1 Introduction: Bones and the Beast

Reconstructing historical languages from texts is a bit like trying to recreate a dinosaur on the basis of fossilised bones. A surviving skeleton – not necessarily complete – may give the rough outline of the animal, but to go further a palaeontologist needs to fill in the gaps between what survives: add on the sinews and muscles and organs and skin. The results will be a combination of fairly straightforward inference and more speculative guesswork, informed by the analogy of how living creatures today are put together.

In the realm of medieval languages, the bones are the surviving written forms. In the languages I treat in this book, these writings are in various alphabetic scripts, which give some idea of what sounds were said in what sequence: elements which can be arranged to give the basic skeleton of the sound inventory. But these scripts typically give no direct indication of units such as syllables or features such as stress, the prosodic connective tissue of phonology. These features can sometimes be inferred and reconstructed more indirectly, often from the effects they have on vowel alternations and changes, or from the roles they play in the metrical systems of poetry.

This is the kind of reconstruction I will attempt in this book. Most of the chapters start with a synchronic approach, using phonological or metrical evidence to build up a picture of the prosodic system of a particular linguistic variety at a particular period: trying to get as good a picture as possible of Tyrannosaurus Rex or Albertosaurus, each on its own terms. This is only the first step, however, and the larger story I want to trace is diachronic, concerning the prosodic history of certain Germanic languages over time: to reconstruct the evolutionary history of the Tyrannosauridae family over time, as it were. This historical dimension will be more or less in focus depending on the chapter, but with an overall synthesis attempted in the conclusion.

In this book, I deal with the prosodic systems of several stages of English and Norse between roughly 500 and 1300, with an eye to what comes before and after this span. These two languages are historically related, both developing from Proto-Germanic, a language that has left no written documents, probably spoken sometime in the last half-millennium BC.¹ There are a number of other Germanic languages, some with extensive medieval records, but I largely limit myself to these two for a couple of reasons. I do turn to Gothic, as the only substantive East Germanic language, as needed, but since this language does not have a long history of reliable records, it is impossible to trace its later prosodic history. It also lacks any useful texts in verse, which means I can’t take my preferred approach of comparing phonological and metrical developments. For these reasons, I rely on Gothic mainly for what it can tell us about the Proto-Germanic point of departure for Germanic prosody in general.

This leaves two other major branches within Germanic: North and West. The former was spoken at first largely between the North and Baltic seas, eventually spreading across much of Scandinavia and the islands of the northern Atlantic. The earlier stages of North Germanic are known primarily through alphabetic inscriptions, especially on stone and metal objects, surviving examples of which date back as far as the 2nd century AD. Only about a millennium later, from the 12th century on, do substantial manuscript records start to appear. In manuscript sources, West Norse is the best attested variety, with Iceland in particular producing by far the greatest volume of surviving texts – including those recording the vast majority of attested alliterative verse.

The term ‘Norse’ is potentially vague or ambiguous. Some use it to refer to any variety of North Germanic before the modern period, while others limit it much more narrowly to western dialects from after the Viking Age. Typical Anglophone use tends to allow ‘Norse’ to take in the Viking Age as well as the later Middle Ages, and to cover all dialects of North Germanic (hence terms such as ‘East Norse’ and ‘West Norse’, the latter being tautological under more restrictive definitions). I am more interested in linguistic continuity than arbitrary periodisation, but in general, by ‘Norse’ I mean the language of the later Viking Age through that of the later medieval manuscripts – in practice, roughly 900–1300. The periods before this may be called ‘Early Runic’ (until the 6th century) and (though this is not a standard term) ‘Transitional Runic’. For the later stages, I do concentrate on West Norse evidence, as this is where most of the poetic evidence happens to come from. I sometimes use the terms ‘classical’ or literary Norse to refer to the language attested in West Norse manuscripts of, especially, the 13th century. Because of the nature of the surviving sources, I follow standard practice in taking the Icelandic variety as my default point of reference for classical Norse.

Among the West Germanic languages I deal almost exclusively with English, a language attested in this period mainly from the island of Britain. To cover all of the West Germanic languages in appropriate detail would be a vastly grander project, and English has several features that recommend it as a representative case study within this sub-family. For one thing, it has the longest history of attestation (though High German comes in a close second), with useful records reaching back to the 7th century. It also has by far the most substantial tradition of alliterative poetry of any Germanic language, with direct attestations of poems found from the early 8th through the 16th centuries. This allows for a relatively full treatment of both phonological and metrical developments over a lengthy span of time. Beyond these general considerations, there are several points where English happens to provide specific interesting evidence of prosodic behaviour.

With English as with Norse, I should add a brief note on labels. The term ‘medieval English’ is meant to emphasise the basic continuity across the period, but conventionally a strong division is made between ‘Old’ and ‘Middle’ English, with the dividing line being drawn anywhere between 1066 and 1200. I do use these traditional labels in a neutral chronological sense, since they are so deeply entrenched in the scholarship, but conceptually these terms should not be taken seriously in the slightest. The appearance of a sharp break between the periods is an illusion created by changing philological contexts and the appearance in writing for the first time of dialects whose earlier history is poorly attested. In many ways, changes within the ‘Old English’ or ‘Middle English’ periods are often far more significant, and I frequently distinguish ‘early’ and ‘late’ stages of both periods. These are not intended as sharp breaks, and their exact import depends on what aspect of the language is under discussion, and in what dialect, but roughly the following scheme will serve for this book: ‘early Old English’ is anything before 750 or so, and ‘late Old English’ most things after 850; ‘early Middle English’ is before around 1250, and ‘late Middle English’ after roughly 1350. The gaps between these phases are intentional, to highlight that I am trying not to speak of sudden breaks and transitions.

This book falls into three broad parts. After this short introduction, there are two further introductory chapters: one on the phonological frameworks that I use to understand prosody in these languages (chapter 2), and another on the metrics of alliterative verse in English and Norse (chapter 3). These are both rather technical fields whose frameworks and terminology may not be familiar to non-specialists. Since not every reader is likely to be a specialist in both, and since I would like this book to also be useful to scholars of English and Norse who may not be familiar with either, I have tried make these introductions slightly fuller than they might have been.

After these introductory chapters, I deal with medieval English first. One pair of chapters investigates early Old English: chapter 4 dealing with the evidence of phonological change and chapter 5 turning to the testimony of alliterative verse as found in Beowulf. This is followed by a similar pair of chapters, 6 and 7, on early Middle English, again dealing respectively first with phonological evidence and then with metre. Finally, chapter 8 carries the discussion in fairly broad terms slightly beyond my main chronological focus and into later Middle English.

The final portion of the book turns to Norse. After a short preliminary on syllable structure (chapter 9), I cover the evidence of phonological changes from Early Runic through to classical Norse (chapter 10). I then follow this with two chapters on metre, one focusing on the general prosodic evidence provided by the fornyrðislag metre (chapter 11), and the other concentrating on one specific set of parallel metrical restrictions found in both skaldic dróttkvætt and fornyrðislag (12).

Throughout the book, I follow the trail of the prosodic unit known as the bimoraic trochee, which is in some ways the main character in the diachronic story that emerges. If you are not already acquainted with the bimoraic trochee, it will be introduced shortly, in the next chapter, and elaborated on extensively for the remainder of the book. By way of conclusion in the final chapter, 13, I outline the general fortunes and fates of this prosodic form from Proto-Germanic down through the later medieval period.

As indicated above, my intent is that this book will be in conversation with several different readerships: linguists studying the diachrony or synchrony of prosodic systems in general, metricists (whether comparative or Germanic), and those whose interests lie in the poetics of medieval English and Norse literatures. That there is something to be gained on all sides by considering all such apparently disparate approaches together is well demonstrated by the excellent recent study of Viking Age poetry by Heslop (2022), or the comparative approach to Norse and earlier English verse-craft and aesthetics taken by Frank (2022). In the words of Roman Jakobson (1985: 375), ‘I believe in the mutual salutory significance of linguistics and philology’ (italics original), and that the ‘interplay of linguistic theory and philological art’, perhaps above all in the realm of poetic metre, can still be a source of inspiration and inquiry.