2. Background to Quartz Research

The history of Irish flaked stone research is primarily a history of Irish flint research. Even though a range of Irish Stone Age lithic raw materials, including quartz, has been recognised for a long time (e.g. Knowles 1889), flint has played a predominant role in the minds of antiquarians and archaeologists (see also Driscoll this volume). The north-east of Ireland is the almost exclusive home of in situ deposits of flint – often, but incorrectly, classed as Antrim flint – and a strong tradition of flint collecting has led to the accumulation of tens of thousands of lithics from that region, with intensive collecting and trading of lithics during the 19th and early 20th centuries (see Woodman et al. 2006). For a long time the distribution of Stone Age, but especially Mesolithic, activity in Ireland was seen as a primarily north-eastern phenomenon and the Antrim flint deposits were seen as the lynchpin of Stone Age settlement patterns (e.g. Macalister 1949). In fact, in areas where other raw materials were noted, the supposed lack of a trade in flint was seen as evidence of the cultural backwardness of the peoples of those areas (see Brunicardi 1914), or a lack of a local source of flint was given as a reason for the surmised early adoption of metal (see Macalister 1949).

More recently, the recognition that the island of Ireland was extensively settled from the Early Mesolithic onwards (Woodman 2003) has run alongside an increased awareness of the diverse suite of materials utilised for stone tool production (see discussion in Woodman et al. 1999). Nevertheless, materials such as chert or quartz are often still regarded as a substitute for flint (e.g. Herity 1987; Woodman and Scannell 1993) as opposed to valid raw materials in their own right, and analytical categories and typologies are primarily based on long-standing categories often derived from flint. In Knutsson's (1998) terms, and in parallel to some areas of Scandinavia, it would appear that in Ireland an 'unconscious projection of the framework of categories for flint' has had considerable impact on our understandings of prehistory. While the use of quartz as a raw material for stone tools in Ireland has been especially under-acknowledged, it has primarily been seen in terms of its ritual or symbolic attributes, as it is a common find in megalithic tombs and other monuments, and also as part of megalithic architecture (Bergh 1995; O'Brien 1999; for parallel distinctions between approaches to quartz see Warren and Neighbour 2004).

Worldwide, diverse approaches have been utilised to try and provide some analytical purchase on the material (e.g. Dickson 1977; Barber 1981b; Knutsson 1988; Bisson 1990; Saville and Ballin 2000; Cornelissen 2003). A review of such research has highlighted that the difficulties of quartz analysis are not easily resolved, much of which concern what Callahan (1987) described as the gravel effect – many quartz assemblages can at first glance appear to be comprised of amorphous pieces, not easily recognised as humanly modified or forming 'tools'. Outside of quartz research, the processualist movement in archaeology called for a greater degree of quantification in lithic studies in order to present the research on a more rigorous scientific base (see Lyman and O'Brien 2004). This call led to many researchers moving away from analyses based on formal typological characteristics of lithics, and to examinations of the tools and debitage products' technological characteristics; these were to be analysed as part of a technological package in order to develop models of prehistoric societal behaviour, and this emphasis on technology as opposed to typology occurred along with an increase in knapping experimentation and the study of fracture mechanics. Even though quartz was the predominant raw material in various parts of the world, such studies had not been carried out on quartz assemblages because quartz was perceived to be an intractable material to analyse; the apparent irregularity in the fracture pattern of quartz made such studies as attribute analysis and reduction sequences inefficient, or futile, in terms of the results. Therefore, quartz assemblages, and quartz regions, lagged behind in such studies.

One of the chief difficulties in analysing quartz, related to the 'gravel effect', is caused by the expectations that researchers have of what lithic assemblages should be like – as Knutsson (1998) outlined, it is the conventions of archaeological training that invariably shape ideas of what is expected of the archaeological record. Knutsson (1998) noted that, as elsewhere, Scandinavian students learn lithic classifications based on a flintcentric research tradition and, consequently, the Northern Scandinavian (primarily non-flint) lithic industries were perceived as 'rough' in comparison to those of Southern Scandinavia and the Continent – '[o]n a subconscious level, this mode of thinking has also been projected onto society as such, which at times has even been apprehended as retarded'. Knutsson outlined how previous Scandinavian research into quartz assemblages, which were based on formal types derived from flint assemblages, led researchers to equate quartz flake fragments with flint 'tools' that had a similar form, therefore bracketing these assemblages into incorrect cultural traditions. Conversely, Gramly (1981) noted that when quartz assemblages appeared not to include certain types of implements, these assemblages would be incorrectly excluded from the geographical distribution of a culture.

These issues go to the traditional heart, and life-blood, of archaeological endeavours – that of typology. The birth of modern archaeological research occurred at a similar time to that of the assertion of evolution as the predominant scientific and social paradigm among the classes that formed the bulk of archaeological researchers: as Lucas (2001, 80) has put it, typology was the archaeological equivalent of evolution. With the subsequent rise of the Culture paradigm, tool typology thence could define cultures and chronologies; the rise and fall of tool types were seen as witnesses to the ebb and flow of cultures. The utility and validity of tool typology was, of course, debated over the years, in terms of what the types actually represented and meant to both the original users and subsequently the archaeologists (see Adams and Adams 1991). A significant debate revolved around the Bordes–Binford debate (Bordes and de Sonneville-Bordes 1970; Binford 1973), where tool types were seen by Bordes as signifying differing (cultural) stylistic preferences or by Binford as differing activities or functions. As Tomášková (2005, 82) has noted, this debate did not question the actual types themselves or how they were defined, but rather what these defined types represented.

Stone tool typologies were for the most part built on morphological characteristics of artefacts, with differing retouch an especial characteristic in defining tool types. In defining and naming an artefact type, form and presumed function, were, and are, often used – for example, a 'disc scraper' (Woodman et al. 2006, 159) is defined by its retouch, its convex shape, along with its perceived function of 'scraping'. In other cases functional and technological criteria are used, such as the burin (for discussion on burins see Tomášková 2005, 83-4). Other categories used by researchers are those of 'formal tool' and the 'expedient tool', as well as the category of 'utilised flake' (for discussion on the difficulties with this latter category see Young and Bamforth 1990) – these categories are not always used in the same manner by researchers, and can mean significantly different things; their meaning is not always explicitly stated and must be inferred (hopefully correctly) by the context of use. Quartz lithics in particular often do not lend themselves to formal typological studies because retouch can be difficult to recognise, and often is not even there in the first place.

In the lithic studies literature there is a general division between typological studies and technological studies (e.g. Minzoni-Deroche 1985; Callahan 1987; Lindgren 1998; Inizan et al. 1999; Andrefsky 2001; Tomášková 2005; Ballin 2008); typological studies are generally geared towards results – the finished artefact; these 'finished' artefacts also include core 'types' in typologies. Technological studies are not restricted to typing 'tools', but concern an entire assemblage, including the 'waste' or 'debitage'/'debris', to understand the mode of manufacture. Of course, the technological studies also use the same methods of typing in their analyses, hence technological studies are sometimes described as debitage typological analysis (e.g. Andrefsky 2001), or typotechnological analysis (e.g. Cornelissen 2003; Ballin 2008). Researchers who focus on the technological aspects of lithics are generally critical of 'pure' typological studies, because lithic assemblages were dynamic entities and what we analyse are the end points of this dynamism – using morphological characteristics to define types, and hence cultural types, does not take into account that a certain type may be the result of resharpening or reuse, and may originally have been morphologically similar to a different type; consequently, morphology can only show the last phase of tool type, and not its original 'type' (e.g. Flenniken and Raymond 1986; Dibble 1991; Clarkson 2005). In addition, while a finished artefact may appear similar to another finished artefact, the mode of manufacture may have been significantly different, or two 'types' may have been part of a sequence of manufacture (Knutsson 1988).

In terms of the quartz research, two broad camps can be discerned – between those who argue that a separate typology is necessary (e.g. Barber 1981a; Knutsson 1988), and those who maintain that quartz can be analysed in a typological framework devised for flint (e.g. Bisson 1990; Saville and Ballin 2000). For the latter group, a separate quartz typology would in effect get in the way of an easy and coherent typology of stone tools that can be compared to assemblages of other raw materials or in mixed assemblages, and the emphasis of these researchers is primarily focused on the analysis of 'tools', with tools being tools if they conform to attributable types, and especially show evidence of retouch. On the other hand, the researchers that have called for a separate typology for quartz have done so with the recognition that the fracture mechanics of quartz entail that fracture characteristics seen in materials like flint do not necessarily occur on quartz and that the prehistoric users took advantage of the differing fracture mechanics in selecting pieces for use; therefore a schema devised with the fracture mechanics of the material as the lynchpin is crucial. Only with this in place can different raw materials be compared.

Ballin (2008) has been highly critical of the Scandinavian work that has called for a separate typology for quartz; Ballin wants to be able compare assemblages across raw material types regardless of the fracture mechanics of the raw materials and how these affected the manufacture and use of stone. He is explicit in his stance of what a 'tool' is, and how to construct typologies – 'a quartz artefact is not a tool unless it has the distinctive retouch generally associated with a particular tool type' (Ballin 2008). From this, we can see that Ballin takes the restrictive view of a tool being a tool only if it is retouched, which goes against much work in the last 30 years of lithic analysis (of quartz and non-quartz), especially from use-wear analysis, as well as ethnographic accounts (for examples of unretouched artefacts shown to have been used as tools see examples in Man 1883; White and Thomas 1972; Hayden 1979; Flenniken 1981; Symens 1986; Knutsson 1988; Odell 1994; Banks 1996; Kozlowski et al. 1996; Read and Russell 1996; Finlayson and McCartney 1998; Briels 2004; Hardy 2004; Setzer 2004; Shott and Sillitoe 2005; Akerman 2006). Consequently, his interpretation becomes a vicious circle – if a quartz artefact does not conform to a particular tool type's retouch, it can be discounted as a tool, and therefore becomes lumped as 'debitage'.

This issue is precisely what the Scandinavian research he disagrees with attempted to prevent: the forcing of quartz into a flint framework of analysis. And Knutsson's (1988) analysis of mixed quartz and flint assemblages shows that by examining them from a technological perspective, with a clear picture of the fracture mechanics involved for each type of raw material, they can be compared, and can only be compared directly taking the fracture patterns into consideration. What is of interest of course are Ballin's (2008) five points where he suggests why quartz assemblages differ from non-quartz assemblages and appear to have less 'formal' tools; here, he notes that quartz fractures differently, implying a need to understand the fracture mechanics of the material, and admits that quartz may have been used without retouch as tools; consequently, according to his own logic, direct comparison with flint assemblages may not be suitable, and contradicting himself, he concedes that 'tools' are not restricted to retouched tool types as devised by archaeologists. Of course, part of this negative attitude by Ballin towards a separate typology stems from the distinction made earlier between a 'typology' implying a 'tool' typology, and that of a debitage typology; Ballin is more critical of the former than the latter, and wants a similar 'tool typology' in order to allow the easy comparative analysis of mixed material assemblages.

The debate here, although seemingly academic, is absolutely central to the development of approaches to quartz – as it asks us to consider little less than the purpose of our analytical typologies and the kinds of comparisons they facilitate. Saville and Ballin 2000 are clearly correct in as much as a quartz-based typology should not endlessly recreate new names for old objects: a barbed and tanged arrowhead in quartz is a barbed and tanged arrowhead in any material. However, and crucially, the radically different fracture properties of quartz imply that a different understanding of past technologies is needed. Following Knutsson (1998) there is considerable danger that a focus on formal properties, in the absence of detailed technological models, will lead to the misidentification of supposedly significant artefact types; conversely, as noted by Gramly (1981), a lack of formal types in a quartz assemblage when compared to a flint assemblage can be interpreted erroneously as resulting from a different group of peoples, instead of the same peoples approaching the material differently. The stance taken in this project is that certain, highly formalised artefacts may have direct typological comparanda in alternative materials, and that in these instances such relationships should be highlighted. However, it will also be critical to understand those objects in terms of the technical sequences, the chaîne opératoires, which have led to their formation – and that these sequences may not be comparable across materials. Such an understanding can only be generated through detailed understanding of the properties of varied materials.


© Internet Archaeology/Author(s) URL:
Last updated: Wed Jun 10 2009