Tuna revisited
The place-name Tuna, the diffusion of linguistic innovations and the emergence of central places in Northern Europe during the first millennium A.D.
Information finns ännu bara på engelska.
Per Vikstrand and Stefan Brink
The names in tuna have for a long time been associated with exceptional archaeological findings such as gold hoards and boat graves. During the last decades, the identification and conceptualisation of the of the Iron Age central place has created a wholly new context for these names. In our study we will use this new knowledge and reassess the meaning and age of tuna.
The names in tuna revisited after 50 years
It is today nearly fifty years since Karl Axel Holmbergs thesis on the names in tuna was published. These names – such as Tuna, Sollentuna and Torstuna – had by then already gained attention from both linguists and archaeologists. Although the etymology of tun as ‘fence, fenced area, enclosure’ was clear enough, the names had specifics that suggested something more and they also displayed a stunning correlation with aristocratic findings and sites from the Iron Age. There were thus lofty expectations on Holmberg’s thesis. But he was deeply influenced by the positivist paradigm of the mid-20th century, which in some veins of the research community of Scandinavian studies had developed into a denial of all societal complexity in ancient times. On the basis of an immaculate empirical presentation of the tuna-names, he reached the somewhat surprising conclusion that there was nothing special with them at all. They just designated ordinary paddocks or pens from bygone days.
Holmberg’s thesis can be regarded as an extreme product of the research trends in onomastics (name research) at his time. As for the future it was clearly a cul-de-sac. It is significant that the starting point for the modern discussion of the names in tuna is not Holmberg’s thesis but rather Thorsten Andersson’s public examination, printed as a review in the journal Namn och bygd (1968). Andersson’s main argument is that we on purely linguistic grounds can deduce a special meaning of tuna. His points in case are the two names Sollentuna and Vallentuna, which only can be understood as ‘the tuna of the people of Soland’ and ‘the tuna of the people of Valand’ ̶ Soland and Valand being two ancient names of communities in central Sweden. These two names, then, must have designated places that were important for the community in its entirety.
In the 1970s, Lars Hellberg incorporated the tuna-names in his theory of the organisation of a presumed Svea kingdom. Although out of touch with the historical discussion of his time (which by then had abandoned the idea of such a state-formation) his observation of clusters of significant names such as Tuna, Karlaby, Rinkaby, Husaby etc. were of immense importance. Hellbergs theory portended the development of the concept of central places within archaeology, such as Gudme, Sorte Muld, Stentinget, Uppåkra, Helgö, Borre and Gamla Uppsala. In the 1980s and 90s Stefan Brink merged these two theories together and demonstrated how the place names could supply the archaeological central places with a linguistic dimension, providing tantalizing glimpses into the life of such a locus.
The archaeological evidence for the tuna-names being associated with central places has kept piling up. Already in 1938, Manne Eriksson could establish that the famous boat grave field in Vendel in Uppland was situated on the grounds of a now lost village called Tuna. Tuna in Badelunda parish in Västmanland became famous in the 1950s when an extremely gold-rich woman's grave from the 3dh century A.D. was discovered. At this Tuna there are also boat graves and one of Sweden's largest grave mounds – Anundshögen – can probably be connected to the site. Inhumation in boats was an exclusive and deviant practise, reserved for a few powerful families. There seems to be a connection between this burial custom and Tuna-places, possible testified also by Tuna in Alsike and Tune in Østfold in Norway, with two ship burials at Visterflo a few kilometres away.
Gold objects from the Roman period or the Migration period (400-550 A.D) are common findings from tuna-places in central Sweden. From later years, extensive excavation at Ultuna outside Uppsala has disclosed an ritual area of an enigmatic and otherwise unknown character, perhaps a holy field (åker) of a sort frequently referred to in place-names. Of the utmost importance is the archaeological investigation of a whole tuna-settlement at Gilltuna in Sweden. In Denmark, a large settlement of some 20 farms has recently been excavated at Tune in Sjælland, dated to the midst of the Iron Age. This Tune is situated in one of the most prominent Danish example of a place-name environment indicating centrality.
The aim of the project
The flourishing discussion of central places in Scandinavia has provided a wholly new context in which to discuss the intriguing tuna-names. However, from a linguistic/onomastic point of view not much ground has been gained since the time of Holmberg’s thesis. No major investigation has been launched to reassess the names according to new theories and new empirical knowledge. The aim of this project is to be such an investigation and to study the names in tuna as part of the very special historic and linguistic context of the central places. A further aim is to investigate if the diffusion of names in tuna can be understood as part of a much more far-reaching wave of linguistic change, related to the emergence of the Proto-Scandinavian language. So far back in history only place-names can provide source-material for such a study.
A contextual approach
One of the reasons for our shortcoming in pinpointing a denomination of the names in tuna depends on a tendency to regard them in isolation. But modern onomastics underlines the importance of coherence and linguistic environments. A place-name element as tuna appears in several contexts:
The context of the name
Each compounded place-name constitutes its own micro-context. As the first element – the specific – normally specifies the second element – the generic, much can be learned by analysing the specifics and the relation between specifics and generics. The specifics of the tuna-names includes potentially significant categories such as names of Old Norse gods (Ulltuna, Torstuna, Fröstuna) and inhabitant designations (Simtuna, Runtuna, Sollentuna, Vallentuna). But many times they are ambiguous or obscure (e.g. Ärentuna, Svintuna, Voxtuna, Fituna, Gilltuna etc.) and they have never been systematically investigated. Such an investigation would provide an important foundation for the assessment of their onomastic meaning and origin.
The context of the name environments
The names in tuna often occur in conjunction with other, specific names, making up varying constellations of names and name elements. Frequent examples are Karlaby, Rinkaby, Husaby, Gillberga, vi, helig, sal, skepp, böte etc. These name-environments are immensely important for the understanding of the names. Linked to Iron Age central places, they seem to reflect a prestigious nomenclature connected with these sites. It seems reasonably that we are dealing with a kind of name-diffusion of both social and geographical character. By analysing the name environments at central places in various parts of northern Europe, it ought to be possible to map patterns of influence and the dispersal of lexical innovations in this network of power. This is important as the tuna-names – at least superficially – seem to defy the normal lines of influences and spread from east to south and west.
The context of language
Roughly in the same period as the central places emerges the language in Scandinavia undergoes a series of profound changes that separates it from other Germanic languages and give birth to Proto-Nordic. The emergence of Proto-Nordic can hardly be understood without a sociolinguistic perspective. But although the splitting up of the Germanic language has been discussed with references to macro-historical events, such as the migration of the Goths and the settling of the Anglo-Saxons in Britain, few attempts have been made to relate it to sociolinguistic factors in the societies involved in the process and to modern notions of language change.
It has been noticed that the second generation of central places, emerging from c. 550 A.D., shows similarities in their layout, with a festival hall supplemented by a smaller and often enclosed building for ritual purposes. There are also similarities as to which functions are connected to them and how they are organised. It seems to have existed something of a Scandinavian standard for how a central place should be constructed; with the terminology of Durkheim they seem to represent Pan-Scandinavian représentations collectives. This presupposes extensive connections between central places. We should probably regard them as a network through which ideas of societal organisation, rituals, warfare, agriculture etc. but also of poetry, songs and narratives of different kinds could be transferred. This transfer of ideas was – as argued for above – most certainly accompanied also by a transfer of linguistic change and innovations. In this respect, one important group might have been the young warriors. At least during the Viking Age it seems to have been usual for young men to join foreign chieftains or king as retainers, and as such they might have acted as agents for linguistic innovations.
The European context
The central places have primarily been discussed in a Scandinavian setting, but models have been sought in the late Roman empire as well as in the Frankish realm. The earliest central places, such as Gudme and Uppåkra, display strong Roman influences in their material culture and it has been suggested that they were inspired by Roman villas and – later on – Merovingian elite residences. Modern research underlines that Scandinavia was actively involved in European history already during Roman times. Thus, it seems a priori more probable that we are dealing with a European phenomenon than an exclusively Scandinavian one.
This holds probably true also for the linguistic setting. Common for the Germanic speaking societies of early Europe appears to have been the comitatus, the band of warriors that had pledged loyalty to their leader. The comitatus-phenomena seems to be embedded in a special linguistic setting, including an elaborated system of hierarchaly ordered titles. One of these was karl, originally meaning ‘freeborn man’. In central Sweden, the place-name Karlaby ‘the settlement of the karlar’ is common at central places. In England karl appears – rather suggestive – in combination with tun in the common name Charlton. In Frankish areas there are faint but evident traces of the word in place-names as Karlebach and Karlstadt. Karlburg am Main has been regarded as an important stronghold in the Merovingian conquest of East Frankia. But the administration of the Frankish realm was early Romanised and the vernacular terminology lost, thus the linguistic footprint of the central places has mostly vanished. In Scandinavia, however, we have the benefit of the periphery. Well-preserved place-name environments, absence of any linguistic super stratum and extremely well-preserved ancient monuments makes it possible to study structures that are lost in other parts of Europe, among them the central place. This then becomes an objective of European importance.
A Celtic connection?
In enhancing the European perspective, it might be justified to once again raise the question if tuna as a place-name element might be influenced by the Celtic names in dūnon ‘stronghold, fortified settlement’. This was suggested by several scholars in the beginning of the 20th century, paying special attention to the Swedish place-name Sigtuna, which could correspond to the Celtic place-name Segdūnon ‘strong fortification’. The idea has since been dismissed, partly because the dominating trend in Scandinavian studies during the 20th century emphasised internal factors of language change. The early Scandinavian central places has, however, been compared with the partly contemporary Celtic oppida-settlements. These settlements do often have names in -dūnon. The linguistically trained archaeologist Sune Lindqvist argued already in 1918 that tun could have been used as a “replacement” for -dunon when reproducing Celtic place-names, in the same way as Old Norse gardher was used as a replacement for Slavic gorod ‘town’ in names like Holmgardher (Novgorod). Considering that the Roman version of the -dūnon -names ended in -dunum, it is perhaps not too far fetched to imagine that the dative plural tunum of tun was used as a replacement by the Scandinavians. Keeping an open mind, I believe this is at least is a possibility worth investigating.