1868 and 1872. Thereafter came a division of the so-called Roman Iron Age (a.d. 0– 450) into two parts, formulated mainly by the Danish archaeologist sophus müller (1846– 1934).

Antiquarian/archaeological research changed from being a political and didactic instrument to an empirical, critical science, and archaeology was finally free from the literary historical analogies. It was no longer a matter of writing a country’s history; instead, it was important to organize and systematize the country’s nonhistorical source material, and the leaders in the change were Thomsen and Nilsson.

Typology

The chief characteristic of the last half of the nineteenth century was the division of the three-age system into subperiods and the formulation of culture-historic guidelines. Find-contexts played an important role. Within Iron Age research, historical analogies were used in the sense that historically datable, Continental material was comparable with Scandinavian conditions. Numismatics also had a high priority in Iron Age research. Around the year 1870, the nature of archaeological discourse changed again, with chronological work directed toward making shorter time divisions while the intensity of research and the amount of source material increased. The increase in material was seen to be connected with the intensification of agriculture, as well as with increased industrialization in general. By this time, researchers and museum scholars both had a specialized university education. It was in this context that a new archaeological discourse on typology was formulated.

The background of typology was empirical. Changes in form had been described by oscar montelius (1843–1921) and hans hildebrand (1842–1913) as “development” in accordance with the language of the time. The idea that development was continuous and unbroken was thus the basis of typology. This meant that development could also be observed and traced without the support of find-contexts. With the aid of typology, correct chronological conclusions could be drawn entirely on the basis of changes in form, and in this area, statistics were an important aid. Typology was, however, formulated in a diffuse way and the term consequently has a broad definition.

It was to mark these new methodological directions that Hildebrand formulated the typological method, but typology was of less importance for him than it was for Montelius (Gräslund 1987, 97). Montelius’s famous Tidsbestämning [Dating in the Bronze Age] (1885, English edition in 1986) was based, to a greater extent, on different find-contexts than on typology. Determining time sequences in the Bronze Age did not require the development of typological analysis. Thus, Montelius worked not only typologically but also, and especially, by empirically combining different find-contexts.

Müller criticized Montelius for neglecting the relationship between the context of artifacts and the typological series and implied that it was impossible to classify artifacts typologically without taking these contexts into account. In other words, Montelius did not differentiate between typology and find-context analysis.

Yet it was Montelius himself who popularized his image as a typologist. After Darwin’s works were published in Swedish, Hildebrand and Montelius began to hint at a connection between typology and Darwinism. However, unlike Hildebrand, Montelius never referred to the theory of natural selection, and as a result, the relationship to Darwinism is vague (Gräslund 1987, 104). Typology belonged more to a pre-Darwin theory of development, a way of thinking that had existed during the eighteenth century, a fact that Nilsson emphasized strongly. But typology also belonged to neo-Kantism and to the advance of positivism within historical research. Hildebrand and Montelius never claimed that Darwinism had given rise to typology (Gräslund 1987, 105). Rather, typology was associated with Darwinism only after the relationship became apparent, and it should therefore be viewed as a scientific-political move.

The most decisive “discovery” of the time was, however, Montelius’s dating of the Bronze Age and indirectly also providing dating for the Stone and Iron Ages. A more exact dating of the Stone Age had to wait, however, for the scientific methods that first became possible during