University of Virginia Library

Search this document 


expand section 
expand section 
collapse section 
expand section2. 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 

expand section 


Unless he was prepared to make his own fair copies a medieval author was dependent on scribes for the dissemination of his work. A few authors have recorded their irritation at the shortcomings of this manner of 'publication', none more strongly than Chaucer in his Wordes unto Adam, His Owne Scriveyn,

Under thy long lokkes thou most have the scalle
But after my makyng thou wryte more trewe,
So ofte a-daye I mot thy werk renewe,
It to correcte and eek to rubbe and scrape,
And al is thorugh thy negligence and rape.[1]
Chaucer describes for us in general terms what the situation was: scribes were inaccurate through carelessness and hastiness and their work required


Page 67
constant revision and correction by the author.[2] In the case of John Capgrave we have what may well be a unique opportunity to examine just such a situation through concrete manuscript evidence, by collating a rejected copy (by an anonymous scribe) of part of Capgrave's The Solace of Pilgrimes with the author's own copy. Two aspects are of major importance. From the point of view of textual criticism in general it will be instructive to observe the extent and the nature of the errors which made a piece of copying unacceptable to an author. From the point of view of Middle English textual transmission it will be of interest to observe the linguistic changes made by a scribe working in an author's scriptorium.

I have shown elsewhere what was involved for at least one medieval English author, Capgrave, in producing manuscripts fit for presentation.

After he had first composed a work, the author then added fresh material arising from further thought or reading. Thus it could take some time to compile a finished work. The resulting 'original' manuscript was too untidy for presentation. A fair copy therefore had to be made. Sometimes considerable haste had to be employed in completing a presentation copy for a particular purpose. Either the author or another scribe (probably supervised by the author) could make this copy. Whether the copy was made by another scribe or not may have depended on the availability of a scribe whom the author considered at least as proficient as himself. [If the author was to make the fair copy he decided upon a script appropriate to the manuscript's destination.] If another scribe was to make the fair copy, the author went through his 'original' manuscript occasionally writing instructions so that the scribe could avoid any obvious blunders. Sometimes the author himself copied a difficult piece such as a diagram, even when the main body of the text was copied by a proficient scribe. The new manuscript was then carefully and extensively revised by the author. When the 'patron' was known, the author marked out certain passages for special attention, and occasionally added his personal mark at the end of the text. Lastly the manuscript was bound before being sent to its destination. [All this could take place in the author's scriptorium.][3]

Part of the evidence for these conclusions comes from three manuscripts which contain, in the case of [3], all that survives of, and in the cases of [7] and [8], fragments of Capgrave's 'The Solace of Pilgrimes'. These manuscripts are:[4]


Page 68
[3] Oxford, Bodleian Library, MS Bodley 423, fols. 355-414 (SC 2322e);[5]
[7] Oxford, All Souls College, MS XVII,[6] fols. i-ii and pp. 221-24;[7]
[8] Oxford, Balliol College, MS 190,[8] fols. 116-19.
Manuscript [3] is the holograph manuscript of Capgrave's 'The Solace of Pilgrimes', a description of Rome (c.1451),[9] written in the author's hand but not signed by him (the end is missing), and contains corrections by the author.[10] Manuscripts [7] and [8] are also from the author's scriptorium. Both contain Capgrave's Latin work 'De Fidei Symbolis' (c.1462) written by different scribes but corrected by the author; in addition the front endleaves of [8] contain a portion of the same text in the author's handwriting. The endleaves at both the front and the back of [7], fols. i-ii and pp. 221-224 (fol. ii and pp. 221-222 are ruled but contain no text), together with those at the back of [8], fols. 116-119 (fols. 116-117 are ruled but contain no text), form a consecutive portion of a copy in a fairly good current hand, not the author's, of Capgrave's 'The Solace of Pilgrimes'.[11] The text in these fragments is evidently unfinished since there is no rubrication (except for chapter-headings, which were evidently written in red ink as the scribe went along) and illuminated or flourished capitals have not been provided in the spaces left for them. The portion of text corresponds to the text in [3], the author's copy, as follows:


Page 69
[7], pp. 223-224 = [3], fol. 364r (line 27) to fol. 364v (line 33);
[7], fol. i = [3], fol. 364v (line 33) to fol. 365r (line 35);
[8], fols. 118r-119v = [3], fol. 365r (line 35) to fol. 366v (line 2).
Since fol. 364 is the tenth leaf of [3] and since one leaf of [3] contains almost exactly 1.7 times the quantity of text on a leaf of the other copy I deduce that we have what would have been fols. 17-24, i.e. all four sheets of the third quire, of another copy of 'The Solace of Pilgrimes', but that this copy was abandoned at what would have been fol. 20v (the centre of the quire) and put aside as waste (see Fig. 1). For convenience I shall refer to the version of the text in [3] as A, the author's, and the version in [7] and [8] as S, the scribe's.


From the circumstances of survival, as endleaves in manuscripts produced in the author's scriptorium, it is evident that the copying of the S-version must have been discontinued and set aside on the author's instructions. Since there are no corrections or alterations to S as it has survived we must assume that Capgrave made his decision to terminate it on the basis of checking the now lost quires 1 and 2. Nevertheless it is probable that the reprehensible features of those quires are continued in the surviving third quire.

The precise relationship between the two versions, A and S, has never been examined. Hingeston, who did not know of the existence of manuscript [3], the A-version, thought that the S-fragments were in the author's handwriting (Chronicle, p. 356), a view now universally discounted. Because of what Hingeston called 'the . . . carelessness of the spelling' and 'the incompleteness of many of the sentences' he thought the fragments 'were only first and rough copies', i.e. a rough draft. This view was contradicted by Bannister, who formed the impression that the S-fragments 'formed part of a late copy which must have been made from dictation' (in Solace, ed. Mills, p. xiv). In order to find out whether these earlier views are true I have made a complete collation of the S-fragments with the corresponding text of A. The new evidence thus revealed shows that both the earlier views are erroneous.

On a priori grounds it is probable that the S-scribe, working in Capgrave's scriptorium, was copying from an authorially approved exemplar, probably one in the author's handwriting; that exemplar could


Page 70


Page 71


Page 72
have been A but need not necessarily have been so since A is itself a fair copy and its existence presupposes an authorial original. As we might expect the relationship of S to A is close. This closeness is suggested firstly by the following sentence.[12]              
The kyng suppo|sed as it was þat  The kinge supposed as | it was þat 
þe cite had sent him for to gete  þe cyte had sent hym forto gete 
grace of þe | kyng and anon as he  grace of | þe kinge and a none as he 
say him with a grete ire and a  say hym with a grete | ire and a 
grete oth | he seide þese wordis  grete oþe he seide þese wordis 
(365v/27-30; Mills 31/7-9)[13]   (118v/16-19)[13]  
The only differences are in the spellings kyng/kinge, him/hym, cite/cyte, anon/anone, oth/oþe, differences which reflect merely the interchangeability of i/y, th/þ, and a preference in S for spellings with final e. Secondly, the close relationship of S to A is suggested, even more strongly, by the fact that spelling-features that are unusual in A reappear in S, where they are also unusual: for example, A's neuly (with u instead of usual w) at 366r/1 (31/22) reappears in S (119r/6), A's f (instead of usual ph) in profecie (365v/38; 31/16) reappears in S's profe|cie (118v/29-30), and A's y (instead of usual e) in by for' (365v/39; 31/17) reappears in S's by fore (118v/31).

In fact, the relationship of S to A is as close as it could be: S is copied directly from A. There are three kinds of evidence which lead to this conclusion.

(1) A and S share some common errors. The c in A's vescal 'Vestal' (364v/27; 28/7) reappears in S's vescall' (224/23). The first four letters of A's guynosopistis 'Gymnosophists' (365v/3; 30/17) reappear in S's guynesopistes (118r/15); for other Middle English spellings of this word see MED Genosophis. Wrong word order (not before myth) in A's þat þing whech he had not him\self/ne not myth haue (365v/10; 30/23-24)


Page 73
occurs also in S's þat þinge | whiche he had not hym selue not myght haue (118r/24-25). (This wrong word order was perhaps responsible for S's omission of ne.) And the omission of the word him (after vndirtok) in A's Tho þei vndirtok of his pri|de [14] (365v/11; 30/24-25) occurs in S's Tho thay vnd|yrtoke of his pride (118r/26-27) as well. This evidence could mean that A and S derive from a common exemplar or that S is copied from A. For reasons already explained A can hardly be a copy of S.

(2) At least one error in S is most plausibly explained as having been induced by the line-division in A: S has all' þe places (ir/4) for A's Alle þe|se places (364v/35-36; 28/15). If this explanation is correct S must be a copy of A.

(3) One correction in A is wrongly incorporated in S in such a way that the conclusion that S was copied from A is inescapable. At 365r/31 (30/4) A has þe pryuy ∧ of the world | ∧ þingis, where þingis occurs in the right-hand margin after the end of the line and the carets, in red ink, indicate that the word is to be inserted in the line after pryuy (see Plate I). S has þe pryuy of þe worlde þi|inges (iv/26-27): see Plate IIa. Evidently S must have been copied from A, and, when it was, the word þingis was regarded as belonging at the end of the line after which it occurs in the right-hand margin. Possibly S was copied from A before A was rubricated (hence the failure to take account of these carets in red ink—they were not yet there) but this possibility seems unlikely as a caret in brown ink at 365v/11 was also ignored (see (1) above and n. 14).

There are three possible reasons why Capgrave rejected the work of the S-scribe, (a) textual, because there were too many copying mistakes, (b) linguistic, because there were too many scribal alterations of spellings and forms, and (c) palaeographical, because its presentation was unsatisfactory. These reasons are arranged in probable descending order of importance: if (a) applies, (b) and (c) probably do not, except possibly as corroborating factors. To take them in ascending order of importance: (c) could hardly apply since the handwriting of S, though no example of calligraphy, is legible and reasonably tidy—the scribe's greatest fault (not a very serious one) seems to have been an inability to keep to his ruled lines, folios 118v and 119r having 32 lines to a page instead of the usual 31, and folio 119v having 34;[15] (b) could apply—these nonerroneous


Page 74
scribal alterations are discussed in the third part of this article—but probably does not, as (a) certainly applies.

Since A is a corrected authorial copy and S was copied directly from it and then rejected, all in the author's scriptorium, the information yielded by collating the two versions may be of use to those concerned with textual transmission in general. It should provide some guidance as to the kind of errors that were most likely to occur when a work was copied for the first time (primary errors), and their relative frequencies, as well as exemplifying a degree of inaccuracy which was beyond an author's tolerance. It should also provide insight into the nature and extent of those alterations that are purely scribal. Part II of this article is concerned with the scribe's textual errors, part III with his linguistic changes.