Version classiqueVersion mobile

What is Authorial Philology?

Giulia Raboni
Paola Italia

4. European Examples

Marco Presotto, Sònia Boadas, Margherita Centenari, Francesco Feriozzi, Carmela Marranchino, Olga Beloborodova, Dirk Van Hulle et Pim Verhulst

Note de l’auteur

Unless otherwise stated, all the information found in this chapter is from the Jane Austen’s Fiction Manuscripts website. I would like to thank Kathryn Sutherland for her help with writing this chapter, and Carmela Marranchino for helping with the sample edition.

Texte intégral

4.1 Lope de Vega’s La Dama Boba

1Lope de Vega’s vast theatrical oeuvre is one of the most expansive bodies of texts from early modernity in Spain, and, for some time now, his more than four hundred plays have been the focus of the methodical analysis of philologists attempting to create a complete critical edition. The PROLOPE Project was founded by Alberto Blecua in 1989 to tackle this monumental undertaking, along with numerous efforts to improve the philological understanding of his literary legacy (​). The textual tradition consists of forty-four autograph comedies, many editions authorized by the playwright as part of his own project to publish his work, and also a wide selection of single editions and copies of all types, which drastically complicate the task of the editor. In addition to this, any plan to provide a critical edition must also consider the peculiarity of theatrical works of the Spanish Golden Age, intended primarily to be performed rather than read.

2According to the conventions of the time, the playwright, called poeta or ingenio, sold his original to the owner of the theatre company, who bought all rights to its use and could change the text at will to suit a performance in a given context. Once the theatrical run was over, the play’s manuscript could be sold to an editor to be printed, inevitably including all the changes implemented that the text had undergone ‘on stage’ during its life, including corrections by the theatre company manager, censor or others connected to the performing arts. The increased popularity of theatre and the rise in demand for plays made it necessary to develop an organized theatrical text editing system, especially for an acclaimed writer such as Lope de Vega, so that manuscripts could be produced quickly and according to the needs of the moment. Even if the documentation in this regard is unfortunately scarce, given the ephemeral nature of the intermediary steps, we can assume that the author generally composed his dramatic works with the following step-by-step writing method:

  1. script in prose, which tended to already be divided into acts;

  2. draft in verse, in which Lope de Vega transported the contents of the prose version, developed the poetic compositions (at least partially), and organized the polymetric structure of the work;

  3. clean copy destined for sale.

3However, this method resulted in numerous variants, especially because of the incessant changes brought about by the author, who was never satisfied with his results and always ready to update his texts, regardless of whether they were drafts or ‘final’ versions.

4The autograph manuscript of La Dama Boba, dated 28 April 1613 and now in the collection of the Biblioteca Nacional de España (Madrid) with shelfmark Vitr/7/5 (​viewer.vm?id=0000051826), is an interesting example of the playwright’s creative process. Despite being a carefully drafted copy, destined for a prestigious theatre company very close to Lope, many of its pages show the corrections and changes of the author. These markings often make it possible to reconstruct different phases of his writing process, even if they were added mainly while revising the final text. Lope perhaps had a script in prose, or more likely a draft in verse, and, in the act of making a clean copy, he re-wrote entire sequences from a structural and chiefly poetic and stylistic point of view, testifying to the author’s tireless creativity and constant perfectionism. In other words, Lope did not stop at copying what presumably appeared in the draft, but spent time improving and building upon the text even while transcribing it into a version fit for sale. It is thus interesting to reconstruct those changes in an attempt to retrace the steps taken in creating the text.

5The play came to be known over the following centuries exclusively via the text published by Lope de Vega himself in 1617, which is quite different from the autograph version because the author, by his own admission, could not use it while preparing the edition, having to fall back on a copy that evidently was flawed. Rudolph Schevill can be credited with publishing the first edition of the autograph manuscript. Schevill provided a diplomatic transcription in his The Dramatic Art of Lope de Vega, together with ‘La Dama Boba’ (1918) with an apparatus of variants and ample room dedicated to the changes made by the author. To do so, he included the ‘deleted’ fragments in the autograph that represent different creative phases, inserting them in parentheses in the edition. Starting from this publication, the autograph manuscript has always been used as a base-text for subsequent editions. A focus on the creative process is also seen in the text edited by Eduardo Juliá Martínez in 1935 (Lope De Vega 1935: 283–449), which includes a diplomatic transcription of crossed-out verses, without comments. Recent editions destined for the general public within popular series of Spanish classics often refer to this peculiarity of the textual tradition, even if it is a secondary aspect of the editorial project.

6The most up-to-date modern printed edition in terms of textual criticism is that by Marco Presotto, published as part of the PROLOPE Project (in Lope de Vega 2007: 1293–466). In the criteria of selection, Presotto has included a description of the characteristics of the manuscripts if considered important for the textual tradition. The system adopted is a symbolic one that refers to François Masai’s model (Masai 1950: 177–93), albeit with a few minor changes. Similar to previous academic editions, this criterion only makes it possible to report that which appears in the manuscript, and does not offer indications on the genesis of the corrections and the various writing phases. Although the diplomatic transcription of the corrections based on Masai’s system is reliable in that it leaves little room for interpretive errors, offering a direct description of what appears in the document, it ends up being an approach that is too cautious and out of tune with the work of a textual critic. After all, the job of a textual critic is to create a working hypothesis that connects all the information, as Gianfranco Contini’s definition reminds us. To overcome these limits and propose various hypotheses about the writing process, the same research group published a digital edition of the play in 2015 (​), as part of the creation of a digital archive for the textual tradition of the work. The transcription of the autograph manuscript includes an attempt to represent the corrections through different colours, the chronological numbering of individual phases and, where possible, interactive annotations that display the times the text was edited, listing them chronologically according to the editors’ hypothesis. In the end, the digital edition is undoubtedly an improvement compared to static printed texts, and XML-TEI encoding is a solid base for further developments. However, as the time of writing, it should be considered only partially adequate in terms of the way complex sequences are displayed. Indeed, the changes and corrections are not always easy to read or understand due to overlapping colours and a lack of uniformity in the display across different browsers, producing undesired issues even in the graphic layout. In this sense, Paola Italia and Giulia Raboni’s filologia d’autore model offers a rather interesting tool due to its greater stability.

7The example included here comes from the second act of La Dama Boba (Fig. 8), containing comic dialogue between the two main leading ladies.

8Only the hand of Lope de Vega appears and the ink is always the same, demonstrating that the text was largely conceived in its final version directly on the pages of the definitive copy. For comparison, the modernized version and the diplomatic notation apparatus using the system included in the PROLOPE edition appear below (for the digital edition:​aplicacion.html#). The same apparatus, but this time relating the genesis of the text according to the filologia d’autore model comes next, offering a detailed account of the creative process just as it appears in the manuscript. Given the complexity of the corrections described, however, philological notes are still necessary to provide readers with more information about the textual critic’s hypothesis. The result is quite satisfactory and undoubtedly innovative compared to previous models.

Fig. 8 La Dama Boba, 1613 (Vitr/7/5, f. 29r, num. 7, vv. 1422–1452),​viewer.vm?id=0000051826

4.2 Percy Bysshe Shelley’s Poems

9The vast majority of Percy Bysshe Shelley’s poems were published posthumously, or at least without his direct supervision, a fact that has complicated the work of critics and experts who study and publish the author’s texts, and continues to have a notable impact on Shelleyan philology. Of the more than 400 compositions that are found in complete collections today, only about 70 or so were printed in Shelley’s lifetime. These can be subdivided into verses found in editions and anthologies overseen by the poet (such as, in addition to the juvenilia, the 1813–1816 editions of Queen Mab; Alastor; Laon and Cythna, 1817 and 1818; The Cenci, 1819; Oedipus Tyrannus, 1820; Epipsychidion, 1821 and Adonais, the pastoral elegy composed upon the death of John Keats and edited in 1821, too), and those published in editions which Shelley did not so supervise, as they circulated during the long periods in which he was abroad (such as the collection Rosalind and Helen, 1819; Hellas, 1822; and the famous drama in verse, Prometheus Unbound, published first in 1820 with the addition of other poems, among which was the Ode to the West Wind).

10The gradual constitution of the author’s poetic canon is thus made up of different phases, and its complexity can be explained not only in relation to the adventurous lifestyle of Shelley and his circle, but also in light of the fluctuating critical appraisal of his writings throughout the nineteenth and twentieth centuries (Rossington and Schmid 2008) and in terms of the evolution of philology as a discipline across the Channel (Reiman 1972; Everest 1989; Fraistat 2000).

11Proceeding in order, the first such phase is the work done by the author’s second wife, Mary Wollstonecraft Shelley, on the unpublished autograph manuscripts left by the poet upon his sudden death (Shelley died in a shipwreck in the summer of 1822 off the coast of Viareggio, Italy). Despite the unwavering hostility of his family members, especially his father Sir Timothy Shelley, Mary Shelley immediately put her expertise to work on her husband’s tangled manuscripts, in preparation for the 1824 publication of Posthumous Poems of Percy Bysshe Shelley. Though printed in a 500-copy edition by John and Henry L. Hunt in London, it was partially withdrawn from the market upon the insistence of the author’s father. Later, Mary’s collaboration during the preparation of The Poetical Works of Coleridge, Shelley, and Keats, the first anthology published in Paris by the Galignani brothers in 1829, was not enough to stop other people from following in their footsteps, and for at least fifteen years, numerous pirated editions of the poems were published. After 1838, when some of those restrictions were withdrawn, Mary was decidedly freer to dedicate her time to the publication of her husband’s work. As such, in 1839 Edward Moxon managed to publish the four-volume Poetical Works of Percy Bysshe Shelley (dedicated to Percy Florence, the couple’s young son), enriched by Mary’s insightful editorial Notes and followed by two volumes of unpublished prose works in the same year.

12This 1839 edition (which strengthened the results of the one before it but which also inherited its alterations, such as changes to its form and organization, promotion to text of fragmentary lyric poems, etc.) was so successful that it became the starting point for almost all subsequent anthologies of Shelley’s poetry up until the twentieth century, with just two exceptions: Shelley Memorials (1859), published in London by Smith, Elder and in Boston by Ticknor and Fields, and Relics of Shelley edited by Richard Garnett (1862), both of which added new texts to the canon.

13In the meantime, the marriage of Sir Percy Florence Shelley and Jane Gibson St John resulted in the transfer of the family archive (guarded over by Mary Shelley until then) to Boscombe Manor, purchased by Percy Florence for his mother in 1849. It then became the home of Lady Jane, who — in particular after the death of her mother-in-law — dedicated herself at length and with authentic devotion to the memory (or rather worship) of her famous relative. She issued a series of biographies and memoirs, and perhaps most importantly managed the author’s papers, later even donating some of them to the Bodleian Library at Oxford.

14The new availability of some of the manuscripts originating from the family’s collection at the end of the century thus sparked a fresh wave of publications, which were, in a way, the continuation of those of Mary Shelley and the numerous projects that Lady Jane had worked on. The Victorian editions of the Poetical Works by William Michael Rossetti, brother of Dante Gabriel Rossetti and one of Garnett’s collaborators, and by Harry Buxton Forman are prime examples. The Rossetti edition came out in 1870, only to be re-edited with corrections in 1878, with extensive changes to the metre, diction, grammar and punctuation of the texts. The Buxton Forman edition, which appeared in 1876–1877, offered a significantly more conservative approach. It was, in essence, the first scholarly publication of Shelley’s poetry, not just because it scrubbed the verses of the numerous typographical/editorial corruptions accumulated over decades of unchecked dissemination and proposed a return to the author’s original versions, but also because it was based on an advanced knowledge of Shelley’s style and linguistic working methods (Buxton Forman 1876).

15However, even this effort was still limited by the lack of exploration of a large quantity of autograph materials — materials that, up to the post-war period, were almost hidden from the public. As a result, the quality of the collections and comments released throughout the first half of the twentieth century was negatively impacted (like the 1911 edition from Charles D. Locock, who had nevertheless access to some of the Boscombe originals in 1903, extracting many new readings that were gathered in Thomas Hutchinson’s Oxford Standard Authors Edition, the latter becoming the reference text for Shelley’s poems for quite a long time). The Julian Edition by Roger Ingpen and Walter E. Peck came next, the only one collecting all prose, poetry and letters together, released in ten volumes between 1926 and 1930. It became almost immediately obsolete due to the excessively eclectic choices made by the editors and the broader access to the manuscripts that had become possible in the meantime. For the first time since Mary Shelley’s efforts, the handwritten documents finally returned as the true protagonists in the ‘third phase’ of Shelleyan philology.

16Indeed, Lady Jane’s donation to the Bodleian Library (1893–1894) was followed by that of other materials, including some important notebooks, by Sir John Shelley-Rolls (in 1946), a direct descendant of the family upon the heirless passing of Percy Florence. More recently, in 2004, the library acquired the Abinger papers, which include many of Mary Shelley’s manuscripts and most of the author’s correspondence, thereby reuniting the entire Boscombe collection at Oxford. Today, these documents are available to researchers, as are the less numerous but no less significant papers scattered in collections outside of the United Kingdom, including those in the Huntington Library (California), the Library of Congress (Washington, DC), and the New York Public Library (where the rich Pforzheimer Collection is held).

17Given the ease of access to the sources in more recent years, textual criticism of Shelley’s work has seen rapid growth since the 1960s, all while in the absence of a complete critical edition of Shelley’s poetry (Rossington 2013; Rognoni 2018: cxxxvi-cxxxvii). The main reason for such a state of affairs is to be found in the very nature of the author’s manuscripts and in the complexity of the compositional mechanisms that Shelley’s papers reveal.

18In particular, the poet’s notebooks, about thirty in total, contain drafts and re-writings that constitute different editorial states of the same works, often incomplete, or developed alongside other compositions, or even dating back to moments that are difficult to determine because they were so far apart. One further problem then concerns the speed and conditions with which the very writing of the texts took place: Shelley mostly wrote while travelling, indoors and outdoors, sitting still or even in movement, perhaps on a boat or in a carriage. The result is a notable sense of compositional disorder expressed on pages handwritten with cursive ductus and penmanship that is so broken down it is almost indecipherable, with numerous erasures, corrections, interlinear insertions or re-writings of portions of the text, one overlapping the other. And because notebooks were his favourite place to jot down ideas, they often intermingle with poetic fragments or glosses, comments, titles, quotes, calculations or even sketches of trees, mountains, buildings and faces. Moreover, the sheets of the notebooks, bound along the upper margin, were used by the author from both ends, without apparent regard for their material proximity. Quite often, different drafts of the same composition, including some clean copies, are spread across one or more different writing supports. Lastly, these issues are compounded by those deriving from the mass dispersion (or destruction) of many of the indicator-links that would have connected the first drafts or revisions of the poems to their definitive publication in printed volumes as arranged by Shelley (autograph or apograph copies, corrected drafts, etc.); not to mention that the author’s practices included the frequent use of loose sheets of paper, many of which are now lost (Everest 1989: xxii–xxvi).

19Faced with such a complex situation, the greatest and longest lasting effort by Anglo-American scholars has therefore consisted in providing the public with facsimiles of Shelley’s manuscripts (and now, in many cases, digital scans, like those that can be found on​divisions/​carl-h-pforzheimer-collection-of-shelley-and-his-circle, or​digital/​collection/​p15150coll7/​search/​searchterm/​percy%20shelley). A robust forty-one volumes containing images, transcriptions and commentaries on specific parts of the text (though selective and based on widely differing standards) make up the three main diplomatic editions currently available to philologists (Shelley 1985–1997; Shelley 1986–2002; Shelley 1961–), almost all thanks to the efforts of Donald H. Reiman. These tools certainly facilitate the work to be done, as they make it possible for the reader to check the hypotheses formulated by critics for each text. However, for obvious reasons, they cannot be considered the result of a systematic, genetic-reconstructive study.

20Also dating to the 1960s, at least in its original conception coming from Geoffrey Matthews, is the first veritable critical edition, in the modern sense, of the writer’s poetry, The Poems of Shelley, a collection that was published as part of the Longman Annotated English Poets (1989–) series. Originally divided into three volumes, today it has reached five total, though the last has yet to be released. The collection is actually one of two complete editions of Shelley’s poems currently being worked on: other than the Longman edition, another series has been ongoing since 2000, The Complete Poetry of Percy Bysshe Shelley (CPPBS), published by Johns Hopkins University Press and divided into eight volumes, with only the first three having been printed.

21These editions deserve a closer look. However, at least a brief mention must also be made of the Norton, the Oxford World’s Classics and the Penguin Classics anthologies, alongside the Italian publication edited by Francesco Rognoni for Meridiani Mondadori (the largest selection of texts to date). Despite not being critical editions, they do constitute the main vehicles of the widespread dissemination of Shelley’s work in recent years (often complete with optimal philological information resulting from a round of manuscript verifications).

22In the Longman publication, the poems are published starting with a close-up examination of the handwritten and printed versions, and providing a systematic modernization of Shelley’s spellings. The texts are arranged in chronological order according to the first writings, thereby abandoning the order imposed by Mary Shelley’s Poetical Works. Each poem is preceded by a short introduction, which provides indications on the dates and the occasions of the drafts, up to the work’s publication, information on the available sources (bibliographic and otherwise) and, where necessary, on the work’s reception. The text is then printed and accompanied, at the bottom of the page, by an apparatus that includes both explanatory notes and comments and minimal lists of substantial variations taken from the autograph manuscripts and the most important printed publications (or re-issues). Contrary to that printed in the introduction to Volume 1 of the series (Everest 1989: xxvi), it should be mentioned that manuscript readings are not ‘given in full’. Instead, they mainly correspond to the last developmental stages that can be deduced from the work carried out by the author on the drafts. Formal and punctuation variants appear here and there, but only when deemed relevant to the way the text is understood. Thus, a large number of the variants found in the originals are missing. In addition, the intermingling of critical and philological levels certainly does not help the legibility of the apparatus, and the chronological order, the true innovation of the Longman version, obviously has its own set of issues: Shelley often worked on multiple texts at the same time, interspersing long compositional pauses (weeks, even months) with corrections made to that same text. Meanwhile, the fragments are, by their very nature, very difficult to date and the writings revised later on are placed in conventional positions, seeing as they are established only by their editorial forms.

23CPPBS, on the other hand, is quite different in its layout and textual criticism solutions, more clearly separating the handwritten text from the printed version, while reserving limited space for the former. Even the criteria that determine the presentation of the texts are different: here, the poems are first of all distinguished between published and unpublished. The published works are arranged in collections, according to the sequence of the author’s editions, while the unpublished works, private because they never went beyond the draft stage, or because they were circulated exclusively within a close circle of friends and never meant to be printed, are mostly grouped according to the moments in which the poet’s life can be sub-divided, spanning from 1803, the year of his lyrical poetry debut, to 1822. Within each section, then, the series follows either the desire of the author as shown by the reference editions, or a more generic chronological order of composition, particularly reliable when deduced from the available correspondence. However, the element that is actually new here is the base-text, which is a snapshot of one of its editions, conceived by the poet to be submitted to readers in a certain moment in time, and reworked as little as possible by the collection’s philologists who, unlike those at Longman, opted for strictly conservative criteria in regards to linguistic form and punctuation. The result is a four-level structure: text; apparatus at the bottom of the page, limited to the different printed versions and used to justify any corrections to the text; Primary Collations with the selective annotation of authorial variants and those found in printed editions (identified through the comparison with other Shelleyan editions or with Mary Shelley’s collections, which are supposed to preserve authorial variants); and Historical Collations including the variants that can be inferred from the most important nineteenth and twentieth century editions of poetry, helpful in tracing the historic fate of the texts. The picture is completed by a few supplementary sections, or proper appendices, where different materials appear occasionally: revisions for reprints, lists of errata, alternative versions of the same composition (as is the case, for instance, of the Mont Blanc poem, 1816, and others), or partial transcriptions of autograph fragments, including vertical apparatuses which are helpful for examination of the texts. As the American edition treats the poems published during Shelley’s life in Volumes 1–3, it has the benefit of systematically documenting the printed versions, though it does not handle the manuscripts in the same way. For this reason, the Longman version is still in the lead, though the supplements to the edition remain indispensable.

24The last twenty-five years seem to have marked a new era in the study of Shelley’s poetry. On the one hand, the current state of Shelleyan philology is firmly rooted in the belief that a close examination of the originals is an indispensable premise for each attempt at a critical study of the texts and, because of this the large diplomatic undertakings mentioned above are favourably received, as are those that today are multiplying in the landscape of the Digital Humanities, such as the​ online platform developed in 2015 by Neil Fraistat, Elizabeth Denlinger and Raffaele Viglianti, which will host the complete digitized versions of the manuscripts of both Mary and Percy Shelley in the future, in addition to those of her father, William Godwin, and her mother, Mary Wollstonecraft. On the other hand, it seems that the manuscripts struggle to become part of the apparatuses of the reference editions, both for practical (space and problems with finding a system of representation) and theorical reasons. The traditional dilemma (which Mary Shelley herself grappled with) between the need to respect the author’s last wishes, reflecting, in a ‘historical’ edition, the intentions that guided Shelley during each step of textual composition and revision, and the importance of guaranteeing readers a text that is comprehensible. The outcome is twofold: offering excessive freedom in altering mostly unfinished texts, and placing the available records in a hierarchal order which clearly prioritises the printed versions, thereby confirming a choice that is not exclusive to Shelley’s texts, but generally applied to Anglo-Saxon textual criticism practices, in which the field of Textual Bibliography and the critical models offered by Greg (Greg 1950) and Tanselle (Tanselle 1998) seem to dominate.

25To correct that situation, for some years now, especially among the scholars connected to the CPPBS, there have been a few attempts at filologia d’autore editions. It is worth mentioning them in conclusion: in 2012, an appendix to the third volume of the series was published, containing the genetic edition of the longest of Shelley’s poems, Laon and Cythna (1817, later revised and reprinted in 1818 with the title The Revolt of Islam). The text presented on paper, accompanied by printed variants and the main collations, is followed by a supplement relative to the corrections handed down by seven Bodleian autographs, plus other rough drafts and copies that make up the history of the work (six fragments from the Pforzheimer Collection and another six from the British Library, National Library of Scotland, Trinity College in Cambridge, Texas Christian University and the University of Texas, Austin; with the addition of six more prints), even if the introduction specifies: ‘this is not a complete record of draft variants; it does not include every repeated cancellation, stray letter, or indecipherable word or phrase in these MSS. But it does reproduce large sections of Shelley’s draft material for L & C in an attempt to trace his thought process and the evolution of his language, imagery, and political, social, and philosophical ideas’ (Neth 2012:​reference/​laon_cythna/​introduction.html).

26The online edition, explicitly meant to have an ancillary function with respect to the paper version does not reproduce the text of the poem, to which it refers only by adopting the same numbering of the cantos, stanzas and lines. The apparatus is vertical and explicit: the explanatory editorial interventions, in italics, are frequent, alternating with the record of the variants, mostly described topographically and followed by the complete list of the record abbreviations involved. Brackets indicate the portion of the text involved in the variant, but also conjectural transcriptions (with the addition of question marks for the indecipherable readings); underlining and deletions are rendered via marks respecting those of the author. Going from top to bottom corresponds to the passage between references, or they identify correcting phases within the manuscript itself (though here and there the apparatus is synthetic and implicit).

27The following example (in which the text has been taken from CPPBS: III, 133–34) is the ninth stanza of canto I (verses 73–81, containing the representation of an eagle and a serpent engaged in an allegorical struggle between revolution and oppression). It accounts for a solid argument for a panorama of research that is more developed than in the past and finally aimed at recovering the workshop-like aspect of the compositional process that, until now, has been partially overshadowed in the field of textual criticism.


A shaft of light upon its wings descended,

And every golden feather gleamed therein—

Feather and scale inextricably blended.


The Serpent’s mailed and many-coloured skin

Shone thro’ the plumes its coils were twined within

By many a swollen and knotted fold, and high

And far, the neck receding lithe and thin,

Sustained a crested head, which warily


Shifted and glanced before the Eagle’s stedfast eye.


entire line canceled in pencil Bod2

shaft] written above gleam and beam Bod2

its] it’s Bod4

descended, ] descended Bod2

descended Bod4


line preceded by From that T above [? The] things Bod2 therein—] therein Bod2


Feather] after The eagles Bod2 scale] scale, Bod4


thro’] thro Bod2

through 1839 1840

plumes] wings Bod2

plumes; 1834 1839 1840

twined] below wreathed Bod2


entire line preceded by Voluminously Even as a waterfall among the woods Bod2

By] In Bod2

swoln and] swoln & in pencil below fold voluminous canceled in pencil Bod2

knotted fold, ] knotted fold below gathered fold all in pencil Bod2 and high] & wide high Bod2


stedfast] steadfast above [? ing] Bod2

steadfast 1839 1840


Around, ] Around Bod4

around, ] around Bod2 Bod4

circles] revolutions above circles Bod2


clang] ing of clanging written over separate word of scream, ] shrieks above scream Bod2

Eagle] eagle Bod2

Fig 9 Laon and Cythna, Bodleian MS. Shelley adds. e.19, f. 14 (see Bod2; published in BSM, XIII, p. 32)

4.3 Jane Austen’s The Watsons

28Jane Austen’s fiction manuscripts represent one of the earliest surviving conspicuous dossiers of materials for a British novelist, covering a large time span in Austen’s life, roughly from 1787, when she was eleven, to 1817. Since all these manuscripts, be they drafts or fair copies, testify to works that were never published (and in the case of Persuasion to an alternative, unpublished, ending), it is widely believed that the author routinely destroyed the manuscripts of her published novels once they were printed, as was common practice in the period. Austen’s manuscripts underwent two major dispersals, one in 1845 when at the death of the author’s sister they were dispersed among surviving family members, and another in the 1920s when they began to enter auction houses to be divided among multiple British public and private collections.

29These manuscripts have been the object of a digital edition directed by Professor Kathryn Sutherland, Jane Austen’s Fiction Manuscripts (​), designed to reunite, order and preserve the materials. The website edits all the manuscripts respecting the author’s use down to her graphic usage, with only minor normalization as regards punctuation. The reproduction of all the manuscript pages can be seen side-by-side with the corresponding transcription, so that the user can at all times double-check the readings of the edition or recover those graphic elements that inevitably get lost in the transcription. While the edition is strongly diplomatic, the chronological succession of the corrections is included as well. The transcription is indeed faithful to the page, reproducing erasures, interlinear corrections and carets as they are on the page; however, in cases of particularly complex, multi-layered corrections, the user can hover with the cursor on the relevant correction to ‘reveal’ the earlier phases, and many of the corrections are accompanied by footnotes that reconstruct particularly difficult elaborations of the manuscript in a fashion akin to that of authorial philology. Sutherland’s team indeed included individuals with an Italian/European philological background, and she herself cultivates philological interests, and therefore produced an edition in which material data and interpretation are productively intertwined.

30This interpretative element is even more apparent in the printed version of the same edition (Austen 2018), where a number of new textual notes have been added. The printed edition was later incorporated in the Oxford Scholarly Editions Online (OSEO) portal. OSEO brings the edition back to the digital realm and overcomes a technical limit of the original Fiction Manuscripts website, now ten years old and partially reliant on discontinued Adobe Flash software, with the consequence that parts of it might soon become inaccessible. While this edition is extremely commendable and is from many points of view superior to the original Fiction Manuscripts, the passage to a physical edition sacrificed much of the astute interactive presentation of the original project, reverting it to a more ‘classic’ edition, despite preserving the synoptic presentation of the website. The evolution of technology is indeed a problem that one should keep in mind when working on digital editions, as all technical solutions used are bound to become dated. Because of this, one either has to keep their project constantly updated from the technical point of view (which can however be quite difficult and costly), or to also publish a physical edition to ensure its long-term usability at the cost of the benefits of the digital form. However, this should not discourage one from attempting digital editions altogether, especially considering how newer technologies such as HTML5 were conceived with the specific intention of avoiding drastic changes such as the discontinuation of Flash software.

31Both the original Fiction Manuscripts and the Scholarly Editions Online version are nevertheless an excellent example of the multiplicity of perspectives offered by a digital edition. Editions such as this have been defined as ‘paradigmatic’ by Elena Pierazzo, Technical Research Associate of the project (2014: 4–5), with reference to the ‘paradigmatic axis’ of possible views from which the user is able to choose.

32For this example, we are examining a draft to which Austen had not given a title, and which was named The Watsons by one of its first editors, James Edward Austen-Leigh. The 17,500-words long fragment was written around 1804–1805 in Bath, after Pride and Prejudice, Sense and Sensibility and Northanger Abbey had already been drafted. The Watsons is a case of codex unicus: the only witness of the work is a sequence of forty-four-leaves distributed across twelve small homemade booklets which were split into two parts in 1915 when the first six leaves were sold at a charity sale. The first six leaves are now held in New York (Morgan Library & Museum MS. MA 1034), while the rest is in Oxford (Bodleian Library MS. Eng. e. 3764). Some time after 2005, while still in private ownership, the second booklet of the manuscript (eight pages) went missing, so that the pictures found on the Fiction Manuscripts website and on its printed equivalent remain the only witnesses of those pages (Austen 2018: 1).

33The manuscript pages are written in a neat hand covering the entirety of the page and leaving little room for large-scale revision other than lines of extra text crammed into narrow interlinear spaces. For more extensive revision (occurring in three parts of the manuscript), probably as part of a single correctional campaign, Austen applied separate ‘patches’ of a different paper which she carefully cut and pinned to the sheets where the correction/addition was intended to be placed.

34The text is the beginning of a story, either a novella or, more likely, a longer novel, which was never finished, but whose narrative outline was known to Austen’s sister (Austen 2018: 5–6). The author perhaps anticipated later division into chapters. As it stands, the text is separated in the draft only by lines or by wider spacing. Virginia Woolf considered The Watsons a model of Austen’s writing strategy, believing its dry style to be due to her habit of writing the text in a bare, factual fashion, and revising the sentences later to ‘cover them with flesh and atmosphere’, not unlike what Gadda used to do (see section 3.6). According to Sutherland, Woolf’s remark is true, but in The Watsons’ case, Austen also did the opposite — when reviewing the text, she used to ‘remove the flesh and on occasion expose the bones’, not unlike what she later did when revising Sanditon (Sutherland 2005: 140). It might be that it was the almost cynical tone adopted in the draft that led Austen to interrupt her work on the novel. Other theories have however been proposed. According to her niece, Fanny Caroline, the reason for giving up on The Watsons was the sudden death of the writer’s father; Austen-Leigh instead believes that it was due to disgust for the excess of social degeneracy that the continuation of the novel would have depicted; and Sutherland also advances the hypothesis that the cause might simply have been discouragement for the fact that despite having sent Northanger Abbey to the publisher more than a year before, it had not been printed yet (Sutherland 2005: 129–30). However, The Watsons was not disowned in its entirety, as materials from it were re-used in later novels (Sutherland 2005: 147).

35The Watsons has been edited as part of longer literary works twice. Austen’s niece Catherine Hubback indeed based the first five chapters of her novel The Younger Sister (1850) on the incomplete novel, while an apocryphal completion, which included the entirety of Austen’s text with reworked punctuation, was written and published in 1928 by Hubback’s granddaughter Edith Brown in association with her husband. The text was instead published by itself in the second edition of James Edward Austen-Leigh’s Memoir of Jane Austen (1871: 297–364), and was edited again by the scholar Robert William Chapman in 1927. Both editions however normalized spellings and did not account for the corrections, with only Chapman giving a brief explanation of the characteristics of the manuscript in an appendix (Pierazzo 2016: 14–15).

36A good example of how Austen worked in revision is in the evolution of the description of Mr Edwards’s townhouse. Here is how it appears in the Fiction Manuscripts website (​manuscripts/​qmwats/​b2-3.html).

37The diplomatic nature of the edition is clear from the example, where one can find even purely graphical characteristics such as the long s (ſ) that was a normal feature of handwriting at the time. An element of interpretation can however be found in the second note, where the editors give an account of the order of the overlapping corrections within the same sentence.

38Here we will propose a specimen of how the same passage would look in an edition that follows the principles of authorial philology, to produce which we will also use the information found in the footnotes of the 2018 edition, using the latest text as base-text and preserving its graphic particularities even down to grammatical errors (cf. ‘Mr. E.s House’ instead of ‘Mr. E.’s House’). We will follow the representational criteria detailed in sections 2.5 and 2.6, but to make it easier for the reader to identify the many implicated variants found in the passage, I will represent them using a smaller font, as the Colli-Italia-Raboni edition of Manzoni’s Fermo e Lucia does (see section 3.4):

[b 2-2] The old Mare trotted heavily on, wanting no direction of the reins to take the right Turning, & making only one Blunder, in proposing to stop at the Milleners [b 2-3] before she drew up towards Mr. Edward’s door. — Mr.E. lived in the best house in the Street, & the best in the place, if Mr. Tomlinson the Banker might be indulged in calling his newly erected House at the end of the Town with a shrubbery & sweep in the Country. — Mr.E.s House was higher than most of its neighbours with windows on each side the door, the windows guarded by posts & chain the door approached by a flight of stone steps. — “Here we are — said Eliz: — as the Carriage ceased moving — safely arrived; – & by the Market Clock, we have been only five & thirty minutes coming. —which I think is doing pretty well, tho’ it would be nothing for Penelope. — Is not it a nice Town? — The Edwards’ have a noble house you see, & They live quite in stile. The door will be opened by a Man in Livery with a powder’d head, I can tell you.”

[b 2-2] heavily] written over stupidly [b 2-3] the place… Banker]1the town, if the Banker from which2T his… Country.]1his new House at the end of the Town in the Country, which however was not often granted. from which2T {newly… House] from newly erected one House} was higher … steps.]1was of a dull brick colour, & an high Elevation –aa flight of stone steps to the Door, & two windowsba flight of stone steps with white posts, & a chain, divided by a flight of stone steps.2was higher than most of its neighbours with two windows on each side the door, & five the windows guarded by a chain & green posts & chain, the door approached by a flight of stone steps. from which T nice] written over prett<y> They live… in stile.] from They live quite in stile I assure you with] written over &

39The edition that we have attempted here gives an idea of the advantages and disadvantages of editing the text using the methods of authorial philology — the chronological information that can be derived from the manuscript is made explicit at the cost of most information on topography, while a readable text is established without the arbitrariness of the old editions, and without sacrificing the strata of corrections that the manuscript attests. Thanks to the possibilities offered by paradigmatic (digital) editions, the reader could potentially be able to determine the level of interpretation s/he desires to see on his/her screen, from a minimum (photographs), to the genetic edition of the manuscript page, to a philologically-established text with apparatus. In this way, the materiality and spatiality of the page is both preserved and transcended at the same time.

Fig. 10 Jane Austen, The Watsons, 1804–1805 (Oxford, Bodleian Library, MS. Eng. e. 3764, b.2-3),​manuscripts/​qmwats/​b2-3.html

4.4 Marcel Proust’s À la recherche du temps perdu

40The analysis of Proust’s manuscripts, which has steadily increased since 1962 (when the Bibliothèque nationale de France (BnF) purchased a few autographs that had been in the possession of his heirs for years), is an exemplary case of French critique génétique (see section 1.4) in action. At the same time, the reconstruction of the history of the text of À la recherche du temps perdu has resulted in a fruitful, constant, theoretical and methodological reflection within the discipline, including recent digital developments. It is exemplary due to the complexity of Proust’s modus operandi; the centrifugal structure of the novel that takes place through a proliferation of textual units without following a firmly established plan; the rich, varied handwritten documents; and even the problems surrounding its publishing, partially imposed by Proust (up to his death on 17 November 1922) and partially compromised posthumously by the arbitrary ‘restoration’ done by his heirs and publishers.

41In 1912, Proust submitted the typewritten documents of the first part of a novel titled Les intermittences du coeur (according to the original plan) to various publishers, dividing it into two parts: Le temps perdu and Le temps retrouvé. After being rejected by Fasquelle, by André Gide on behalf of Gallimard’s Nouvelle Revue Française (NRF), and by Ollendorf, the first volume appeared with the title Du côté de chez Swann, published by Bernard Grasset at the expense of the author in 1913. Although it was to include another two volumes, Le côté de Guermantes and Le temps retrouvé, their publication was interrupted by the outbreak of WWI. Publication was resumed only in 1919 by NRF, which released a new edition of Du côté de chez Swann while the author was still alive (which varied from Grasset’s version), and the books À l’ombre des jeunes filles en fleurs (1919), Le côté de Guermantes I and II (1920–1921), and Sodome et Gomorrhe I and II (1921–1922). La prisonnière, the correct typewritten version of which Proust managed to send to Gallimard in time, was issued posthumously, with the subtitle Sodome et Gomorrhe III (1923). Other posthumous publications included Albertine disparue (1925) and Le temps retrouvé (1927), edited by Proust’s brother Robert and by scholars Jacques Rivière and Jean Paulhan who, worried about ensuring the text was comprehensible, presented a contradiction-free follow-up novel as if it had been completed by Proust himself, all while quietly implementing their heavy-handed interventions behind the scenes.

42This textual arrangement was canonized by the subsequent editions published by Gallimard in the Bibliothèque de la Pléiade. The first was issued in three volumes edited by Pierre Clarac and André Ferré (1954), with the second-to-last part of the novel maintaining the original title of La fugitive, which had been previously changed by Proust to Albertine disparue to avoid confusion with the equally-titled translation of a work by Rabindranath Tagore performed by Madame Brimont. The second was issued in four volumes edited by Jean-Yves Tadié (1987–1989) that were enriched by a series of esquisses, that is, a selection of old drafts extracted from the Cahiers manuscripts offered as a simplified transcription. However, the massive editorial reworking of the posthumous volumes of La recherche became clear in 1986, upon access to the typescript of Albertine disparue, corrected by Proust just before his death and intentionally ignored by the early editors. It was then published in 1987 by Nathalie Mauriac Dyer and Étienne Wolff. The numerous substantial variants that Proust added to the text of Albertine disparue, in particular the removal of 250 pages, surpass the last version attested to by the Cahiers manuscripts (Cahiers de mis au net XII–XV), which the Gallimard editions were based on. They also create undeniable narrative disjunction with the subsequent Temps retrouvé, highlighting the unfinished nature of the novel (Mauriac Dyer 2005). By demonstrating the arbitrary nature of the text in the posthumous volumes, this discovery has caused an identity crisis for the entire published corpus of La recherche, previously untouchable on its lofty pedestal, and has fed into the debate between genetic and textual criticism on the handling of an incomplete, partially posthumous work. The different editions are representative of different positions in the debate: in editing the 1992 Champion edition, Jean Milly published the long version of Albertine disparue, typographically differentiating Proust’s subsequent changes; on the other hand, in order to adhere as much as possible to the state in which Proust left the novel when he died, Mauriac Dyer suggested publishing — after the last volume that appeared in the author’s life, Sodome et Gomorrhe II — the typewritten documents of La prisonnière and Albertine disparue corrected in 1922, which were to be parts of a Sodome et Gomorrhe III, followed by their complete preliminary texts, i.e., the Cahiers di mis au net VIII–XV, and by the last notebooks in the series, XVI–XX (Mauriac Dyer 2007). In 2017, for the Classiques Garnier edition, Luc Fraisse published the long version taken from the Cahiers manuscripts, providing the variants in a critical apparatus. In addition, by making note of the changes imposed by the first publishers, Fraisse provided a more complete analysis of Jean Paulhan’s correcting approach, often aimed at flattening the lexicon and Proustian syntax in favour of greater clarity.

43In the 1970s, the formation of the Proust team at the Institut des Textes et Manuscrits Modernes (ITEM) led to the steady increase in genetic studies on the Proustian corpus, facilitated by the acquisition of the author’s papers by the BnF. After the 1962 purchase of eighty-two notebooks owned by the author’s niece, Suzy Mante-Proust (who also gave the library other papers in 1977), in 1983 the BnF acquired thirteen notebooks from the collection of Jacques Guérin (Fau 2013: 135–36). Among those notebooks, a few stand out: first, a series of seventy-five pieces containing the drafts of Contre Sainte-Beuve and of La recherche written between 1908 and 1922 (Cahiers de brouillon), and second, a series of twenty notebooks of clean copies that had often been reworked (Cahiers de mise au net), marked with Roman numerals I–XX by Proust himself and enriched by the famous paperoles, sheets of paper glued to the original pages (Contini 1947). This second series is datable between 1915 and 1922 and contains the text of the second part of La recherche, from Sodome et Gomorrhe onward. The collection, digitized on Gallica and freely consultable (, also includes jotter-notebooks, typescripts often made with the involvement of assistants and servants (Brydges 1984 and Pugh 2000), printing proofs with the corrections of the author and others, and loose pieces of paper.

44The dynamism of Proust’s writing is reflected in the magmatic appearance of the autographs. First Proust used the recto of the paper, then proceeded to eliminate and add text, which he initially inserts between the lines, then in the margins and on the verso, resorting to the paperoles for particularly long additions exceeding the material limits of the paper. Sometimes additions were connected to the text with reference symbols ranging from simple to complex (crosses, circles, flowers, a face in profile, etc.). These variations should be distinguished from the author’s notes written by Proust on the lines of the paper (interrupting the narrative thread), in the margins, or more frequently on the verso of the paper. They contain critical and aesthetic musings, programmatic instructions, and references, which can be written at the same time as the words on the page (as is the case with the notes on the lines of the paper), or derive from a subsequent re-reading (Herschberg Pierrot 2007). It is thus essential to define the systematic relationships that are established between the text, corrections, and author’s notes so that we may identify different layers of variants (of which, however, a diplomatic approach does not offer diachronic representation, contrary to authorial philology). The cross-references between one notebook and another, demonstrating the tightly-woven interdependence of different documents in Proustian working methods, are quite significant, characterized by constant re-writing and re-reading that spans multiple notebooks, where similar or even diametrically opposed narrative nuclei are reworked more than once, without being dated or numbered, until the novel is fine-tuned, which happens more on the level of the typewritten documents and the printing proofs (Brun 2011). Many notebooks, then, are piecemeal, with numerous pages cut entirely or in part and reused by Proust to save time when transcribing passages that had slowly taken shape over time.

45A systematic plan to restore Proust’s notebooks was launched in 2008 with the critical-genetic edition of Cahiers 1 à 75 de la Bibliothèque Nationale de France overseen by Nathalie Mauriac Dyer and co-published by BnF and Brepols. Two volumes are dedicated to each notebook: the first contains a facsimile of the manuscript and useful ‘diagrams of textual units’ that indicate in numbered boxes the blocks of primary text (1, 2, 3, etc.), additions (1A, 1B, 1C, etc.) and Proust’s notes (nr., in French notes de régie), with the scope of helping readers trace the process of writing the manuscript and broadly reconstruct the succession of narrative sequences, without however pinpointing the exact chronology of the variants (Mauriac Dyer 2008a: 159–60). The second volume, on the other hand, contains the transcription of the manuscript. By highlighting the limits of a linear transcription, considered unable to convey the multiple layers of Proust’s writing, often difficult to arrange in a hierarchy or set to a precise timeline, the recent publishers have adopted a diplomatic transcription that respects the layout of the words on the page. The rendering of the autograph manuscript is conservative, respecting Proustian spelling and punctuation. Corrections and annotations have been signalled with special diacritical marks: minimal interlinear additions are placed between angle brackets (< >); first and second degree crossed-out text has been indicated by single or double strikethroughs, or by x’s, depending on the author’s markings; when a word has been written over another or if a variant takes up part of the previous reading, the two readings are separated by a forward slash (/); an asterisk (*) indicates a hypothetical interpretation; square brackets ( []) mark the input of the publisher. The transcription is correlated by historical/critical, codicological and genetic notes, which mark the sections derived from previous Cahiers or continued within the same notebook (Mauriac Dyer 2001–2002).

46Special attention was paid to the reconstruction of the original physiognomy of the notebooks, from which Proust ripped out parts of and even entire pages, reusing them elsewhere. The research done on the paper fragments was based on a careful analysis of their physical qualities, such as the sizes of the fragment and the shape of the tear, the type of paper, the ink, the ductus, and even the textual continuity that can be detected between the various fragments (Mauriac Dyer 2008b: 100–02). A project of this scope is incredibly demanding, considering the extensive dispersion that the ripped-out pieces have been subjected to, as shown by the extreme case of the manuscript of À l’ombre des jeunes filles en fleurs, spread among fifty copies of a luxury edition (Wise 2003).

47The following example of a genetic edition is that of Cahier 46, datable to 1914–1915 (André 2009). In the first part of the notebook, Proust introduces the Albertine character to the narrator’s worldly Parisian setting, while, starting on folio 57r, he outlines the events of the Deuxième séjour à Balbec. The following excerpt offers a brief portrait of Albertine, transcribed from Cahier 46, NAF 16686, 52v (​ark:/12148/​btv1b6000131k/​f59.item.r=NAF%2016686):

48The diplomatic transcription conveys the placement of the words on the paper while also making the autograph more accessible and legible, to the detriment of the chronological reconstruction of the editing process. It is worth mentioning that the Proustian example is particularly complex, and formalizing Proust’s multi-layered process according to the principles of authorial philology could be too burdensome and unmanageable on the whole. Diachronic formalization could however be applied to single passages with the benefit of providing a more accurate reconstruction. For isolated segments, it would also be useful to reference, each time, the final text with the aim of facilitating a comparison that would be both orderly and in order, as the critical editions of the individual Cahiers are published.

49As an example, the previous passage of Cahier 46, revised in the manuscript, is presented below according to the principles of authorial philology. The last reading confirmed by the manuscript appears first and, in the apparatus, the variants formalized by the different compositional phases:

Il n’était pas pâle comme il semblait quelque fois dehors mais comme le jour où j’avais voulu l’embrasser paraissait un beau globe rose, tant un sang vif et clair transparaissait sous sa peau vernie, tant les pentes de ses joues étaient courbes et douces. Le regard glissait sur elles et sur son front jusqu’aux premiers contreforts de ses beaux cheveux noirs naturellement soulevés qui ici saillaient en massifs, là se creusaient en anfractuosités sans qu’en haut cessât de se poursuivre la chaîne ondulée et ininterrompue de leurs crêtes.

Il n’était… douces]1Elle n’était pas pâle comme elle semblait quelque fois dehors; un sang vif col<orait> et clair colora<it> transparaissait dans ses joues lisses; sa figure tout entie<re> →2Il n’était pas pâle comme il semblait quelque fois dehors mais rosi de ce rose do<> verni et t<> ˹d’une matinée (from des matinées) d’hiver partiellement ensoleillée ˹qui m’avait (from que j’a<vait>) tellement tenté à Balbec (with mais… Balbec not deleted) →3Il n’était pas pâle comme il semblait quelque fois dehors maisasemblait un beau globe rosebcomme il m’avait paru le jour où j’avais voulu l’embrasser comme le jour où j’avais voulu l’embrasser paraissait un beau globe rose, tant un sang vif et clair transparaissait sous ses joues vernies, tant les pentes de son visage étaient courbes et douces →3T{sa peau vernie] from ses joues vernies ses joues] from son visage}
Le regard] before Et Et elles venaient expirer et… front] interlinear insertion naturellement… crêtes]1massés dont les ondulations soulevaient comme des chaî<nes> →2que leur ondulation soulevait, creusait, relevait en massifs une chaîne de massifs onduleux →3naturelle<ment> naturellement soulevés, creusés, relevés en une chaîne de massifs onduleux →4T{ici] inserted en massifs … crêtes] inserted in the left margin là] inserted sans] before t< > ondulée] before in< > ininterrompue] before ind< >}

50For reference, the definitive text in Le côté de Guermantes II is below:

J’aurais bien voulu, avant de l’embrasser, pouvoir la remplir à nouveau du mystère qu’elle avait pour moi sur la plage, avant que je la connusse, retrouver en elle le pays où elle avait vécu auparavant; à sa place du moins, si je ne le connaissais pas, je pouvais insinuer tous les souvenirs de notre vie à Balbec, le bruit du flot déferlant sous ma fenêtre, les cris des enfants. Mais en laissant mon regard glisser sur le beau globe rose de ses joues, dont les surfaces doucement incurvées venaient mourir aux pieds des premiers plissements de ses beaux cheveux noirs qui couraient en chaînes mouvementées, soulevaient leurs contreforts escarpés et modelaient les ondulations de leurs vallées, je dus me dire […]

51As can be seen in the example, this method aims to create a legible body of text of which the entire process of textual correction is recorded in the apparatus, in this case genetic, according to a diachronic and systematic approach that gives priority to the chronology of the markings rather than their layout. This alternate formalization could, perhaps, constitute a useful addition to the model used for the French genetic edition, providing the reader with a truly complete image of the document that doesn’t exclude a priori an amplified interpretive intervention by the philologist.

52Using folios 46v-49r from the same Cahier 46, Julie André and Elena Pierazzo have prepared the Proust Prototype, the first prototype of a genetic digital edition of La recherche according to the XML-TEI standard (​proust_prototype). Pages are displayed in pairs, with the verso of each folio next to the recto of the next folio, where the textual units are grouped into numbered zones distinct from Proust’s instructions and notes, following a layout that echoes the ‘diagrams of textual units’ of the genetic editions on paper. By clicking on the image, the superimposed diplomatic transcription appears. The user can read the manuscript by following the order in which the different textual sequences were written (writing order) or according to the order of the manuscript’s final version (reading order). The greater or lesser degree of certainty of the position of a zone within the sequence is indicated by varying degrees of chromatic intensity of the background of the transcribed zones: the darker the colour of an area, the more hypothetical its place in the sequence (André and Pierazzo 2013; André 2016).

53The first complete digital edition of a Proustian document, on the other hand, was created in 2015 by Nathalie Mauriac Dyer, Françoise Leriche, Pyra Wise and Guillaume Fau. It is the edition of the Agenda 1906 — a deep red leather Kirby, Beard & Co. day planner for the first quarter of 1906, though Proust used it only later on — that became part of the BnF collection in 2013 (​editionsbnf/​ 1457 ).

54Of the 80 pages that make up the Agenda, only 29 include notes by Proust, some private (concerning, for example, his spying on Agostinelli around Paris) and some relating to work. The latter were made at different times and for the most part a posteriori, and focus on specific, historic, botanical, medical, and other matters that required verification and follow up. According to their chronology and their particular purpose, documentary and scheduling notes can be identified, datable between late spring and/or summer of 1909; verification notes taken on the occasion of the clean copy of Combray which are from late summer/early autumn of 1909; and verification notes and reminders jotted down when correcting the drafts of Du côté de chez Swann in June/July 1913. The diplomatic transcription is followed by a conservative linear transcription linked to the genetic and critical notes that, where necessary (exploiting the connections that define the digital environment), refer to the reproductions of all the cited Proust collection documents, or of other digitized archives (Wise 2017, Leriche 2016).

55In the digital environment, it is worth mentioning the Corr-Proust project, which launched the digital publication of the writer’s letters under the direction of ITEM, the Université Grenoble Alpes and the University of Illinois at Urbana-Champaign, where most of Proust’s letters are kept ( The digital interface contains a two-columns display divided in different tabs that can be viewed side by side, depending on the user’s needs. The tabs contain the digital images of the letter, a diplomatic transcription of the text and a standardized transcription; notes; and documentary, archival and bibliographic information. The corpus can be filtered on various levels, according to different search criteria, such as date, place, people, etc.

56These initial, meaningful digital experiments, along with the development of knowledge about Proust’s work promoted by the paper edition of Cahiers, are the premise of the ambitious digital ‘Hyper-Proust’ proposed by Mauriac Dyer: a single digital place, interoperable on multiple levels that holds the publication of the Cahiers and of all the materials in the BnF’s Proust collection, the text of La recherche published by Proust, along with letters and the writer’s library (Mauriac Dyer 2008a: 168–69). Such a tool would have infinite potential: it could, for example, make it easier to find and compare the various compositional phases of the same passage among multiple manuscripts and typewritten and printed documents, a feature that would be particularly advantageous over the printed reproduction of the single documents that make up each of Proust’s avant-textes.

4.5 Samuel Beckett’s En attendant Godot / Waiting for Godot

57Genetic Beckett studies have a long history that stretches far beyond the so-called ‘archival turn’ of the 1990s, although the accessibility of manuscripts was certainly an issue in the first decades of Beckett scholarship (1960s-70s). Things changed radically when he donated a large number of his manuscripts to the University of Reading’s archive (UoR) in 1971. Thanks to the efforts of scholars like James Knowlson, John Pilling and later Mark Nixon, the archive has grown throughout the years and now holds the world’s largest collection of Beckett manuscripts. Other large repositories are the Harry Ransom Humanities Research Center in Austin, Texas (HRC) and Trinity College Dublin (TCD), Beckett’s alma mater.

58Even this short list of the largest collections points to one of the biggest problems in genetic Beckett studies, namely the enormous geographical spread of archives and holding libraries. A number of important collections are kept in the United States: Washington University in St Louis (WU), Syracuse University in New York (SU), Indiana University (IU), Ohio State University (OSU), and The University of California, San Diego (UCSD). L’Institut Mémoires de l’édition contemporaine (IMEC) and the Bibliothèque National (BnF) are the two most important repositories in France, where the bilingual Irish author lived for most of his life. The scattered nature of Beckett’s legacy was one of the chief reasons, along with preservation, for the establishment of the Beckett Digital Manuscript Project (BDMP) in 2011 (​). The purpose of the BDMP is to reunite the manuscripts of Beckett’s works in a digital way, and to facilitate genetic research: by offering transcriptions of Beckett’s manuscripts, tools for bilingual and genetic version comparison, a search engine and an analysis of the textual genesis of his works.

59As a digital resource, the BDMP uses the affordances of the digital medium to the fullest by foregrounding and visualizing Beckett’s rich and layered intertextuality. In this connection, the Beckett Digital Library (BDL) is a crucial feature. The module consists of the ‘extant’ library (the books that are still in Beckett’s apartment in Paris and in a few other collections) and the ‘virtual’ library (the books we know Beckett read, based on information in letters and reading notes, but which no longer survive). Both the extant and the virtual library contain links to relevant pages in the genetic editions, and — conversely — one can find references to source texts in the individual modules and enter the library from there (see example below).

60Although the focus of the BDMP lies mostly on the endogenesis (the succession of draft versions) and exogenesis (the author’s use of external source texts), it also catalogues and collates different editions of Beckett’s works, for which the Bibliography feature, compiled by Breon Mitchell, provides exhaustive bibliographical information. In some cases, the publication history is marked by a complex epigenesis (the continuation of the genesis and revision after the first publication). The example discussed below — the genesis of Beckett’s most famous play En attendant Godot/Waiting for Godot — is a case in point.

61Samuel Beckett began writing Godot on 9 October 1948, finishing it some four months later on 29 January 1949.2 Together with 9 December 1948, these are the only three dates recorded in the squared ‘Avia’ notebook of the original manuscript, now held at the BnF in Paris. The play was completed in the middle of what Beckett referred to as a ‘frenzy of writing’ or ‘siege in the room’ to his biographer James Knowlson (1996), a sustained period of intense composition in French which yielded another play, four stories, a novella and three novels. Beckett began Godot ‘as a relaxation’, to get away from the ‘awful prose’ he was writing (Colin Duckworth, in Beckett 1966: xlv). The only other handwritten material in French are a few lines of dialogue in the ‘Tara MacGowran’ notebook (OSU). Two typescripts are mentioned in Beckett’s letters, but only an annotated playscript based on the second one survives (Morgan Library, New York), which was also used for a radio recording of the play for Michel Polac’s Entrée des auteurs on RTF’s Club d’Essai in 1952. Two ‘prompt’ copies of Minuit’s first edition (TCD; IMEC) were heavily used during rehearsals for the premiere at the Théâtre de Babylone in January 1953 and, lastly, a fair copy Beckett made for manuscripts dealer Jake Schwartz (HRC) in 1959 is largely identical to the published text.

62Following the play’s success in Paris, serious offers started coming in from the USA. Worried about the fate of his text in the hands of others, Beckett made his own English translation. Its original manuscript was likely thrown away, but a first typescript was sent to Harold L. Oram, the financial backer, and American publisher Barney Rosset (Grove Press, New York) in June 1953. Despite Beckett’s warning that it was only a rushed first draft, it was retyped, duplicated and disseminated for negotiations, but no copy of this playscript has yet come to light. Beckett also sent his first typescript to Britain, for theatre impresario Donald Albery and director Peter Glenville, who likewise had copies made. This playscript did survive, in four versions, all with different annotations — none in Beckett’s hand — and some with unique but probably unauthorized variants. The fourth copy was shown to the Lord Chamberlain’s office, who censored it for performance (British Library). To complicate matters even more, Alan Simpson wanted to stage Godot at the Pike Theatre in Dublin, asking Beckett for a text in November 1953. This time, he made sure to send a copy of his second, revised typescript, which makes it the only English draft to have survived, be it with missing pages (TCD). We know from Beckett’s letters to his American publisher that copies of the first and second typescript were given to the New York Public Library in 1961, but these went missing in 1970. Another fair copy manuscript of the English text was made for Schwartz in 1959 (HRC), but this was again based on a published edition, the American one.

63Godot’s publication history, in both French and English, is long and complicated. The original French edition (1952) appeared before the play’s premiere, so it did not include any of the changes that were made for performance. A few small cuts and additions were implemented in the second impression (1953), which was further updated with substantial excisions for paperback and hardback reissues in 1970 and 1971. Another annotated copy that informed this revised edition was the one Beckett made for the Godot revival at the Odéon Théâtre in 1961, directed by Roger Blin and featuring the famous tree designed by Alberto Giacometti (IU). By this time, various English editions had appeared. The American version (1954), was based on Beckett’s second typescript of the translation and is thus more advanced than the French text. The British edition, however, published by Faber and Faber in 1956, was based on Beckett’s by now obsolete first typescript translation, as well as the various playscripts that were used for the play’s UK premiere. In addition to unauthorized variants, introduced by Glenville, Albery and set designer Peter Snow, it also printed the bowdlerized text. Since Beckett was very unhappy with this situation, a new edition was published in 1965. He had carefully marked up a copy of the American edition for this purpose (Columbia University), but instead of setting the text anew from this document, Faber just incorporated Beckett’s corrections into their original type, so that many of the old variants were left intact. It was therefore not the ‘definitive edition’ it claimed to be.

64The editorial model we developed for the BDMP is based on a text-oriented approach to the transcriptions in XML, following the guidelines of the Text Encoding Initiative (TEI). The rationale behind this choice is that, since we also provide scans of the document, this text-oriented approach is complementary. The disadvantage of a topographic transcription (in a digital context) is that the transcription does not translate the facsimile image into a searchable text, but into another (unsearchable) image, produced by means of graphic software like Photoshop. In the French tradition of critique génétique (genetic criticism), this form of transcription would accord with the principle of donner à voir (made for looking), as opposed to the principle of donner à lire (made for reading). The latter approach (applied in the BDMP) considers it the role of the transcriber to provide a text that facilitates the reading and therefore tends to linearize the textual features of the manuscript. The linearized transcription still leaves open many possibilities to mark visual particularities as well. For instance, the difference between typed text and handwritten annotations can be rendered by means of different fonts. The line breaks are respected, and the blank space where Beckett did not immediately find the right word can just be marked as such. If the linearized transcription is presented in parallel with a (digital) facsimile, the combination (à voir + à lire) is greater than the sum of its parts, as it shows the translation of toposensitive facsimile into chronosensitive linearized sequence.

65For genetic editions, one of the most powerful tools is the possibility for users to diachronically compare segments (<seg>) of the text across versions. To enable this type of genetic research, it is helpful to number the segments and it is up to the editor to decide what the size of these segments will be. The BDMP works with the sentence as a unit of comparison, broadly defined as a syntactic unit that ends with a full stop, an exclamation mark or a question mark. The first edition serves as a ‘base-text’ (or ‘anchor text’) that determines the numbering of the sentences. In case a segment never made it into the published text, and therefore does not correspond with any sentence in the base-text, a solution for the numbering is to take the number of the preceding sentence that did make it in and add a vertical bar | followed by a second numbering. For instance, [0014|001] means this is a segment that did not make it into the final text and is situated in this manuscript as the first segment after the segment corresponding to sentence 14, which did make it in.

66By choosing the page as a division (<div>), it is also possible to link the XML transcription to the digital facsimile. In the BDMP, this coupling of text and image happens at the level of the ‘zone’, a flexible textual unit of about a half dozen lines (depending on the context). The content of the zone can be efficiently linked to the corresponding sentences in the XML transcription. An advantage of this unit’s size is that it facilitates the image/text visualization enabling the immediate comparison of the topography of the facsimile (document-oriented) with the linearized transcription (text-oriented). The image/text view is the most frequently used way of reading the transcriptions. The zone can be drawn on the facsimile with a simple, free software tool like ImageJ, and in the BDMP the four coordinates necessary to encode this rectangular selection in the XML are stored in a <div> element.

67The transcription work focuses on the microgenesis of ‘intradocument variation’ (layers of writing within this one draft), while the macrogenenesis opens up the scope across versions (‘interdocument variation’). Because segments are numbered, the digital architecture of the genetic edition can be designed in such a way that it retrieves all versions of one particular sentence and visualizes them in a ‘synoptic sentence view’. It allows users to compare versions, i.e. what Donald Reiman dubbed ‘versioning’ (1987: 167–80), but at the level of the sentence, which facilitates comparison. Since Beckett wrote in two languages and translated most of his own works, the edition offers the possibility of bilingual version comparison. In the synoptic sentence view, the sentences in the manuscript that did not make it into the final text, such as segment [0014|001] above, are highlighted in bold type and linked to the preceding sentence that did (here segment [0014]). To turn this form of versioning into the equivalent of a critical apparatus of textual variants, new developments in (semi-) automatic or computer-assisted collation (Juxta, CollateX, HyperCollate) enable editors to highlight variants. By integrating Collatex into the edition, the BDMP offers not only editors but also users the chance to compare differences between versions by means of a collation engine, even in the manuscripts (including cancellations and additions), in both the French and the English texts.

  • 3 This passage is quoted from the French notebook of En attendant Godot, module no. 6 in the Beckett (...)

68The following manuscript passage, edited according to the principles of authorial philology, illustrates how Beckett revised, cancelling and substituting fragments between the lines, sometimes using the facing verso for additions (both in superscript below). In the digital genetic edition of the play, passages in the manuscripts that refer to source texts in Beckett’s personal library are annotated and linked to the relevant pages in his books.3


Je me rappelle les cartes de la Terre Sainte.


En couleurs.


Très jolies.


La mer morte était bleu pâle.


J’avais soif rien qu’en la regardant.


Je me disais, c’est là que nous passeronsirons passer notre l’x lune de miel.


Nous serons heureux.



Tu aurais dû être poète.



Je l’ai été.


Ça ne se voit pas?





Qu’est-ce que je disais… .


Comment va ton pied?



Il enfle.



Ah oui, cette histoire dudes larrons.


Tu t’en souviens?






tu sais où ça vient.



Tu sais où ça vient.


(un instant)



Ça sent l’Evangile.



Quelle histoire?



Mais je viens de te le dire.


Des L’histoire des larrons.



Quels larrons?



Les deux voleurs crucifiés en même temps que Jésus l’autre. L’un fut sauvé et l’autre (il cherche le contraire de sauvé) damné.



Sauvé de quoi?



Des enfers?



Je n’ai jamais pu blairer l’Evangile.


Philological notes:

690151 Estragon was still called ‘Lévy’ in the manuscript, which suggests a Jewish heritage or ancestry. The name’s potentially political motivation notwithstanding, the author may have decided to change it in order to avoid accusations of Jewish stereotyping in the aftermath of World War II and the Holocaust. Beckett was in Germany from 1936–1937, witnessing the rise of Nazism, and he later helped the Resistance from Paris and Rousillon after German forces invaded the city (see 0181|001 below).

700155 Lévy’s comment that the very look of the Dead Sea’s pale blue colour made him thirsty, added on the facing verso page, may be inspired by similar maps of the Holy Land in the author’s personal library, for example in his Italian Bible (see Fig. 10).

710156 The author changes ‘nous passerons’ into the more colloquial ‘nous irons passer’. This revision also introduces an extra element of movement by adding the verb ‘aller’, which contrasts with the characters’ immobility. The English (self-) translation likewise makes use of the verb ‘to go’: ‘that’s where we’ll go for our honeymoon’ (Beckett 2010: 8).

720158 The sentence ‘nous serons heureux’ was added inline as an afterthought, which makes it seem more desperate and suggests they are not happy now.

730167 The correction of singular ‘du’ to plural ‘des larrons’ suggests the author was first thinking of using a different word, or perhaps just one thief instead of two, which would have made Vladimir seem even more unreliable (see 0180-0181 below).

740169|001-|008 The author first deleted Vladimir’s question about the story of the two thieves’ source, as well as Estragon’s answer that it stinks of the Gospels, after a pause for thought. He then added the question again between the lines, but now Estragon answers with another question instead, a pattern that continues in the following exchange (see 0180-0181 below). The author’s revision of ‘Des larrons’ to ‘L’histoire des larrons’ also creates more repetition, another central theme in the play. Estragon’s deleted comment may connect to Beckett’s annotation of Luke 23: 34 in his ‘schoolboy’ Bible, which mentions the crucifixion of the two thieves (see Fig. 11).

750179 By obscuring Jesus as ‘the other’, the author seems to suggest that Vladimir does not remember his name, adding to the general breakdown of memory in the play, or is reluctant to name him. In the playscript and the published text, the term was replaced with ‘le Sauveur’ (FP, 07r; 1952, 17), which emphasizes the biblical function of Christ rather than his name. Vladimir also refers to Godot as a saviour in the manuscript — ‘qui dit nous délivrer’ (FN, 04v) — but this explicit identification with Jesus was cut, possibly because it promoted a Christian reading, which the author discouraged.

760180-0181 Further emphasizing the characters’ failing recollection, as well as their lacking knowledge of the Bible or New Testament, in this facing-page addition the one forgets what the two thieves were saved from and the other guesses it was from hell, which gets them no closer to the truth. In the playscript and the published version, Vladimir sounds more assured and Bible savvy: ‘De l’enfer’ (FP, 07; 1952, 17).

770181|001 Lévy’s Jewish-sounding name would have explained why he cannot stand the New Testament and the Gospels, since the Torah or Pentateuch only includes the first five books of the Old Testament. The fact that it was later changed to the more French-sounding name Estragon may be connected to the omission of this sentence from later drafts and versions, although the comment would still have retained blasphemous connotations.

Fig. 11 La Sacra Bibbia ossia L’Antico et il Nuovo Testamento: versione riveduta sui testi originali, Società Biblica Britannica e Forestiera, 1924 (n.p.), Beckett Digital Library,​library/​SAC-BIB.html?page=map&zone

Fig. 12 The Holy Bible: Containing the Old and New Testaments, Translated out of the Original Tongues and with the former Translations diligently compared and revised, by His Majesty’s special command, Oxford University Press, n.d. (p. 862), Beckett Digital Library,​library/​HOL-BIB-1.html?page=862&zone=1


3 This passage is quoted from the French notebook of En attendant Godot, module no. 6 in the Beckett Digital Manuscript Project (BDMP6, FN, 05r).

Table des illustrations

Légende Fig. 8 La Dama Boba, 1613 (Vitr/7/5, f. 29r, num. 7, vv. 1422–1452),​viewer.vm?id=0000051826
Fichier image/jpeg, 129k
Fichier image/jpeg, 288k
Fichier image/jpeg, 388k
Légende Fig 9 Laon and Cythna, Bodleian MS. Shelley adds. e.19, f. 14 (see Bod2; published in BSM, XIII, p. 32)
Fichier image/jpeg, 107k
Légende Fig. 10 Jane Austen, The Watsons, 1804–1805 (Oxford, Bodleian Library, MS. Eng. e. 3764, b.2-3),​manuscripts/​qmwats/​b2-3.html
Fichier image/jpeg, 70k
Fichier image/jpeg, 50k
Légende Fig. 11 La Sacra Bibbia ossia L’Antico et il Nuovo Testamento: versione riveduta sui testi originali, Società Biblica Britannica e Forestiera, 1924 (n.p.), Beckett Digital Library,​library/​SAC-BIB.html?page=map&zone
Fichier image/jpeg, 111k
Légende Fig. 12 The Holy Bible: Containing the Old and New Testaments, Translated out of the Original Tongues and with the former Translations diligently compared and revised, by His Majesty’s special command, Oxford University Press, n.d. (p. 862), Beckett Digital Library,​library/​HOL-BIB-1.html?page=862&zone=1
Fichier image/jpeg, 178k


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search