header
published quarterly by the university of borås, sweden

vol. 27 no. 3, September, 2022



Citing methods literature: citations to field manuals as paradata on archaeological fieldwork


Isto Huvila, Lisa Andersson and Olle Sköld


Introduction. This article investigates how researchers cite methods literature, and to what extent and how these citations could function as a form of paradata i.e., descriptive data on research processes.
Method. Citations to two prominent field manuals were retrieved using Scopus; full-texts were obtained for analysis.
Analysis. Descriptive statistics and qualitative content analysis were used.
Results. Field manuals are cited both for compliance and contrast to clarify procedures and actions, understanding of what is considered conventional and extra-ordinary, to elucidate work processes in broader terms, and to explain concepts and what is common disciplinary knowledge. Even if literature use seems indicative of work procedures, a citation to a method cannot necessarily be considered as direct evidence of what was done in reality.
Conclusions. Citations to field manuals can function as a complementary form of paradata to other information on how archaeological work has been conducted. However, rather than forming a standalone corpus of evidence, they can be expected to function best if combined with other indicators. A citation to a specific methods text can be indicative of certain patterns of work or presence of a shared scope of relevance with other works citing the same text.

DOI: https://doi.org/10.47989/irpaper941


Introduction

Researchers can report on the research methods used in specific studies in multiple ways. In many disciplines, it is conventional to include a section in research reports that provides a descriptive narrative of the methods, materials and procedures used in the study. Apart from explaining what was done, it is also conventional to provide citations to methods literature. Similarly to citations in general (Wouters 2014), also citations to methodological texts and manuals can serve multiple purposes (Thompson 2010).

In spite of the thinkable variety of motives and disciplinary differences in the prevalence of citing methods literature, there is not much doubt that perhaps the most typical reason for citing these types of texts is the one noted in the methods literature itself (e.g., Schneider 2015; Ravitch and Carl 2020), that is, to explain why and how the procedures of collecting, managing and analysing a particular set of ’material’ (or, data) chosen and used in the particular study should be considered adequate and relevant. In this sense, the citations function as paradata i.e., data about the process of how the data came into being and how it was used to achieve the reported results (Couper 2000; Huvila 2012).

Citations can complement other forms of paradata (e.g., manual descriptions of procedures, automatically extracted and collected process data) as evidence of the appropriateness of research procedures, increase trust in the results, and improve the (re)usability of earlier research results and data (cf. Huvila 2022). Some of the potential advantages of using citations as paradata are that they are plentiful and reasonably easily available for analysis, and there is a large corpus of established methods, tools and theories relating to citing and citations developed in the context of bibliometrics research. So far, in contrast to their potential significance, there is, however, little earlier research on to what extent the citations in practice can inform of the origins and underpinnings of the data.

The aim of this article is to address this research gap and provide insights into 1) how researchers cite methods literature when they are reporting their research, and 2) to what extent and how these citations could function as paradata, i.e., serve as information about how the research was conducted. The investigation is based on a study of how archaeologists cite two prominent archaeological field manuals in research publications and how these references are indicative of data collection, management and use procedures. Archaeology provides an interesting context for this type of inquiry as a data and data collection driven (Hodder 1999; Khazraee Afzali 2014) interdisciplinary field in the nexus of humanities, social sciences and sciences with a strong tradition of data and data collection intensive research.

Literature review

Citations and methods literature

Determining the meaning of citations, including the criteria, quality and reason for citing, is one of the major questions of bibliometrics scholarship (Borgman and Furner 2002). On a macroscopic level, they have become an increasingly central element of the infrastructure of scholarly communication, its impact and the practice of research itself (Wouters 2014; Cronin and Sugimoto 2014).

Citations have many functions (Garfield 1996). Ingwersen has noted that they are ‘footprints of information interaction[s]’ and ‘manifestations of degrees of utility of methods, results and ideas’ (Ingwersen 2012, p. 1). Further, they can be seen as indications of the relevance of cited works, at least at some point in a research or writing process (Borgman and Furner 2002), and a part of the process of argumentation (Carrascal 2014; Gullbekk and Byström 2019) and social practices of performing scholarship (Gullbekk and Byström 2019).

Experienced researchers and readers of scholarly literature tend to have a working understanding of the meaning of citations in their own areas of expertise and how to cite others but the conventions are not necessarily easy to articulate or navigate (cf. Gullbekk and Byström 2019). In addition to disciplinary differences even within seemingly similar clusters of research fields, the function of citations vary from case to case within a text and between different types of texts written for different audiences (Hammarfelt 2016), similarly to how different types of texts and resources are cited differently. This has been linked to the diverging roles different genres play in scholarly communication (Bazerman 1988; Swales 2004). The length and comprehensiveness of books mean that they are cited in a broader variety of ways than journal articles (Glänzel et al. 2016). As Hammarfelt (2011) shows further, the citation profiles of the different parts of books also vary.

Citation analysis has been used to a certain extent in earlier research to analyse methodological literature. A part of the studies has focused on the mapping of the use of research methods in specific disciplines. Patil and Kant (2014) have investigated methods used in knowledge management research and Partington and Jenkins (2007) have analysed citations to research methods literature in organization studies. The latter provides a template for structuring citation usage in academic research, however, primarily for educational rather than strictly analytical purposes. Bricker (1988) analyses accounting literature and notes that a large part of references out of the research field are citations to methodological and theoretical work.

There are also studies of handbooks, which have certain similarities with methods literature (Milojević et al. 2014). Closest to the present study, Brughmans (2014) has investigated citation networks related to formal network analysis in archaeology. The findings show that citation networks are fragmented and many archaeologists have traditionally focused on rudimentary methods rather than the state-of-the-art. Sinclair (2016) conducted a comprehensive analysis of all archaeological literature indexed in the Web of Science data and published in 2004-2013. The study identified a large theory and interpretation cluster, and several clusters relating to specific scientific methods but none specific to field methods.

Others have focused on analysing the influence of particular methods and texts. For instance, Titscher and colleagues (2000) have surveyed the popularity of different text analysis methods in social science literature using keywords based approach whereas Thompson (2010) analyses citations to a seminal article by Golder (2000) that describes and exemplifies the use of the historical method in the context of marketing research. Her analysis of the most cited articles citing Golder shows that the text has been cited for its empirical findings and the method, or parts of it, as they have been used in later studies, and to discuss the method, its strengths and weaknesses.

Archaeological field manuals

The appearance of field manuals is linked to the evolution of archaeology as a scholarly rather than antiquarian discipline. In Anglo-American archaeology, theory and method remained tightly interconnected until the early postwar years. During the first decades of the 20th century, only a few field manuals were published whereas the number has proliferated since the late 1940s (Collis 2013). Simultaneously, the theoretical and methodological discussion drifted apart to a degree that began to raise concerns about the lack of theory in field archaeology (Lucas 2012). Even if some of the manuals have had an impact on fieldwork practices around the world (Collis 2013), as Aitchison (2017) argues, archaeology is highly parochial and local. Therefore, the different manuals tend to have fairly distinct orientations to particular national practices (e.g., Hester et al. 2009 for the US, and Joukowsky 1980 for US overseas archaeology; Westman 1994 for British urban archaeology; for German handbooks, see Davidović 2009). In some cases, international collaborations and campaigns have imported foreign practices to specific archaeological sites. Sometimes, direct influences spread when archaeologists study abroad and return to their home countries (e.g., Barretto-Tesoro, 2013). More often, the influences are mediated and adapted to the local practices.

At its most basic level, the field manual should provide an introduction to field archaeology for the person who does not want to display ignorance and incompetence during the first days of an archaeological excavation’ (Dyson 1981, p. 78). As Dyson continues, it can also provide information on what ’archaeologists actually do’ and provide guidance to experienced fieldworkers to develop their skills (Dyson 1981, p. 78).

In parallel with the calls for comprehensive texts, the attempts to write comprehensive prescriptive manuals have also been criticised (e.g., Stanford 1981; Straus 1981) for the inherent impossibility of the task. An experienced archaeologist should know everything covered in a generic manual and needs only specialist publications for learning purposes (Dyson 1981). The ambivalent attitude that underlines the importance of manuals and criticises their shortcomings revolves undoubtedly around the attitude Roskams (2001) has described as manual worship. A further, perhaps not too surprising, related issue is that in spite of their perceived importance, the actual practice is not necessarily the one described in the manuals as Sellers (1973) suggests in a humorous but fundamentally very serious essay on the gap between explicit and implicit expectations and practices.

While archaeological manuals have been both praised and criticised, the manual of field archaeology as a literary genre evades a concise definition beyond a basic-level shared understanding reflected, for instance, in the above-quoted passage in Dyson’s (1981) text. There is a considerable variation in, for instance, for whom they are written (e.g., students, experts, field directors), what is their primary purpose (e.g., to function as a textbook, cheatsheet, handbook), and to what extent their audience and aims are articulated and plausibly met or not (cf. e.g., Pavel 2011; Straus 1981; Dever 1981).

In addition to generic texts, there are numerous manuals and handbooks describing specialised techniques from archaeological surveying to soil analysis (Carver 2009), and manuals or guidelines issued by individual research projects, and local and regional archaeological authorities. In regards to generic field manuals, individual authors have referred to slightly different sets of texts as classics (see e.g., Carver 2009 and the lists of recommended readings in Hester et al. 2009; Joukowsky 1980; Roskams 2001 and others). In some cases, a particular volume might be popular as a textbook (e.g., Barker 1993; Joukowsky 1980) but not necessarily representative of field practices in a particular country (Thorpe 2012). Some volumes, like Joukowsky (1980), have been described in different sources as textbooks (Payne 1982; Gaughwin 1982) or as manuals (e.g., Spoerl 1982; Dyson 1981; Dever 1981; Straus 1981; Stanford 1981; Thomas 1981) whereas some, as Davidovic (2009) notes of German field manuals, have been written specifically for field directors, not students.

As a whole, the genre of field manuals appears to meander somewhere between being and not being a textbook. The recurring question in the reviews of field manuals of the audience of the texts is symptomatic of this ambiguity. In best cases, field manuals are suggested to be useful for both newcomers and specialists (e.g., Spoerl 1982) but all too often the critics struggle to see whether they would be relevant for either of the groups (e.g., Pavel 2011; Straus 1981; Dever 1981).

Theoretical considerations: paradata and citations

The concept of paradata refers to data that describes processes (Couper 2000). It is comparable to metadata that describes data (Pomerantz 2015) but also closely related to terms context (e.g., Faniel et al. 2013), provenance and provenance metadata (e.g.,Doerr et al. 2016; Huggett 2012 cf. the earlier popular notion provenience i.e., the origins or birthplace of an object, Buchanan 2016) as a specific kind of information relating to the origins, context and processes pertaining to the earlier life of data or information.

The importance and usefulness of knowing about processes and origins of data have been acknowledged for some time both in archaeology and other domains as a key constraint of research with secondary data. Analysing excavation documentation created by someone who has not excavated at a particular site is ‘a very difficult endeavour’ (Demoule 2011, p. 8) not necessarily because of the lack of data or metadata, but because there is not enough contextual knowledge to interpret them properly (Voss 2012; Faniel et al. 2013).

A central problem with acquiring and keeping useful paradata is the difficulty of foreseeing what future users would exactly need to know and that manual documentation of scholarly processes is very time-consuming. Therefore, it has been proposed that a fruitful course of action to approach the problem would be to use automatically collectable paradata and forensic post-hoc approaches to extract paradata from data and metadata to reduce the amount of manually created paradata (Huvila 2022).

The suggestion that a citation to a methodological text in a (scholarly) publication is linked to particular understandings of (scholarly) processes is derived from citation theory. In citation theory (Leydesdorff and Wouters 1999), a citation to a methods text can be hypothesised to be a direct pointer to a method i.e., what was done. However, it is also conceivable that these citations might have other functions as well, and (or) that the citations to a particular modus operandi is more indirect. In a parallel sense, as a part of the formalised paratext of a scholarly work (Leydesdorff and Wouters 1999), a citation is a sign of membership in and codification of particular scholarly communities (Leydesdorff 2011), epistemic frameworks (Hyland 2013), or invisible colleges (Crane 1972) indicated by the presence of co-citations (Farideh 2009) to particular methods or methodological texts.

Similarly, a presence of co-citations could sign that multiple authors share the same or similar ‘documentation ideals’ (Börjesson 2016, p. 674). Börjesson writes that documentation ideals concern ‘what documentation should be like, for what purposes, and for whom’ (Börjesson 2016, p. 674). In comparison to the literary formulations found in official texts like information policies (Börjesson 2016) or methods literature, the ideals are closer to practitioners and their everyday work. Similarly to Börjesson’s (2016) analysis of documentation ideals and information policies in development-led archaeology, the present study investigates how archaeological documentation is conditioned by literature. More precisely, the focus of this study is on how citations to field manuals can be understood as formal but explicit and implicit (cf. Börjesson 2016) expressions and links to certain (documentation) ideals of documenting archaeological practices.

Material

The present study is based on a qualitative analysis of citations to two prominent English-language archaeological field manuals. The first, Archaeological Site Manual, widely known as the MoLAS (for Museum of London Archaeology Service) manual, is a reference work originally published by the Museum of London to function as a field manual for archaeological investigations conducted in London. The first edition of the text was published in 1980 with second, third and fourth editions respectively in 1990, 1994 and 2002. The second field manual included in the analysis is Martha Joukowsky’s A Complete Manual of Field Archaeology (Complete Manual) published in 1981 as a literal attempt to to provide a complete reference for conducting fieldwork in archaeology.

The two field manuals have become well-received not only in their countries of origin but also to a certain extent, around the world (Collis 2011; Hodder 1997) as two popular albeit very different types of reference works on archaeological field methodology. They have been referred to as being canonical or standard (Thorpe 2012; Masur et al. 2014) works in archaeological field methodology even if, as discussed later in this article, there are limits to the degree their recommendations have been followed in practice.

The Archaeological Site Manual is essentially a ’site’-specific manual for archaeology in London and as such resembles many other project field handbooks produced by archaeological projects and organizations around the world from larger research projects to contract archaeology operators and archaeological administrative bodies. The main objective of these texts is to standardise archaeological field practices within a given organization or area. At the same time, however, this Manual has become a standard reference to a specific fieldwork approach known as the >single-context method developed in the UK from the 1960s onwards, especially in the context of urban archaeology (Collis 2011). Even if the method is far from being a standard approach to conducting fieldwork, especially outside of the UK, it has inspired many archaeologists and different versions of it have been adopted in use around the world.

In contrast to the Archaeological Site Manual, the Complete Manual was written as a generic reference work even if it draws heavily from its author’s experiences and background in classical archaeology. Thorpe (2012) suggests that it is more representative of fieldwork conducted by American archaeological missions abroad rather than, for instance, of fieldwork in the US or in another individual country. This applies even if the Complete Manual contains a lot of US-specific references to excavation and funding opportunities (Thomas 1981).

Even if it has been criticised, both at the time of its publication (e.g.,Stanford 1981; Thomas 1981; Straus 1981; Spoerl 1982; Payne 1982) and later, (e.g.,Buccellati 2017) for omissions, an urge to standardise all aspects of archaeological fieldwork (Stanford 1981), and for being atheoretical, it is has considered by many to be a ’canonical’ (Thorpe 2012, p. 42) ’tour de force... advising people how best to proceed with the practicalities of archaeological fieldwork’ (Thomas 1981, p. 670), and a ’virtual Larousse Gastronomique of field archaeology’ (Dever 1981, p. 86).

The two field manuals were selected for the present study based on a pre-study of citation profiles of popular field manuals and textbooks using Google Scholar and Scopus. The total volume of such literature is large. To exemplify the extent of English-language literature, Caraher (2017) listed twenty-five excavation manuals he found using a Google search that were available online in 2017. The Parks Canada Archaeological Recording manual (2005) lists twenty-two general, mainly British, North American and French archaeology field manuals and general volumes in its list of recommended readings. Archaeology textbooks do also contain long, partly overlapping lists of manuals and other textbooks (e.g. Greene 1998; Roskams 2001). However, as Thorpe (2012) remarks, many of the texts are used only locally in specific projects, regions and parts of the world. In addition, even if some texts have been longer-lived than others, many of them have been disseminated as grey literature rather than as published books (Caraher 2017). In this respect, the two manuals are examples of manuals with a broader than usual scope of citation and use.

The principal reasons for selecting the particular two manuals were their recognised status in archaeology, (consequent) comparably high citation counts, and distinct profiles (as explained above). As Thomas (1981, p. 671) notes of the Complete Manual, ‘[t]here will always be room in a field library for this book’ and even if ’the crew might not refer to it often... it’s nice to know it’s there’. Even if Joukowsky sees novices as the principal audience of her book, neither of the two are introductory textbooks proper (cf. Watson 2019), or specialist handbooks discussing very particular types of archaeological activities (e.g., processing of specific types of finds, conducting fieldwork in particular contexts). They both also focus on the practical aspects of working in the field rather than on theory. As such, they can be expected to have been cited at least to a certain degree as indications of how work was done, making it meaningful to enquire if these citations could be treated as a form of paradata on the processes of how archaeological fieldwork is conducted. The conceivable citation profile of project, organization and area specific manuals, textbooks and theoretical work is likely to be different and less interesting to the aims of the present study.

Selecting two different types of works was anticipated to increase the variety and breadth of insights into citation practices. The fact that they are not entirely new can be assumed to lead to a broader and more diverse citation profile than with newer texts.

Methods

The citation data for the main study with the Archaeological Site Manual (N=73) and the Complete Manual (N=60) were collected using Scopus in October 2019 to retrieve citations in scholarly publications. A list of the analysed publications is in Appendix 1. Individual texts are referred to in the reporting using codes (indicated in the Appendix) in brackets consisting either of A or C (referring to the relevant manual) and an index (e.g., A1 or C54). A cursory comparison between Scopus and Google Scholar data did not reveal any obvious substantial differences between the citation patterns in the two reference sources that would have invalidated the use of Scopus as a data source. The main difference was the higher citation counts retrieved in Google Scholar (227 for the Complete Manual, 126 for the Archaeological Site Manual). Other citation databases (Social Science Citation Index, the Book Citation Index) were considered but abandoned because of their lack of coverage of texts citing the two investigated texts. The full-texts of the citing literature were retrieved through the library system of a major research university where the main author of the article was working at the time of the study.

The material was analysed using descriptive statistics (frequencies) and qualitative content analysis to understand how researchers cite methods literature when they are reporting their research, what are the functions of the citations, and consequently, to what extent and how these citations could function as paradata, i.e., serve as information about how the research was conducted. The analysis covers the text authors’ country of origin and patterns of collaboration, text audience and disciplinary context, the citations’ functions, and the texts’ themes.

These aspects together provide the grounds to discuss citation practices and how citations can function as paradata. The qualitative coding of the material was done by the main author of the article and conducted iteratively first to identify preliminary categories and to consolidate them later in the course of the analysis using writing (cf. Richardson 2000) as an explicit method of inquiry. The analysis, including the reading of the texts, took approximately 120 hours. A sample of 20% of the material was reanalysed one month after the original analysis for quality control. The reanalysis led to minor revisions in the descriptions of the categories and themes, not in abandoning old or introducing new ones.

It is important to note that the chosen method limits the study to a small subset of published scholarly and scientific texts. The coverage of humanities-orienteddisciplines is limited in all major citation databases (Hammarfelt 2016), an issue that also pertains to archaeology (cf. Brughmans 2014) and Scopus. The exclusion of the quantitatively largest genre of archaeological texts, investigation reports (Börjesson 2016), is another shortcoming in the analysed material. The first omission was not considered to be critical because the present study does not aim to provide a comprehensive mapping of all citations to particular works, but rather to understand the patterns and functions of citing a particular type of literature based on a qualitative rather than quantitative analysis.

Grey literature was excluded for two principal reasons: first, the poor representation of grey literature in citation databases would have made it difficult to produce a dataset, which would have been comparable to the one with published texts. Secondly, a preliminary exploration of a corpus of reports from three European countries confirmed a working hypothesis that stemmed from reading earlier literature (Donnelly 2016; Börjesson 2015). It seems that the number of citations to field manuals in archaeological reports is very low and consequently, it is unlikely that these citations would function as a significant form of paradata in that particular context.

Analysis

Authors

Even if the analysed sample was fairly small, several of the texts citing the Archaeological Site Manual were written or co-written by the same authors. It suggests that particular individuals might have a habit of citing a field manual in their texts. Eleven authors had participated in writing two texts, and one author three. With the Complete Manual, nine individual authors had participated in writing two different texts in the material but none of them more than that.

Unsurprisingly, even if the evidence is admittedly anecdotal, most of the texts referring to the Archaeological Site Manual have at least one author with an affiliation in the UK, the country of origin of the manual (Table 1) whereas the Complete Manual was popular among authors with a US affiliation. It is equally unsurprising that three Israeli authors have cited the Complete Manual considering that its author is known for her fieldwork in the Middle East.


Table 1: Country affiliations (with N=1 for at least one of the two field manuals, retrieved from Scopus) of authors of the texts included in the analysis.
No. of texts referring to
CountryArchaeological Site Manualthe Complete Manual
United Kingdom4014
United States1227
Germany7-
Australia6-
Argentina5-
South Africa51
Spain41
Canada32
New Zealand31
Sweden31
Bosnia and Herzegovina21
Czech Republic2-
Turkey21
China-3
Israel-3
Finland-2

Audience and disciplinary contexts

Both manuals were cited in texts that explained archaeological methods for non-archaeologists and in texts that can be described as intra-disciplinary for archaeology itself. The citations from outside of archaeology relate often to the use of archaeological methods in other disciplines (e.g., [A66][C26][C35][C45]). Correspondingly, field manuals were cited in cases when methods borrowed from other disciplines (e.g., soil analysis in [A49] and [A12]) were applied in archaeological research. In these cases (e.g., [A13][A18][A60][A66][C26]), the citations could serve the purpose of explaining the usual procedures of archaeological work, or how an investigation was conducted in the specific case described in the text (e.g., [A2]). Table 2 lists the disciplinary subject areas assigned to the analysed texts in Scopus. The most frequently used categories were unsurprisingly social sciences (59 for the Archaeological Site Manual and 34 for the Complete Manual) and arts and humanities (58 the Archaeological Site Manual, 36 the Complete Manual), the disciplinary domains where archaeology is conventionally categorised. The presence of a fairly large number of other disciplines shows the interdisciplinary interest in archaeological research and the interdisciplinarity of the field itself.


Table 2: Subject areas of texts included in the analysis.
Subject areaArchaeological Site Manualthe Complete Manual
Agricultural and Biological Sciences 2 -
Arts and Humanities 58 36
Biochemistry, Genetics and Molecular Biology 1 1
Chemistry 1 -
Computer Science 3 9
Decision Sciences 1 -
Earth and Planetary Sciences 9 5
Engineering 3 4
Environmental Science 3 2
Immunology and Microbiology 1 -
Materials Science 1 -
Neuroscience 1 -
Psychology 1 -
Social Sciences 59 34
Business, Management and Accounting - 1
Medicine - 3
Nursing - 1

Themes

It was possible to identify certain thematic variation in the literature that contain citations to the two manuals. Partly, the analysis resulted in three broad categories of archaeological texts that cited them. First, a group of texts consisting of reports of fieldwork results (for examples, see category 3 in Table 4) cite the manuals predominantly to show how a particular investigation was conducted. A second major category includes theoretical and methodological texts that discuss, for instance, the implications of particular field techniques or archaeological practices in general, propose new methods or tools, or advocate the use of specific techniques in particular contexts (e.g., [A1][A8][A71][C8][C15][C28]). Third and finally, some of the non-archaeological texts cite the two texts as an information source of archaeologists’ work practices (e.g., [A2][C3][C25]). The Archaeological Site Manual seems to have been cited to a greater extent as a general reference to the single-context method whereas the Complete Manual has been cited for its descriptions of specific procedures.

A parallel to a thematic categorisation of the texts that cite the Archaeological Site Manual and the Complete Manual, it was possible to see a contrast in what types of texts cite respectively the Archaeological Site Manual and the Complete Manual. The pattern was obvious in close reading of the texts but could also be discerned in the keywords assigned to them. Table 3 lists author-assigned keywords (occurring more than once) retrieved from Scopus and used to describe the texts. Multiple texts citing the Archaeological Site Manual have been assigned keywords stratigraphy (9) and excavation (5), but also interpretation (2), excavation methods (2), and digital archaeology (3), that all relate to the approach and perspective of the texts i.e., what kind of archaeological research they are relating to.

In contrast, many of the geographic (e.g., Andes, Argentina) and thematic keywords (e.g., geoarchaeology, rock art, taphonomy, training, sediment analysis) are related to the empirical research contexts of the individual authors with more than one publication in the material. The higher proportion of fieldwork reports among the texts citing the Archaeological Site Manual is reflected in the keyword frequencies in comparison to the Complete Manual where all terms used more than once relate to archaeology and its subfields, and different tools and techniques. The same applies also to the frequencies of keywords in Table 3 and categories in Table 4. Citations to the Archaeological Site Manual are focused on pointers to the stratigraphic method (Category 1 in Table 4) and its applications (Category 3) whereas the Complete Manual was used as a generic source of knowledge and details of archaeological fieldwork (Categories 2 and 5).


Table 3: Author-assigned keywords (occurence in the dataset N=1) retrieved from Scopus used to describe the analysed texts.
Archaeological Site Manualthe Complete Manual
KeywordNKeywordN
Arch(a)eology 11 Archaeology 11
Stratigraphy 9 Excavation 3
Excavation 5 3D Scanning 2
Article 3 Archaeological Excavations 2
Digital Archaeology 3 Ceramics 2
Andes 2 Cultural Resource Management 2
Argentina 2 Data Reduction 2
Comparative Study 2 Forensic Archaeology 2
Excavation Methods 2 Forensic Pathology 2
Geoarchaeology 2 Interactive Computer Systems 2
Interpretation 2 Mathematical Models 2
Iron Age 2 Stratigraphy 2
Later Stone Age 2 Three Dimensional 2
Lesotho 2 Virtual Reality 2
Micromorphology 2 Visualization 2
Middle Horizon 2
Radiocarbon Dating 2
Rock Art 2
Sediment 2
Sediment Analysis 2
Settlement History 2
Swahili Coast 2
Taphonomy 2
Training 2

Functions of citations

Even if the citations in the two analysed texts had many similarities, the Archaeological Site Manual was cited as a specific reference to the single-context approach of field archaeology (in contrast to others), whereas the Complete Manual was cited typically as a source of what is commonly known about archaeological field practices. In some cases, the Complete Manual was cited for the archaeological subject matter described in the text.

The analysis revealed five primary functions of citations that are summarised in Table 4. It is worth observing that the categories are not exclusive and in several cases, a particular citation could be classified to have several parallel functions. For instance, in [A41] it appears that when the authors explain that ‘collected artifacts were catalogued following conventional methods’ (followed by a citations to the Archaeological Site Manual), they were simultaneously describing what was done at a particular excavation (category 5) but also what is considered to be a common practice in archaeology (category 2). Considering this, the number of citations to the two manuals per function given in Table 4 should be interpreted as indicative of their probable primary functions rather than as absolute figures. However, keeping this in mind, the analysis shows that the Complete Manual has been cited much more frequently as an information source on archaeological common knowledge and practices, whereas the Archaeological Site Manual has been cited to describe or refer to specific practices and work procedures.


Table 4: The five primary functions of citations identified in the analysis. All citations to texts in the quotations are given in standardised Harvard form as (Archaeological Site Manual) or (Complete Manual) for the sake of the clarity of reading. The occurrences of mentions (columns three and four) in the two manuals have been counted as interpretations of the primary functions of the citations.
FunctionExampleNo. in Complete ManualNo. in Archaeological Site Manual
1. Specific practices in field archaeology ’Cleaning protocols (Semenov 1964; Complete Manual; Sease 1987) used during stone tool curation call for treatments that may interfere with DNA and protein recovery.’ [C41]; ’In the West, standards for single context recording based on the Harris Matrix were already established more than years ago, notable that of the Museum of London Archaeology Service (MoLAS)’ [A68] 3 13
2. Regular practices in field archaeology ‘Archaeological field work involves the use of various specialized tools. (Complete Manual)’ [C2]; ’Since the nineteenth century, many practitioners have sought to observe and objectively document the world, be it the changing colours of soils or similarities of form. Archaeologists are trained in technical practices as a means of rendering things objective and allowing comparative analyses (e.g., Archaeological Site Manual).’ [A37] 15 5
3. How we did what we did ’In order to allow for good control over the excavation process, which involved the removal of soil in equal layers to prevent damage, remains were buried and excavated by means of archaeological techniques (shovels, brushes and trowels).’ (Complete Manual) [C3]; ’On the excavations we used a modified version of single context recording system as devised by the Museum of London Archaeological Service and widely used by field archaeologists trained in the UK or Ireland (Archaeological Site Manual). This system emphasizes excavator interpretation and allows for a greater sharing of on-site responsibilities.’ [A56] 3 18
4. Reference to a concept explained in the text ’The concept of assemblage in archaeology is in many ways a very loose term, used in various ways, but two of its most common meanings are a collection of objects associated on the basis of their depositional or spatial find—context (e.g., midden assemblage) and a collection ’Of one type of object found within a site or area (eg. pottery assemblage), often also referred to as an industry (eg. Complete Manual: 279; Carver 2009: 224; also see Joyce & Pollard 2010).’ [C24]; ’Many of the observational terms in Table 1 are used in other recording systems (e.g., those derived from the Archaeological Site Manual).’ [A1] 4 3
5. What is generally known in archaeology ’A report is a compilation of the description of an archaeological excavation or a survey process, a survey of the related literature, and an interpretation of the results of the investigation (Complete Manual).’ [C25] 11 0

To give a summary of the categories, first, the texts were cited to refer to specific practices in field archaeology such as a particular technique or an aspect of archaeological fieldwork and documentation. Such practices could be, for instance, a computer-based recording system used at the Museum of London [A72], a particular type of recording sheet [A66], a method of describing soil [A49], a burial excavation technique [C40] or a cleaning protocol [C41].

Second, the citations were used to make remarks on the regular practices in field archaeology i.e., how archaeologists act in general (e.g., [A60][A71]). For instance, [C46] suggests that the classification of pottery is done by categorising ceramic finds using two-dimensional drawing and measurement techniques, [C13] refers to typical health concerns relating to archaeological fieldwork, [A43] explains how the single-context method has become a common practice at urban excavations [A43], and [A37] how ’[a]rchaeologists are trained in technical practices as a means of rendering things objective and allowing comparative analyses’.

Third, citations to the manuals are used to explain how we did things i.e., how the author or a particular group of archaeologists conducted their fieldwork. These types of references were typical in texts reporting fieldwork results, and considerably more common in citations to the Archaeological Site Manual. The authors could explain, for instance, how pits were dug at a particular excavation [A7], what was the excavation strategy [A30], and what excavation protocols were adopted at a particular site [A34]. A citation to either manual could suggest conformity to a specific approach or technique (e.g., [A25][A31][A32][C31]).

Using the Archaeological Site Manual, it was common to refer to the single-context excavation method described in the manual (supposedly) in its entirety (exceptions e.g., [A61][A31) whereas with the Complete Manual, the citations tended to refer to individual techniques (ranging from surveying techniques [C29] and post-excavation analysis techniques [C11] to the management of pottery finds [C23][C28]) rather than to the manual’s approach as a whole (exceptions [C9][C45][C57]). Citations could also suggest that the adopted approach was ’based’ (e.g., [A67]) on, ’followed’ (or ’sequimos’ in [A55]), ’derived’ [A2], used ’a modified version’ of [A56], or ’modified and expanded upon’ [C36] a method explained in the text rather than ’used’ (e.g., [A5][A21][C40]) or ’employed’ [A4] it.

A third alternative was to use the citation to contrast the chosen (often characterised as a better alternative) or typical (characterised explicitly or implicitly as inferior, e.g., [A3][A53], or merely different, e.g., [A25][A29]) approach in a specific context and the approach described in a manual. For instance, [A29] contrasts the system described in the Archaeological Site Manual with the one used in North American urban archaeology.

Fourth, the citations could make a citation to what is generally known in archaeology. It could be used to explain what is an archaeological report [C25] and what are their ’key components’ [C26], that historical maps are important sources for understanding the past [C24], or that accurate and rapid measurements are crucial in archaeological fieldwork’ [C56]. These citations could be found only in relation to the Complete Manual.

Finally, citations are used to make a reference to a concept explained in the text. These could be the notion of ’context’ (e.g., [A3][A45][A52][A60]), ’assemblage’ [C24], ’field’ [C39], ’attribute’ [C47], or a broader set of, or instance, terms used to describe observations [A1][C19].

In addition to the five categories, one citation to the Complete Manual referred to a map published in the book on early population in North America [C44]. A small number of texts lacked citations and referred to the texts only in their bibliographies [C37][C38][C27][C55][A26], or in a list of recommended readings [C30]. In two cases the citation was apparently misplaced and was intended to cite another text [C1][C35]. In twelve cases, the Complete Manual was cited as a general example of archaeological field manuals or literature.

Discussion

Citation practices

The findings confirm several earlier observations and remarks on methods literature and in particular on archaeological field manuals. In general, the analysis suggests that it is relatively uncommon to make explicit citations to field manuals, especially when reporting fieldwork results. The low number and diversity of citations to the two well-known manuals suggest that they are cited when individual authors experience a need to make a specific point by providing a pointer rather than that citing them would be considered relevant or necessary by default.

Citations had also clearly multiple parallel functions, for example, to explain what is common practice and what was done during a particular excavation. This means that any identified primary functions of specific citations are at the most indicative. However, keeping this in mind, it was possible to see many similarities but also several differences in the citation profiles of the two texts. It was possible to discern variation in the expressions used to refer to the literature relating to how the cited works were used, from likely rather instrumental ‘use’ to ‘modifying and expanding upon’ them. The analysis shows similarly that the texts citing the Archaeological Site Manual were usually described with both in-total fewer and fewer unique keywords. A plausible reason is that the Complete Manual has a broader scope and it is cited by a more heterogeneous group of researchers, an observation that is supported by a qualitative reading of the individual citations and their contexts.

Both Watson (2019) and Buccellati (2017) criticise that there is little convergence between technical field manuals and theoretical literature. More specifically, field manuals hav e been criticised for lacking theoretical consideration or depth (Thomas 1981). The present findings confirm this observation in that the citations to the field manuals were unmistakably related to technical rather than theoretical considerations. At the same time, however, it was apparent that the field manuals embody to a certain degree a distinct kind of ‘practical theory’ (Cronen 2001) of archaeological field practice. This is especially apparent with the Archaeological Site Manual and how it is used as a reference to the ’theory’ and practice of the archaeological single-context method but it also to an extent applies to the Complete Manual.

Based on the analysis, it seems that citing a field manual is relevant especially when it is representative of the specific field practices discussed in the text but when the modus operandi is different from what a reader might expect. As Morgan and Wright (2018) note of archaeological drawings, those who are familiar with local conventions have no difficulties in interpreting them without explicit metadata. In contrast, it is difficult or impossible for others. A perceptible documentation ideal that appears to guide the use of citations is to use them as markers for practices that are at odds to what a seasoned reader would be inclined to expect.

As a whole, from the citing perspective, manuals might suffer from an image problem as the presentation text of Carver’s volume Archaeological investigation (2009, p. i) suggests: ‘[t]his is no plodding manual but an inspiring, provocative, informative and entertaining book’ that does, at the same time, intend to be ’a companion for a newcomer to professional archaeology’ from student to a ’fully fledged professional’. Quoting other, ‘[a] comprehensive guide to field archaeology is needed today’ (Spoerl 1982) and it may be useful to have in hand during an archaeological investigation (cf. Thomas 1981 on the Complete Manual) but not necessarily there to be cited without a very particular reason to refer to a specific technique or idea. The eclecticism of the archaeological practices of citing methods literature is further illustrated by the fact that similarly to some earlier field manuals (Thompson 2010), also the Complete Manual has been cited both for its archaeological content and method descriptions.

A comparison of how the two manuals are cited for different purposes shows that the citations to the Complete Manual unfold as more technical than those to the Archaeological Site Manual. The Complete Manual seems to work as a handbook that helps to appreciate and understand ‘the complexity of archaeology’ (Spoerl 1982, p. 249) both within and outside of the discipline but the specific procedures need to be integrated into the situation-specific research strategies rather than considered as a checklist of a series of technical procedures (Spoerl 1982). Even if it has been cited as a description of regular practices in field archaeology and what is generally known about archaeology, as the findings show, the Complete Manual is cited in fieldwork contexts primarily as a technical reference. This supports earlier remarks that even if Joukowsky’s work undoubtedly belongs to the canon of field manuals, the comprehensive approach described in the Complete Manual has never been embraced to a comparable extent in its entirety (Thorpe 2012).

Similarly, even if the analysis shows that the Archaeological Site Manual also has readers around the world from the UK to France, Italy (Collis 2011), Sweden (Dell’Unto et al. 2017), Kazakhstan (Dawkes and Jorayev 2015) and the Philippines (Barretto-Tesoro, 2013), the specific steps described in the text have not become a global standard (Aitchison 2017). However, in contrast to the Complete Manual, the Archaeological Site Manual appears to have emerged as a popular reference of the idea of the single-context method – especially in contexts where it is not supposed to be the norm. It is used to establish a relationship with the epistemic framework of that particular approach (cf. Hyland 2013), and (literally) ’used’ as a monographic reference of an entire fieldwork approach whereas the citations to the Complete Manual tended to focus on specific techniques and details rather than to Joukowsky’s approach as a whole.

Citations as paradata

A major problem with assuming that citations to methods literature can function as paradata is the potential discrepancy between cited works and the actual practices. Even if a particular manual would be representative of local practice, it is not necessarily followed to the point. The varying degrees of expressing conformance to the texts echo historical studies of archaeological recording practices (e.g., Pavel 2010), and how ideals, techniques and details of archaeological documentation practices can have a tendency to spread much wider than complete sets of procedures.

Similarly to how the earlier literature criticises that the methods described in the Archaeological Site Manual (Davies and Parker 2016) or other manuals (Sellers 1973) are seldom followed to the point, the analysed citations to the Archaeological Site Manual and the Complete Manual evince a lack of total compliance. In the studied material, citations are used to refer to general principles and specific procedures (e.g., [A17], [A20]) that were applied or adapted in a local context to form a creolised version of the line of action described in the literature. Both the unspecific blanket citations and the selective use of specific pointers to particular techniques and procedures suggests what Sellers describes as developing alternative field practices ‘to get their work done and reach their objectives’ (Sellers 1973, p. 140).

Instead of indicating compliance with a set of canonical instructions (Amerine and Bilmes 1990), the specificity and style of citations insinuate that the archaeologists were improvising and engaging in ’metagaming’ (Huvila 2013) their work as it is represented in the formal descriptions. This is not surprising considering the practical nature of archaeological knowing and how learning archaeology by digging (Dyson 1981) on the field has been emphasised as the only way to develop ’a sense of judgment, which is the key to conducting successful archaeological experiments’ (Straus 1981, p. 106). As Dyson suggests, even if manuals have their place and can serve as a convenient repository of appropriate references, they are only the beginning of a complex archaeological learning experience (Dyson 1981).

Another problem with using citations as descriptive paradata is that citations and citing change over time. As Wouters (2014) remarks, they evolve as a part of the evolution of scholarly knowledge infrastructures. Still, a citation marks awareness of a particular work, and by proxy, a method and its significance, even if the practices described in the literature and applied in practice would differ from each other. From this perspective, the citations should not be discarded directly but used together with other indicators that ’have different meanings in the social and/or intellectual organization of the sciences’ (Leydesdorff 2011, p. 73) such as the country and region where an investigation has been carried out, where the excavating team and its leaders have received their education, where the results are published (e.g., a particular publishing outlet, or broader informational genre) and for what type of an audience, and what directives (including contracts) have guided the work.

Even if using citations as paradata has certain limitations, the results do still suggest that they have potential to reduce the need for and to complement manually created paradata. In the functional categories 1-3 and to a certain extent in 4 (Table 4) citations are indicative of work procedures, their theoretical and procedural underpinnings and how they are conceptualised by the authors of the citing texts. It is unlikely that the manuals have been followed to the point. Nonetheless, it is within reason to assume that the presence of citations and formulations suggesting compliance, contrast, and influence all point to a specific significance and exceptionality of the cited text and approach.

According to Ingwersen’s theory of polyrepresentation, in information retrieval, the overlap of multiple different structures of representation, for instance, user-written search terms and index terms, keywords in full-text or citations, that point in the same direction can be considered as an indication that a particular text is relevant for the user (Ingwersen 1994).

Building on the original proposal, it has been suggested that polyrepresentation could indicate the presence of a common scope of relevance (Huvila 2016). From this perspective, a repeated co-occurrence of citations to a specific field manual together with other cognitively (using Ingwersen’s term) diverse evidence that suggests particular ways how the fieldwork was conducted (or is suggested to be conducted) might point towards a certain modus operandi (how things are done) and a related documentation ideal (cf. Börjesson 2016) that unfolds a lieu of ’practical theory’ of how things should be done in a perfect world.

Considering the variation and complexity of archaeological practices, it is likely that the possibilities to recognise functions of citations or the specifics of how the cited works have been used are limited. To a certain extent, the citations that can be classified into functional categories 1, 3 and 4 (Table 4) might have some potential in this respect. They all refer to specific concepts and activities with direct relevance to the citing text and therefore can provide cues of its both practical and theoretical underpinnings i.e., (to a certain degree) what was done and (especially) how the doing and its results were conceptualised. Also, the citations to complete manuals typically cited for work procedures (e.g., Archaeological Site Manual) and specific sections in other works (e.g., in quadrant excavation method, [C40] or pottery analysis [C33] in the Complete Manual) cited for the same purpose seem promising in this sense. However, rather than to function as an independent instance of paradata, it is conceivable that the principal value of citations is likely to be in providing useful complementary information to automatically or manually compiled human-readable process summaries at least as long as the current citation practices remain prevalent.

Conclusions

The main conclusion of this study is that citations to methodological texts can serve several different purposes and by functioning as a source of complementary paradata, provide multiple insights into archaeological practices. Methods literature is cited to clarify specific procedures and actions, understand what is considered conventional and extraordinary, to elucidate work processes in broader terms, and to explain concepts and what is common disciplinary knowledge. Even if the literature use is beyond doubt indicative of work procedures, a citation to a method cannot necessarily be considered direct evidence of what was done in reality (cf. Plutniak and Aguera 2013).

In contrast, a cited work can be expected to play a particular role as a point of reference or inspiration in the described activity and how the citing author prefers to frame it. The context of the citation in the text and how it is formulated (whether a method is ’used’, ’employed’ or if it serves as a ’basis’) can provide further cues of the function of the citation. Even if the present findings do not directly support the idea that citations, their specific function and relation to the actual work process could be harvested and classified automatically for producing structured descriptions or paradata on research work, they show that the citations have the potential to complement the polyrepresentative model of research practices. As such, if combined with other indicators, a particular type of citation to a specific text can be indicative of specific patterns of work, and function as a valuable additional piece of paradata on the reported research work.

Acknowledgements

We would like to thank two anonymous reviewers for their invaluable comments on earlier versions of this text. This work has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme grant agreement No 818210 as a part of the project CApturing Paradata for documenTing data creation and Use for the REsearch of the future (CAPTURE).

About the authors

Isto Huvila is Professor in Information Studies at the Department of ALM (Archives, Libraries, Museums), Uppsala University in Sweden. Huvila chaired the recently closed COST Action ARKWORK and is directing the ERC funded research project CAPTURE. His primary areas of research include information and knowledge management, information work, knowledge organization, documentation and social and participatory information practices. He can be contacted at isto.huvila@abm.uu.se.
Dr. Lisa Andersson works as a researcher at the Department of ALM at Uppsala University in Sweden. Her research focuses on research information including research information management systems, data descriptions, data publishing and use. She can be contacted at lisa.borjesson@abm.uu.se.

Dr. Olle Sköld is a senior lecturer at the Department of ALM and the director of Uppsala University’s Master’s Programme in Digital Humanities. His research is characterised by a broad interest in the ALM field, research data creation and use and digital humanities. He can be contacted at olle.skold@abm.uu.se.

References


How to cite this paper

Huvila, I., Andersson, L., & Sköld, O. (2022). Citing methods literature: citations to field manuals as paradata on archaeological fieldwork. Information Research, 27(3), paper 941. Retrieved from http://InformationR.net/ir/27-3/paper941.html (Archived by the Internet Archive at https://bit.ly/3eHrACk) https://doi.org/10.47989/irpaper941

Appendix 1: Analysed publications

The references include the code (e.g., A1) used in the article text, authors, year, title, publication title, volume, issue, pages, publisher, an eventual DOI identifier, and either ASM to refer to the Archaeological Site Manual or CMF to the Complete Manual of Field Archaeology.



Check for citations, using Google Scholar