According to Lee Erickson, in a recent essay, Walter Scott’s refusal of the poet laureateship in 1813 and the publication of his first novel, Waverley, a year later were sure signs of ‘a coming shift in the readership’s taste’, and he goes on to trace a turnabout in the relative popularity of poetry and prose fiction in the later Romantic period.  As we are now aware, however, Scott’s first novel was written in at least two stages: Scott himself, in the first chapter of Waverley, suggests 1805 as the starting-point, but evidence survives of a clear intervention in 1810, which—as I have argued elsewhere—may represent the true inception of the project (put bluntly, it is not unlikely that Scott started the novel in Autumn 1810). Scott would then have known about the popular success of Jane Porter’s Scottish Chiefs (published March 1810),  which offered a clear signal that Scottish subjects could be profitably used in the novel, and by 1810 he would have had a much clearer view of the potential of the ‘national tale’ as developed by Sydney Owenson and Maria Edgeworth. In fact, a cynical view would be that by predating his intervention at 1805, Scott constructed a literary history which placed him at the helm.
Another perspective here is provided by Ina Ferris’s excellent study of the Waverley Novels as institution, The Achievement of Literary Authority (Ithaca, NY and London: Cornell University Press, 1995), which sees Scott as the main agent in a wholesale shift from a generally female-authored ‘common novel’, to a new kind of ‘serious/historical novel’, which allowed main subjectivity to enter into a female genre without compromising ‘masculinity’. Ferris’s thesis on one level echoes earlier claims by feminist literary historians such as Dale Spender that the novel was predominantly a ‘female form’, while tending to corroborate the more specific allegation made by Gaye Tuchman, in Edging Women Out (London: Routledge, 1989), that the Waverley Novels somehow represent an untypical male capturing of the novel form. But Ferris also more broadly points to a larger underbelly of new readers, connected with commerce and manufacturing, who were already turning in increasing numbers to fiction. At what point the novel became ‘unstoppable’ is a moot point, but the statistics shown below in Fig 1, based on the number of new items in J. R. de J. Jackson’s Annals of English Verse (New York: Garland, 1985) against Andrew Block’s The English Novel 1740–1850 (1939, rev. 1961; London: Dawson, 1968), indicate that the first year in which the output of fiction outnumbered that of poetry was 1810 (which is followed by a more general overtaking in the 1820s, as shown in Fig 1).
Poetry vs. Fiction, 1780–1829 [Fig 1]: Here it is necessary to own up to some manoeuvring in order to get the figures: comparing different genres by output is, of course, always likely to bring some mixed results. But most damaging of all, in the present instance, is the unreliability of Block’s catalogue (the by-product of a career as an antiquarian bookseller in London), which jumbles together chapbooks, shilling shockers, miscellanies, non-existent ‘ghost’ titles, works which on examination prove not to be fiction, and other such flotsam, alongside mainstream novels. More immediately pertinent is Block’s habit of ‘rounding off’ dates to the nearest decade in cases of uncertainty, which partly explains the unusually large number of ‘1810’ items. A number of gender-selective studies have also appeared since the 1980s, but these raise their own difficulties: clearly women wrote much more fiction than was ever given credit for before, but (to put it crudely) a lot from how much?
It was such issues relating to Romantic fiction that nearly twenty years ago led me to start compiling a catalogue of fiction titles between 1780 and 1830, in what now looks the incredibly quaint form of a card index file. This was assembled in an eclectic way, making use of modern sources as they became available, such as the Eighteenth-Century Short-Title Catalogue (ESTC), as well as a variety of documents contemporary with the period (over fifty circulating library catalogues, William Bent’s London Catalogue of Books, review listings, and so on). The index finally settled down as a collection of some 2,897 items, representing—I then felt confident—more than 90% of original titles in this period. It allowed a number of special observations to be made about the production of fiction—for instance, the spread of publishers, and gender of authors—but always with the proviso that in many cases information was not based on copies seen at first hand. This was brought home to me by Mervyn Janetta, editor of The Library, when I wanted to add a Checklist to an article about the publisher J. F. Hughes—he argued that only bibliographies based on copies actually examined at first hand carried weight—it was he too who used the word ‘quaint’ about card index files.  So, I shuffled back to Cardiff, feeling terribly provincial, and stopped working on it …
The project then revived, unexpectedly, in 1990 through contact with the collection of English novels in the Library of Corvey Castle and with Projekt Corvey at Paderborn University Schloss Corvey is near Höxter on the River Weser, about thirty minutes’ drive from Paderborn (itself a cathedral town in Westphalia). ‘Die Fürstliche Bibliothek’ (Princely Library) is an aristocratic family library, containing about 67,000 volumes, mainly in German, French, and English, with a tailing off circa 1834. One striking feature of the collection is the large number of English novels belonging to the Romantic period, which is quite exceptional in view of the fact that in Britain fiction was more often borrowed than bought, and even when bought rarely preserved in libraries. Projekt Corvey is co-headed by Professors Rainer Schöwerling and Harmut Steinecke (a Germanist), and since the late 1980s has been involved in processing the contents of the library. Books were brought over in single vanloads to the project room in Paderborn for processing in three ways:
- All title pages were Xeroxed
- Bibliographical details for each item were then taken, and sent to the central German computing system at Cologne
- Microfiches were made of the full text of each title.
When I first arrived in May 1990, the cataloguing of the first phase, focusing on belles lettres, was under way, and I was kindly given a microfiche containing all the title information compiled for the Cologne central computing system.
Through collation of the Corvey titles against my original card index file, it was possible to make a number of generalisations about the Corvey holdings in relation to the output of fiction in the Romantic period. The serious purchasing of English titles clearly began in the mid-1790s. By the early 1800s, the library appears to have been taking about 80% of production, with a regular intake of fifty new novels annually. In the 1820s, accessions reached a new level, with the library in two single years (1822 and 1829) actually taking all but one of the novels in my index file. This exercise also threw light on the kinds of novel that are absent from the library. Prominent here are translations into English of works previously published in French or German. Other omissions include subscription novels and works ‘published for the author’ (e.g. Sense and Sensibility), which for commercial reasons were not pushed by the book trade; titles issued by publishers who were not fully established (e.g. Henry Colburn in his early days), or by those who were never respectable (e.g. J. F. Hughes), and scurrilous titles which might be taken to indicate a bawdy content. Also discovered by this process were some fifty to sixty titles which had been previously unknown to me, and comparison against sources such as the Nineteenth-Century Short-Title Catalogue (NSTC) and the National Union Catalog (NUC) suggested that a small but significant proportion of these are probably unique to Corvey. 
A decision was taken to compile a new Bibliography to replace Block, bringing together my original file, the holdings of Corvey Castle, and also involving collaboration with James Raven, the compiler of British Fiction, 1750–1770, whose team would deal with 1770–99. This lead to a contract with Oxford University Press for English Novels 1770–1830: A Bibliographical Survey of Prose Fiction Published in the British Isles, in two volumes (1770–99 and 1800–29)—and, delivery is due in 1998. All entries in the new Bibliography whenever possible are based on physical examination of a first edition—this has been achieved in all but a handful of cases for 1800–29—and they record the full title, details of authorship, and imprint information as they appear on the original title-page. Below are some typical entries from the year 1820, which will give an idea of procedure.
#TIT# ZELICA, THE CREOLE; A NOVEL, BY AN AMERICAN. IN THREE VOLUMES.
#PUB# London: Printed for William Fearman, Library, 170, New Bond Street, 1820.
#COL# I 243p; II 254p; III 309p. 12mo. 21s (ECB).
#REV# ER 35: 266 (Mar 1821); WSW II: 41.
#CAT# Corvey; CME 3-628-47473-6; ECB 654; NSTC 2A10533 (BI BL).
#NOT# ER gives ‘Madame de Sansée’ as the author. For another work probably by the same author see Entry 1823: 12.
#AUT# [BARHAM, Richard Harris].
#TIT# BALDWIN; OR, A MISER’S HEIR. A SERIO-COMIC TALE. IN TWO VOLUMES. BY AN OLD BACHELOR.
#PUB# London: Printed at the Minerva Press for A.K. Newman and Co. Leadenhall-Street, 1820.
#COL# I vi, 245p; II 270p. 12mo. 11s (ECB, ER, QR).
#REV# ER 34: 509 (Nov 1820); QR 24: 276 (Oct 1820).
#CAT# Corvey; CME 3-628-47091-9; ECB 36; NSTC 2B7767 (BI BL, C, O; NA MH).
#NOT# Dedication ‘To Anybody’, signed ‘G.H.E.’. Copy at Harvard (*EC8.B2395.8206) includes author’s MS revisions, in preparation for a 2nd edn.
#AUT# BARRON, Edward.
#TIT# THE ROYAL WANDERER, OR SECRET MEMOIRS OF CAROLINE: THE WHOLE FOUNDED ON RECENT FACTS, AND CONTAINING AMONG OTHER THINGS, AN AUTHENTIC AND HITHERTO UNPUBLISHED ACCOUNT OF COURT-CABALS, AND ROYAL TRAVELS. BY EDWARD BARRON, ESQ. EMBELLISHED WITH ENGRAVINGS.
#PUB# London: Printed and published by H. Rowe, 11, Warwick-Square, Paternoster-Row, 1820.
#COL# 860p, ill. 8vo.
#CAT# NN CK.Barron; NSTC 2B9759 (NA DLC).
#NOT# Preface dated ‘January 1st’. A ‘secret history’ of Princess Caroline, and distinct from The Royal Wanderer, ‘by Algernon’, 3 vols.—see Entry 1815: 15. The copy seen is bound with the same author’s The Wrongs of Royalty; Being a Continuation of the Royal Wanderer—see 1820: 15. Collates in fours.
Further edn: 1823 (NSTC).
All entries are contained within a predetermined mask, with seven fields signalled by ‘flags’ which will be later edited out.
#AUT# is the author field. Where not known, ANON is given, and these appear arranged alphabetically by title at the beginning of each year. Where the author’s name does appear on the title-page, this is given in the form it appears there, and with additional information, such as expanded Christian names, appearing in square brackets. Where the author does not appear on the title-page, but has been identified, the name is given though contained within square brackets.
#TIT# gives accurate details of the imprint title-page, with punctuation exactly as there, though always in capital letters. In the very few cases where the first edition has not been located, and the entry reconstituted from secondary materials, this is indicated by an asterisk at the beginning.
#PUB# The first-named main place of publication is always recorded at the beginning of this field, followed by a colon: the rest of the imprint is then given exactly as it appears, though a comma is always used to precede the date.
#COL# details collation of the text. Page numbers are given for each volume, followed by format, which is determined by counting leaves between signatures. Price information is also given here, the main sources for this being review listings, and, in the case of vol. 2 of the Bibliography, the English Catalogue of Books, 1801–1836 (ECB). 
#REV# delineates review information, but—in the case of post-1800 entries especially—this is not to be taken as a guide to review material as such: there are several extant guides, such as William S. Ward’s Literary Reviews in British Periodicals (New York: Garland, 1972 etc.). Rather, it points to our main sources of information about price and date. Vol. 2 follows on from its predecessor by featuring at its start the Monthly and Critical reviews, though these are superseded in due course by the Edinburgh and Quarterly respectively. Most of the information in the latter two cases comes from the ‘Novels and Romances’ section in their lists of New Publications at the end of each number, though when full reviews are given this is noted. When additional reviews are to be found in Ward (WSW), this is indicated also.
#CAT# returns cataloguing details, and always begins with the library source and shelfmark number for the copy used. When Corvey is the source, this appears as such (i.e. ‘Corvey’); in other cases the abbreviation system employed by the ESTC is used (C for Cambridge, E for National Library of Scotland, etc.). This is followed by the ISBN number of the Corvey microfiche, where applicable. Also, in the CAT field, reference is given to the English Catalogue of Books; and to an NSTC number, along with the holding libraries listed there. The tag ‘xNSTC’ is used to indicate that a copy is not to be found in NSTC.
#NOT# provides pertinent notes on the title: the foreign source work, for example, for translated novels, is always given where known; information concerning authorship and publication history found in preliminaries is also provided here, though this practice has been approached optionally rather than systematically. Each entry ends by given a brief record of further editions: British and Irish editions to 1850; the first American edition; and the first French and German translations (with title if differing interestingly); also details are provided of modern facsimile editions.
Since 1990 Cardiff and Paderborn teams have been working to compile entries for every novel known to have existed between 1800 and 1829 inclusive. As a rule, we have excluded non-standard works such as miscellanies, shorter tales, children’s literature, and religious tracts. Usually, the copy at Corvey was used for our entries, unless there are good reasons for not doing this (e.g. an imperfect copy, not a first edition). In cases where Corvey couldn’t provide the copy (approximately 28% of cases), we have usually gone first to the British Library, then (if not available there) to the other main copyright libraries (Bodleian, National Library of Scotland, Trinity College Dublin, etc.). An invaluable help here was NSTC (the second series of which (1816–70) was being completed as we ourselves were proceeding, and now, of course, both series I and II are available on CD-ROM). Having a fullish core collection held electronically also allowed cross-checking of the titles given as ‘by the author’ in our #TIT# line, which produced a number of fresh novels then locatable through online facilities such as Blaise and the OCLC database. A fast laptop (a thing unheard of when we started) also aided stack checks against specialist collections at Aberdeen and Bristol. When a title was not locatable in Britain and Ireland, we turned to the USA with our list of remain titles to be found. A full check was made against the card index file at the Houghton Library, Harvard University, left by Sidney Greenough—a kind of American Block. There were also visits to other US libraries, notably to the large holdings at Urbana (Illinois), and the Sadleir-Black collection at the University of Virginia. By now we were experiencing diminishing returns, and thirty or so titles that we know to have existed but were unable to find have been ‘reconstituted’ from available secondary evidence.
The total file (all three decades) is now closed, with 2,256 titles in all for the years 1800–29 inclusive. This research has been used as the basis for the creation of a dynamic database, designed by Anthony Mandal in collaboration with Peter Garside, using Microsoft Access 97—this time employing the latest technologies, rather than other, quainter methods! It is now possible, for the first time, to draw statistical conclusions from what we have gathered together over the past two decades. As well as this, the database is fully searchable, and contains complete details of the texts, and additional material which it was impossible or unnecessary to include in the original bibliography. Such fields as authorial status (whether the author published pseudonymously, anonymously, or with an authenticated name), publishing concerns (enabling the analysis of the fortunes of the major publishing houses over the three decades), and full gender categorisations have been included. The second phase is underway, which includes ascription to each entry of Genre/Style (e.g. Gothic, Historical, Domestic, etc.) and Narrative Structure (e.g. Epistolary, Direct Narrative, etc.), as well as pricing statistics for the various volume-sizes of the novels.
Examples of what is being achieved with this combination of dedicated scholarship and modern advances in IT and data-manipulation is provided in some of the sample graphs contained below.
This graph shows that the new level of output achieved during the late 1790s was largely sustained during the 1800s. Output in the 1810s, however, is 3% lower than in the 1800s, but then builds up into the period of most output during the earlier 1820s. One particularly noteworthy feature is the fact that the largest year of all is not 1810, but 1808. Another interesting figure is the trough which occurs midway in the 1810s (the time of the earliest Scott novels). Why? This could be a lot to do with the cost of paper, which rose during this period. Also noteworthy is the disruption of the steady rise in the 1820s, as novel production was buffeted (but not seriously dented) by the financial ‘crash’ of 1826 in the publishing trade.
Fig 3 compares women’s novels as published by the most prolific concerns of the period 1800–29: Minerva, Longmans, Henry Colburn, J. F. Hughes, and Whittakers. Here, works by women are taken to include female-implied (‘By a Lady’, ‘By Lady —’, etc.) and pseudonymous novels, as well as those whose authors are named on the title-pages or have been subsequently identified. The blue cylinders show women’s output as a percentage of the total production by each concern. The magenta cylinders again are percentages of the publishers’ total output, indicating how many novels written by women were published anonymously (i.e. no name or variant thereof given on title-pages). What is of interest is the similarity between Minerva and Longmans, the top two publishers: both have a female authorship of between 54% and 55.5%, of which only around 15% were published anonymously. Compare this with the more male-inclined Henry Colburn and Whittakers, which only had a female authorship of around 30%, half of which was published anonymously. Finally, J. F. Hughes, a far more controversial figure, while not having such a high total output of female-penned works as Minerva or Longmans, only published around 4.3% of his female authors anonymously.
What this might be taken to indicate is that the publishers of the earlier part of the period (most significantly, Hughes) were more inclined to openly publish female-penned works as such, while concerns which appeared at a later date (Colburn did not begin publishing until 1807) not only published fewer works by women, but were even less inclined (proportionately speaking) to advertise when they did publish them.
The final graph presents details of the output of the Minerva concern (under William Lane, Lane, Newman & Co, and then A. K. Newman & Co) over the three decades. The gender divisions represent the status of the works as deduced from the title-page (see above). The terms correspond with the Bibliography’s #AUT# field: ‘Named’ texts indicate an authenticated name given in the title-page; ‘Identifed’ texts are based on scholarly discoveries, deduced associations (i.e. ‘By the author of …’), or established ascriptions; ‘Implied’ texts are based on unidentified pseudonymous works or those whose title-pages indicate some gender type (e.g. ‘By a Lady’, ‘By a Reverend’, etc.); ‘Unknown’, unsurprisingly, indicates no further substantial knowledge of authorial gender is available. What is noteworthy of Minerva’s output over the three decades is the precedence of Female Named texts, which is in keeping with the data of Fig 3. Furthermore, female-penned works outnumber male-penned works by 23% (approximately 121 novels), although the average output of women’s novels had dropped from 64% in the latter half of the 1810s to around 51% during 1825–9. Of their average yearly production of novels, Minerva managed just over twenty-one books per year in the 1800s (best year, with twenty-eight novels, was 1805), compared with slightly more than fourteen per year in the 1820s (worst year, with nine novels, was 1829). Perhaps this emphasis on the female market explains the demise of Minerva in the 1820s, with their rather ‘unfashionable’ touting of female-penned works at a time which anticipated the dominance of the Victorian male author in the light of Scott’s phenomenal achievement.
One last point brings us finally back to Scott. As noted in Fig 2, as far as the output of new titles is concerned, 1810 it turns out is not the optimum year; rather, it is 1808, where we have a clear peak of 111 titles. Of these, forty-one are by male authors, fifty by women, and twenty remain to be identified. The titles themselves are a mixed bunch—at the head used to be Atrocities of a Convent, until examination of the only surviving copy at UCLA revealed that it was authored by Thomas Rickman (and is more a radical satire than a Gothic potboiler). There are, however, a few genuinely salacious titles—The Noble Cornutos, The R—l Stranger, The Royal Sufferer, Royal Intrigues—reflecting a royal scandal (between the Prince Regent, later George IV, and his wife, Princess Caroline) not unlike the one current now. In the middle of this squarely sits Hannah More’s Cœlebs—an attempt at moral re-armament from within the enemy camp if ever there was one!  Also among the few more heavy titles is Joseph Strutt’s antiquarian novel, Queenhoo Hall, which was completed by Scott for the publisher John Murray.  Murray was then angling for a share of Scott’s poetry; Scott was trying to offload a number of projects, including a collected set of British Novelists, which were meant to keep James Ballantyne’s press busy and help launch John Ballantyne as an Edinburgh publisher (both concerns were secretly owned by Scott). Early in 1809 James Ballantyne (acting as Scott’s literary agent), met Murray at Boroughbridge in Yorkshire and took a memorandum of the meeting which has survived. One project itemised there is ‘New poem’, against which is recorded Murray’s comment, ‘Most certainly’. Another, called ‘Anonymous work’, is undeniably Waverley. Scott in this new light can thus be seen embarking on a career as a novelist at a time when fiction, if at a low ebb in reputation, was undeniably reaching a high watermark in terms of output.
1. Lee Erickson, ‘The Poet’s Corner: The Impact of Technological Changes in Printing on English Poetry, 1750–1850’, in English Literary History 52 (1985), 902. Reprinted in his The Economy of Literary Form: English Literature and the Industrialization of Publishing, 1800–1850 (Baltimore: Johns Hopkins Press, 1996).
2. A second edition of Scottish Chiefs (1811) can be found in the Corvey Microfiche collection at Cardiff, ISBN CME 3-628-48361-1.
3. Q.v., Peter Garside, ‘J. F. Hughes and the Publication of Popular Fiction’, Library 9 (1987), 240–58.
4. Cardiff University holds both the NSTC (1st and 2nd series) and NUC for inspection on the ground floor of the Arts and Social Studies Library. The ESTC is also avaliable for consultation at any of the computer bays in the ASSL.
5. The English Catalogue of Books, Preliminary Volume, 1801–1836, edd. Robert Alexander Peddie and Quintin Waddington (1914; New York: Kraus Reprint Corporation, 1963).
6. So rapid was its success, this title was not acquired for the Corvey collection until the 14th edition of 1813! Again, this is available for inspection, under CME 3-628-47303-9.
7. Queenhoo Hall is also held on the Corvey microfiche: CME 3-628-48681-5.