The inclusion of explicit references to the work of other authors is ... a central feature of academic research writing, helping writers to establish a persuasive epistemological and social framework for the acceptance of their arguments' (Hyland 1999, 344).
Although the process of citation is accepted as central to the practice of academic communication, there remain many uncertainties about why and how citations are made (Smith 1981; Cronin 1981; 1984; Leydesdorff 1987; 1998; Nicolaisen 2007). Research on the citation process has informed us about the academic and sociological context in which any knowledge claim is made through the many ways in which citations are used and the network of relationships they draw between documents' authors and sources is created (Price 1965). Citation(s) are used to show influence or assign credit to others in the academic community for the use of their influential 'intellectual property', reinforcing the norms of acceptable scientific behaviour (Kaplan 1965). In so doing, citations make clear one's own original contribution and intellectual identity (Hellqvist 2010). Not all citations, however, recognise influence in the same way; some citations mark out the area of contribution of the knowledge claim through both positive and negative comment on earlier work (MacRoberts and MacRoberts 1989). Citations may also serve to convince a reader of the reliability of the claim and/or claimant by persuasion (Gilbert 1977), through polite rhetorical practice (Myers 1989) or by demonstration of sufficient understanding of relevant background research – citations as a form of costly signalling of authorial expertise to potential reviewers (Nicolaisen 2007). The number of citations per research article or per 1000 words of text has been claimed to be higher in the social sciences and humanities, and lower in the hard sciences, to be more common in the academic literature of academics from some countries rather than others, and to be increasing through time (Ucar et al. 2014). The number of citations also reflects the degree to which the author presumes that her readers can either share her common assumptions of disciplinary knowledge, or cannot since problem areas are more diffuse in topic and history (Hyland 1999; Becher and Trowler 2001). Initial studies, examining publications in anthropology (Lutz 1990) and sociology (Davenport 2007), have also indicated that there is a degree of gender bias in the pattern of citation, with women being significantly less cited by their male colleagues. Finally, the reason for, and context of, the citation of any piece of academic communication changes with the life history of the knowledge claim it makes and the shifting understanding of that claim within an academic community. Longitudinal studies based on publications in the sciences suggest that, early in the life history of a knowledge claim, citations might reference the particular data or specific arguments of any claim (Cozzens 1988), while at a later date, citation to the broader aspects of any claim may act as signs or concept symbols (Small 1978) that serve to mark a claimant's position in relation to particular ideas, schools of thought, or even methodological approaches.
Before presenting the network maps of archaeology, a brief discussion of citation is required since it is the fundamental process by which these networks are created. Academic communications, as exemplified by journal articles and monographs, are in essence a form of knowledge claim (Cozzens 1985; Gilbert 1976) made within a specific community of scholars – a discipline or a smaller specialism or specialty. This claim might be to a new understanding of the relationship between elements within a research context, or to the presence or discovery of new information (or material finds in an archaeological context), or to the existence of new links to other areas of research or the value of new forms of appreciation of our evidence. In support of any knowledge claim, the defining feature of scientific communication is the use of citations (Price 1965; Smith 1981; Cronin 1984). At the most general level of description, citations recognise the influence of a prior publication (the cited document) and its author(s) upon the research process that led to the current publication (the citing document). They are part of the formal accounting process of research (Cronin 1984, 6).
Following the extraordinary impact of the long-term recording of citations upon scientific practice, there has been a need to understand exactly what citations record and why or how they are made. The earliest research on citations concentrated on understanding the citation process (See Box 2) as a reflection of the tacit rules of science as a social and academic practice (Smith 1981; Cronin 1981; 1984; Hellqvist 2010; Hyland 1999; Leydesdorff 1987; 1998; 1998; Nicolaisen 2007). The citation process binds documents (individual knowledge claims) together through the links made from the citing to the cited document (Price 1965). As the aggregated number of citing and cited documents increases, single relationships grow into networks where individual documents become nodes and their citation links become the edges between them. Since documents usually possess both authors and sources (journals/edited books) any network of citations can be represented as edges between nodes in terms of different document attributes, such as authors or sources. The patterning of the nodes and edges in any network of knowledge claims reveals how any discipline, or smaller domain (such as a specific research area within a discipline), is structured either internally and/or externally through the links made to other domains and disciplines. When drawn, this becomes a visual map of the complex relations between the largest of networks of academic knowledge claims, and if mapped across different time intervals we can see an emerging picture of the changing shapes of knowledge. The visual mapping of bibliometric data and the citation relations between documents is now the most effective tool for visualising the shape and structure of the sciences and social sciences (Börner 2010; Börner and Polley 2014; Börner et al. 2003; Boyack 2004; Boyack et al. 2005; Samoylenko et al. 2006), for understanding the publication structure and development of terminology of disciplines and smaller research domains (e.g. Dolfsma and Leydesdorff 2010; Gobster 2014; Goldstone and Leydesdorff 2006; Shiffrin and Börner 2004; Skupin 2004; Yuan et al. 2014), for plotting the geographical spread of research topics, and for exploring the nature of interdisciplinary and multidisciplinary responses to research questions (e.g. Leydesdorff and Persson 2010; Klavans and Boyack 2014). This work has been facilitated by the development of specialist analytical and mapping software that can read data gathered from the citation indices, explore and visually represent/map the essential shape of a discipline and its knowledge structures (Cobo et al. 2011).
Most citations-based studies, such as those cited above, focus on communication practices in the sciences and, less often, the social sciences: analyses of citation practices in the humanities, although growing in number, remain relatively infrequent when compared to the sciences (see review by Ardanuy 2013). For anthropology, there are a number of studies of citation behaviour examining issues such as the primary journals used, the referencing of literature outside of that recognised as classically anthropological, the nature and number of references cited, the potential fragmentations of author networks in the discipline and library structures to support its research, and gendered patterns of citation (see Clark and Clark 1982; Choi 1988a 1988b; Lutz 1990; Michalski 1999; Robinson and Posten 2005). Citation studies for archaeology are rare and usually focus on the analysis of a method of practice in the discipline, or a distinct form of citation. For example, Sterud (1978) used citations to explore the growth of the processual approach to archaeology by tracking citations to documents with titles that indicated an interest in matters of social process. More recently, citations have been use to examine the gendered nature of power relationships and academic prestige in archaeology as evidenced by differential citation (see Victor and Beaudry 1992; Beaudry and White 1994; Hutson 2002; 2006; and the discussion 'Gender and the Intellectual Base of Archaeology' below). The most recent use of citations for the analysis of archaeology is by Brughmans (2013; 2014) who has examined making use of network analysis.
Until now, however, there have been no examples of the use of aggregated citations data to develop an overview of the discipline of archaeology and allow the relationships within the discipline to be explored in detail. When reflecting upon their discipline, archaeologists usually produce accounts of its structural development and history that take the form of narratives; intellectual histories in which authors seek to account for intellectual developments through a type of 'thick description' in which an author's expertise allows him or her to identify influence and effect through an examination of the detail of the language used and their own familiarity with the personal history of the participants: see McNairn (1984) and Trigger (2006) for well-respected examples. Such studies trace a path through a set of publications aided by the interests and concerns of the author as guide. Sometimes they will follow clearly recognised paths of inference as revealed in citations from one publication to another; at other times they rely on the overarching knowledge of the author to draw out links to other publications or sources of inference that might not be explicitly recognised. If, however, we accept that the exponential growth in the number of published outputs surpasses the ability of any individual, or even small groups, to read, then such studies are increasingly difficult. There is something valuable and important to be gained in presenting an overview of archaeological research networks based on a mapping of the aggregated citations data across all research publications, as has been done for other science disciplines.
Internet Archaeology is an open access journal. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.