The Early Runic Inscriptions

← 11 | 12 → Introduction

In this book, I attempt to locate and prove the existence of West Germanic features in a set of runic inscriptions assumed to belong within a Northwest Germanic language. According to most scholarship, these WGmc features are not supposed to appear in the language until a later date than that of the inscriptions. In order to understand the early evolution of the Germanic languages, this introduction will present the history of Germanic dialectology from its beginnings until currently. Part I, composed of Chapters 1 and 2, will deal with the first corpus analysed: the earliest Danish runic inscriptions. The inscriptions in this corpus date anywhere from the second to the fifth century. Once the inscriptions are investigated and the WGmc features presented, in Part II (Chapters 3 and 4) I will compare the results to early younger inscriptions in two WGmc-proper corpuses: those of the Anglo-Saxon and South Germanic territories, dating from the fifth to seventh centuries.

The analysis of these three sets of inscriptions occupying different areas and centuries will give me an overview of the evolution of WGmc from the earliest appearance of dialectal features in Denmark to the establishment of different WGmc varieties (Ingvaeonic and non-Ingvaeonic) in the modern territories of England and Germany. The previous analysis will also present sociolinguistic and anthropologic points, which will be dealt with in Chapter 5. Here I will show how the appearances of certain features will give evidence for the purposes of early runic carving and the evolution of these purposes as well as inscription typologies. I will also present wider implications from the linguistic analysis regarding societal customs in the early Germanic areas such as gender roles in runic inscribing.

The conclusions drawn from the linguistic and extra-linguistic analysis of the three sets of inscriptions will strongly suggest that the language of the South Scandinavian/northern SGmc areas during the first four centuries of our era presented features divergent with the NWGmc paradigm, but which can only be expected in a living language in a non-homogeneous community of speakers in turmoil. That these divergences point to WGmc early traits will be justified by a potential closer linguistic relationship of Denmark with its sourthern neighbours than with northern Scandinavia. The Conclusion will summarise these ideas as well as all the significant linguistic and extra-linguistic points developed throughout the chapters.

A few words concerning methodology are in order. Regarding the sources, only the earliest inscriptions showing potential WGmc features will be presented and further scrutinised. This means that a high percentage ← 12 | 13 → of the Danish inscriptions in the Older Futhark will be part of the present work, and only the pre-eighth century set of the Anglo-Saxon and SGmc inscriptions. These inscriptions have been chosen because they represent the earliest stages of the WGmc branch.

Regarding the media used for the analyses of the carvings, the inscriptions dealt with in this monograph have been examined in picture or in person, but none have undergone microscopic analysis. For each inscription, I will present the transliteration and a reading or different reading possibilities. In each analysis, all likely possibilities of interpretation of the appearance of the unexpected feature will be taken into consideration, and each possibility will be assayed for potential problems. The process of elimination of potential interpretations or linguistic claims will be done by an examination of the archaeological content of the object, the mode of carving, the comparison to similar objects or linguistic material and/or the comparison to trends visible in other inscriptions. I will present the least problematic and most encompassing possibility as the most likely to be correct.

State of the question

It may seem obvious that when we set out to write a grammar of a language, the definition of that language is one of our first tasks. Further, that since we are dealing with language, the criteria used in our definition should be linguistic. These criteria should delimit the language geographically, that is, from other languages, and chronologically, that is, from earlier and later stages of the same language. Yet in spite of these minimum requirements for an adequate definition, our handbooks are not precise in their definitions or explicit in regard to criteria used for the definitions they give.¹

Lehmann’s complaints about the inaccuracy of the definitions of Proto-Germanic seem completely appropriate for summarising the state of the literature regarding Germanic dialectology. From the appearance of the first treaties on Germanic diachronic divisions in nineteenth-century Germany and Scandinavia to the most recent literature on the topic, there has been little consensus on the state of the Germanic languages after the “Common Germanic” period (i.e., after the migrations of the Goths in the second century AD), whose existence is a matter of disagreement, as with most periods of Germanic history.² The history of the literature covered in this introduction ← 13 | 14 → will be that concerning the evolution of the Germanic languages from the beginnings of the Christian era to the middle of the Migration Age. The information presented in this section will be fundamental for understanding the historico-linguistic background to which the later chapters will attempt to contribute. Given that there is no complete agreement on what the language of the oldest runic inscriptions is, this chapter will present the reader with the different solutions that have been given to the problem of nomenclature at this stage of the development of the Germanic languages. Also, I will introduce my linguistic position regarding the language of the earliest corpus and explain how I plan to contribute to modern Germanic dialectology and runic studies.

From fairy tales to modern linguistics: two centuries of Germanic dialectology

Kufner begins his chapter on “The Grouping and Separation of the Germanic Languages”³ by commenting on the difficulties encountered by historical linguists when trying to determine the various branches into which Proto-Germanic split. Although it is more common in current literature to claim that Proto-Germanic splits into three branches (North, West and East Germanic), Kufner indicates that the impossibility of agreeing on the subject is based on the scarcity of and chronological differences among the physical sources for the study of these emerging languages; the already existing disagreement about the degree of unity of Proto-Germanic (many scholars consider it a conglomerate of dialects); the level of mutual influences that the different Germanic languages may have had after their rupture from Germanic (a high degree of influence between the two languages may create conflicting groupings, some traits being understood as inherited and not borrowed) and, finally, the diversity amongst the methodologies and criteria followed by scholars.⁴

These considerations will help to evaluate the major hypotheses regarding the break of Common Germanic into the different Germanic languages. However obsolete, some of the following theories will give us a good overview of how the study of the subject has evolved since the mid-nineteenth century until now.

← 14 | 15 → Earlier studies tended to separate Germanic initially into West and East branches (henceforth WGmc and EGmc), amalgamating the later division of North (NGmc) and EGmc into a single East branch. Johannes Schmidt claimed that NGmc could not be included in the EGmc group without being included at the same time in the WGmc one,⁵ thus rejecting the contemporary partition of the Strammbaum theory and the idea of a unified WGmc group.⁶ The concept of a WGmc unity and independence from the two other branches was later defended by Karl Müllenhoff and August Schleicher, the latter of whom is the precursor of the tripartite division which is more in accordance with current arguments.

References to Germanic dialects from twentieth-century linguists, usually of German origin, date the split of WGmc and NGmc to as late as the eighth century. In his Germanische Sprachwissenschaft,⁷ Richard Loewe initially classified all Scandinavian Languages within a “Nordische” subgroup, whereas he gave English, Frisian, Dutch, Lower German and High German separate identities as dialects, since in his chapter on the division of Germanic he states that “Von den Dialekten dieser Völker stehen sich das Angelsächsische, Friesische und Deutsche gegenüber dem Gotischen sowohl wie dem Nordischen einander so nahe, daß sie als seine einzige westgermanische Gruppe erscheinen”⁸. Even though he does not reject Schleicher’s Stammbaum tripartition, he is influenced by Johannes Schmidt’s idea that “das nordische ist sowohl ostgermanisch als westgermanisch, es bildet den ubergang vom gotischen zum angelsachsischen, das angelsachsische und friesische den vom gotischen zum altsachsischen.”⁹

When analysing the literature of the early Germanic scholars, Wilhelm Grimm and some of his contemporaries play a fundamental role. Although his importance in the field of Germanic linguistics cannot be disputed, Wilhelm Grimm’s account on the evolution of the runic alphabet reverses the chronology applied by modern scholars to the different stages of runic writing. Thus, Grimm considers the 16-character futhark as the oldest one due to its being the simplest and the one containing the fewest runes,¹⁰ the 24-rune futhark being then a later evolution in which new runes have been added to represent the new developments in the language. Only some years later, Jakob Bredsdorff will refute Grimm’s theory by merely arguing that simplification is a much more common diachronic feature than addition. However erroneous his chronology was, Grimm did achieve a more accurate description of the language represented by the 24-character script. Through the study of the five rune stones known to him, Grimm reached the conclusion that they presented a “mixture of Scandinavian and ‘German’ runes”¹¹, though no further successful analysis of the five inscriptions was achieved.

It was not until Jakob Bredsdorff attempted to transliterate the Gallehus Horn inscription through the previous analysis of the newly-discovered Vadstena bracteate that we can date the starting point of the debate about the phonemic realities represented by the Older Futhark.¹² This attempt was not completely successful, since Bredsdorff made use of sound values of the Anglo-Saxon futhark and applied them first to the Vadstena bracteate and then to the Gallehus inscription, which resulted in some transliteration inconsistencies. Some of these inconsistencies (such as the rendering of m for R) led him to place the language of the inscription as a close-to-Gothic Norse early variant,¹³ even though he had used the script of a Western Germanic language to decipher it.

Bedsdorff’s reading of the inscription was challenged several years later by Peter Andreas Munch, who tried to make the transliteration more consistent, firstly by transcribing as /a/ (the use of the Anglo-Saxon futhork had resulted in a and o for ). Munch did agree on the Gothic character of the inscription, since he maintained as m finale, even though some years before he had already transcribed the symbol in the Tune stone as a final z, which attempted to represent a “sound which was intermediate between R and s”.¹⁴

Many others of the Scandinavian inscriptions were ascribed to the EGmc language group during the first half of the twentieth century, although ← 15 | 16 → part of this supposedly EGmc corpus found its way into some of the most important works of the second part of the century. It was not until the late nineties, with the work of Lena Peterson,¹⁵ that some of the inscriptions that were considered to be Gothic (the Mos lancehead, the Vimose buckle, etc.) were demonstrated not to be EGmc.

Although, as we have seen, the twentieth century saw many supporters of a Scandinavian component for the oldest runic inscriptions, there were also scholars who “looked with scepticism or outright rejection upon the alleged Scandinavian origin of the older runic idiom”.¹⁶ These specially criticised former readings of the Gallehus horn where had been transliterated as R, suggesting a Norse connection, and preferred to use z. Even though it is difficult to examine early runic inscriptions with the purpose of extracting linguistic evidence for dissimilarities between the formants of the Germanic family, some differentiating traits (however small the quantity) may be obtained by close analysis. An instance of a trait that has been supposed to be particular to one of the three branches is the usage of the rune at the end of a word. Some authors have claimed that its usage at the end of the word to represent - R is proof for the inscription to be Scandinavian, and not WGmc represented -z in Common Germanic, but it was devoiced in East Germanic. The process of rhotacism from-z to - R occurred mainly in NGmc, but there are grounds to assume that the same process happened in WGmc before the sound was finally dropped. Makaev states that the date for the dropping of this sound in WGmc is not clear, and refuses to create a distinction between the usage of in NGmc and WGmc.¹⁷ What he does not consider is the possibility of the inscriptions representing a NWGmc variety, the aforementioned relation between and R being a trait of one single language and not a similarity between two different varieties. Makaev also rebukes the existence of any single dialectal trace in the inscriptions containing (normally considered as Scandinavian) in contrast to (WGmc). He argues against this hypothesis by claiming that the inscriptions containing are too late and that earlier inscriptions containing ← 16 | 17 → this rune may have been carved in degradable materials for which we have no evidence.¹⁸

The linguistic interpretation of this 15th rune, , is one of the most heated debates in the history of runology. This problematic grapheme plays a central role in the debate over the provenance of the oldest runic inscriptions, and more particularly, of the linguistic origins of the Gallehus inscriptions.

The history of the transliteration of the algiz rune is one of political and linguistic antagonism between Germany and Scandinavia, the reason of which was the conflict over the duchies of Schleswig-Holstein, ruled by Denmark and claimed by Germany on the basis of previous Germanic population in Jutland. As late as the mid-nineteenth century, the 15^th rune was being translated with the value the grapheme had had in the younger futhark, m ( was used in the younger futhark for previous , whereas had evolved into and also changed phonological values). The result of this transliteration is the apparent use of WGmc dative plural forms as -*gastim and *holtingam in the Gallehus horn.¹⁹

The first transcription of as other than the younger futhark m was Peter Andreas Munch’s transliteration of the rune in the inscription on the Istaby stone as -R. He suggested that differed from only in that it occurred in final position.²⁰ However, eight years after the publication of his first transliteration,²¹ Munch wrote an article about the Tune stone in which he claimed that could not represent the -R Nordic sound either, since there is already a grapheme representing that phoneme.²² He concluded that “one might therefore be most inclined to assume that here it designates an intermediate sound between R and S which the Goths seemed to have designated with Z, the Slavic peoples who used the Latin letters with Rz.”²³ Some years after the publication of this article, Ludvig Wimmer, who generally accepted Munch’s theory, started transcribing the rune as capital R, as a ← 17 | 18 → “Nordic reflex of [z].”²⁴ The custom has been retained until today, but only to refer to final in younger Scandinavian inscriptions where total rhotacism has occurred. Antonsen sees this as an attack on German linguistic claims after the loss of Schleswig-Holstein to Prussia. He also complains that many contemporary scholars still participate in these nineteenth-century tensions by using Wimmer’s interpretation of as R for their transliterations. Antonsen claims that “[r]ecognition of the fact that they are not written in ‘Proto-Nordic’, however, will free future scholars from the necessity of finding ‘Nordic’ solutions to the problems they present, opening them up to a broader Germanic and Indo-European perspective, which is necessary for their proper understanding.”²⁵