The Collation

Research and Exploration at the Folger

The use of paragraph marks in early 16th-century Flemish editions

(UPDATE January 23: In editing this post, I inadvertently inserted an inaccurate use of “Dutch” as a modifier in the post title; I’ve now updated it to the correct “Flemish”. SW.)

The Folger Shakespeare Library has very strong Continental holdings. The first time I did research in this library, I was pleased to find a number of editions printed by the first “Belgian” printer, Dirk Martens, also known as Thierry or Theodoricus Martens (c. 1447–1534). Martens published his first book in 1473 as the associate of Johannes van Westfalen. During his remarkably long career, spanning more than five decades, Martens moved four times: the first time in 1493 from Aalst, a town about 20 miles west of Brussels, to Antwerp; in 1498 to the university town of Louvain; in 1502 back to Antwerp; and ten years later again to Louvain, where he ceased his activities in 1529. In their 2009 biography of Martens, Renaud Adam and Alexandre Vanautgaerden list 269 extant editions, as well as 32 editions that either have been lost or never existed. ((Renaud Adam and Alexandre Vanautgaerden, Thierry Martens et la figure de l’imprimeur humaniste. Une nouvelle biographie, Turnhout: Brepols 2009.))

Martens is an important printer in addition to being the first. In 1503, he published a work by Erasmus for the first time; during his rich career, he was responsible for a total of 68 editions authored, translated, or edited by the prince amongst the humanists, including 32 first editions. ((Adam & Vanautgaerden, Thierry Martens, p. 100.)) Erasmus was not the only humanist with whom Martens collaborated. Late in 1516, for instance, he published the editio princeps of Thomas More’s Utopia, edited by Pieter Gillis and proofread by the Louvain professor and poet Gerardus Geldenhauer. ((About Geldenhauer, see P.J. Blok & P.C. Molhuysen, Nieuw Nederlandsch biografisch woordenboek. Deel 6. Leiden: A.W. Sijthoff 1924, pp. 550–554. Available online at: And from a typographical point of view, the importance of Martens can hardly be overestimated. He was the first printer in the Southern Netherlands using Greek characters, in 1591 1519, and at the turn of the century, he was the first to start using roman type. In 1518, Martens also introduced Hebrew type, and four years later italics.

The Folger possesses seven books emerging from Martens’s printing shop: one from 1501, two from 1509, one from 1517 and three from (approximately) 1519. The title page of the oldest edition shows a very rudimentary design: four justified lines of roman type of a single size on a furthermore blank page mention nothing other than the authors and titles of the text. The 1509 title page of Giovanni Pico della Mirandola’s Aureae epistolae combines a larger black letter with roman type, all printed in red, and has a large woodcut illustration in the middle, printed in black. 

title pages of the 1501 and 1509 Martens editions
title pages of the 1501 and 1509 Martens editions

The title page of the 1517 edition of Aesop’s fables bears a woodcut border in the shape of an architectural gate and features at the bottom the name of the printer. The 1519 title pages show a clear attempt to lay out the title in a triangular pattern, with a large first line spanning the entire measure of the type area, which is followed by shorter, centered lines of a smaller body. ((The purest form of this idea can be found in John Fisher’s Eversio munitionis adversus unicam Magdalenam, which I discussed in an earlier Collation post dealing with the single vine leaf.))

title pages of the 1517 and a 1519 edition
title pages of the 1517 and a 1519 edition

However different all seven title pages may seem, to my eye they had one thing in common: they all lacked paragraph marks, or pilcrows, a frequent typographical feature in this period of Dutch printing. This absence is consistent with the twenty or so title pages reproduced in Adam and Vanautgaerden’s biography. Only two title pages, both produced in 1515, bear a printed paragraph mark, in both cases marking the beginning of a new element of information on the title page. ((Adam & Vanautgaerden, Thierry Martens, ill. 19 and ill. 35.))

The 1501 title page holds the promise of a paragraph mark, albeit not a printed one but one to be added by a rubricator. The first line of the title is not accidentally but deliberately indented, as becomes clear when one compares different copies of the work. On the first page of text in the copy in the Antwerp heritage library Hendrik Conscience, one can see why there are blank indentations: the layout of this early sixteenth-century edition is still faithful to the traditional conception of  books which had to be finished by hand.

comparison of two copies of the first page of text (fol. A2v) of the 1501 Martens
comparison of two copies of the first page of text (fol. A2v) of the 1501 Martens

I had expected to find more printed paragraph marks in Martens’s editions; why weren’t they there? In an attempt to solve that question, I analyzed a database of the bibliographical descriptions of the first volume of Wouter Nijhoff’s and M.E. Kronenberg’s Nederlandsche bibliographie van 1500 tot 1540 (NK). I reviewed their descriptions of dated books printed in Flanders, 1,270 in total, and I noted if paragraph marks appeared in the title (fol. 1 recto), or elsewhere in the book, at least according to the descriptions in NK.

A description in volume 1 of Nijhoff-Kronenberg showing the use of paragraph marks in the title and elsewhere in the book.
A description in volume 1 of Nijhoff-Kronenberg showing the use of paragraph marks in the title and elsewhere in a book.

Image 09 Graph Paragraph Marks

As the black linear trendline indicates in the graph above, about 26% of title pages in the period 1501–1540 bear at least one printed paragraph mark. The strongest variance shows in the period 1516–1520 (17%) and in the five years before that (33%), the reasons for which are not yet clear. As far one can deduce from the descriptions in NK, paragraph marks are used, at least once, elsewhere in books in 50% of the cases, but much more often in the beginning of the century than forty years later. Again, their presence is the strongest in the period 1511–1515 (89%), but the overall trend is a shrinking one.

A survey of sixteenth-century folio bibles in the vernacular printed in Antwerp confirms this observation. Whereas pre-1540 bibles show ample use of paragraph marks as a means to add structure to the text, later bibles mostly drop this feature. ((Goran Proot, “Designing the Word of God: Layout and Typography of Flemish 16th-Century Folio Bibles Published in the Vernacular,” in De Gulden Passer 90:2 (2012), pp. 143–179.)) As a rule, paragraph marks become exceptional in seventeenth- and eighteenth-century Flemish imprints. They are clearly a characteristic feature of fifteenth-century books and those produced in the first half of the sixteenth century, but they seem to lose their original function as a means as a structuring element in the text.

But this does not explain why this typographical feature is so rarely found in Martens’s editions. The following tables explain this a little bit more.

Table 1. The use of paragraph marks on the title page and elsewhere in dated, Belgian editions according to the descriptions in vol. 1 of NK, according to format

title page elsewhere
folio 35% 81%
quarto 24% 51%
octavo 26% 44%

Table 2. The use of paragraph marks on the title page and elsewhere in dated, Belgian editions according to the descriptions in vol. 1 of NK, according to language (only mono-lingual editions)

title page elsewhere
Dutch 44% 85%
Latin 16% 32%
French 52% 67%

Table 3. The use of paragraph marks on the title page and elsewhere in dated, Belgian editions according to the descriptions in vol. 1 of NK, according to type face (editions can be printed in more than one major type face)

title page elsewhere
black letter 43% 82%
roman 15% 28%
italics 3% 7%

Table 1 indicates that paragraph marks appear more often in folio editions than in works of other formats, and Dirk Martens published the bulk of his editions as quartos. Table 2 shows that paragraph marks are used twice as often in vernacular works (either in Dutch or in French) than in Latin books, and Martens’s production is mainly in Latin. Finally, Table 3 makes clear that books printed (mainly) in black letter show the most intensive use of paragraph marks, and in the period 1501–1540 Martens used primarily roman type for his publications. After all, he introduced that type face in the Southern Netherlands.

To conclude, that Martens is not using paragraph marks very often is clear from the profile of his back list. We now also better understand the distribution of the use of this typographical feature stemming from the manuscript tradition and making its way, handwritten or in print, into hand-press books. In contrast to the single vine leaf, which is primarily linked with books in Latin, the paragraph mark is closely connected with large formats, editions in the vernacular, and the use of black letter. At least, that is the case in Flanders, but it is obvious that the situation differs from region to region, but exactly how, and, foremost, why, has not yet been found out.


Leave a Reply

  • (will not be published)