Incunabula: A Guide to Fifteenth-Century Printed Books, 1450–1500

Incunabula are often introduced simply as books printed before 1501. The definition is correct, but it conceals much of what makes fifteenth-century printing historically important. An incunable belongs to the first half-century in which European texts could be reproduced mechanically in hundreds of substantially similar copies, yet many of the habits surrounding its production, decoration, sale and use still belonged to the manuscript world from which printing emerged.

This is why incunabula reward examination as physical objects rather than as old editions alone. Type, paper, watermarks, rubrication, illumination, binding, manuscript annotations, ownership inscriptions and even defects can recover information about where a book was made, where it was sold, who read it and how it travelled. Two copies of the same edition may consequently have very different historical significance, even though their printed text was produced from the same setting of type.

The scale of the surviving material is considerable. The Incunabula Short Title Catalogue, or ISTC, maintained by the British Library and hosted by the Consortium of European Research Libraries, now contains more than 30,000 bibliographical records for fifteenth-century printing, together with holdings information from collections around the world.[1] That corpus is large enough to reconstruct patterns of production and distribution across Europe, but incomplete enough to remind us that the books which survive are not a neutral sample of everything that was printed.

What are incunabula?

The conventional definition covers material printed from movable type in Europe up to the end of the year 1500. It includes substantial folio Bibles and legal commentaries, but also quartos, small devotional works, indulgences, calendars, proclamations and other single-sheet printing. The category is therefore broader than the modern idea of a “book”, while blockbooks and other forms of printing not produced from movable type are normally treated separately.[2]

The word itself is much younger than the books it describes. It derives from the Latin incunabula, associated with a cradle or beginnings, and was applied retrospectively to the early period of typography. Bernhard von Mallinckrodt used the expression prima typographicae incunabula in his 1640 history of printing, while Cornelius van Beughem's Incunabula typographiae of 1688 helped establish the term bibliographically.[2]

The boundary at 1500 is a scholarly convention rather than a technological event. Printers did not change their presses, type or page design on 1 January 1501, and a book printed in 1502 may be typographically almost indistinguishable from one produced in 1498. The term remains useful because it creates a manageable corpus for bibliography, but specialists should resist treating 1500 as a historical rupture in the development of printing.[2]

This problem becomes especially obvious when dealing with undated books. Fifteenth-century printers did not always give a year, place or even their own names, and different local calendars could also complicate apparently explicit dates. Modern catalogues therefore frequently use approximate formulations such as “about 1480”, “not after 1476” or “between 1495 and 1500”, reflecting the actual state of the evidence rather than an artificial precision.

The printed book did not replace the manuscript overnight

The earliest printers were producing books for readers who already knew what a serious book was supposed to look like. Their natural visual model was therefore the manuscript codex. Early typefaces reproduced familiar Gothic bookhands, abbreviations remained abundant, biblical and legal texts continued to be arranged in multiple columns, and printed commentaries often reproduced the sophisticated hierarchy of main text and surrounding gloss familiar from manuscripts.[3]

The relationship went much further than imitation. Many incunabula left spaces for initials that were expected to be supplied by hand after printing. Rubricators added red or blue initials, paragraph marks, headings and capital strokes, while wealthy purchasers could commission full illumination, borders, coats of arms or miniatures. Printed guide letters were sometimes placed inside the blank spaces to tell the rubricator which initial was required.[3]

The result was a genuinely hybrid object. The press supplied the repeatable body of the text, but the book could remain materially unfinished when it left the printer's workshop. Decoration, binding and sometimes even corrections or missing passages were subsequently supplied by other craftsmen, often in another city.[4]

This makes the familiar opposition between “manuscript” and “printed book” misleading when applied too rigidly to the fifteenth century. Printing transformed the reproduction of texts, but manuscript labour remained embedded within the production, completion and use of printed books. An elaborately illuminated incunable may contain hundreds of pages produced mechanically while nevertheless presenting itself to its first owner as an individually finished luxury object.

Printing changed scale before it created uniformity

The economic novelty of movable-type printing lay in the ability to distribute the cost of composition and presswork across many copies. A manuscript normally had to be copied again if another copy was required, whereas a printer could produce hundreds of impressions from the same setting of type. Known fifteenth-century print runs vary greatly, but runs numbering several hundred copies were entirely normal, and larger commercial projects could exceed them.[3]

That multiplication created a new problem which manuscript production had not faced on the same scale: distribution. Producing several hundred books was only profitable if purchasers could be found for them. Printers, publishers, merchants and booksellers therefore developed networks capable of moving books between cities, across political boundaries and into institutions and private libraries far removed from the press where they had originated.[5]

Latin books were especially suited to long-distance trade because they could address educated readers across much of Europe. A legal commentary printed in Venice could find buyers in German universities; a theological text produced in Strasbourg could travel to monasteries in Poland; liturgical books commissioned for a particular diocese might be printed hundreds of kilometres from the place where they were intended to be used. The geography of printing and the geography of readership were therefore never identical.

This matters when cataloguing incunabula as Polonica. A fifteenth-century book does not require a Kraków imprint to have substantial Polish significance. Works by Polish authors, books printed for Polish dioceses, volumes imported into Polish monasteries and universities, and copies carrying early Polish provenance can be historically more important to Poland than many books physically printed within the Kingdom.

What was actually printed in the fifteenth century?

The surviving corpus can give a misleading impression of fifteenth-century demand. Large theological, legal and liturgical books survive relatively well because they were expensive, substantial and frequently entered institutional libraries. Schoolbooks, popular devotional texts, calendars, indulgences, political ephemera and practical publications were more likely to be worn out, discarded or recycled.[3]

Modern statistical work has made the scale of this problem much harder to ignore. Jonathan Green, Frank McIntyre and Paul Needham showed that the distribution of surviving copies suggests a substantial population of editions which have disappeared completely, while later modelling has reinforced the likelihood that many thousands of fifteenth-century editions are now lost.[6] Small-format and ephemeral printing is particularly vulnerable, so the surviving incunable corpus should never be treated as a complete representation of fifteenth-century reading.

The problem can be expressed simply. A substantial scholastic folio preserved in fifty libraries was not necessarily fifty times more commercially successful than a school text surviving in a single fragment. The latter may originally have been printed in hundreds of copies and then consumed through repeated use.

This distinction is particularly important for rarity. A book represented by one surviving copy may indeed have been printed in a very small edition, but it may instead belong to a category with an exceptionally high rate of destruction. Bibliographical rarity tells us how many copies are known now; it does not by itself reveal the original size or cultural reach of the edition.

How an incunable was constructed

The vocabulary used to describe an incunable still comes largely from the physical structure of the handpress book. Folio, quarto and octavo are formats, determined fundamentally by the way a printed sheet was folded and imposed, rather than simple statements of height. A large quarto and a small folio can therefore overlap considerably in physical dimensions.

Printed sheets were folded into gatherings or quires and sewn together for binding. Signatures, usually letters accompanied by numbers or other marks, increasingly helped the binder place the gatherings and leaves in the correct sequence. Registers appearing at the end of some books recorded the sequence of signatures and provided another means of checking completeness.[2]

Pagination was by no means universal. Many incunabula have neither printed page numbers nor foliation, making bibliographical collation essential when determining whether a copy is complete. A professional description therefore records the sequence of gatherings and leaves rather than relying upon a simple page count.

Paper provides another body of evidence. Most surviving incunabula were printed on paper, while some editions also contained copies or portions printed on parchment for more expensive presentation. Watermarks can sometimes help establish paper stocks, chronology or geographical connections, although they should be used in conjunction with typographical and documentary evidence rather than treated as an independent dating mechanism.

The binding requires similar caution. A fifteenth-century printing in a contemporary binding preserves evidence very close to the moment of first ownership, but incunabula were rebound constantly over the following five centuries. A seventeenth- or nineteenth-century binding may still be historically important, but it belongs to a later stage in the life of the copy and should not be allowed to suggest a false picture of its original appearance.

Why copies of the same incunable can be radically different

The mechanical reproduction of text did not create identical books in the modern industrial sense. Press corrections could alter individual sheets during a print run, and some editions survive with typographical variants or differing states. More conspicuous differences arose after printing, when individual copies received different decoration, bindings and manuscript additions.

One copy might have plain red initials; another could contain elaborate painted decoration and gold. A third might never have been rubricated at all, leaving empty initial spaces and printed guide letters visible. These distinctions can identify regional workshops, levels of expenditure and sometimes the intended market for particular groups of copies.[7]

Annotations create an even greater divergence. University students structured texts with headings and pointing hands; theologians added references; owners translated difficult vocabulary; readers corrected printing errors or compared the text against another exemplar. Greek passages which a printer could not set were sometimes supplied by hand, while damaged or missing leaves might be reconstructed in manuscript.

Binding history can preserve equally valuable evidence. Several separately printed works bound together at an early date form a Sammelband, and the association between those works may reconstruct an individual reader's interests or a particular institutional curriculum. When such a volume is broken up, information about early ownership and intellectual context can be lost even if the individual printed works become easier to sell separately.

The implication is fundamental to incunabula research: the edition and the copy are different historical units. The edition belongs to printing history and textual transmission; the individual copy belongs additionally to the histories of reading, ownership, collecting and libraries.

How anonymous and undated incunabula are identified

Modern bibliographical precision can disguise how little identifying information many early books originally supplied. The title page developed gradually and was not yet a universal feature. Printer, place and date may instead appear in a colophon at the end, while many books contain no explicit imprint information at all.[2]

Bibliographers consequently developed methods for identifying books from their typographical material. Different printing houses owned identifiable fonts, initials, woodblocks and printer's devices. Measuring type and comparing distinctive letterforms allows apparently anonymous printing to be associated with a particular workshop or group of books.

The Proctor-Haebler method became central to this work. Type can be classified partly through the design of the capital M and through measurements based on the height of twenty lines, allowing unknown material to be compared with type used in signed or securely dated editions. The modern Typenrepertorium der Wiegendrucke continues this tradition digitally and now records thousands of fifteenth-century types alongside initials, printer's marks and other typographical material.[8]

Adolf Rusch of Strasbourg offers a particularly revealing example. Because his early books were unsigned and undated, the printer was long known bibliographically as the “R-Printer” from the distinctive majuscule R found in his roman type. Comparison of typography and documentary evidence eventually associated this anonymous group with Rusch, although the precise relationship between Rusch and Johann Mentelin in some productions continues to be treated carefully in specialist catalogues.[9]

Copy-specific evidence can also help date an edition. A dated ownership inscription proves that a book existed before the recorded acquisition; dated rubrication may provide an equally valuable terminus ante quem. The ISTC record for Nicolaus de Lyra's Postilla super totam Bibliam, printed at Strasbourg by Johann Mentelin, for example, notes a rubricator's date in one copy and a purchaser's date in another as evidence for dating the edition not later than 1472.[10]

ISTC, GW and the modern bibliography of incunabula

No serious work on an incunable should rely upon a single general library catalogue. The fundamental starting point is normally the Incunabula Short Title Catalogue, which functions as an international union catalogue for fifteenth-century printing and records more than 30,000 editions together with surviving holdings.[1] Its stable ISTC identifiers have become standard references in auction catalogues, institutional records and scholarship.

The Gesamtkatalog der Wiegendrucke, usually abbreviated GW, approaches much of the same corpus through a more detailed bibliographical tradition and remains indispensable for identifying editions, variants and typographical attribution. The Staatsbibliothek zu Berlin describes the GW as covering more than 30,000 editions in over half a million recorded copies, while its associated Typenrepertorium provides the specialised typographical evidence required for anonymous and unfirmed printing.[8]

Individual collection catalogues can provide far more detailed copy descriptions. The Bodleian's six-volume catalogue, commonly cited as Bod-inc, records textual and bibliographical information together with provenance, decoration and copy-specific characteristics. Similar catalogues for major institutional collections remain valuable because they often preserve observations that cannot be compressed into a union database.[11]

For Polish holdings, Incunabula quae in bibliothecis Poloniae asservantur, prepared under the direction of Alodia Kawecka-Gryczowa, remains a central reference and is normally abbreviated IBP. The history of that project is itself significant: work begun before the Second World War made it possible after 1945 to reconstruct both surviving holdings and catastrophic wartime losses in Polish incunable collections.[12]

Copy counts from these resources need interpretation. An ISTC holding list is an extraordinary research tool, but it should not automatically be translated into a claim that exactly that number of complete copies exists. Holdings may represent fragments, incomplete volumes or multiple copies at one institution, while private holdings and uncatalogued material remain harder to trace.

Provenance has changed the study of incunabula

Traditional incunabula bibliography concentrated heavily on identifying editions, printers and dates. Those questions remain essential, but modern research increasingly treats the surviving copy as historical evidence in its own right. The development of Material Evidence in Incunabula, or MEI, has been particularly important in shifting attention towards ownership, decoration, binding, annotation, prices and the geographical movement of individual books.[13]

MEI allows evidence to be associated with particular periods and places, making it possible to reconstruct the movement of books rather than merely identify where they are held today. A Venetian book now in London might have been decorated in Padua, owned by a Polish cleric in the sixteenth century, transferred to a religious institution and dispersed after a nineteenth-century suppression. None of that history is visible from the imprint alone.

Prices and methods of acquisition can be equally revealing. Early inscriptions recording purchase prices allow historians to compare books with wages and other commodities, while donations and institutional transfers reconstruct patterns of library formation. An erased ownership inscription, cancelled monastic stamp or superseded shelfmark may therefore deserve as much attention as a decorative woodcut.

This has substantial consequences for the antiquarian market. A copy whose edition is comparatively well represented institutionally may possess exceptional provenance or annotation, while a bibliographically rarer copy may remain relatively anonymous as an individual object. Edition rarity is therefore only one component of significance.

Incunabula and the beginnings of printing in Poland

Printing reached Kraków very early in the spread of the new technology. Kasper Straube, a printer from Bavaria, worked there during the 1470s, and his Almanach Cracoviense ad annum 1474, printed in 1473, is regarded as the earliest surviving print produced on Polish soil. The unique surviving copy is held by the Jagiellonian Library.[14] 

Straube's surviving output also includes three substantial Latin books, among them Juan de Torquemada's commentary on the Psalms, Francis of Platea's work on restitution and usury, and works of Augustine. His production illustrates the international character of early printing perfectly: a German printer using a new technology in Kraków produced Latin texts for an educated market whose intellectual culture extended far beyond the Kingdom of Poland.[14]

Kraków became important again in the history of scripts when Szwajpolt, or Schweipolt, Fiol printed Church Slavonic liturgical books in Cyrillic around 1491. His Oktoich belongs among the earliest Cyrillic incunabula, and Fiol's workshop occupies an important place in the history of Slavonic printing.[15] The surviving books also demonstrate that fifteenth-century printing in Kraków cannot be understood solely through Latin Western Christianity; the city's presses were already participating in the textual culture of Orthodox Eastern Europe.

Gdańsk provides another striking case. Konrad Baumgarten established a press there in 1498, and his Agenda sive Exsequiale sacramentorum of 1499 was prepared for the diocese of Vilnius by a Vilnius canon. It survives as an imposing red-and-black liturgical incunable and is an excellent example of the geographical complexity of early Polonica: printed in Gdańsk, commissioned for Lithuanian ecclesiastical use, and produced within a book trade whose craftsmen and materials moved freely across Central Europe.[16]

The same complexity applies to books printed abroad for Polish use. Kraków liturgical books were commissioned from Nuremberg printers, while large quantities of scholastic, theological, legal and classical printing entered Polish institutions from Venice, Strasbourg, Nuremberg, Basel and other centres. Early Polish book history is therefore a history both of local printing and of imported books.

What makes an incunable important?

Age alone is an inadequate measure. Every incunable is early in the history of European printing, but some survive in hundreds of copies while others are known only through a fragment. Importance can arise from the text, the edition, the printer, the geography of publication, typography, illustration, provenance, annotation, binding or a combination of these factors.

A first printed edition of an important classical or medieval text has a different bibliographical significance from a routine reprint. A provincial press represented by only a handful of surviving editions may be historically important even when the text itself is common. An extensively annotated copy can preserve evidence for teaching or readership that cannot be recovered from an untouched copy of the same edition.

Completeness also needs intelligent interpretation. Missing text leaves generally matter, but the loss of blank leaves or later endleaves does not have the same consequence. Conversely, a fragment can be of exceptional importance when it is the sole surviving witness to an edition, as frequently happens with schoolbooks and other ephemeral printing.

Condition should therefore be described rather than reduced to a crude hierarchy. Contemporary rubrication, staining caused by genuine early use, manuscript corrections and an unrestored binding may all be historically informative. Aggressive nineteenth- or twentieth-century restoration can sometimes remove precisely the evidence that now matters most to book historians.

Provenance can transform an otherwise familiar edition. A volume from a medieval monastery, the working library of a Renaissance scholar or an early Polish institutional collection occupies a different historical position from a copy without traceable early ownership. The strongest incunabula frequently combine bibliographical importance with a copy history that can be reconstructed across several centuries.

Incunabula as historical evidence

The study of incunabula has changed substantially since bibliographers first attempted to enumerate every fifteenth-century edition. Identifying the book remains fundamental, but current research asks wider questions about production, distribution, price, readership, annotation, institutional ownership and the survival of texts. Databases such as ISTC, GW and MEI make it possible to combine evidence that was once scattered through hundreds of printed catalogues and individual libraries.[1][8][13]

The most productive approach nevertheless begins with the physical book. Databases can identify an edition and list known holdings, but they cannot substitute for examining paper, type, decoration, binding, annotations and provenance. Significant evidence continues to be discovered because an owner wrote a date in a margin, a binder reused an older fragment, or a bookseller and cataloguer recognised that a supposedly ordinary copy carried traces of a much earlier history.

Incunabula consequently occupy a particularly productive position between medieval manuscript culture and the mature early modern book trade. They record the moment when texts became reproducible at unprecedented scale while books remained individual objects shaped by scribes, illuminators, binders and readers. The press created the edition; five centuries of ownership created the copy that survives.

For collectors and institutions, this is why the most interesting incunable is not necessarily the one with the earliest date or the smallest number of recorded copies. The decisive question is what survives in that particular book that cannot be recovered elsewhere. Once the printed text, material evidence and provenance are read together, an incunable can become a document not simply of early printing, but of the people and institutions through whose hands the new technology passed.

See our collection of available incunabula here.

Bibliography

[1] British Library, Incunabula Short Title Catalogue Data Release 2025, London: British Library, 2025; John Goldfinch and Karen Limper-Herz, “The Incunabula Short Title Catalogue (ISTC),” in Cristina Dondi, ed., Printing R-Evolution and Society 1450–1500: Fifty Years that Changed Europe, Venice, 2020, pp. 899–909. 

[2] National Diet Library, Japan, “What are Incunabula?”, “Definitions of Incunabula and Calendars” and “History of Incunabula Studies,” in Incunabula: Dawn of Western Printing. 

[3] Cristina Dondi, ed., Printing R-Evolution and Society 1450–1500: Fifty Years that Changed Europe, Venice: Edizioni Ca' Foscari, 2020.

[4] Lotte Hellinga, Texts in Transit: Manuscript to Proof and Print in the Fifteenth Century, Leiden and Boston: Brill, 2014.

[5] Lotte Hellinga, Incunabula in Transit: People and Trade, Leiden and Boston: Brill, 2018.

[6] Jonathan Green, Frank McIntyre and Paul Needham, “The Shape of Incunable Survival and Statistical Estimation of Lost Editions,” Papers of the Bibliographical Society of America, 105, no. 2 (2011), pp. 141–175. 

[7] University of Glasgow, Glasgow Incunabula Project, “Decoration,” “Manuscript Initials” and related indexes of rubrication and manuscript intervention in fifteenth-century books.

[8] Staatsbibliothek zu Berlin, Gesamtkatalog der Wiegendrucke and Typenrepertorium der Wiegendrucke. The latter continues the Proctor-Haebler tradition of identifying fifteenth-century type through measurable and diagnostic letterforms.

[9] University of Glasgow, Glasgow Incunabula Project, records for the Strasbourg “R-Printer”; see also the Bodleian printer index for the continuing attribution of the group to Johann Mentelin and Adolf Rusch.

[10] Incunabula Short Title Catalogue, ISTC in00133000, Nicolaus de Lyra, Postilla super totam Bibliam, Strasbourg: Johann Mentelin, not after 1472. The record notes dated rubrication and purchaser evidence used in establishing chronology.

[11] Alan Coates, Kristian Jensen, Cristina Dondi, Bettina Wagner and Helen Dixon, A Catalogue of Books Printed in the Fifteenth Century Now in the Bodleian Library, 6 vols., Oxford: Oxford University Press, 2005; see also the Bodleian's account of its holdings and cataloguing history.

[12] Maria Bohonos and Eliza Szandorowska, under the direction of Alodia Kawecka-Gryczowa, Incunabula quae in bibliothecis Poloniae asservantur, Warsaw and Wrocław: Ossolineum, 1970–1993; Michał Spandowski, “Polish collections of incunabules destroyed, scattered and dislocated in the time and as a result of World War II,” Rocznik Biblioteki Narodowej, XLIV (2013).

[13] Consortium of European Research Libraries, Material Evidence in Incunabula. The database records ownership, bindings, decoration, manuscript annotations, prices and other copy-specific evidence attached to ISTC records.

[14] Jacek Partyka, 550 lat druku w Polsce: katalog wystawy, Kraków: Biblioteka Jagiellońska, 2023; Muzeum Narodowe w Krakowie, “Początki druku na ziemiach polskich. Pierwsi drukarze i ich książki.”

[15] Aleksander Sokolyszyn, “Sweipolt Fiol: The First Slavic Printer of Cyrillic Characters,” American Slavic and East European Review, 18, no. 1 (1959), pp. 88–94.

[16] Biblioteka Narodowa, More Precious Than Gold, entry for Agenda sive Exsequiale sacramentorum, Gdańsk: Conradus Baumgart, 10 June 1499; see also the Encyklopedia Gdańska entry for Konrad Baumgart.