The earliest book discovered in which appeared indicia which may properly be termed marks of quotation was printed in 1516 at Strasbourg, Alsace (then in Germany), by Mathias Schurer. It was “De Vitis Sophistarum” by Flavius Philostratus. The marks consisted of two commas in the left hand margin of each page outside the regular type measure. They were placed at the beginning of each line in which a quoted passage appeared, and were evidently added after the page was set up, because their alignment varies greatly.

☛ Concerning Quotation Marks by Douglas C. McMurtrie, New York: privatly printed, 1934, p. 4. Read the whole essay: PDF.

[UPDATE–Dec. 4, 2012] See the end of this post for an additional reference to Prof Ruth Finnegan’s highly relevant book ‘Why Do We Quote? The Culture and History of Quotation’ (Open Book Publisher, 2011)

This 7 pages essay by American typeface designer and historian Douglas Crawford McMurtrie (1888-1944) was first published as an article in the Gazette of the Grolier Club in April of 1926 (no. 8). It was than privately reprinted in a limited edition of 200 (or so it says on the brochure). Surprisingly, original copies are still sold online (at the time of writing): see for example Oak Knoll and AbeBooks.

As it happens, copies of De Vitis Sophistarum were digitized by the Munich Digitization Center (MDZ) and are available via their Digital Collection: see Flavii Philostrati de Vitis Sophistarum Libri duo. Sure enough, one can find the marks of quotation mentioned by McMurtrie on several pages. Below is an example of a six lines quotation found on page 29:

In his major study on the uses of quotation La Seconde main, ou le travail de la citation (1979) ―which remains untranslated― Antoine Compagnon is tempted to see the first occurrence of quotation marks in the French edition of Petrus Ramus’s Dialecticae partitiones published by André Wechel at Paris in 1555:

En revanche, dès la première édition française de la Dialectique, parue en 1555 chez André Wechel à Paris, toutes les citations insérées dans le texte sont distinguées par la typographie: quand elles sont en vers, elles sont imprimées en italique, l’ensemble du texte étant en romain, et le nom du traducteur (le plus souvent Ronsard) figure au bas du passage; quand elles sont en prose, exclusivement celle de Cicéron, une virgule retournée [turned commas] se trouve en marge, à hauteur des lignes où la citation commence et se termine, deux virgules retournées [two turned commas] en marge des lignes intermédiaires. Ces indicateurs qui annoncent les futurs guillemets, sous la forme qu’ils prendront au siècle suivant, représentent une innovation capitale. Il n’y a que je sache, pas d’antécédent à une telle précision dans la citation: début et fin repérés sans équivoque, auteur et traducteur désignés. (Paris: Seuil, 1977, p. 246-247)

A copy of the 1555 French edition of Dialectique by Pierre de La Ramée (as he was also known) is available at BNF’s digital library Gallica. Here’s what an eleven lines quote from Ciceron looks like (from page 10):

Compagnon goes on quoting the work of “Douglas C. Mac Murtrie”, but he is referring to a much earlier and much shorter article (about a page and a half): “Origins and Development of the Marks of Quotation” also available online (The Library, Volume s4-II, Issue 2, September 1921, pp. 133-134, PDF, access may be restricted). In this 1921 “memorandum” –as he calls it– McMurtrie basically informs the readers about his intention “to study and report upon […] the origin and development of the marks of quotation”. He therefore invites readers ―printers, bibliographers, or librarians― to send him “any suggestions which may aid in [his] research, or any notes on the origin or history of quotation marks in English printing”. In a concluding paragraph, he quickly sketches the actual state of his research:

Research up to the present date indicates that the marks of quotation had their origin in France some time about 1580-1590, two commas or turned commas in the margins being used to indicate cited passage. (p. 134)

Based on this statement and on the fact that Ramus’s Dialectique was published prior to the period suggested by McMurtrie, Compagnon is tempted to see in the Dialectique the very first book to make use of quotation marks:

Or la Dialectique de Ramus fut publiée quelque trente ans auparavant [thirty years before “1580-1590”], où figure précisément ce signe: il serait tentant [it would be tempting] de considérer que ce livre fut à la fois le premier ouvrage philosophique en langue française et le premier à recourir aux guillemets, les deux phénomènes étant d’ailleurs coordonnés puisque ce sont des morceaux traduits d’une langue étrangère qui sont signalés. (p. 247)

Ironically, in his later essay Concerning Quotation Marks which I quoted at the very beginning of this entry, McMurtrie offers this piece au cautionary advice:

It is never safe to say just when the first use was made of any particular typographic device, for no sooner is the statement made than someone discovers an earlier appearance. (pp. 3-4)

As we know, McMurtrie goes on mentioning the existence of quotation marks in a book ―De Vitis Sophistarum– published 39 years prior to Ramus’s Dialectique. Furthermore, McMurtrie also identified the oldest known French book to make use of quotation marks: it is Geofroy Tory’s Champ fleury published at Paris in 1529, 26 years before Ramus’s Dialectique (see Gallica and the illustration below). Campagnon was wise to be only tempted by the idea of declaring the Dialectique the very first book to use quotation marks.

In 1983, a certain C. J. Mitchell published a much more elaborated paper in the very same publication in which McMurtrie had published his memorandum of 1921: “Quotation Marks, National Compositorial Habits and False Imprints” (The Library, volume s6-5, issue 4, December 1983, pp. 359-384, PDF, access may be restricted). Although the 26 pages paper is more specifically concerned with the introduction of quotation marks “in characteristically different ways in different parts of eighteenth-century Europe” it does briefly address the problem of the origin of this typographic device:

The origins and development of quotation marks are obscure. The occasional use of marks of some kind seems to be a very ancient custom, but modern practice, systematic and largely obligatory, seems to have followed upon the invention of printing. Certainly, medieval scribal methods such as indicating citations by underlining, sometimes in coloured ink, were not readily transferable to the new technology.

Exactly when quotation marks were first used in printing is not known. In 1922 [sic] Douglas C. McMurtrie appealed in The Library for information on the first use and subsequent development of quotation marks. Later, after receiving ‘numerous helpful letters’, he wrote that ‘the earliest book discovered in which appeared indicia which may properly be termed marks of quotation was printed in 1516 in Strasbourg, by Mathias Schurer. It was “De Vitis Sophistarum” by Flavius Philostratus’. Within six years, in 1522, Aldus used quotation marks in Italy, in Libri V de Asse by G. Budaeus, and in 1529 they appeared in Paris, in Geoffrey Tory’s [sic] Champ Fleury. (pp. 362-363)

Mitchell adds the following note to this summary, linking the emergence of quotation marks to technical aspects (same page):

The early 1500s was when printers began to move away from printing works intended to resemble manuscripts, and quotation marks thus seem to be an aspect of that liberation: Colin Clair, A History of European printing (London, 1976), writes both that the liberating process took nearly a century from the invention of printing (p. 21), and also that it took only until around 1520 for printed books to free themselves (p. 126).

Contrary to McMurtrie, Mitchell provides numerous references to support his research. His essay alone would be very interesting, but the bibliographical references in his paper makes it a truly precious document in regard to the development of the quotation mark as a typographical device.

Finally, one cannot avoid mentioning Robert Bringhurst now classical The Elements of Typographic Style (Vancouver: Hartley & Marks, Second Edition, [1992]1997). In the short history of quotation marks he proposes, Bringhurst adds an explanation regarding “intended block quotation”. As Antoine Compagnon did, he too puts the appearance of the first mark of quotation around 1550, a bit late in light of McMurtrie discoveries:

Typographers got by quite well for centuries without quotation marks. In the earliest printed books, quotation was marked merely by naming the speaker ―as it still is in most edition of the Vulgate and King James Bibles. In the High Renaissance, quotation was generally marked by a change of font: from roman to italic or the other way around. Quotation marks were first cut in the middle of the sixteenth century, and by the seventeenth, some printers liked to use them profusely. In books from Baroque and Romantic periods, quotation marks are sometimes repeated at the beginning of every line of long quotation. When these distractions were finally omitted, the space they had occupied was frequently retained. This is the origin of the intended block quotation. Renaissance block quotations were set in a contrasting face at full size and full measure. (p. 86)

[UPDATED INFORMATION–Dec. 4, 2012] A few days after I published this post, I became aware of Prof Ruth Finnegan’s book ‘Why Do We Quote? The Culture and History of Quotation’ (Open Book Publishers, 2011). Concerning the origins of the quotation mark, it is by far the most exhaustive study I’ve found. Section 4.1 of the book is particularly relevant to the subject at hand here: “What are quote marks and where did they come from?” (pp. 80-94). Prof. Finnegan starts her inquiry way before the “double inverted commas” of the 16th century with the use of the diplē in ancient Greece:

So a second route to follow is to ask where the quote marks that are so highly visible today came from. What is the origin of our familiar inverted commas? The answer would seem to lie in a little graphic sign used by ancient Greek editors to draw attention to something noteworthy in the text.

>

Shaped like an arrowhead, it was known as the diplē (‘double’) from the two lines that formed its wedge shape. Its prime purpose was not, as now, to enclose quotations but to act as a marginal signal for drawing attention to some particular aspect of the text. In the Western written tradition, this basic shape seems to have been the root of the variegated plenitude of modern forms.

Its vicissitudes make up an intriguing story. In early Greek manuscripts the diple was a kind of all-purpose marginal mark to indicate something noteworthy in the text, linguistic, historical or controversial. It drew attention to such features as a dubious or variant reading that might need amending or elucidating, the start or end of a section, a new episode, the interpretation of particular word or phrase, or a passage to be commented on. (p. 86)

She then shows how this arrowhead mark evolved into a great variety of graphical marks, something resembling an elongated comma. She also provides visual evidence from manuscripts produced between the 7th and the 9th century:

The diple sign went on changing and varying, going through an extravaganza of shapes and usages. Usually set along the left-hand margin (occasionally the right), their shapes and numbers varied with geographical region, sometimes with the individual scribe. They often looked very unlike the original wedge- shaped diple (some idea of the many variants can be conveyed by the examples in Fig. 4.9). They came as dots (single, double, multiple), as squiggles, curves, horizontal lines, double or single s-like flourishes, r-shapes, V’s or Y’s with or without dots inside (a diple turned on its head), a cross, double or single comma-like curves, a single or double horizontal stroke sometimes with dots above and below, sometimes in red – and yet other permutations. A reversed diple with dot was used in sixth-century gospels in Syriac, pairs of sinuous strokes in the left margin marked quotations from scripture in an eighth- century example, and a sixth-century insular manuscript showed citations by ‘a tiny flourish (7-shaped or like a bird in flight) to the left of each line’ (Lowe 1972: 4). There was also the idiosyncratic version known as the ‘English form’: a series of one, two or three dots followed by a comma, sometimes in red. (pp. 88-89)

She also addresses the emergence and (mechanical) standardisation of the double-inverted commas at the beginning of the 16th century:

By the early sixteenth century however comma-shape quotation marks, successors of the diple, were being cut in metal type and, after various experiments with shape and placing, printing houses were beginning to construct their own standard type pieces for reproducing them. They were being more plentifully used, at least by some printers, by the seventeenth century, mostly as doubled semicircular signs, now becoming known as ‘inverted’ or ‘turned’ commas. The signs gradually became more systematised and standardised: the use of type increasingly ensured a greater consistency than in the variegated handwritten manuscript scripts. (p. 90)

Fortunately for anyone interested in those matters, Prof Ruth Finnegan’s book was published under a Creative Common license (CC BY-NC-ND 2.0). One can read it online for free or download a complete PDF copy for a small fee (about $8 USD).

Previously here, indirectly related to this matter: “Curator’s Code”: some remarks on authorship and source attribution