 198 tokens, follow the former pattern. They account for 85.7% of second declension names. Alternatively, the number of names with zero inflection is rather low in terms of types, only six, but relatively high in terms of tokens, 33, especially given the high attestation of the name Theodor (x19) (see (8)). The occurrences of these names are distributed thus: Theodor (x19), Aidan (x6), P(i)p(p)in (x4), Crist (x2), Coloman, and Eleuther. Interestingly, none of those names are original Latin names. On the one hand, Crist is fully assimilated into the Old English system and always shows Old English inflections (see also Ruiz Narbona 2023a, p. 8). The same seems to be the case with the widely attested Greek name Theodor as its zero-inflected accusative forms show, even if this name is not as systematically assimilated as Crist (see, for example, Theodorus in (Bede 4 B9.6.6 [0017 (1.254.15)] and [0027 (1.256.11)])). As for Eleuther, based on evidence from the Old English Martyrology (Ruiz Narbona 2023a, p. 9), where forms such as Peter or Thelesfor are attested alongside Petrus and Thelesforus, there seems to be a tendency for names whose root ends in -r to have no ending in the nominative. Notice that both Theodorus and Eleutherius always have their corresponding -us ending in the Latin text. On the other hand, Aidan and Coloman are original Irish names, whereas Pippin is Frankish. Their zero-inflected nominative form is attested in Bede’s Latin version7 (see, for example, (9a) and its Old English translation (9b)). It seems to be the case that these foreign names were given a special treatment in Medieval Latin based on their original forms at least as far as the nominative inflection is concerned. This is followed in the Old English version. It is important to notice, however, that these names retain second declension Latin inflections in the rest of the grammatical cases in the original Latin text.

(8)	Þa cwom he Theodor (nominative) biscop to his cirican in Contwara burg.
	‘Then bishop Theodore arrived in his church in Canterbury’
	Bede 4 B9.6.6 [0037 (1.258.1)].

(9)	(a) Sed et ipse antistes Aidan (nominative) non plus quam XII post occisionem regis, quem amabat, …
	(Colgrave and Mynors 1969, p. 260).
	(b) Ge eac se seolfa biscop Aidan (nominative) nales ma þonne þy twelftan dæge æfter þæs cyninges slege, þone he lufode, …
	‘And bishop Aidan himself, not more than twelve days after the death of the king whom he loved…’
	Bede 3 B9.6.5 [031300 (12.198.14)].

The category of third declension names is represented by eight types and 19 tokens. Names in this group have the form of their Latin counterparts, e.g., Iohannes (x9), Clemens (x2), Felix (x2), Daniel, David, and Michael. The only exception to the rule is exemplified by two Latin n-stems, Agatthon (Bede 5 B9.6.7 [0465 (17.460.1)]) and Tilmon (Bede 5 B9.6.7 [0211 (11.418.6)]). These names have an -n in the nominative, whereas Classical Latin would have just -o, Agatho and Tilmo. This phenomenon is also attested in the Old English Martyrology (Ruiz Narbona 2023a, p. 10).

In addition to the aforementioned groups, there are also a number of names in the OEHE that are indeclinable in Latin, namely Deusdedit (x3), Aaron, Adam, Beniamin, and Iacob. None of them have any nominative inflection in the Old English text.

All in all, the results concerning the nominative case show that the scribes in the OEHE followed Latin models rather closely. The only major exception to this trend is the group of second declension names that have no inflectional ending, like Old English strong nouns. The fact that some of these names are not either Latin or Greek may have contributed to their assimilation into the Old English system. Additionally, other names that are well known for the scribes and their readership such as Crist and Theodor seem to have been adapted to Old English in most cases.

4.3. Accusative

Latin names in the accusative are not as widely attested as their nominative counterparts. However, they appear in enough numbers to allow for a description of some major tendencies regarding their inflections. The total number of attestations amounts to 35 types and 56 tokens.

First declension names in the accusative do not abound in the OEHE. There are only eight types, each of them attested only once, with the exception of the two forms of Tobias, i.e., Tobiam (Bede 5 [0160 (8.408.10)]) and Tobium (Bede Head [0122 (5.22.21)]); the expected regular first declension form of Tobias is the former. In the latter, the name seems to have undergone a change in declension given that most masculine names belong to the second declension (see, for example, accusative Thomum in the Old English Martyrology (Ruiz Narbona 2023a, p. 11)). The rest of the first declension names attested in the accusative are the following: Attilan (Bede 1 [0154 (10.48.9)], Blædlan (same passage as Attilan), Caiphan (Bede 5 [0365 (15.442.29)]), Esdram (Bede 5 [0598 (22.482.21)]), Neemiam (same passage as Esdram), Thomam (Bede 3 [0457 (14.220.1)]), and Satanan (Bede 5 B9.6.7 [0364 (15.442.26)]). As for their inflectional endings, two alternatives are recorded. Four of the names have the ending -am, corresponding to the first declension Latin accusative. The other four have -an. The explanation behind this inflection is not as straightforward as in the case of -am. The -an inflections exemplified above may have two different origins. On the one hand, Caiphan and Satanan seem to be examples of the first declension accusative marker displayed by Greek names in Latin (see, for example, Satanan in (10)). On the other hand, neither Atilla nor Blædla are original Greek names. Therefore, a plausible explanation for the -an ending could be an adaptation to the Old English weak declension. However, since these names are not attested elsewhere in the OEHE, this hypothesis must remain tentative.

(10)	Et ait illis videbam Satanan (accusative) sicut fulgur de caelo cadentem.
	‘And he said to them: I saw Satan like lightning falling from heaven’
	Luke 10.18 Vulgate8.

As for second declension names, they represent the largest number of examples, as expected, with a total number of 21 types and 41 tokens including Crist (x4), Augustinum (x4), Albanum (x2), or Paulinus to name but a few.

With regard to inflectional endings, the second declension Latin inflectional ending -um is clearly predominant with 17 types and 31 tokens, representing 75.6% of all second declension names attested in the accusative in the OEHE. These include Mellitum (x6), Iustum (x4), or Adrianum. Notice that, as was the case with Tobium, the only accusative form of nominative Eutychius, Eotycem (Bede 4 [0370 (19.312.8)]) has undergone a change in declension and displays a typical third declension ending.

In spite of the clear predominance of Latin -um, alternative endings are also attested, although their relevance is minor. Nine tokens, namely Crist (x4), Theodor (x3), Albinus (11), and Paulinus (Bede 2 [0275 (10.136.10)]) present ‘unmarked’ nominative-like endings. This should not be surprising as pointed out by Baker (1998, p. 198) since Latinate nominative forms seem to function as a default ending and may appear in syntactic contexts where they are not expected. As for the endings themselves, Albinus and Paulinus show the typical second declension nominative ending -us discussed above, whereas Crist and Theodor (see (12) for an example in context) have a zero ending. The fact that the latter names have a zero ending in the nominative and also in the accusative seems to reinforce the idea that they have been assimilated into the Old English inflectional system since strong names in this language tend to show no overt marking in these grammatical cases. However, it is important to point out that the form Theodorum also occurs once in the OEHE (Bede 3 [0128 (5.170.31)]. Finally, one of the examples of names in the accusative, Bonefatio (13), has an unusual -o ending corresponding to the Latin second declension dative. This phenomenon, although rare, is not entirely uncommon (see, for instance, Bately (1980, pp. cviii–cix), Baker (1998, p. 198), or Ruiz Narbona (2023a, p. 13)). Additionally, it may have been influenced by the dative form in the original Latin passage: de necessariis ecclesiae Anglorum cum apostolico papa Bonifatio tractaturus (Colgrave and Mynors 1969, p. 146).

(11)	Ðurh Albinus (accusative) swiðost ic geðristlæhte þæt ic dorste þis weorc ongynnan.
	‘And it was chiefly through Albinus that I was encouraged to commence this work’ BedePref [0014 (4.10)].

(12)	Ða forðferdum Gearomonne biscope bæd he Theodor (accusative) biscop, þæt he him & his leodum biscop funde & sealde.
	‘When, on the death of bishop Gearuman, he begged bishop Theodore to look out and assign him a bishop for his people’ Bede 4 [0063 (3.260.22)].

(13)	Ond he ða wæs smeagende mid þone apostolican papan Bonefatio (accusative).
	‘And he took counsel with the apostolic pope Boniface’ Bede 2 [0101 (4.108.8)].

The description of third declension names in the accusative is rather straightforward. The four types and five tokens belonging to this group have the expected Latin ending -em, e.g., Clementem (Bede 2 [0094 (4.106.20)], Iohannem (Bede 4 [0382 (20.314.17)] and Bede 5 [0471 (17.460.26)]), and Pharaonem9 (Bede 4 [0028 (1.256.14)]). The only occurrence worth commenting on is Danielum (14), a third declension name that, just as some other previously discussed names, has undergone a change in declension and displays a second declension inflectional ending.

(14)	& Westseaxan onfengon Danielum (accusative) & Aldelmum.
	‘And the West Saxons received Daniel and Ealdhelm’ Bede Head [0130 (5.24.7)].

In addition to the names above, the indeclinable name Deusdedit is also attested in the accusative case in the invariable form Deosdedit (Bede 5 [0619 (20.246.13)]).

The results regarding the accusative case are characterized by their homogeneity since, with very few exceptions, most names have their corresponding Latin accusative ending. First declension names have -am, or -an in original Greek names, second declension ones have -um, and third declension ones show -em. This choice of ending seems sensible as it allows scribes to mark direct objects overtly without using any additional syntactic means such as determiners. There are some minor exceptions to this pattern though. One involves an apparent adaptation to the Old English inflectional system. Thus, names such as Attila have a weak accusative ending while Crist or Theodor remain unmarked. Finally, other attestations show an ending that does not correspond to the syntactic context where they are found, as is the case with nominative-like Albinus and dative-like Bonefatio. Given their very rare occurrence, they cannot be said to conform to any type of pattern.

4.4. Dative

Names in the dative case are more widely attested than in the accusative, although their numbers fall behind the nominative and genitive cases with a total of 51 types and 127 tokens. As far as inflectional endings are concerned, there is a clear difference between the first and third declension groups, on the one hand, and second declension names on the other. While the former have exclusively either Latin endings, in the case of first declension names, or Old English ones, in the third declension names, variation is the norm in the original second declension names.

Only two first declension names are recorded in the dative case, e.g., Andreæ and Iiba. Both have a Latin ending, as mentioned above, although only -æ in Andreæ (Bede 2 [0082 (3.104.25)]) corresponds to an original Latin first declension ending. The -a ending in Iiba, however, is typical of the nominative. As was already pointed out above, the appearance of names with nominative-like inflections in non-nominative contexts is not rare. Additionally, Baker (1998, p. 198) offers a phonological explanation that may shed light on the use of the ending -a instead of -ae. He argues that scribes may have felt that the spelling -e, the typically post-classical first declension Latin ending, had qualities closer to that of the vowel schwa, and therefore, the spelling -a, as in Iiba, may have been a better choice for the Latin sound represented by -æ. Finally, it must be mentioned that case marking in the passage where Iiba appears is especially inaccurate (15). While it is clear that all proper names in the passage are in a dative context as shown by the phrase þæm ereticum ‘the heretics’, none of them is appropriately marked as such. Þeodorum has an accusative ending, Theodoreti a genitive one, and Iiba a nominative -a.

(15)	ond eft in Constantinopoli se fifta seonoð wæs gesomnad, in Iustinianes tide þæs geongran caseres, wið þeodorum & Theodoreti & Iiba (dative) þæm ereticum & heora larum.
	‘And again at Constantinople the fifth synod was assembled, in the time of the emperor Justinian the younger, to oppose Theodorus and Theodoret and Iba, the heretics and their doctrines’
	Bede 4 [0370 (19.312.8)].

In contrast with first declension names, third declension names in the OEHE consistently have an Old English dative ending, namely -e, typical of the strong declension. This group consists of 11 tokens and six types. They are the following: Daniele (x4), Iohanne (x2), Nerone (x2), Agathone, Davide, and Felice. To conclude with names that do not belong to the Latin second declension, indeclinable Deusdedit (BedeHead [0081 (4.16.29)]) has no inflectional ending in the dative.

Variation is the norm within dative second declension names in the text under study since inflectional endings do not conform exclusively to either Old English or Latin patterns. Interestingly, attestations are also quite evenly distributed. Out of the total 42 types and 113 tokens of second declension names, 18 types and 53 tokens, that is roughly 46.9%, have the Old English dative strong ending -e. They include Criste (x19), Theodore (x7), Aidane (x3), Benedicte (x2), or Claudie (x2). In contrast, 45.1% of names, i.e., 29 types and 60 tokens, have a Latin ending. There is variation within this group too. While the majority of names, 45 out of the above-mentioned 60 tokens, show the expected second declension dative -o, e.g., Augusto (x8), Iusto (x3), or Laurentio (x3), eight tokens have -i. The latter corresponds to the third declension i-theme dative ending. However, this deviation from the expected norm is not unprecedented. As was seen in the analysis of names in the Martyrology (Ruiz Narbona 2023a, p. 14), names that have an i in the last syllable of the root tend to have a dative ending -i, thus Augustini (x7), Gregorii (x2), or Paulini (x3). The only exceptions to this rule in the OEHE are Pauli (Bede 2 [0079 (3.104.21)]) and Theodoreti (see (15) above).

One of the objectives of this paper is to shed light on the factors that may contribute to variation in the choice of inflectional endings. There are several factors that may be involved in this variation. A working hypothesis of this study is that inflection might be name-dependent. The data show that this is clearly not the case as several of the names attested in the dative are inflected following the Latin model in some attestations and the Old English one in others. For an example, refer to Adriano/Adriane, Petro/Petre, or Laurentio and Laurentie in (16) and (17).

(16)	Þæt Mellitus æfter Laurentio (dative) arcebiscop geworden wæs.
	‘That Mellitus became archbishop after Laurentius’ BedeHead [0042 (2.12.9)].

(17)	þe se ilca papa Gode þam leofan ærcebiscope Laurentie (dative) […] sende:
	‘which the pope sent to archbishop Laurentius, well-beloved of God’
	Bede 2 [0105 (4.108.15)].

Given the fact that Miller’s edition of the OEHE is based on several of the surviving manuscripts, variations in inflectional endings in each of those manuscripts might be behind the different dative endings in the edited text. The analysis of the data, however, proves that this is not so, as names with Latin and Old English dative markers appear in extracts from different manuscripts. A paradigmatic case is that of Adriano and Adriane, which appear in both the preface (examples (18) and (20)) and in Book IV (examples (19) and (21)) based on MS Ca and MS T, respectively. The same can be said of Honorie (Bede 1 [0148 (10.46.26)]) and (Bede 2 [0317 (12.142.21)]) or Laurentio (BedeHead [0042 (2.12.9)]) and (Bede 2 [0137 (5.112.25)]), where the former attestation is from MS Ca and the latter from MS T.

(18)	Ðæt þam eaufæstan abbude Adriano (dative) Albinus æfterfyligde.
	‘That Albin succeeded the pious abbot Adrian’ BedeHead [0132 (5.24.15)].

(19)	& swa ætgædre mid Adriano (dative) þam abbude in Sexta Kalendas Junias wæs to Breotone onsended.
	‘And so he was dispatched along with abbot Adrian to Britain on the 27th of May’
	Bede 4 [0024 (1.256.3)].

(20)	ðeodorus wæs to arcebiscope gehalgod & mid Adriane (dative) þam abbude wæs to Breotone onsended.
	‘Theodore was consecrated archbishop and sent with abbot Adrian to Britain’
	BedeHead [0081 (4.16.29)].

(21)	Wæs in þa tid sum munuc in Rome, se wæs cuð <Adriane (dative)> þæm abbude, þæs noma wæs Theodorus.
	‘There was at that time a monk at Rome named Theodore, well known to the abbot Adrianus’ Bede4 [0017 (1.254.15)].

Another relevant factor that may influence the choice of inflection is context, both textual and syntactic, as demonstrated in the Martyrology in relation to the genitive case (see Ruiz Narbona 2023a, pp. 15–16; 2023b, p. 51). The analysis of the textual context in the OEHE does not shed light on the issue of variation since none of the endings seem to be restricted to a particular context, such as headings, chapter titles, or the like. As for the syntactic context where dative names occur, an in-depth analysis of the results shows that Latin and Old English forms can appear in exactly the same contexts. Examples (19) and (20) above illustrate this perfectly since Adriano and Adriane are both preceded by a preposition that governs or may govern the dative, mid ‘with’, and followed by the noun phrase þam abbude in which both the determiner and the noun are clearly marked as dative. In fact, all dative names in my corpus of examples appear in very similar contexts. They are either preceded by a preposition that governs dative (16), accompanied by a noun phrase including a determiner in the dative (17), or both, as in the examples mentioned above. There are examples of the dative absolute construction too (22). The fact that names in the dative case in the OEHE always appear in contexts in which the dative marking in the name is reinforced by additional morphosyntactic elements that clearly mark the function of these names may have contributed to variation as the dative inflection may have been regarded as secondary with respect to the other morphosyntactic elements. As is the case with most of the phenomena discussed in this paper, this is not alien to other texts, as a similar situation can be described in the Martyrology, where inflection in the accusative names varied but was always reinforced by the presence of the determiner þone, which seems to have acted as the primary means of accusative marking (Ruiz Narbona 2023a, p. 13).

(22)	Ðæt ricsiendum Honorio (dative) Gratianus & Constantius wæron on Breotene acende.
	‘That in the reign of Honorius, Gratianus and Constantinus were born in Britain’
	BedeHead [0011 (1.6.25)].

Finally, it must be mentioned that there is a minor group of second declension names that do not take the inflectional endings -o or -e. Some of them undergo a change in declension and have either -a, Aurelia (Bede 1 [0045 (4.32.1)]) or -æ, Fursæ (Bede 3 [0424 (14.214.24)]) and Gregoriae (Bede 2 [0091 (3.106.9)] endings typical of the first declension. This unexpected change from a masculine to a feminine declension is also detected in Orosius as pointed out by Bately (1980, p. cviii). Other names show Latin endings that do not correspond to the dative. For example, Constantinus (Bede 1 [0115 (8.42.13)]) has a nominative ending, whereas Arerum (Bede 4 [0370 (19.312.8)] and Theodorum (15) an accusative one. As already mentioned above, examples like these are not rare, although they are not numerous and do not constitute any type of regular pattern.

All in all, the analysis of the dative names in the OEHE has determined that variation is pervasive. While the first and third declension inflections are modeled after Latin and Old English, respectively, inflections from both languages are equally predominant in second declension names. The study of the context where these names appear points to the fact that dative marking on names may have been regarded as secondary since case marking is reinforced by other morphosyntactic means in all examples.

4.5. Genitive

The grammatical case that presents the largest set of data is the genitive. This is, in part, due to the high number of attestations of Cristes (x160), which represent more than half of the total number of 55 types and 297 tokens. The inflectional endings in this group stand out for the coexistence of both Latin and Old English for which an explanation is offered at the end of this section.

The first declension data set consists of five types and nine tokens. Six of these tokens have a Latin ending. On the one hand, Andree (Bede 5 [0499 (18.466.6)]) and Luce (Bede 5 [0602 (22.482.26)]) show the expected post-classical first declension -e. On the other hand, four tokens display a typical nominative-like inflection. Three of them, Andreas (x2) and Tobias (BedeHead [0135 (5.24.24)]), have -as, the first declension nominative ending of names that entered Latin through Greek, while Tobius (Bede 5 [0600 (22.482.23)]) seems to have undergone a declension change as in the accusative (see Section 4.3 above). The three remaining first declension tokens have an Old English inflection. Columban (x2) is adapted to the Old English weak declension as most feminine names and also Old English names such as Beda, while Esaies (Bede 3 [0525 (17.230.20)] takes the strong declension inflection -es.

The results regarding second declension names are similar in that Latin-inflected names are widely attested alongside those that have the Old English inflection -es, although the latter are predominant. As was the case with first declension names, the 29 types and 52 tokens that make up the second declension genitives with Latin inflections can be divided into two groups. The first one consists of five types and eight tokens, namely Gregorii (x3), Martini (Bede 1 [0235 (15.62.2)]), Maurici (Bede 2 [0002 (1.94.6)]), and Pauli and Petri (both in Bede 1 [0420 (17.90.18)]), which have the corresponding Latin inflection -i. Contrary to what was detected in the Martyrology (Ruiz Narbona 2023a, pp. 15–16), the attestation of -i genitives in the OEHE is rather limited and does not seem to follow the same textual restrictions as their counterparts in the Martyrology, where they only appeared in the introduction to each martyr’s story. In contrast, the majority of Latin-inflected genitives in second declension names, 24 types and 45 tokens, have the nominative-like ending -us, e.g., Paulinus (x3), Petrus (x3), Anatolius (Bede 5 [0612 (22.484.6)]), Dioclitianus (23), or Germanus (BedeHead [0078 (3.16.24)] to name but a few.

(23)	Be Dioclitianus (genitive) rice, & þæt he cristene men wæs ehtende.
	‘About the reign of Diocletian, and that he persecuted the Christians’
	BedeHead [0006 (1.6.16)].

Widely recorded as they are, Latin -us genitives are outnumbered by names with the Old English strong genitive inflection -es, with 208 tokens and 18 types. This is due to the fact that Cristes appears 160 times in the text. However, even if Cristes is left out of the equation, the results show that names with the genitive -es, such as Constantines (x2), Paules (x3), or Petres (x21), with 120 tokens are more prevalent than those with Latin inflections with 97 tokens.

Finally, it must be mentioned that seven examples of second declension names do not conform to the cases described above. Þeodoris (Bede 2 [0087 (3.106.2)]) and Euticetis (Bede 4 [0359 (19.310.3)]) have undergone a change in declension and present the Latin third declension genitive ending -is. Additionally, there is a group of five names that have the ending -e, e.g., Anastase (Bede 5 [0616 (22.484.10)]), Criste (Bede 4 [0197 (9.286.1)]), Marce (Bede 5 [0601 (22.482.25)]), Paule (24), and Petre (Bede 4 [0375 (20.314.1)]). None of them share a similar textual or syntactic context. Given that Crist’s genitive form is Cristes throughout all Old English texts, it may simply be the case that the scribe may have failed to add the -s to the genitive form in these particular instances.

(24)	forþon þe he ær sceare hæfde eastleoda þeawe Sancte Paule (genitive) þæs apostoles.
	‘For he previously had the oriental tonsure after the mode of the apostle Saint Paul’ Bede 4 [0022 (1.254.30)].

As was the case with previous groups, third declension names can take Latin or Old English genitive inflections. Notice, however, that no nominative-like endings are found in this group. Seven tokens and four types show the Latin genitive ending -is. These are the following: F/Vocatis (x4), Ismahelis (Bede 5 [0591 (22.482.11)]), Samuelis (Bede 5 [0593 (22.482.15)]), and Salomonis (Bede 5 [0596 (22.482.19)]). Names with the Old English strong genitive inflection -es appear in similar numbers with six types and 10 tokens including Moyses (x3), Agothones (x2), Iohannes (x2), Danieles (BedePref [0014 (4.10)]), Felices (Bede 5 [0614 (22.484.8)]), and Michaheles (Bede 5 [0018 (2.388.5)]).

In addition to the groups discussed above, there are also three occurrences in the genitive case of originally indeclinable names. Abaccuc (Bede 5 [0599 (22.482.22)]) and the already mentioned Deosdedit (Bede 3 [0632 (21.248.8)]) have no inflectional endings. However, by contrast, Isaces (Bede 5 [0591 (22.482.11)]) takes the predominant Old English strong ending -es.

The results concerning the genitive may seem difficult to interpret at first given the fact that the use of original Latin genitive endings is attested alongside the widespread Old English inflection -es in addition to a large number of names with a Latin nominative inflection. As for the latter, contrary to the case with the accusative or dative, nominative-like forms are attested in large numbers almost equivalent to the predominant Old English inflection. Therefore, they cannot simply be regarded as slips of the pen on the part of scribes. When faced with the abundance of -us genitives in the Enchiridion, Baker (1998, pp. 195–96, 198) concluded that the choice of Latin endings must have been influenced by the Old English inflectional system, in this case, based on formal similarities since like the Old English -es, these nominatives, as well as the Latin third declension genitive inflection -is, consist of a vowel plus the sibilant -s.

Except for the scarcely attested Old English -an inflection in Columban and original Latin genitive endings such as -e (Luce) and -i (Gregorii), all genitive endings discussed above consist of a vowel plus -s. This distribution of the evidence suggests that -s functions as the genitive marker, with the preceding vowel acting as a thematic vowel, with -a, -u, and -i attaching to the first, second, and third declension names, respectively. The only exception is -es, which is the only original Old English inflection that can be used in any class. Following this line of argumentation, then, genitive inflectional endings in masculine names in the OEHE can be reduced to two. First, the predominant ending consists of a thematic vowel + -s, and second, an original Latin genitive inflection such as -e or -i.

5. Conclusions and Further Research

This paper set out to analyze the inflections of Latin proper names in the OEHE with the aim of shedding light on the distribution of Latin and Old English inflections and the underlying patterns governing the seemingly arbitrary choice of inflectional endings, while at the same time providing insights into the dynamics of Old English nominal morphology and how it can interact with the Latin system.

In general terms, the results showed a considerable degree of variation with regard to the adaptation of Latin proper names to the Old English system. The few feminine names recorded as well as first declension names in -a, such as Attila and Columba, are integrated into the Old English weak declension and, therefore, have the inflections -a in the nominative and -an in the rest of the grammatical cases. These results match those of Ruiz Narbona (2023b); however, they must be considered tentative at best, given the scarcity of the data.

The situation concerning the majority of masculine names differs considerably. Although, on the whole, the inflectional paradigm of these names results from the interaction and mixture of both Old English and Latin patterns, the latter seems more relevant. This contrasts with the data analyzed by Ruiz Narbona (2023a) in the Old English Martyrology, where Old English inflection configured the general inflectional pattern.

The most significant factor underlying the choice of inflectional ending is grammatical case. Direct cases, that is, the nominative and the accusative follow Latin patterns almost exclusively. The predominance of Latin inflections in the nominative was already anticipated by Funke’s (1914, p. 144) mixed loanwords. The use of Latin inflections in the accusative, however, offers more interesting insights as to the adaptation of proper names. In the Old English masculine strong declension, which attracts original second declension Latin loans, both the nominative and accusative have a zero inflection and must be disambiguated by resorting to different morphosyntactic means, for example, the unambiguous accusative determiners þone and þisne. Such is the case of Latin proper names in the Martyrology (Ruiz Narbona 2023a, p. 12). The use of the Latin system, thus, offers an advantage in that it allows scribes to mark names as direct objects without any further additional means. In fact, Baker’s (1998, p. 191) results and the preliminary analysis of Ælfric’s Lives of Saints (Ruiz Narbona, forthcoming) support the view that this trend seems to be more widespread than the one in the Martyrology. Nevertheless, it must be pointed out that complete adaptation to the Old English pattern is still possible. As discussed above, familiar names such as Crist or Theodor, which feature prominently in the OEHE, can appear with no inflectional endings in the nominative or accusative as Old English strong nouns, although not necessarily in all occurrences.

Whereas direct cases are dominated by Latin inflections, the dative and genitive are characterized by the combination or fusion of Latin and Old English endings, even if in rather different ways. As for the dative, the analysis shows that inflections from both languages are evenly distributed. Neither the analysis of textual contexts nor the manuscript where forms are attested seems to be linked to the preference of one inflectional ending over the other. However, all attestations do share an important feature in that the dative case is never solely marked on the proper names themselves but also through additional means, mostly an Old English determiner in the dative. As a consequence, similar to the case with the accusative inflections in the Martyrology, my conclusion is that Old English morphosyntax seems to act as the primary dative marker, while the freely distributed endings -o or -e are regarded as secondary markers, which is contrary to what is found in other texts (Ruiz Narbona 2023a, pp. 14–15).

Inflections in the genitive case are perhaps the most interesting ones in so far as they show an example of the actual fusion of both Latin and Old English inflections and not just the use of both as in the dative case. Evidence suggests that the most widespread genitive inflection in Latin proper names in the OEHE is the result of the reinterpretation of the Old English genitive -es and several Latin inflections ending in -s, which are not necessarily original genitive inflections such as -us as already pointed out by Baker (1998, pp. 195–96). This shows that interaction between inflectional systems can go beyond the use of the inflections of one language or the other. This highlights that fusion of both systems is possible and that formal issues should be considered as a likely trigger for such fusional inflections. Furthermore, it is interesting to point out that -s would end up becoming practically the sole available genitive case marker in later periods of the language. Thus, the study of the reinterpretation of the vowel plus -s ending at this early stage might serve to shed light on the origins of this development.

The results are also relevant as far as the Old English inflectional system itself is concerned. If, as proposed by Gneuss (1996, p. 4), the power of attraction of loanwords can be regarded as a determining factor in assessing the dominant status of a given declensional type in a language, it is clear that both the strong and weak declensions are already dominant during the early period as all Old English inflectional endings discussed in this paper are only taken from these two declensions.

All in all, it must be borne in mind that the data analyzed in this paper represent the inflections of Latin proper names in just one Old English text. As has been discussed throughout this paper, the results differ when compared with other works. This shows that alternative systems to the one described here are possible. It would be insightful, then, to analyze how the inflections of these names operate in other texts in order to explore further factors that may contribute to the configuration of a different inflectional paradigm. Likewise, the analysis of later texts such as Ælfric’s Lives of Saints, which also features numerous Latin names, might provide relevant information with regard to the evolution of the adaptation of these names in the late Old English period.