I

Such revisionist ambitions can only be countenanced because of the brilliance of the scholar whose labours they build upon. Robert Darnton's contributions to the fields of pre-revolutionary French history, book history, sociology, the history of ideas and, more recently, digital humanities have been profound and inspirational. It is testimony to the range of his work that the research presented here, which concentrates on the archives of the STN, directly relates only to a section of his path-breaking academic output.Footnote ¹ Building upon the publications of Jean Jeanprêtre, Charly Guyot, and Jacques Rychner, Darnton brought wider international recognition to the STN papers with his important 1968 and 1971 articles ‘The Grub Street style of revolution: J.-P. Brissot, police spy’ and ‘The High Enlightenment and the low-life of literature in pre-revolutionary France’.Footnote ² These pieces laid the foundations of his ‘Grub Street’ thesis, describing the pre-revolutionary trials and tribulations of writers like Jacques-Pierre Brissot and Charles Théveneau de Morande as they lashed out in desperation against the twin evils of the Old Regime and the entrenched mediocrity of the post-Voltairean High Enlightenment of the 1770s and 1780s. Their arguments were revisited, expanded upon and confirmed in Darnton's outstanding 1982 collection The literary underground of the Old Regime. Readers were treated to a parade of the most colourful and exceptional characters to be found in the STN dossiers: booksellers, printers, pamphleteers, and peddlers with remarkable stories that, taken together, revealed a compelling composite picture of a world flooded with illegal literature. This ‘sketch of characters in motion’ seemed almost stranger than fiction. Mercifully, because it was accompanied by the promise of forthcoming systematic study, it required only a temporary suspension of disbelief.Footnote ³

True to his word, Darnton presented the deferred statistical evidence in his acclaimed 1995 monograph The forbidden best-sellers of pre-revolutionary France.Footnote ⁴ Rather than trusting in the importance of the canon of classics venerated and endlessly reshuffled by dusty professors, he insisted, we should endeavour to get closer to the reading experiences of eighteenth-century Frenchmen by examining what they actually read. Counting books was thus thrust to the heart of the argument, and Darnton had counted a lot of books. A mere sampling of the STN's illegal trade in France had yielded a total of 28,212 copies requested in 3,266 orders. The results of this survey were considered sufficiently important to be presented in a companion volume of statistics, The corpus of clandestine literature in France, 1769–1789. This outsized appendix listed 720 illegal books that, taken together, offered a ‘fairly complete’ view of the underground literature available. Moreover, the relative importance of each its titles, based on STN demand, could be gauged. The resultant ‘best-sellers’ tables – of titles, authors, and genres – were trumpeted as having as much accuracy as their modern equivalents.Footnote ⁵ The works that populated their upper echelons validated, modified, and extended Darnton's previous studies. The pre-revolutionary French book market appeared awash with texts by relatively uncelebrated authors like Louis-Sébastien Mercier, Pidansat de Mairobert, the baron d'Holbach, and Morande. There were of course naysayers. Jeremy Popkin wrote an insightful early review, and 1998's The Darnton debate contained much cutting, if largely theoretical, criticism.Footnote ⁶ But the doubters could not compete with Darnton's archival legwork. In eleven summers and two winters spent in distant Neuchâtel he had read 50,000 letters, found 720 illegal books and recorded 28,212 copies ordered. The weight of numbers – the ‘hard evidence’ as Malcolm Cook put it – was on Darnton's side.Footnote ⁷

These numbers warrant some explanation. Darnton's 720 books were compiled from four sources: a sample of the STN's orders, publishers' catalogues of illegal stock, inventories of titles seized in police raids on bookshops, and the registers of Parisian customs confiscations. Of the 720 titles, 457 were ordered from the STN, and thus could be ranked as ‘best-sellers’ according to demand. Anything not ordered was by methodological implication presumed to be in little demand. The size and range of Darnton's sample appeared to justify such an approach. His orders were drawn from the records of sixty-three booksellers whose businesses collectively covered the four corners of France, from Marseilles to Bordeaux, Rennes to Nancy, as well as most points in between. These traders varied in size from itinerant peddlers like Planquais and Blaisot, who roamed the Loire valley, to major established dealers like Rigaud, Pons & Company in Montpellier and Charmet (and later his widow) in Besançon. Although some of their number only requested books once or twice, many were in regular contact with the STN and, collectively, their orders spanned the two decades before the French Revolution.Footnote ⁸ Whilst the STN's order books, the livres de commissions, would have proved an invaluable starting point to extract the data relating to these clients, they cover neither the whole period nor all of the traders in Darnton's sample. To arrive at the complete picture, he had to work through the dossiers individually, extracting orders from these booksellers' messy, incomplete, and sometimes incomprehensible letters. This was formidable archival work.

It was also bold. To recreate pre-revolutionary France's forbidden best-sellers from the records of a lone Swiss publishing house without recording a single book sale required two leaps of faith. Supply had to equal demand and the STN archives needed to be representative. Darnton admitted to having lost sleep over these issues, yet the eventual courage that he showed in his methodological convictions was reassuring. He substituted ‘ordered’ for ‘sold’ throughout his twin manuscripts; he insisted that the STN's sales did not excessively favour their own publications; he assured readers that, because the STN succeeded in fulfilling the vast majority of their orders, his statistics showed ‘supply as well as demand’.Footnote ⁹ To be clear, he claimed that his list betrayed what illegal books went into bookshops everywhere in France.Footnote ¹⁰ This sureness rested upon two pillars. First, because a large part of the print-runs of new editions were immediately swapped amongst allied publishing houses, the stock available to eighteenth-century publisher-wholesalers tended to be extremely similar. The overlap between the STN's catalogues and those of other publishers, as well as the fact that the society accumulated over 1,500 works by 1787 despite having printed less than 300, supported this view. Indeed, Darnton argued, a kind of invisible, floating stock was available to all major border traders from the Low Countries to Switzerland. By placing an order with one or two of them, retailers within France could secure whatever they wanted. Second, because these foreign wholesalers did not allow returns, their French clients tended to order only what they could guarantee to sell.Footnote ¹¹ In short, the STN could supply almost anything, and its clients only ordered saleable stock.

But, through their efficient trading in the taboo, could these Neuchâtelois, alongside their allies and competitors, have contributed towards causing the French Revolution of 1789? The question exists because so many revolutionary actions and discourses, from patriotism to paranoia, appear rooted in ideas taken from books. Of course, pinning the chaotic events of the Revolution to the previous reading matter of its actors seems, at times, mildly absurd. As the butcher Desnot hacked off the head of the Bastille's governor the marquis de Launay with his pocket knife, was he thinking of Mercier? Probably not, but, so the story goes, the radical ideas circulating throughout the late Enlightenment French public sphere might have at least made his actions thinkable.Footnote ¹² The value systems of the would-be revolutionaries had somehow slipped out of step with the political realities of the Old Regime. As such, the question of what Frenchmen read before the French Revolution, ever since Daniel Mornet first raised it academically in 1910, has been of central historical significance.Footnote ¹³ The STN archives, the most complete to have survived, offer an excellent opportunity to address this question.Footnote ¹⁴ It is surprising, therefore, that for all the controversy Darnton's work has provoked, very little has been written about his methodological approach towards, and representation of, these papers. In Le rayonnement d'une maison d’édition, the published acts of the 2002 colloquium organized by Darnton and Michel Schlup, a parade of excellent historians discuss the STN's trade from Brussels to Lunéville, Poland to Italy, without ever quite addressing fundamental questions of the meaning and representativeness of the datasets that they offer.Footnote ¹⁵ General histories of the period too, Thomas Munck's recommendable The Enlightenment: a comparative social history, for example, uncritically repeat Darnton's statistics as the ‘real best-sellers’ of the illegal trade.Footnote ¹⁶ Digging deeper into these figures, however, reveals some surprising results.

II

Despite having been ordered, 144 of the 457 ‘STN’ works in Darnton's Corpus of clandestine literature never passed through the hands of the Neuchâtelois booksellers during their entire trading existence between 1769 and 1794. Frédéric-Samuel Ostervald, Jean-Elie Bertrand, and their colleagues never saw Les aventures de la madonna et de François d'Assise; they may have never heard of Le passe-partout de l'église romaine; they traded in neither Le chien après les moines nor its apparent table-turning sequel Les moines après le chien. For sure, these 144 works do not include the STN's supposed underground top-sellers, but they do include a number of works that registered significant demand. 100 copies of La foutro-manie were ordered and never sent, 100 of Les plaisirs de tous les siècles et de tous les ages, 76 of Voltaire's L’évangile du jour and 75 of Les entretiens de l'autre monde. To put these numbers in perspective, 100 copies would have been enough to appear joint 75th on Darnton's master best-sellers table, had it not been cut short at number 74.Footnote ¹⁷

If nothing else, this initial foray into the figures suggests that the relationship between supply and demand warrants investigation. First, however, a word about the methodologies that lie behind these numbers. Darnton's intuitions concerning supply matching demand are supported by archival evidence of two sorts. The STN's sales patter to certain distant booksellers, found throughout the dossiers, included claims to be able to supply all known works.Footnote ¹⁸ Such claims often accompanied dubious promises that orders would be supplied on time and in good condition, that editions would be printed on the finest quality paper, and that their prices could not be beaten. These were things that needed to be said, even if they were to be taken by book trade insiders with a pinch of salt. Nonetheless, the claims were made. In addition, the relationship between demand and supply is explicit in the STN's order books, the livres de commissions.Footnote ¹⁹ On the left-hand folio, the STN recorded orders as soon as they were received by post; on the right-hand folio, once the shipments had been prepared, they registered the actual works sent out in response to each commission. The extent of the correlation, unfortunately, appears to be within the eye of the beholder. For Darnton they usually matched. For this author, who has also run through the entire set, this is rarely the case.

Thankfully, our differing visual impressions are of little consequence because the supply and demand question can be approached systematically. The French Book Trade in Enlightenment Europe (FBTEE) project has reconstructed the STN's entire knowable book dissemination and acquisition business from the firm's 1769 establishment to its final transactions in 1794. To do this, we have not focused upon the livres de commissions, which are incomplete in several ways, but instead have taken a holistic approach to the archive, employing the STN's daybooks, stock books, client accounts, order books, and, where necessary, letters. Whilst there are significant lacunae in each of these sources, the STN's double-entry accounting system ensures that where one document is lost its data can almost always be recovered elsewhere. Playing an extraordinary game of archival hopscotch, and endlessly testing the patience of the Neuchâtel archivists, it has been possible to extract the distribution records of over 400,000 examples of the nearly 4,000 identifiable editions that the STN traded. These numbers include the entirety of the society's dealings with France. By comparing this new supply dataset with Darnton's demand sample, allowing compensations for the different methodologies, we can begin to understand better the complex relationship between the titles that were ordered and the books that were sent.Footnote ²⁰

The first lesson may be considered timeless – never take salesmen at their word. The majority of the 144 never-sold illegal works can be put down to French booksellers calling the STN's bluff. An order that the agent Malherbe forwarded from his client ‘Desbordes’ in September 1772, for example, listed a series of forbidden books, including the Le passe-partout de l’église romaine and Mercier's Jean Hennuyer, évêque de Lisieux, not to be found in the Neuchâtelois' illegal catalogues. The request also contained several of the baron d'Holbach's anti-clerical and materialist titles that were only available in Dutch editions at the time.Footnote ²¹ As was their wont, the STN simply never realized the order. Similarly unfulfilled requests for never-stocked works were received from all over France: from Pavie in La Rochelle (L'homme dieu, Le superstitieux); Charmet in Besançon (Histoire d'une détention de trente-neuf ans, Le philosophe ignorant); Manoury in Caen (Aux mânes de Louis XV, Lettres … du chevalier d'Eon); Blouet in Rennes (Timée de locres, Le bannissement des jésuites); Chevrier in Poitiers (Discours historique sur l'Apocalypse, Lucrèce); and from elsewhere. The champion, however, was undoubtedly Mauvelain in Troyes who chanced his hand at scores of such works, from Jean-Baptiste Renoult's Les aventures de la madonna et de François d'Assise to Jean-François Cailhava de l'Estendoux's Le soupé des petits-maîtres. Many of the most scandalous titles in Darnton's list, including the entirety of the much-contested pre-revolutionary Marie Antoinette corpus (Les amours de Charlot et Toinette, Essai historique sur la vie de Marie Antoinette d'Autriche, Passe-temps d'Antoinette, and Vie d'Antoinette) were ordered by and never sent to Mauvelain.Footnote ²² None of the above works were ever traded by the STN. Some of the more obscure titles booksellers ordered, such as the Requête au roi pour la suppression des moines, probably never existed.Footnote ²³

Seventy further forbidden works that definitely did exist, and were recorded by Darnton as registering no demand from the STN, were traded by the society. Take, for example, James Rutledge's Le bureau d'esprit. The corpus of clandestine literature records no ‘sales’ of this work; it is listed due to its confiscation by the Parisian police in 1780.Footnote ²⁴ Darnton gives its publication details as ‘Liège, 1776’, noting that at least one other edition was published by the following year. The ‘Liège, 1776’ edition, however, was a false imprint, realized in Neuchâtel by the Société typographique themselves. The Neuchâtelois printed 1,143 copies of this edition, a relatively large run, on 4 March 1777. By 15 February 1778, less than a year later, they had only 448 copies left in stock. In the intervening period they sent the work to Besançon, Lyons, Marseilles, and many other towns across France and Europe. They sent copies to five traders in Darnton's sample: Bonnard in Auxerre, Charmet in Besançon, Caldesaigues in Marseilles, Gerlache in Metz, and Manoury in Caen. Similarly, Jacques Le Scène-Desmaisons's Contrat conjugal, ou loix du marriage de la répudiation & du divorce, appears in Darnton's list only because it was once confiscated in Versailles. The STN printed 4,166 copies of this text in September 1782 and sent all but 25 of these directly to Pierre Duplain l'aîné in Paris. A final example: Georg Jonathan von Holland's Réflexions philosophiques sur le système de la nature, a Protestant-leaning response to d'Holbach's atheism, registered no demand and a single confiscation in Darnton's survey. The work was printed twice by the STN, as witnessed by a chunky dossier describing its publishing history in vivid detail.Footnote ²⁵ The society sold 1,229 copies of the work, 219 of which were dispatched directly into France.

The 214 cases (144 plus 70) counted above represent only the extremes. More sophisticated analysis further exposes the extent of the gap between supply and demand in the STN accounts. Many works listed in The corpus of clandestine literature, such as Louis-Gabriel Dubuat-Nançay's Les maximes du gouvernement monarchiques, were traded by the STN, but never to France. In this particular case, the society's travelling salesman picked up a copy from Emmanuel Tourneisen in Basle on 21 July 1787, which he promised to Charles-Marie Toscanelli in Turin on 15 December 1787. The only other copy the STN ever saw, of unknown providence, was offered to Wolfgang Gerle in Prague the following August. Other works on Darnton's list, for example Henri-Joseph Du Laurens's Abus dans les cérémonies, were not traded until after 1789, and even then never to France. Scores of illegal works that the STN did sell but that do not appear at all in Darnton's survey, including Delauney's Histoire d'un pou françois and Jean-André Perreau's Le roi voyageur, must also be considered. Foremost amongst such omissions are the remarkable, and highly illegal, books and pamphlets that the STN printed for the Genevan economist Théodore Rilliet de Saussure as he tried to restore his fallen reputation by accusing his former wife of having admitted, on their wedding night, to having previously borne an incestuous love-child to her brother. Rilliet's accusation was especially spicy because the sibling in question was the baron de Planta, a shady character protected by the powerful cardinal de Rohan. The STN printed more copies of Rilliet's slanderous publications than Darnton recorded demand for all 720 of his forbidden books combined.

Unsurprisingly, Rilliet's scandalmongering rides high in the FBTEE database's new STN best-sellers tables based on sales. In all their details, Darnton's demand statistics bear little resemblance to these new lists. D'Holbach's Histoire critique de Jésus-Christ, for example, comes 18th on his master table, registering total demand of 327 copies in 36 orders and the full gamut of catalogue appearances and police and customs confiscations. Darnton's genre-specific tables show it to have been the second best-selling pre-revolutionary anti-religious treatise. The STN actually traded in a maximum of 55 copies of the work, several of which were not sent to France. It was their 927nd best-selling title overall and their 516th to France. Thérèse philosophe, one of the centrepieces of The forbidden best-sellers, registered similarly disappointing sales. It was the STN's 438th best-selling work, representing about 0·03 per cent of their combined legal and illegal sales. It was their 353rd strongest selling work to France across the period. Such wild variations between the demand survey and the STN's actual sales are typical of most books listed within The corpus of clandestine literature in France. Certain classics like Voltaire's Candide shoot up the new lists; the (regrettably) much-discussed Morande disappears almost entirely from view; STN printings including L'homme sauvage and Soldat citoyen rise inexorably; many of the baron d'Holbach's largely Dutch-produced anti-religious texts descend towards apparent obscurity. Yet, if we temporarily overlook easily explained outliers like the Rilliet texts, some Necker pamphleteering, and Protestant works of marginal legality, the STN's overall top-ten clandestine best-sellers read:Footnote ²⁶

1. 1. Tableau de Paris, Louis-Sébastien Mercier – 14,065 (sales)1

2. 2. Contrat conjugal, Jacques Le Scène-Desmaisons – 4,164

3. 3. Histoire des deux Indes, Guillaume-Thomas-François Raynal – 3,684

4. 4. Questions sur l'Encyclopédie, Voltaire – 2,511

5. 5. Dissertation sur l'établissement de l'abbaye de S. Claude, Charles-Gabriel-Frédéric Christin – 2,504

6. 6. Requête au Conseil du Roi, Simon-Nicolas-Henri Linguet – 2,201

7. 7. Dieu, Voltaire – 2,187

8. 8. Système de la nature, D'Holbach – 1,972

9. 9. De la verité, Jaques-Pierre Brissot de Warville – 1,930

10. 10. L'an 2440, Louis-Sébastien Mercier – 1,904

A more limited enquiry of illegal sales to France reads:

1. 1. Contrat conjugal, Jacques Le Scène-Desmaisons – 4,164

2. 2. Histoire des deux Indes, Guillaume-Thomas-François Raynal – 2,488

3. 3. Tableau de Paris, Louis-Sébastien Mercier – 1,915

4. 4. Dissertation sur l'établissement de l'abbaye de S. Claude, Charles-Gabriel-Frédéric Christin – 1,806

5. 5. Requête au conseil du roi, Simon-Nicolas-Henri Linguet – 1,599

6. 6. Les confessions, Jean-Jacques Rousseau – 1,480

7. 7. De la verité, Jaques-Pierre Brissot de Warville – 1,460

8. 8. Théorie des loix criminelle, Jaques-Pierre Brissot de Warville – 1,115

9. 9. Un independant à l'ordre des avocats, Jaques-Pierre Brissot de Warville – 1,038

10. 10. Le soldat citoyen, Jacques Antoine Hippolyte de Guibert and Joseph de Servan de Gerbey – 1,030

Specialists will remark that these top-tens do not wildly differ from those presented in The forbidden best-sellers and The corpus of clandestine literature. Mercier, Voltaire, d‘Holbach, Raynal, and Brissot all occupy plum positions. L'an 2440 was indeed a STN clandestine favourite. Supply, despite all the evidence presented above suggesting the contrary, appears here to have relatively closely mirrored demand.

III

At this point, the current author must also admit to having passed several uncomfortable nights intrigued and concerned. Two questions weighed heavily. First, why, when the middle and lower reaches of Darnton's list seemed to bear no relation to the STN's actual clandestine sales, were many of the ‘best-sellers’ holding relatively firm? Second, more worryingly, did not the extent that the supply and demand link had been broken simply undermine a supply-led methodological approach to the archives? These new supply-side figures are riddled with biases. They favour works available when the STN's business was at its peak around the turn of the 1780s. They contain batch printings like those produced for Rilliet, which ‘sold’ from the STN's perspective, but which were sent almost exclusively to the author and, in this particular case, hardly reached market due to his suspicious and untimely death. They speak of every coup de génie and failure of business acumen the Neuchâtelois made over twenty years. Thiers, we might do well to remember, was a typographical society that had racked-up 75,000 livres-worth of bad debts by 1785.Footnote ²⁷ Demand, pure and representative demand of the sort that Darnton claims for the archive, by contrast, is a much more attractive proposition for the general historian. Knowing for which illegal books French readers yearned establishes a significant departure point for an investigation into the pre-revolutionary state of mind. It is the Holy Grail of bibliometric research for this period.

Alas, it is a fantasy to suggest that anything approaching representative demand can be gleaned from the STN archives. The key to resolving my twin apprehensions lay in a strikingly simple realization: booksellers ordered STN books from STN catalogues. Of course the books that the STN heavily promoted and constantly stocked both sold well and were much ordered! Any attempt to uncover general French demand from the STN archives is destined instead to reveal only a noisy and imprecise echo of what Gallic booksellers through the Swiss could supply. As a consequence, the actual sales figures of the STN, despite their biases, appear preferable to the historian. To understand the extent of the corruption in the demand data, it is necessary to meditate for a moment on how the eighteenth-century book trade functioned.

By standard industry practice, printer-wholesalers and retailers periodically drew-up catalogues of their stock holdings, many of which ran to hundreds of pages offering thousands of volumes.Footnote ²⁸ In the STN's case, at least, these catalogues were created during the stock-taking process, and as a consequence were complete right down to the inclusion of obscure lone texts that had been gathering dust in their storeroom for years. Once compiled and printed, traders sent their catalogues to clients across Europe by the post or, whenever possible, had one of their agents or travelling salesmen deliver them personally to guarantee receipt and encourage orders.Footnote ²⁹ If they were lucky, clients made commissions based upon the contents of these catalogues and, presuming the desired works were still in stock, they were duly sent out.

Two important clarifications, however, are necessary to complete this picture. First, when booksellers did not order from the STN's legal or illegal catalogues, they did so from circulars or indications of new or forthcoming stock that the society made in individually addressed letters. Indeed, once a relationship was established between wholesaler and client this was the most common way of trading. The personal touch was preferable because printed catalogues, however regularly supplemented, became quickly out of date and never indicated how many copies of works were available. Eventually, the STN semi-formalized this stock-indication system by creating numbered lists of ‘nouveatés’ to be copied out at the end of each letter that they sent. At any given time several of these lists would be current, and the appropriate one could be chosen to suit the predicted needs or wants of its recipient.Footnote ³⁰ Regular correspondents thus knew the broad contents of the STN's stockroom from their printed catalogues, and which new works the society was printing and receiving in more-or-less real time. Second, for fear of reprisals, the STN deliberately omitted certain highly subversive works that tended to be illegal everywhere in Europe from their printed catalogues. On at least one recorded occasion they advertised such works together in a hand-written catalogue of ‘livres philosophiques’, which they sent out to selected clients alongside their standard catalogue.Footnote ³¹ Darnton located a handful of comparable Swiss hand-drafted ‘livres philosophiques’ catalogues for his study, demonstrating this to have been common industry practice.Footnote ³² Otherwise, the STN simply indicated these exceptionally illegal works to their network of clients by private letter.

Neither of these elucidations changes the fact that clients generally ordered books that they expected the Neuchâtelois held in stock. Nor does another potentially misleading anomaly regarding how the STN handled certain contraband editions. Their 1771 counterfeit edition of d'Holbach's atheist and materialist Système de la nature angered the local community of pastors, and the ensuing political fall-out saw their directors stripped of local political office.Footnote ³³ Once bitten, the Neuchâtelois were more careful with how they dealt with illegal books, and for some time they stopped printing them and avoided holding large stocks. Instead, the society established relationships with allies including François Grasset in Lausanne, Barthélemy Chirol, and Gabriel Grasset in Geneva, as well as with Samuel Fauche in Neuchâtel.Footnote ³⁴ They advertised some illegal works held by these nearby traders as if they were in stock, knowing that they could either pass on the orders or fulfil them without severely holding up the supply chain.Footnote ³⁵ This, it must be emphasized, was an unusual and temporary arrangement that effectively amounted to outsourcing stock holdings rather than meeting unsolicited orders. The famous STN ‘livres philosophiques’ catalogue of 1775, evoked in numerous places as evidence of the society's illegal trade has thus been misunderstood.Footnote ³⁶ Yes, the STN dispatched this hand-written list of 110 highly illegal works to trusted clients alongside their standard catalogue. Yet, they neither stocked nor sold listed works like Nicolas Fromaget's erotic fiction Le cousin de Mahomet, Jacques-Antoine-René Perrin's libertine Les egaremens de Julie, or Jean-Jacques Rousseau's Contrat social around this time. They could simply get hold of them.

Admittedly, on occasion the STN were willing to procure works to order. Such cases rarely included illegal works and instead largely fall into two categories: highly specialized scientific, medical or trade-oriented books sourced for Neuchâtelois clients, or Parisian nouveautés purchased for the local literary society. The Société littéraire de Neuchâtel, which the STN director Samuel Ostervald helped organize, took about 250 such titles in 46 separate orders between July 1771 and November 1778. Many of the works – novels like Henriette Wyndham, ou La coquette abusée or the Anecdotes de la cour et du règne d'Édouard II – were purchased from the Parisian bookseller Pyre of the rue St Jaques soon after their first publication. This link was vital for the literary society: the Swiss publishers were adept at cherry-picking popular new books to counterfeit, but the choice that they offered was insufficient for true bibliophiles. Professionals had equal cause to demand works available only outside the usual channels. The STN supplied their locally based client ‘Vernier’, for example, with the latest editions of the Journal de médicine from Paris, as well as a number of specialist medical texts. Only from the French capital could such remarkable titles like Guillaume-René Lefébure's treatise on venereal disease combined with a recipe for a chocolate aphrodisiac, another that Vernier took, be sourced!Footnote ³⁷

Perhaps the needs of Vernier's patients were sufficiently pressing that the STN took pity, for in general such business was unprofitable. Chasing after obscure texts normally only made sense for Neuchâtelois clients, and then only to show goodwill on behalf of the society or its directors. Indeed, the rarity of the practice throws new light on the nature of the STN's business and the wider eighteenth-century francophone book trade. For sure, as has previously been emphasized, the STN's networks were sufficiently vast that they could procure anything from agents or allies in Amsterdam or Paris, Brussels, or London.Footnote ³⁸ Yet, the book market was extremely competitive, and time was money. Sending out feelers to distant publishing houses in search of obscure titles was a time-consuming, costly, and downright impracticable process. Enquiries and negotiations could take months. Should the requested edition be located, it would have to be sent to Neuchâtel by prohibitively expensive channels without benefiting from the economies of scale of bulk shipments. Large commissions might be delayed or lost on a wild goose chase. Much better practice was quickly to supply titles that were to hand, and ignore those ordered more in hope than expectation. Perhaps, if the directors knew certain editions were readily available in Yverdon, Geneva, or Lausanne, or better still chez the Fauche clan in Neuchâtel, they might seek them out, but there came no guarantees.

For book historians, better understanding the STN's business model points towards two conclusions: the book trade was more regionally variegated than has typically been recognized and, as a consequence, the Neuchâtel archives speak less of distant markets than previously hoped. Darnton's argument that, due to their mutual dependence and frequent swapping of printed sheets, the printer-booksellers of London, Amsterdam, and Neuchâtel quickly finished up with much the same stock does not bear close scrutiny. To root this point in his empirical work, he insisted that his STN ‘best-sellers’ lists showed no signs of excessively favouring the society's own publications. Indeed, he explained, the Neuchâtelois sold far more copies of Mercier's L'an 2440, which they did not print, than of his Jezennemours, which they did.Footnote ³⁹ While this example is true, Mercier's futuristic fantasy was alone amongst their actual top-50 best-sellers to be not at least partially printed in Neuchâtel. Moreover, its appearance can be entirely explained by the fact that the STN's close collaborator and supplier François Grasset, based in nearby Lausanne, published editions of the text in 1772 and 1774. The only other non-Neuchâtel edition to penetrate the top 100 – Mairobert's Anecdotes sur Mme la comtesse du Barry – was also printed by Grasset in Lausanne. Respectable selling non-Neuchâtel editions further down the lists – Samuel Auguste André David Tissot's Onanisme at 157th for example – were invariably published in French-speaking Switzerland. Put simply, the Swiss traded in Swiss editions and, as such, the STN archives are only representative of the Swiss trade.Footnote ⁴⁰

The thoroughness of the new sales dataset, juxtaposed against other sources, confirms this picture by indicating the considerable volume of non-Swiss works that the STN never traded. The Neuchâtelois only handled a fraction of the French editions available in Europe between 1769 and 1789. Pierre Conlon's Le siècle de lumières lists 37,837 original editions (thus excluding counterfeits and reprints) published between 1769 and 1789. Throughout their entire trading history, the STN saw only about 4,000 different editions, two-thirds of which they traded in less than a half-dozen copies. Admittedly, the ratio of forbidden texts handled by the Neuchâtelois was higher than permitted ones. Nonetheless, Darnton's own work on various confiscations lists revealed 263 illegal titles never ordered by his sample of booksellers. To these we might add the above works ordered but not sold, and a potentially much larger number of books neither ordered nor sold. It is currently impossible to determine this number, but the scores of illicit titles that the Parisian bookseller Poinçot found warehoused in the Bastille in 1789 – La none éclairée, ou les délices du cloître (1774) and Les nymphes de Châteldon et de Vichy (1785) to name but two – indicate its significance.Footnote ⁴¹ Thus, especially given the STN's low trading volumes and sporadic holdings for most titles, a bookseller ordering a random contemporary edition, legal or otherwise, had only a miniscule chance of it being in stock. Even Mauvelain eventually grasped a more nuanced understanding of the trade and stopped playing such unappealing odds.

We might follow his lead and move towards a polycentric understanding of the eighteenth-century book trade. A hypothesis can be suggested for the illegal sector. Two distinct zones of clandestine production emerged beyond France's borders, profiting from relative press freedom and corrupt entry points. The first of these was centred upon the Swiss Romande – notably the towns of Geneva, Lausanne, Yverdon, and Neuchâtel – which grew in importance throughout the century and benefited from Protestant traditions, the vitality of the Lyons–Basle thoroughfare, low labour costs, and a plentiful paper supply.Footnote ⁴² The concentration of printing-shops and auxiliary services situated there permitted the frictionless swapping of books between houses.Footnote ⁴³ Traders exchanged the products of the region's presses amongst themselves so extensively that they all ended up with similar stocks to trade on international markets. Bound together towards a common destiny, they learnt to operate as a cartel: they shouldered shipping and production costs mutually; promoted each other's wares through instruments like the Journal helvétique; and shared information concerning the literary marketplace.Footnote ⁴⁴ Nonetheless, patchwork Romandie was culturally and politically a world away from overheard whispers in the cloisters of the Palais Royal in Paris, the blackmailing bustle of Grub Street, London, or Friedrich Nicolai's upscale book boutique with important clientele in Berlin.Footnote ⁴⁵ The Dutch towns of Amsterdam, Leiden, and The Hague, although past their seventeenth-century prime, remained better placed to take advantage of proximity to Europe's cultural and political heartlands. Together with Brussels, Maastricht, Liège, Bouillon, and London itself, they formed a powerful northern European zone of clandestine book production that functioned similarly.Footnote ⁴⁶

Naturally, this bicentric model somewhat simplifies complex realities. Thierry Rigogne's work suggests that the eastern généralités of Metz, Alsace, and Franche-Comté, as well as the southern book trade centres of Avignon and Nice, are worthy of serious attention.Footnote ⁴⁷ The depth of Swiss links with Italian and Spanish producers and consumers has similarly been discussed elsewhere.Footnote ⁴⁸ We might also consider underground Parisian and regional printing, and the effects of the international book fairs of Leipzig and Frankfurt on the flow of proscribed books.Footnote ⁴⁹ Nonetheless, only the ‘northern’ and ‘Swiss’ markets had the density of producers and the book trade infrastructure (including networks of translators, engravers, typographers, paper-makers, and middlemen) to enjoy something approaching mutual independence. Books produced in Yverdon were instantly available throughout Romandie; those fabricated in Maastricht could quickly be found throughout the northern zone. Yet, locating the latest Swiss edition in Amsterdam, or Dutch nouveatés in Lausanne, was a hit-and-miss affair. Tastes differed, guaranteed best-sellers were certain to be counterfeited locally in both markets, and transport and insurance costs were prohibitively high reasonably to consider swapping stocks. During twenty-five years of trading, the STN dispatched just sixty-eight books to Amsterdam and never received a single sheet in return; their travelling salesmen wrote-off attempts to woo traders in such hostile markets as futile.Footnote ⁵⁰ Of course, significant consignments of books, often tied to exclusive distribution deals and shipped via Nancy, Cologne, or Frankfurt, were in both directions.Footnote ⁵¹ Yet, those expectations only serve to prove the rule: the STN never found a reliable and profitable way of marketing Swiss books in the north and, in turn they barely traded Dutch, Parisian, or London imprints.Footnote ⁵² Altogether moving beyond the ‘floating stock’ thesis and embracing new models more finely attuned to complex realities thus now appears unavoidable. It will require a whole new approach.

IV

In his recently republished essay ‘Lost and found in cyberspace’ Darnton shares his fantasy of writing an e-book. Over the decades, he has filled enough shoe boxes with index cards of STN material as to be ready to produce a total history of the book as a force on the eve of the French Revolution. The notes and unpublished manuscripts, he insists, open up new ways of thinking about the history of ideas, economics, politics, and society. His proposed electronic edition will be arranged according to a multi-layered pyramidal structure, with hyperlinks that will enable readers to choose their own pathway through the text. Having navigated the topmost narratives and mini-monographs, they will arrive at selected documents and bibliographies, as well as historiographical and pedagogical material. Readers may, he promises, even be treated to background music! The concept appears to offer both a personal curated tour of selected highlights of the STN archives and the opportunity to rummage through Darnton's drawers of unpublished material.Footnote ⁵³

The FBTEE project offers a radically different vision, albeit one equally rooted in the desire better to understand the eighteenth-century francophone book trade by opening up the Neuchâtel archives using digital technologies. The project's centrepiece is the aforementioned database of the STN's trade, populated with 70,584 transaction events, each of which can be pinpointed to a precise date and associated with a client fixed in geographical space. Together, these events record the acquisition and dissemination of 410,074 copies of 3,987 bibliographic entities. Where possible, full descriptions of these books are related to each transaction, as are extensive details concerning the clients and transactions themselves (for example street addresses and trade routes). These pieces of data are structured in over 100 interlinked tables, allowing the scholarly community to address complex questions in granular detail. Researchers interested in, for example, the volume of Dutch editions that the STN dispatched to the Lorraine in the early 1770s, or the number of Swiss illegal libels that the society sent to Paris in the run up to the Revolution of 1789, will find precise answers to their questions.

The rewards are especially rich for enquires focused upon the Neuchâtelois and Swiss trade. The database exposes the authors, books, and genres in which the STN specialized (notably novels, histories, and political works); the society's reach (pan-European, but particularly weak in Paris and the Low Countries save for commissioned editions); the routes their shipments took (conspicuously, their absolute dependence on Lyons via Geneva to reach the French market); and their fertile but sometimes fractious trading relationships with other French-Swiss maisons d’édition (including the Fauche's in Neuchâtel). Yet, with care, investigations can be extended further. First, understanding the peculiarities of the STN's trade has allowed the FBTEE team to incorporate tools that begin to compensate for biases in the dataset. Queries can be restricted by time (to minimize corruptions resulting from changing market conditions and the STN's evolving business); space (to gauge variations in regional trade patterns from national averages); source data (to compensate for lacunae in the archives); publication place (to address the overrepresentation of Neuchâtel and Swiss printed works); work type (to eliminate anomalous commissioned editions); and, finally, client type (to distinguish the wholesale from retail trades). Using these instruments, researchers can tackle questions of the representativeness of the STN figures directly and, in the process, build towards more constructive enquiries concerning the wider trade. Of course, it is impossible entirely to eliminate a speculative element by using such techniques and, as such, second, the data structures, processes, and software of the FBTEE are open source and intended to facilitate further research. Europe's libraries teem with potentially compatible bibliometric sources. Comparable printers’ archives have survived (although none as complete as the STN's); hundreds of extant bookseller's catalogues show the reach of various editions; confiscations and customs records abound; journal and newspaper reviews, advertisements, bookplates, and even recorded instances of reader's responses are all readily available. Although alone these sources lend themselves most readily to esoteric enquiries, collected and juxtaposed against the STN dataset, they would provide a detailed picture of the dissemination of books in the late eighteenth century.

Realizing this vision will be contingent upon an international collaborative effort to harness digital humanities approaches and apply them rigorously to the related fields of the geography of the book and intellectual geography.Footnote ⁵⁴ For now, by returning to the topic in hand, we might simply hope to demonstrate the potential value of such a research programme. Faced with a fiendishly complex set of account books and limited technology, Darnton was forced in his 1995 volumes to present his empirical findings more reductively than would now be the case. His work relied heavily on the immutable containers ‘1769–1789’ and ‘France’, which concealed wild spatial and temporal fluctuations in the data.Footnote ⁵⁵ The twenty-year time block is especially problematic. First, the STN's business evolved radically over these two decades. Their operations in 1785, when they stocked 1,500 works, were unrecognizable from their modest business of the early 1770s; after two years of trading, the Neuchâtelois still held only the ten works they had managed to print. Thus, statistics for books published in 1770 are not usefully comparable with those that appeared a decade later. Second, the extension of the period to 1789 is curious; a close examination of the booksellers in Darnton's sample reveals that his statistics do not cover the key period after 1785 at all.Footnote ⁵⁶ Third, such a lengthy time block may be considered inappropriate for cultural products that enjoyed such short shelf lives. The STN managed to liquidate the majority of the print-runs of editions that they published within a couple of years. Thus, touting works published in the early 1770s as pre-revolutionary best-sellers is problematic. As for space – while STN-distributed clandestine books did indeed reach all four corners of the kingdom, treating France as a unitary block conceals vital anomalies. Aside from the job-printing of works by the likes of Brissot and Mercier in the early 1780s, for example, the Parisian market appears to have been of negligible importance to the STN's model of selling illegal (and indeed legal) books. The Neuchâtelois shipped more than a dozen copies of only twenty-eight of Darnton's forbidden books to the city of lights during their entire trading history.

For the broadest conclusions of Darnton's work, such issues of space and time are perhaps only minimally consequential. The principal thesis of The forbidden best-sellers, that the diffusion of ‘livres philosophiques’ undermined the legitimacy of the Old Regime, relies upon the weight of empirical evidence presented during its opening chapters rather than its minutiae.Footnote ⁵⁷ In France, throughout the two decades that preceded the French Revolution, a large body of politically and religiously subversive titles are shown to have been highly sought after. This much is substantially corroborated by the FBTEE database, which reveals that the STN dispatched 117,031 (28·5 per cent of a total of 410,074 envois) books of dubious legality, 52,699 (12·6 per cent of total shipments) of which were directly sent to France.Footnote ⁵⁸ Further, the brilliance of Darnton's book lies beyond the bibliometrics, particularly in the sophisticated ways that he blends elements of diffusion and discourse analysis to link his empirical findings to his thesis outlining a collapse of value systems under Louis XV and Louis XVI. Where literary scholars have emphasized individual works, Darnton stresses intertextuality and tries to understand how contemporaries derived meaning by slotting newly encountered volumes and discourses into pre-existing cultural frames.Footnote ⁵⁹ Where previous dissemination studies relied upon crude models of causation, he provides two important and influential communications circuit models that represent his books as only one part of wider print and literary culture markets and machinery. Subversive words resonated and, drip-drip, began to precipitate a revolution.

The devil, however, is in the FBTEE dataset's capacity for detail. Darnton's ‘Grub Street’ thesis has never been simply about broadly defined illegality. In The forbidden best-sellers, he emphasized the slow but steady delegitimizing effect of works of political slander like Mairobert's Anecdotes sur Mme la comtesse du Barry. The recently published The devil in the holy water further sharpens the focus, concentrating the role that a relatively small group of anti-ministerial texts played in the creation of the political myth of despotism. Whilst the somewhat amorphous empirical evidence of The corpus of clandestine literature was successfully marshalled in support of the ‘Grub Street’ argument, the sharper FBTEE evidence at times appears to stand in opposition. Erotic works comprised only 10·7 per cent of the STN's illegal sales and those railing against despotism just 3·6 per cent. The STN saw neither Les bohemiens nor the wildly successful Le gazetier cuirassé before the French Revolution; they handled just 26 copies of Le diable dans un bénitier. Thus, whilst a quarter of the STN's envois were indeed illegal, only a tiny minority of these might be described as the work of an embittered scandalmongering bunch of hacks. Instead, the themes of a higher Enlightenment – philosophy (32 per cent), religion (24 per cent), politics (20 per cent), social issues (20 per cent), current affairs (17), and social mores (17 per cent) – dominated the society's illegal shipments.

A half-century-long Neuchâtelois odyssey has thus, for the moment at least, taken us back to square one. Even in counterfeiting Romande, illegal texts were only a part of the story. Highly salacious ones were a mere footnote. The STN evidence cannot be leveraged to support the letter or spirit of the ‘Grub Street’ thesis, which remains viable only in its pre-statistical, case-study-based form. The Neuchâtel archives – to repeat – can only take us so far in the quest towards understanding what Frenchmen read before the French Revolution. Of the 117,031 illegal copies that the STN sold, 78,304 (66·9 per cent) were identifiably produced in Neuchâtel, a further 11,328 in Switzerland. Just 4,994 (6·4 per cent) can be identified as having been produced in France, the Netherlands, Britain, Germany, Italy, Avignon, or anywhere else in Europe. Nonetheless, the approach pioneered by the FBTEE database gives three new hopes for better understanding the trade.

First, geography matters. Books that did not become international best-sellers published in multiple cities – be they works of political scandal like Morande's Le gazetier cuirassé or more sober scientific studies like Dupré de Lisle Traité sur le vice cancéreux – are worthy of attention. Such books might have enjoyed a considerable contemporary impact or, regardless, may be valuable to the historian trying to peek into the literary culture of their times. These, after all, were the bread and butter of publishing: a matter of tens of thousands of editions and millions of physical volumes during our period. Naturally, individual copies of any given edition, sent by the post or amongst assorted trade shipments, flew far from their printer's nest. As such, contemporary bookseller and private library catalogues often boasted a rich assortment of titles from distant presses. Yet, the STN data suggests that the intra-market trade was much more complicated than has hitherto been recognized.Footnote ⁶⁰ Transporting regular books across a continent was a tricky business proposition: they were volumetrically inefficient (bulky and of limited value); fragile (subject to water damage and wear); and highly, unevenly, and sometimes capriciously regulated (the legality of many texts was especially difficult to judge in an international context).Footnote ⁶¹ As such, viewed as material commodities, they suffered an appreciable ‘friction of distance’, as physical transport costs, insurance premiums, and risks all disadvantaged long-distance trade. Moreover, even when a potentially economically viable trading model could be found, contemporary views of commerce as a battle of competing commercial states meant that finding buyers and ensuring their payment was no simple matter.Footnote ⁶² Stock, in short, did not gracefully ‘float’ around Europe, it ground its way down highways and drifted along waterways, haemorrhaging profitability and approaching fierce competition by the kilometre.

Such quantifiable and predictable commercial realities mean that, given enough data, we might try to model Europe's intellectual geography in new and exciting ways. Perhaps texts like Le gazetier cuirassé, steeped in coded political discourse that must have rendered it unintelligible to most readers, captivated only certain circles within the ‘northern’ markets.Footnote ⁶³ Extending the argument, maybe authors too had measurably different impacts across geographic space. The baron d'Holbach is a case in point. Between 1766 and 1773, employing Marc-Michel Rey's Amsterdam presses, d'Holbach flooded the market with an unprecedented campaign of twenty-eight works of anti-clerical and atheist propaganda.Footnote ⁶⁴ Yet, the Neuchâtel evidence suggests that only the Système de la nature, which the STN published at the bequest of a Brussels dealer and, spotting an opportunity, over-printed and distributed for their own profit, circulated in significant numbers in Switzerland at the time. Further, where substantial intra-market differences between works, authors, or discourses can be discerned, by reverse engineering the communications circuit, we might just be able to draw insights into the literary culture – the street-corner gossip, salon discussion, and tavern chitter-chatter – that shaped them. We might better understand that not all areas were equal in the republic of books.

Second, a more temporally sensitive approach promises to be equally revealing. What might be made of the fact that the STN evidence suggests that Voltaire and the baron d'Holbach, touted as pre-revolutionary best-sellers, sold phenomenally well during the 1770s but rarely troubled readers during the following decade?Footnote ⁶⁵ Typically, the STN exhausted their print-runs quickly: their 2,500 copy run of Voltaire's 1771 Questions sur l'Encyclopédie, for example, was 60 per cent spent after a year and sold out after four years. Of course, such a fact hardly negates the potential impact of the patriarch of Ferney on the intellectual climate that preceded (and perhaps issued onto) the French Revolution. Models including Darnton's ‘communications circuit’ help us to understand that ideas and discourses bounced around various forms of media long after their parent works became scarce on the shelves of Europe's bookstores. Yet, nothing should prevent an equally empirical approach to the newspaper reviews, journal articles, police statements, confiscations, interdictions, reader responses, and other such bibliometric events that would allow us to truly see the communication circuit in action. In the process, alongside better understanding the relationships between the various forms of media, we should be able better to disaggregate short-term flash-in-the-pan successes from those works that had a more lasting legacy.

Third, and finally, alongside encouraging scholars to scrutinize sequences of spatial and temporal data, a new approach would allow a more precise focus upon the quantifiable impact of specific space-time events. These are potentially many: the first appearance of an Enlightenment classic; new book-trade laws or environments; the death of an important actor (Frederic II's August 1786 passing, for example); or political or social events with literary consequence, like the infamous Jean Calas execution of 1762. A single example should suffice. Since Darnton's early work, there has been a tension between his emphasis on illegal trafficking – of the night time sorties across treacherous mountain paths variety – and his insistence that such smuggling was expensive and impractical.Footnote ⁶⁶ Examining the STN sales data before and after significant changes to the regulatory environment allows us to test the place of border smuggling empirically for the first time. First, the French state introduced new book trade regulations on 12 June 1783 which effectively outlawed the international trade by demanding that all foreign shipments were to be inspected in Paris before being forwarded to their final destinations.Footnote ⁶⁷ Second, in October 1784, triggered by the Fauche's attempts to smuggle the comte de Mirabeau's pornographic Libertin de qualité through the Jura mountain chain, thus flouting the new laws, Neuchâtel was singled out for particularly close surveillance and direct diplomatic pressure.Footnote ⁶⁸ In the wake of these twin blows, the STN's trade to France almost entirely dried up. The Swiss booksellers complained bitterly; they sent a delegation to Versailles; they tried to diversify into job printing for Parisian publishers; and they searched for alternative routes via the Atlantic and North Sea ports. But none of this changed the fact that their French provincial trade had been obliterated; there was no way to get books through. Not, at least, until Christmas 1786, when word spread that the widow Barret had found a corrupt way of getting shipments through to her operation in Lyons.Footnote ⁶⁹ Smuggling – of the mountain passes variety – is a red-herring that was of no real consequence to the international trade in illegal books.

V

The technology and methodological approach of the FBTEE project points the way towards these new pistes. Here, it has allowed us to understand the relationship between supply and demand in the STN archives; to re-draw the STN's ‘best-sellers’ list based on sales; to challenge the prevailing view of the nature of the eighteenth-century francophone book trade; to suggest bicentric and polycentric models for future study; and to point the way forward regarding the practicability and benefits of such an approach. So, an unanswered question remains: if a non-existent book is never sent to an atypical bookseller, could it cause a revolution? Perhaps not. But if Mauvelain's anomalous orders help to focus our minds on the need for more spatially and temporally sensitive approaches to the eighteenth-century book trade, they may not have been entirely in vain.