Printer Friendly

The survival of Ceolfrith's 'Tobit' in a tenth-century insular manuscript.

We learn from the histories of the house that, after becoming abbot of the twin foundation of Monkwearmouth and Jarrow in 6go, Ceolfrith ordered three great pandects of the Vulgate bible to be made.(1) In 716, one of these, with a dedicatory inscription added and with Ceolfrith in attendance, was taken for presentation to St Peter's in Rome. The abbot died during the journey but the book reached its destination and was to survive (though not at St Peter's) as the |Codex Amiatinus', our earliest complete Vulgate bible and a magnificent testimony to the codicological, scribal, and textual accomplishments of late seventh-century Northumbria.(2) All that survives of the two other pandects, written for use in the home churches and probably completed before Amiatinus,(3) is a total of twelve leaves and some fragments of a thirteenth from one of them, all now in the British Library.(4) The copyists at Monkwearmouth-Jarrow seem to have had at hand good exemplars for most of the Vulgate books, usually Italian, and probably acquired by Abbot Ceolfrith himself or by Benedict Biscop during their several journeys to Rome, but there were exceptions. One of these was the book of Tobit, which seems to have been available in a Vulgate version of Italian origin which had become in many places corrupt. The text had therefore to be |improved', apparently without the aid of other Vulgate models, and consequently Ceolfrith's Tobit, as preserved in Amiatinus, contains many distinctive readings.(5) Not surprisingly, this text was used by Bede for the citations in his little commentary In Tobiam,(6) but no other evidence of its further dissemination, as far as I am aware, has ever been noticed.(7) However, such evidence does exist, in a manuscript probably written in Cornwall in the first half of the tenth century, or possibly in the late ninth, and now bound in the composite codex, Bodley 572 (S.C. 2026) in the Bodleian Library, Oxford. It contains a version of Tobit which preserves to a remarkable degree the distinctive Ceolfrithian text created at the turn of the seventh century in Northumbria.

I. Bodley 572

Bodley 572 comprises five separate manuscripts in a total of 107 folios, containing various theological, devotional, and regulatory texts, and has long been of interest to scholars.(8) The contents may be summarized as follows.

1. A single leaf containing most of a Missa propria Germani episcopi, in an insular hand (fo. 1).

2. An Expositio missae, in a continental hand (fos. 2-13) and the Liber Thobiae, in a mixture of continental and insular hands (fos. 14-25).

3. Two |epistles of Augustine', the second, in fact, a homily by Caesarius of Arles, written by an insular scribe, Bledian (fos. 26-40).

4. A Latin conversation lesson, with vernacular glosses (fos. 41-50).

5. Various devotional and regulatory texts (fos. 51-107).

The first four manuscripts, comprising the first fifty folios, all appear to belong to the tenth century and are of insular origin, although they show a great variety of scripts. The fifth was written in France in the ninth century. The whole codex was eventually at St Augustine's, Canterbury, but may not have been in its present form before the second half of the fifteenth century. The manuscript containing Tobit comprises three complete quaternions. The text is on fos. [14.sup.r]-[25.sup.v] and is written in several hands, both Caroline and insular, and often badly, but it is complete and includes the Hieronymian prologue beginning |C(h)romatio et Heliodoro episcopis'. The Cornish provenance (if not origin) of this section of the manuscript is suggested by the presence of three Old Cornish glosses. Although a date in the tenth century remains probable, for the present a more cautious assessment of 850-950 is desirable.(9)

Particular problems face the textual critic of the Bodley Tobit, for the text is untidily written by its several scribes and errors are frequent. Many of these have been corrected, a few of them more or less contemporaneously, but most in a variety of hands of the eleventh and twelfth centuries, which bear witness to the codex's early peripatetic existence. The line between error and what may be scribal idiosyncrasy is sometimes a fine one, but I have noted some sixty obvious mistakes, ranging from discordant verbal or nominal endings to disruptive omissions of words or phrases.(10) Erroneous omissions of single words include diebus (4: 3) and Deum (13: 12). Two long accidental omissions are confitemini Domino filii Israhel et in conspectu gentium laudate eum (13: 3) and sed et cum timore Domini nuptiarum conuiuium (9: 12). The latter is a clear example of homoeoteleuton, triggered by the preceding accesserunt ad conuiuium; the omitted words have been added above the line in what may be a more or less contemporary hand. Further long omissions occur in 1: 15 and 3: 15, where et monita salutis dabat eis and aut certe desuper terra eripias me, respectively, have been lost; but in both cases the narrative is not disrupted and it is possible that the omissions were already in the Bodley scribes' exemplar. Among incorrect verbal, nominal, and adjectival forms are liberatur for liberabitur (3: 21), sufficerant for sufficerent (6: 6), dare for daret (7: 11), recipies for recipias (9: 3), liberet for liberat (12: 9), expectu for aspectu (12: 21), and sepultis for sepultus (14: 2). Errors involving omission of letters include pra[n]dium (2: 1), conue[r] to (3: 14), perdu[c]ere (5: 14), no[c]te (8: 4), itine[re] (11: 1), blasfema[ue]rint (13: 16), and Ni[ni]uen (14: 2). There is dittography in conuiuiuium (7: 9), superiiori (11: 5), and potueerunt (12: 21). Other misspellings include iussus for iustus (3: 2), tum for tuum (3: 23), toliam for Tobiam (8: 23), and lonquinquo for longinquio (13: 14).

Many of the eccentricities and inconsistencies in spelling which occur are characteristic of manuscripts in insular script.(11) They frequently involve either the simplification of double consonants or the gemination of single consonants, as in seccantes (prologue), possitus (1: 2), pretermitas (4: 6), comittetur (5: 27), ingresus (5: 10, 11; 6: 18 etc., but cf. ingressus 9: 8), suficiebat (5: 25; cf. sufficerant 6: 6), dexterram (7: 15), obtulli (12: 12), and missit (12: 14, 20). Vowel confusion occurs, as in cremen (4: 13), recipit (9: 6), and doobus (14: 1). A fourfold pattern of inconsistency in a single word is shown by chirographum (4: 21), chirografum (4: 22), cirografum (5: 3 and 9: 6), and cirographum (9: 3).(12) In the prologue, agiographa was written originally as opogripha, and altered later to apogripha.(13) 13 Not unexpectedly, proper names have caused particular problems, resulting often in unique forms, some of which have been corrected by contemporary or later hands. Thus, Senhec for Sephet (1: 1; cf. Am. Sephec), Salminasar for Salmanas (s) ar (1: 2, where the scribe has added ar above the line, and 1: 13, 18), Sincharif for Sennacherim (1: 21),(14) Asmodiu for Asmodeus (3: 8). Tobia(s) is more often than not spelled without an h, but compare Thobia(s) (which is the invariable form in Amiatinus) in the prologue, 2: 17; 4: 1; 7: 5; 8: 2, etc. Surprisingly, there is consistency in the spelling of Gabellu(s) (1: 17; 9: 3, 6, 8, etc.), although the form with the doubled consonant is very rare in other Latin witnesses (and Amiatinus invariably has Gabelus).

We have no way of knowing for certain how reliable the exemplar used by our Bodley scribes was, but their obvious lack of competence in writing (and presumably in understanding) Latin makes it likely that a majority of the errors was theirs. If this is the case, however, it becomes difficult to assess a further important stratum of readings in the manuscript. These are the variants which are not grammatically erroneous and which may even produce plausible alternative versions of the expected text but which are nevertheless unique among manuscripts of Tobit, at least according to available collations.(15) What are we to make, for instance, of putaret for nutaret in 7: 11 (Hier. et cum nutaret et non daret ullum petenti responsum), or of amicis for unicis in 8: 19 (Hier. misertus es autem duobus unicis)? It is highly probable that such readings were the unwitting creation of the careless or incompetent Bodley scribes. Yet the possibility remains that at least some of them were in the Bodley scribes' exemplar and that they represent traditions established, however erroneously, earlier in the text's transmissional history. In the analysis and discussion which follow, therefore, I have given the status of |variant' to all readings in which grammatical proprieties are observed and where, arguably, narrative sense remains intact.

II. The Bodley-Amiatinus Connection

The following passage from the Bodley Tobit, selected more or less at random, will serve to illustrate the close textual connection between Bodley and Amiatinus and also the nature of the differences. The passage (Tob. 10: 1-4, on fo. [22.sup.v]) is given in parallel with both the Amiatinan version and the probable Hieronymian version.(16) Words and phrases which it is useful to compare in the three versions are italicized; those in which Bodley and Amiatinus agree to the exclusion of the Hieronymian and all other collated versions are, in addition, underlined. My Amiatinan extracts are reconstructed from the critical apparatus of the Rome Biblia Sacra and are presented according to the conventions employed by the editors of that work in their main text. The Bodley extract is presented in the same way, although in the manuscript the text is written continuously, with a variety of punctuation marks.(17) As in all further citations from the manuscript in this article, abbreviations are expanded silently and capital letters given to proper names, but errors and orthographical peculiarities are retained.
    Jerome                Amiatinus            Bodley 572
1. cum uero moras       cum uero moras       Tunc uero moras
   faceret Tobias       faceret Thobias      fecere tobias
   causa nuptiarum      nuptiarum causa      nuptiarum causa
   sollicitus erat      sollicitus erat      sollicitus tobias
   pater eius Tobias    Thobias pater        pater eius erat
   dicens               eius dicens          dicens
   putas quare          putas quare          putas quare
   moratur filius       moratur filius       moratur filius
   meus aut quare       meus aut quare       meus aut quare
   detentus est ibi     detentus est ibi     detentus est ibi
   putasne Gabelus      putasne Gabelus      putasne gabellus
2. mortuus est et       mortuus est et       mortuus est et
   nemo illi reddet     nemo illi reddet     nemo illi reddet
   pecuniam             pecuniam             peccuniam
3. coepit autem         et coepit            et cepit
   contristari nimis    contristari nimis    contristari nimis
   ipse et Anna         ipse et Anna         ipse et anna uxor
   uxor eius cum eo     uxor eius cum eo     eius cum eo
   et coeperunt ambo    flentes eo quod      flentes eo quod
   simul flere eo       non die constituto   non die constituto
   quod die statuto     reuerteretur filius  reuerteretur filius
   minime               eorum ad eos         eorum ad eos


reuerteretur filius

eorum ad eos
4. flebat igitur mater  dicebat igitur cum   Dicebat igitur
   eius                 multo dolore         cum multo dolore
   inremediabilibus     mater eius           mater eus


lacrimis atque

dicebat
   heu heu me fili      heu heu me filii     heu me fili mi ut
   mi ut quid te        mi ut quid te        quid te missimus
   misimus              misimus              peregrinari
   peregrinari          peregrinari
   lumen oculorum       lumen oculorum       lumen oculorum
   nostrorum            nostrorum            nostrorum
   baculum              baculum              baculum
   senectutis           senectutis           senectutis
   nostrae              nostrae              nostrae
   solacium uitae       solacium uitae       Solacium uite
   nostrae              nostrae              nostrae
   spem posteritatis    spes posteritatis    spes posteritatis
   nostrae              nostrae              nostrae


In the distribution of variations, this passage is characteristics of the Bodley Tobit as a whole. There are no concentrations of particular sorts of variation which would suggest a separate textual history for different parts of the book. Obvious scribal errors in the Bodley version of the passage are facere for faceret (10: 1) and cm for eius (10: 4), and examples of insular orthography are peccuniam (10: 2) and missimus (10: 4). The other variations may conveniently be categorized as follows:

1. Readings which are unique, as far as is known, to Bodley 572. Thus, tunc for cum in 10: 1 is not an Amiatinan reading and is not to be found in any other collated manuscript. The variant version of Hier. erat pater eius Tobias, although paralleled in Amiatinus in respect of the ordering of the last three words, is unique in its placing of erat at the end of the phrase. In these two cases, no semantic change is involved, but there are many other unique variations throughout Bodley (such as those in 7: 11 and 8: 19, noted above), where there is alteration in meaning, sometimes radical.

2. Readings which are paralleled only in Amiatinus (and, where the relevant passages are used, in Bede's commentary on Tobit).(18) Such readings, underlined in the passage, range from the simple substitution of one conjunction for another to the restructuring of whole clauses. In 10: 3, et c(o)epit for coepit autem is not in itself a remarkable emendation and its correspondence with Amiatinus might be coincidental, but this can hardly be the case with the two complex readings in 10: 3 and in 10: 4.

3. Readings which are paralleled in Amiatinus but are not unique to that manuscript, occurring in each case as variants in at least one other manuscript of Tobit and sometimes in a number. Thus, spes for spem in 10: 4 occurs in all the Theodulfian bibles and in manuscripts U, C, and [[sigma].sup.T].(19)

4. Readings which are not paralleled in Amiatinus but which do, again, occur in other manuscripts. Thus, heu for heu heu in 10: 4 occurs in [L.sup.a] (as a correction),(20) S, all the Alcuinian bibles, and a number of others.

Some initial idea of the relative importance of each category of reading in the Bodley Tobit may be gained by comparing the number in each category, which I have noted throughout the text, including the prologue:
1. Unique readings                                          141
2. Exclusive Bodley-Amiatinus readings                      140
3. Non-exclusive Bodley-Amiatinus readings                  142
4. Non-Amiatinan, non-unique readings                        66
Total readings                                              489


It will be seen that some three-fifths of the variant Bodley readings examined (the sum of categories 2 and 3, Viz. 282) are shared with Amiatinus, and almost half of these (140) occur in no other manuscripts. However, even these figures grossly understate the debt of Bodley 572 to Ceolfrith's Tobit. As the extract has shown, what I have defined as a |reading' is a self-contained unit of difference, which may be a single word or a phrase or a whole sentence, and such a sentence may itself comprehend several types of variation, including word order, syntax, and vocabulary. It is precisely this sort of |compound' variant reading in Bodley which is almost exclusively Amiatinan. The chances of coincidental parallelism here are usually remote, whereas, as we shall see, many of the variant readings not shared with Amiatinus (in the majority of which the unit of difference is a single word) are quite likely to have arisen accidentally during transmission of the text.

Among the variants for which Bodley and Amiatinus are the only witnesses, I have noted twenty-nine important compound readings. In the following selection of them, the Hieronymian version is given first,(21) then the Bodleian version. The Amiatinan version in each case is identical with the latter, except where indicated. 2: 10 ueniens domum/in domum suam ueniens 5: 14 dixit itaque illi Tobias/Tobias ei dixit [Am. Thobias) 6: 4 quod cum fecisset adtraxit eum in sicco et palpitare coepit ante

pedes eius/quod cum fecisset palpitare cepit in siccum ante

pedes eius [Am. coepit] 6: 11 sed neque masculum neque feminam ullum habet alium praeter

eam/et neque masculum neque feminam absque ea ullum habet 6: 18 continens esto ab ea/contine te ab ea 8: 11 et abierunt pariter ut foderent sepulchrum/cum quibus abiit

fodit sepulcrum [Am. cum quibus abiit et fodiit sepulchrum] 8: 16 et benedixerunt Deum Raguhel uidelicet et Anna uxor eius et

dixerunt/et benedixit Deum Raguel cum uxore sua dicentes

[Am. Raguhel, dicens] 9: 8 fleuit itaque Gabelus et benedixit Deum et dixit/et flens

Gabellus benedixit eum dicens [Am. Gabelus, Deum dicens] 10: 7 ut procul uideret eum si fieri possit uenientem/ut procul uidere

eum posset uenientem 11: 5 unde respicere poterat/inspicere cupiens 11: 10 unde contigit ut exurgens/quapropter exurgens [Am. exsurgens] 11: 14 et sustinens quasi dimidiam fere horam coepit albugo ex oculis

eius quasi membrana oui egredi/et sustinens fere dimedium hore

spatium cepit ex occulis eius quasi membra oui albugo egredi

[Am. dimidio horae spatio coepit, oculis, membranum] 14: 4 reliquum uero uitae suae/ceteris temporibus uitae suae

It will be seen that where there are differences between Bodley and Amiatinus in these readings they are always minor, some of them being merely orthographical. In 9: 8, Bodley's eum for Deum is clearly erroneous, and may be blamed on the Bodley scribe: there is no obvious antecedent for the pronoun in the preceding cola, and the result is confusion. In 8: 11, we may assume that et was lost in simple error also. In 11: 14, Amiatinus is unique in its use of the ablative to express the duration of time;" the Amiatinan membranum, on the other hand, is a frequently-occurring alternative for the plural membrana (mistakenly written membra in Bodley), particularly in the earlier manuscripts." In the case of the expression of time, the Ceolfrithian ablative may have been modified during transmission to Bodley, perhaps by a scribe or corrector unhappy with the unusual choice of case. However, it is equally possible that the ablative reading was never in the Ceolfrithian exemplar and that the variation occurred unilaterally in the copying of Amiatinus. The idiosyncracies of Amiatinus and the probable origin of the Bodley text in one of the other two pandects are discussed further below.

The many one-word readings in which Bodley coincides with Amiatinus to the exclusion of other manuscripts include the following: latinorum for latinis (prologue, before auribus), gentium for gentilium (1: 12), fortassis for forsitan (3: 19, before uiro), exproprationem [Am. exprobratione] for probatione (3: 21), corroborabitur [Am. conr-] for coronabitur (3: 21), domino for deo (5: I3), tulerunt for sustulerunt (6: 6), reponi for servare (6: 7), discumbere Om. (7: 9), demonem [Am. dae-] for daemonium (8: 3), ducat for perducat (10: 11), uenerunt for peruenerunt (11: 1), praestularetur [Am. praesto-] for specularetur (11: 6), pretioso [Am. prae-] for candido (13: 22), and cognati for cognatio (14: 17). In addition, I have noted forty cases of parallel variation in the use of conjunctions, adverbs, pronouns, or prepositions (such as quoniam for quomodo in 9: 5, illa for alia in 11: 18, and the omission of semper in 2: 13), and a further thirteen cases where verbal forms are shared (such as lacrimabantur for lacrimatae sunt in 7: 8, sciens quia for scias enim quoniam in 11: 8). There are twenty-four exclusive Bodley-Amiatinus variations in word order, thirteen of them involving three or more words. In 12: 4, for instance, de omnibus quae ablata sunt medietatem asumere sibi [Am. allata, assumere] occurs for medietatem de omnibus quae adlata sunt sibi adsumere. There is further word order variation in some of the compound variant readings noted above.

Clearly the value of individual variants as evidence of textual connection varies enormously. While the parallel replacement of seruare by reponi would look significant in any circumstances, the loss of verbal prefixes (as in tulerunt, ducat, uenerunt) and variation in the naming of the deity (Domino for Deo) are not in themselves impressive as sourcing evidence. However, when the sheer number of such minor correspondences is taken into account, and is added to the evidence of the compound readings, their significance cannot be ignored. We may indeed ask at this stage whether there are any readings exclusive to Amiatinus, and therefore presumably representative of Ceolfrith's recension, which are not found also in Bodley. In fact I have noted only fifteen, excluding the few apparent copying errors identified below. None is of the distinctive type which we have seen to be so characteristic of the Amiatinus-Bodley connection. They are propter blasphemia sua for propter blasphemiam suam (1: 21),(24) potuisset mori for mori potuisset (4: 1), in diem for in die (4: 10), superbiam + autem (4: 14), respondit patri suo et om. (5: 1), quidem om. (5: 3); cf. Bodl. quid, with em added above the line in what may be a contemporary hand), ut dum adhuc uiuo recipias for dum adhuc ut uiuo recipias (5: 4; cf. Bodl. dum adhuc uiuo ut recipias, a word order shared only with two Alcuinian manuscripts), requiris for quaeris (5: 17; cf. Bodl. inquaeris), ueniant for eueniant (6: 15), qui sunt om. (6: 16), coniugia for coniugium (6: 17),(25) et om. (8: 4, before cras), in om. (10: 10, before puellis), Dominum for eum (14: 11; cf. Bodl. Deum)(26) and quinta generatione for quintam generationem (14: 15; cf. Bodl. quartam generationem).

Some of these may well be original Ceolfrithian readings which were in the exemplar from which the Bodley version derived but were lost or modified during transmission to it. Bodley's unique inquaeris is more likely to have derived from requiris than from quaeris (5: 17), while Deum (dm) was probably an accidental modification of Dominum (dnm), not eum (14: 11).(27) Yet it is probable that some of the variations, especially the six omissions, were peculiarities only of the Amiatinan copy of Tobit. They may not have been in the Ceolfrithian exemplar at all and thus may not have reached the other two pandects. Whatever the line of transmission from Monkwearmouth-Jarrow to Bodley 572, Amiatinus itself is scarcely likely to have had any direct part in it, having left Northumbria for Italy in 716. The probable role of one of the other two pandects is discussed in the concluding section. It must be emphasized, however, that Amiatinus was written in general with great accuracy.(28) Some dozen original errors were corrected more or less contemporaneously, at least one of them by the scribe himself (who made good his omission of ab in 2: 13) and the others apparently in a different hand. I have noted only five further Amiatinan readings which can be said with confidence to be erroneous and which remained uncorrected. They are particula for particulam (6: 8), Thobiam for in obuiam (11: 10),(29) uenient repeated (13: 14), sternuntur for sternentur (13: 22),(30) and filios for filiis (14: 14).(31) Some idea of the likely scale of textual variation between Amiatinus and the other two pandects may be obtained from a comparison of the text of the one which survives in part in the British Library, as noted above, with the corresponding Amiatinan text. The results of this collation cannot be presented in full here, but they show that in the folios from 3-4 Kings, in a volume of text totalling approximately half as much again as the whole of Tobit, Amiatinus and its sister pandect differed originally at only some thirty points. All the variations are minor, and some are errors which were subsequently corrected.

III. The Independent Element in Bodley 572

While the evidence for a direct textual connection between, the Bodley Tobit and Ceolfrith's version is thus overwhelming, there remains the substantial number of readings it appears to be independent of Ceolfrith's text. The readings are fairly evenly distributed throughout the book. They may be divided into those which are paralleled in one or more other known copies of Tobit and those which are apparently unique to Bodley 572. I have noted 66 of the former and 141 of the latter, and together they account for a little more than two-fifths of all the readings examined.

It is in the category of paralleled readings that we would expect to find signs of any influence on Bodley's text of textual traditions other than the Ceolfrithian. In fact, the evidence of the 66 such variants (among some 500 readings which have been examined throughout the text) is not impressive, for two-thirds of them involve only minor variants, such as different forms of pronouns, prepositions, or conjunctions, or alternative verbal, nominal, or adjectival endings. The rest include two simple variations in word order, seven omissions or additions of nouns or verbs, and thirteen substitutions of one verb or noun for another. Examples of shared readings are infantia + sua (1: 10; also in [[gamma].sup.A]),(32) auertatur a te for a te auertatur (4: 7; also in F and K), a om. (4: 11; also in E, before correction), ingres(s)a + est (8: 15; also in one non-Turonian Alcuinian manuscript), recta for recte (10: 11; also in F and, before correction, in K), Deum for eum (13: 4; also in S and in W), 13 and habitabunt for inhabitabunt (14: 8; also in U and one twelfth-century Spanish manuscript). The most noteworthy readings are ne + forte (8: 12; in U only), nomen for semen (9: 11; in [L.sup.a] only),(34) recepit for suscepit (12: 3; also in F, K, as a correction, two non-Turonian Alcuinian bibles, and a twelfth-century Spanish bible), duobus for duodecim (14: 2; also in [L.sup.a], before correction, in W, and, as a correction, in S; and cf. duo in K, before correction, and C), and quartam for quintam (14: 15; also in U, as a correction, S and W). It will be noted that most of these readings are restricted to eighth-century witnesses (the exceptions being the two verb variations, which are of a sort quite likely to occur spontaneously) and they may represent fairly common early Vulgate textual traditions, although none occurs, of course, in Amiatinus. The variant in 14: 2, duobus, is of particular interest. The higher figure, duodecim, which is incorrect, is the ubiquitous Old Latin version and derives from the Greek of the Codex Sinaiticus.(35) Duodecim occurs in a majority of bibles copied before the twelfth century (including all the Theodulfian and most of the Alcuinian), after which duobus is almost invariably found. The editors of the Rome Biblia Sacra have restored the latter as the Hieronymian reading, which thus differs from that of Amiatinus and, presumably, the other Ceolfrithian pandects. Bodley's use of duobus may thus indicate deliberate emendation at some point in its text's transmission. The same explanation may account for the variant nomen for semen in 9: 11. This is unlikely to have been a palaeographical error (though it could have resulted from mishearing, if the text were dictated). The appearance of both nomen and duobus in the eighth-century [L.sup.a] raises the possibility that these (and perhaps other) readings derive from older textual traditions which somehow came to influence Bodley. Such external influence is considered further below.

Of the sixty-five paralleled Bodley readings, eighteen occur only in one other collated manuscript, and eleven in only two manuscripts. Among the remainder, a number have a fairly wide distribution, as the above examples have shown. Overall, E is the most persistent witness to Bodley readings, with twenty-six parallels (two of them unique to E; five were later emended). S follows with twenty-two readings (only one unique to S, and four emended), W with twenty, F with nineteen, U with eighteen (five of them unique to U, and five emended), and K with sixteen (six emended).(36) The Theodulfian and Alcuinian traditions witness twelve and nineteen, respectively, with readings shared in only four cases. Only four readings are shared with L (each as a correction) and three with the restored [L.sup.a]. The figure for E is not dramatically larger than those of the other manuscripts and no significance can be attached to it, especially as it includes none of the more noteworthy of the independent readings listed above. Thus no major pattern of correspondence is discernible between Bodley and any of the continental manuscripts, and many of the parallels may be coincidental.

It is left to consider the persistent stratum of readings in Bodley - 141 in all - which have no parallels in the manuscripts collated in the Rome Biblia Sacra or in others which I have examined. May we assume them to be in general the result of accidental emendation of the Ceolfrithian text during its long transmission, or do they offer evidence of deliberate revision at some stage during that transmission, perhaps making use of a version of Tobit in an independent (and otherwise unknown) textual tradition? Analysis of the 141 variants involved shows that only fifteen involve simple variation in the grammatical form of nouns or verbs (such as mortuum for mortuos in 2: 8, exigitis for exegisti in the prologue, and expauesceret for expauescens in 6: 3). Eleven are variations in word order, with transposition of two words in five cases, and up to five words in the others.(37) Of twenty-two omissions, fifteen are of pronouns, prepositions, or conjunctions; the others are cotidie (1: 3), et monita salutis dabat eis (1: 15), aut certe desuper terra eripias me (3: 15), fuerit (3: 21, after tribulatione), sanum tibi (5: 20), uxorem (6: 13), and turbati sunt et (12: 16).(38)

Nearly a quarter of the unique Bodley readings (thirty-three) are additions, all but two of them being single words. Many of them, including sixteen pronouns, prepositions, or conjunctions, add definition or a sense of completion to the passages in which they occur; thus, superioribus+partibus (1: 1), in sinistra + parte (1: 1; cf. Hier., Am., etc. in sinistro), orationem + suam (3: 12), ad te + domine (3: 14), illum + scio (5: 2),(39) recipias eam + pecuniam (5: 4; cf. Hier., Am., etc. recipias ea), and et + sit (13: 1). Others more positively amplify the sense, while not essentially altering it, as the following examples show.

4: 13 fornicatione + abstinere. The almost invariable Vulgate version is adtende tibi fili mi ab omni fornicatione. Only in one other manuscript, in a correction to L, is a second verb used, and it is again abstinere, but the syntax is different: adtende tibi fili et abstine ab omni fornicatione.

5: 5 inuenit + cito. This is an appropriate amplification of tunc egres sous Tobias inuenit iuuenem splendidum.

5: 22 uale + dicere. This is an uncharacteristic addition, for it oddly muddies the sense of fecit Tobias uale patri suo et matri (thus Am., Bodl., with U and probably E; cf. Hier. patri et matri suae), by implying that Tobit's farewells are made through an intermediary.

5: 24 + utinam. The adverb is prefixed to numquam esset ipsa pecunia, thus reinforcing the emphatically negative sense of numquam.

11: 5 delonginquo + uidere uenientem. The addition occurs in a passage which describes how Tobit's mother positions herself to watch for his return, and the Bodley version of the colon preceding the addition is known otherwise only in Amiatinus: inspicere cupiens delonginquo; cf. Hier. unde respicere poterat de longinquo. We might have expected the Bodley addition to include an accusative pronoun also. Bodley's curious and unique treatment of the colon preceding this one in 11: 5 is discussed below.

13: 9 uero + dilectabor. This addition (corrected, perhaps contemporaneously, to delectabor) occurs within a version of the passage which is otherwise known only in Amiatinus: ego uero et anima mea in eo laetabitur; cf. Hier. ego autem et anima mea in eo laetabimur. The singular verb in Amiatinus is inappropriate after the double subject. The Amiatinan version (in eo laetabitur) was used also in [L.sup.a], before correction to in Deo laetabimur, a reading which occurs also in [sigma].sup.T].

14: 3 annorum + uixit antequam and recepit + lumen. These additions, along with a unique word substitution (septuagenarius for sexagenarius), occur in a passage which summarizes the chronology of the elder Tobit's loss and subsequent recovery of his sight. The first addition quite alters the syntax of the first part of the passage, which in Jerome's version (shared by Amiatinus) is succinct and balanced: quinquaginta namque et sex annorum oculorum lumen amisit sexagenarius uero recepit; cf. Bodl. .l. nanque et .vi. annorum uixit antequam occulorum lumen amisit. Septuagenarius uero recepit lumen. The age at which Tobit regained his sight was not specified in any of the Greek or Old Latin versions. The first figure, which gives the age at which Tobit became blind, was subject to considerable variation.

The greatest number of the unique variants in Bodley (sixty-six, or nearly half of the total) involves the substitution of one word or phrase for another. Some substitutes are superficially similar enough in form to the correct word or phrase for us to be fairly certain that they arose through inattentive copying or the copying of corrupt or barely legible passages. Others are |visually' quite different, so that deliberate emendation may be a more likely explanation. A little fewer than a fifth of variants involve simple replacement of one preposition, conjunction, or pronoun by another (such as de for a in 2: 10 and quoniam for quia in 3: 5).

Most of the others are verb or noun substitutions.

In the case of the verbs, the simple loss or addition of prefixes is often involved, or one verb is replaced by another (or by a verbal phrase) which is semantically similar. Sometimes, however, the substitutes have quite different meanings, although these are generally (and sometimes surprisingly) appropriate, or at least plausible, in the context. The full list of verb substitutions is: uenisset for deuenisset (1: 11), regresus [sic] for reuersus (1: 21), perierunt for paruerunt (2: 22; the variant occurs also in one thirteenth-century Paris bible), seruasti for conseruasti (3: 19), inq[uae]ris for quaeris (5: 17; cf. Am. requiris), exortaretur for hortaretur (7: 9), putaret for nutaret (7: 11), alligauit for religauit (8: 3), et adora for deprecemus (8: 4; cf. Am. et oremus, with E and U), sol luceat for inlucescat (8: 14; cf. Am. lucescat), deberet for deuiniret (8: 24; cf. Am. eveniret, before correction to deveniret), docuit for fecit (12: 3), abscondere for condere (12: 8), stamus for adstamus (12: 15), credite for cantate (12: 18; cf. Am. decantate, with a number of other MSS), and exaltanti for excitauit (13: 23; cf. Am. exaltauit, with most other MSS).

The unique noun and adjective substitutions, which include some of Bodley's most interesting readings, are de cubili for de accubitu (2: 3; the use of cubile is barely suitable in the context of an eating-place), patentia for spes (2: 16), opera for iudicia (3: 2), temtationem for tempestatem (3: 22), flumen for fluuium (6: 1), iaecores [corrected in another hand to iaecoris] for cordis (6: 8), ciuitate for captiuitate (7: 3), ignis for uiuos (8: 2), amicis for unicis and Deus for Domine (8: 19), et ut leui gradu for et lento gradu (11: 3), in superiiori [sic] cubiculo montis for in supercilio montis (11: 5), lumen celi et filium meum for Tobiam filium meum (11: 17), spiritali for inuisibili (12: 19),(40) in uoluntate sua for ciuitatem suam (13: 19), and septuagenarius for sexagenarius (14: 3).

Other unique substitutions not included in the categories above are ea for ita (2: 17), omnes qui for eos mox ut (3: 8),(41) quoniam in itinera for et omnia itinera (5: 8), illam tradam for eam non tradam (7: 14), Post nos for iter nostrum (11: 3)(42) ego enim for et enim (12: 18), atque for omnes (14: 9), and ex quo die for ex eo (14: 12). Finally, there are a few textual simplifications, such edere et for aut edere aut (2: 21; cf. Am. edere aut), ambo for pater scilicet et filius (12: 5) and Deo for tam Deo quam hominibus (14: 17).

In trying to assesss the signifcance of these unique Bodley variants, we meet again the problem of how to distinguish between error and deliberate emendation. The ability of careless scribes to produce plausible alternative readings is well illustrated in 2: 22, where Bodley has perierunt for paruerunt. The reading occurs in a colon which is shared without significant variation by all collated versions, including Amiatinus: Hier. manifeste uana facta est spes tua et elemosynae tuae modo paruerunt. In this correct version, which uses pareo, the sense is ironic: Tobit's wife suggests to him that the result of his persistent almsgiving is now only too apparent, in that his reward has been nothing but a succession of troubles. To use pereo instead does not (arguably) contradict the fundamental meaning, for Tobit's almsgiving has been indeed quite wasted (or so it seems) and its intended effects destroyed. It may be noted that the same copying error was made by the scribe of one of the early thirteenth-century Paris bibles.(43) Again, in 7: 11, putaret for nutaret may be just about acceptable, if puto is understood absolutely, in the sense of |consider', thereby offering perhaps a plausible substitute for the correct |hesitate'. On the other hand, it is difficult to make sense of intrauit for nuntiauit in Hier. reuersa nuntiauit bonum nuntium (8: 16),(44) which must therefore be classed as an error and has been omitted from the above list. So also has the reading cadit for calida in Hier. ex nido hirundinum dormienti illi calida stercora insiderent super oculos eius fieretque caecus (2: 11).(45) However, although cadit was doubtless the result of scribal carelessness, it is interesting to observe that, had a corrector or subsequent copyist made a simple adjustment of its termination to concur with insiderent, a credible variant version of the colon would have been established.

Yet even if we need not doubt that many of the simpler Bodley variants originated accidentally, through inattentive copying, it is scarcely likely that all can be attributed to the Bodley copyists themselves, whose deficiencies have been noted. These copyists are even less likely to have produced the more complex variants in the above lists, those which seem to have involved a conscious process of |editing' or those which may have been reconstructed from illegible or corrupt passages in the available copy text. It therefore seems likely that the Bodley scribes worked from an exemplar already containing a large proportion of the unique variant readings, although we can only guess in each case when, and under what circumstances, they were established. A large-scale effort to revise the text at some point against other non-Ceolfrithian Vulgate texts can probably be ruled out; it is inconceivable that such revision would have left untouched so many Ceolfrithian idiosyncracies. Nor does it seem likely that our Bodley text represents a recension which was originally independent of Ceolfrith but was thoroughly revised at some stage, using a Ceolfrithian exemplar. It is hard to allow that, if such revision had been thought desirable and the available Ceolfrithian version had been considered authoritative, the independent text would not simply have been abandoned and the Ceolfrithian used in toto for further copying.

The independent Bodley variations which cannot simply be dismissed as accidental include many of the additions which, as noted above, seem to represent a deliberate attempt to clarify, such as abstinere (4: 13) and uixit antequam (14: 3). In the case of emendations like quoniam in itinera for et omnia itinera (5: 8) and in uoluntate sua for ciuitatem suam (13: 19), an origin in corrupt versions of the original phrases is easy to imagine, but the reconstructions, if such they be, are not accidental but are skilfully made. Deliberate alteration of readings seems to have occurred in some instances, such as patientia for spes (2: 16). Bodley's iaecor(i)s for cordis in the colon containing Raphael's advice to Tobit in 6: 8 (Hier. cordis eius particulam si super carbones ponas fumus eius extricat omne daemonium)(46) may be a deliberate alteration also, made on the analogy of the sequel in 8: 2, where Tobit acts on the advice and iecur is used (Hier. recordatus itaque Tobias sermonem angeli protulit de cassidile suo partem iecoris posuitque eam super carbones uiuos).(47) It is surprising that similar emendations (whether of cordis to iecoris in 6: 8 or of iecoris to cordis in 8: 2) were never apparently made in any other Vulgate manuscript. Jerome's version seems to have been unique in its use of only one organ of the fish in these two instances; the Old Latin, Greek, Hebrew, and Aramaic versions consistently have heart and liver.(48) Possible explanations of two other emendations may be suggested. The substitution of lumen caeli et for Tobiam in 11: 17 (uideo lumen caeli et filum meum for Hier., Am. uideo T(h)obiam filium meum) echoes the wording of the promise made by Raphael shortly before, in 11: 8 (Hier., Am., Bodl. et uidebit pater tuus lumen caeli et in aspectu tuo gaudebit [Bodl. in om.]), and looks like a deliberate amplification. In 11: 5, the extraordinary substitution of superiiori cubiculo for supercilio (Hier. Anna uero sedebat secus uiam in supercilio montis; Am., Bodl. uiam + cot(t)idie) repeats a phrase from 3: 10: Hier., Am., Bodl. ad hanc uocem per (r) exit in superiori cubiculo domus suae. May we see here the reflex of a scribe, perhaps working from dictation, in whom supercilio prompted unconscious amplification to a remembered phrase which had compelling phonic similarities? Or did an emender expand deliberately, perhaps being familiar with the sense of cubiculum as a raised sitting place (one of its classical uses) rather than as an upstairs room? The theory that an over-zealous emender worked through this part of Tobit at some time is supported later in the same verse (11: 5), where Bodley, as noted above, makes the unique and unnecessary addition of uidere uenientem to a colon which is otherwise uniquely Amiatinan in form.

Finally, there is reason to suspect that at least one of Bodley's unique readings, and possibly a second, was influenced by a recension of Tobit which followed an Old Latin textual tradition, at least in part. In 3: 2, Bodley's opera for iudicia in Hier., Am. et omnia iudicia tua iusta sunt has no parallel in Vulgate manuscripts, where variations in the clause are few."" However, opera for iudicia is used invariably in the Old Latin versions. In three of these the noun is then qualified, not by iusta, but by magna (et omnia opera tua magna sunt), which is the sense conveyed in the Aramaic and Hebrew versions also,(50) but a fourth has, like Bodley, iusta. This is in |Sangermanensis 15', now Paris, Bibliotheque Nationale, lat. 11553, the surviving (though incomplete) second volume of a bible written before 820, probably at Saint-Germain-des-Pras.(51) It transmits in the main a very good Vulgate text, but Tobit, Judith, and 1 Maccabees 1-14: 1, along with an incomplete series of canticles, are Old Latin in form. The Gospels show a mixing of traditions, with the Old Latin element strongest in Matthew and then decreasing. In 3: 2, lat. 11553 and Bodley thus concur with the Greek version of the Codex Sinaiticus,

(52) The second case of possible Old Latin influence occurs in 8: 2, where Bodley substitutes ignis for uiuos in et posuit super carbones uiuos, although in other respects this version follows Amiatinus (cf. Hier. posuitque eam super carbones uiuos). Bodley's variant is again without parallel in the collated Vulgate versions. Most Old Latin witnesses differ little from the Vulgate here,(53) but once again lat. 11553 is an exception, providing the single recorded occurrence of ignis for uiuos. Caution is necessary, however. The phrase carbones ignis could have been well known to the Bodley scribe from several scriptural locations, notably Pss. 17: 13 and 139: 11, and accidental substitution of ignis for uiuos cannot be ruled out.(54) The possibility that other |unique' variants in Bodley are owed to Old Latin sources remains open. Full collations of all the Old Latin versions of Tobit which survive are not yet available and we must wait for the Beuron edition.

IV. Conclusion

The overwhelming debt of the Bodley Tobit to Ceolfrith's Tobit is beyond dispute. Yet the large number of unique readings, which cannot all be the result of error, and of rare readings, at least some of which have apparent Old Latin antecedents or parallels in earlier Vulgate manuscripts, suggests that a deliberate effort to |improve' the text was made at some point during its transmission from Northumbria to the Cornish codex. It may be that the text had fallen into a state of partial corruption similar to that which had been suffered earlier (if present assumptions are correct) by the Italian text which formed the basis of Ceolfrith's version at Monkwearmouth - Jarrow. The materials at hand for use in the later improvements may have included a version of Tobit transmitting some type of |mixed' textual tradition influenced, in varying degrees, both by early Vulgate and by Old Latin texts. Although evidence for the late survival of such traditions in the insular area comes largely from Ireland, the use of a mixed, or even largely Old Latin, biblical tradition also in Wales as late as the fourth quarter of the ninth century is confirmed by citations in Asser's biography of King Alfred.(55)

The textual history of the Brodley Tobit is unlikely ever to be known in full, and only a more precise dating of the manuscript will reveal just how long that history was. However, some circumstantial evidence enables us provisionally to reduce the chronological gap in our knowledge, while at the same time moving the focus of our attention from Northumbria southwards and westwards to the Hwicce. As noted above, there survive in the British Library, folios from one of the two Ceolfrithian pandects which were deposited in the home churches of Monkwearmouth and Jarrow. One of these is more likely to have provided the original source of the Bodley text than Amiatinus itself, which left England in 716. There is good reason to believe, from the evidence of cartulary leaves associated with these surviving folios, that the pandect from which they came was at Worcester in the eleventh century. Further, it is probably to be identified with the impressive bible which, according to Worcester writers, was given to the church at Worcester by Offa towards the end of the eighth century.(56)

There is of course no way of proving that the third Ceolfrithian pandect, now totally lost, did not also leave its Northumbrian home for some other part of the insular area and become a source for further copying. It is also conceivable that the original, improved Ceolfrithian exemplar of Tobit (which presumably was used in the copying of each of the pandect versions, and by Bede for his commentary In Tobiam) was used at Monkwearmouth - Jarrow in the production of other individual copies of Tobit or of part-bibles containing it, from which other branches of transmission may have arisen. However, in the absence of alternative evidence, the church of late eighth-century Worcester may be suggested as the likely starting point for the individual history of our Bodley Tobit text. If the copy from which the text descends was indeed made from a Ceolfrithian pandect in Worcester the relations between that important centre and other churches in the south and west including Wales, may be a profitable area for future research.(57) (1) Historia Abbatum auctore Baeda, ch. 15, and Historia Abbatum auctore Anonymo, ch. 20, Venerabilis Baedae Opera Historica, ed. C. Plummer, 2 vols. (Oxford, 1896), I, 379 and 395, respectively. (2) On the Codex Amiatinus, see esp. E. A. Lowe, English Uncial (Oxford, 1960), pp. 8-15, 18, and Pl. VIII-IX; D. H. Wright, |Some Notes on English Uncial', Traditio 17 (1961), 441-56; R. L. S. Bruce-Mitford, The Art of the Codex Amiatinus, Jarrow Lecture 1967 (Jarrow, 1968), repr. Journal of the British Archaeological Association 3rd ser. 32 (1969), 1-25; S. Berger, Histoire de la Vulgate pendant les Premieres Siecles du Moyen Age (Paris, 1893; repr. New York, 1958 and Hildesheim, 1976), pp. 37-8; and B. Fischer, Lateinische Bibelhandschriften im fruhen Mittelalter, Vetus Latina: Aus der Geschichte der lateinischen Bibel 11 (Freiburg, 1985), pp- 9-34 and 67-9. (3) Historia Abbatum auctore Anonymo, in Plummer, Baedae Opera, I, 395. Lowe, English Uncial, p. 19, has suggested that some evolution in the script took place between the earlier pandects (for which we have some manuscript evidence: see following note) and Amiatinus. See also Wright, |English Uncial', 442-3. Bruce-Mitford's argument (Art of Amiatinus, p. 7) that all three pandects were complete by about 700 is not convincing, and a terminus ante quem of 716 remains necessary. But cf. also R. Gameson, |The Cost of the Codex Amiatinus', Notes and Queries 235 (1992), 2-9. (4) No. 293 in H. Gneuss, |A Preliminary List of Manuscripts written or owned in England up to 1100', Anglo-Saxon England 9 (1981), 1-60. The remains are divided between three manuscripts, Add. 37777 (one leaf, with part of 3 Kings), Add. 45025 (ten leaves and fragments of an eleventh, with more of 3 Kings and parts of 4 Kings), and Loan 81 (a single leaf, only discovered in 1982 and containing part of Sirach). On the Additional manuscripts, see N. R. Ker, Medieval Libraries of Great Britain: a List of Surviving Books, 2nd ed. (London, 1964), 104-5; and E. A. Lowe, Codices Latini Antiquiores, 12 vols. (Oxford, 1934-71), 11 (2nd ed., 1972), no. 177, and English Uncial, pp. 8-9, 19 and Pl. X. On Loan 81, see B. Bischoff and V. Brown, |Addenda to Codices Latini Antiquiores', Medieval Studies 47 (1985), 317-66, at 351-2; A. G. Watson, Medieval Libraries of Great Britain edited by N. R. Ker: Supplement to the Second Edition (London, 1987), p. 40; M. B. Parkes, The Scriptorium of Wearmouth-Jarrow, Jarrow Lecture 1982 (Jarrow, 1983), p. 3; repr. in his Scribes, Scripts and Readers (London, 1991), pp. 93-120; and R. Marsden, The Text of the Old Testament in Anglo-Saxon England, Cambridge Studies in Anglo-Saxon England (forthcoming). Circumstantial evidence links Loan 81 with Add. 45025, for both were used in the sixteenth century for binding documents from the estate of Sir Francis Willoughby of Wollaton, while all three manuscripts have decorated chapter numbering added by the same late thirteenth- or early fourteenth-century scribe. See I. Atkins and N. R. Ker, Catalogus Librorum Manuscriptorum Bibliothecae Wigorniensis made in 1622-23 by Patrick Young, Librarian to King James I (Cambridge, 1944), pp. 77-9, and Bischoff and Brown, loc. cit. (5) On the textual origins of Amiatinus, see Fischer, Bibelhandschriften, pp. 20-1, 29-34, and 67-9. The possibility that |Ceolfrith's' Tobit reached Northumbria already in its emended state (or partly so) should not, I believe, be ruled out. The evidence against it is largely negative, in that no trace of the emendations has been found outside the insular area. Old Latin influence is unlikely, for the known old versions of the emended passages differ in most cases radically both from Jerome's version and Ceolfrith's. On the complex textual history of the Latin and Greek versions of Tobit, see esp. D. C. Simpson, |The Chief Recensions of the Book of Tobit', JTS 14 (1913), 516-30; A. Schulte, Beitrage zur Erklarung und Textkritik des Buches Tobias, Biblische Studien 19 (Freiburg, 1914), pp. 4-33; R. H. Charles, The Apocrypha and Pseudepigrapha of the Old Testament in English, 2 vols. (Oxford, 1913), II: Pseudepigrapha, pp. 174-6, 178; and R. H. Pfeiffer, History of New Testament Times: with an Introduction to the Apocrypha (New York, 1949), p. 258-84. (6) Edited by D. Hurst, in CCSL 119B (Turnhout, 1983), pp. 1-19. In his praefatio (no page number), Hurst calculates that In Tobiam was composed between 720 and 730; see also M. L. W. Laistner, A Hand-List of Bede Manuscripts (Ithaca, 1943), p. 78. Fischer, Bibelhandschriften, p. 31, has suggested that Bede may himself have helped with the textual emendations made to Tobit. (7) Except in the Gospels, there is little evidence for the subsequent influence of the Amiatinan Vulgate text or its exemplars, either within the insular area or outside it; Fischer, Bibelhandschriften, pp. 69 and 130-1. (8) N. R. Ker, Medieval Libraries of Great Britain: a List of Surviving Books, 2nd ed. (London, 1964), p. 46 and Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957; reissued with suppl., 1990), pp. 376-7. For detailed descriptions, see R. W. Hunt, et al., A Summary Catalogue of Western Manuscripts in the Bodleian Library at Oxford, 7 vols. (Oxford, 1895-1953; rev. ed. Munich, 1980), II, 170-4; A. G. Watson, Catalogue of Dated and Datable Manuscripts c-435-1600 in Oxford Libraries, 2 vols. (Oxford, 1984), I, 19; and W. M. Lindsay, Early Welsh Script (Oxford, 1912), pp. 26-32. On the codex's non-biblical contents, see M. Lapidge, |Latin Learning in Dark Age Wales: Some Prolegomena', Proceedings of the Seventh International Congress of Celtic Studies, ed. D. E. Evans, J. G. Griffith, and E. M. Jope (Oxford, 1986), pp. 91-107. (9) For this interim assessment I am grateful to Dr David Dumville, who is preparing a paper on the palaeography of the manuscript. (10) In my analysis of the Bodely Tobit my Vulgate reference text is the reconstructed Hieronymian text (Hier.) of the Biblia Sacra iuxta Latinam Vulgatam Versionem ad Codicum Fidem, ed. H. Quentin et al., 18 vols. (Rome, 1926-), VIII: Libri Ezrae, Tobiae, Iudith (1950), pp. 169-209, hereafter abbreviated to BS VIII. My references to variations from this text in surviving Vulgate manuscripts rely entirely on the editors' extensive critical apparatus and I use their manuscript sigla; see n. 19, below. The many corrections and emendations in Bodley 572 warrant a separate study and they have not been noted here, except where they are of particular interest. (11) Lindsay, Early Welsh Script, p. 32, has drawn attention to some of the more frequent examples of idiosyncratic spelling. See also his extensive list of abbreviations used, pp. 29-32. (12) Cf. Amiatinus: chirografum (4: 21, 22 and 5: 3) and chirographum (9: 3, 6). (13) Although the spelling of agiographa (which is the Amiatinan form) varies among the collated manuscripts, none has apocrypha, or a variation on it. (14) Recorded Vulgate variations include Sennacherib, the form to which Bodley's version in 1: 21 has been turned by a later corrector. In a previous occurrence of the name in 1: 18, the first part, Senna, has been written correctly but cherib has been written above later, though not apparently by the same hand which made the correction in 1: 21. (15) A full list of the collated manuscripts of Tobit is given in BS VIII, pp. 165-6. (16) BS VIII, p. 196. (17) Periods are commonly marked by a comma with two points superimposed. Low, medial and high points are also used, but it is not always clear which position the scribe has intended. (18) None of the citations in Bede's In Tobiam shows variations which would indicate significant textual differences between Bede's exemplar (which very likely was the same as that used for the pandects) and Amiatinus. (19) The following Vulgate manuscript sigla (as used in BS VIII) are adopted in this article:

L Vatican City, Biblioteca Apostolica Vaticana, Pal. lat. 24 (ca 700) [L.sup.a] the restored portions of L (saec. [viii.sup.2]; see following note)
K   Cologne, Dombibliothek, 43 (saec. viii)
U   St Gall, Stiftsbibliothek, 6 (saec. viii)
W   Stuttgart, Wurttembergische Landesbibliothek, HB. II, 35 (saec. viii/ix)
E   Paris, Bibliotheque Nationale, lat. 11533 (830-850)
F   Munich, Bayerische Staatsbibliothek, Clm. 6225 (saec. ix; from Friesing)


[[gamma].sup.A] Milan, Biblioteca Ambrosiana, E. 53 inf. (saec. ix; Biasca Bible)
S   St Gall, Stiftsbibliothek, 14 (saec. ix)
C   La Cava dei Tirreni, Archivio della Badia, 1(14) (saec. ix; Codex Cavensis)


[[sigma].sup.T] Madrid, Biblioteca Nacional, Vitr. 13-1 (saec. x; Codex Toletanus) (20) The original portions of Pal. lat. 24, which account for a total of about a third of its Tobit text (and are designated L), were written in southern Italy around 700. Restoration was done at Lorsch during the second half of the eighth century ([L.sup.a]) and two further leaves were added later. Tob. 1: 1-2: 20 has not survived in either an original or a restored form. See Lowe, Codices, I, no. 68, BS VIII, p. xii, and Fischer, Bibelhandschriften, pp. 176 and 372-3. The manuscript is among our earliest witnesses of Tobit and was treated as an archetype by the editors of BS VIII. (21) In a few cases, other Vulgate manuscripts differ slightly from this version, but in none does the variation suggest a connection with the Amiatinan version. (22) All the Theodulfian bibles use the ablative, but in a different form of the phrase: quasi dimidia fere hora. (23) In insular manuscripts, confusion between u and flat-topped a can occur easily. (24) The creation of neuter plurals from first declension nouns is common in Late Latin, and the plural blasfemia itself was used by Prudentius in his widely-known Psychomachia (715). The Dictionary of Medieval Latin from British Sources, ed. R. E. Latham et al. (Oxford, 1975-), records one instance of blasphemium from the mid-eleventh century. (25) I have judged the plural noun to be an acceptable variation, rather than an error, in view of the colon's plural subject (hii ... qui coniugia ita suscipiunt). (26) Bodley shares Deum with K (as a correction), U, [[gamma].sup.a], the Alcuinian and some later manuscripts. (27) Compare, however, Deum for eum in 13: 4, listed below, where there is no record of the variant Dominum. Bodley's Deum in 14: 11 is paralleled in a number of manuscripts, including U, [[gamma].sup.a], and all the Alcuinian. (28) This assessment is based on the information supplied in the critical apparatus of BS VIII. (29) Hier. occurrit in obuiam filio suo. The Amiatinan version would have been acceptable, had the proper noun Thobias been in concord with the dative noun and pronoun, as indirect object of occurrit (cf. Am. dedit Thobiae in 8: 24). Bodley has here occurrit obiam, omitting in and misspelling obuiam. The latter error occurs also in L, where original uenit obiam is |corrected' to occurit in obiam. It could be argued that the variant Thobiam in Amiatinus is more likely to have arisen in the copying of an exemplar having in obiam than one having the correct in obuiam. (30) The preceding and following verbs in 13: 20-2 are all in the future tense. The error is also in [[gamma].sup.A]. (31) This error occurs in an addition to the Hieronymian version, found in Amiatinus, Bodley, and many later manuscripts. (32) This reading occurs also in a Paris bible of the early thirteenth century. There are isolated late parallels also for some of the other readings which follow, but these have not been noted. A few, such as recepit in 12: 3, were adopted in later editions, including the Clementine. (33) See also above, P. 12 and n. 27. Five more of the readings presently under consideration involve the naming of the deity. (34) The frequent replacement in Vulgate manuscripts of nomen by semen in Deut. 25: 7 may be compared. (35) The Sinaitic text is one of two main versions printed in Septuagixta: Id est Vetut Testmuentum Graece iuxta LXX Interpretes, ed. A. Rahlfs, 2 vols. (Stuttgart, 1935); sw I, 1036. For the Old Latin versions, see P. Sabatier, Bibliorum Sacrorum Latimae Vertiones Antiquae seu Vetus Italica, 3 vols. (Rheims, 1743-49), at I, 709-43. (36) Thirteen Bodley readings are shared by K and U, eleven by E and F, eleven by E and K, and eight by E and U. (37) In one case, in 10: 1, the Bodley version comes very near to the unique version in Amiatinus (see the illustrative passage, above). (38) A long omission such as that in 3: 15 may well have occurred at the stage of copying from a majuscule manuscript, where the words would probably have formed a complete line. Two other substantial omissions, each disrupting the narrative sense and without doubt accidental, were noted above in the discussion of copying errors. (39) Hier., Am. ille me nescit neque ego illum quod signum dabo illi. Cf. [[L.sup.A] illum + noui. Another tradition, which became the dominant one, has et ego illum [later eum] ignoro for neque ego illum; see BS VIII, p. 181. (40) There is a radical change in word order also, with Amiatinus and Bodley agreeing in this respect against all other manuscripts: Bodl. sed ego de cibo et potu utor spiritali [Am. inuisibili] qui ab hominibus uideri non potest; cf. Hier. sed ego cibo inuisibili et potu qui ab hominibus uideri non potest utor. (41) Bodl. occiderat omnes qui ingrediebantur; cf. Hier. occiderat eos mox ut ingressi fuissent. In its use of ingrediebantur [-batur, before correction], Bodley parallels Amiatinus, which otherwise follows Jerome; the reading occurs in no other manuscripts. (42) Again, Bodley has modified the Amiatinan version (shared with U and an eleventh-century Italian manuscript), viz. nos without the preposition. (43) The Mazarine Bible, [omega].sup.M] in BS VIII. (44) In 8: 15-16 Bodley follows in general the exclusive Amiatinan form of the verses, including et reuersa [nuntiauit] nuntium bonum. (45) Bodley has harundinum, corrected by a later hand, insederent for insiderent, and in for super (the latter occurring also in U). Amiatinus shares insederent but otherwise follows Jerome. (46) Bodley otherwise follows Jerome; so does Amiatinus, except for erroneous particula. (47) Am., Bodl. Thobias, et posuit, eam om. On Bodley's variant version of carbones uiuos, see below. (48) Although it seems unlikely that an Old Latin version influenced the emender of our Bodley text at this point (for presumably he would then also have adopted the |two organ' formula), attention is drawn below to further pecularities of the Bodley version of this passage which may indeed suggest such influence. (49) See BS VIII, p. 175. The only notable variation is uera for iusta in corrected L, corrected K, and F. (50) For the Old Latin versions, see Sabatier, Latinae Versiones Antiquae, 1, 715. The three manuscripts with magna are Paris, Bibliotheque Nationale, lat. 3564 (saec. ix) and lat. 11504, formerly |Sangermanensis 4' (copied in 822), and Vatican City, Biblioteca Apostolica Vaticana, Vat. Regin. lat. 7 (saec. ix med.). The Aramaic and Hebrew versions are printed, with English translations, in A. Neubauer, The Book of Tobit: a Chaldee Text from a Unique Manuscript in the Bodleian Library with Other Rabbinical Texts, English Translations and the Itala (Oxford, 1878); see PP. 5 and xxx, and 21 and xlix. (51) See Fischer, Bibelhandschriften, pp. 81-9 and 147-8. The manuscript is no. 7 in the Beuron edition of the Vetus Latina and has the siglum [gamma] in Charles, Pseudepigrapha. (52) Rahlfs, Septuaginta, 1, 1008. (53) Characteristic are Bibliotheque Nationale, lat. 3564 and Biblioteca Apostolica Vaticana, Vat. Regin. lat. 7, with et imposuit super carbones uiuos; Sabatier, Latinae Versiones Antiquae, 1, 730. (54) The phrase is identical in the Roman, Gallican, and |Hebrew' versions of the Latin psalter. It occurs also in 2 Sam. 22: 13, 4 Esd. 16: 54, and Rom. 12: 20. (55) On Ireland, see P. Doyle, |The Latin Bible in Ireland: its Origins and Growth', in Biblical Studies: the Medieval Irish Contribution, ed. M. McNamara, Proceedings of the Irish Biblical Association 1 (Dublin, 1976), PP. 30-45, esp. at 34-7, and M. McNamara, |The Text of the Latin Bible in the Early Irish Church; Some Data and Desiderata', in Irland und die Christenheit, ed. P. Ni Chathain and M. Richter (Stuttgart, 1987), PP. 7-55. On Assser, see W. H. Stevenson, Asser's Life of King Alfred (Oxford, 1904), pp. 61, 86 and xciv-xcv, and S. Keynes and M. Lapidge, Alfred the Great (Harmondsworth, 1983), p. 258, n. 159, and p. 273, n. 241. (56) See I. Atkins, |The Church of Worcester from the Eighth to the Twelfth Century: Part II. The Familia from the Middle of the Tenth to the Beginning of the Twelfth Century', Antiquaries Journal 20 (1940), 1-38 and 203-29, at 3 and 220-3, and P. Sims-Williams, Religion and Literature in Western England, 600-800, Cambridge Studies in Anglo-Saxon England 3 (Cambridge, 1990, pp. 182-3. (57) In preparing this article I have been greatly indebted to Dr David Dumville, who first drew my attention to Bodley 572 and who generously shared his knowledge of insular manuscript. Prof. Michael Lapidge offered valuable criticism of an earlier draft.
COPYRIGHT 1994 Oxford University Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1994 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Marsden, Richard C.J.
Publication:The Journal of Theological Studies
Date:Apr 1, 1994
Words:10710
Previous Article:Rediscovery of Creation: A Bibliographical Study of the Church's Response to the Environmental Crisis.
Next Article:Arius in modern research.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters