8.2 — The Printing Press

Before printing, a book had to be copied by hand. A scribe working full time produced perhaps two or three substantial books a year, and each copy introduced errors that the next copy inherited and added to.

A printing shop in 1500 could produce several hundred copies of a work in a few weeks, every copy identical.

The estimated number of books produced in Europe in the fifty years after Gutenberg exceeds the total produced in the previous thousand. By 1500 there were printing shops in over 250 European towns and something on the order of twenty million volumes in circulation, in a continent of perhaps 80 million people.

Nothing else in this volume changes the cost of an activity by that factor in one lifetime.

What Gutenberg actually invented

Not printing. Woodblock printing was centuries old in China and known in Europe for playing cards and devotional images.

Not movable type. Bi Sheng made clay type around 1040, and Korea was printing with cast metal type from the thirteenth century (Chapter 7.9).

What he invented was a system, and each part solved a specific problem.

The hand mould. This is the heart of it. A reusable adjustable mould into which molten metal is poured to cast a single letter, then opened and reset for the next. It produces identical type in quantity, quickly, with the same body height so every letter sits at exactly the same level on the page. Uneven type height means uneven inking and an unreadable page.

The alloy. Lead, tin and antimony. Lead alone is too soft and shrinks as it cools; antimony expands slightly on solidifying, which forces the metal into the fine detail of the mould and gives a sharp letter, and hardens the alloy enough to survive thousands of impressions. Getting this alloy right is a genuine metallurgical achievement.

The ink. Water-based scribal ink beads up on metal. He used an oil-based ink, closer to oil paint — soot or lampblack in varnish — which adheres to metal type and transfers cleanly to paper.

The press. Adapted from the screw press used for wine and olives, applying even pressure across a whole forme at once.

And paper. Chinese papermaking had reached Europe through the Islamic world by the twelfth century (Chapter 7.2), and by 1450 European paper mills — using water power and rag pulp — supplied it cheaply. Printing on parchment requires a flock of sheep per book. Printing is only economic on paper, and the two technologies had to meet.

The Gutenberg Bible, around 1455, was printed in about 180 copies. Gutenberg was then sued by his financier Johann Fust and lost his press and his types to him, which is a fittingly modern ending: the inventor went bankrupt and the investor took the business.

Why it exploded in Europe and not in China

Chapter 7.9 raised this and here is the answer.

The alphabet. European printing needs perhaps 200 to 300 distinct sorts including capitals, punctuation and ligatures. Chinese requires several thousand characters minimum for ordinary text. Casting, storing, finding and redistributing that many sorts is enormously more expensive and slower, and skilled compositors are harder to train.

So woodblock suited Chinese better, and China printed a great deal by woodblock — but woodblock has different economics: high cost per new title, very low cost per reprint. It favours reprinting established classics, which is precisely what Chinese printing largely did.

Movable type favours short runs of new titles, which is what pamphlets, controversy and news are. The technology matched the alphabet, and the alphabet was the accident.

And the market differed. Europe had a rapidly growing lay reading public, competing states with no central censorship, and a religious controversy about to break out (Chapter 8.7). China had a state whose examination system defined the canon, whose bureaucracy licensed printing, and which had no equivalent of a continent-wide argument in which each side needed to publish faster than the other.

What it changed

Standardisation and error correction

A hand copy is unique and its errors are unique. A print run is identical, so an error in one copy is an error in all of them — and it can be corrected in the next edition, and the correction reaches everyone.

This is the crucial epistemic effect and it is easy to miss. Scholarship before printing degraded with each copying generation. After printing, texts improve — errata are collected, editions are compared, and a corrected edition supersedes the previous one. Elizabeth Eisenstein's argument is that this cumulative correction, rather than the volume of books, is what made printing revolutionary.

It also made citation possible. Page numbers, indexes, title pages and tables of contents developed as printed conventions, and they only make sense when everyone's copy has the same text on the same page. A scholar can now write "see page 47" and be understood by a reader a thousand kilometres away. Chapter 8.8's science depends entirely on this.

Vernacular languages

Printers printed what sold, and what sold was books in the language people spoke.

This standardised languages. A printer had to choose one dialect's spelling and grammar, and the chosen version became the standard. Luther's German Bible substantially created standard written German; the King James Bible and the printers of London did the same for English.

And it hardened the boundaries between languages that had previously shaded into each other across regions — which fed directly into the national identities of Chapter 10.6.

The Reformation

Chapter 8.7 covers it. The printing connection is direct.

Luther's Ninety-Five Theses, posted or sent in 1517, were printed and circulated across Germany within weeks. Luther became the most published author in Europe — his pamphlets, in German, short, cheap, and often illustrated with woodcuts for the illiterate, sold in enormous numbers.

Previous reformers had been suppressed. Jan Hus was burned in 1415 and his movement contained; Wycliffe's followers were suppressed in England. The difference in 1517 was that the argument could not be contained, because suppressing a man is possible and suppressing a printed pamphlet already in ten thousand hands is not.

And the Church printed back, which made it a public argument conducted in front of an audience — the first mass media controversy.

Science

Chapter 8.8's revolution is not possible without printing.

Diagrams could be reproduced exactly. A hand-copied anatomical or astronomical diagram degrades badly; a printed one does not. Vesalius's anatomy of 1543, with its extraordinary illustrations, and Copernicus's book of the same year, both depend on it.

Results could be published, dated, and claimed. Priority disputes are a nuisance and they are also evidence of a system in which discovery is public and attributed.

And correspondence networks formed around printed works, producing by the seventeenth century the scientific journal and the learned society.

Politics and news

Pamphlets, broadsheets and eventually newspapers. The first regularly published newspapers appear in Germany around 1605.

Governments responded with censorship — licensing, indexes of prohibited books, and prosecution — and it worked imperfectly because Europe was politically fragmented. A book banned in France was printed in Amsterdam and smuggled back. The Dutch Republic became the printing house of Europe precisely because of its relative tolerance, and Chapter 8.9's Enlightenment ran on that.

The costs

A book about the printing press should include the other side.

The witch hunts were a print phenomenon in part. The Malleus Maleficarum, a manual for identifying and prosecuting witches published in 1487, went through dozens of editions and spread a systematic demonology far more widely and consistently than manuscript could have. The great European witch persecutions coincide with the print era.

Confessional hatred was amplified. The Reformation's pamphlet war included vicious propaganda on all sides, and the wars that followed killed on a scale Europe had not seen (Chapter 8.7).

And antisemitic literature was mass-produced, with real consequences.

The general point is neutral and worth stating plainly: a technology that lowers the cost of spreading an idea lowers it for every idea. Chapter 12.8 and Chapter 16.4 apply the same observation to the internet, where the pattern has repeated with recognisable precision.

The Islamic world's delay

Chapter 7.2 flagged this and it deserves the detail.

Printing in Arabic script was restricted in the Ottoman Empire for roughly three centuries. A decree of 1485 and its successors banned or restricted printing in Arabic and Turkish; the first Ottoman Turkish printing press producing books was established only in 1727 and operated intermittently.

Three reasons are advanced. Religious objection to mechanical reproduction of the Qur'an and of sacred names. The scribal guilds, which were large, organised and politically connected, and whose livelihood was threatened. And the aesthetic and cultural centrality of calligraphy, which is a major art form in Islamic civilisation in a way it is not in Europe.

Printing in other scripts was permitted — Jewish, Armenian and Greek presses operated in Ottoman territory from the fifteenth and sixteenth centuries — which shows the restriction was specific rather than a general hostility to the technology.

The consequence compounds. Three centuries in which European book production grew exponentially and Ottoman book production did not is three centuries of divergence in the accumulation and correction of knowledge. This is one of the more concrete and less speculative explanations offered for the divergence of Chapter 8.10.

And India's case is comparable but different. Printing reached Goa with the Portuguese in 1556 and was used mainly for missionary work. Indian script printing developed substantially only from the late eighteenth century, and a manuscript culture of extraordinary scale continued — Indian manuscript collections are among the largest in the world. The oral transmission traditions of Chapter 3.11 also meant that the most valued texts were carried in memory rather than in books.

Where this shows up in your life

Everything you read is a descendant of this. The page number, the index, the title page, the standard spelling, the edition number, the copyright, and the very idea that a text is a fixed thing that can be cited.

Typography's vocabulary is from the hand press: uppercase and lowercase from the cases the type was stored in, leading from the strips of lead used to space lines, and font from a complete set of one size and style of type.

And the mechanism is the one to carry forward. A drop in the cost of copying and distributing information produces, in order: more of it, more standardisation, more error correction, more argument, and more political disruption — in that sequence. It happened with print, and Chapter 12.8 shows it happening again in the last three decades with a compressed timeline and the same shape.

What the next page covers

The other Chinese invention that reached Europe did something quite different. Chapter 8.3 covers gunpowder and the military revolution — why cannon made medieval castles obsolete within decades, what the new star-shaped fortifications did about it, why the resulting arms race made war so expensive that states had to build tax bureaucracies to pay for it, and how that fiscal pressure is one of the main engines that produced the modern state.