Playing with numbers: Latin v. the Vernacular, and the transformation of the year 1200

Some of the recurring questions about this project, posed by historians as well as by colleagues in literary studies, relate to numbers. How many chronicles were there? Are we dealing with a few isolated snippets of narrative to which far too much attention has been paid already? How do they fare compared with charters, for instance? Similarly, what was the relationship between Latin and the vernacular? Is the emphasis on Latin texts not excluding a considerable corpus of materials that should legitimately be included in any discussion of historical culture? Well, to answer these questions, I did some number crunching. With surprising results.

Thousands upon thousands: the dominance of Latin

Latin chronicles, annals etc. from the high Middle Ages survive in truly stupendous numbers. Auguste Molinier’s Le Sources de l’histoire de France (1902), the most recent and complete survey of its kind for France, lists roughly 1,250 for the period 987-1194. The figure has to be taken with a pinch of salt: it includes, for instance, accounts of the crusades or from Norman Sicily (France, it seems, was wherever anyone Francophone set foot: though even Molinier did draw the line at including England). It also contains later texts discussing events or people active during this period. Still, Geschichtsquellen.de, a repertory of sources for the western (later Holy Roman) Empire also lists 1,200 vitae, gesta, chronicles and annals for this period. Given the nature of the project, Geschichtsquellen includes texts written in Italy and the Czech Republic, in addition to those from the medieval German regnum (very roughly, Germany, Austria, Switzerland, Belgium, Luxemburg, the Netherlands, and the eastern parts of France). A rough guess would be that we are looking at about 800-900 texts from France, and about 1000-1100 from the regnum.

How do these figures compare? Richard Sharpe’s Handlist of the Latin Writers of Great Britain and Ireland before 1540 (2002) lists 5,200 authors – not texts! – for the entire Middle Ages. He obviously did not limit himself to historical narratives. A reliable estimate would be that 200-400 texts (rather than authors – an important distinction to which I will return in another blog) would fall into this category. Not every region was as prolific. Medieval Nordic Literature in Latin, ed. Borgehammar, Friis-Jensen, Mortensen, Ommundsen (2012), for example, only includes about 200 items for the entire Scandinavian Middle Ages, again not limited to historical narratives. The Medieval Welsh Chronicles project lists 10, but does not cover saints lives, for instance, or annals. The situation for Hungary and Poland is not much better, though the crusaders states were rather well served, not least because of the sheer number of crusading chronicles that were written across Europe. The section of western sources in the Recueil des historiens des croisades runs to five densely printed folio volumes. I have not been able to find a comparable repertory for Italy or Iberia.

Overall, then, we may assume that roughly 3,000 – 3,500 historical narratives in the broadest sense survive from the eleventh and twelfth centuries across western Europe. That is a lot, especially when set alongside other types of evidence. For example, only about 3,500-4,000 charters issued by western emperors survive for the period before 1200. The forthcoming edition of the Angevin royal acta (grants and letters issued by the kings of England 1154-1216) is said to contain about 5,000 items. Yet again, vast variations existed. Only about 300 charters issued by the Norman kings of Sicily (1130-98) remain, and roughly 800 for the crusader kings of Jerusalem (1099-1291). A lot more can be found in the archives of bishoprics and monasteries, but charters by lay princes are equally rare. For Henry the Lion (1129-95), for instance, duke of Saxony and Bavaria and thus ruler of territories several times the size of England, only 300 charters remain extant. Latin accounts of the past were an exceptionally popular genre of writing.

That impression is reinforced when we compare Latin to vernacular materials. Geschichtsquellen’s 1,200 Latin items stand alongside 11 in German, 7 in Hebrew, and 1 each in French and Frisian. And the site is rather catholic in its definition of what constituted a ‘historical source’: it counts didactic texts, legal treatises, letters, bible commentaries, florilegia, liturgical materials, and so on. It lists 55 different categories of text. It does, however, largely exclude texts that would traditionally have been labelled ‘literature’ (the problems with this distinction are best discussed on another occasion). Yet even if those are taken into account, the overall picture does not change very much. By far the largest corpus of extant vernacular ‘literary’ manuscripts from the twelfth century was composed in Middle High German, with roughly 200 items. A mere 66 survive in Northern French (mostly in Anglo-Norman, and hence likely to be of English provenance).[1] With very few exceptions (pre-Conquest England being one), Latin remained the dominant language of literary production, and narratives of the past a popular outlet.

That is not what I had expected to find. It certainly calls into question the emphasis that scholars in the Anglosphere place on French vernacular writings. Their numbers pale compared to the German materials, and both are mere drops in the vast ocean of Latin. Yet, once we parse these numbers apart a bit further, we can see that a momentous shift occurred in the thirteenth century.

The transformation of the year 1200

The number of 14 Middle High German items listed on Geschichtsquellen is misleading. Two are contained in the same manuscript (which moreover consists mostly of Latin materials). More intriguing still is the extent to which most of the items listed were composed in the thirteenth century, but included because they revised, continued or translated earlier texts. They include poems by Walther von der Wogelweide (c. 1198-1228); Keiser und Keiserin, an ode to Emperor Henry II and St Kunigunde, written at some point between 1200 and 1240; a legendary account of the order of St John, produced at some point in the thirteenth century; a Life of St Ulrich, also produced in the thirteenth century; a genealogy of the dukes of Austria, translated from the Latin in the thirteenth century; and a fifteenth-century continuation of a set of eleventh-and twelfth-century notes on the history of the monastery of Adelsberg. Which leaves a mere 7 items in the vernacular written between 950 and 1200.

This makes all the more noticeable the change that occurred after 1200. For the period 1201 to 1500, 666 texts written in Middle High German are listed on Geschichtsquellen. That’s almost a hundredfold increase. There also were more Latin texts (2,150), though, once other vernaculars are taken into account, their percentage of the total output decreased from 98% to c. 70%. This should certainly warn against overestimating the popularity of vernacular texts (they still constituted but a fraction of the total historiographical output). But it nonetheless shows that Latin's overwhelming predominance had weakened.

Equally striking is the relatively low rate of the overall increase: from 1,220 to 2,969. That’s rather little compared to what has been estimated for overall manuscript production in Europe during this period, which multiplied rather more dramatically. Now, there are reasons why the number of ‘historiographical’ texts seems to slow in relative terms. For instance, there still are hundreds of texts across Europe that await being edited. The situation is especially bad in England (again, one for a future post), but other regions of Europe are different in degree, not substance. The Geschichtsquellen figures thus need to be taken with a pinch of salt, and a lot more number crunching needs to be done on repertories of other genres (sermons, miracle and exempla collections, and so on). In fact, another reason for this relative slowdown may well have been that writings about history found other outlets, notably exempla collections (raw materials for preaching) and sermons.

With all these provisos in mind, it proves nonetheless fruitful to compare genres of historical writing. Some clearly became less popular. Texts classed as ‘hagiography’ decreased from 258 to 171 – a decline even more precipitous once viewed as share of overall production (from c. 20% to just over 5%). Others maintained their relative status. There are, for instance, 51 world chronicles for 950-1200, and 117 for the later period – roughly in line with the wider pattern of increasing production. Others flourished: dynastic histories (27 v. 91); urban chronicles ( 9 v. 141); texts on secular law (8 v. 135); contemplative writings (3 v. 13); and Legendaria, that is, often vast compilations of saints’ lives and miracle lists (2 v. 13).

These shifts point to broader changes. The increase in dynastic and urban histories reflects the emergence of new political actors, but also their increasing access to the resources and modes of writing history. Geschichtsquellen does not have a general search function for the likely background of authors (e.g. whether they were monks, secular clerics, knights, notaries, etc.; though, where available, that information is contained in the descriptions of individual texts – a task for another occasion). But the broader pattern certainly suggests that historiographical production shifted to newly emerging groups of literate notaries, merchants, lawyers, urban and dynastic officials and so on. Similarly, while, for the earlier period, vernacular did not denote lay authorship, the relative distribution of languages suggests a change. Of 141 later medieval urban chronicles, less than a third were written in Latin (43 out of 141). A similar distribution emerges from writings on secular law (32 out of 135). Those may turn out to be coincidence, but urban chronicles and writings on secular law are also exactly where one would expect to find greater lay engagement. We’ll see.

Now, one of the questions this project asks (and, no, I do not yet have an answer, though preliminary thoughts have been published here) is to what extent the widening use – and the widening range of users – of writing brought about different notions of what history should look like, how it ought to be written, what constituted a reputable topic, and who would count as a reliable source. There certainly are recognisable trends in the thirteenth century. We do, for instance, see merchants, notaries, officials etc. appear as informants, and even as patrons of writing. Many of the themes popular in and constitutive of regnal history in particular surface in urban histories – such as the sudden popularity of a Trojan connection). One of the topics to explore next will be how English royal officials used the past. Intriguing snippets survive of Henry III’s courtiers writing history, and of using historical precedent. In the 1250s, for instance, the king explained that he was keen to ensure that one of his trusted officials became bishop of Ely, as the island had frequently used by Danes when attacking England. Those had largely ceased after c. 1080. Comparable moments information include the use of a reputable corpus of authorities when writing ‘literature’. The Nibelungenlied (c. 1200), for example, used Procopius and other canonical texts to construct its almost entirely fictional account of Siegfried and of Kriemhild’s revenge. That was enough for the Hungarian chronicler Simon de Keza (c. 1278) to use the Nibelungenlied as a source for the history of Attila’s campaigns in the fifth century. And so on. It is this mixture of innovation and continuity that makes the thirteenth-century materials so fascinating. Crunching numbers does help make some of these shifts stand out more clearly, but also throws up a whole new set of questions.

[1] Rodney Thomson, ‘The place of Germany in the twelfth-century Renaissance’, in: Manuscripts and Monastic Culture: reform and renewal in twelfth-century Germany, ed. Alison I. Beach (Turnhout, 2007), 19-42, at 27-8.