University of Virginia Library

The Growth of Robert Thornton's Books
by
Ralph Hanna III [*]

Robert Thornton was a gentryman who lived at East Newton, about fifteen miles due north of York City. Over an indeterminate but obviously protracted period, probably in the second quarter of the fifteenth century, Thornton spent a substantial amount of his leisure as a collector and copyist of Middle English literary works. Thornton's tastes were extremely catholic, ranging from romance narrative to religious prose and medical recipes; his copying is particularly important in providing us with texts known from no other sources, most notably the alliterative poem Morte Arthure.[1]

Thornton's literary remains, as we know them, survive in two large miscellanies, Lincoln Cathedral MS. 91 and British Library MS. Additional 31042. The separateness of these two codices is, however, deceptive. Both were prepared in the same format, on quires of rather uneven size formed of broad-sheets folded in folio; moreover, five paper-stocks appear in both volumes, an indication that the copying was, at least in part, contemporaneous. Although Thornton clearly composed his volumes out of booklets or fascicles,[2] each established on a broadly generic basis, no one has yet made any effort to speculate exactly what procedures Thornton followed in creating these codices.

In this essay I hope to offer such an hypothesis about how Thornton worked. I want to begin by offering a concise statement of the production data known about the two codices—the limits of the quires and booklets, the nature of the paper in each. From this data and some initial linguistic information about two of Thornton's exemplars, I believe it is possible to construct an account, albeit a messy one, of how Thornton proceeded in producing his various booklets.

LINCOLN CATHEDRAL 91 [3]

• Booklet I (the prose Alexander)
• quire 124 (-1, -2, -3, -4, -23) ff. 1-19

  ---

  52

  
  
  
  six full sheets of stock K (1+16, 2+15, 3+14, 4+13, 5+12, 6+11) two full sheets of stock E (7+10, 8+9) three unwatermarked half-sheets, ff. 17, 18, 19
• quire 224 (-1) ff. 20-42 four full sheets of stock K (20+41, 28+33, 29+32, 30+31) seven full sheets of stock E (21+40, 22+39, 23+38, 24+37, 25+36, 26+35, 27+34) one half-sheet of stock E, f. 42
• quire 318 (-11, -12, -13, -14, -15, -16, -17, -18, all cancels) ff. 43-52 four half-sheets of stock K, ff. 43, 45, 46, 49 four unwatermarked half-sheets, ff. 44, 47, 48, 50 one unwatermarked full sheet (51+52)
• Booklet II (romances)
• quire 416 ff. 53-68 eight full sheets of stock B (53+68, 54+67, 55+66, 56+65, 57+64, 58+63, 59+62, 60+61)
• quire 518 ff. 69-86 six full sheets of stock B (69+86, 70+85, 71+84, 72+83, 73+82, 74+81) three full sheets of stock L (75+80, 76+79, 77+78)
• quire 616 ff. 87-102 eight full sheets of stock L (87+102, 88+101, 89+100, 90+99, 91+98, 92+97, 93+96, 94+95)
• quire 722 (-1, -22) ff. 103-122 six full sheets of stock G (103+122, 104+121, 105+120, 106+119, 107+118, 109+116) two full sheets of stock C (111+114, 112+113) one full sheet of stock M (110+115) one unwatermarked full sheet (108+117)
• quire 822 (-12) ff. 123-143 ten full sheets of stock L (123+143, 124+142, 125+141, 126+140, 127+139, 128+138, 129+137, 130+136, 131+135, 132+134) one half-sheet of stock L, f. 133
• quire 922 (-5, -18) ff. 144-163[4]

  ---

  53

  
  
  
  ten full sheets of stock B (144+163, 145+162, 146+161, 147+160, 148+159, 149+158, 150+157, 151+156, 152+155, 153+154)
• quire 1016 (-16) ff. 164-178 two full sheets of stock K (165+178, 166+177) two full sheets of stock E (167+176, 168+175) three full sheets of stock A (169+174, 170+173, 171+172) one half-sheet of stock K, f. 164
• Booklet III (religious materials, mostly in prose)[5]
• quire 1120 ff. 179-198 ten full sheets of stock L (179+198, 180+197, 181+196, 182+195, 183+194, 184+193, 185+192, 186+191, 187+190, 188+189)
• quire 1224 ff. 199-222 twelve full sheets of stock L (199+222, 200+221, 201+220, 202+219, 203+218, 204+217, 205+216, 206+215, 207+214, 208+213, 209+212, 210+211)
• quire 1320 (-1, -2, -17, -18, -19, -20) ff. 223-236 six full sheets of stock O (225+236, 226+235, 227+234, 228+233, 229+232, 230+231) two half-sheets of stock O, ff. 223, 224
• quire 1418 (-1) ff. 237-253 eight full sheets of stock G (237+252, 238+251, 239+250, 240+249, 241+248, 242+247, 243+246, 244+245) one unwatermarked half-sheet, f. 253
• quire 1530 (-1; -10, -11, -12, cancels) ff. 254-279 nine full sheets of stock C (254+278, 255+277, 256+276, 257+275, 258+274, 259+273, 260+272, 261+271, 262+267) two full sheets of stock L (263+266, 264+265) two half-sheets of stock C, ff. 268, 279 two half-sheets of stock N, ff. 269, 270
• Booklet IV (medical recipes)
• quire 1638 (-17, -22, -38) ff. 280-314[6] fifteen full sheets of stock P (281+314, 282+313, 283+312, 284+311, 285+310, 286+309, 287+308, 288+307, 289+306, 290+305, 291+304, 292+303, 293+302, 294+301, 297+298)

  ---

  54

  
  
  
  one unwatermarked full sheet (296+299) one half-sheet of stock P, f. 300 two unwatermarked half-sheets, ff. 280, 295
• quire 17? (seven fragments) ff. 315-321

BRITISH LIBRARY ADDITIONAL 31042 [7]

• Booklet I (Cursor Mundi excerpts)
• quire 1? (a six leaf fragment) ff. 3-8 two leaves of stock A, ff. 3, 8
• quire 224 ff. 9-32 twelve full sheets of stock A (9+32, 10+31, 11+30, 12+29, 13+28, 14+27, 15+26, 16+25, 17+24, 18+23, 19+22, 20+21)
• Booklet IIa (The Northern Passion, romances)
• quire 322 (-22) ff. 33-53 ten full sheets of stock B (34+53, 35+52, 36+51, 37+50, 38+49, 39+48, 40+47, 41+46, 42+45, 43+44) one unwatermarked half-sheet, f. 33
• quire 420 ff. 54-73 six full sheets of stock C (55+72, 56+71, 57+70, 58+69, 59+68, 60+67) three full sheets of stock D (61+66, 62+65, 63+64) one full sheet of an unidentified stock (54+73)
• quire 526 (-5, -8, -26) ff. 74-96 four full sheets of stock E (76+95, 77+94, 78+92, 79+91) five full sheets of stock F (80+89, 81+88, 82+87, 83+86, 84+85) one unwatermarked full sheet (75+96) one half-sheet of stock E, f. 74 one half-sheet of an unidentified stock, f. 90 one unwatermarked half-sheet, f. 93
• Booklet IIb (religious verse)
• quire 6? (six leaves) ff. 97-102 four leaves of stock E, ff. 97, 99, 101, 102
• quire 718 (-1, -10, -11) ff. 103-17 six full sheets of stock G (103+116, 104+115, 105+114, 106+113, 107+112, 108+111) one half-sheet of stock G, f. 110 two unwatermarked half-sheets, ff. 109, 117
• quire 86+1 (+7) ff. 118-124 two full sheets of stock G (118+123, 119+122) one full sheet of stock E (120+121) one half-sheet of stock E, f. 124

---

55




• Booklet III (Richard the Lion-Hearted, verse infancy of Christ)
• quire 922+1 (-20, -21, -22, +23) ff. 125-144 eight full sheets of stock H (128+143, 129+142, 130+141, 131+140, 132+139, 133+138, 134+137, 135+136) two half-sheets of stock H, ff. 125, 127 two unwatermarked half-sheets, ff. 126, 144
• quire 1024 ff. 145-168 twelve full sheets of stock J (145+168, 146+167, 147+166, 148+165, 149+164, 150+163, 151+162, 152+161, 153+160, 154+159, 155+158, 156+157)
• Booklet IV (two alliterative debates)
• quire 1118 (-14, -15, -16, -17, -18, most cancels) ff. 169-181 four full sheets of stock J (174+181, 175+180, 176+179, 177+178) three half-sheets of stock J, ff. 170, 171, 173 two unwatermarked half-sheets, ff. 169, 172

To this information about format one needs to add what is known of the sources of Thornton's exemplars. Long ago, Angus McIntosh demonstrated that Thornton derived both the alliterative Morte and the prose work called The Privity of the Passion from an exemplar which was probably copied in southwestern Lincolnshire.[8] In addition, McIntosh was able to identify linguistically a second Thornton exemplar, this one from the area near the junction of Yorkshire, Lincolnshire, and Nottinghamshire. From this second exemplar, Thornton derived a large number of works, including the romances Octavian, Isumbras, Diocletian, Percival of Galles, and The Siege of Milan; the prose Abbey of the Holy Ghost; and the two alliterative debates Winner and Waster and The Parlament of the Thre Ages.

One's thinking about how Thornton acquired materials and how he went about copying them can begin with a salient fact. As George Keiser has pointed out,[9] the two works which Thornton acquired together in the Lincolnshire exemplar both open booklets in the manuscript now at Lincoln Cathedral (the Morte appears on ff. 53-98v, The Privity on ff. 179-189). This seems to be more than merely fortuitous, and probably the Lincolnshire exemplar was acquired by Thornton very near to the start of his copying career.

If one looks more closely at the second booklet of the Lincoln ms., into which Thornton copied the alliterative Morte, another interesting pattern emerges. The copying begins with a quire entirely on the paper-stock B; this is succeeded by a quire on stock B mixed with stock L and next by a quire

entirely on stock L. The Privity of the Passion, together with a succession of further short religious pieces, appears in the first of two full quires entirely of stock L at the head of Booklet III of the Lincoln codex.

From this disposition of paper-stocks, one can make two inferences. First, the pattern of papers at the start of Lincoln Booklet II suggests Thornton was running out of stock B at the time of copying: the gradual emergence of stock L as the normal form in which to write the codex seems predicated on a steadily diminishing supply of the other stock. This fact implies that Thornton copied the alliterative Morte first; only when this text (and stock B) was finished did he turn to The Privity and the pieces of religious prose which follow it.

A second inference is possible because stock B does not occur uniquely in Lincoln Booklet II. Other texts which appear on stock B must be contemporary with the alliterative Morte in copying, since they must preceed the exhaustion of this stock. Two such patches occur in the codices: quire 9 later in Lincoln and quire 3, which opens Booklet II in the London ms. As the two codices are now bound, quire 9 appears near the end of the same booklet as the Morte; however, this quire appears at some point in its history to have been independent and to have begun at the current f. 154. It could then, originally, have formed a single-quire fascicle.[10]

It appears that Thornton acquired the texts on these folios—The Awntyrs off Arthure (literarily, a pendant to the Morte, yet lacking the spelling forms which typify the Lincolnshire exemplar) and The Northern Passion—at a time very near his gaining the Morte. But both relevant quires are comprised totally of stock B; in consequence, their production may in fact have been completed by the time Thornton began copying from the Lincolnshire exemplar. Indeed knowing The Awntyrs and recognizing it as dependent upon an alliterative Death of Arthur may well have stimulated Thornton into seeking out an exemplar like the Lincolnshire. It is further conceivable that The Northern Passion and the text which follows it, The Siege of Jerusalem, may have had some loose attachment to The Privity. One of the few parallel copies of this last work survives in a Yorkshire manuscript with The Siege—a Robert H. Taylor ms. at Princeton.[11]

Having identified the Lincoln ms., Booklets II and III, as materials likely begun early in Thornton's career as copyist, one can move through these early booklets to examine their possible connections with other materials in the manuscripts. In this regard, study of the romances fascicle, the second in the Lincoln volume, suggests that copying proceeded contemporaneously on this fascicle and the religious prose. Both Lincoln quires 8 and 11-12 share stock L. However, by the time Thornton finished off the booklet of devotional prose, apparently very little of stock L remained: two sheets occur,

stuck into quire 15, as if they were odd leaves only useful to fill out a quire.

The completion of the romance booklet seems to indicate other materials were being copied contemporaneously. Quire 7 contains two sheets of stock C, which appears elsewhere in the London Booklet IIa (quire 4, also romances) and at the end of the Lincoln devotional book (quire 15). In addition, quire 10, which concludes the booklet, is on a mixture of stocks A, E, and K. This seems to place its composition after London Booklet I, where the Cursor Mundi extracts appear entirely on stock A; the mixture of stocks E and K recalls, and presumably indicates composition contemporary with, Lincoln Booklet I, where the Life of Alexander appears.

The romances in both manuscripts appear to have been copied straight on, since many of them reflect a single source, the Doncaster exemplar. Thus Thornton would have copied Octavian, Isumbras, and Diocletian (Lincoln ff. 98v-122v) in direct succession; at this point, with the acquisition of at least one other exemplar with desirable texts,[12] he seems to have decided to join the independent fascicle, Lincoln quire 9, with these other romances. This change involved refolding the quire to catch runover from these texts following on Diocletian (St. Christopher, Degrevant, and Eglamour; with some intervening blank pages, ff. 147-153v, filled in later). Thornton then used the folios following The Awntyrs off Arthure to begin Percival, another text he had discovered in the still-retained Doncaster exemplar.

Extension of the romance quire in the London ms. also involved copying from the Doncaster exemplar. The Siege of Milan shows the same dialect forms as the other romance texts; presumably, given the paucity of Charlemagne materials in Middle English, the adjacent romance Roland and Otuel reflects a common exemplar (even though it lacks distinctively Doncaster spellings). And ff. 271-276v of the Lincoln MS., containing The Abbey of the Holy Ghost, is also on stock C and a text from the Doncaster exemplar; it comes at the end of the Lincoln devotional book, although, as we will see, it need not for that reason have been copied last.

The production of the romance quire in the London ms. may have been less straightforward than the Lincoln romances. Quire 4 may have begun as fourteen leaves, largely on stock C, originally a size calculated to handle only The Siege of Jerusalem. Thornton would have produced a larger quire by suppletion, inserting into the original leaves extras which extend the writing area; this extension involves the unique examples of stock D and enabled Thornton to continue on with the Charlemagne materials. At the end of the subsequent quire, the fifth in the London ms., Thornton may have used a few originally blank leaves (ff. 94-96) to copy in short poems.

Also contemporary with the romance books is Booklet IV of the London ms. This fascicle contains the two alliterative debates, which Thornton again obtained from the Doncaster exemplar. This quire is written on stock J, which indicates production possibly a little later than the preceding Booklet III. This textual section begins with a quire all of stock H, then succeeded by one completely of stock J.

The devotional fascicle of the Lincoln ms. (Booklet III) may have involved considerable exemplar shifting. Some of this material appears to have had rather narrow ranges of circulation or to have circulated as units quite distinct from some of the other materials included. This feature suggests that Thornton may have had access to some small booklet-like exemplars or may have made eclectic choices of groups of texts reaching him from discrete sources.

Several such units may be pointed out. At the end of quire 12, and overlapping the boundary into quire 13, is a brief sequence of Walter Hilton texts—Angels' Song, Medled Life, and an excerpt from the Scala (ff. 219v-230).[13] A second such unit occurs at the end of quire 13 and again crosses the quire boundary, the sequence of texts known elsewhere as Gracia Dei (ff. 233v-250v). A third group of texts with an independent pattern of circulation are the brief bits of Rolleana on ff. 192-195. And finally there is the group of Rolle lyrics which the Lincoln ms. shares with Longleat 29 and Cambridge Univ. Library Dd.v.64.

But although Thornton may have relied on a sequence of exemplars, the copying of this portion of his texts seems to have been generally straight-forward. The only exception to this statement is Lincoln quire 15. This unit, as George Keiser has persuasively suggested, may have begun as an 18 (ff. 253a [now lost], 254-261, 271-279), all on stock C. To follow Keiser, this would have been an independent fascicle, beginning at f. 271 with The Abbey. Eventually leaves were added to the quire, after it had been refolded, in order to accommodate St. Jerome's Psalter (ff. 258v-269v), a text copied only later. These additions were of two sorts—first a pair of sheets of stock L were quired within the extant quire. That there are only two sheets of L suggests that the additions occurred at the very end of Thornton's supply, after the completion of quires 11 and 12. The other addition, two half-sheets on the unique stock N, is probably secondary: Thornton may have had trouble gauging how much space he needed to finish off the Psalter and associated prayers, his unique piece of copying in Latin. The extreme abbreviation of his exemplar might have confounded his best efforts at estimating how many leaves the piece would occupy.

Only two portions of Thornton's copying remain undiscussed. Quires 16 and 17 of the Lincoln ms. are utterly independent and on a single unique paper-stock. These may well be much older than the remainder of the manuscripts and indeed Thornton's first essay at copying. The paper-stock is of a

type which would have been available from 1413 on; in contrast, the remainder of Thornton's papers, insofar as they are datable, suggests that the scribe's main work was in the period roughly 1430-49 or perhaps slightly later.[14]

Booklet IIb of the London ms. is written on stocks G and E. This mixture suggests copying during the extension of the Lincoln romances, where stock G appears in quire 7. This copying would also be contemporary with use of stock E at the very end of the romances (Lincoln quire 10) and with the copying of Lincoln Booklet I. Additionally, such copying would be absolutely contemporary with the extension of the devotional materials into quire 14 of the Lincoln ms. (a quire fully on stock G). This reconstruction suggests that retailoring quire 15 may have been one of the last steps in the preparation of the manuscripts.

To summarize, one can cast these inferences into a tabular form which indicates the history of Thornton's copying efforts. The distribution of booklets into groups put into different codices presumably was a decision made at the end of production, if the separate manuscripts indeed reflect any of Thornton's behavior at all and are not later arrangements.[15] Copying proceeded in the following order:

• 1. Lincoln Booklet IV, perhaps many years before the remainder.
• 2. More or less contemporaneous undertaking of four new booklets, two from the Lincolnshire exemplar: Lincoln Booklet II (quires 4-6), Lincoln Booklet II (quire 9, at this point an independent fascicle beginning at f. 154), London Booklet IIa (quire 3), and Lincoln Booklet III (quire 11). The last is probably slightly later than the rest; London Booklet IIa and Lincoln quire 9 probably predate Lincoln quires 4-6.
• 3. Extension of Lincoln Booklet III, derived from a variety of exemplars, continuing on an indeterminate basis throughout the following steps (quire 14 is as late as step 9 below).
• 4. London Booklet III could have been transcribed at any time before the copying of London Booklet IV from the Doncaster exemplar (5 below). Given the absence of overlap with paper-stocks used in the productions described in 1-3 above, I have placed this production at the latest possible point: it could be contemporary with some portions of 5.
• 5. Copying of texts derived from the Doncaster exemplar, more or less contemporaneously, into four booklets, two new and two extensions of preexisting units: Lincoln Booklet II extended (quires 6-7), Lincoln Booklet III

  ---

  60

  
  
  
  (quire 15, at this point an independent fascicle beginning at f. 271), London Booklet IIa extended (quires 4-5), London Booklet IV.
• 6. During this period, from unidentified exemplars, Thornton acquired the texts of Lincoln Booklet II (quire 8); he subsequently reversed quire 9 to handle their text runover.
• 7. Before he relinquished the Doncaster exemplar, Thornton copied London Booklet I from a different exemplar.
• 8. Further copying from the Doncaster exemplar to conclude Lincoln Booklet II (quires 9-10).
• 9. Copying of two new units, Lincoln Booklet I and London Booklet IIb, is contemporary with the preceding step.
• 10. Lincoln quire 15 reversed so as to join the remainder of Booklet III, and St. Jerome's Psalter copied into its new and expanded first half.

This chronology scarcely can be argued to bring order out of chaos. Rather, it suggests a particularly complex set of procedures: typically Thornton seems to have worked contemporaneously on four or five emerging fascicular manuscripts. He was remarkably flexible in his methods. None of the larger booklets was absolutely completed until Thornton arrived at the end of all his copying: each was capable of extension and of melding with other units, so long as new texts could be acquired. The only rule, not a particularly rigorous one, which Thornton seems to have followed was that each fascicle contained works which were generically homogeneous.[16]

But if the chronology as I sketch it seems complicated, one should remember that my model in fact must simplify. It does so by a silent assumption that Thornton's use of paper is motivated, rather than arbitrary or irrational. I assume that Thornton received his materials from the paper merchant in lots comprised of a single stock and that he routinely used these stocks straight through until they were exhausted. In my account, Thornton only mixed stocks when his supply was running out. The alternative to this view would be a proof, based on an extensive survey of surviving English paper manuscripts, that mixed stocks (and especially mixtures within single quires) represent a commonplace and expected feature.

Second, my account surely simplifies in presenting only the copying Thornton performed which has survived. There is no reason not to expect extensive losses, which would presumably add to the difficulty of any chronology of copying.

This analysis, besides providing a chronology, also suggests some conclusions about Thornton's access to sources, his exemplars. Minimally, he must have been able to get to his hands on at least fifteen separate manuscripts, possibly as many as twenty. These would have included:

• 1. an archetype of the medical recipe book[17]
• 2. the Lincolnshire archetype
• 3. the Doncaster archetype
• 4. between four and eight archetypes for various materials in Lincoln Booklet III
• 5. an exemplar of The Awntyrs off Arthure
• 6. an exemplar of The Northern Passion, possibly with The Siege of Jerusalem attached (both perhaps with exemplar 2 above)
• 7. two separate exemplars for the works of London Booklet III
• 8. three separate exemplars for the longer texts of Lincoln quires 8 and 9
• 9. an exemplar of Cursor mundi
• 10. an exemplar of the Life of Alexander
• 11. an indeterminate number of exemplars for briefer verse texts, most of these religious, written into London Booklet IIb and used as page fillers elsewhere (notably within Lincoln quire 9 and at the end of quire 10); some may have overlapped with other exemplars used elsewhere, Thornton putting materials into different fascicles on generic grounds.

These speculations suggest a considerable acquisitiveness. However, there is some suggestive evidence that this acquisitiveness was practiced within a very narrow range and that Thornton participated in a lively local literary culture. McIntosh has noted Thornton's slackness at converting materials written in other dialects into his own. That most of the materials in the two manuscripts do not seem to show signs of having come from other dialects implies that Thornton obtained them locally, written in a dialect homogeneous with his own. The works may not have been composed in that dialect; but their circulation within it implies a considerable provincial literary culture in north Yorkshire, one about which we know very little indeed.[18] It may be that this effort at suggesting the order of Thornton's copying will allow further inferences about that culture.

Notes