Per Vikstrand, The Institute for Language and Folklore (SOFI), Uppsala

The aim of the project

The Swedish landscape is to a great extent characterised by place-names from the Iron Age. These names constitute an invaluable source material, the use of which, however, is made difficult by the lack of a reliable chronology. The chronology now used is based mainly on research carried out during the first half of the 20th century. It is principally based on tax rates during historic times, presupposing that larger villages with a higher tax rate are older than smaller villages with a lower tax rate.

The method is based on the assumption of a solid and even expansion of the settled area over time. Modern research has, however, revealed that this has not been the case. The great clay-flatlands of central Sweden upheld extensive dwellings already during the Early Iron Age. At the transition between Early and Late Iron Age, that is around 500 AD, many of these dwellings were abandoned and the settlement concentrated to higher grounds.

The aim of my project has been to relate the Iron Age place-names to the new picture of the Iron Age landscape. A more direct objective has been to re-establish a chronology for different types of place-names, thus promoting the use of place-names in landscape-related research. The area of investigation has been limited to Central Sweden, that is the provinces of Närke, Södermanland, Uppland and Västmanland.

The three main results

The nomenclature and concepts used for describing the name-landscape of the Iron Age must be revised. During the Early Iron Age settlement names seem to appear on two different levels, on one hand as names of individual settlements, "farms", on the other hand as names of small domains comprising several settlements. While the individual settlement names have a punctate reference, the domain names refer to an area. This means that the prerequisites for dating differ between these two kinds of names. For the individual settlement names, strict demands have to be applied as to the close spatial connection between settlement and name. As regards the domain names, on the other hand, all ancient monuments and settlements within the domain have to be taken into consideration.

Place-name chronology does not seem to be possible to describe in terms of a simple succession of different name types. Rather, the name types seem to overlap each other chronologically. The choice of name type made by the name giver is thus a qualitative choice. However, there are tendencies to chronological differences between name types. While the names in hem might transgress the border to the Bronze Age, certain names in by seem to be established as late as the 12th or even 13th century.

The principal types of settlement-names of Central Sweden all seem to be well established already during the Early Iron Age. This includes names in by, -hem, inge, -stad and names in the plural (Berga, Sanda, Lunda etc.). Especially the Roman Iron Age (0-400 AD) stands out as a formative period for the name-landscape, while the names of later periods, not least the Viking Age, are harder to get to grips with. In central areas of settlement, such as the large plains of central and southwest Uppland, rather few names seem to have been coined during these periods. Thus, one can no longer claim Late Iron Age as a "cautious" and normal dating for prehistoric names in these areas.

New questions which the project has raised

As the border between the Bronze and the Iron Age no longer stands out as an insurmountable barrier, the question of how far back in time the settlement names could have been established is more open today than earlier. We must face the crucial question if there could be continuity from names of Bronze Age territories to domain names of the Early Iron Age. This study supports the idea that certain names in hem could have their origin as names of Bronze Age territories, but further investigations into this question are needed.

Place-name types have been regarded as having a limited period of production and one, clearly defined meaning. However, this study points at the possibility of very long periods of production for most of the principal name types. This means that we must consider the possibility of meaning-change in place-name elements, as it is very common that the meanings of words have shifted during the course of time. Such a possibility is discussed for the names in by, which seem to have been productive for at least a millennium. It is conceivable that the word in the oldest names has a meaning 'prepared land (for agricultural purposes)', while it in later names has its more well-documented meaning of 'settlement'. Such shifts of meaning are possible or even probable also for other name-elements, perhaps especially for the names in stad, and requires further investigations.

A question of more interdisciplinary character is how the domains of the Early Iron Ages have been configured. That individual settlements have been incorporated in larger units seems rather unproblematic, but we have few models or theories about how these larger units or domains have been structured. This question must be discussed jointly between place-name scholars, archaeologists and others dealing with landscape-history.

Publications

I have chosen to publish the results of the project as a larger study with the title Järnålderns ortnamn. Bebyggelsenamnens uppkomst och ålder i Mellansverige (The place-names of the Iron Age. The origin and age of settlement names in Central Sweden). This study will be published on-line with open access at the website of the Institute for Language and Folklore (www.sofi.se) during the spring of 2012. A printed version will be available later the same year. The main results will be presented at the 15th Scandinavian congress of onomastics in Denmark in June 2012. Parts of the project have also been discussed at seminars at the department for Scandinavian Onomastics at Uppsala University.