As noted, those few earlier citation-based studies in archaeology examined power relationships related to gender inequality and academic prestige (Victor and Beaudry 1992; Beaudry and White 1994; Hutson 2002; 2006): specifically, whether archaeological research published by women was undervalued by a process of preferential citation to research by men - as has been claimed for anthropology (Lutz 1990). Initial research indicated that women received fewer citations (Victor and Beaudry 1992), but that over time the number of citations to research by women was increasing (Beaudry and White 1994). A detailed study by Hutson (2002) based on citing documents published within a ten-year period in four sample journals (American Antiquity, Journal of Field Archaeology, Ancient Mesoamerica and Southeastern Archaeology) concluded that while the level of citation of research by women was lower than might be expected if the percentage of citations to men and women was to match the respective percentage of research outputs published by them, both men and women cited publications by women equally. Hutson (2006) also showed that self-citation was more commonly practised by archaeologists than other researchers in the humanities, but no more so by men than women; rather it was older researchers who cite their own work more commonly than younger researchers, perhaps because they had more work to cite.
The maps presented here (Figure 3, Figure 4) allow us to expand upon these earlier studies by looking across the discipline as a whole rather than a select sample. While the citations in the network map of authors have been aggregated, making it impossible to examine differences in citations made by male and female authors in the same manner as Hutson (2006), the network map of authors opens a window upon two other important aspects of gender and citation practice that have not yet been considered within the discipline. These are (i) the relative standing of men and women in terms of citation numbers in the discipline as a whole, and, (ii) the relative standing of men and women as highly cited researchers within the various thematic clusters. This allows some critical reflection on the discipline and follows up on a long discussion in the sociology of science on 'the Matthew Effect' – the practice by which intellectual recognition through citation accumulates to certain prominent or established researchers while other researchers whose contributions might have been equally significant in terms of intellectual merit receive less recognition (see Merton 1968; 1988; Zuckerman 1977). Here I have examined the number of citations received by a named author, and the clustering of these authors into thematic groups to consider the relative recognition given to men and women through citation. As noted above, it is important to remember that the data presented here only include the first authors of any work.
In accord with the Matthew Effect, a graph of citation numbers by author for the fifty most cited (first-) authors shows that there are a small number who receive significantly more citations/intellectual recognition than others - Binford, Stuiver, Hodder, Lyman, Reimer, Bronk-Ramsey, Ambrose, Renfrew, Kirch, and then the number of citations starts to decrease at a steady rate. There are just six women (12%) in this list – Piperno, Stiner, Wadley, Meskell, Buikstra and Villa, reinforcing an impression that intellectual recognition through citation is more generally accorded to men. Looking more widely to the most cited 250 (first-) authors, the number of women increases (19%) but it is still less than might be expected - estimated at 25% by Hutson with respect to his sample (2002). Taking the discipline as a whole, therefore, earlier observations about the under-recognition of research by women appear to be supported.
|Cluster number||Thematic cluster grouping (of authors)||No. of authors in cluster||Sample analysed|
|No. of authors in sample||Men/Women in sample||% of women in sample|
|2||Ancient environments and carbon dating||219||22||19/3||14%|
|3||Theory and interpretation in Europe and Old World||215||21||19/2||10%|
|4||Archaeology of Central and Eastern Mediterranean (including Classical World)||202||21||19/2||10%|
|5||Archaeology of Australasia and Pacific||143||15||14/1||7%|
|6||Materials analysis (esp. ceramics)||141||16||15/1||6%|
|7||Archaeology of mobile societies in North America||138||16||15/1||6%|
|8||Archaeology of Early State Societies in the Americas (N and C America)||127||13||7/6||46%|
|9||Archaeology of Early Settled Societies in the Americas (esp. N. America)||117||11||10/1||9%|
|10||Archaeology of diet (including the origins of domestication/agriculture)||111||11||8/3||27%|
|11||Archaeology of human remains (incl. forensics and fossil hominins)||106||11||10/1||10%|
|13||Archaeology of Africa||89||10||8/2||20%|
|14||Early occupation of the Americas||85||10||8/2||20%|
|15||Archaeological survey (esp. remote sensing)||59||10||8/2||20%|
|16||Archaeology of South American civilizations||58||10||8/2||20%|
|17||Dating (esp. thermoluminescence and magnetics)||36||10||7/3||30%|
|18||Archaeological chemistry (esp. lipids analysis)||26||10||6/4||40%|
|19||The Bronze Age in the Near East||18||10||9/1||10%|
|Archaeology as a whole discipline||2229||222||190/32||14%|
The network map of authors (Figure 4), however, allows us to look at the actual networks of communication between clusters of authors across the discipline as a whole, and in so doing we can see that intellectual recognition given to women varies markedly by cluster (see Table 10). Within three communication networks women are the most cited researchers (8 – Archaeology of Early State Societies in the Americas – Rosemary Joyce; 11 – Archaeology of human remains – Jane Buikstra; 14 – Archaeology of the early occupation of the Americas – Dolores Piperno). Examining the overall level of citation of women's and men's research on a cluster by cluster basis reveals there is considerable variation (taking a sample the top 10% most-cited authors in networks of 100 authors or more; and a minimum sample of the ten most cited authors in smaller clusters): in 3 networks (clusters 5, 6 and 7) women are represented at a low level (6-7%) while in 5 networks (clusters 8, 10, 12, 17and 14) women are more represented (27% to 45%). Unfortunately, it is still difficult to determine how this observed level of representation relates to an expected level of representation. Hutson (2002) bases the expected level of citation on the numbers of publications produced by women through the period of the citing half-life of his sample journals (10 years) using articles published over a three-year period in nine source journals. This procedure cannot be replicated across the whole of the discipline of archaeology where the number of documents cited is as many as 250,000 in a ten-year period. We might, however, use the percentage of women employed in universities as a proxy measure for potential citation. Using surveys for the years 1997-98, 2002-03, 2007-08 and 2012-13 for the UK (Aitchison 1999; Aitchison and Edwards 2003; 2008; Aitchison and Rocks-Macqueen 2013) and for 2012 for the USA (Surface-Evans and Jackson 2012) the percentage of women in academic posts increases from 35% to 46% in the UK, and is measured at 49% in the USA. On the basis of this data only two communication networks (cluster 8 – Archaeology of Early State Societies in the Americas, cluster 18 – Archaeological chemistry) show a level of recognition for research published by women that appears to match their actual participation, with possibly three other clusters (12 – Isotope analysis; 17 – Dating, esp. thermoluminesence and magnetics; 10 – The archaeology of diet) coming close. The higher representation of women as cited authors in communication networks associated with scientific analysis perhaps reinforces an observation made by Joan Gero, who noted that women in archaeology were most active in laboratory-centred research activities rather than excavation/fieldwork related activities (Gero 1988). Significantly, the network map of authors presented here provides the most effective mechanism for the identification of actual communication networks within which patterns of gender-differential citation might be examined in detail.
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.