University of Virginia Library

ERRORS IN THE MALORY ARCHETYPE:
THE CASE OF VINAVER'S WIGHT AND
BALAN'S CURIOUS REMARK
by
RALPH NORRIS

THE text of Sir Thomas Malory's Morte Darthur is based upon two authoritative
witnesses: the Winchester Manuscript and the incunable printed by Wil-
liam Caxton in 1485. 1 Scholarship, primarily by Eugène Vinaver, has established
beyond reasonable doubt that the two versions descend from a common arche-
type rather than one from the other and that the arche-
type was not Malory's own autograph. 2 This situationallows textual scholars to reconstruct more of
Malory's exact words from the errors and interference that are an inescapable
part of the transmission of medieval romance than would otherwisebe possible. 3
Not only does each witness allow for a check upon the other, but in conjunction
with Malory's sources they can at times allow scholars to correct the archetype
and glimpse the exemplar that must lie beyond it.

The cases discussed below show that the scribe of the archetype occasionally
misread his exemplar. One has long been familiar to Malory scholars from the
introduction to Vinaver's magisterial editions; the other is a more recent contri-
bution to Malory studies.However, neither scholar seems to have come as close
as possible to the reading of the exemplar of the archetype, and the new critical
edition by P. J. C. Field retains Vinaver's readings atthese points. 4 Although no
such textual problem can ever be solved with certainty, probable solutions can
be found that create clauses much more characteristic of Malory's usual style and
that, therefore, are more likely to take us a little closer to the ideal of recovering
Malory's exact words.

I

In addition to being Malory's most important twentieth-century editor, Eu-
gène Vinaver was a significant exponent of the theory of editing medieval lit-
erature that is associated with his mentor, Joseph Bédier. This theory calls for
an editor to choose a single manuscript, the best manuscript so far as it can be
determined, and only to emend manifest errors and readings that could not
possibly be correct. 5 Because ofthis, Vinaver's three-volume edition The Works
of Sir Thomas Malory took a conservative approach towards emendation of the
text. This edition is based upon the Winchester Manuscript but also uses the
Caxton to fill Winchester's lacunae and to shed light on various textual cruces.
Furthermore, despite his allegiance to Bédier's motto, conserver le plus possible,
réparer le moins possible, 6 Vinaver was, to his credit, willing to recognize a number
of occasions that seemed to demand conjectural emendations, that is to say,
emendations to the text that produce readings that are not found in either me-
dieval witness. 7

One of the more notable examples of this kind of emendation occurs in the
"Book of Sir Tristram" section. Because this error is the basis for Vinaver's argu-
ment that the archetype of Winchester and the Caxton was not Malory's origi-
nal autograph, he discusses it at length in the introduction to his edition: "The
following sentence in Caxton, totally unintelligible as it stands, but reproduced
without comment by all modern editors, should suffice to prove this." 8

And the meane whyle word came vnto sir Launcelot and to sir Trystram that sire Carados
the myghty kynge that was made lyke a gyaunt / that fought with sir Gawayn and gaf hym
suche strokes that he swouned in his sadel / and after that he took hym by the coller /
and pulled hym oute of his sadel / and fast bounde hym to the sadelbowe / and so rode
his wey with hym toward his castell / 9

Vinaver asserts:

This reading is the result of two mistranscriptions: the words made lyke a gyaunt thatare
clearly a corruption of some such phrase as made lyke a gyaunt whyght (or whycht); some early
scribe, mistaking the final t for e, made whyght (or whycht) into whyche, and a later redactor,
probably Caxton himself, changed whyche to that. The first of these two errors is found in
the Winchester MS.; hence it must have occurred in the common source of our texts, and
this source (X) could not, therefore, have been Malory's own manuscript. 10

The reading of Vinaver's edition is, therefore,

And meanewhyle worde com to sir Launcelot and to sir Trystramys that kynge Carados,
the myghty kynge that was made lyke a gyaunte which<t>, fought wyth sir Gawayne and
gaff hym suche strokys that he sowned in his sadyll. 11

Vinaver makes fought into the verb of the main clause ratherthan of a subordinate
clause, thus producing a complete sentence, and this resolves one of the three
differences in wording between the two versions. Field endorses this reading
with a small change in the spelling of wyght and a further emendation that will
be discussed below.

However, brilliant as this proposed solution is, it is unlikely to be correct. In
the first place, Malory is not likely to have used the word giant as an adjective,
as Vinaver's emendation would require. Nowhere else in the Morte Darthur does
Malory use giant in this way. 12 Nor is it probable that Malory used the word wight
to refer to Carados. Malory certainly knew this word in its noun form, referring
to a person or being; however, it is not an active part of his vocabulary. Infact,
it appears only once elsewhere in the Morte Darthur , in the Roman War section,
in which the narrator describes the giant of Mont Saint Michel as "the foulyst
wyghte that ever man sye" (157.9). 13 In this instance, Malory was probably in-
fluenced by his source, which for the Roman War section is the alliterative Morte
Arthure. Although the sole surviving manuscript of this poem does not contain this
word at the corresponding place, it appears at a slightly earlier point: an old lady
whom Arthur encounters curses the giant saying "Weryd worthe þe wyghte." 14

Defenders of Vinaver might concede the influence of the poem in this case
but argue that this use of wight supports the idea that Malory used it to describe

Carados, who was made like a giant. However, despite the fairly large number
of giants and huge knights in the Morte Darthur , Malory never selects this word
elsewhere. Nor does his source for the section containing the passage in question,
the prose Tristan , seem to contain anything that would suggest this reading.
Although no form of wight could be expected in an Old French text,one of
the interesting aspects of Malory's adaptation of his sources is his tendency
to choose English words that sound like words of different meanings in his source. 15
Vinaver's argument would be much stronger, therefore, if Malory's source con-
tained a word with a similar sound that might have brought this rarely-used
word to Malory's mind. The editio princeps of the prose Tristan is based upon a
manuscript that is very close to Malory's source, 16 and at the equivalent point
in the story, it calls Carados, "vng geant que on appelloit karadoes le sire de la
douloureuse tour." 17 The semantic range of wight could include geant, but Malory
always uses the English derivative for this word, as he does here as well. There-
fore, because neither Malory's normal usage nor the source suggest it, the odds
seem to be against Malory's actually using the word wight in this passage.

If Vinaver's emendation is therefore rejected, we return to the question of
what Malory actually wrote. Since Vinaver first proposed this idea, several other
editions of the Morte Darthur in addition to Field's have appeared that contain
the passage in question, yet, despite their many differences, these editions, by
Janet Cowen, James Spisak, Helen Cooper, and Stephen Shepherd, all agree
in their solution to this problem. Their texts each read, "Carados …that was
made like a giant fought with Sir Gawain," 18 thus removing the relative pronoun
that Vinaver believed had originally been wight. This is a sensible solution. It
produces a clear, intelligible reading at the priceof removing a single word, and
the reading thus created could not differ greatly from the meaning that Malory
had intended. 19 This solution seems to imply that arrhythmia had caused the
scribe of the archetype to intrude an unwanted word into his copy. Arrhythmia,
however, is generally a hypothesis of last resort, so we should try to exhaust all
other possibilities before reaching this conclusion. In fact, looking at the context
suggests something besides an isolated error.

Vinaver's solution was characteristically ingenious, and the way he presented
it makes it seem even more so. In his introduction he creates the impression,
though without doubt inadvertently, that the two witnesses have only one sig-

nificant difference, that and which, and that he has explained that difference and
solved the unintelligibility of the passage at a stroke. However, comparison of
the passage in the two versions shows that they have three differences of equal
magnitude to the that / which discrepancy. In the Winchester, the passage reads:

And [___] meane whyle worde com to sir launcelot & to Sir Trystramys that kynge Ca-
rados the myghty kynge that was made lyke a gyaunte whyche fought wyth Sir Gawayne
and gaff hym suche strokys þt he sowned in his sadyll & after þt he toke hym by þe coler &
pulled hym oute of his sadyll & bounde hym faste to the sadyll bowȝe & so rode his way
wt hym towarde his castell. 20

Again, the Caxton version:

And the meane whyle word came vnto sir Launcelot and to sir Trystram that sire Carados
the myghty kynge that was made lyke a gyaunt / that fought with sir Gawayn and gaf hym
suche strokes that he swouned in his sadel / and after that he took hym by the coller /
and pulled hym oute of his sadel / and fast bounde hym to the sadel bowe / and so rode
his wey with hym toward his castell / 21

These discrepancies seem to show that the process that created them was more complex
than any single mistake.

If we take these discrepancies one at a time, the following suggests itself.
Malory probably wrote "And the meane whyle" rather than "And meane whyle."
In the 67 instances of his use of meanwhyle, Malory always precedes it with the,
or, in thirteen cases, this, except for the possibility of this disputed case. The next
discrepancy involves whether he wrote "Sir Carados" or "King Carados." The
fact that the phrase "the mighty king" follows in both versions argues in favor of
sir, since it is unlikely that Malory would feel the need to qualify "King Carados"
by explaining that he was a mighty king. 22 Probably, therefore, Malory wrote Sir
Carados as he sometimes did, and the Winchester scribe, who was more accus-
tomed to seeing this name preceded bythe royal title and whose eye may have
caught the word king in close proximity, made a mistake.

This mistake would be the easier for a scribe to make because in the Morte
Darthur there are three characters with this name. 23 The Carados of this scene,
brother of Sir Tarquin and lord of the Dolorous Tower, is not said to be a king
except in this episode, and in this episode, according to the Winchester, Carados
is called "King Carados" twice and "Sir Carados" four times. 24 In the Caxton,
he is referred to only as "Sir Carados." 25 Again excluding this disputed instance,
Malory refers to a characterby this name 36 times, 12 as "Sir Carados" and
24 times as "King Carados," and before this instance, the statistics are 12 and
six respectively. This fact shows that the scribe would be twice as likely to have

written King. 26 Either the Caxton regularized its copy or the Winchester scribe
varied his, or perhaps both happened.

There is some evidence that both may have happened. In 1498 Caxton's
successor, Wynkynde Worde, printed a new edition of the Morte Darthur, and
recently de Worde has been shown to have occasionally used Caxton's exemplar
to supplement Caxton's copy. 27 Surprisingly, this may, in fact, be the case for this
page as well: de Worde's edition deviates from the Caxton in eight instances on
this page, and of those instances Winchester agrees with de Worde fourtimes
and with the Caxton three. 28 None of the variations is very great, and all of them
together might be coincidence. If so they are de Worde's personal variation and
have no bearing on the issue, yet if they are not, then de Worde's edition gives
our conclusions some support.

One of the correlations between Winchester and the de Worde involves Ca-
rados. In the passage under discussion, de Worde agrees with the Caxton against
Winchester in calling Carados Sir Carados the mighty king; however, the de Worde
and Winchester both call this character King Carados against the Caxton's Sir
Carados when they refer to him next. If de Worde were following Caxton's copy
text for this page rather than the Caxton itself, it would confirm the conclusions
reached above. 29

Now we must account for the grammatical mistake that led Vinaver to draw
attention to this passage in the first place and answer the question of the relative
pronoun. Both versions agree that there is a relative pronoun, and, therefore,
either "which fought with Sir Gawain," "that fought with Sir Gawain" or a phrase
with a word that a scribe could mistake for one of the two must have been in the
exemplar of the archetype. That appears far more often in the Morte Darthurthan
which, at 8,426 occurrences to only 158 respectively. 30 In both previous instances,
we have seen that if the argument is sound the advantage goes to the Caxtonver-
sion, suggesting perhaps some problem with the exemplar of the Winchester or
with the attention of the scribe. 31 Together these facts argue in favor of that; how-
ever, Vinaver was probably right to say that Malory did not write either one.

The construction of the passage in question as preserved by its two witnesses
contains multiple subordinate clauses, one embedded into another, and this is
very unusual for Malory. Although one could not say that Malory never uses
subordinate clauses, paratactic constructions of simple clauses joined by coor-
dinating conjunctions, usually and or but, are much more common. 32 Of course,
this very fact might account for Malory occasionally losing his way when creating
a clause with a complex construction, yet any rational solution that removed a
double subordination would have some inherent plausibility.

The alternatives suggested by textual criticism are misreading and internal
contamination, that is reminiscence of words already written or anticipation of
those later on the page. Vinaver's attempt to restore the passage posits misread-
ing, and in this respect Vinaver is clearly correct. Internal contamination occurs
when, after looking away from the exemplar, the scribe resumes at the wrong
point and introduces words from elsewhere on the page into his copy. Although
it appears from the above that Vinaver failedto determine the exact misreading,
there is no reason to suspect that the that (or possibly the which) could have come
from any other place on the page nearby. 33

The most likely solution is that the misread word was had. The exemplar of
the archetype probably read,

that Sir Carados the myghty kynge that was made lyke a gyaunte had fought wyth Sir
Gawayne and gaff hym suche strokes that he sowned in his sadyll.

The scribe of the ar-
chetype would have misread had for þat. The scribe of the archetype would have taken the descender and front loop of the h for a þ, correctly
read the a and then,perhaps already having the word that in mind, mistook the
stroke of the d for that of a t. 34 There is evidence that the exemplar of the Caxton
was irregular and hard to read, 35 and perhaps the exemplarof the archetype was
as well. The Winchester scribe would then have altered that to which, and the
Caxton compositor would have reproduced faithfully from his exemplar, which is
in keeping with the accumulating evidence that the Caxton is the more accurate
witness for this passage.

This explanation is no more radical than the one Vinaver proposed. Al-
though misreading þ for h is not a commonly recognized scribal error, neither is
Vinaver's suggestion of misreading t for e. Moreover, this emendation is more
plausible because it creates a clause more typical of Malory. The reading thus

created is also more logical grammatically: because two levels of past tense are
involved in the narrator relating that a character heard news of a yet prior event,
the narrator's indirect discourse would naturally be in the pluperfect tense.

This conjectural emendation also has some support from Malory's source.
The relationship of Malory's writing to its sources can vary from almost literal
translation to very free retelling within the same tale, butin this case, Malory
seems to be working somewhere between these two extremes. The prose Tristan
reads here,

il [i.e. Gallehaut] estoit alle eu royaume de logres ou le roy artus auoit tenu court a
londres a vne penthecouste et y auoit este vng geant que on appelloit karadoes le sire
de la douloureuse tour / et celui par sa force auoit prins monseigneur gauuain et lauoit
emporte. 36

Although Malory's version is clearly not anything like a direct translation, he is
describing the same action. Therefore the French par sa force auoit prins could have
led Malory to write had fought.

One last point to consider before reaching our final conclusions about this
passage is the emendation of "word com to sir Launcelot" to "word come to
sir Galahalt" that P. J. C. Field introduced in his revision of Works. 37 Although
Winchester and the Caxton agree on reading Launcelot, Field is almost certainly
correct to emendas he does. If the reading of Launcelot were correct, the result
would be a sequence of events conspicuous for being both oddly-narrated and
implausible. Launcelotwould reenter the story for the first time since the end of
"The Tale of Sir Gareth" to hear these tidings, apparently apart from Tristram,
and then happen upon Carados and Gawain "by fortune" (333.15) before Cara-
dos could bring Gawain to his castle. In a tale of wonder, plausibility can stretch
far, and it istrue that we do not know what distance Carados had to travel with
his captive, but the story gives us no reason to think that rumor is running faster
than the hooves of Carados's horse.

In this instance, Malory's sources are no help. The story is related very dif-
ferently in the prose Tristan and in its own source, the prose Lancelot. In these
versions, the action is narrated directly to the reader rather than through the
intermediary of uninvolved characters hearing the news. Lancelot is present dur-
ing the abduction and pursues Carados, but he is not able to free Gawain until
after Carados has taken him to this dungeon. 38

Malory's story itself, however, shows that Field must be correct. The story
of Gawain's abduction and rescue is told to Tristram and Galahalt at Galahalt's
court before Tristram's departure. The story is framed on one side with the words
quoted above and on the other with, "So this same tale was tolde to Sir Gala-
halte and to Syr Trystrames, and Sir Galahalte sayde, 'Now may ye hyre the
nobles that folowyth Sir Launcelot.'" 39 The episode, therefore, ends as it must

have begun, with Sir Galahalte and Sir Tristram hearing one of the innumerable
stories of the nobility of Sir Launcelot, not with Launcelot and Tristram hearing
the story of Gawain's abduction.

Just how the error of Launcelot for Galahalt occured takes us to the limits of
what scholarship can discover. Perhaps, as Field suggests, the mistake is a "men-
tal echo caused by the preceding alternation of Sir Trystram and Sir Launcelot," 40
or, to paraphrase Field in another connection, because authors do not always
write what they intend, there must be a real possibility that Malory himself in-
advertently wrote Launcelot. If so, the reading that Malory intended will have ex-
isted only in Malory's mind until Field put it into print. 41 Or, finally, perhaps the
scribe of the archetype was confused by seeing Tristram and Galahalt together
immediately after Tristram was said to have taken his leave, 42 and if so, perhaps
he made the change to try to restore sense.

In any case, if the above arguments are sound then Malory's original
should be,

And [the] meanewhyle worde com to Sir Galahalt and to Sir Trystramys that [Sir] Carados
the myghty kynge that was made lyke a gyaunte had fought wyth Sir Gawayne and gaff
hym suche strokys that he sowned in his sadyll, and after that he toke hym by the coler
and pulled hym oute of his sadyll and bounde hym faste to the sadyll bowȜe, and so rode
his way with hym towarde his castell. 43

The possibility that the scribe of the archetype could misread one small word
for another, and in particular misread the stroke of a d for a t may shed light on
another small mystery of the archetype of Winchester and the Caxton.

II

In a study of corrected mistakes in Winchester, Takako Kato demonstrated
that, contrary to the impressionone might get from the apparatus of Vinaver's
edition, the Winchester scribes were not human automata, only reproducing their
text accurately or making only unconscious errors. 44 They also sometimes noticed
mistakes in their copy and tried to correct them.

One of her examples comes from the story of Balin. Balin has recently been
released from King Arthur's prison when he encounters his brother, Balan. The
Winchester version records Balan'swords as follows: "I am ryȝt glad of your
delyueraunce and of youre dolerous presonment." 45 Kato rightly observes that
the inclusion of the and makes nonsense of the passage, which certainly is why
the scribe deleted it, 46 but that the same word appears without the correction
in the Caxton: "I am ryght glad of your delyueraunce and of youre dolorous

prysonement." 47 Kato draws convincing conclusions about how most of the er-
rors she discusses probably originated, but her discussion of this error is less
extensive than the others, leaving the reader with the impression that the Win-
chester scribe restores Malory's original intention. However, as above, something
must have led the scribe of the archetype to make this mistake.

Here internal contamination can be ruled out, if not as swiftly as in the case
above. It is true that the scribe of the archetype could possibly have been misled
to write delyveraunce and by the words delyverde and which appear in the line below
in Winchester and two lines below in the Caxton. 48 Yet if that were so he must
have paused after delyveraunce, returned to his exemplar and found delyverde and
resumed his copying but realized his mistake after writing the single word and.
Then he must have returned to the correct place and continued without either
adapting the wording to tidy up the mistake or even, like his successor of the
Winchester manuscript, deleting the word. Such a situation would be most un-
likely in a case of internal contamination.

By way of comparison, to illustrate a more probable instance of internal
contamination, consider thissimilarly illogical sentence from "The Tale of Sir
Launcelot:" "Than Sir Launcelot made his complaynte unto the kynge, how he
was betrayed, and how he was brother unto Sir Lyonell, whyche was departed
frome hym he wyste not where" (197.34 – 198.1). This reading has been retained
by every edition based upon the Winchester, nor has any editor commented
on what appears to be an example of unintended humor: Launcelot complain-
ing that he has recently been betrayed and also complaining that Lyonell is his
brother. A look at the Caxton text illuminates how the Winchester scribe came
most probably to produce this reading. The Caxton reads, "Anone syre launcelot
made his complaynte vnto the kynge how he was bytrayed And how his broder
syre lyonel was departed from hym / ne nyst not where /" 49 The Winchester
scribe apparently committed an error of saut du même au même and returned to the
first how in the phrase in the line above the one that he intended and recopied "he
was," only realizing his mistake as he reached the word "betrayed." The scribe
then altered the rest of the sentence to compensate, an example of what Field
has called invisible mending. 50

The textual situation is different in the example from the story of Balin. Field
believes that this is a case of an intruded and, which "W often and C and the
archetype sometimes use … to cure textual difficulties, real and imaginary." 51
Given that the scribe ofthe archetype shared this tendency with his successor
of the Winchester Manuscript, this is a neat solution to this mystery. However,
it begs the question of what textual difficulty, an imagined one presumably, the
scribe of the archetype might have thought he was solving by adding a word that
destroys the sense of what he was copying.

A better solution is that the scribe of the archetype misread a word in his
exemplar without realizing it, and the misread word would almost certainly be
out. One can see easily how such a misreading could take place. Mistaking the
loop of the o for an a, the scribe would naturally interpret the double minims of
the u as an n, and finally mistake the stroke of the t for a d, the same error, in the
opposite direction, as above. This hypothesis would produce as the reading of the
archetype's exemplar, "I am ryght glad of youre delyveraunce out of youre doler-
ous presonment." This reading is idiomatic, makes very good sense, and is typical
of Malory's style. Malory never uses the word presonment other than this once, but
he refers to knights' release (or otherwise) from prison fourteen times and always
uses the construction oute of preson, never, say, merely delyverde of preson 52

The similarity of this misreading on the part of the scribe of the archetype to
the misreading of had for þat is remarkable. In both cases short functional words,
an auxiliary verb and a preposition, have been mistaken for another word, and
both contain a similar d/t confusion. Each of the two cases provides supporting
evidence for the validity of the other.

III

The effect of considering these cases together is to affirm both the importance
and the dangers of conjectural emendation in editing Malory's text. Kato's article
provides outstanding examples of how this principle can solve textual mysteries,
and even Vinaver, a textual conservative, recognized its usefulness.

Vinaver's intelligence and industry stand as an inspiration to all later Malory
scholars. His grand editions remain monuments of twentieth-century scholar-
ship and will doubtless continue to be the standard by which new editions of
Malory's work will be measured. And as this article shows, his conclusionsabout
the existence and nature of the archetype from which Winchester and the Caxton
independently derive remain solid.

However, in the cases discussed above, Vinaver appears to have been misled.
In the instance from "The Book of Balin," Vinaver's theory of editing prevented
him from considering the implications of the corrected mistake in his copy text:
because the corrected reading was not impossible, he apparently did not give
much thought to the fact that the Caxton contains the same mistake. 53 On the
otherhand, his emendation to wight in "The Book of Tristram" section actually
demonstratesa danger of conjectural emendation that Bédier and Vinaver sought
to avoid with their best text system of editing. Perhaps swayed by the excitement
brought about by examining the then newly-discovered Winchester manuscript
with its promise of solutions to mysteries, Vinaver introduced into the standard
edition of a major author an erroneous reading of his own creation. 54

Of these two pitfalls, however, the narrowness of Vinaver's theory is the more
harmful because it focusesattention away from the analysis of textual variants,
which is indispensable to textual criticism. Vinaver'scaret brackets alert the
reader to the hand of the editor. In all such cases the scholarly reader must judge
the validity of the emendation, which will lead to progress either as confirma-
tion or refutation. Yet the missing word in the story of Balin threated to vanish
from all consideration into the tiny type at the bottom of the page until rescued
by Kato's study, a small loss perhaps, but a distortion of a great author's syntax,
nevertheless. Applying a less narrow theory of editing and possessing a greater
appreciation of the critical importance of both major witnesses, we are in a posi-
tion to continue to recover more of Malory's exact words than ever before. Field's
new edition is a great step forward in this endeavor; it improves on Vinaver's text
in a multitude of closely-reasoned variants. This paper argues for afew more in
the ongoing effort to safeguard Malory's work from, as has been memorably put,
time's wallet of oblivion.