[In the following essay, Zier examines Lombard's use of the Glossa ordinaria in composing the Sentences.]

Every student of medieval theology knows the importance of Peter Lombard to the history of that field: the Libri quatuor Sententiarum served as an official textbook in theology for hundreds of years. What is less well known and less well documented is the Lombard's impact on the study of the Bible and on the development of the Glossa ordinaria.

One of the first of Beryl Smalley's publications¹ considered the possibilities of the Lombard's biblical commentaries, in particular on Isaiah; and toward the end of her career, she began to uncover his work on the Gospel of Luke.² One of the more obscure corners of the picture she began to paint was Peter's relationship to the Glossa ordinaria, which became the basis of the cursory reading of the Bible throughout the Middle Ages.

The story of Peter Lombard and the Glossa ordinaria starts with the necrology of the cathedral chapter of Notre Dame, where we read that Peter gave to the chapter “omnes libros eius glosatos, scilicet: Novum Testamentum, totum; in Veteri Testamento: Psalterium, quinque libros Moysi, quatuor maiores Prophetas, duodecim minores, Cantica, Iob, Hester, Thobiam, Iudith, librum Sapientie, Ecclesiasticum.”³ But from the necrology alone we know nothing of the nature of Peter's glossed books of the Bible. The connection between Peter and the Glossa ordinaria reached a high-water mark with the publication in 1933 of the History of the Vulgate in England,⁴ in which the author argued at some length, based on the many verbatim parallels between the Gloss and the Sentences, that the Lombard was in fact responsible for the entire Glossa ordinaria. More recently, Ignatius Brady found himself of two minds. At the time of the publication of his edition of the first two books of the Sentences in 1971, he rejected the suggestion that the Lombard had composed any biblical commentaries other than on the Psalms and Paul; but upon the publication of the last two books in 1981, he came to recognize that Peter had commented on, or at least glossed, several other books in the canon.⁵

Wherein lies the truth? The evidence is admittedly thin. The books that the Lombard bequeathed to the chapter have disappeared over time.⁶ Although there may be a few manuscripts containing a biblical gloss attributed to Peter Lombard, none of the ascriptions is authentic. What we do have are several comments by Peter Comestor and Stephen Langton: Comestor cites the Lombard's teaching in some detail on the Gospel of Luke, and Langton refers in his biblical commentaries to readings of the Lombard.⁷ Perhaps more important for understanding Peter's relation to the Gloss is the use to which Peter put it in the Sentences. Altogether, the pattern that emerges from this evidence suggests that it was Peter's use of his own glossed books of the Bible, probably a more or less final form of the Gloss as we know it, coupled with the prestige of his Sentences—of which his anonymous quotations of the Gloss were an integral part—that helped to normalize the text that became the Glossa ordinaria.

In this investigation I have focused on two areas: the Lombard's readings cited by Stephen Langton in his commentaries, and the Lombard's use of the Gloss on Genesis in composing his teaching on Creation in book 2 of the Sentences.

STEPHEN LANGTON AND PETER LOMBARD

In his prolegomenon to the edition of books 3 and 4 of the Sentences, Brady considers a handful of passages in Stephen Langton's commentaries on Isaiah, noted by Smalley in the 1930s, where Langton cites a reading of the Lombard on Isaiah,⁸ none of which is to be found in the Sentences. In the way he refers to the Lombard's text Langton gives us the kind of precision that makes a comparison with other known manuscripts of the Gloss quite instructive. For this collation I have considered six manuscripts from Paris, Bibliothèque Nationale; four from Cambridge, Pembroke College (from the library of Bury St. Edmunds); and the recently reprinted Rusch edition.⁹

The first and perhaps most interesting example comes from Langton's commentary on Isa. 7. Although the Gloss gives a nod to the historical meaning of these verses, we find primarily a Christological reading of the text, which, in the present instance, raises the issue of Christ's human psychology. At Isa. 7.15, “Butyrum et mel comedet, ut sciat reprobare malum et eligere bonum,” Langton comments on the interlinear glossula, “nunquam vitiorum patet affectibus”: “Notat quod ista interlinearis non est de originali; et ubi nos habemus in interlineari affectibus Lombardus habet aspectibus. Ex quo verisimile est quod littera sit corrupta, et vera littera sit ista: que nunquam virorum patet aspectibus.”¹⁰

Of the issues regarding the humanity of Christ, his human will received considerable attention throughout much of the medieval period. And the tendency of several theologians in the twelfth century was to “dehumanize” it, that is, to consider it in such a way as to attribute capacities to it that simply did not apply to the will of other human beings. In this they followed the teachings of Augustine, some going so far as to assert that the non posse peccare normally reserved for the deity applied to Christ's human will as well.¹¹

With the reading of the Gloss—that Christ was never afflicted with the desires of temptation—it is possible to allow that Christ may have contemplated temptation. This would be in keeping with the Lombard's teaching on this point: that while Christ may have contemplated temptation (propassio), he was not moved by it (passio).¹² But the Lombard reads “aspectibus” where the Gloss reads “affectibus.” The connotation of the Lombard's reading, “patet aspectibus,” could imply that Christ was never exposed (emphasis on “patet”) to vice or that he never tolerated the presence (emphasis on “aspectibus”) of vice. It may be that this alternative was intended simply to reinforce the justice of Christ's judgments. The precise connotation seems somewhat vague; but in any case, it could well be in keeping with the direction of the Lombard's Christological teaching in general: to soften, though not to change drastically, the patristic position that while Christ possessed a human will, it was nevertheless untouched by the weaknesses of the postlapsarian human condition. Alternatively, Langton's version, “uirorum aspectibus,” seems to have the virtue of emphasizing the constancy of Christ's judgment without necessarily robbing Christ of the essential human experience of temptation.

What can the witnesses of the Gloss tell us? The manuscripts are all roughly contemporary, that is, second half of the twelfth century. C appears to be the earliest, or at least exhibits the earliest type of format: a single column of biblical text written first on each page, then the glossulae added at a second stage, with the result of a relative lack of economy of space. C, together with G, omits this interlinear glossula altogether. A and B, arguably among the earliest witnesses,¹³ read with the Lombard as reported by Langton (“nunquam uitiorum patet aspectibus”). Four of the manuscripts, E, F, H, and I, together with R, read with the Gloss as reported by Langton (“nunquam uitiorum patet affectibus”).¹⁴ Two witnesses, D and J, give a third reading: “nunquam uitiorum paret affectibus,” another attempt, perhaps even more successful than the Lombard's own efforts, to clarify the text in a Lombardian direction.¹⁵ Taken as a whole, the witnesses suggest that the volatility of this particular glossula reflects the lack of consensus on the human psychology of Christ that characterized the theology of the twelfth century.¹⁶

In a second example, at Isa. 8.1, “Et dixit Dominus ad me: Sume tibi librum grandem, et scribe in eo stylo hominis,” Langton comments that this refers, among others things, to a virgin who would bear a son:¹⁷ “Et virgo promittitur paritura filium. Talem enim litteram ponit Ieronymus, et est sensus: virgo promittitur paritura filium. Lombardus sic habet: Virgo paritura promittitur et filius nasciturus.”¹⁸

The text in all ten manuscripts is virtually identical with the text of the Gloss reported by Langton. But the Rusch edition adds the Lombard's reading: “et filius nasciturus.” Peter's reading, if it is indeed his and not just a textual variant, emphasizes the son to be born, rather than the virgin who will give birth, once again focusing on a Christological emphasis for the text.

The other Lombardian variants cited by Langton similarly reinforce the first meaning of the text as Christological. The manuscripts and Rusch agree with the Gloss readings cited by Langton, against the Lombard, with only two minor exceptions.¹⁹

From these examples Langton seems clearly to use the Lombard as something of a corrector to the readings of the Gloss. To this it should be added that the readings he cited were almost certainly from the glossed books that the Lombard had bequeathed to the cathedral, to which Langton had access at the end of the twelfth century. But as to the text of the Lombard's books, on the whole there can be almost no doubt that it is little more than a variation on the manuscript tradition of the Glossa ordinaria that we can identify today. In these two examples the differences between what may be unique readings, that is to say, additions to the manuscript readings, and simple variants on the manuscript readings amount to one rather volatile glossula and one additional phrase: “et filius nasciturus.” While this may seem somewhat insignificant, it is precisely this kind of addition that is representative of the final stages in the “redaction” of the Gloss.²⁰

PETER LOMBARD ON CREATION

Although Peter may not have left his own written commentaries on Scripture (other than on the Psalms and Paul), there is the evidence, especially from Peter Comestor, that the Lombard did expound orally at least the Gospel of Luke, and probably other books of the Bible as well. As we have seen, it is not at all unlikely that the Lombard's commentaries would be little more than a few additional comments or annotations added to the Gloss as he found it in his own library. This, in fact, is largely what we find when we examine his use of the Gloss in the Sentences. And although each of the four books of the Sentences is liberally peppered with quotations from the Gloss, none exhibits as many extensive citations as found in book 2, which includes the Lombard's discussion of Creation.

Marcia Colish has pointed out that Peter's treatment of Creation is unusual with respect to his methodology; that is, in this discussion Peter relies much more heavily and extensively on all of his sources to make his point, weaving together the usual patristic authors, together with references to the Gloss, Hugh of St. Victor, and the Summa Sententiarum.²¹ This characteristic, however, only serves to clarify the mechanics of his use of the Gloss and provides an excellent opportunity to see Peter using the Gloss, which, after his own glosses on the Psalms and Paul, was his favorite source.²² For the collation here I have used the Rusch edition (R) together with ten manuscripts, eight from the Bibliothèque Nationale, and two from Pembroke College, Cambridge.²³ What emerges is not entirely unexpected: the manuscripts (twelfth-thirteenth century) tend to present one version of the text, and R (fifteenth century) a slightly different one. Where Peter's citations vary from the established text of the Gloss, it is usually the result of an amplification or of a summation. The first example is a classic case of the former. Peter amplifies Alcuin's opinion as cited in the Gloss regarding Creation (here and below Peter's amplifications are indicated in boldface, and his unique alternative readings are in boldface italics):²⁴

Quatuor modis < … > operatur deus < … > primo in Verbo omnia disponendo, secundo in materia informi quatuor elementorum, de nihilo eam creando, unde: Qui vivit in aeternum creavit omnia simul, omnia scilicet elementa vel omnia corpora materialiter simul creavit; tertio per opera sex dierum varias distinxit creaturas; quarto ex primordialibus seminibus non incognitae oriuntur naturae, sed notae saepius reformantur ne pereant.²⁵

Here Peter uses Alcuin's opinion to sum up the debate over the nature of Creation as simul or per intervalla temporum. If Peter's additions are surpressed, the text of the Gloss alone leaves the reader with the question that initiated the discussion in the first place: how is it that God created everything simul, yet distinguished the variety of creatures by means of the works of six days? By way of a summary Peter will amplify the text on the basis of the discussion he has reported in the previous five chapters: it was the unformed matter of the four elements that was created simul ex nihilo; but the forms of creatures were only added to those elements over time. The three phrases found only in the Lombard's version of the citation might be considered in a small way his “gloss” on the Gloss.

A second example exhibits a slightly more subtle relationship between the text of the Gloss and Peter's use of it, and addresses the question of the location of the light created on the first day (again, the boldface text represents Peter's additions; italicized text represents alternative readings shared only with Rusch; boldface italics have been used to indicate unique alternative readings):²⁶

Si autem quaeritur ubi est facta lux illa, cum abyssus omnem terrae altitudinem tegeret, dici potest in illis partibus facta, quas nunc illustrat solis diurna lux. Nec mirum lucem in aquis posse lucere, cum etiam nautarum operatione saepius illustrentur; qui in profundum mersi, misso ex ore oleo aquas sibi illustrant; quae multo rariores fuerunt in principio quam modo sint, quia nondum congregatae fuerant in uno loco.²⁷

According to Augustine, that first light appeared in those parts that the daily light of the sun now illuminates, even in the waters, which at that time would have been much more attenuated since they had not yet been gathered into one place. Peter has added a couple of phrases by way of clarification (“lux illa,” “facta,” “fuerant”), expanding the sense of what is a rather tersely worded passage. Similarly his unique variants show an editorial hand at work: the first, “dici potest,” softens the rather bold phrase in the Gloss, “patet quia”; and the transposition near the end of the passage, “fuerunt in principio quam modo sint,” seems to make smoother sense of the text than the Gloss's “quam sint modo fuere in principio.”²⁸ It is perhaps worth noting where Peter's reading agrees only with Rusch against all the other manuscripts. In the present case this amounts to the transposition of “facta est” and the use of “lucere” in place of “splendere.” Although minor variants, they suggest the nature of the relationship between Peter's text of the Gloss and the received text of the fifteenth century.

A third example from Peter's discussion of Creation reveals a common text of the Gloss that the Lombard appears then to have edited to his purposes. The question under discussion is this: where would the waters have been gathered that covered the earth?²⁹

Si autem quaeratur ubi congregatae sunt aquae, quae totum texerant spatium usque ad caelum, potuit fieri ut terra subsidens concavas partes praeberet, ubi fluctuantes aquas reciperet. Potest etiam credi primarias aquas rariores fuisse, quae sicut nebula tegerent terras, sed congregatione esse spissatas.³⁰

Peter has emphasized the totality of the phemonena described in this glossula by using the phrase “totum spatium” in place of “omnes partes terrae” and transposing “texerant” (all other witnesses read: “omnes partes terrae usque ad caelum texerant”), and by reading “ubi” where all the other witnesses read “quibus.” It is almost as if he were striving for a more global mode of description. That interest is continued in the citation that follows immediately in the text:³¹

SENTENTIAE

Cumque multa constet esse maria et flumina, in unum tamen locum dicit aquas congregatas, propter continuationem omnium aquarum quae in terris sunt, quia cuncta flumina et maria magno mari iunguntur.

Ideoque, cum dixerit aquas congregatas in unum locum, deinde dicit pluraliter: congregationesque aquarum, propter multifidos sinus earum, quibus omnibus ex magno mari principium est.

GLOSSA ORDINARIA (RUSCH)

Cum multa constet esse maria, in unum tamen locum dicit aquas congregatas,

quia cuncta magno mari iunguntur.

Si qui lacus in semetipsis stricti videntur occultis tamen meatibus in mare revolvuntur, fossores quoque puteorum hoc probant quia omnis tellus per invisibiles venas aquis repleta est < dixit eas congregatas in unum locum, nunc uocat pluraliter congregationes aquarum propter multifidos sinus earum>

quibus ex mari principium est.³²

Here we can see fairly clearly an editorial hand at work: with his modifications and rearrangement, it seems as though Peter wants to be more precise in his designation of the “waters” and to emphasize the single source of them all: the one place into which they are said to be gathered includes not just the oceans and seas, but the rivers as well, all of which are ultimately connected, and hence legitimately described as “one place.” The specific examples given in the Gloss to support the argument are passed over in silence, and the phrase “ideoque … sinus earum” has been inserted from an adjacent glossula³³ and reinforces Peter's gobal focus. This is perhaps the place to note that the Lombard seems to depend on the Gloss for this kind of “historical” information—all the more striking as the Glossa ordinaria on Genesis is predominantly spiritual, with allegories and moralities filling the page. Yet Peter is clearly uninterested in the spiritual sense, at least for the purpose of his Sentences, reflecting perhaps the understanding that dogmatic theology can be founded only upon the literal/historical sense of Scripture, an understanding codified, for example, by Thomas Aquinas in his Summa theologiae.³⁴

In any case, Peter's approach to the Gloss reveals a conservative, respectful attitude that preserves for the most part the integrity of the text. To be sure, he sometimes reworks a sentence or two in keeping with the focus or emphasis of his overall argument; and occasionally he may reshuffle the order of the text as he finds it. Still, it is important to recognize that Peter has chosen not simply to summarize the text of the individual glossulae that he has selected, but to excerpt them faithfully, even if anonymously. And although it would be overstating the case to read the Sentences as a gloss on the Gloss, his liberal use of the Gloss did help to color the focus of his thinking in the Sentences. What appears, then, is an intermeshing of the Gloss with the Sentences in such a way that the former serves as the natural complement to the latter. Indeed, Peter incorporated in the Sentences citations from the Gloss on virtually all the major protocanonical books of the Bible:³⁵ of the approximately 175 references to the Gloss indicated in Brady's apparatus fontium, only 3 come from the Gloss on books of the Bible that do not appear among the books bequeathed by Peter (Joshua, 4 Kings, Proverbs).³⁶

These patterns reinforce the textual patterns noted above, suggesting a close relationship of the Gloss to Peter's Sentences. It seems likely that it was this interconnectedness that led by the end of the twelfth century to the standardization of the text of the Gloss more or less according to the version of it that was known to Peter Lombard. Peter himself may even have had a small role in fixing the text of the Gloss: perhaps only a phrase here and there—but, then, that is the difference between early and late versions of the text. And although there seems no longer to be a “smoking gun” to prove this point, the more we learn about the manuscript tradition of the Glossa ordinaria, the more we will understand, no doubt, how closely linked were all the endeavors of Peter Lombard to the study of the Bible in the Middle Ages.

Notes

1. Coauthored with G. Lacombe, “The Lombard's Commentary on Isaias and Other Fragments,” New Scholasticism 5 (1931), 123-62.

2. “Peter Comestor on the Gospels and His Sources,” Recherches de théologie ancienne et médiévale 46 (1979), 84-129.

3. Ignatius Brady, ed., Magistri Petri Lombari Parisiensis Episcopi Sententiae in IV libris distinctae, 2 vols. (Grottaferrata, 1971, 1981), 2:21*.

4. H. Glunz, History of the Vulgate in England (Cambridge, 1933), esp. pp. 213-45.

5. Brady, Sententiae, 1:115*-117*, 2:20*-52*.

6. Ibid., 2:20*-23*.

7. Ibid., 2:45*-47*, taken from Smalley and Lacombe, “The Lombard's Commentary on Isaias,” pp. 132-34; Beryl Smalley, “La Glossa ordinaria: Quelques prédécesseurs d'Anselme de Laon,” Recherches de théologie ancienne et médiévale 9 (1937), 365-400. See also Smalley, “Peter Comestor on the Gospels.”

8. Brady, Sententiae, 2:45*-46*.

9. BN MSS lat. 144 (A), 145 (B), 146 (C), 147 (D), 148 (E), and 11954 (F); Pembroke College MSS 58 (G), 59 (H), 60 (I), and 146 (J); Biblia Latina cum Glossa ordinaria (R), ed. Adolph Rusch (Strasbourg, 1480/81), facsimile reprint prepared and introduced by Karlfried Froehlich and Margaret Gibson.

10. Brady, Sententiae, 2:45*. The quotation comes from Langton's In Isaiam, BN lat. 14417, fol. 181r. Brady cites it from Smalley, “La Glossa ordinaria,” p. 400. The passage can be found at A, fol. 17v; B, fol. 24v; C, fol. 12v; D, fol. 23r; E, fol. 22v; F, fol. 23r; G, fol. 16v; H, fol. 16r; I, fol. 11r; J, fol. 17r-v; and R, 3:16a. (“This interlinear gloss is not from Jerome's commentary; and where we have between the lines ‘desires,’ the Lombard has ‘appearances.’ From this it seems likely that the text is corrupted and the true meaning is this: that < his judgment > is never accessible to the sight of men.”)

11. Marcia Colish, Peter Lombard, Brill's Studies in Intellectual History 41/1-2 (Leiden, 1994), pp. 443-48.

12. Ibid., p. 444. Peter's teaching can be found at Sententiae 3.16.1-2.

13. B exhibits an excellent system of reference marks linking glossulae to biblical text, and A exhibits a rather less thorough reference system and a series of flourished, but not truly illuminated, capitals. …

14. Among these are the most datable of the manuscripts: the French manuscript F was given to St.-Germain-des-Prés by Peter of Poitiers (d. 1202), the second person to hold the chair in theology at the cathedral school after Peter Lombard; and the English manuscript I contains flyleaves taken from a Hebrew liturgical codex. The Jews were expelled from Bury in 1190, suggesting a date for the manuscript of about that time.

15. Interestingly, J seems to have been corrected from the Lombard's reading: “patet aspectibus.”

16. Before leaving this glossula, it is worth noting that the alternative proposed by Langton did not require a great deal of imagination: indeed, the phrase “que numquam uirorum patet aspectibus” can be found at the end of a marginal glossula that appears in every one of the witnesses consulted, often immediately adjacent to the biblical display text, “Butyrum et mel. …” The focus of this marginal glossula is the virgin (“alma”) who is conceiving, and the connotation of “alma,” and it concludes with the etymology “alma apud hebreos adolescentula dicitur, uel abscondita, idest apocrifa, que nunquam uirorum patet aspectibus.” Could it be that the Lombard (or whoever introduced this interlinear glossula) meant to link mother and son in this way?

17. Brady, Sententiae, 2:46*. Although Langton cites Jerome here, the words follow more closely Jerome as reported in the Gloss (where Langton cites authors directly, he usually uses some such formula as “in originali”). This passage comes from Langton's In Isaiam, BN lat. 14417, fol. 182v, via Smalley and Lacombe, “The Lombard's Commentary on Isaias,” pp. 132-34.

18. The text can be found at A, fol. 18v; B, fol. 25v; C, fol. 13v; D, fol. 24r; E, fol. 23v; F, fol. 24r; G, fol. 18r; H, fol. 16v; I, fol. 11v; J, fol. 18r; and R, 3:17a. (“And it is promised that a virgin will bear a son. For such is the text that Jerome gives, and this is the sense: it is promised that a virgin will bear a son. But the Lombard has: it is promised that a virgin will bear, and a son will be born.”) The collation yields:

    uirgo paritura promittitur filium] ABCFGHIJ, uirgo paritura filium promittitur transp. DE, uirgo paritura filium promittitur et filius nasciturus R.

19. H, fol. 17v, omits a portion of a glossula on Isa. 8.9, even as Langton notes that the Lombard does. A later hand has added it in the margin beyond the marginal gloss. Indeed, this manuscript has been heavily corrected, and secondary glosses have been added throughout. In the same comment Langton notes that the Lombard deletes “vincimini” from the biblical text, as does E at fol. 24r: Brady, Sententiae, 2:46*. Again, these variants in the Lombard's book seem to reflect nothing more than textual variants attested in the manuscripts themselves. It remains to be seen whether any, or rather, how many, manuscripts of the Gloss might be found to corroborate the Lombard's readings: it seems likely that one might eventually discover a manuscript containing all the Lombard's variants.

20. See Mark Zier, “The Manuscript Tradition of the Glossa ordinaria for Daniel and Hints at a Method for a Critical Edition,” Scriptorium 47 (1993), esp. pp. 13-15.

21. Colish, Peter Lombard, pp. 336-42.

22. Brady, Sententiae, 1:119*.

23. BN MSS lat. 63 (K), 65 (L), 66 (M), 67 (N), 367 (O), 368 (P), 369 (Q), and 370 (S); and Pembroke College MSS 47 (T) and 142 (U).

24. Sententiae 2.12.6, ed. Brady, 1:388-89. The passage can be found at K, fol. 2r; L, fol. 2v; P, fol. 2v; Q, fol. 2v; S, fol. 2v; T, fol. 2r; U, fol. 1r; and R, 1:6b. This glossula, found as a protheme, is lacking in N; M and O have no prothemes at all.

25. As noted, the use of boldface distinguishes Peter's text from the common text of the Gloss. The other variants are all relatively insignificant:

    informi] om. L et R primordialibus seminibus] primordiis seminalibus ante corr. L notae] nocte ante corr. P,QS reformantur ne pereant] ne pereant reformantur transp. omnes codices Glossae, etiam codices manuscripti Sententiarum LMNX (uide Brady, Sententiae, 1:389).

    The last variant has little textual significance, but the Lombard's transposition does make the sense of the passage smoother.

26. Sententiae 2.13.3, ed. Brady, 1:390. The text can be found at K, fol. 3r; L, fol. 5r; M, fol. 1v; N, fol. 1v; O, fol. 2r; P, fol. 4r; S, fol. 4r; T, fol. 4r; U, fols. 2v-3r; and R, 1:10a. Q seems to be missing the folio that would have contained this glossula.

27. The collation demonstrates, incidentally, the stability of the text of this glossula. Apart from the Lombard's editorial changes, it amounts to little more than a handful of unique (and largely inconsequential) variants:

    terrae altitudinem] transp. L tegeret] regeret S diurna] diuturna M nautarum] naute U, uel -tarum supra scriptum U profundum] aque add. L illustrant] illustrent L quae] tunc add. R loco] sunt add. PS, add. et del. U esse spissatas] transp. L.

28. There is a third variant at “profundum,” where all the other witnesses read “profundo,” except O, which reads “profundum” with the Lombard. This is likely little more than a textual variant and tells us very little about the Lombard's editorial decisions.

29. Sententiae 2.14.8, ed. Brady, 1:398. The text can be found at K, fol. 4v; L, fol. 7r; M, fol. 2v; N, fol. 2r; O, fol. 3v; P, fol. 5v; Q, fol. 4r-v; S, fol. 5v; T, fol. 5v; U, fol. 4r; and R, 1:11b. Again, the boldface text represents Peter's additions; boldface italics indicate unique alternative readings.

30. Peter's unique alternative readings are discussed below. The balance of the collation yields only two variants:

    sunt] sint PQSU fluctuantes] fluentes KMNOT, fluitantes R, fluantes corr. ex fluctuantes U.

    Again, the text of the Gloss appears to be relatively stable.

31. The text can be found at K, fol. 4v; L, fol. 7r; M, fol. 2v; N, fol. 2r; O, fol. 3v; P, fol. 5v; Q, fol. 4v; S, fol. 5v; T, fol. 6r; U, fol. 4r; and R, 1:11b. Peter's additions are indicated in boldface. Unique alternative readings are given in boldface italics.

32. The collation of the manuscripts with Rusch (the righthand column) yields the following variants:

    constet] constat L tamen locum] om. N, transp. PQS, tantum locum T aquas] esse add. L aquas congregatas] transp. N si] sed L qui] quis Q, autem add. KMNO semetipsis] seipsos K, semetipsos MNT, semetipsis corr. ex semetipsos O, semetipso T reuoluuntur] reuoluitur K hoc] hec P quia] quod KMNOT tellus] telles K eas] esse add. LPQSU unum locum] transp. LMOPQSU nunc] tamen nunc PS, tunc non Q multifidos] multimodos L est] om. U.

    Seven of the sixteen variants derive from the portion of the text not found in the Sentences (“Si … tellus”).

33. In P and U these two glossulae can be found side by side. In M they “flank” the display text, with one of the glossulae “snaking” between the lines and nearly “touching” the other; …

34. Thomas Aquinas, Summa theologiae I,10, ad primum. Here Thomas is responding to the opinion that a multiplicity of scriptural senses leads to confusion and weakens the basis for argument. Thomas, addressing the relationship of literal and spiritual meanings, replies: “Et ita etiam nulla confusio sequitur in sacra Scriptura, cum omnes sensus fundentur super unum, scilicet litteralem. Ex quo solo potest trahi argumentum, non autem ex his quae secundum allegoriam dicuntur. …”

35. For the handful of books in the canon for which he did not possess a copy of the Gloss (Joshua, Judges, Ruth, 1-4 Kings, 1-2 Chronicles, Ezra-Nehemiah, 1-2 Maccabees, Proverbs, and Ecclesisastes), it is likely that another of the Paris masters chose his own copy, and that this process filled in the gaps in the Lombard's collection. Peter Comestor and Peter of Poitiers are likely candidates here. Peter of Poitiers's copy of the Gloss on Joshua (BN lat. 11945) is virtually identical, for example, with several other copies of that work in the Bibliothèque Nationale (MSS lat. 76, 77, 394, 395, 396).

36. On the other hand, Peter possessed the glossed text of several (albeit minor) biblical books that he did not cite in the Sentences: there seems to be no reference to the Book of the Twelve, except for Amos, Hosea, and Joel; of the Pauline canon he does not refer to Colossians, 1 and 2 Thessalonians, nor to the Pastoral Epistles; of the Catholic Epistles, he cites only 1 and 2 Peter and James. It is perhaps worth noting that he does not use the Gloss on any of the apocrypha.

An earlier version of this paper was presented in May 1994 at the Twenty-Ninth International Congress on Medieval Studies at Western Michigan University. I wish to thank the Social Sciences and Humanities Research Council of Canada, without whose material support this research would not have been possible. I also wish to acknowledge the support and encouragement of my mentors, especially Leonard Boyle, whose skill guided my initial enquiries and fixed the principles of my research, and Edouard Jeauneau, whose patience has trained me over many years in the painstaking work of manuscript studies and in the art of teasing from the text the mind of the author. I am also grateful for the collaboration of other colleagues, in particular Karlfried Froehlich and Ann Matter, whose interests and insights have enriched my investigations into medieval biblical studies. I am also indebted to the recent work of Margaret Gibson on the gloss: “The Twelfth Century Glossed Bible,” Studia patristica 23 (1989), 232-44; and “The Place of the Glossa ordinaria in Medieval Exegesis,” in Ad litteram: Authoritative Texts and Their Medieval Readers, ed. Mark D. Jordan and Kent Emery, Jr. (Notre Dame, Ind., 1992), pp. 5-27. Finally, I wish to thank Marcia Colish and Nancy Spatz for their comments on the present work.