Note: The data I have used for the following charts comes from three sources: the Short Title Catalog of Pollard & Redgrave for the period 1475-1640,  the Wing catalog for the period 1641-1700, and the English STC (ESTC) for the period 1701-1800, all in machine-readable format. Each one of those sources represents the most comprehensive and up-to-date source of information for their respective period. In a few months, it will be possible to rely upon a single comprehensive source of data in electronic format for a great variety of statistical uses, the English STC. Although the data itself is not my own, I must take complete responsibility for any factual error or error of interpretation contained in the text and analysis of those charts. 

A detailed count of the number of ESTC-scope printed items per year is available at the following location: stcdates.html

1. Surviving recorded titles - 1475-1640
2. Close up on the years 1545-1553
3. Surviving recorded titles - 1600-1700
4. Surviving recorded titles - 1701-1800
5. Surviving recorded titles (5-year increments) - 1475-1800
6. Surviving titles (1-year increments) - 1600-1800

The first charts included in this page represent the number of surviving, catalogued titles per year, roughly divided into chunks of a century each. Although it is tempting to see an exact projection of historical reality in this data, it is important to be aware of some of the factors that can cause significant distortions.
First, the survival rate of books is not constant from one period to another. That survival rate is naturally and proportionally lower for the more remote periods than it is for later periods. Also, we can suppose that proportionally more books survive from the more "interesting" or troubled periods of history.
It is likely, for instance, that the extraordinary surge in publication during the Civil War that is observed in Chart 2 has not one but two main causes: an actual increase in the number of published items, and the extraordinary work of George Thomason, a London bookseller, who helped preserve almost every scrap of printed material published in the course of his carreer. The Thomason collection, as it is now known, is currently housed at the British Library, and according to the ESTC records 15018 items. Another important collection of Civil War-related material,  now  kept at Worcester college, Oxford, was gathered by the General Clarke.
Finally, another crucial factor to take into account are the cataloguing rules and constraints imposed upon bibliographers. Regular and spectacular peaks in production for the 18th century period did not occur with clockwork regularity every five years. Those peaks, clearly observable in our data, are not, of course, indicative of actual historical publication booms, but of the large number of undated publications (about 20%), which catalogers have to estimate to the best of their ability, thus giving multiples of five and ten an artificial boost.

This first chart covers the period from the introduction of print by Caxton in England in 1473 or 4 to the year 1600. It shows a steady increase in the output of the press, with a noticeable peak in production during the troubled years of the break between Henry VIII and Rome and of that king's succession. This is only one of several spectacular accidents of history that directly and significantly impacted the output of the press throughout the period covered.  In the space of a little over a century, the output of the press goes from zero to about 400 titles per year. Overall, not much, for an age of Renaissance. A man or woman living immediately before the Civil War could easily have read the whole of the English output of the press of any given year within that same year.  This begs the question of imports and manuscript production. How many foreign books did English people read? After an initial period of encouragement, legal restrictions upon imports are set into law in the mid-16th century in order to protect the budding English industry. Yet, paradoxically, it was Thomas Berthelet, most probably an expatriate Frenchman  like Richard Pynson before him, who, as the King's printer, printed those statutes. Even as late as the eighteenth century, sales catalogues of private libraries show on average quite a large proportion of imported books. Despite the lobbying efforts of booksellers, culture remained very much an international affair throughout the hand-press period. In the case of Castiglione's 'The courtier', for instance, the ESTC records 18 editions, out of which 6 in Italian and English, 9 in Latin, and only three in English only.  The authors of the Histoire de l'edition francaise reckon that the printed production in France outnumbered manuscript editions only about a century after the introduction of print in that country. I have no data available for Great Britain, but it would certainly be interesting to compare the two figures.   The secret of the survival of printing in Western countries could probably be found in the distribution of publications into various categories of genre and purpose during that period. The important proportion of official documents, in particular, demonstrates a symbiotic relationship between printers and the nascent absolute monarchies of Europe. The STC, for instance, lists some 80 editions of the "The boke of iustices of peas" for the 16th century, a book that was of crucial importance in the Tudor's efforts to assert Royal authority in the remote provinces of the kingdom.  England, we are tempted to say, London, is virtually the sole producer of printed books until the early 18th century for Great Britain. For the STC period (1475-1640), London publishers account for about 29200 titles out of a total of a little under 35,000. The other "centers" of publication in Britain are far behind: Oxford contributes 850, Edinburgh about 875, Cambridge 570, and Dublin 231 titles to the STC.

The 17th century was an unhappy century. Some parts of Germany lost up to  two thirds of their population. Ireland did not fare very much better. England was torn by the Civil War and half a century after the regicide, the trauma of those years was still very much present in everyone's consciousness. This perhaps helps explains why the Glorious Revolution was relatively bloodless. In the 1726 edition of Chamberlayne's Present State of Great Britain, the author still notes that "never so many good and bad books were printed as during the late distracted times". According to our chart, that remark was in no way inflated. Yet, in spite of this spectacular outburst of publications in the 1640's, and after a short period of turmoil, the return of Charles II to the throne in 1660 brings things back to normal. The steady long-term upward progression of the printed output - from about 400 titles a year in 1600 to about 2,000 titles a year in 1700 - resumes where it left off in 1640, almost as if nothing at all had happened in the intervening years.
If we were to take into account every single published item today, that figure of 2000 titles is probably much inferior to the published production of a single day in 1998 England. Yet, that figure is also vastly superior - particularly if we consider that the average number of copies per edition reaches maybe 1,000 by 1700 -  to the production of the previous century.
Together with the inflation in the number of published titles and pages comes a significant decrease in quality. Fonts get worn out; printers cram more text per page, reduce margins; the size of the printed sheet becomes larger, lines of text longer, the fonts smaller; everything contributes to make print less legible, at the same time as the printed word becomes more in demand. Details matter less when you are reading about the latest news of the war. Quality suffers when booksellers are sure to sell their stock.  Their excuse is the greater - and cheaper - good of the reading populace.
Quality is the price England pays for its national treasure: freedom of the press. An advertising war goes on in earnest to draw the honest people's attention to the latest, freest perspective on the affairs of the time. A visiting Spaniard in the late 17th century claimed that more was to be found on an English book's titlepage than in the book itself. He wasn't far from the truth. The titlepage is the reflection of the book as a whole, its own self-reflection. The English book of that time chose to represent itself as free of any petty decoration.
The decorated book, the pretty book, the collector's book pretty much died in England together with Charles I. The commodities sold by 17th-century English booksellers are not precious objects, but precious, serious, radical and sometimes even seditious, ideas. The entire English industry, in a way, is Puritan and revolutionary in spirit, in its almost complete disdain for material form. Significant efforts to improve the quality of English print will only come in the second half of the 18th century.

 
Although published in the 1720s, the true inaugural monument of the 18th century is Alexander Pope's The Dunciad. The subject of that poem is the transfer of Poetry from the Court to Smithfields, from the aristocracy to the populace. Pope was, like Voltaire, an "aristocrate de la plume" with a territory to defend but no actual title of nobility. After a century of free - and often anonymous - ideas, comes a century of authors, some more venal than others. The theme of the Dunciad was not particularly new -  poets, especially minor poets -  had long been complaining in print about information overload. That particular subject would go on evolving in various forms, most noticeably in the arcane poetry of Mallarme, whose famous phrase "Tout finit en un livre" has an awfully ambiguous ring. Should we translate it as: "Everything ends in a book", "Everything ends up in a book", "Everything ends with a book", or "Anything will end up in one of those books"? Ironically, when Pope's poem finally gets published, the prosperity is over: the publishing bubble burst together with the South Sea one and a few others. The profession of author, newly protected by the first copyright law in 1710, is not doing that well any more. For the next fifty years, from 1720 to about 1770, a slump in production is clearly apparent. It is made worse by the fact that, like the Walpole administration, it never seems to really end. The successful authors of the period, like Richardson or Fielding, have a full-time job on the side.
Successful or semi-succesful booksellers, like the "infamous Curll", print cancel titlepages to hide long unsold sheets under a new veneer, and sell a lot of "Miscellanies" for exactly the same reason. The "miscellany" as a genre may have helped create a new notion: that of the classic Modern author. A classic, by definition, is a work worthy of being repeatedly, regularly, published. Until the 1700s, this privilege did not extend to modern works. The profession of antiquarian booksellers, in the 17th century,  was a respected and lucrative one partly - or mostly - for the reason that modern works were soon out of print and therefore very hard to find. Most second editions in the 17th century represent, rather than a sign of success, an attempt to finish off a stock of sheets that sat in the shop for much too  long - sometimes up to 70 years.
The miscellany is a similar concept, but it also tries to make new soup with the old broth. It glorifies the memory and virtues of a modern author, who is de facto worth re-publishing, as long as the words "Second edition" do not appear on the titlepage. "A new miscellany" rings - and quite possibly sells - better, given the growing appetite for reading from a middle class population who can spare just enough for an old edition of Swift but not enough for a new one of Virgil. Later on in the century, the idea of the miscellany evolves in the direction of multi-volume sets of complete works of authors whose copyrights have expired.
In spite of the 1720s lasting slump, the output of printed material keeps on the rise and one of the most striking aspects of the 18th century is the diversification of subjects and targeted readerships. Women, children, but also accountants, farmers, apothecaries, or servants rapidly become after 1750 as many specialized or even "niche" markets. What had once been the possession of the few now concerned everyone. How and why did this happen? No doubt there are immediate reasons for the expansion of the book market to new categories of readers. The first one of those is simple: there were more readers, i.e. more people, and more people who could read. The demographic expansion and the development of literacy can and should be seen as direct factors for the take off of the 1770s, all the while remembering that there is no definite causal effect between a larger population and a larger reading population.
But, in order to fully understand the definite qualitative change in the printed production of the 18th century, one has to take, I think, a bird's eye view.

This final chart gives us a global picture of the printed material from Great Britain, dependencies and North America that has survived in public libraries organized by number of titles per decade. The overall mathematical shape is clear: it is an exponential curve. Even if we compensate for the fact that more items are likely to survive from the 18th than they are from the 16th century, and are cautious about the distorting lens of cataloguing procedures, it is very unlikely that historical reality could have been very much different.
What historical factors actually contributed to give this curve its particular shape? Was it an economic or technological expansion, or a geometric rate of growth in the European population? None of those possible reasons seem totally applicable, although they may have contributed to the trend to some extent.
If external factors are not satisfactory, we must turn to internal factors, that is to say, factors inherent to the press itself and its economic laws. For instance, the mechanization of printing can be seen as the fruit of technological progress, but it can also be seen as the result of a momentum of the work which the press itself built up. By the end of the 18th century, the hand-press had reached its full potential and could no longer adequately satisfy the demand which it had, for centuries, helped to create. After three centuries of technological stability, printers in England and elsewhere had finally out-reached the potential of the hand-press. Before the 19th century, there would have been no point investing in research to improve a machine that performed quite well and well under its potential in most situations. By 1800, however, the potential profits of technological innovation are such that actual investments are made.  There is no reason to think, in fact, that this same mechanism was not at work in Gutenberg's own time. In the 15th century, the traditional "manuscript factories" had reached the limits of their potential and were faced with an increasing demand that could no longer be adequately satisfied.
There is another reason why we cannot think of the mechanization of the press as the simple result of technological improvements. Nothing that touches publishing can be simple, because it is publishing that brings ideas and knowledge to the people who devise new inventions. The hand-press, for centuries the only means of mass-production, helped mass-produce ideas that later on disseminated the concept of mass-production in other fields. This self-sustaining movement is the basic idea behind an exponential progression.
Whole areas of study are or should be concerned with this self-feeding mechanism: books that help people to learn how to read, for instance. The fact that those books not only create another reader - potentially avid for more reading - but can also be shared and create not only one, but possibly many new readers. The fact that the more readers read, the more books are published, and the cheaper, consequently, those books become, thus encouraging new layers of the population to acquire literacy. The fact that publications often call for their published refutation, which in turn calls for a response to the refutation. This mechanism - very much alive in the 17th century - is still at work today.
One neglected area of inquiry in the history of the book, in my opinion, is the study of books as process, as part of a unifying, evolving, living procedure. What role does print play in communication, in tying, splicing, replicating ideas, in generating new, slightly or radically deviant copies of themselves? How does the publishing industry, individual writers or booksellers evolve survival rules for ideas and themselves?

Everything begins and ends in a book. That circularity however is not perfect. Although motivated by a similar impulse, the hand-press, the steam-press and today's computers are not, in their effects or means, equivalent, except in the sense that they all serve the purpose of increasing the availability of information. Widely different quantitative measurements do not simply represent "more of the same", but a qualitatively different environment. A second edition is qualitatively, if not textually, different from a first.
Our modes of thinking, today, in full knowledge of the fact that we cannot possibily hope to learn but an infinitesimal fraction of what is there to learn, cannot be the same as those of the 16th or 17th century learned man or woman. The process of reading, of decoding  information in our age must be unique to us.
The new opportunities now opened to us by the tremendous and unique work accomplished by the various teams involved in the compilation of the STC, Wing and ESTC catalogs, and the electronic media, will help us investigate new avenues of knowledge and understand better one of the most rewarding purposes of history: understanding ourselves in the ever-changing light of our own present.