"What is the History of Books?" Revisited

Having accepted the invitation to revisit my essay of 1982, “What Is the History of Books?”, I find that I can do it only in the first person singular and therefore must ask to be excused for indulging in some autobiographical detail. I would also like to make a disclaimer: in proposing a model for studying the history of books twenty-four years ago, I did not mean to tell book historians how they ought to do their jobs. I hoped that the model might be useful in a heuristic way and never thought of it as comparable to the models favored by economists, the kind in which you insert data, work it over, and arrive at a bottom line. (I do not believe that bottom lines exist in history.) It seemed to me in 1982 that the history of books was suffering from fissiparousness: experts were pursuing such specialized studies that they were losing contact with one another. The esoteric elements of book history needed to be integrated into an overview that would show how the parts could connect to form a whole—or what I characterized as a communications circuit. The tendency toward fragmentation and specialization still exists.

r How do books come into being? r How do they reach readers? r What do readers make of them?
But to answer those questions, we need a conceptual strategy for bringing specialized knowledge together and for envisioning the field as a whole.
When I reflect on my own attempt to sketch such a strategy, I realize that it was a response to the sense of interconnected problems that struck me much earlier, when I first began to work in a publisher's archives. Looking backward from the present also serves as a reminder that my essay of 1982 does not do justice to the advances in book history that occurred during the following quarter of a century. It has been reprinted and debated often enough for its inadequacies to be visible. So I do not propose to rewrite the essay, but I would like indicate how it might be improved and to explain the experience in the archives from which it originated.
I first took the plunge into the papers of the Société typographique de Neuchâtel (STN) in 1965 and immediately found myself studying the history of the book without knowing it. The term did not exist then, although the pioneering volume by Henri-Jean Martin and Lucien Febvre, L'Apparition du livre, had been available since 1958. I came to Neuchâtel looking for something else: information about Jacques-Pierre Brissot, the leader of the "Brissotins" or Girondins during the French Revolution, who published most of his works before 1789 with the STN. But when I began to follow Brissot's trail through the papers of his publisher, I discovered a subject that seemed more important than his biography, namely the book itself and all the men and women who produced and distributed it under the Ancien Régime.
It was not that I felt disappointed by the 160 letters that Brissot exchanged with the STN. On the contrary, they provided the most vivid and detailed picture of the relations between an author and his publisher in the eighteenth century that I had ever encountered, and I eventually published all of them on the Internet. But Brissot's dossier looked small in comparison with the 50,000 other letters in the STN archives-letters by authors, booksellers, paper millers, shipping agents, smugglers, wagon drivers, compositors, and pressmen; letters scrawled by such unlettered persons that they had to be sounded out and read aloud to be understood; letters that revealed a whole human comedy behind the books.
The most exciting kind of history in 1965 was known as "history from below." It was an attempt to recapture the experience of ordinary people, especially those in the bottom ranks of society, and to see the past from their perspective. They had never made it into history books, except as the faceless "masses" summoned to produce revolutions or die of famine at appropriate points in the narrative. As a graduate student at Oxford, I had sympathized with this kind of history, but I had never attempted to write it. The archives in Neuchâtel opened up the possibility of doing for the obscure men and women in the world of eighteenth-century books what E. P. Thompson, Richard Cobb, Georges Lefebvre, and George Rudé had done for workers, peasants, and sans-culottes. Even intellectual history, I thought, could be studied from below. Authors in Grub Street deserved consideration as much as famous philosophers. I still find this perspective valid, although I also believe that the past should be studied from above, from margins on the side, from every possible angle. In that way, it might be possible to create what the Annales historians used to call histoire totale. But I had not absorbed much Annales history in 1965. I first encountered it during the late 1960s through contact with Pierre Goubert and François Furet. In 1972 I became friends with two book historians connected with the Annales, Daniel Roche and Roger Chartier, and I have worked with them ever since. But that came later. The book came first. I got to know it through the archives in Neuchâtel, although it was not what I had been looking for and it turned out to be very different from anything I had expected.
Of course, I had seen plenty of books from the eighteenth century, but I had never taken them seriously as objects. I studied the texts embedded in their pages without asking questions about the material itself. Once I waded into the archives of the STN, all sorts of questions arose, notably about paper. To my surprise, paper occupied a large proportion of the publishers' correspondence, far larger than fonts of type and presses. (I will use the term "publisher," rather than "bookseller" or "libraire-imprimeur," despite its anachronistic character.) The reason became clear when I reconstructed the costs of production from the STN's account books. Paper represented 50 per cent of the costs for producing an ordinary octavo at a typical pressrun of a thousand copies-and 75 per cent of the costs of the Encyclopédie.
The letters of the paper millers themselves opened up another perspective. They are full of talk about the weather: "The weather is turning malicious." "I curse the weather." Why? Because if it rained too much, the water got muddied, spoiling the "stuff" (water mixed with pulped rags) that went into the paper. If it did not rain enough, the water wheel would not turn adequately. Moreover, bad weather provided an excuse for failing to furnish batches of paper on time. It turned out that printers often commissioned special batches, or "campaigns," as they called them, when they took on important jobs. They set their production schedule-and sometimes the hiring and firing of workers-according to delivery dates specified by contracts with the paper suppliers. The contracts required intense negotiations, not only over timing but also about the price, quality, and weight of the reams. Conditions were different in cities like Lyon and Paris, where large stocks of paper were readily available, thanks to specialized middlemen (marchands papetiers). But Swiss printers had to draw supplies from millers scattered everywhere in eastern France and western Switzerland, a vast area in which three different measures of weight were used along with different kinds of currency. Specie was chronically scarce, so printers occasionally paid in barrels of wine or other commodities. Bills of exchange varied in value according to the trustworthiness of the signatures on them. They could be traded at varying discounts or collected at their maturation date, usually through negotiations at the four annual fairs in Lyon. Printers tried to dump inferior bills of exchange on millers, just as millers withheld their best rags from the stuff destined for printers. And the bargain-hunting on both sides took the form of threats to shift business to a more accommodating supplier or client.
From a miller with two vats on a Jura mountainside to a moneychanger in the hurly-burly of Lyon, the human topography was extraordinarily complex, and it left a great deal of room for fraud. Millers often cheated by slipping extra sheets into their reams. Why extra sheets? I wondered. The protests from the STN revealed the answer: the millers produced inferior sheets by diluting their stuff, so they needed to use more than 500 sheets in their reams in order to make the agreed weight. The printers therefore had to weigh the reams that they received, to count the sheets in them, and to send back letters full of complaints and demands for rebates. The millers replied in tones of hurt pride and indignationor, when beaten down, with excuses: the weather primarily but also special circumstances: "My vatman was drunk." The notion of paper as an item under continuous negotiation-contracts for campaigns negotiated before delivery and renegotiated afterward-took me completely by surprise. As far as I know, it has never been recognized by bibliographers or historians of printing.
It also bears on the question of reader reception. If you read advertisements for books in eighteenth-century journals, you will be struck by the emphasis on the primary material of literature: "Printed on the best-quality paper from Angoulême." That line of salesmanship would be unthinkable today, when readers rarely notice the quality of the paper in books. In the eighteenth century they often found splotches made by drops from a poorly held deckle or bits of petticoat that had not been properly pulped. Remarks about paper turn up so often in the letters of booksellers-and even a few readers, though the STN rarely heard from individual consumers-that I think a peculiar paper consciousness existed in early modern Europe. It must have died out with the advent of machinemade paper from pulped wood in the nineteenth century. But in earlier times people looked at the material substratum of books, not merely at their verbal message. Readers discussed the degrees of whiteness, the texture, and the elasticity of paper. They employed a rich esthetic vocabulary to describe its qualities, much as they do for wine today.
I could go on and on about paper, but the point I want to make concerns something different: the complexity built into the everyday activities of publishers. They inhabited a world we cannot imagine unless we read their archives and study their business from the inside. Their correspondence shows them struggling with the intricacies of problems in many aspects of their trade. They could not concentrate exclusively on one problem, because each element of their business bore on all the others and the parts functioned simultaneously to determine the success of the whole. The daily or weekly tabulation of entries in their account books-elaborate registers from which I could redo the arithmetic in order to follow their reasoning-reminded them that they had to coordinate a wide variety of interrelated activities so that, when stock was taken and the accounts were balanced, they would have a profit. Their pattern of behavior corresponded to the diagram, inadequate as it was, that I produced in "What Is the History of Books?".
In order to drive this point home, I would like to mention some other aspects of publishing that surprised me when I studied the STN archives and that have not been assimilated, as far as I can tell, in the history of books. For example: Smuggling When seen through the letters of smugglers, it turned out to be quite different from the razzle-dazzle I had imagined. Smuggling was a major industry-in many trades, notably textiles, as well as books-and it was organized in different ways. The most sophisticated variety went under the name of "insurance." Self-designated "insurers" negotiated contracts with publishers, guaranteeing to get illegal books to secret entrepôts across the French border in the Jura Mountains for a percentage of their wholesale value. If a shipment was seized by a flying squadron of the customs (employees of the Ferme générale, a tax-gathering corporation, not officials of the state), the insurer would reimburse the shipper for its full cost. He employed teams of peasants to do the actual work, lugging the books on their backs in sixty-pound packs (fifty pounds when the mountain passes were clogged by snow.) If caught, they could be branded with the letters GAL for galérien or "galley slave" and sent to row in the prison galleys of Marseilles for nine years or more.
Distribution and sales These functions took many forms. I was particularly impressed with the importance of sales reps (commis voyageurs, or traveling agents of publishers). I had thought that they did not exist before the nineteenth century, but I found that they honeycombed France under the Ancien Régime, performing all sorts of tasks. They sold books, collected bills, arranged shipping, and inspected all the bookshops along their itineraries. Every important publisher employed them. They often crossed paths with one another, stayed in the same inns, and exchanged trade secrets in evenings spent over a pichet of wine and a roasted pigeon. Some of their shop talk appears in their letters and diaries. One sales rep of the STN spent five months on a horse, stopping by nearly every book shop in southern and central France. When he arrived in a shop, he would take its measure and run through questions set for him in his diary: How much credit could be extended to the bookseller? (Ask the local tradesmen.) What was his character? ("Solidity", the most desirable quality, meant he could be relied upon to pay his bills on time.) Was he a family man? (Bachelors aroused suspicion, but a married man should not have too many children: they might drag him into debt.) When the sales rep returned to Neuchâtel, he had acquired an incomparable knowledge of conditions in the book trade. His reports supplemented the letters of recommendation from businessmen and allies in the trade that arrived at the publisher's office every week. Taken together, they provided crucial information for adjusting sales strategies to the complex human topography of the publishing business.
Literary agents They did not exist in the modern sense, as representatives of authors. In the eighteenth century authors generally received a cash payment for their manuscript or a certain number of printed copies, if they got anything at all. Royalties and translation rights did not exist. But all important Frenchlanguage publishers located outside Paris needed a representative to look after their interests at the heart of the publishing industry. The Parisian agents wrote regular reports about the state of the book trade, political conditions, the reputations of authors, and the latest books that were creating a buzz among professional insiders. In some cases, the reports form a running commentary on literary life, and they can be read as sources for a historical sociology of literature.
Piracy France was surrounded by publishing houses that pirated everything that sold well within its borders. Although I cannot prove it, I believe that more than half the books that circulated in pre-Revolutionary France-works of fiction and non-fiction but not professional manuals, religious tracts, and chapbooks-were pirated. But piracy differed substantially from what it is today. The modern concept of copyright did not suit the conditions of early modern publishing, except in Britain after the copyright law of 1710. Everywhere else, rights to copy were determined by privileges, which extended only as far as the jurisdiction of the sovereign who issued them. The Dutch and Swiss publishers looked like pirates to the French, but they appeared as solid businessmen at home. They conducted market research, calculated risks and profits with professional expertise, and sometimes formed alliances, sealed by treaties, in order to beat competitors to the market while sharing costs and risks. I found several contracts among the sociétés typographiques of Lausanne, Bern, and Neuchâtel, negotiated after intense bargaining, which committed each publisher to print a proportion of the books and to provide a corresponding amount of the capital investment. Such joint enterprises force one to rethink the economics of early modern publishing and to reconsider the nature of piracy itself, for pirated books were rarely intended to be exact copies of the originals. Printed on relatively cheap paper, stripped of illustrations, abridged and adapted without concern for the integrity of the text, they were aimed at the broader, poorer sectors of the reading public.
Swapping Publishing alliances also took the form of agreements to swap books. After printing an edition of a thousand copies, a publisher often exchanged a hundred or more with allied houses in return for an equivalent number of sheets, which he selected from their stock. In this way he could maximize the variety of works on offer in his own general stock (livres d'assortiment) while minimizing the risks involved in the diffusion of his main products (livres de fond). But swaps involved complex calculations, which concerned the quality of the paper, the density of the type, and estimations of demand. Skill at swapping could determine the success of a publisher.
Demand Because of the prevalence of swapping, publishers tended to evolve into wholesalers. Clusters of allied houses carried similar backlists, and everyone raced to the market with pirated editions when word spread about a potential bestseller. In contrast to the "blockbusters" of today-huge editions put out by a single company-bestsellers in the eighteenth century were produced simultaneously in small editions by many publishing houses. A publisher who arrived late on the market or who miscalculated the demand for an ordinary, "mid-list" book could be punished with a heavy loss. So producers took elaborate measures to sound the market, using their sales reps, their Paris agents, and above all their commercial correspondence. By building up a network of reliable, savvy customers among booksellers, a publisher received constant advice in a stream of letters that arrived every day from wholesalers and retailers scattered over a wide area, sometimes all of Europe. To follow the arrival of letters, day by day and town by town, is to watch the ebb and flow of literary demand.
Politics But demand could not be supplied freely, because all sorts of political obstacles stood in the way. A publisher located across the French border had to keep informed about shifts within the Direction de la librairie and among the police and the inspectors of the book trade in provincial cities. Conditions varied enormously from place to place and from year to year. The rules of the game shifted substantially at the national level during critical periods, such as the lobbying to influence the new règlements de la librairie in 1777. The provisions of the 1777 edicts could easily be studied from their printed texts. But only by reading the correspondence of booksellers can one gauge their effects. I was surprised to discover that the edicts did not transform the conditions of the trade and that they were far less effective than an unknown order, issued by the foreign minister to customs officials on 12 June 1783. The order required all shipments from foreign suppliers, no matter what their destination, to pass through Paris and to be inspected by the officers of the Parisian booksellers' guild reinforced by the tough Parisian inspector of the book trade. A shipment from Geneva to Lyon therefore had to make a ruinous detour to Paris. With one stroke of the pen, this measure destroyed most of the trade between the provincial booksellers and foreign publishers. Letters from the provincial dealers prove that it produced a crisis that lasted until the Revolution but that had never been noticed by historians of the book trade, because they had confined their research to printed documents and administrative sources.
I could cite many more examples of the surprises that struck me while working in the archives of the STN-and then by comparing those findings with material available in the main sources in Paris: the Collection Anisson-Duperron, the papers of the Chambre syndicale de la Communauté des libraires et des imprimeurs de Paris, and the archives of the Bastille. What impressed me most was the need of a publisher to keep several balls in the air while the ground was shifting beneath his feet. He might be negotiating terms for new campaigns of paper, recruiting workers for his printing shop, settling a contract with an insurer at the French border, firing off directions to a sales rep in deepest France, modifying his view of the market according to information from his agent in Paris, laying plans to pirate promising new works, arranging swaps with half a dozen allied houses, adjusting his list in conformity to advice received from dozens of retailers, and trimming his business strategy to suit the vagaries of politics, not only in Versailles but in other parts of Europe-all at the same time. He also had to consider many other factors, such as the possibility of purchasing original manuscripts from authors (a hazardous undertaking, because they sometimes sold the same work under different titles to two or three publishers), the availability of specie in the quadrennial fairs of Lyon, the dates of expiration of outstanding bills of exchange, the changing rates of tolls on the Rhine and the Rhône, even the date when the Baltic was likely to freeze over, forcing him to send shipments to St Petersburg and Moscow overland. It was his ability to master the interrelation of all these elements that made the difference between success and failure. Therefore, when I attempted to picture the system as a whole, I tried to bring out its interconnections, not merely from the publisher's viewpoint but as it affected the behavior of everyone in the system. My diagram hardly did justice to the complexities, but it brought out the way the parts were linked, and I think it conveyed something of the nature of book history as it was experienced by the men (and also many women-la veuve Desaint in Paris, Mme La Noue in Versailles, la veuve Charmet in Besançon) who made it happen. * * * Those impressions, first registered in 1965, determined the character of the model that I put together in 1982. Every once in a while since then I receive a copy of another model that someone has proposed to substitute for mine. The pile of diagrams has reached an impressive height-and a good thing, too, because it is helpful for researchers to produce schematic pictures of their subject. Rather than review them all, I would like to discuss one of the best, a model proposed by Thomas R. Adams and Nicholas Barker in "A New Model for the Study of the Book" published in a volume edited by Nicholas Barker, A Potencie of Life: Books in Society (London, 1993).
Adams and Barker base their analysis on what they call a "bibliographical document" rather than a book. That approach makes room for ephemeral printed matter, an important consideration, as printing shops depended heavily on small jobs and special commissions. In practice, however, Adams and Barker concentrate on books, and their proposal for enlarging the scope of my diagram makes it more adaptable to conditions that prevailed after the first decades of the nineteenth century. Although I thought my diagram could be modified to suit later periods (I never intended it to be applied to books before Gutenberg), I had in  mind primarily publishing and the book trade during the period of technological stability that stretched from 1500 to 1800-thus my decision to emphasize the role of binders, who were especially important in an era when publishers usually sold books in unbound sheets or in gatherings that were stitched together but not sewn.
In place of the six stages in my diagram, Adams and Baker distinguish five "events": publication, manufacture, distribution, reception, and survival. By doing so, they shift attention from the people who made, distributed, and read books to the book itself and the processes through which it passed at different stages of its life cycle. They see my emphasis on people as a symptom of my general approach, one that derives from social history rather than from bibliography and is aimed at the history of communication instead of the history of libraries, where books often find their ultimate resting place. I find those points valid. In fact, I cannot work up enthusiasm for any kind of history that would be emptied of human beings. So I still would stress the importance of studying the activities of book people in order to understand the history of books. When I examine the finer points in the argument of Adams and Baker, they seem to do the same. For example, they intend the first box in their diagram to represent the decision to publish-a decision made by people, though it determines the creation of the book as a physical object. At the same time, they underplay the role of authors. I stressed authorship in the first of my boxes, intending in that way to open up book history to what Pierre Bourdieu described as the "literary field"(champ littéraire)-that is, a set of relations determined by lines of force and regulated according to rules of the game accepted by the players.
The last box in the Adams-Baker diagram, "survival," represents a significant improvement over mine. I had made room for libraries, but I failed to take into consideration the reworking of texts through new editions, translations, and the changing contexts both of reading and of literature in general. Adams and Baker make their point effectively by citing the example of The Pilgrim's Progress, which first appeared as a chapbook, later came out in deluxe editions, and finally took its place in the canon of classics as an inexpensive paperback read by students everywhere. Peter Burke's study of Castiglione's Courtier is another example of excellent book history that is difficult to accommodate in my diagram. Because I tried to picture the interrelated stages in the life cycle of one edition, I did not do justice to phenomena such as preservation and evolution in the long-term history of books. I wonder, however, if a flow chart can capture the metamorphoses of texts as they pass through successive editions, translations, abridgments, and compilations. By concentrating on a single edition, my diagram at least had the advantage of tracing steps in a concrete process, one that connected authors with readers through a series of clearly linked stages.
Finally, I should acknowledge fields in book history that defy the urge to draw diagrams. Iceland had a printing press nearly a century before the Pilgrim Fathers set foot on Plymouth Rock. But it turned out nothing but liturgies and other ecclesiastical works required by the bishops in Skálholt and Hólar. Secular printing did not begin until 1773, and even then it was confined to a small shop in Hrappsey. (I am drawing here on the work of Icelandic book historians such as Sigurdur Gylfi Magnusson and David Olafsson.) Iceland never had any bookshops between the sixteenth century and the mid-nineteenth. It also had no schools. Yet by the end of the eighteenth century the population was almost entirely literate. Families in farms scattered over an enormous area taught their own children to read-and the Icelanders read a great deal, especially during the long winter months. Aside from religious works, their reading matter consisted primarily of Nordic sagas, copied and recopied over many generations in manuscript books, thousands of them, which now form the principal collections in Iceland's archives. Iceland therefore provides an example of a society that contradicts everything in my diagram. For three and a half centuries, it had a highly literate population given to reading books, yet it had virtually no printing presses, no bookshops, no libraries, and no schools. An aberration? Perhaps, but the experience of the Icelanders may tell us something about the nature of literary culture throughout Scandinavia and even in other parts of the world, especially in remote rural areas where oral and scribal cultures reinforced each other beyond the range of the printed word. * * * The example of Iceland suggests the importance of venturing off the beaten path that connects great centers like Leipzig, Paris, Amsterdam, London, Philadelphia, and New York. And whatever one makes of the Icelanders, it must be admitted that diagrams are merely meant to sharpen perceptions of complex relationships. There may be a limit to the usefulness of a debate about how to place boxes in different positions, provide them with appropriate labels, and connect them with arrows pointed in one direction or another. When I reflect on how I could have improved my essay, I think less of my diagram than of the need to take account of the impressive advances made in the history of books since 1982. Rather than attempting to survey them all, I would like to concentrate on four and to indicate how they have affected my own research.
First, I should mention the reorientation of bibliography wrought by D. F. McKenzie, a friend who taught me a great deal, not only by his writing but also through our collaboration in a seminar at Oxford. McKenzie did not reject the techniques of bibliographical analysis developed a century ago by Greg, McKerrow, and other masters of the discipline. He used them to open up a new area of investigation, which he called the sociology of texts. "Sociology" sounded like a declaration of war to some of the bibliographers who heard or read McKenzie's Panizzi lectures of 1985. But he employed it in an effort to extend rigorous bibliographical analysis to questions about the ways texts resonate through the social order and across the ages. In one of his most influential studies, he showed how the character of Congreve's plays was transformed from the scrappy, bawdy quartos of the late seventeenth century to the stately classicism of the 1710 octavo edition. Although the texts remained essentially the same, their meaning was modified by page design, new modes of presenting scenes, and the typographical articulation of all the parts. John Barnard has incorporated McKenzie's interpretation in a broad account of the emergence of a literary canon through editions of Shakespeare, Dryden, Congreve, and Pope. The book, in all its physicality, therefore appears as a crucial element in the development of literary culture in Augustan England-and, beyond literature, as an ingredient in the consumer society and the ethos of politeness that characterized middleclass life throughout Britain in the eighteenth century. In a similar series of studies, Peter Blayney has extended bibliography into the sociocultural history of Elizabethan England. Were I to rewrite my essay, I would try to do justice to this rich strain of scholarship.
A second strain that I would emphasize usually goes under the name of paratextuality. It has occupied bibliographers for generations and more recently has engaged literary theorists insofar as it has become increasingly important in concrete studies of texts. After roaming through this literature, I found myself paying far more attention to the way title pages, frontispieces, prefaces, footnotes, illustrations, and appendices work on the perceptions of the reader. Burlesque footnotes appear everywhere in eighteenth-century books. One of my favorites says simply, "Half of this article is true." It is up to the reader to discover which half. Devices such as that invite the reader to play a game, solve a puzzle, or decode a riddle. I have become fascinated with romansà clef, a very popular genre in the eighteenth century. To make sense of them, you have to read on two levels, moving back and forth between the narrative, which can be perfectly banal, and the key, which makes the story come alive by means of "applications"(a key term for the Parisian police) to current politics or social issues. The history of reading now looks far more complex than I had originally imagined. Of the many kinds of reading that developed in early modern Europe, one that I think deserves special attention is reading as game-playing. You find it everywhere, in libels, novels, and literary reviews, which constantly invite the reader to penetrate into secrets hidden between the lines or beneath the text.
Another important element in understanding the way in which books relate to the world around them is to be found in the concept of intertextuality. Put so abstractly, such words can sound unduly pretentious, but both paratextuality and intertextuality convey a common concern with the way seemingly extraneous elements-whether internal, like typography, or external, such as borrowings from other texts-shape the meaning of a book. Historians of political thought have long studied tracts by Machiavelli, Hobbes, and Locke as part of a running debate marked by other tracts. Each work, as they see it, belongs to a collective discourse and cannot be understood in isolation. While surveying eighteenthcentury libels, I kept running across passages that I thought I had read somewhere else. When I traced them back to their sources, I was surprised to find the same anecdotes recounted in nearly the same words scattered everywhere in books, pamphlets, and journalistic chroniques scandaleuses. A case of collective plagiarism? The word existed two centuries ago, but "plagiarism" hardly describes the practice of writers scribbling in Grub Street. They lifted passages from each other's works, added material picked up in cafés and theaters, stirred well, and served up the result as something new. Bestsellers like La Vie privée de Louis XV and Anecdotes sur Madame la comtesse du Barry contain the same anecdotes culled from a large variety of the same sources. In the seventeenth and eighteenth centuries, unlike today, "anecdote" meant "secret history." The term, derived from Procopius and other writers in ancient Greece and Rome, referred to hidden incidents from the private lives of public persons, things that had really happened, though they might be distorted in the telling, and that therefore revealed the inadequacies in official versions of events. Anecdotes made up the basic elements in all sorts of illegal literature, and they could be concocted in endless combinations. I have come to think of libelous books as by-products composed from pre-existing chunks of information that were available to any hack who needed to make some money and to any political agent intent on character assassination. Libels were cobbled together out of material scattered through the information systems of the Ancien Régime. To make sense of them, it is crucial to study the system itself-that is, to concentrate on intertextual combinations rather than on the book as a self-sufficient unit.
Finally, I would like to emphasize the importance of comparative history. It is preached more often than practiced, but a few historians-Roger Chartier and Peter Burke, for example-have demonstrated the value of following books across the boundaries of languages and countries. In my own research since 1982, I have tried to compare censorship as it was practiced in three authoritarian regimes during three centuries: Bourbon France, colonial India, and Communist East Germany. The comparisons demonstrate that censorship was not a thingin-itself, which can be monitored like a radioactive particle in a blood stream, but rather an ingredient in sociopolitical systems, each of which operated according to its own peculiar principles. A macro-analysis of publishing and the book trade throughout eighteenth-century Europe could turn up more revealing results. Germany and Italy lend themselves to comparison, because both were fragmented into small political units while a national literature was flooding a single, large-scale market. The opposition between Frankfurt and Leipzig led to the modernization of the trade in Germany. It entailed a shift from a system dominated by the exchange of books (Tauschhandel favored in Frankfurt) to one stimulated by payments in cash (Barhandel practiced increasingly in Leipzig), and it resulted in the victory of publishers in Leipzig and Berlin who paid significant advances to important authors, notably Goethe. Perhaps Milan began to eclipse Venice in the same manner. The Italian Enlightenment certainly spread from strongholds in the north, such as the philosophers grouped around Il Caffè in Milan. France and England provide even more fruitful possibilities of comparative analysis. The Stationers' Company monopolized the trade in London in a manner similar to the Communauté des libraires et des imprimeurs in Paris. Each oligarchy stifled publishing in the provinces, and in each case the provinces retaliated by forming alliances with foreign suppliers. Edinburgh, Glasgow, and Dublin flooded England with cheap, pirated editions, just as Amsterdam, Brussels, and Geneva conquered the market in France. Of course, the political conditions were different. The English enjoyed something close to freedom of the press, despite the repressive effect of prosecution for seditious libel, while prepublication censorship and the book police inhibited the French trade, despite the opening up of legal loopholes such as permissions tacites (permission to publish books without official approbation by a censor). Were economic conditions more important than the formal rules imposed by political authorities? I am inclined to think so. Moreover, the rules of the game began to change at the same time in both countries. The case of Donaldson v. Beckett in 1774 freed the English market in a way similar to the French edicts on the book trade of 1777. The raids on the German market by Austrian pirates could be compared with the foreign attacks on the trade in England by the Scots and the Irish and in France by the Dutch and the Swiss. By combining such comparisons with a study of the evolution of copyright throughout Europe, it might be possible to develop an overview of tendencies in the history of books on a large scale.
Other book historians would propose other agendas for future research. These remarks are necessarily idiosyncratic and egocentric, for that was the nature of the assignment: to reassess an article I wrote in 1982. This exercise has of necessity taken me back to 1965, but I hope that it also can help to focus attention on the opportunities that will exist beyond 2007.