Skip to main content

Chapter 25 | Analysing Documents

Handbook
Download PDF
Analysing Documents
AmandaCoffey

Most qualitative research takes place in settings that are ‘documented’ in various ways. That is to say, many social settings are self-documenting and there is considerable methodological potential to study the documentary realities of social worlds. What can be included as a ‘document’ in social research covers a potentially broad spectrum of materials, both textual and otherwise. There are, of course, ‘official’ records of various kinds – organizational and ‘state’ documents designed as records of action and activity (such as large data sets and public records). There are also everyday documents of organizations and lives – notes, memoranda, case records, email threads and so forth; semi-public or routine documents that are at the heart of everyday social practice. There are also private papers of various kinds that we can also treat as documentary data or evidence – for example, diaries, testimony, letters and cards. But we can go further than that in defining what might count as a document for social research purposes – maps, photographs (see Banks, Chapter 27, this volume), newspaper reports (see Hodgetts and Chamberlain, Chapter 26, this volume), autobiographies, novels, advertisements and paintings can all be considered documents that tell of settings, organizations, times and lives. In contemporary times, documentary materials also now encompass a wide range of technological, digital and social media – for example, email conversations (see Marotzki et al., Chapter 31, and Kozinets et al., Chapter 18, this volume), SMS text messaging, websites, social networking sites and hypermedia. All kinds of documents are routinely written, produced, read, consumed, stored, circulated and used in everyday social life and practice. Indeed, documents can be thought of as the ‘physical traces’ of social settings (Webb et al., 2000); as data or evidence of the ways in which individuals, groups, social settings, institutions and organizations represent and account for themselves. Documents provide a mechanism and vehicle for understanding and making sense of social and organization practices or, as May describes, ‘documents, read as the sedimentations of social practices, have the potential to inform and structure the decisions which people make on a daily and longer-term basis: they also constitute particular readings of social events’ (2001: 176).

Qualitative researchers have not always recognized the analytical potential of studying written documents and textual recordings as research data; indeed in many qualitative accounts of social settings there is often little or no mention of the documentary realities of those settings. Of course, documentary research has a long history within social science more generally. Many early sociological thinkers (including for example Marx, Weber and Durkheim) gathered and analysed documents as part of their empirical and theoretical practice, as did many later scholars of social science (such as Foucault and Bourdieu). In this chapter I argue that qualitative research can be enriched by a careful and critical attention to the gathering and analysis of documents, of various kinds, in various modes and through various media. This might usefully include a close reading of documents themselves, but also include developing an understanding of the ways in which documents are authored, produced, used and consumed. Thus this chapter explores the potential of documents as social research data, and considers some of the methodological and technical aspects of analysing documents as a way of understanding social practice.

Defining Documentary Data

Documents are pervasive in organizational and social life – consider, for example, a typical or ideal-type organization, such as a private business, a public sector organization, a school, hospital, car manufacturer, university or accountancy firm. It is hard to imagine such a modern kind of social organization without recourse to its routine documentation – administrators, managers, accountants, lawyers, civil servants, managers and practitioners are all, routinely and extensively, involved in the production and consumption of everyday documents and texts as part of their daily work. If we wish to understand how organizations and social settings operate and how people work with/in them, then it makes sense to consider social actors' various activities as authors and audiences of documents. And, of course, there are differing levels of formality and informality in the production and intention of such documents, which can include official brochures, records and minutes, but also other physical traces such as email threads and conversations. Textual records also embody individual actions, interactions and encounters within social settings. ‘People-processing professions’ for example, such as medicine, nursing, teaching or social work, routinely compile documents of professional-client interactions (case notes, medical records, care plans, school reports). These written records can be used to inform future action, and are themselves drawn upon in the more formal recording (and documentary) mechanisms of official statistics, performance indicators, efficiency league tables and similar constructs.

Qualitative explorations of a range of social settings have included some attention to the production and consumption of such documentary data. Examples include studies that have incorporated analyses of school reports (Woods, 1979), medical records (Rees, 1981), classifications of causes of death (Prior, 1985), coroners' records (Fincham et al., 2011) and health visitors' case records (Dingwall, 1977). Indeed there are many research questions and settings that arguably cannot be investigated adequately without reference to the production and use of documentary materials. For example, it would be impossible to study the everyday work and occupational culture of a profession such as actuaries without addressing the construction and interpretation of documentary artefacts such as the life-table (Prior and Bloor, 1993); and difficult to understand the modern university and higher education more generally without studying their documentary realities – for example, prospectuses, committee minutes, accounts of research performance and student feedback are all part of the ways in which universities ‘do’ their work (Atkinson and Coffey, 2010). As Bloomfield and Vurdabakis (1994) point out, textual communicative practices are a vital way in which organizations constitute ‘reality’ and the forms of knowledge appropriate to it. But as I have noted above, formal organizational documents form only part of the documentary reality in which and through which organizations, social settings and lives are represented, lived and told.

Like organizations and other social settings, individual and collective lives are also marked by, with and through documents of various kinds. Alongside organizational and professional records of lives, for example as told through medical notes or school reports, are a myriad of other kinds of ‘life document’. Thus documents and their analyses can also be utilized to understand personal lives and experiences, and to place biography within and in relation to social context. As Plummer reminds us:

the world is crammed full of human, personal documents. People keep diaries, send letters, make quilts, dash off memos, compose auto/biographies, construct web sites, scrawl graffiti, publish their memoirs, write letters, compose CVs, leave suicide notes, film video diaries, inscribe memorials on tombstones, shoot films, paint pictures, make tapes and try to record their personal dreams. … They are all in the broadest sense ‘documents of life’. (2001: 17)

Plummer contends that social science should treat such life documents seriously both as resources for understanding complex social life and as topics of analysis in their own right. Just as organizations and social settings have documentary realities so, too, do individuals, families and other social groups. Thus paying analytic attention to documents can shed light on the intimate and the personal as well as the public, organizational and corporate.

It is also helpful to make a distinction between documents that are ‘found’ in the process of social research (documents that exist prior to, and not because of, the research) and documents that are ‘made’ as part of the research (produced explicitly for the research to hand). So, for example, during social research we might gather together documents that pre-exist in a setting, that are there ‘anyway’ as part of the everyday order of the setting, but we might also ask participants to keep a diary or paint a picture or construct a webpage as part of the research process itself. This distinction between unsolicited and solicited documents (Scott, 1990) is useful in helping to explore the social context and circumstances of documentary production. Such a distinction, however, should not distract us from recognizing the characteristics (and therefore the analytic potential) of all documents as constituting social science data or evidence.

Documents, then, are literary, textual or visual devices that enable information to be shared and ‘stories’ to be presented. Thus, all documents are, in that sense, artefacts that are created for a particular purpose, crafted according to social convention to serve a function of sorts. It is this social production (and indeed consumption) of documents that gives them analytical affordance. At the same time it means that we need to be quite clear about what documents can and cannot be used for in social research. Documents are ‘social facts’, in that they are produced, shared and used in socially organized ways. They are versions of reality, scripted according to various kinds of convention, with a particular purpose in mind. This is equally true of the most public of record or the most private of diary. Documents construct particular kinds of representations using particular kinds of textual (and often, too, non-textual) convention. Documents should not be seen as replacements for other kinds of data. We cannot, for instance, learn through written records alone how an organization actually operates day by day. Similarly, we cannot treat documents – however official or otherwise – as firm evidence of what they report. This observation has been made repeatedly about data from official sources, such as statistics on crime, suicide, health, death and educational outcomes (Cicourel and Kitsuse, 1963; Sudnow, 1968; Atkinson, 1978; Roberts, 2003; Maguire, 1994; Macdonald, 2008; Scourfield et al., 2012). This understanding means we should always be reflexive in how we treat documents as social data.

The recognition of the existence of documents as social facts (or constructs) alerts us to the necessity to treat them seriously in social research (Prior, 2008). Documents can tell us a lot about a social setting or an individual life. However, we have to approach the analysis of documents for what they are and for what they are used to accomplish. This means paying attention to the knowledge that documents ‘contain’ about a setting, but also examining their role and place in settings, the cultural values attached to them, their distinctive types and forms. The analysis of such documentary evidence can form an important part of broader ethnographic studies of everyday life; documentary analysis may also be employed as the main method for qualitative research in its own right (Prior, 2003). In either event it is important to establish a methodological framework for documentary analysis. In the remainder of this chapter I outline some strategies for approaching the qualitative analysis of documents. This is certainly not intended as a comprehensive review of all analytical strategies or approaches (see Silverman, 2006; Prior, 2011), nor a technical manual of prescriptive techniques. Rather I introduce approaches to the systematic analysis of documentary data and discuss some contexts of their use.

Analytical Strategies

A useful starting position for the analysis of documents in social research is that documents are socially defined, produced and consumed. Thus in looking at documents analytically, we need to examine the processes of production and consumption – be they technical, linguistic or conceptual – as well as the content contained within documents. In that sense we might think of documents as resources (i.e. as information repositories, telling us about a setting, an organization, an event, or a person), but also as artefacts for exploration in their own right. A document in and of itself can tell us something about the social setting. If we understand documents as accomplishments, as products with purpose, then it naturally follows that analysis should seek to locate documents within their social as well as textual context. Documents then are resources to be ‘mined’ but also topics to be studied.

In keeping with most other kinds of qualitative data analysis it is entirely possible and appropriate to undertake a thematic analysis of documentary data. Following on from the analytical conventions of content analysis (see Schreier, Chapter 12, this volume) or ‘code-and-retrieve’ (Seidel and Kelle, 1995; see Gibbs, Chapter 19, this volume), documents can be read in terms of their content meaning. Practical strategies for this kind of analysis can vary from almost quantitative measures (counting instances, for example) through to the kinds of thematic analysis supported by Straussian approaches to coding and grounded theorising (Strauss, 1987; Strauss and Corbin, 1990; also see Thornberg and Charmaz, Chapter 11, this volume). Thus we can approach documents in terms of the frequency of words, phrases or other elements or characteristics. We can index and code data to identify key themes and thus generate theoretical categories and identify patterns. It is not my intention here to provide detailed instruction in or critique of thematic or content analysis. There are good overviews in this volume and elsewhere on such approaches. Suffice to say that such approaches to the analysis of documents focus on the product (the document as information resource or vessel), and not on the processes of production per se. For many researchers who use documents in social research this approach to analysis may be entirely appropriate, particularly if documents are being used primarily to provide background information or context. However, for our purposes in this chapter I want to argue that the analytical potential of documents as social data is in also understanding the circumstances of production and the receiving (reading) of the document as an artefact of the setting under study. Thus it is important that we bring to bear analytical strategies that enable the meaning-making of documents to be subjected to critical scrutiny – analytical approaches that recognize documents themselves as ways in which social actors make sense of social worlds. Hence we also need to be concerned with intended meanings (and thus with the authorship and function of documents) and received meanings (recognizing the importance of readership and audience, and the ways in which documents are interpreted by intended and unintended audiences). Documents, as social artefacts, have narrative structures and are imbued with cultural ways of telling (see Esin et al., Chapter 14, and Winter, Chapter 17, this volume). They draw upon and conform to various genres, in terms of style, structure and language. They employ visual signs, literary devices and other symbols to present and display meaning. Documents are also rarely, if ever, produced and read in isolation from other documents. In adopting this more semiotic approach to documents we can explore relationships and meanings within a text and in relation to other texts. It is helpful here to distinguish between what documents ‘look like’ (i.e. language and form), what they ‘do’ (i.e. purpose or function) and how they are related (i.e. intertextuality between documents).

Language and Form

Documentary constructions of social reality – documents – depend upon particular uses of language and form. Documents will constitute and conform to particular genres with specific styles and conventions. These are often marked by quite distinctive use of language and structure. Documents may use specialized language (which might be referred to as a linguistic register) associated with particular domains of everyday life, and will draw on culturally recognized ways of telling (what we might refer to as narrative structures – see Esin et al., Chapter 14, this volume). Occupations, for example, often have distinctive language (with specialized vocabularies and narrative forms), as do particular kinds of organization or cultural activity. We can therefore learn a lot about such settings by paying particular attention to these structures, registers and forms. We can often recognize what sort of document we are looking at simply from its distinctive use of language or the way in which it is presented. We can, for example, recognize a theatre review, or a university prospectus, or a personal diary entry from their characteristic styles – both linguistically and stylistically. At an elementary level we can recognize that ‘official’ or public reports are crafted in language that differs from everyday, spoken, language use (see Toerien, Chapter 22, this volume). Similarly we can distinguish between the register and form we might use to draw up a shopping list and the register and form we might use in an obituary or other similar kind of semi-public announcement. Indeed, culturally understood registers and narrative forms are precisely the kinds of devices that are used to construct, and make distinctive or special, modes of documentary representation.

In approaching the analysis of documents in this way, it is helpful to adopt an interpretative standpoint (see Willig, Chapter 10, this volume). The initial task is to pay close attention to the question of how documents are constructed as distinctive kinds of artefacts or productions. It is therefore appropriate to pay close attention to the textual organization of documents, and the semiotic and narrative qualities of the materials within the documents. Important analytic questions in this context are: What kind of reality is the document creating? How is the document accomplishing that task? In undertaking such analyses we can draw on a repertoire of analytical techniques and resources – drawn for example from formal approaches to the study of language and structure – such as narrative analysis (see Esin et al., Chapter 14, this volume), discourse analysis (see Willig, Chapter 23, this volume) and semiotics. For our purposes here it is most appropriate to outline some of the general features of such approaches in terms of their specific application to documentary sense-making. When we look at a document, therefore, we can ask questions about the role and use of language as well as other discursive practices. Many documents will display a distinctive register: that is, a distinctive and specialized use of language associated with a particular context or domain. It might be associated with a particular group, occupation, activity or organization, or with a distinctive kind of intellectual field, or an esoteric pursuit. It implies a general feature of language in social life: distinctive uses of language (written and spoken) are associated with, and are constitutive of, specific social contexts. Thus in developing an interpretative understanding of documents as topics to be studied, we are interested in language, words and phrases, and also in the systems of convention that guide the ordering and structure of the text. Hence we are interested in the ways in which the messages (the meanings or social realities) are produced and articulated by an author or authors to an audience (or audiences). The look and feel of a document can thus tell us something about the social setting or social practice under consideration.

Studying the Function of Documents

As well as analysing the form and content of documents it may also be useful to consider the ways in which documents function and have function. That is, we can explore the ways in which documents are used and have use in everyday life and social context. We might usefully make a distinction here between intended and received messages or functions. What purpose is the document intended to serve by the author or authors, and how is the document read, understood and used by audiences or readers? As well as asking analytical questions in relation to ‘how’ the document is ‘constructed’, it is therefore also appropriate to ask how documents ‘function’ in everyday activities and thus how they help to construct everyday realities in their procurement and usage. In other words, what is the document doing?

A way of approaching this kind of question is to think in terms of what the linguistic philosopher John L. Austin described as speech acts. This refers to the fact that language does not merely describe events or states of affairs. It also creates or performs them. When you make a promise or utter a threat, you are not using language to describe something else; you are using the language to accomplish the act itself. In just the same way documents can be seen not (just) as describing an event, organization, emotion or state of affairs, but also as helping to create them (see Toerien, Chapter 22, this volume). In doing so documents deploy discursive or rhetorical devices – to create plausible accounts and to construct believable versions of reality; in other words, documents persuade. Rhetoric in this context is not being referred to in a negative way, nor does it imply wrongdoing. Rather it is an acknowledgement that a document can be conceptualized as an act of persuasion – and as such, and in line with any other act of persuasion, depends on rhetorical devices to describe, explain and justify. Moreover, the social actors who write documents and the social actors who read (and evaluate) them bring to bear their knowledge – often tacit – of the conventions that go into their production and reception. Writers develop and display a working knowledge of the register(s) of their own professions, or organizational setting or cultural activity or intimate life. Readers, too, bring to bear a repertoire of conventional understanding to interpret and make sense of documents. Indeed, the phrase ‘making sense’ is especially apposite when we think of the ways in which documents are interpreted and come to be understood. Making sense is a socially organized activity of interpreting documents. This suggests that the interpretation of documents is an active process. Thus documentary sources do not transparently describe or reveal goings on or states of affairs. They help to construct and display them; and that construction requires the active participation of readers as well as writers. Reading documents and making sense of their contents requires readers to bring their own assumptions and understandings to bear. The culturally competent reader will ‘know’ how to use documentary sources to create the organizational reality they purport to describe. Knowledgeable readers will know something of the cultural features of the organization or cultural setting, and thus will be well placed to use what Mannheim called – appropriately – the ‘documentary method’ (see Bohnsack, Chapter 15, this volume). That is, the text is used to furnish indications or provide physical traces of what the reader interprets or understands as the social reality. Thus readers will read into the text what might reasonably be assumed to be the case, given a shared stock of tacit knowledge about this organization or social setting or intimate life and how they typically function. Typical cases are interpreted in terms of their typical manifestations, and their typical rhetorical representations. Thus documentary realities are built, consolidated and confirmed.

This focus on the functions of documents can be usefully developed and illustrated by exploring the ways in which documents do various kinds of ‘work’. Consider for example what a school or university prospectus is setting out to achieve, or a social work case report, school report or suicide note. We might usefully use terms such as ‘to persuade’, ‘to validate’, ‘to justify’. A classic example of this is Garfinkel's seminal commentary on the analysis of clinical hospital records (Garfinkel, 1967; also see Eberle, Chapter 13, and Bohnsack, Chapter 15, this volume). Garfinkel's actual starting point was the use made of clinical records by social researchers, who appeared to be able to make practical use of clinical records as data in order to make sense of how hospital clinics work. Garfinkel argued that the researchers were only able to make sense of those records by ‘reading into’ them what they had already come to understand about clinics as particular kinds of organization. The clinical records themselves were messy documents, but culturally competent readers of these records – for example, clinicians, administrators or indeed researchers – were able to make sense of them by bringing to bear prior assumptions and cultural understandings. In this way documents can be seen to presuppose a community of readers and writers who share a common stock of knowledge and taken-for-granted assumptions. The analysis of documents can therefore examine those cultural and organizational features that are implicitly invoked when records and documents are produced and used. In invoking the documentary method in this context we are returning again to an earlier claim – whereby everyday social actors use a generally understood methodology to interpret documents as ‘physical traces’. In this sense documents are signs or symbols through which social actors infer underlying patterns or states of affairs, and to which social actors add and embellish with their common-sense knowledge.

Intertextuality and Authority

Documents do not construct domains of documentary reality as individual, separate activities. Documents refer – however tangentially – to other realities and domains. Moreover, documents refer and are connected to other documents. This is especially the case for particular kinds of organizational settings and their systems of accountability via documentation, though can be just as applicable to other kinds of social or cultural settings. The analysis of documents and documentary reality must, therefore, look beyond individual texts as artefacts, and also ask how (and in what ways) they are related. That is, we can recognize that, like any system of symbols and signs, documents often make sense because they have relationships with and to other documents. Thus we can pay attention to the intertextuality of documents: that is, their relational qualities and what these can reveal about the setting under investigation. The concept of audit is useful here. If we consider the basic mechanics of audit, then it starts to become quite easy to grasp the point and significance of systematic relations between documents. One of the root metaphors of an audit is that of the audit trail. Conventionally defined audits, for example of firms and organizations, carried out by accountants or auditors, place great emphasis on the audit trail. Audit trails trace each document and statement presented in organizational accounts to other documents contained in the audit file (the preparation of papers for an audit). There is an assumption that references can and should be made to other documents; indeed it is through these references and trails that decisions, accounts and everyday practices are documented and justified. An auditor's task is to establish the extent of these relationships and intertextualities, in order to account for and make sense of the process and practice of the organization. These relationships between documents are usually based on elementary – but significant – principles. They include the principles of sequence and hierarchy, which form part of the constitutive machinery whereby organizations produce and reproduce themselves. From a general analytic perspective, therefore, we can see that documentary realities do not rely on particular documents mirroring and reflecting a social reality. Rather, we can think of a semi-autonomous domain of documentary reality, in which documents reflect and refer to other documents.

We can analyse such documentary realities in various ways. The term ‘intertextuality’ derives from contemporary literary criticism and is used, in that context, to refer to that fact that literary texts (such as novels) are rarely free-standing pieces, nor do they just or only refer to a fictional world. Literary texts, in their very nature, refer to other texts, albeit sometimes implicitly. This can include other texts of the same genre, or other kinds of textual product (such as journalism or biography). Texts can therefore be analysed in terms of these intertextual relationships, tracing the dimensions of similarity, comparison, contrast and difference. We can examine, for example, how conventional formats are shared between texts, and thus how they construct a uniform style. We can note how texts are linked as sequences of documents, and seek to understand the nature and meaning of those sequences. We can also examine how relations between documents reveal temporality – documents can often provide a temporal sequence or structure to the organization or setting or life, though will not necessarily describe the passage of time as experienced as an everyday phenomenon by the individual actor(s) concerned. Documentary sources can hasten time, slow time, ‘trouble’ time and even suppress time – lifting events out of the flow of lived experience, and recording them in decontextualized language and formats of a documentary record. Intertextuality thus alerts us to the fact that documents are usually part of wider systems of distribution and exchange. Documents circulate through social networks and organizations, and in doing so help actively to construct those networks and organizations.

Documents move, flow and exchange because they can be used to decontextualize and recontextualize events. We can transform things, events, activities and lives by incorporating them into texts. By writing an event, activity or life in a documentary format, we translate them from the specific and the local, and make of them ‘facts’ and ‘records’ which take on an independent existence. Some texts become ‘official’, and can become ‘proof’ of events and roles. This is an argument that was made by Latour and Woolgar (1986) in relation to the production of scientific facts and findings through the production of scientific papers. Latour and Woolgar suggested that scientific ‘facts’, represented in documentary form, achieve an independence of their original site of production (e.g. the peopled and relational research group or laboratory), and take on an independent existence. In other words, the academic paper recounting or claiming the scientific discovery actually removes that discovery from the process of discovery (with the people, personalities, luck, judgement, risk and failure that might have been involved). A similar observation can be made about the routine circulation of minutes of meetings, which those of us who regularly attend meetings will know are only ever a partial and scripted version of events. The professional audit report of a business organization serves a similar purpose. The audit report becomes the documentary reality, superseding other files, records and memories.

One should also note here questions of authorship and readership in the analysis of documents and relations between documents. It is important to address authorship (whether actual or implied) and readership (again whether actual or implied) if one is to understand the system of production, exchange and consumption of documentary materials. Documents are usually ‘recipient designed’ (see Toerien, Chapter 22, this volume). That is, they are produced with readers in mind and will therefore reflect implicit assumptions about who will be the reader. This implied reader does not, of course, have to be an actual individual person. The implied recipient can correspond to what George Herbert Mead referred to as the ‘generalized other’. Indeed it is a basic tenet of interactionist social analysis that social actors monitor and shape their actions in the light of generalized others' imputed responses and evaluations. When a document is created it is often in the light of the kind of readership expected or being written for. And as analysts of documents we need to be attuned to addressing this. Equally, while it is self-evident that a person or a group must actually author documents (since they do not write themselves), that does not always imply a social recognition of ‘authorship’. Indeed, it is part of the facticity of many documents (particularly but not exclusively ‘official’ documents) that they are not identifiably the work of an individual author. Anonymity itself can be part of the production of documentary reality. For example, while there may be an implied ‘ownership’ of a document – such as the originating administrator or department – official materials usually do not have visible social actors expressing opinions. It is important therefore to inspect texts for indications of authorship, or its absence. In that sense, too, we can look for how documents claim whatever authority may be attributed to them. In simple terms ‘texts must be studied as socially situated products’ (Scott, 1990: 34), with socially situated authors and readers.

An Exemplar – Documentary Analysis in Practice

This chapter has argued that, in analysing documents as social research data, we should be mindful, not only of what the documents might contain in terms of information or content, but also of how they are structured and the functions to which they are (or might be) put. Moreover, I have noted that documents are rarely present in isolation from other documents, and explorations of relationships between documents can be analytically fruitful. In this section I turn to an exemplar that draws on a multifaceted approach to the analysis of documents in social research. A recent example of the qualitative analysis of documents in social research is the work of Scourfield and colleagues on suicide and the sociological autopsy (see Scourfield et al., 2012; Fincham et al., 2011). This choice of exemplar is apposite of course, as official records (documents) of suicide were used by Durkheim in what is still heralded as a watershed development within sociology (Durkheim, 2002 [1897]). Drawing on 100 suicide case files from a UK coroner's office, Scourfield and his colleagues set out to develop a sociological approach to the study of suicide, examining individual suicidal lives in their broader social context, and with recourse to the social construction of knowledge. The project focused on the myriad of documents contained in the coroner's files, thus drawing on a rich diversity of documentary data, including: ‘forms filled out by coroner; scribbles by the coroner on file wallets; police statements from witnesses and significant others; forensic pathology reports; medical letters and reports, especially psychiatric ones; suicide notes; mobile phone records; photographs of corpses; letters to the coroner and newspaper clippings’ (Scourfield et al., 2012: 467). Aside from noting the important role of the coroner's office in putting together or constructing these interrelated files of evidence, Scourfield and his colleagues also reveal the ways in which these documentary artefacts can be analysed as sites for the creation of identity; ‘we are concerned with relationships: how they extend into documents and how they constitute different kinds of persons and identities, during someone's lifetime and beyond’ (Fincham et al., 2011: 65). The project explored the ways in which documents as ‘evidence’ are constructed by parties to the suicidal life (and death), by those living and those now dead, and the ways in which documentary data can be used to explore how knowledge about suicide is constructed by professionals, families and publics.

In terms of our interest here in the analysis of documents in qualitative research practice, this project is an excellent exemplar in a number of important ways. First, as highlighted above, the project demonstrates some of the sheer richness and diversity of documentary possibilities for social research, drawing as it does on many different kinds of documentary artefact, all present within a single file. Second, the project demonstrates the ways in which documents can be analysed in terms of thematic content, to reveal patterns, sequences and absences. The project was unusual in that it sought to draw on a qualitative thematic analysis of whole cases (rather than extracts), which also enabled the generation of quantitative analyses that went beyond the individual case. Code-and-retrieve was used to analyse across as well as within different genres of documents, as well as providing some quantification of the qualitative data. Third, the project focuses particular attention on the form and structure of documents and the analytical possibilities of understanding such characteristics and qualities between and across different kinds of documents. For example, the project examined official and professional reports of the living person (e.g. as patient or client) and the dead person (as body or corpse). The project also examined witness statements:

[S]uch statements were not verbatim records of the interview. Rather the events recounted in them had been consecutively ordered and the narrative itself had been shaped by the need to be concise and to the point. In order to achieve this aim, the statements were drafted first by a police officer using institutional conventions of language and content. These documents' hybrid production process and the need to fulfil institutional and legal requirements, lent the accounts a shared appearance. (Fincham et al., 201 1: 76)

These contrast sharply to individual suicide notes – which ‘vary greatly in their form and content’, are the only documents written by the deceased in the file, and the only documents ‘produced before the deceased died’ (Fincham et al., 2011: 80–1). Moreover, this project reveals the ways in which documents can be analysed in terms of the functions, intended or otherwise, that documents can perform. For example, they reveal the ways in which documents justify decisions or verdicts, display professional hierarchies or exercise agency (in this case, after death).

The coroner's file on the suicidal life and death, and the analyses of multiple cases, also allow for a sociological understanding of intertextuality in practice. The diversity of documents, and the exploration of their pre-and post-death production, provide an opportunity to explore the relationships between and within documents – within and across cases. In doing so the project provides ways of looking at the complex documentary realities of suicidal lives and action, and the ways in which these realities are imbued with socially situated meaning and authorship. And in doing so the project articulated and worked with the opportunities heralded by documentary analysis, as well as noting some of the limitations of seeking to understand social worlds through documents. As the authors note, ‘we accept that evidence about suicide, including documents in coroners’ files, is produced under specific circumstances which affect how it should be read, but maintain that such evidence aims to establish something about an externally verifiable social world' (Fincham et al., 2011: 52–3). In the final section of this chapter I elaborate a little further on some of the challenges of documentary analysis, as well as looking towards further development in this area.

The Limitations of, and New Possibilities for Working with Documents

Working with documents in social research means paying careful attention to the ways in which documents are classified and conceptualized. What counts as a document, and what meaning we attach to a document, is a complex and multifaceted task. There are various ways in which we might usefully classify documents for social research purposes, and such classifications help to distinguish between kinds of documents, provide opportunities for thinking across documents and enable us to recognize some of the possible limitations of documents. For example, Scott (1990) makes the distinction between primary, secondary and tertiary documents, and in some contexts this might be useful. Primary documents are materials produced by those experiencing events or settings first hand, secondary documents are constructed as a representation of an event (by others), and tertiary documents include such things as catalogues, references and the grey literature. Another useful classification might be private and public documents – distinguishing aspects of the intended purpose and function of documents, though not necessarily accessibility for social research. And, as noted earlier, we can also think about the ways in which documents exist regardless of a particular research project, or are constructed as part of a research project. Such classifications might tell us very little about the authority and authenticity of documents, nor their accessibility for the social researcher. But they do alert us to the ways in which authority and authenticity are claimed in documentary form. In all cases what is key to the analysis of documents for social research is that documents do not necessarily consist of descriptions of the social world that can be used directly as evidence of that social world. All documentary accounts are just that – a constructed account rather than necessarily an ‘accurate’ portrayal of complex social reality. Documents construct their own kinds of reality – of people, places, organizations and other social settings. It is therefore important that we approach documents as texts and as representative of the practical accomplishments involved in their production. That is, documents are resources and topics for investigation – and produced according to conventions that are themselves part of a documentary reality. Thus it is important that we ask appropriate questions about documents and what they can and cannot reveal about the social world. Rather than ask whether a document offers a ‘true’ account, or whether it can be used as ‘valid’ evidence about a research setting, it is more fruitful to ask questions about the form and function of documents themselves. We should also examine documents for their formal properties. As noted in this chapter, it is important to consider the ways in which documents tell and persuade, and there is analytical affordance in exploring the linguistic registers and rhetorical features of texts as documents of persuasion. In doing so we need to think about documents in relation to their production (authorship) and their consumption (readership), but one should note that in textual terms these are not necessarily just coterminous with the particular individual social actors who write and read. We need to pay close attention to the implied readers, and to the implied claims of, in some cases anonymous, authorship. The analysis of documents requires considerable reflexivity (see May and Perry, Chapter 8, this volume) on the part of the researcher, both in relation to understanding the possibilities and limitations of documents as artefacts and representations of social life, and in the selection (see Rapley, Chapter 4, this volume) of documents for analysis. As we know, documents do not refer transparently to the social world. Their referential value is often in their intertextuality – their relation to other documents or texts. Indeed in many settings and in relation to many events we can identify semi-autonomous domains of texts and documents that refer primarily to one another. A dense network of cross-referencing and shared textual formats can create powerful documentary realities. Indeed some of the limitations of working with documents – the often complex and hidden relations between documents, the selectivity of documentary accounts, the prescriptive and often formulaic structure of documents, the functional purposes to which many kinds of documents are put (intentionally and unintentionally) – are also the same characteristics of documents that make them such a rich source of data for sustained and creative analysis.

A limitation of this chapter has been that I have primarily focused, albeit often implicitly, on the analysis of written texts – documents of words. The approaches I have suggested for analysis – thinking about documents as resources (or repositories of information), as structures (with narrative form and convention), with purpose and function, and in relation to other documents – are equally applicable to other kinds of documents or material artefacts. Thus it is important to acknowledge that the analysis of documents can be developed beyond the confines of the written text. Indeed, as was noted at the beginning of this chapter, documents, as physical traces of social settings, can be thought of in much broader terms, offering considerable potential for innovative research practice. Documents can and do incorporate visual materials that construct and present the social world in pictures as well as words. The analysis of visual materials, including photographs (see Banks, Chapter 27, this volume), moving images (see Mikos, Chapter 28, this volume), maps, drawings and other art practices, is a growing field of scholarly activity within qualitative social science, and there are now both established and increasingly innovative methods for visual analysis. Moreover, in the contemporary digital age, there is considerable scope for developing our understandings of social worlds and social life through scholarly analysis of new and emergent forms of documents. Information technology has created new possibilities for communication and representation, allowing us to think about documentary realities in ever expansive ways. For example, the Internet and World Wide Web, electronic communication, SMS text messaging and social networking sites have broadened the scope and genres of documents potentially available for analysis (see Kozinets et al., Chapter 18, and Marotzki et al., Chapter 31, this volume). And it is increasingly the case that such documents can be seen both as multimodal (encompassing for example text, sound (see Maeder, Chapter 29, this volume) and pictures (see Banks, Chapter 27, this volume) and as dynamic or fluid – that is, documents can be thought of, not as static or fixed representations, but as increasingly shifting and changing. Digital technologies create new possibilities too for authorship, readership and connections between documents, and indeed serve to disrupt some of the documentary distinctions – for example, what counts as public or private, primary or secondary, is complicated by virtual social worlds, working with different ‘realities’ of time and space. In that sense documents and the documentary realities they create are likely to become more rather than less important to the accomplishment and understanding of everyday life and to social research.

Further Reading
Fincham, Benjamin, Langer, Suzanne, Scourfield, Jonathan and Shiner, Michael (2011) Understanding Suicide: A Sociological Autopsy. Basingstoke: Palgrave Macmillan. http://dx.doi.org/10.1057/9780230314078
Plummer, Ken (2001) Documents of Life 2. London: Sage.
Prior, Lindsay (2011) Using Documents and Records in Social Research. London: Sage.
References
Atkinson, J. Maxwell (1978) Discovering Suicide: Studies in the Social Organization of Sudden Death. London: Macmillan.
Atkinson, Paul and Coffey, Amanda (2010) ‘Analysing documentary realities’, in DavidSilverman (ed.), Qualitative Research: Theory, Method and Practice. London: Sage. pp. 56–77.
Bloomfield, Brian P. and Vurdabakis, Theo (1994) ‘Re-presenting technology: IT consultancy reports as textual reality constructions’, Sociology, 28(2): 455–78. http://dx.doi.org/10.1177/0038038594028002006
Cicourel, Aaron and Kitsuse, John (1963) The Educational Decision-Makers. New York: Bobbs-Merrill.
Dingwall, Robert (1977) The Social Organization of Health Visitor Training. London: Croom Helm.
Durkheim, Emile (2002 [1897]) Suicide: A Study in Sociology. London: Routledge. http://dx.doi.org/10.1522/cla.due.sui2
Fincham, Benjamin, Langer, Suzanne, Scourfield, Jonathan and Shiner, Michael (2011) Understanding Suicide: A Sociological Autopsy. Basingstoke: Palgrave Macmillan. http://dx.doi.org/10.1057/9780230314078
Garfinkel, Harold (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Latour, Bruno and Woolgar, Steve (1986) Laboratory Life. Princeton, NJ: Princeton University Press.
Macdonald, Keith (2008) ‘Using documents’, in NigelGilbert (ed.), Researching Social Life. London: Sage. pp. 194–210.
Maguire, Michael (1994) ‘Crime statistics, patterns and trends’, in MichaelMaguire et al. (eds), Oxford Handbook of Criminology. Oxford: Oxford University Press. pp. 322–75.
May, Tim (2001) Social Research: Issues, Methods and Process. Buckingham: Open University Press.
Plummer, Ken (2001) Documents of Life 2. London: Sage.
Prior, Lindsay (1985) ‘Making sense of mortality’, Sociology of Health and Illness, 7(2): 167–90. http://dx.doi.org/10.1111/1467-9566.ep10949063
Prior, Lindsay (2003) Using Documents in Social Research. London; Sage.
Prior, Lindsay (2008) ‘Documents and action’, in PerttiAlastuutari et al. (eds), The Sage Handbook of Social Research Methods. London: Sage. pp. 479–92.
Prior, Lindsay (ed.) (2011) Using Documents and Records in Social Research. London: Sage.
Prior, Lindsay and Bloor, Mick (1993) ‘Why people die: social representations of death and its causes’, Science as Culture, 3 (3): 346–74. http://dx.doi.org/10.1080/09505439309526355
Rees, Colin (1981) ‘Records and hospital routine’, in PaulAtkinson and ChristianHeath (eds), Medical Work: Realities and Routines. Farnborough: Gower. pp. 55–70.
Roberts, Gareth (2003) Review of Research Assessment. Bristol: Higher Education Funding Council for England.
Scott, John (1990) A Matter of Record. Cambridge: Polity Press.
Scourfield, Jonathan, Fincham, Ben, Langer, Suzanne and Shiner, Michael (2012) ‘Sociological autopsy: An integrated approach to the study of suicide in men’, Social Science & Medicine, 74: 466–73. http://dx.doi.org/10.1016/j.socscimed.2010.01.054
Seidel, John and Kelle, Udo (1995) ‘Different functions of coding in the analysis of textual data’, in UdoKelle (ed.), Computer-aided Qualitative Data Analysis: Theory, Method and Practice. London: Sage. pp. 52–62.
Silverman, David (2006) Interpreting Qualitative Data,
3rd edition
. London: Sage.
Strauss, Anselm L. (1987) Qualitative Analysis for Social Scientists, Cambridge: Cambridge University Press. http://dx.doi.org/10.1017/CBO9780511557842
Strauss, Anselm and Corbin, Juliet (1990) Basics of Qualitative Research: Grounded Theory Procedures and Techniques,
2nd edition
. Thousand Oaks, CA: Sage.
Sudnow, David (1968) Passing On. Englewood Cliffs, NJ: Prentice Hall.
Webb, Eugene, Campbell, Donald, Schwartz, Richard and Sechrest, Lee (2000) Unobtrusive Measures: Nonreactive research in the Social Sciences, revised edition (
1st edition
, 1966). Thousand Oaks, CA: Sage. http://dx.doi.org/10.4135/9781452243443
Woods, Peter (1979) The Divided School. London: Routledge & Kegan Paul.

Copy and paste the following HTML into your website