V and U in 17th-century Flemish book titles

For many years bibliographers in Flanders have been speculating about the use of “V” in the place of “U” on title pages of early modern hand-press books. For the occasion of this blog post, I decided “TO TAKE VP THE GAVNTLET” in figuring out whether my friend Diederik Lanoye was right when he insisted that “everything happened in the sixties!”

A Latin title page from 1602 (left) and one from 75 years later (right): the former has “V” representing “U” while the latter does not.

When it comes to books printed in Flanders, we all agree that title pages from the beginning of the 17th century have a different disposition than those printed a century later. One of the things that change is the use of the uppercase “V” and “U” throughout the title. In the beginning of the century, compositors often put a “V” on the composing stick where modern spelling would expect a “U.” In addition, “W” is often represented by a double “V.” The combination of both habits sometimes leads to, for us, very strange word images, such as “VROVVV” for “vrouw” (woman), or “TROVVV” for “trouw” (faithful, true; fidelity). Of course, this phenomenon does not go by unnoticed, especially when one has to render diplomatic transcriptions of these titles to enter them in an online bibliography. This V-for-U practice slowly fades away and becomes rare by the end of the 17th century. But exactly when does it happen? And why? 

Optimistically, I reckoned that a couple of hours of data-crunching would suffice to solve this question. What I needed was a database of bibliographic descriptions maintaining the use of “v” and “u” in title words, even when modern orthography would replace the “v” with “u.” Among others, the German retrospective bibliography VD16–VD17–VD18, the Dutch Short Title Catalogue Netherlands (STCN), and the Short Title Catalogue Flanders (STCV) keep this usage in their title transcriptions. I decided to use STCV and asked the project manager for a download of the entire dataset. ¹

The download I received consists of 19,857 descriptions of hand-press books published in Flanders between 1473 and 1861, of which 11,754 editions date from the 17th century. In order not to complicate things, I retained only those descriptions with a main title on the title page and with a title-page imprint showing a date, which let me use the date featuring in the fingerprint to sort my records. After this operation, my dataset consisted of 7,389 titles from the 17th century with a known date of publication.

In total, the titles contain 97,026 individual words, of which 18,681 (about 19%) have one or more “v”s. The final step in this process concerned the marking of all those instances in which “v” was used instead of “u”—probably one of the most dreary hours I’ve spent in service of the Folger. As a result, 6,861 words with “v” representing “u” were marked, almost 37% of all v-words.

Title page of a 1655 Antwerp edition showing title words with “v” where modern orthography would require “u” (e.g. “SCRIPTVRE” instead of “SCRIPTURE”)

There are several possibilities for analyzing this data. For now, I will only discuss the trends over time relating to the individual v-words.

Graph 1 shows for each period of five years how many v-words of the total number of v-words for that period use a “v” instead of a “u.” This collection includes v-words in Latin (8,125) in addition to v-words in Dutch (9,655), French (1,656 about 1600), English (49), Italian (176), Spanish (315), and other languages. Until the 1650s, this fraction seems to remain more or less stable, and from 1661-1665 onwards, this number begins to drop at a steady rate. For the period 1656-1660, I counted 423 v-for-u words out of a total of 966 v-words or 44%; in the period 1696-1700 this number was as low as 36 out of 809 v-words.

The second graph indicates a significant difference between v-words in Latin and Dutch titles. In Dutch titles, only 13% of the v-words have “v” representing “u,” whereas this number is as high as 61% for Latin words. Of course in Flanders, Latin-language works are mainly printed with roman type and (until the 1660s) works in Dutch mainly in black letter. But in the second half of the 17th century, most printers selected roman type for text, even for Dutch-language books. ² But, as I have explained elsewhere, there is an important difference between the type used for text for the book proper and types used on title pages. ³ From the 1540s onwards, Dutch-language title pages begin to mix different type families, such as black letter, roman type (both lower and upper case), small caps, italics (both lower and upper case), and civilité. Consequently at the beginning of the 17th century, roman type, small caps, and italics are no longer a rarity on Dutch-language title pages and often appear next to black letter, which begins to fade away on title pages from the 1620s onwards. The steady increase of other type families on Dutch-language title pages can explain the increase of v-for-u words from about 15% in 1601-1605 to about 21% in 1636-1640, probably the cause of the sudden bump in graph 1 in this period.

But after 1665, the portion of Dutch v-for-u words ceases to increase—on the contrary they begin to disappear, and by the turn of the century they are almost totally gone. Interestingly, the same trend can be observed in Latin, but the tipping point seems to come about a decade earlier. In the period 1651-1655, Latin v-for-u words drop from 70% at a rate of about 1% per year. By 1700, only 11% of all Latin v-words substitute “u” for “v”.

It seems safe to conclude that Latin sets the tone and that Dutch follows the trend with some delay and with a lesser intensity. This corresponds with other typographic phenomena, for instance the use of the Single Vine Leaf during the first forty years of the 16th century. But that does not explain why the change happens. Of course, in Latin books “v” representing “u” on title pages are not odd at all as they come out of epigraphic customs. But around the middle of the 17th century, that cultural link to antiquity begins to erode, until it seems to be completely broken by the beginning of the new century. Again, we are confronted with an incremental trend change, and it is difficult to tell who got this process going. For the time being, we have to content ourselves with the conclusion that, for Dutch titles, “everything happened in the sixties.”

Updated February 10: The frequency of v-words in various languages has been modified since the original numbers didn’t add up to the total number of counted words; this minimal change doesn’t affect the general findings.