Dating localizing medieval scripts

The criteria for dating presented here have been developed progressively in the systematic pan-European effort to identify dated and datable manuscripts, an enterprise that began in the mid-twentieth century. The project, known under the generic name of the Catalogue of Dated Manuscripts, intends to offer a solid footing for studies in paleography and the history of texts, indeed, for history in general, by creating a repertory of dated and datable manuscripts that provide reliable chronological and geographic benchmarks. For five decades the enterprise has moved forward without interruption, library by library, methodically if somewhat irregularly, depending upon the country and problems that arise with such publications. (1) The long duration of the work has given time to refine methods step by step, so that today it is relatively easy to give a broad view of the principle indications for dating that can be found in medieval manuscripts.

Chemists and physicists have not yet provided practical, nondestructive, or reliable tests for dating manuscripts. We thus depend on two types of indicator: explicit indicators, which must be treated critically, and implicit indicators, which have to be flushed out.

I Colophons

First among the explicit indicators is the colophon. The luckiest case for the researcher is obviously one where the scribe is charitable when he finishes his work. He himself informs us of his name, the date, the place where he is writing, and while he is at it, supplies a bit of information on the time it took and the circumstances surrounding his work. This type of indication is utterly exceptional before the tenth century. In most cases, only the name of the scribe or the patron is mentioned, and it is by cross-referencing this information with other indicators that we manage to date the volume. In these cases we speak of a "subscription." We reserve the term "colophon" for mentions that give fuller information, including the date and the place.

One of the best-known examples of a subscription is in the Maurdramnus Bible, one of our earliest witness to the new Caroline minuscule in the late eighth century: "Ego Maurdramnus abbas propter Dei amorem et propter conpendium legentium hoc volumen fieri jussi" (I, Maurdramnus, abbot, had this volume made for the love of God and for the benefit of the readers). (2) It is known that Maurdramnus was abbot of Corbie from 772 to 780. The Bible can thus be dated with a precision that is unusual for this period.

The use of colophons remains rare until the thirteenth century. They are a bit more frequent at the end of the Middle Ages but are not found in all or even most manuscripts. At Laon, the last library to be studied for the Catalogue of Dated Manuscripts in France, (3) colophons appear in one fifth of the manuscripts from the fifteenth century and in a third of those from the sixteenth century.

A manuscript at Soissons (4) (figure 1) has a dated colophon representing the rare ideal case where the scribe gives his name and the date but also the place of transcription: "Ce livre est a Jehan Thoulouse bouttiller en la viconte de l'eaue de Rouen escript l'an mil quatre cent soysante et quatre" (This book belongs to Jehan Thoulouse, cup-bearer in the viscounty of the waterways (5) of Rouen, written in the year 1464).

Of course, a colophon is not necessarily reliable out of hand. Two examples suffice to show that one has to be a bit cautious before rejoicing at having fallen upon an explicitly dated manuscript. In another manuscript at Soissons (6) (figure 2), the colophon reads: Fuit hic liber compilatus et completus Rome, anno Incarnationis dominice millesimo quadringentesimo quadragesimo sexto, VII kal. Marcii [23 February 1447 n. st.] pontificatus domini Nicolai pape quinti [Nicolas V, 1447-1455] anno primo, per reverendum patrem dominum Bernardum de Rosergio, prepositum Tholosanum (This book was compiled and finished in Rome, in the year of the Incarnation of our Lord, 1446 [old style], on the seventh kalends of March [23 February], in the first year of the papacy of Pope Nicolas the fifth, by Lord Bernard de Rosergue, provost of Toulouse).

The verb used in this colophon, compilare (which usually means the composition of the text and not its transcription), should arouse distrust. It should be the date of the work. But the colophon has clearly been corrected. Would the date of the work have been modified to coincide with the date of the transcription? Unfortunately, the answer is no. We know of at least two other copies of the same work that have also been modified so as to give the same date. (7) These corrections are simply the result of the fact that the work was finished on the very day Pope Eugenius the fourth died; Nicolas V was elected two weeks later, on March 10. The scribe or the corrector has thus hesitated between two names, and the dates have been changed accordingly. But the date refers to the composition of the work and not the scribal transcription, which in this case took place slightly later, in the second half of the fifteenth century.

Manuscript 456 in the municipal library at Laon (figure 3) has a subscription that gives the name of the scribe. On folio 95v, one of the five scribes who transcribed this volume has written: "[e]go Signorinus condam Petroboni Bentivegne de Bononia hoc opus scripsi" (I, Signorinus, son of the late Petrobonus Bentivegne of Bologna, wrote this work). Here there is nothing ambiguous in the word used: scripsi normally refers to the copy. Besides, as luck would have it, there are two other manuscripts copied by the same scribe. (8) Unfortunately, the script of the two other manuscripts has absolutely no similarities with that of the manuscript at Laon. Which colophon is lying? Alas, it is probably ours, as suggested by the layout. The colophon is not set off or detached but written like the rest of the text, and it begins with a guide letter, like the other paragraphs. This is relatively rare for a subscription. Furthermore, the text continues in the same script on the following folio. This mention is thus very certainly copied from the model. The scribe must have written it mechanically, without realizing it did not originally belong to the work he was copying.

In order to evaluate the reliability of a colophon, one must carefully consider the vocabulary. Scribere is not the same as componere or compilare. The layout and placement on the page are also vital. The more the colophon is integrated with the text either physically or in its formulation, the more it should be viewed with suspicion.

II The Ex Libris as a Possible Indicator of Origin

Just the opposite is true of the ex libris, which is another explicit indicator. But the ex libris cannot be considered as a valid indicator of origin unless it appears to have been written at the same moment as the transcription or immediately afterward and is thus fully integrated with the text. It is inconceivable that a scribe would copy either by error or by caprice the ex libris of his model into another manuscript. The function of an ex libris is too important and its formulation too recognizable. It cannot be mistaken for a part of the work copied.

On the other hand, a new ex libris may quite naturally be inscribed in a book each time it changes owners. The ex libris is therefore not a reliable indicator of origin unless it is clearly contemporary with the copy, indeed, by the hand of the scribe of the text. If it appears to be more recent or can be grouped with other identical mentions in manuscripts from the same library, it must be treated as the work of a librarian, which, even if it is close in time to the transcription of the book, does not automatically give us the place of production. A few examples will illustrate the problem.

The first is an ex libris for Saint-Vincent at Laon. (9) It is in the hand of the text and is a reliable indication of origin. We have here a rare bird, a dated ex libris that is also a colophon: "Hic liber Sancti Vincentii Laudunensis Cenobii, anno millesimo centesimo tricesimo sexto scriptus est." (This book belongs to the monastery of Saint-Vincent at Laon. It was written in the year 1136). The colophon-ex libris is followed by a traditional formula cursing the thief which is frequently added to such inscriptions: "Si quis illum quolibet ingenio ab ecclesia alienare voluerit, iram Dei et ipsius sancti in tremendo Judicio incurrat. Fiat Fiat, amen" (If anyone, in any way whatsoever, takes this book from the church, may he suffer the wrath of God and the saint on the terrible day of judgment).

Another original ex libris is found in a book from the Cistercian abbey of Vauclair, (10) near Laon (figure 4). Although not by the scribe of the manuscript, this ex libris uses the same ink as the text and respects its layout: here the scribe has been very careful not to write beyond the original ruling. Furthermore, this ex libris does not resemble the usual inscriptions written by the librarian of the abbey that occur in many volumes from the same abbey. The book was therefore clearly written for--if not in--Vauclair itself.

In two other manuscripts from Vauclair, manuscripts 42 and 108 at Laon (figure 5), the ex libris has been inscribed on the inside cover of the binding and thus has no relation to the copied text. The same ex libris occurs in closely related forms in several manuscripts from Vauclair. These inscriptions are clearly the work of the librarian and therefore give no information about whether the book was copied at the monastery or not.

Because of its function, the ex libris is a fragile indicator. It tells us about the identity of an owner at a given moment, but a book produced in the twelfth century may easily have been borrowed, exchanged, stolen, sold, or bequeathed half a dozen times before it arrived at the place where we find it today and thus may have known as many different owners and places. If we can usually depend on the colophon as an indicator, we find that in most cases it is difficult to use the ex libris for localizing production.

III Historical Indications and Additional Elements

In many cases, it is the examination of converging indicators that permits the construction of a logical proposal for dating and localizing a volume. Dating is generally made possible by the confrontation of a date implicitly or explicitly contained in the text (a terminus post quem) with other, added, datable elements that furnish the terminus ante quem.

This is the case of manuscript 132 from the municipal library in Soissons (11) (figure 6), which contains the Chronicle of Hugh of Saint Victor, compiled year by year. The chronicle was written by Hugh a little after 1130, but this copy is slightly later. In the most recent part (folio 429), additions were made from the beginning, and these allow us to date and localize the volume. A close examination of this leaf provides enough information for a precise dating:

1. The death of the author is mentioned in the margin by the original scribe: "in the year of the Incarnation, 1141, on the third ides of February (February 11) Master Hugh of Saint Victor died. He was the author of this work." Here is our terminus post quem: the copy must be after 1141.

2. The list of popes in Hugh's chronicle, which originally stopped with Honorius II (who died in 1130), has been brought up to date by a contemporary hand through the papacy of Eugenius III, who died in 1153. The volume was thus copied before his death; 1153 is our terminus ante quem. Here we already have a very satisfactory approximate date: 1141-1153. One can even go a little further. The years in the chronicle are entered by the original scribe up to the year 1145. This detail may give us the exact date of the copy.

3. With regard to the transcription, one will note the presence of a number of additions fully integrated in the text concerning the city of Soissons. For the year 1115 we read: "Lupus insanus infra urbem Suessionensem discurrens multos interficit" (A rabid wolf prowling around Soissons killed many people). Such additions obviously betray a local origin.

4. On the following folios there are one or two obits, one of which is that of abbas Theobaldus who dies in 1195. An abbot by that name died in that year at the abbey of Saint-Crepin in Soissons. That monastery is perhaps the place of origin of the volume.

In other cases, the insertion of texts concerning current events gives us clues for dating. Manuscript 64 (figures 7a, 7b) from the municipal library at Soissons (12) contains several short texts, all copied by the scribe of the main text and concerning a rather precise period:

1. A list of bishops of Auxerre on folio 1 ends with the name of Guillaume de Gretz, who was bishop of Auxerre from 1279 to 1293.

2. A list of the peers of France on folio 4v does not mention the peerages created in 1297 by Philippe le Bel; these peerages include the duchy of Brittany and the counties of Artois and Anjou.

3. At the end of the volume (on folio 219v-220) there is a satirical poem addressed to King Philippe le Bel (who reigned from 1285 to 1314); the poem is datable to 1290 because of its contents, which mention the youth of the king, who acceded to the throne at the age of seventeen in 1285, and it also speaks of the military campaign in Aragon from 1290 to 1291.

These little texts would be of no interest to the original owner if they were not contemporary with the people and events they allude to. What is the point of a list of bishops or peerages if it is not up to date? These additional elements form a group of texts that have great current interest and thus provide dependable indications for dating the volume. The date of the book must be after 1290 because of the terminus post quem given by the poem; it is probably before 1293, the date of death of Guillaume de Gretz, and it is definitely before 1297, when the three new peerages were created.

Happily for us, texts like these (lists of kings, popes, or bishops, local history, and so on), which are usually copied into blank spaces before or after the main work, can be found in many different contexts and not just historical works--while MS 132 contains a chronicle, MS 64 contains a work of biblical exegesis. It is worth recalling, however, that certain types of manuscripts, such as Bibles or patristic works, are rarely datable. Their transcription is usually as anonymous as it is careful.

IV Liturgy

Liturgical books offer fertile ground for discovering dating mechanisms because throughout the Middle Ages, the annual cycle of liturgical offices was constantly brought up to date, especially with regard to the celebration of saints' days. As these grew in number, they obliged the occasional reorganization of texts and calendars to make way for newcomers. It is thus not unusual to find dating elements among these revisions in liturgical manuscripts. Among the saints who furnish material for dating because they are the object of a widely venerated cult, one can cite the canonizations of Saint Thomas of Canterbury in 1173, Saint Bernard in 1174, Saint Francis in 1228, Saint Louis in 1297, and Saint Yves in 1348. One can also mention the Fete-Dieu, which was universally recognized in 1311 by a decision at the Council of Vienna, which is itself explicitly mentioned in several manuscripts. But this feast should be handled with extreme care, for it was already celebrated in certain dioceses as early as the mid-thirteenth century but is still not mentioned in a number of liturgical manuscripts from the fourteenth and fifteenth centuries. If the presence of these feasts in a liturgical manuscript furnishes a terminus post quem, their absence does not constitute evidence for dating. It can only confirm a dating arrived at on other grounds.

The study of dating mechanisms in the liturgy of the Cistercian order is particularly interesting. The evolution of the Cistercian sanctoral is known in great detail, (13) thanks to studies of the many documents drawn up by the annual general chapter of the order. Furthermore, the liturgical directives of the chapter were put into practice very rapidly, as illustrated in manuscript 213 at Laon (figure 8). In the office for the sick and dead, Saint Bernard is inscribed by the original hand with a double invocation in the litanies, in conformity with the Cistercian liturgy from 1200 onwards. On the other hand, the name of Saint Thomas of Canterbury, added to the litanies of the Cistercian order in 1214, is a later addition. The volume was thus copied between 1200 and 1214. Other rather delicate changes instituted by the chapter and immediately inserted as additions allow us to date certain liturgical manuscripts within a year or two. The feast of Robert de Molesme, for example, was instituted in 1222 but displaced in 1224; some manuscripts copied between these two dates bear witness to the change.

We should not be lulled into false hopes. Most medieval manuscripts are neither dated nor datable. Chemical analyses, as I said in my introduction, are both difficult to apply and marred by uncertainty,14 which renders them of little value. However, the tiny proportion of properly dated manuscripts furnishes, by its chronological, geographic, and typological diversity, dependable touchstones that, although they are not distributed harmoniously through time, do not leave any great chronological gaps, except for the eleventh century, which is constantly underrepresented.

Confronted with a manuscript bereft of any objective indicator, we are obliged to compare the formal aspects of its writing and decoration with other manuscripts that are correctly dated or localized in order to formulate a hypothetical dating that can usually attain the precision of a quarter century (or, for the more ancient and formal hands, only a half century). To correct and refine the temporal indicators provided by scripts, it is essential to consider the decoration, even in its most summary forms, and this merits a specific approach.

Institut de recherche et d'histoire des textes (CNRS), Paris

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7a OMITTED]

[FIGURE 7b OMITTED]

[FIGURE 8 OMITTED]

NOTES

The original version of this paper was first translated into English by Patricia Stirnemann, whom I thank most sincerely for her help and kindness.

(1.) Published catalogues of dated manuscripts are listed online at http://www.palaeographia.org/cipl/cmd.htm.

(2.) Amiens, Bibl. mun., MS 11, folio 96.

(3.) Manuscrits dates des bibliotheques de France, t. 2: Laon, Saint-Quentin, Soissons (forthcoming). Most examples discussed here are drawn from manuscripts that have been studied during the preparation of this volume.

(4.) Soissons, Bibl. mun. MS 226, folio 200v; cf. Manuscrits dates, 154 and pl. 125.

(5.) The "Viconte de l'eau" is a jurisdiction governing waterways, and here it covers the Seine and its tributaries.

(6.) Soissons, Bibl. mun. MS 56, folio 229; cf. Manuscrits dates, t. 2, 139.

(7.) Toulouse, Bibl. mun. MS 385, and Vatican, Vat. lat. MS 1021.

(8.) Chantilly, Musee Conde MS 754, and BNF lat. MS 6368, where the scribe calls himself Signore.

(9.) Laon, Bibl. mun. MS 404, folio 1; cf. Manuscrits dates, t. 2, 97, and pl. 28.

(10.) Laon, Bibl. mun. MS 31, folio 120; cf. Manuscrits dates, t. 2, 10, and pl. 47.

(11.) Manuscrits dates, t. 2, 151, and pl. 34.

(12.) Manuscrits dates, t. 2, 140, and pl. 106.

(13.) B. Backaert, "Levolution du calendrier cistercien," Collectanea ordinis Cistercensium reformatorum, 12 (1950): 81-93, 302-315; 13 (1951): 108-127. Information concerning the evolution of the sanctoral and the liturgy of the major religious orders is now available online at http://www.univ-Nancy2.fr/MOYENAGE/UREEF/MUSICOLOGIE/ outils.htm.

(14.) Using carbon dating, for example, may give an approximate date, but the terminus post quem and terminus ante quem would be separated from each other by more than a century.