A Novel Geometric Morphometric (GMM) Application to the Study of Bronze Age Tutuli

DANISH JOURNAL OF ARCHAEOLOGY 2019, VOL 8, 1-24 1 A Novel Geometric Morphometric (GMM) Application to the Study of Bronze Age Tutuli Christina Vestergaard1,3 and Christian Steven Hoggard1,2 1 Department of Archaeology and Heritage Studies, School of Culture and Society, Aarhus University, Moesgaard Allé 20, 8750 Højb- jerg, Denmark. 2 Department of Archaeology, Faculty of Humanities, University of Southampton, United Kingdom. 3 Corresponding author ([email protected]) ABSTRACT ARTICLE HISTORY In this paper we examine the morphological diversity of the tutuli object group from the Ear- Received 18 February ly Nordic Bronze Age onwards, an often over-looked artefact. With a significant presence 2019; throughout this period, and their widespread geographic and temporal distribution throughout Accepted 24 June 2019 the Nordic Bronze Age, tutuli are of great interpretive potential. Currently, only a few studies, KEYWORDS focusing on the morphological diversity of tutuli have been published, consisting of accepted Artefact analysis; Small decades-old typologies. The objective of this paper is first and foremost methodological, exa- finds; Quantitative mining two research questions grounded on the classification and periodisation of tutuli. Spe- analysis; Scandinavia; cifically, through an analytical and exploratory framework this article examines whether the Ornaments breadth of archaeological tutuli shapes conform to the classificatory system of Montelius’ typology, and whether a temporal relationship exists between specific tutuli types and shapes. Introduction “the label tutuli (singular tutulus) was designated early on to distinguish some small circular plates, which From the beginning of the earlier Nordic Bronze Age have an eye or crossbar on the underside and a more (NBA henceforth), ranging from c. 1700-1100 BCE or less protruding tip on the upper side” (Figure 1). (following Nørgaard 2018), an expansion in the va- Another exception is Nørgaard (2018: 234-236), riety of material culture, owing to the introduction who thoroughly analysed and discussed the pro- of the ‘new’ raw material, is witnessed. Through the cesses behind the crafting of tutuli, in addition to a medium of bronze, and its distinct new structural broad collection of other Early NBA object types. In properties, an assortment of different object types their form, tutuli vary considerably, from small and were then made possible. With such an abundance flat-plated morphologies to cone-shaped and even of new material culture, many (often smaller) objects hemispherical shapes (Figure 2). And given their are frequently over-looked in archaeological stu- abundance throughout the NBA, spatially throug- dies. One such often-overlooked example to appear, hout Denmark, Sweden and Northern Germany, which is the focus of this article, is the tutulus. and chronologically throughout the entire Early It is now accepted within archaeological litera- NBA, as well as their sheer quantity, with over a ture on the NBA that tutuli are, to some degree, thousand examples recorded in Denmark and Nort- overlooked (Nørgaard 2018), and only superficially hern Germany alone (Aner et al. 1973, 1976, 1977, considered when present within large hoards (e.g. 1978, 1981, 1986, 1991, 1995, 2001, 2005, 2008, Aner 1962; Antiq. Tids. 1849-51; Frost 2008) or in 2011, 2014; Aner and Kersten 1979; Aner, Kersten extraordinary burials (e.g. Bergerbrant 1999; Boye and Neuman, 1984; Aner, Kersten and Koch 1990; 1896; Clausen 1990). An exception to this rule is Aner et al. 1993), there is considerable interpretive Kristiansen (2013), who provides a general intro- potential in their analysis. (Figure 1 and 2) duction into jewellery from Scandinavia. Kristian- Originally, tutuli were thought to have been de- sen (2013:758) introduces tutuli by explaining, that signed for practical purposes with initial interpreta- 2 Christina Vestergaard and Christian Steven Hoggard Figure 1. A cone-shaped tutulus (top left) and a flat-plated tutulus (top right). The lug on the underside of the cone-shaped tutulus is a crossbar (bottom left), while the flat-plated tutulus has an eye (bottom right). Not to scale. Photographs by Christina Vestergaard. tions highlighting their function as shield-buckles 1999, 152; Broholm 1944, 107). In one particular (Rafn 1856). This viewpoint, however, stemmed example, Broholm (1944, 119) argues that when from an incorrect interpretation of remains from an tutuli, and specifically flat-plated tutuli (Figure 2: Early NBA burial at Buddinge, Sealand (Kristiansen Type A), were used by men, these functioned as cape 2013; Rafn 1856). Excavations at Buddinge revealed buttons, while for women they fulfilled a solely aes- a Bronze Age individual, with wooden fragments thetic role. Nevertheless, with a lack of debate in the positioned on the torso, which were in turn inter- last few decades, and the absence of rigorous empi- preted as the remains of a shield; it was hypothe- rical frameworks, the functional and stylistic debate sised, that the tutuli bound the edges of the ‘shield’ on tutuli remains open. together, thus providing a functional interpretation Here, we wish to focus on the strength of tutuli (Rafn 1856, 362). However, later investigations es- classificatory schemes, that is to say how tutuli are ca- tablished that the wood belonged to a wooden cof- talogued by archaeologists, and the degree of success fin rather than a shield (Kristiansen 2013, 758). Ex- in these morphological-based classifications. From cavations between 1878-1883 at Hesselagergaards the later part of the 19th century onwards, typolo- Mark on the Danish island of Bornholm also ques- gical approaches were intgral to how archaeologists tioned this notion, when four tutuli were recovered understood the Bronze Age and later prehistory in in situ and attached to remains of textiles. Sehested general. Hildebrandt (1866) was the first to apply a (1884, 51) argued that the tutuli could not be in- typological method on archaeological material, al- terpreted as parts of a shield as no wooden remains beit with limited engagement with the archaeologi- were recovered. Accordingly, a decorative and a cal material (c.f. Gräslund 1987). While establishing more style-centric interpretation was provided (Se- a typology, Almgreen (1967) notes that Hildebrand hested 1884, 51). Yet, several other function-based (1866) did not consider typological connections as interpretations followed, including the use of tutuli a proxy for chronology. Building on from this, Mon- as clothing buttons and/or as beltware (Bergerbrant telius (1872) published his first typological frame- DANISH JOURNAL OF ARCHAEOLOGY 2019, VOL 8, 1-24 3 Figure 2. Montelius’ four types of tutuli. Type A: a flat-plated tutulus (called belt discs by Nørgaard (2018, 234)); Type C: a cone-shaped tutulus; Type D: a hemispherical tutulus; Type E: a protruding tutulus with tip. Montelius’ Types B and F (early and late belt plates respectively) are not accounted for in this article. The complete typology also accounts for belt buckles from NBA IV and NBA V (Montelius 1885, pl. 2 and 3). work, encompassing a variety of material culture (1971) focused on the Lünenburger Heide (Lower including fibulae, hanging vessels, shaft-hole axes, Saxony) region of Germany, while Kersten (1936) knifes and swords. Following this, Montelius (1875, focused broader. To further complicate the scena- 253) noted explicitly that the more important fea- rio, the starting points for the different typologies tures of antiquities could be used to distinguish bet- are quite different as Montelius (1885) considered ween different periods of the Bronze Age (Gräslund the morphology of the tutuli, while Kersten (1936) 1987), emphasising that Bronze Age burial customs, constructed his groupings (and sub-groupings) and changes in these customs, could underpin a through both the morphology and the decoration chronological system within the Bronze Age. Resul- of the objects. ting from these object typologies, Montelius (1885) Given the lasting impact of Montelius (1885) on established six periods (from NBA I to NBA VI). the periodisation of the Bronze Age (Bergerbrant See Gräslund (1987) for further information on 2007; Gräslund 1987; Hornstrup et al. 2012; Knei- these subdivisions. sel 2013; Kristiansen 2013; Vandkilde 1996; Vand- A number of other typologies have also been kilde et al 1996), the framework’s temporal scope, constructed, expanding from Montelius’ mid-19th and the explicit focus on tutuli, this classificatory century classification (Baudou 1960; Kersten 1936; system is perhaps the best starting point, and most Laux 1971), grounded on artefact variability and applicable method, for analysing the shape of larger their presence and/or absence; these are all of diffe- tutulus datasets over multiple regions and periods. ring detail, and geographical and chronological fo- While so, the classificatory success is unknown; and cus. While Baudou (1960) only considered the Late it is unknown how idealised shapes account for the NBA, Montelius (1885) paid particular attention to nature of variation as witnessed in the archaeologi- the Early NBA. Meanwhile, Laux (1971) and Kers- cal record. Furthermore, and building on from this, ten (1936) consider both the Early and Late NBA with an abundance of examples ranging throughout periods, but of differing geographic scope: Laux the NBA, it is unknown how tutuli shapes conform 4 Christina Vestergaard and Christian Steven Hoggard to the different periods

Load more