Arthurian Intertextuality[1]
Our topic is current methods in the study of
intertextuality in Arthurian literature. The emergence of this formidable term
in the last decade is a welcome sign. In the preceding decades, a number of
forces had combined to focus much literary study on individual, unaccommodated
texts. The New Criticism, Russian and other Formalism, Structuralism,
Hermeneutics and Reader-Response criticism, and the preoccupation of the
Deconstructionists with the self-reflexivity of texts, had all abetted a
decline in the study of literary interrelations. They discredited some earlier
philological study, source- and analogue-hunting, and some of the fuzzier kinds
of comparative study as ‘flights from the text.’ As Donald Maddox says,
scholars left off listening at the seashell’s lip for echoes of ancient seas
and turned to examine the convolutions of the conch itself.
Now, with the term intertextuality,
Semiotics has conferred a new benediction on the bookishness of books, and the
study of sources, influences, archetypes, allusions, and interrelations may go
forth and multiply. Of course a number of other notions have been smuggled into
the baggage of intertextuality, and some Arthurian scholars may have difficulty
with them. Along with the familiar fact that books absorb and transform other
books, the theorists sometimes also look upon nature, culture, and personality
as ‘texts’, the mental residue of verbal experience. They explore key words for
the infinite modalities of language; they worry discourse free of its
referential bonds, and they read the inviting spaces between the lines. If,
with all this playfulness, we can still tell a conch from a cauliflower,
perhaps we can return to the pleasures of setting one shell beside another in
taxonomic, evolutionary, or ecological arrays, as well as searching for those
ghostly gastropods resulting from la différence.
My thesis, which I hope will please
you, is that medievalists too have something rich to add to the intertextual
heap. There is, as yet, no settled conceptual vocabulary of intertextuality,
although Michael Riffaterre, Julia Kristeva, Jonathan Culler and others each
have their own. At this point, I think it is still useful to respect this
fluidity and toss in our own distinctions. Medievalists know that the
literature they habitually deal with is inherently intertextual in ways that
the theorists, accustomed to the conventions of print media, insufficiently
appreciate. Because works transmitted orally and by manuscript are unavoidably
variable, artists and artisans working within those media developed great
feeling for both the frustrations and the opportunities. For them a manuscript
could represent anything from an authoritative text to be respectfully handed
on intact to an open invitation to intervene in its improvement. Our modern
terminology for this range of possibilities is divided in origin between
textual and literary criticism. But in actuality it forms a continuum, at one
end of which we say that one text copies another, consults it, contaminates it,
conflates it and so on toward the other end, where we say that texts continue,
rehandle, adapt, borrow from, influence and echo one another.
Scholars know that several interesting
intertextual effects follow from the characteristic medieval mode of literary
production. Books are made from other books, literally, at all levels.
Principally, of course, through copying by hand, making literature a matter of
both model text and inherent variation. New works also sometimes proceed by an
accumulation of minute variation. And when they are most innovative, they
pretend to be most traditional, masquerading as the translation of some old
book. The games are endless by which medieval authors, narrators, and scribes
adduce authority and allude to other real and imaginary works, other branches
of the tale, in order to mark units of narrative, gain credibility, avoid
responsibility, and constrain future editors.
My main concern here is not,
however, to inform the theorists of what medievalists can add to their arsenal,
but to promote among students of Arthurian literature the best that their
colleagues have to offer and to pass on the criticism that they have made of
work that fails to heed these insights, although this may delight you less.
Take, for example,. The theorists’
interest in lost intertexts. For them it is one thing to interpret Mallarmé’s
silences, but for medievalists, who are desperately aware of how much early
publishing has perished, it can mean another. Not only have Arthurian scholars
been listening to unheard melodies since Bedier and before; following his
example, they have set out to fill the silence with reconstructions. But
whether Bedier assembled his prose archetype of the romance of Tristan and
Iseult for its power and beauty, we continue to make reconstructions to
account, intertextually, for the problems in the extant works.
Bedier was a pejorist, whose
belief that things keep getting worse is shared by such men as the poet Robert
Graves, who invented the term iconotropy to cover the process by which
later mythographers endlessly misunderstood the pre-Hellenic images of the
White Goddess and her doomed consorts. Roger Sherman Loomis was another
sometimes pejorist, who resorted to reconstructing the privileged Celtic
antecedents of things he thought Chretien and Robert de Boron could often only
garble. But among the more recent reconstructors, it is not gloom that
motivates, but assumptions about the direction and strength of intertextual
vectors from Latin to vernacular, from Celtic to French, from types like the
Fair Unknown to individualized protagonists, from religious to secular works.
And not the other way around.
Let me recall for you three recent
examples. Ulle Lewes published in 1978 The Life in the Forest: the Influence
of the Saint Giles Legend on the Courtly Tristan Story. She notes the
intriguing similarities between texts of the 12th century
Anglo-Norman Vie de Saint Giles of Guillaume de Berneville and the
courtly romances of Tristan and Iseult, particularly respecting their life in
the forest. To account for this intertext, she rejects the Virgilian topos of
the locus amoenus as a common source for both -- because it lacks the
friendly deer as a transition device from the wild to the civilized state.
Instead of accepting this animal as a familiar thing of Norman real
life, she assumes that the saint’s life influenced Thomas. This then requires
her to reverse the received chronology and to attempt to reconstruct Thomas’s
lost episode by conflating Gottfried, the Tristramsaga and the Middle
English Sir Tristrem as Bedier had done before. But this reconstructed
text has no basis in the manuscripts of Thomas and rests on a suppositious
ladder of influence down from Latin hagiography to courtly vernacular
entertainments. We get a hierarchical Middle Ages even when it isn’t there.
A more complex intertextuality is
asserted by Claude Luttrell in his book The Creation of the First Arthurian
Romance: a Quest (1974). Here again the assertion is of influence one way,
from Latin to French authors, in this case from Alain de Lille’s Anticlaudianus
to the Erec of Chrétien de Troyes. Again, a radical redating is
required, pushing the whole oeuvre up to 1184-6. Tony Hunt believes the
debt to be to the writings of the School of Chartres in general (BBSIA, ’78)
although he is inclined to accept the redating. The second major text on which
Chrétien is supposed to have relied is
an oral conte d’aventure of the Fair Unknown type, which Luttrell
reconstructs from incidents in Erec and other romances. Here the
reviewers have found that many of the intertextual similarities between Alain
and Chrétien should not necessarily be taken as signs of dependence, that
differences have been minimized, some details are either lacking in Erec
or are available in classical literature, the chansons de geste, or real
life. The intertextual elements brought forth do not, in the words of Barbara
Sargent-Bauer, “appear conclusive enough to bring about a drastic revision in
our evaluation of Chrétien’s originality and influence.”
Finally, some time ago D.D.R. Owen’s
The Evolution of the Grail Legend (1968) based a large part of its
argument on an attempt to reconstruct the Book of Count Philippe of Flanders on
which Chrétien claimed to basing his romance of Perceval. A
thorough-going Celticist, Owen traces the combination of an adventure story,
again of the Fair Unknown type, with a wholly secular Grail, as symbol of the
Chivalric ideal, from scattered elements in Irish myth and later sagas down to
a proposed common source for three works: the 12th century Dream
of Macsen Wledig, Renaut de Beaujeu’s Bel inconnu, and Count Philippe’s livre. This last is
supposed to be an Anglo-Norman redaction of a bilingual Welsh minstrel’s legend
centered on Caer Seint (Carnarvon). Owen has been criticized for specifically
rejecting the evidence of the Perceval manuscripts, in which the Hermit
Uncle is not reconstructable in the Caer Seint legend. Others have argued
vigorously for the authenticity of both this and the Gauvain portion of
Chrétien’s last romance. My point is that this reconstruction not only, like
the others cited here, neglects to work within the intertextual possibilities
offered by attention to manuscripts, it goes so far as to spurn it.
We do, of course, have several
excellent models of the appreciation of those manuscript possibilities in
action. I am thinking particularly of the work of some of the most eminent
disciples of Eugène Vinaver, who must be classed not with the pejorists but with
the meliorists in inclining to the belief that things often get better. And I
think it was this habit that set him over against Loomis as much as any
substantive doubt about Celtic antecedents. In his writings Vinaver gets nearly
mystical about the emergent and organic unities of the great prose cycles. For
him, Ferdinand Lot’s metaphors of sparterie, tapestry, and interlace
became transcendent forces evoking works whose meanings for their
contemporaries were far more important than the loss of original significance
or its corruption by confused and Christian iconotropic processes. Vinaver also
believed that the later Middle Ages saw, not the disintegration of these
tastes, but their replacement by a new aesthetic preference for shorter
romances with self-contained narratives following the adventures of a single hero. Between the Vulgate Cycle and
Malory’s Morte Darthur there are, as we all know, a number of
problematic manuscripts, such as the Huth Suite du Merlin and Michel
Gonnot’s Livre de Lancelot, and it was in dealing with the relations of
these texts and other parts of the Matter of Britain that Vinaver’s students
Fanni Bogdanow and Cedric Pickford considerably advanced our appreciation of
the subtleties of Arthurian intertextuality.
They each have pursued a different
aspect of Vinaver’s thought and inevitably differed over the interpretation of
these problem texts. Pickford observed that in MS B.N. fr. 112, the scribe
Michel Gonnot had gathered into a coherent tale some episodes involving Alexandre
l’Orphelin that had been scattered in earlier performances, more or less
interrelated, and related to other versions of other texts, with which they
might be bound. Both Pickford and Bogdanow knew that literature often proceeds
by a kind of dialectic of outdoing, undercutting, answering back, and
forestalling all of these. As more medievalists sensitize themselves of these
processes, we will find them making additions to the more familiar kinds of
intertextual studies.
My remaining examples of stasis and
progress come from recent Malory studies. The decades-long debate between
Vinaver and the students of R.M. Lumiansky lapsed into a kind of stand-off,
between those who thought Malory wrote one book or eight tales, with the advent
of structural, contextual, and stylistic studies. I will scan a few of these
for the purpose of noting an interesting convergence on an intertextual issue.
A while back, Stephen Knight
asserted in a little monograph on The Structure of Sir Thomas Malory’s
Arthuriad (1969) that both Vinaver and Lumiansky were wrong. The former for
dwelling on sources and thereby obscuring “the truly critical issues,” and the
latter for imposing modern novelistic unity in the triple scheme of failure in
love, religion, and chivalry. Knight finds instead a binary structure. In the
first 60% of the book, nearly through Tristram, there is a linear series
of more or less disconnected episodes bespeaking a juxtaposition of a “Gothic”
method of narration, while the latter 40% is steadily concerned with aggrandizing
and humanizing a single hero, Launcelot. This new thematic ordering is seen as
a change in literary intention.
The work of Elizabeth Pochoda, Arthurian
Propaganda: le Morte Darthur as an Historical Ideal of Life (1971) also
criticizes Vinaver’s predisposition to source study, as if Malory had no ideas.
She finds the intertext of his political notions in the works of John of
Salisbury, Bracton, and Fortescue. Her own main source, however, is The
King’s Two Bodies of Ernst Kantorowicz, which supplies her belief that
Malory criticized Arthur because his kingship was neither independent of his
fellowship nor immortal as an office. Pochoda finds this criticism inherent in
the structure of the Morte. Although she claims that Malory intended an
exposé of Arthurian weakness for the outset, she also finds the work falling
into two parts. Tales I through V, or the first half of Tristram,
present Arthurian chivalry as a valid political model to be imitated in actual
life, but the second half shows that society as inadequate. Reviewers have
noted her focus on failures and disregard of successes and find curious her
claim that Malory eliminated the role which fate played in his sources,
especially in view of Arthur’s portentous dream and the little serpent that
pops out of the Salisbury Plain. Malory, they insist, inherited Arthur’s fall,
and Pochoda ignores, for example, Elspeth Kennedy’s study of the place of
political and social ideas in the Vulgate Cycle. And she ignores the mediating
intertexts between the legal sources and later romances.
Mark Lambert’s Malory: Style and
Vision in “Le Morte Darthur” (1975) argues that ideas are quite secondary
to Malory’s mastery of an earnest style appropriate to his vision of
knightliness as the center of civilized experience. This combination of
alertness to aesthetics and to nobility of manner is also reflected in what
Lambert sees as the structure of the Morte, and he too sees a bifurcated
book. The ideal of chivalry is presented as attractive in the first six tales,
but is qualified in the last two by a new sense of tragedy. Again unlike
Pochoda, Lambert emphasizes the roles of fate and chance here rather than moral
responsibility, although unlike Moorman and company, he feels this tragedy to
be multicentric rather than organic.
We have here then, three modern
critics who variously disparage detailed intertextual study in their pursuit of
Malory’s personal notions of theme, politics, ethics, and aesthetics. And,
while they arrive at opposite views of these, they do seem to converge in a
two-part structure to the “Whoole Booke,” although they locate the dividing
point somewhat differently. They each have their characterizations, in literary
terms, of these two parts. The first section is episodic in method and sets out
to uphold Arthur as an ideal of knightliness, while the second part centers on
Launcelot and the flaws in Arthur and the codes which guide his doomed kingdom.
Choose your own doom.
Does it not seem at lest as relevant
to an account of the binary conception to inquire into the compositional
history of the Arthurian Cycle? This is the view taken by Larry Benson in Malory’s
Morte Darthur (1976), where he begins by sketching Malory’s certain
intertextual knowledge of at least four cycles -- the prose Lancelot (Lancelot,
Queste, Mort Artu ), and the expansion of that cycle into the
Vulgate Cycle; the cyclic version of the prose Tristan, and the Roman
du Graal, as well as the acquaintance he shows with the Perlesvaus,
whose author promises a continuation, presumably cyclic. The two-part
structure, which also governs Benson’s discussion, is therein these cycles. The
prose Lancelot is separated from the Estoire and the Merlin,
those retrospective sequels that grew out of it and the reworkings of Robert de
Boron. Bogdanow’s Roman du Graal, which unites the extension of the Merlin
and parts of the prose Tristan in a literarily-conceived New Arthuriad,
still exists as a two-part division of the evidence: the Cambridge and Huth MSS
contain the first part, and the Iberian Demandas contain the second.
Whether the Roman du Graal is platonically real or not, it does not
exist whole in any surviving medieval book, and Malory’s own bifurcated book
may be one consequence.
Benson utilizes this intertextual
knowledge, as well as the work of Bogdanow and Pickford, to arrive at what some
reviewers have found to be a judicious, less gloomy, and contextually sound
view of Malory. Benson sees Malory’s respect for chivalry as neither tragic nor
nostalgic, and he finds in the massive Tale of Tristram Malory’s best
work by fifteenth century standards. This view would accord with the
conclusions of Emmanuèle Baumgartner, Tatiana Fotich, and Renée Curtis, who
variously argue, in part on the basis of deeps study of the MSS, for the
evolution of a great work of fiction.
I could go on. Medieval culture
provides many models of discourse through which the events in the Arthurian
romances may be interpreted (or, in the various vocabularies of intertextual
study: recuperated, naturalized, motivated, or made vraisemblable). And recent
scholarship has explored many of these intertextual relations: romances have
been read against the texts of the Bible, the church fathers, the Latin
classics, the scholastic doctors, the law, chronicles, hagiography, personal
history, Hebrew and even Iranian texts. Some have even urged reading them in
the light of other romances. But my sense is that what has made some
medievalists especially alert to new intertextual possibilities is their
appreciation of what happens when books are begotten of books by busy and
loving hands.