Wessex Parallel WebTexts


about the project
what is mouvance?
case study
index of first lines


What is mouvance? 


London, British Library, Cotton MS Cleopatra C.vi, f. 4Editors working on the vernacular literatures of the Middle Ages may be faced with very different problems from the editors of classical works. 

Traditionally, the task of the editor of classical works has been seen as the examination and (where possible) comparison of surviving manuscripts to identify and eliminate those features of their texts which are scribal rather than authorial, in order to 'reverse the process of transmission and restore the words of the ancients as closely as possible to their original form' (Reynolds and Wilson (1974), p. 212).

For some medieval vernacular works, this approach is arguably appropriate. There is no doubt, for instance, that Chaucer in the late fourteenth century saw the verbal integrity of his work as important, and felt that it was threatened by the process of scribal transmission. Taking leave of his work ('Go, litel bok') at the end of Troilus and Criseyde, he hopes that the lack of standardization in the spoken and written English of his time will not erode its poetic quality in the course of transmission ---
And for ther is so gret diversite 
In Englissh and in writyng of oure tonge,
So prey I God that non myswrite the,
Ne the mysmetre for defaute of tonge ...

( Troilus and Criseyde, 5. 1793-6, ed. Benson (1988), p. 584)
--- and elsewhere he wishes a scalp infection on Adam, his personal scribe, if he doesn't improve the accuracy of his copying in future:
Adam scriveyn, if ever it thee bifalle
Boece [Boethius] or Troylus for to writen newe,
Under thy long lokkes thou most have the scalle
But [unless] after my makyng [composition] thow write more trewe.
So ofte aday I mot thy werk renewe,
It to correcte and eke to rubbe and scrape,
And al is thorugh thy negligence and rape
(Chaucers wordes unto Adam, his owne scriveyn, ed. Benson (1988), p. 650).

However, the textual transmission of some other medieval vernacular works suggests less concern for the textual integrity of the original work, and a less clearly-marked distinction between the functions of author and scribe. Cerquiglini (1989), p. 57, has argued that 'L'oeuvre littéraire, au Moyen Age, est une variable ... Qu'une main fut première, parfois, sans doute, importe moins que cette incessante récriture d'une oeuvre qui appartient à celui qui, de nouveau, la dispose et lui donne forme' ('The literary work, in the Middle Ages, is a variable ... The fact that one hand was the first is sometimes, undoubtedly, less significant than this constant rewriting of a work which belongs to whoever recasts it and gives it a new form').


Zumthor's concept of mouvance

The concept of mouvance was formulated by Paul Zumthor in the second chapter of his study of medieval French poetry, Essai de poétique médiévale (Zumthor, 1972). Zumthor noted the contrast between the relatively fixed texts found in manuscripts of the works of some named late-medieval French poets (Charles d'Orléans, Guillaume de Machaut) and the much more common medieval combination of authorial anonymity (or near-anonymity) and a high level of textual variation, which might involve not only modifications of dialect and wording but more substantial rewriting and the loss, replacement, or rearrangement of whole sections of a work.  He used the term mouvance to describe this textual mobility.

Zumthor argued that anonymity and textual variation were connected: medieval vernacular works were not normally regarded as the intellectual property of a single, named author, and might be indefinitely reworked by others, passing through a series of different 'états du texte' ('textual states') (p. 72). The modern emphasis on 'textual authenticity' (i.e. the attempt to reconstruct the author's original as the only authentic version of the text) was therefore anachronistic as an editorial approach, ignoring the 'mobilité essentielle du texte médiéval' ('the essential mobility of the medieval text', p. 71). To avoid this anachronism, our concept of the medieval 'work' (oeuvre) needed to be redefined.

Diagram: 'work' and 'text'

Zumthor: diagram of distinction between 'work' and 'text'

(based on the diagram in Zumthor (1972), p. 73)

Zumthor emphasised that the term 'work' (oeuvre) in this diagram should not be identified with the archetype in the traditional editorial diagram of manuscript relationships, the stemma. It represented not the historical antecedent of the surviving manuscripts, but 'l'unité complexe ... que constitue la collectivité des versions en manifestant la matérialité; la synthèse des signes employés par les "auteurs" successifs (chanteurs, récitants, copistes) et de la litteralité des textes ... L'oeuvre est fondamentalement mouvante'  (p. 73) ( 'the complex unity constituted by the collectivity of its material versions; the synthesis of the signs employed by the successive "authors" (singers, reciters, copyists) and of the literality of the texts .... The work is fundamentally mobile'). The 'work' was not static, a chronological starting-point for the process of manuscript transmission, but dynamic, passing in the course of its transmission through phases of growth, transformation, and decline.

Zumthor explained mouvance as a product of the oral culture of the Middle Ages, an 'intervocal'  (as opposed to 'intertextual') network offering access to a variety of possible resources for poetic composition; the different realizations of a 'work' reflected a continuing interaction between written and oral culture at each stage of transmission (see the section on 'Intervocalité et mouvance' in La lettre et la voix (Zumthor (1987)), pp. 160-8).


later developments

Some more recent textual theorists have explored similar issues, questioning the textual 'authority' of the medieval vernacular author and, more generally, examining the ways in which the creation of a 'work' might be seen as a communal process extending over time rather than than an individual act of literary production. 

Jerome J. McGann, for instance, argued in A Critique of Modern Textual Criticism (McGann (1983)) that even modern literary works 'are fundamentally social rather than personal or psychological products' (p. 43). 'The fully authoritative text is . . . always one which has been socially produced; as a result, the critical standard for what constitutes authoritativeness cannot rest with the author and his intentions alone' (p. 75). The role of printers, editors, even friends, in the production of successive stages of a literary work needs to be taken into account; and the printed version of an author's draft may offer opportunities not only for contamination but for decontamination ('Authors' works are are typically clearer and more accessible when they appear in print', p. 41).

Two books, one French, one English, can be used to illustrate approaches to the specific problems of editing medieval works ...


Cerquiglini, Éloge de la variante (1989)

Bernard Cerquiglini, in his brief, witty, and incisive 'critical history of philology', Éloge de la variante (Cerquiglini (1989)), offers a post-structuralist sequel to Zumthor's work. He avoids Zumthor's 'beau terme de mouvance' because of its association in Zumthor's work with oral culture, preferring the term variance (p. 120, n. 19). His emphasis is less on the relationship between oral and written than on the relationship between the varying written realizations of medieval vernacular works, and the implication of their variance for the medieval concept of textual authority. Like Zumthor, he sees this purposeful variation (the scribe's 'intervention consciente', p. 79) as intrinsic to the transmission of medieval Romance works. Cerquiglini argues that it is anachronistic to see works  of this kind as the intellectual property of a single author, textually fixed at the 'moment unique où la voix de l'auteur, que l'on suppose, se noua à la main du premier scribe, dictant la version authentique, première et originelle' (p. 58) ('[the] unique moment when the imagined voice of the author linked itself with the hand of the first scribe, dictating the first, authentic, and original version' ). In taking all manuscript variants as errors, the editors of medieval vernacular works have misunderstood their essential nature: 'Dans l'authenticité généralisée de l'oeuvre médiévale, la philologie n'a vu qu'une authenticité perdue' (p. 58) ('In the generalized authenticity of the medieval work, philology has seen only a lost authenticity').

Cerquiglini also questions (referring to the discussion of Kane and Donaldson's edition of Piers Plowman in Patterson(1985); see further Conclusions, below) the assumption in much textual criticism of an 'auteur transcendant ... [qui] tranche absolument, par l'unicité de sa conception, l'opacité de son oeuvre (argument de la lectio difficilior), la qualité de sa langue, avec la diversité scribale, ignorante et sans dessein, qui pluralise l'oeuvre, en banalise l'expression, appauvrit la langue' (pp. 90-1) (a transcendent author ... absolutely distinguished, by the unity of his conception, the opacity of his work (argument of the lectio difficilior [i.e.that the more difficult MS reading is likely to be the original one]), the quality of his language, from the scribal diversity, ignorant and unplanned, which pluralises the work, makes its expression banal, impoverishes its language').

He argues that not only traditional stemmatic editing but the 'best-text' method advocated by Bédier (1928), who preferred the conservative editing of a single MS text, assimilates the medieval work 'au texte autorisé, stable et clos, de la Modernité' ('to the authorised, stable and closed text of the modern era'). Stemmatic editing offers an 'illusory reconstruction' of the work, 'best-text' editing 'only snapshots' (p. 101); both marginalize (literally as well as metaphorically) its actual variance, relegating it, in selected and fragmented form, to the apparatus criticus at the foot of the page. Parallel-text editions are an inadequate solution to the problem; Cerquiglini sees information technology as the way forward, since the interactive, multidimensional space of the electronic edition can offer both more textual information than the printed edition and easier comparison of different versions of a work. The production of works electronically in itself mirrors medieval conditions of production more accurately than print technology: 'L'écrit électronique, par sa mobilité, reproduit l'oeuvre médiévale dans sa variance même' (p. 116) ('electronic writing, by its mobility, reproduces the medieval work in its actual variance' ).


Machan, Textual Criticism and Middle English Texts (1994)

A similar approach, applied to the editing of Middle English rather than Old French works, appears in Tim William Machan's Textual Criticism and Middle English Texts (Machan, (1994)). 

Machan argues that the editing of Middle English works has been dominated by a powerful 'Humanist' tradition of textual criticism, 'lexical' (seeing the work as essentially a verbal construct), 'idealist' (giving the authorially-intended work precedence over its specific documentary realizations), and equating the authorial with the authoritative text. The manuscript evidence, however, suggests that this approach is anachronistic. The work is characteristically treated as 'a nonlexical, not self-contained res inseparable from the supplements of others' (p. 165): its verba (words, rhymes, etc.) are a less important feature than its res (content), it may be modified or expanded during textual transmission, and its documentary realizations in manuscripts of varying content and layout further modify its meaning for the medieval reader. Textual authority in this period was normally the prerogative of the Latin auctores, and Machan argues that the evident anxiety of some late-medieval vernacular authors (including Chaucer) about the integrity of their texts suggests that they had not yet achieved this kind of authority

Machan suggests that a genuinely historical edition of a medieval work might entail reconstructing 'the work behind a document' rather than an 'authorized text' underlying the surviving documents (p. 184), taking into account the social and cultural framework within which it would have been read, and giving greater attention than at present to the bibliographical codes (e.g. page layout, choice of script, illumination) involved in its documentary realization.


three illustrations

Zumthor applied his concept of mouvance to specific types of medieval vernacular work, more or less closely linked to oral traditions: the chansons de geste, romances, lyric poetry, and the fabliaux. The fluidity of form and content that he noticed in this kind of work can, however, be found in other types of medieval literature (although not necessarily for the same reasons). The illustrations below offer a few specific Middle English examples: from popular romance (Sir Orfeo), religious lyric (see case-study), and works of spiritual instruction (Ancrene Wisse).

popular romance: Sir Orfeo

Most undergraduates studying Middle English encounter this charming short romance in the text preserved in the 'Auchinleck manuscript' (Edinburgh, Advocates' Library, MS 19. 2. 1), probably produced in London c. 1330. 

It tells the story of a harper-king, Orfeo, whose wife Heurodis is carried off by the king of the fairies. He leaves his kingdom in charge of a steward and goes into the wilderness to mourn her loss. After more than ten years of solitary hardship, he sees her riding through the wilderness with a troop of ladies; he follows her through a tunnel in the rock into a beautiful country with a castle built of gold and precious stones. He gains entry to the castle as a harper; there he finds people abducted by the fairies, who were 'thought dead, and are not'. He wins back his wife from the king of the fairies through his skill in harping, and returns to his court still dressed as a travelling minstrel; his steward does not recognize him, but welcomes him nevertheless for the king's sake. Orfeo then reveals his identity and resumes the throne; the faithful steward, overjoyed by his return, eventually succeeds him as king.

The Auchinleck version of this story, however, represents only a single stage in a sequence of narratives extending from classical times to the late nineteenth century, crossing geographical, cultural, linguistic, generic, and stylistic boundaries, and transmitted both orally and in writing. The earlier history of this sequence raises the question of authorship: how far can the anonymous poet who first put the story into Middle English verse be described as its 'author' in the modern sense? Its later history raises the question of authority: how far was the original Middle English version regarded as 'authoritative'? 

As the names of the characters suggest, one source of the Middle English romance was the classical legend of Orpheus and Eurydice. Before the narrative was turned into Middle English, however, it had already merged with legends of the Celtic underworld, probably in Brittany  (see Walter Map's late twelfth-century miscellany, De Nugis Curialium (ed. James (1983)), Dist. 2, ch. 13, and Dist. 4, Ch. 8, for a similar story of fairy abduction, set in Brittany). There are references in French romances of the twelfth and thirteenth centuries to a Breton lai of Orpheus; and the Middle English version is introduced, in two of the three manuscripts, by a prologue (lost in Auchinleck) linking it to the Breton lais. Internal evidence suggests that the Middle English version had as its immediate source an Old French narrative lai (now lost) in octosyllabic couplets. In other words, the Middle English version acknowledges its debt to earlier narrative, and probably owed much to the content and style of its French original (with the possible exception of the narrative motif of the faithful steward); should it be seen as an original work, a translation, or something between the two?

The Auchinleck MS is the earliest of three MSS containing the Middle English version; London, British Library, MS Harley 3810, dates from the early fifteenth century, Oxford, Bodleian Library, MS Ashmole 61, from the late fifteenth century. There are considerable divergences between the three texts, too numerous even to allow an edition in parallel columns (the standard edition, Bliss (1966), simply prints all three separately); in particular, the Ashmole 61 text has numerous omissions, particularly towards the end, probably the result of oral transmission. Even the Auchinleck MS does not necessarily reflect the earliest ME version exactly; its confident identification of Orpheus's original home, Thrace, with the ancient capital of England,
For Winchester was cleped tho
Traciens, withouten no
('For Winchester was then called Thrace, without a doubt')
is not shared by either of the later manuscripts.
A later version of the story still is printed by Bliss (1966), pp. l-li, the ballad of King Orfeo partially transcribed in Unst, Shetland, in the later nineteenth century: the king here is nameless, his wife is called Lady Isabel, and he wins her back by his skill on the bagpipes.
The surviving English versions of the story reflect, particularly in their reproduction of names, the degenerative process of 'Chinese whispers' which is an inherent risk of both written and oral transmission: Orfeo is said in the Auchinleck MS to be descended from 'King Pluto' and 'King Juno', in Harley 3810 from 'Sir Pilato' and 'Yno', and 'Dame Heurodis' in Auchinleck becomes 'Dame Meroudys' in Ashmole 61. But the textual variants are not entirely the product of scribal error; sometimes they reflect the more extensive textual modification caused by oral transmission (for a detailed study of the combination of memorization and recomposition reflected in some MSS of ME romances, see Baugh (1959)), sometimes purposeful adaptation for a different audience.


Medieval religious lyric

De Lisle Psalter, crucifixionDouglas Gray says of the Middle English devotional lyric, 'It was not in its own time a remote "aesthetic" literary form, but was an integral part of the religious life of contemporary society ... more often than not the impulse behind [the lyrics] is quite functional and practical' (Gray (1972), p. 37). 

The case-study provides six different texts of a meditation on the Passion linked to the visual image of the Crucifixion; an echo of the same work can be seen in a stanza of the longer Passion lyric in London, British Library, MS Harley 2253, I syke when Y singe, lines 41-50. Although clearly related, these short texts show a striking degree of variation in content, wording, and even length.

Although some medieval English devotional lyrics have named authors, this meditation is anonymous; its content is highly conventional, and the first-person speaker who appears in four of the six texts is less an authorial voice than a meditative persona for the poem's users. While the overall structure of the argument, and much of the poetic form, are retained in the individual versions, the high level of variation (whether caused by purposeful recasting for aesthetic or other reasons, memorial recomposition, or the rationalization of manuscript problems) suggests that the users and transmitters of the poem did not see its textual integrity as important.


Works of spiritual instruction: Ancrene Wisse

The two previous examples could be seen as illustrating specifically Zumthor's concept of mouvance: both involve works which, although they survive to us in written form, reflect the continuing influence of oral culture in their textual development. 

There may be other causes, however, for textual mobility in medieval works. The early-thirteenth-century Middle English guide for anchoresses, Ancrene Wisse, survives in a number of manuscript versions which sometimes differ very considerably from each other. The work was paraphrased, modernized, and translated into French (twice) and Latin; it was revised---both by its original author and by others---and adapted for different audiences; and its content was selected, rearranged, and incorporated into other devotional works.

The context of Ancrene Wisse  is written rather than oral (see Millett (1993)); although its author is anonymous, the authorial 'I' of the earliest versions seems to be a personal rather than an institutional 'I'; and some at least of its scribes seem to have taken considerable care to reproduce its text accurately (the late-fourteenth-century version in the 'Vernon manuscript' (Oxford, Bodleian Library, MS Eng. poet. A. 1) is one of the most reliable manuscript witnesses). 

London, British Library, Cotton MS Cleopatra C.vi, f. 4The key to its textual variation lies rather in its practical function as a work of spiritual instruction; although it may have been valued for its rhetorical skill, it was nevertheless pragmatically adapted, both by the author and by some of his successors, for changing audiences and changing purposes. This illustration of an early manuscript, London, British Library, Cotton Cleopatra C. vi, shows corrections (probably by the author himself) to the Cleopatra scribe's faulty text, but also his modifications (explanations and revisions) to the original version; a rather later manuscript, Cambridge, Corpus Christi College, MS 402, incorporates more extensive revisions---again probably authorial---for a larger group of anchoresses than the three addressed in the original version. This process of revision and adaptation continued through the Middle Ages, as the work was modified for the use of nuns, of male religious communities, and even of the laity.

The opposition in 'classical' textual criticism between author and scribe cannot be easily mapped on to this kind of transmission. Although there is plenty of evidence in the textual history of Ancrene Wisse (beginning with the Cleopatra MS) for scribal 'negligence and rape', the transmission of Ancrene Wisse seems to have taken place, particularly in the early stages, in an institutional context where the functions of author and scribe (as well as the intermediate functions of editor, reviser, and corrector) were not always sharply differentiated; a possible model for this context is offered by David d'Avray's study of the composition and transmission of mid-thirteenth-century Paris sermons, Medieval Marriage Sermons (d'Avray, 2001). 



The concepts of mouvance and variance have been received --- where they have been received at all --- with considerable suspicion by English textual critics; see, for instance, Nicolas Jacobs' criticism of Machan's approach (Jacobs, 1998), which questions the historical accuracy of the claim that for much of the Middle Ages there was no concept of 'authorship' for vernacular works, and emphasises the distinction between 'creative intelligence' and more low-level kinds of adaptation in the process of textual transmission.

These objections highlight two genuine problems. The first is the fondness of the theorists of mouvance for sweeping, sometimes untenable, generalizations; their approach may offer valuable insights into the textual transmission of certain types of medieval work, but is it really, either historically or methodologically, universally applicable?  The second is the question of textual quality: if  scribal rewritings (as Jacobs acidly remarks) are generally characterized by 'dim-wittedness, literal-mindedness, and triviality' (p. 5), is the editor obliged to renounce the concept of the 'transcendent author' and give them equal weight with authorial readings? Might it not be intellectually more respectable, and aesthetically more satisfying, to follow Kane and Donaldson's approach to Piers Plowman (see Kane and Donaldson (1975)), affirming the 'absolute difference' between scribal and authorial readings, and the need to eliminate the former from the edited text? Patterson elegantly (though not without irony) summarizes the Kane and Donaldson approach: 'The scribes are many, the poet unique; the scribes write the language of common men, the poet composes a language of his own. The poet traces no conventional path but works out for himself the way of genius, and it is the task of his editors to rediscover that way from among the ruins of the manuscripts' (Patterson (1985), p. 97).

The two problems are related: mouvance is more usefully seen as a significant aspect of medieval vernacular literary transmission than as its defining characteristic. The features which distinguish an oeuvre mouvante (authorial anonymity, collective rewriting, influence from oral tradition, textual changes for changing audiences or functions) are more likely to occur in some types of medieval work than others; and the qualitative distinction between an original 'authorial'  input and that of later contributors (scribes or others) to the textual transmission of a work may similarly be more marked in some contexts than in others. The study of mouvance is likely to be more rewarding (for instance) for an editor working  on thirteenth-century Middle English lyrics than for one working on a major late-medieval author.

The theory of mouvance, however, has more general implications for the editors of medieval vernacular works. Patterson, questioning Kane and Donaldson's exclusive concentration on the reconstruction of the author's original, nevertheless sees no alternative but the 'best-text' edition, which he argues is an abdication of the editor's responsibilities, 'an edition that, for all its conservative claims to soundness and reliability, in fact represents an arbitrarily foreclosed act of historical understanding' (Patterson (1985), p. 113). But the rehabilitation of the process of textual transmission implied by the concepts of mouvance and variance offers other possibilities. 

Cerquiglini's 'multidimensional' approach to editing proposes (as in the case-study offered here) simultaneous access to all the MS evidence, with no single 'privileged' text.  Cerquiglini's hope, however, that information technology would open up new possibilities for 'multidimensional editing' has not yet been fully realized. The 'years of grinding labour' (in Peter Robinson's phrase) required for multi-manuscript editions of even comparatively short works, and the difficulty of obtaining continuing funding for large projects, remain a problem. Most of the major electronic editions of medieval English works produced so far have been (in Housman's words) editions in usum editorum 'for the use of editors' --- manuscript facsimiles, transcriptions, and electronic collations (see the section on Electronic resources in the booklist), offering much fuller information on the raw materials for the editorial process than was previously possible, but not the user-friendliness or historical overview of a comprehensive edition.  

But other types of 'multidimensional' editing are possible. D'Avray's recent study of Paris marriage sermons (which might be significantly modified by different preachers, or for different purposes, in the course of transmission) offers a possible model (see d'Avray (2001), pp. 43-45): d'Avray chooses a single manuscript version as a representative of the textual tradition (not necessarily the earliest, and 'not a "best manuscript", though it would be perverse not to choose a good one', p. 40), and edits it in the light of the tradition as a whole, correcting where necessary, and recording significant textual variations and revisions in other MSS in the apparatus criticus or (if extensive) in an appendix.  The forthcoming EETS edition of Ancrene Wisse will follow a similar approach (see Millett (1994)), using Cambridge, Corpus Christi College, MS 402 not only as a relatively correct text incorporating some interesting revisions, but as a vantage-point from which the earlier and later development of the work can be surveyed.

Postscript: for a more recent discussion of the issues involved in editing 'Ancrene Wisse', see the EETS edition, in particular the section on 'Editorial Aims and Principles' in the 'Textual Introduction' to volume 1, pp. xlv-lxi:

Ancrene Wisse: A Corrected Edition of the Text in Cambridge, Corpus Christi College, MS 402, with Variants from Other Manuscripts, ed. by Bella Millett, drawing on the uncompleted edition by E. J. Dobson, with a Glossary and additional notes by Richard Dance, 2 vols., EETS 325, 326 (Oxford: Oxford University Press, 2005, 2006).



home | contents | search | top

Set up by Bella Millett, enm@soton.ac.uk. Last updated 09 September 2014 . Clip from London, British Library, Cotton MS Cleopatra C. vi, f. 4, and image from the De Lisle Psalter, London, British Library,  Arundel 83 II, f. 132r, reproduced by permission of the British Library; no further reproduction permitted.