The subtitle of Alice Rio's fine book is a better indication of its contents than is its misleading title. Even so, rather than a study of Frankish formulae themselves, this is mostly a study of the manuscripts which contain them and their potential value as historical data. Only one section (33-40) reproduces any actual formulae, four of which are printed side by side with documents related to them without any linguistic analysis or translation; fourteen others are quoted in English translation alone (179-80, 203-04, 207 and 223-33). Those interested in the formulae themselves will still need to go to the printed collections, in particular those edited by Karl Zeumer in 1886 (Formulae Merovingici et Karolini Aevi, MGH Leges V) and derivatively by Uddholm in 1962 (Marculfi formularum libri duo, Uppsala). And yet, as Alice Rio shows on many occasions with well-articulated arguments of great perspicacity, Zeumer's presentation of the formulae is as much confusing as enlightening.
What we call "formulae" are not called that in the manuscripts that contain them. They were often called "cartae", just as real documents were (43), and Marculf preferred the word "exemplarium". So it might be more accurate to see them as form-letters, or templates, which professional scribes could refer to when drawing up new documents concerning sales, donations, requests, manumissions, judgements, resolutions of disputes, letters of recommendation, agreements, and many similar topics. Patrick Geary referred to them as "scripts for future performance" (14), but in essence they are decontextualized versions of genuine earlier documentation rather than abstract schemata; brilliantly summarized as "descriptive material turned into normative form" (198; as also happened to the late Imperial Artes Grammaticae, similarly responding to an instinctive need to get things right). This explains why the eventual development of accessible cartularies later removed the need for formulas, as model documents could easily be then found in the cartulary collections and be used for the same purpose. Indeed, it is not always easy in practice to distinguish a formula from a genuine document. And many of them seem to have been included in manuscripts for the benefit of the scribe who put them there and nobody else. Thus Alice Rio looks at the texts from the point of view of the scribe who created them and wished to be able to refer to them later, and the results of an analysis from such a perspective are genuinely illuminating. When we find a manuscript containing many of them, they are not necessarily all together, or organized in such a way that they can be seen as a unit, despite the way that Zeumer collected them into large "formularies" (a concept which is thus a modern invention); they are often added separately into an existing blank space in manuscripts primarily devoted to something else. The idea of formularies is probably thus best left on one side, despite Marculf, and Alice Rio does so, looking at the manuscripts which contain formulas in enormous detail instead. There are 36 of these, 13 of them from the BN in Paris, each of them listed with their complete contents in an Appendix (241-71) and described with astonishing detail and thoroughness in Part II of the main text, "Inventory of the Evidence" (41-164), which is much the longest part. The analysis shows, for example, that formulae apparently produced in one area could be used in another, and that formulae originating two hundred years earlier could still be thought to be useful. The most valuable manuscript for the modern historian is Paris BN Lat.2718, safely localizable to Louis the Pious's chancery, containing what are known as the Formulae Imperiales, and which "can best be characterised as a private notebook" (133); and yet even this one has sufficiently haphazard organization to be useful in practice probably only to the manuscript's owner, who had put the formula texts into spaces there left available after pre-existing texts, rather than to the royal scriptorium in general.
Under these circumstances, it is not surprising to find an "astonishingly diverse manuscript tradition" (64), but even so the fluidity of the genre is remarkable. Scribes in monastic scriptoria "did not shy away from modifying and reordering their texts to suit their own purposes" (76). Alice Rio approves of this. Zeumer, on the other hand, regarded later accretions and variations in the text as errors, as if a formula were a literary text with a correct initial state vitiated by later scribal corruptions, "...picking and choosing what is authentic and what is not in any given manuscript" (113); which rightly appears presumptuous to Alice Rio, who knows that she is more than a match for Zeumer, "dismissing Zeumer's urtext (as we clearly ought to)" (169). The misconception that formulae are and ought to be fixed and unevolving, and thus necessarily also anachronistic and archaic, is banished without any qualification; they are shown to be quite right to evolve, in order to represent better in practice the needs of their users. The evolution may not always have been intentional, as we can see in the way that the word "gesta" seems to have come over time to be thought to mean "archive" (180-83). Indeed, it is hard to see why these formulae should ever have been preserved if they did not evolve with the times, but the probably correct idea that most early medieval scribes and intellectuals were ingenious and hard-working intellectuals is still a hard one to convey in a modern world where "medieval" tends to be thought to imply "closed-minded".
Much of Alice Rio's achievement here consists in demolishing the deductions hitherto made by Zeumer (and others) as regards the dating and localization of original formula versions, of what she consistently calls an "urtext". Marculf himself cannot be located, as he deliberately removed local references in order to decontextualize the model, and can only be roughly dated. This demolition is not a step backward, however, because she also convincingly shows that the original place and time of a formulary text is of no great significance in itself anyway, and if they are viewed as a living part of the intellectual's repertoire they can tell us a great deal about the time at which they were used in the production of new documentation, wherever and whenever they had originated. In this way, and in others, the formulae are reminiscent of glosses and glossaries; the origins of particular lists of glosses also often go way back into the mists, and if Carlotta Dionisotti is right they too originated in the late imperial linguistic repertoire available to notaries and others, while later glossary collections seem relatively as unwieldy and cumbersome as a large formulary.
There is a very good and rather surprising final section (198-237) on the relationship of formulae to written legal texts, whose "relationship to reality is at best uncertain" (202); she argues that "written law was understood more as a guide to be customised than as a rule to be enforced" (207), so cannot be taken as a guide to what actually happened. If she is right about this, it is no wonder that Germanic scholars find the situation easy to misinterpret; and she may be indulging in some wishful thinking here anyway, as she argues, particularly on the basis of the contrast between the evidence from legal texts and from formulas concerning the treatment of slaves, that in reality the operations of legal systems were more humane then they seem to be if we just consider the laws themselves. This thought needs further consideration, and with luck it will receive it from Alice Rio herself.
The subtitle of the book, which includes the word "Frankish", is accurate. In the phrasing of the main title ("The Early Middle Ages" in general) this book continues an unfortunate, and peculiarly Cambridge-based, tradition of assuming that the Frankish world is the only area worth investigating for these centuries. The Visigothic equivalents are mentioned in passing but unstudied, for example. Some of the bibliographical linguistic references are also outdated (e.g. to Keil's edition of Alcuin's De Orthographia, rather than to either of the two twentieth-century editions, 273); and overall, linguists and philologists, while recognizing the intelligence of much of the general introductory discussion of orality and literacy in Frankish society (11-26), will be disappointed that there is so little linguistic analysis of detailed data, rather than the general (and generally justifiable) conclusion that the formulas and the texts based on them were intended to be, and in the event usually were, intelligible to their users. Indeed, somewhat tactlessly, Alice Rio indicates that she sees little point in such detailed analyses: "not only is it impossible to know what spelling Marculf would originally have used, but it is unclear why we should care..." (99); but a clear conclusion that Marculf usually did, or did not, use correct traditional spellings would in fact be of some significance. The cases of the practical reordering of the formulae into texts, those which we are permitted to see in the four brief extracts reproduced at the start, are enough to show that linguistic data of considerable interest are attested there; the updating of some texts is also said to be for "stylistic modifications" (174), unspecified. So there remains a great deal more that can be done. (This relative concentration on manuscripts rather than on their texts is reminiscent of the current state of Priscian studies; a great deal was achieved, particularly but not only by the late Margaret Gibson, to trace the Frankish manuscript traditions of the works of Priscian, but no scholar has recently studied or evaluated in any detail what Priscian actually said.) But this book will be the basis for all future study, and the general "sensitivity to scribal intentions" (187) will be the example to follow. From being seen as dry uninviting schematic and skeletal data, the formulas will henceforth be "treasured for their eccentricities" (240). More generally, the step forward which Alice Rio's book represents is of enormous value to all those interested in the written data of the early Middle Ages in the Frankish kingdoms, and its continually manifested articulate intelligence leads us to hope that from the author the best is yet to come. That could be spectacular.