1

An electronic edition has as many continuities with the past as radical departures. Its fundamental mission of transferring a portion of our cultural heritage to a new medium finds precedent in the humanist projects of Aldus Manutius in the early 1500s, when print was the new technology. Certainly it calls upon the same painstaking scholarship in the service of textual accuracy that has long been the standard of scholarly editions in print. And it recognizes the same need for informed human judgement in selecting and presenting texts for academic study. New media bring change, but they don't dispense with craftsmanship. It is wise, then, not to overstate the impact of computers in any field, especially in the wake of the hi-tech boom of the 90s.

2

What computers do offer, however, is access, whether that is expressed in terms of convenience or volume. An electronic edition can bring to your desktop a facsimile of a rare document, which otherwise you might have to travel hundreds of miles to see. Better yet, it enables you to bring together and compare texts that are physically housed in several different locations. Of course a similacrum is not the real thing, and there will always be some questions that can be answered only by consulting the book itself (happily, the role of the library is by no means supplanted), though consulting an electronic edition can save on a good deal of preliminary legwork. An electronic edition, then, ought to be more than a single electronic text, but rather a collection of electronic texts (and images) which are linked together in meaningful ways.

3

Digital storage, especially in a networked environment, is extensible in ways the printed page cannot be, allowing vast amounts of material to be assembled and combined. An annotated electronic edition can provide not only convenient access to explanatory notes but connections to whole other texts invoked in the annotated passage. Likewise, a scholarly electronic edition can offer what a traditional critical edition cannot, that is, access to the whole body of a number of variant texts that instantiate a literary work.

4

Lyrical Ballads is a good candidate to illustrate what a scholarly electronic edition can do because it exists in so many different forms. When Jack Stillinger warns us that "The Ancient Mariner" went through 18 versions, each representing a distinct authorial intention at a particular time (119-21), it is a wonder that anything at all resembling the "literary work" can be located amid such a plurality of texts. Certainly, it is not clear which one should be taken as authoritative, and constructing an ideal 19th text would only add to the confusion. But perhaps traditional textual criticism's goal of perfecting the text was as much constrained by the limits of the printed page as motivated by any quest for the true work of art. An electronic edition breaks with these past practices in seeking neither to privilege any single version nor to construct an ideal exemplar. Instead, by presenting a number of extant versions each in their entirety, it is in tune with current editorial theory which it seeks to ground in the practice of the new medium.

5

Admitting that a definitive edition is an "impossible ideal" (McKenzie 2; see also Finneran x and McGann 102-04), many editorial theorists now insist instead on a respect for the integrity of each version as the embodiment of evolving intentions at a particular point in time. Such postmodern textual scholars see the work of literature more as "work" in the sense of an activity, and the text generated by such work as a work-in-progress. The practical consequence of such an approach is to acknowledge each stage of the text's development as unique and valuable in itself, and to represent these multiple versions on an equal footing. Editing in the traditional sense gives way to what has been called "versioning" (Reiman 169; Greetham 237), in which textuality comes to be represented more as a series of snapshots rather than a single, composite portrait. The result is a "genetic" edition which documents the historical development of the work, but which takes no single text as its standard.

6

An electronic edition can accommodate this dispersal of textual authority, but (as is often the case with computers) at the risk of information overload. In theory, every version showing even the slightest typographical variant could be made available, but whether this should be the practice in electronic editions runs up against the limits of intelligibility. Some principle of selection needs to exercised on the part of electronic editors, otherwise their readers will be swamped. This electronic edition of Lyrical Ballads provides SGML-encoded texts of the five states of the work as a discrete collection authorized during Wordsworth and Coleridge's lifetimes: both the Bristol and London imprints of the first edition of 1798, and the two-volume editions of 1800, 1802, and 1805. These texts are supplemented by scanned photographic images of all the pages of the various editions, which can be compared with the encoded texts through hypertext links. In addition, this system of hypertextual linkage has been extended to make possible two kinds of navigation through the versions: any two texts may be viewed side by side, and all identified variants may be studied in context by means of a functionality we call "dynamic collation." Our aim is to assemble a serviceable scholarly electronic edition of the printed book, Lyrical Ballads, which preserves two important characteristics of the collection that traditional print editions have made it difficult to discern: first, the complex interaction between authors, publishers, and printers that brought it into being, and, second, the multiplicity of versions of the collection that readers had available to them in the early nineteenth century.

top

7

So that the reader will understand our procedures more clearly, it is necessary to give a brief account of the publication of each of the editions of Lyrical Ballads. This account depends largely upon earlier scholarship, especially the bibliographical and editorial studies of Wells, Foxon, Healey, Reed, Butler and Green; those seeking more detailed information should consult their works. These and other commentaries are listed in our selected bibliography of scholarly studies, which provides links where these exist online.

8

When in 1798 Wordsworth and Coleridge decided to publish a volume of poems, they were not primarily thinking of ways to revolutionize English poetry. Instead, as Mark Reed has shown, they needed money for an extended trip to Germany, where William and Dorothy Wordsworth were to become fluent in German, and where Coleridge would study at German universities (Reed 1965). With this is mind, they turned to Joseph Cottle, the Bristol publisher and bookseller who had already published several volumes of Coleridge's poetry, and could be depended upon to come to the financial aid of fledgling authors. Several proposals were put forward, including a joint publication of the poets' two tragedies, Osorio and The Borderers, a single volume containing "The Ruined Cottage" and "Salisbury Plain," or "Salisbury Plain" and "Peter Bell," and even a two-volume set of Wordsworth's poetical works. But all of these proposals came to naught, and by late May of 1798, Cottle had agreed to publish a collection of ballads, including Coleridge's "Rime of the Ancyent Marinere," Wordsworth's "The Thorn" and "The Idiot Boy," as well as other poems that the two authors had ready for the press. The volume would appear anonymously—according to Coleridge, "Wordsworth's name is nothing—[and] to a large number of persons mine stinks"—and Cottle agreed to advance the poets thirty guineas as payment (Butler & Green 3-12, 43-44).

9

At Coleridge's insistence, typesetting began almost immediately, and the first poem to be set was "The Rime of the Ancyent Marinere." Wordsworth himself was frequently in Bristol "to superintend the printing" (Letters: Early Years 219), and the result was a remarkably well-printed volume: there were just five printer's errors corrected on an errata sheet, and a handful of others that were noticed and corrected in the printing process. By late August, the entire volume was ready for binding, and Cottle bound up a few copies apparently for private circulation. Subsequently, and perhaps because of readers' reactions, the authors made two major changes to their collection: Coleridge's poem "The Nightingale" was substituted for his "Lewti" (probably because the latter poem had already appeared in The Morning Post, and its inclusion might compromise the anonymity of Lyrical Ballads), and a brief explanatory preface, called an "Advertisement," was written and inserted just before the table of contents. A few more copies were then bound up, some of them for the poets and their friends, including Coleridge, Robert Southey, J. F. Pinney, and probably Thomas Lovell Beddoes; some of those copies were what Reed has called "archival copies," in that they included both "Lewti" and "The Nightingale" (Reed 1998). The result is that the first printing of Lyrical Ballads existed in multiple versions almost from the start. Of the fourteen surviving copies with the Bristol imprint, three contain "Lewti" and neither "The Nightingale" nor the "Advertisement," two contain "Lewti," "The Nightingale," and the "Advertisement" (one of these, Robert Southey's copy, has two different tables of contents), and nine copies contain "The Nightingale" and the "Advertisement" (Reed 1998). Of these nine, one also contains, following "The Nightingale," an extra leaf on which is printed Beddoes' "Domiciliary Verses," a poem which parodies The Lyrical Ballads. If Duncan Wu and Reed are right, this was Beddoes's own copy (Wu; Reed 1998).

10

At the last minute, just as the Wordsworths and Coleridge were about to leave for Germany, Cottle decided not to publish the book himself. He was concerned about his growing financial difficulties (for this reason, he quit the bookselling business the next year), and, fearing that the volume would do poorly, he began to look for a London publisher. T. N. Longman was certainly approached but just as certainly refused Cottle's offer. Wordsworth approached the prominent liberal publisher, Joseph Johnson, who had previously published An Evening Walk and Descriptive Sketches, and Johnson agreed to assume publication. But Cottle, probably independent of Wordsworth's efforts and certainly without telling him, offered and sold the copyright to the more obscure firm of J. & A. Arch. Thus, on October 3, 1798, Wordsworth wrote from Germany that he still did not know who his publisher was, and on the next day, October 4, 1798, Lyrical Ballads finally appeared in London bookshops with a title page bearing the London imprint of J. & A. Arch (Butler & Green 14-15). It was composed entirely of sheets printed and bound for the Bristol imprint; only a new title page was added as a cancel.

11

In spite of its haphazard beginnings and a few hostile reviews (most notably by Robert Southey in the Critical Review and Charles Burney in the Monthly Review), the first edition of Lyrical Ballads sold reasonably well, and sales were aided in 1800 by the support of Mary Robinson in The Morning Post,and by a pair of strongly favorable notices in The British Critic and The Anti-jacobin Review (Butler & Green 23-24; Woof 172-175). In the meantime, William and Dorothy Wordsworth and Coleridge had returned to England, the Wordsworths to take up residence in Grasmere, and Coleridge, after spending a few months in London, relocated himself and his family at Greta Hall, near Keswick, a few miles up the road from the Wordsworths. Upon settling in Grasmere, Wordsworth began writing a series of new poems, most of them about places and people in the Lake District, several of which he called pastorals. Together with poems written in Germany, these gave Wordsworth a significant body of new work which he was eager to publish, and Coleridge, still residing in London, acted as his agent. Negotiations were begun with the firm of T. N. Longman for a second edition of Lyrical Ballads, and by June, 1800, Wordsworth was able to write his brother, Richard, that the first edition had sold out, and that Longman had agreed to publish a second edition, for which he would be paid £80 (Butler & Green 23-24). It was to include a second volume of new poems, and a substantial new "Preface." One poem, Wordsworth's "The Convict" was withdrawn from volume I, and in its place, Coleridge's "Love" was added.

12

Over the next six months, Lyrical Ballads (1800) was printed. It is, remarked George H. Healey, "bibliographically the most complex of all Wordsworth's books," and the reasons for its complexity are not far to seek (Healey 6). First, although Wordsworth and Coleridge had contracted with Longman for a new volume of poems, and although Wordsworth himself had a considerable stock of new poems already written, Coleridge's intended contributions (including Christabel and a group of poems to be set in the Lake District) were either unwritten or unfinished. Second, the Bristol firm of Biggs and Cottle was again engaged to print the volume. But this time the authors could not oversee the printing process because they were in the Lake District, hundreds of miles away. Their only means of communication with the printer was by post, and their letters sometimes took weeks to arrive. To compensate for this difficulty, they enlisted the services of the young chemist, Humphry Davy, newly famous for his studies of the effects of nitrous oxide. Davy was asked to check over the printer's manuscripts and the printed proofs, and even to add and correct punctuation. But Davy was by no means privy to the poets' intentions, and thus would not have been able to answer accurately printer's queries, nor was he especially qualified as a proofreader, and, to make matters worse, he was ill much of the autumn of 1800 and was probably not able to offer much assistance. Third, the letters containing the manuscript poems were sent over a period of several months, between late July and December, 1800, and Wordsworth revised poems that had been sent in earlier letters, changed the arrangement of the poems even after the printing process had begun, and even tried to change the title of the volume. At least one letter was lost in the mail, causing further delays. And then, in October, 1800, the poets made the most significant, and most controversial, change of all: they withdrew the longest new poem in the second volume, Coleridge's Christabel, and eventually substituted for it Wordsworth's pastoral poem, "Michael," which, at the time Christabel was withdrawn, had not yet been conceived of. As a result, Biggs and Cottle had to cancel the first printed leaf of the "Preface," which mentioned Christabel, and to delay publication of the new edition until Wordsworth had finished his 485-line poem. The delays meant that Longman could not distribute the volumes in time to take advantage of the 1800 Christmas trade, and Biggs and Cottle had to rush "Michael" through the press. Inevitably, they made a big mistake: fifteen lines of the poem, crucial to the plot, were omitted, and Wordsworth complained that the volume "is throughout miserably printed," while Coleridge fumed over "an infamous Blunder of the Printer" (Butler & Green 26-31, 123-125; Healey 6).

13

Besides complaining, Wordsworth was quick to send Longman corrections to "Michael". On April 9, 1801, less than three months after the volumes appeared, he advised Thomas Poole to send to Longman for a "half a sheet" containing corrections. The phrase "half a sheet" refers either to a printed paste-in which supplies the fifteen lines missing from "Michael," or to printed cancels containing both the missing lines and a new, twenty-seven-item errata leaf, which were usually bound into the second volume of Lyrical Ballads, but may have been issued separately as well. The paste-in survives in two copies, one in the Huntington Library, and the other at Swarthmore College. It is not a handsome thing: the type does not match the rest of the volume in size or font, the inner and outer edges of the paper are torn, not cut, and, in the Huntington copy, it is not even pasted in straight. John Edwin Wells, who purchased the Swarthmore copy in the 1930s, suspected that the paste-in may have been a sophistication added later by a not wholly scrupulous bookseller (Wells, Library). More recently, Butler and Green seem convinced that it represents Longman's earliest, rather crude effort to correct the printer's error (Butler & Green 126). The cancels, on the other hand, are a much more attractive, professional-looking effort. They match the paper, font, and the layout of the original volume, and replace pages 209-212 with a new set of pages, numbered 209-210, *209-*210, 211-212. The four-item errata leaf of the first printing, pages 227-[228] was also replaced, and the new errata leaf contains twenty-seven items, correcting most of the rest of the printer's errors, and offering several new readings as well (Healey 6; Butler & Green 126; Wells, PMLA). If one counts three copies that are currently unlocated, only about a dozen surviving copies of Lyrical Ballads (1800) contain the cancels (of these, one contains just the "Michael" cancels; another just the new errata leaf), which suggests that they were added only after most of the copies had been sold or distributed to other booksellers (Healey 6). In any case, what was true of the 1798 printing was also true of the edition of 1800: multiple versions circulated, as a result either of printer's errors, miscommunications between authors and printers, or the authors' own revisions. Lyrical Ballads (1800) was clearly a work in progress, even as it appeared in published form.

14

In fact, preparations for a new edition began almost as soon as the errors in the 1800 volumes had been corrected. Using two proof copies of the 1800 edition which contained the cancels, the Wordsworths began entering revisions and corrections , keeping one copy for themselves and eventually sending the other to Longman as printer's copy for a new edition. This process began during the summer of 1801, and was complete by April, 1802 (Butler & Green 31-32). Some of the changes were matters of formatting and presentation: the order of the poems was changed, some poems were omitted and others were moved to the second volume, and a new half-title was added after the "Preface," with an epigraph from Quintilian. Other changes were more extensive: a long section was added to the middle of the "Preface," containing the famous definition of the poet ("a man speaking to men") and the comparison of the poet and the "man of science," an "Appendix" concerning "poetic diction" was added at the end of volume I, a note to the "Ancient Mariner" was dropped, and, among the poems, "Ruth" was drastically revised. Two new poems, "Louisa" and "I travell'ed among unknown men," were initially intended to be published in the second volume of the new edition, but were later withdrawn. Drafts towards them survive in a family copy of Lyrical Ballads (1800), now at St. John's College, Cambridge, and a note about the placement of "Louisa" is included in a proof copy of Lyrical Ballads (1800) containing the revisions for 1802, now in the Beinecke Library (Curtis 196-201; Butler & Green xxxiv-xxxv).

15

The third edition of Lyrical Ballads appeared in June, 1802, just in time to achieve notoriety. That year the inaugural number of the Edinburgh Review was published, and in the October number, in a review of Southey's Thalaba, Francis Jeffrey began his long campaign against Wordsworth and the "Lake School" of poetry. This notoriety may have spurred Longman to market the volumes more aggressively: full-page notices, quoting liberally from favorable reviews, began to appear in Longman's advertisements in the backs of better selling works, such as Southey's edition of Chatterton (1803) and Joanna Baillie's Miscellaneous Plays (1804), and adverts for Lyrical Ballads appeared side-by-side with ones for the best-selling poem of the time, Robert Bloomfield's The Farmer's Boy. But Longman's marketing efforts seem to have had little effect : the print run of 500 copies took almost three years to sell out (Butler & Green 32)—at a time when a competing collection, Bloomfield's Rural Tales, Ballads, and Songs, sold thousands of copies in less than a single year— and it was not until October 9, 1805 that the fourth and final authorized edition of Lyrical Ballads appeared.

16

Lyrical Ballads (1805) has historically been considered the least interesting of all the editions of the collection. Few authorial changes were made: no new poems were added, no old ones subtracted, and no new front or back matter was included. Only "Ruth" received extensive revisions, and most of those were reversions to the text of 1800. But though the authors made few significant changes to the text, a change was made by the publisher that subtly affected how the new edition would be read: Longman changed printers. Rather than the firm of Biggs and Cottle, which by 1802 had relocated to London, Longman entrusted the printing of the 1805 volumes to R. Taylor of Shoe-Lane, London. Taylor's firm imposed much more rigorous standards of spelling, capitalization, and punctuation than had been practiced by Biggs and Cottle, it used more modern type, eliminating altogether the archaic ct ligature and the long "s" and, on the whole, Taylor produced the most consistently-printed edition of Lyrical Ballads that we have. Because of this consistency, as well as the avoidance of archaic typography, the 1805 edition is also more modern in appearance: with this printing, Lyrical Ballads looks to its readers less like a quaint holdover of the late eighteenth century, and more like a fully modern book.

17

The 1805 Lyrical Ballads was the last authorized edition of the collection to appear. By 1815, Wordsworth's contributions had been dispersed throughout the first collected edition of his poems, and by the 1820 collection, the phrase "Lyrical Ballads" had disappeared from his title page. Coleridge, too, had republished his contributions in Sybilline Leaves (1817), the same year in which he publicly distanced himself from his collaborator's critical opinions in Biographia Literaria. Pirated editions of Lyrical Ballads, made up of leftover sheets from Longman's various print runs, appeared in 1820, at least one of which is actually a reprint of one volume of Poems, in Two Volumes (1807) (Butler & Green 32; Wells, PQ 398-402; Healey 22). But Wordsworth himself had gone on to other things: to tour poetry, sonnet sequences, and, of course, the unfinished Recluse. And so had Coleridge, bringing to an end one of the most important collaborations in the history of English poetry.

top

18

This edition reproduces full texts of five editions or states of the Lyrical Ballads, encoded in SGML according to the guidelines of the Text-Encoding Initiative. We have used the "lite" version of the TEI document-type definition (DTD), and have made no modifications to it. Throughout, we have used the <L> tag for individual lines, and have included the line number as an attribute value (n="101"). We have also used the <LG> (line group) tag to indicate stanzas and verse paragraphs; the "type" attribute has been used to distinguish stanzas from verse paragraphs (type="stanza" or type="verseparagraph"); we have also numbered stanzas and verse paragraphs, again using the "n" attribute. Thus our texts can be searched in a variety of different ways, depending on the search engine used, and the ingenuity of the user.

19

Our textual transcriptions are based on specific copies of the collection: the Princeton University copy of the Bristol imprint of 1798; and the Simon Fraser University copies of the London (1798) imprint and the two-volume editions of 1800, 1802, and 1805. Our aim is to produce texts of the collection that represent, as closely as possible, the actual texts available to readers in the early nineteenth century, and thus we have transcribed these copies exactly as they appear, even to the extent of preserving some printer's errors. In the case of printer's errors, our rationale is that when these errors are corrected in the printing process, either by cancels, errata sheets, or some other means, they preserve important records of the interaction between publisher and author. When there are uncorrected printer's errors, or errors that can be attributed to type batter, we have silently corrected them, primarily so that they do not impede electronic searches of our transcribed texts. Thus the error "horsebehind," found in some copies of the 1798 edition (including the Simon Fraser copy of the London printing), is silently corrected, as are several errors attributable to type batter on page 204 of the 1798 edition. But the error "fog smoke-white," corrected to "fog-smoke white" in the errata to the 1798 edition, is left uncorrected, as are the errors "te" and "becn" in "The Idiot Boy" (1800), which are corrected with a cancel leaf in some copies. Our aim is also to make possible the kinds of analysis that an electronic environment can provide: electronic searches, performed according to a variety of criteria; side-by-side comparison of various editions; even studies of layout and typography, using digital images of pages of the printed copies. It must be stressed, however, that the focus of this electronic edition is our lightly edited transcriptions of the selected versions. Images are linked to each page only for the purposes of illustrating features of layout and typography that electronic texts are inept at reproducing. Page images may be called up when required, but are not continuously linked to one another as on a spool—microfilm exists for that purpose.

20

Our decision to base our texts on specific copies has led us to make a potentially controversial decision regarding our reading texts. Most print editions attempt to represent authorial intent. Our edition does not. Rather, it attempts to preserve the artifact, the printed book Lyrical Ballads, in a digital form that approximates what it actually looked like in its various material forms. This procedure has meant that, in the case of the 1800 edition, we have chosen to privilege the most commonly available printing of volume two, the one which omits fifteen lines of "Michael." We would like eventually supply a full text of volume two with the "Michael" cancel, based on the copy at the University of Colorado, and we will also supply a text based on the Swarthmore copy of volume two that contains the paste-in correction. But, for this web-based edition, we only reproduced text and page images of the cancels and the paste-in, and have included them in a separate appendix.

21

A similar, though less controversial, problem exists regarding the advertisement leaf published with both the Bristol and London imprints of 1798. When published, all copies of this edition included a leaf advertising fourteen works published by T. N. Longman for Joseph Cottle, including works by Southey, Beddoes, Coleridge, and Cottle himself. This leaf was often removed by nineteenth-century bookbinders. As it happens, the advertisement leaf survives in the Princeton copy of the Bristol imprint; in the Simon Fraser copy of the London imprint it does not. Our transcriptions and page images reflect this state of things. We do not, however, mean to imply that the advertisement leaf was not included in the London imprint: it was.

22

A word needs to be said about the digital images. Our digital images are, for the most part, taken from the Simon Fraser copies. In the case of the Bristol (1798) imprint, we have used the Simon Fraser images of the London (1798) imprint when they do not differ from the Princeton copy. For the cases in which they do differ, we have used images from the Princeton copy itself, or, in the case of "Lewti," images from one of the Yale copies of the Bristol (1798) imprint that contain that poem. Our reasons for this procedure are practical. First, except when sheets containing new material were added, the Bristol and London imprints were made from the same printed sheets, produced at the same time. Thus it is obviously cheaper to use one set of images rather than two. Second, the Bristol imprint is so valuable, and its copies so fragile, that it is irresponsible to subject any one of them to the kind of wear that producing complete photographic facsimiles would entail. In fact, it is best to photograph each copy as little as possible. Since the text of "Lewti" is identical in all the five copies which contain it, and was produced in the same print run, we took those images from the relatively well-bound copy at Yale, rather than further weaken the copy at Princeton, the binding of which is in rather delicate condition.

top

23

Representing the different choices an artist makes at particular times has long been a problem for compilers of genetic editions. The traditional apparatus criticus relies on footnotes which are visually subordinate to some single text, and conceptually therefore variations from some standard. Only with great difficulty can the reader reconstruct the various texts from this scheme, whether by mentally plugging in the textual variants to the printed page or by laboriously re-transcribing the text with the selected variants. Transposing lists of textual variants to the electronic medium would only prolong this difficulty, and perhaps even compound it by adding more heaps of information which it becomes increasingly troublesome to sort through. This electronic edition of Lyrical Ballads instead attempts a new approach to the table of variants which takes advantage of some of the features that distinguish electronic text from print.

24

"Dynamic collation" is proposed as a new form of critical apparatus appropriate to computer-based editions. It is a navigational system of hypertext links that is designed to take advantage of the new medium's potential for animation. In contrast to the stasis of the printed page, hypertext involves a series of apparent "leaps" from one spot to another. Machines such as the Hinman collator have been in existence for some time, and are well-known for the way they compare texts and provide visual cues to changes. Electronic representation, however, allows us to dispense with cranks and mirrors by quickly marshalling parallel passages from the SGML files and displaying their textual variations in the context of the version in which they appear.

25

Using complex perl-scripts that work in the background, our electronic edition replaces static footnotes with the scheme of active comparison we call dynamic collation. What is seen on the screen is an array of four windows displaying the text of any poem as it varies through the four lifetime editions (1798, 1800, 1802, 1805). This parallel display is accompanied by a fifth window on the left which maps the changes in the poem as descriptive hyperlinks. Clicking on a link in this "variant map" causes the text in each of the four windows to leap to the same line where the revision in question may be observed in context. This variant map turns the concept of footnotes inside out, putting the revised texts in the foreground and relegating the reference to an operation on the side. This reconceptualization of how variant readings can be represented in the electronic environment tests in actual practice David Greetham's proposition that "dismembering scholarly apparatus" will be a consequence of the transition to the new medium (Greetham 329).

26

Though the texts in each of the four windows can be scrolled through independently, dynamic collation is intended to be driven by the variant map to the left. The variant map acts a guide to revisions that were made at various stages in the poem's development. It is based on the poem to the extent that it reproduces the text of the poem wherever changes were not made, but whenever a change in any of the versions under consideration is encountered it substitutes a descriptive hyperlink for the variants themselves. Scrolling down the variant map, the reader is thus alerted by a sort of palimpsest that an alteration has been made, and by clicking on the "hotspot" or hyperlink can summon up the parallel passages. Replacing a standard or base text with an abstract variant map ensures the neutrality of versions and at the same time builds on the dynamism of the medium, for the hiatus in reading caused by running across a link cues the reader to click on the spot and look to the right (in the direction of our accustomed flow of reading) to learn the word or character elided and to compare texts. Rather than footnotes which distract attention from a definitive text, the variant map is an abstraction of the poem which does not privilege one version of the text over another and that piques the reader's curiosity by means of gaps in the text to pursue the significance of revisions made in successive versions.

27

Dynamic collation may be regarded as a region in the topography of an electronic edition which is dedicated to close study of textual variation. Reading for pleasure is best accommodated in the region denominated "The Texts". For the sake of convenience, a third region has been provided in which the full texts of any two of the printed versions of Lyrical Ballads may be compared. It is left to the reader to decide on the priority among these.

28

As readers may note certain variations not marked in the dynamic collation scheme, it is appropriate to comment on the variants selected and on how they are designated. Changes in actual wording are marked as [diction], together with the year of the edition affected. When whole lines or stanzas are cut or added, or in some cases moved, these are marked with appropriate descriptions, accompanied as in all cases by the year. Glaring typographical errors, especially when they are distinctive bibliographic features of the printings, are marked as [typo]. The occurrence of the long "s" in the "Ancient Mariner" is marked as such, as this is one of the striking features of the first edition. Changes in spelling, capitalization, and hyphenation are marked throughout, especially as cumulatively they are a distinguishing feature of the 1805 edition. In the interest of avoiding information overload, changes in punctuation have not been marked, except for a number of dashes and exclamation points which have been marked as [punc] purely for purposes of demonstration. On reflection though, if all changes in punctuation were to be accounted for consistently, it would wise to do so in a separate dynamic collation devoted to such variants exclusively.

29

Future upgrades of this electronic edition, then, might include a more elaborate apparatus, although the energies of the editors might be better expended on improving the searchability of the texts. Such improvement, however, would require another layer of SGML markup, and would necessitate considerable deliberation on precisely what features should be encoded so a search engine could find them. In future iterations we hope also to provide digital texts and images of more variant printings of the Lyrical Ballads, including full texts of several of the Bristol (1798) printings, the variant printings of Lyrical Ballads (1800), and the American edition of Lyrical Ballads, published in Philadelphia in 1802. In addition, we hope to provide digital images and searchable transcriptions of the surviving printer's manuscripts of the collection, when the technological means of representing manuscript revisions improves. We shall rely on readers' responses in determining the directions in which we will go.

top

Bruce Graver
Providence College
Providence, Rhode Island, USA
Ronald Tetreault
Dalhousie University
Halifax, Nova Scotia, Canada

Butler, James. "Wordsworth, Cottle, and the Lyrical Ballads: Five Letters, 1797-1800." JEGP 75 (1976): 139-153.

Butler, James, and Karen Green, eds. Introduction. "Lyrical Ballads" and Other Poems, 1797-1800. By William Wordsworth. Ithaca: Cornell University Press, 1992.

Curtis, Jared R. "A Note on the Lost Manuscripts of William Wordsworth's `Louisa' and `I travell'd among unknown men'." Yale University Library Gazette 53 (1979): 196-201.

Finneran, Richard. The Literary Text in the Digital Age. Ann Arbor: U of Michigan Press, 1996.

Foxon, D. F. "The Printing of Lyrical Ballads, 1798." The Library, 9 (1954): 221-241.

Greetham, D. C. Theories of the Text. Oxford: Oxford UP, 1999.

Healey, George Harris. The Cornell Wordsworth Collection: A Catalogue of Books and Manuscripts presented to the University by Mr. Victor Emanuel, Cornell, 1919. Ithaca: Cornell University Press, 1957.

McGann, Jerome J. A Critique of Modern Textual Criticism. Chicago: U of Chicago Press, 1983.

McKenzie, D. F. Bibliography and the Sociology of Texts. Cambridge: Cambridge UP, 1999.

Reed, Mark L. "The First Title Page of Lyrical Ballads, 1798." Philological Quarterly 51 (1998): 230-240.

---. "Wordsworth, Coleridge, and the `Plan' of the Lyrical Ballads." University of Toronto Quarterly 34 (1965): 238-253.

Reiman, Donald. Romantic Texts and Contexts. Columbia: U of Missouri Press, 1987.

Stillinger, Jack. Coleridge and Textual Instability. New York: Oxford UP, 1994.

Wells, John Edwin. "Lyrical Ballads, 1800: A Paste-In." The Library 19 (1939): 486-491.

---. "Lyrical Ballads, 1800: Cancel Leaves." PMLA 53 (1938): 207-229.

---. "Wordsworth's Lyrical Ballads, 1820." Philological Quarterly 17 (1938): 398-402.

Wordsworth, William and Dorothy. The Letters of William and Dorothy Wordsworth: The Early Years, 1787-1805. Ernest de Selincourt, editor; revised by Chester Shaver. Oxford: Oxford UP, 1967.

Wu, Duncan. "Lyrical Ballads, (1798): The Beddoes Copy." The Library 15 (1993): 332-335.