6 Documents and other artefacts,
real and virtual
In the course of this chapter we consider two further kinds of data to which ethnographic
researchers need to pay attention: documentary sources and material artefacts. Both
are easily overlooked. In recent times, there has been so much emphasis, in some
quarters, on the study of face-to-face interactions, on encounters and situations, that
other key features of the social world could be forgotten. For example, many of the
social settings we study are self-documenting, in the sense that that their members are
engaged in the production and circulation of various kinds of written material.
Government departments and many kinds of other organization generate and consume
huge amounts of documentation. These include reports on ‘cases’, financial records,
rule-books, organizational charts, timetables, memoranda, and so on. These may be
on paper or in electronic form, or both. And, often, there are multiple sources of relevant
documentation for any setting or group of people one might study. Given this, we need
to think of contexts as involving documentary constructions of reality (Coffey and
Atkinson 2004): documentary sources construct ‘facts’, ‘records’, ‘diagnoses’,
‘decisions’, and ‘rules’ that are crucially involved in social activities (see Prior 2003,
2004). Moreover, this is not just a matter of words: images can be involved in this
process too, as Ball (2005) illustrates in relation to the practical work of the police
and road users on public highways.
It is equally important to notice that the organization of collective social activity
involves the creation, use and circulation of material artefacts. A great deal of practical
mundane activity is concerned with the manipulation of objects of one kind or another.
Social worlds are created out of material goods as well as from interpersonal
relationships, and meaning inheres in them. They have often been given less attention
even than written documents, but can be equally important.
In this chapter we will consider how these two kinds of material can be understood
within the ambit of an ethnographic approach to the social world.
Documentary sources
Ethnographic work in its various guises has frequently been employed in the investigation
of essentially oral cultures, or at least that is how they have been treated. Whether
they were the non-literate cultures of much social anthropology, or the street cultures
and demi-monde beloved of many sociological fieldworkers, the social worlds studied
by ethnographers have often been largely devoid of written documents other than those
produced by the fieldworkers themselves.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Although it was not the only rationale originally proposed for ethnographic fieldwork
as a method, the fact that the ‘exotic’ societies studied by anthropologists had no
written history was given as a major justification of the method, as well as of the synchronic
functionalist analyses that initially went with it. Rather than attempt to
reconstruct an essentially unknowable past, anthropologists were inclined to concentrate
on the construction of working versions of the present. They turned their backs on
conjectural history. There was, therefore, more than a coincidental relationship between
ethnographic methods and the investigation of non-literate cultures.1
By contrast, sociologists of the Chicago School made rather more use of written
materials, such as the records of social workers, diaries and letters. Indeed, they
sometimes requested written life histories from those they were studying. For example,
Thomas and Znaniecki (1927) in The Polish Peasant in Europe and America – generally
regarded as an early classic of American sociology – relied substantially on written
documents, mainly letters but also a life history. Thomas (1967) employed the same
approach in The Unadjusted Girl (1924). He collected personal documentary accounts,
in the belief that ‘the unique value of the personal document is its revelation of the
situations which have conditioned the behaviour’ (Thomas 1967: 42). In both cases
what we have is a dense accumulation of personal accounts, which were arranged
thematically and juxtaposed in order to draw out the regularities and contrasts in
‘definitions of the situation’ (Thomas 1967: 42).
In a rather similar vein, early use of the term ‘participant observation’ designated
the generation of documents by participants who might in contemporary parlance be
called ‘informants’. For instance, in the research that produced The Gold Coast and
the Slum, Zorbaugh (1929) persuaded people who inhabited the exclusive society of
Chicago’s ‘gold coast’ to generate such ‘inside’ accounts. They were the participant
observers as much as Zorbaugh himself.
Most of the settings in which contemporary sociologists and anthropologists work
today are literate and contain a plethora of documents: not only are their members
able to read and write, but this capacity is also an integral feature of their everyday
life and work (Smith 1987, 1993). In many instances, therefore, ethnographers need
to take account of documents as part of the social setting under investigation. Indeed,
these documents may play a central role in the activities taking place there, in a way
that they generally did not in the settings investigated by early Chicago School
sociologists.
Documents can provide information about the settings being studied, or about their
wider contexts, and particularly about key figures or organizations. Sometimes this
information will be of a kind that is not available from other sources. On other occasions
they may provide important corroboration, or may challenge, information received
from informants or from observation.
Equally important, documents may be of value in stimulating analytic ideas. The
development of generic concepts demands a broad and eclectic reading of textual
sources (formal and informal, factual and fictional) on differing substantive topics.
This means that there are rarely clear demarcations around what documents will and
will not be of value. It also means that wide reading should inform the generation of
122 Documents and other artefacts, real and virtual
1 Of course, this is much less true today; indeed, anthropologists have taken a specific interest in literacy:
Goody (1968, 1986, 1987); Street (1984); Gee (1996); Cook-Gumperz (2006).
concepts throughout the research process, rather than being limited either to the initial
stage of planning or to that of writing up. By and large sociologists and anthropologists
are not conspicuously good at this. The textual variety of an Erving Goffman is a rare
accomplishment. His work shows how documents may play a key role in facilitating
comparative analysis; both generating a sense of how the case(s) investigated are similar
to and different from others, and indicating the limits to any generalizations that it
may be tempting to make on the basis of the study. A much more recent example is
provided by Gupta’s use of ‘a prize-winning novel written by an official of the Uttar
Pradesh (U.P.) state government, and the accounts of corruption by one of the major
social anthropologists of India, F. G. Bailey’ to help make sense of the stories about
corruption in his own fieldwork data (Gupta 2005).
In a literate culture, it is possible to draw on all sorts of documents, both those
generated independently of the research as well as ones specifically elicited by the
researcher. We shall begin with a discussion of documents as ‘secondary’ sources for
the ethnographer, and then turn our attention to a more detailed examination of the
ethnography of settings where the production and use of documents is an integral
feature of everyday life.
Types of document and their uses
There is often a quite bewildering variety of documentary materials that might be of
some relevance to the researcher. These may be distributed along a dimension ranging
from the ‘informal’ to the ‘formal’ or ‘official’. At the informal end of the spectrum
there are many ‘lay’ accounts of everyday life that the enterprising and imaginative
researcher can draw on for certain purposes. These include fictional literature, diaries,
autobiographies, letters, and mass media products. Some of these will be produced in
multiple copies and made widely available, some not.
There are, for example, numerous categories of persons in contemporary society
who publish versions of their own life story:
More than ever before in history, men of affairs, including politicians, military
leaders, and business executives, are intent upon recording their experiences,
personal as well as public, for posterity. In recent decades a number of American
governmental leaders, including those in the military have, after resigning from
their official posts, published their memories or personal accounts in which they
seek public support for causes that the bureaucracy may have rejected during their
period of office.
(Sjoberg and Nett 1968: 163)
In the decades since that observation was made, nothing has changed. The output of
memoirs continues unabated.
There are, too, a large number of first-hand accounts published by less politically
eminent folk, including those drawn from the criminal underworld, and the realms of
sports and entertainment. Similar personal accounts can be found in newspapers and
magazines, or can be culled from radio and television documentaries and chat shows,
or found on the web. These accounts by or about leading scientists, musicians, artists,
celebrities, and others add to the list of contemporary social types represented in
published documents.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 123
Of course, such biographical and autobiographical accounts will rarely, if ever, be
those of the actual people we are studying at first hand. They can, nevertheless prove
valuable resources for the ethnographer. They can be a source of ‘sensitizing concepts’
(Blumer 1954): they can suggest distinctive ways in which their authors, or the people
reported in them, organize their experiences, the sorts of imagery and ‘situated
vocabularies’ (Mills 1940) they employ, the routine events, and the troubles and
reactions, they encounter. Read in this light, they can be used to suggest potential lines
of inquiry and refine ‘foreshadowed problems’.
Documents of this sort have rather particular characteristics. Authors will have
interests in presenting themselves in a (usually) favourable light; they may have axes
to grind, scores to settle, or excuses and justifications to make. They are often written
with the benefit of hindsight, and are thus subject to the usual problems of long-term
recall. Authors have a sense of audience that will lead them to put particular glosses
on their accounts. For some purposes, such considerations must be treated as potential
sources of bias in accounts of this sort. But, looked at from another perspective, the
sources of bias are data in themselves. As we noted in Chapter 5, as important as the
accuracy or objectivity of an account is what it reveals about the teller’s interests,
perspectives, presuppositions, and discursive strategies.
Such accounts can be used, with appropriate caution, for comparative purposes.
They can furnish information (albeit partial and personal) on groups and settings that
are not available for first-hand observation. As a general category of data, published
biographical and autobiographical sources are subject to a further sort of ‘bias’ in that
they tend to over-represent the powerful, the famous, the extraordinary, and the articulate.
But even that can also be a strength, since it is precisely people in such social categories
who are often difficult to research directly.
Since the mid-1990s there has been a considerable resurgence of interest in the
sociological analysis of biographical or autobiographical accounts. While that interest
goes well beyond the scope of ethnographic research, ethnographers can incorporate
many of the insights from this research field (Reed-Danahay 1997, 2001; Plummer
2001; Chase 2005). The growth in scholarly interest reflects a renewed emphasis on
the study of narrative forms, temporality, and memory. It reflects too a focus on the
intersection of the ‘personal’ and the ‘social’ (Erben 1993). As Stanley (1993)
summarizes some of these concerns:
I see a concern with biography and autobiography as fundamental to sociology,
because I perceive the grounds of their sociological interest lying within the
epistemological problematics concerning how we understand ‘the self’ and ‘a life’,
how we ‘describe’ ourselves and other people and events, how we justify the
knowledge-claims we make in the name of the discipline, in particular through
the processes of textual production.
(Stanley 1993: 50)
These sociological perspectives on ‘lives’ and ‘documents’ also often reflect commitments
to a feminist standpoint. Documentary sources may be drawn on to recuperate
the otherwise muted voices of women and other dominated groups, and feminist
scholarship particularly affirms the intersection of the personal and the social (Stanley
1992; Evans 1993).
124 Documents and other artefacts, real and virtual
In the collection and investigation of ‘informal’ documentary materials, the fictional
– even the most popular and ephemeral – can be used profitably. The most banal
(‘pulp’ or ‘pot-boiler’) fiction is often replete with images, stereotypes, and myths
bearing on a vast range of social domains. Indeed, the lack of literary merit characteristic
of such genres reflects the fact that it unquestioningly trades on stocks of common
knowledge and conventional wisdom. Here too, then, we can become sensitized to
cultural themes pertaining to sex, gender, family, work, success, failure, class, mobility,
regional variations, religious beliefs, political commitments, health and illness, the law,
crime, social control, etc. These are not necessarily to be read at face value, as accurate
representations of social reality, but can suggest themes, images, or metaphors. This
is no less true of more ‘serious’ fiction: novels can suggest different ways of organizing
experience, and alternative thematic models. We need not shy away from the careful
use of literary sources. As various authors have pointed out, there is a long and
complex set of relationships between literature and the social sciences (for example,
Lepenies 1988; Cappetti 1993). As Davis (1974) noted in the mid-1970s, ethnographers
and novelists alike find themselves telling ‘stories’.2
Of course, by no means all informal documents are published, in the sense of being
widely available via multiple copies. Unpublished personal diaries, or works of fiction,
are sometimes available. There are also informal documents produced in the field that
can illuminate key social relationships. In her study of girls’ friendships, Hey (1997)
made considerable use of the notes that the girls passed to one another in the course
of lessons. She comments that they ‘represented in their most condensed and dramatic
form’ some of the key themes of her book (Hey 1997: 2).3
At the other end of the spectrum are sources of a more ‘formal’ nature, and these
include other published ethnographies. There is every reason for the sociologist interested
in, say, hospitals or clinics to examine works on a variety of other institutional settings
– schools, courts, welfare agencies, religious houses, police stations, university departments,
or emergency services, for example. The precise selection of settings, and the
lessons drawn from them, will depend on the analytic themes being pursued. Through
such comparisons one might trace the variety of ‘degradation ceremonies’, the conditions
of ‘information control’, or the moral evaluation of ‘clients’. There is, in principle, no
limit to such comparative work, and no prescriptions can be offered. The part played
by serendipitous discoveries and unpredicted insights will be considerable here, as in
all creative work. One must establish the right conditions for serendipity, however,
and that includes attention to sources of many sorts. As Glaser and Strauss (1967)
remark with characteristic enthusiasm:
Theorizing begs of comparative analysis. The library offers a fantastic range of
comparison groups, if only the researchers have the ingenuity to discover them.
Of course, if their interest lies mainly with specific groups, and they wish to
explore them in great depth, they may not always find sufficient documentation
bearing on them. But if they are interested in generating theory, the library can
be immensely useful especially . . . for generating formal theory. Regardless of
which type of theory the theorist is especially interested in, if he browses intelligently
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 125
2 See Chapter 9 for further discussion of parallels between ethnography and literary analysis.
3 Nowadays, such data may no longer be available to the researcher; text messaging is probably used
instead.
through the library (even without much initial direction), he cannot help but have
his theorizing impulses aroused by the happily bewildering, crazy-quilt pattern of
social groups who speak to him.
(Glaser and Strauss 1967: 179)
As in Goffman’s work on topics like ‘total institutions’ (Goffman 1961; see also
Perry 2007), the imaginative use of secondary documentary sources allows for the
elaboration of ‘perspective by incongruity’ (Burke 1964; Lofland 1980; Manning 1980):
that is, the juxtaposition of instances and categories that are normally thought of as
mutually exclusive. Such sources and devices are ideal for heuristic purposes: they
can rejuvenate jaded imaginations and spark off novel conceptualizations. In his or
her imagination the researcher is free to wander at large among diverse social scenes,
gathering ideas, insights, hypotheses, and metaphors.
In addition to the sorts of documentary source we have referred to up to now, in a
literate culture it is possible to emulate researchers like Zorbaugh (1929) and draw on
the ability of informants to generate written accounts specifically for research purposes.
By such means one can gather information that complements other data sources in the
field. Some versions of research have indeed drawn extensively on such indigenous
written accounts. The entire tradition of ‘mass observation’ in Britain rested on the
ability of literate volunteers to produce ‘native’ accounts of everyday life around them
(see Stanley 2001). The revival of the Mass Observation Archive has again depended
on such written documents:
The writing has been generated in response to a call from the Mass-Observation
Archive, repeated at intervals over the years, for people to take part in a form of
collective autobiography. No special skills, knowledge or qualifications are required,
only an enjoyment of writing and a willingness to put thoughts and experiences
on paper in a discursive way.
(Sheridan 1993: 27)
This emphasis on the collection of demotic accounts, characteristic of Mass Observation,
is but one version of wide possibilities for the collection of documentary evidences.
The use of diaries of different types is often an important adjunct to fieldwork. An
influential version of this strategy was advocated by Zimmerman and Wieder (1977),
who used a diary technique in their study of counter-cultural lifestyles. They comment
that, while they were committed to participant observation, there were settings and
activities that remained hard for them to observe directly. They therefore recruited
insider informants, who kept detailed diaries over seven-day periods. Subsequently,
the researchers subjected each informant to a lengthy and detailed interview, based on
the diaries,
in which he or she was asked not only to expand the reportage, but also was
questioned on the less directly observable features of the events recorded, on their
meanings, their propriety, typicality, connection with other events and so on.
(Zimmerman and Wieder 1977: 484)
Narratives are also sometimes solicited by ethnographers. For example, Nygren and
Blom (2001) collected short narratives from social work students dealing with problems
126 Documents and other artefacts, real and virtual
that they had experienced. They note not only the advantages that such data can have,
but also its drawbacks by comparison with interviews. Combining the two methods
could, of course, overcome these.
Solicited written accounts, such as diaries and narratives, are especially useful ways
of obtaining information about the personal and the private.4
When carefully managed,
and with suitable cooperation from informants, the diary, in particular, can be used to
record data that might not be forthcoming in face-to-face interviews or other data
collection encounters. Sexual behaviour is one obvious example. For instance, one
major study among gay males made extensive use of personal diaries in order to obtain
information on the types and frequencies of sexual practices (Coxon 1988, 1996).
Similarly, Davies used personal diaries in her study of student midwives (Davies
and Atkinson 1991). Her research shows some of the anxieties and coping strategies
associated with status passage, as experienced nurses became novice midwives. It is
noticeable from the responses Davies obtained that the students were able to use the
research diaries as a kind of personal confessional, often addressing the researcher
directly about private anxieties, sources of anger, and frustrations. These personal
accounts were complemented by interviews and observations.
Diaries of this sort can also be used to pick up the minutiae of day-to-day social
action. Robinson (1971), in the course of an investigation of the experience of illness,
persuaded a series of married women in South Wales to keep a diary on the health
status of the members of their households. The diaries were kept over a four-week
period. They enabled Robinson to gain some insight into the daily symptomatic episodes
and health-related decisions characteristic of everyday living. Many of the episodes
reported were minor, though by no means insignificant, and could easily have been
overlooked in retrospective accounts produced in, say, interviews or questionnaires.
This sort of procedure has also been used in work on educational settings. Ball
(1981), for instance, used diaries in combination with a range of other techniques,
including sociometric questionnaires on friendship choices. He explicitly notes the
value of combining such data sources:
The sociometric questionnaires failed to pick up the casual friendships that existed
between pupils outside school, and made it appear that they had no such contact.
In addition, they failed to pick up the cross-sex friendships that were established
at this time. Perhaps the notion of ‘friendships’ is too narrow and ill-defined to
account for these other kinds of adolescent relationships. . . . The entries in the
diaries that several of the pupils wrote for me did, however, refer to these contacts.
(Ball 1981: 100)
Research-generated documents of this sort embody the strengths and weaknesses
of all such personal accounts. They are partial, and reflect the interests and perspectives
of their authors. They are not to be privileged over other sources of information,
but nor are they to be discounted. Like other accounts, they should be read with
regard to the context of their production, their intended or implied audiences, and the
author’s interests. Equally, one must note that a written account is not a debased version.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 127
4 An alternative strategy, of course, is to provide informants with audio-recorders that they can use to
record day-to-day events, their own thoughts and feelings, and so on. There is an interesting question
about what differences there could be between written and oral diaries.
Given the historical and intellectual roots of ethnographic work, one can often detect
a romantic legacy that privileges the oral over the literate. It is easy (but wrong) to
assume that the spoken account is more ‘authentic’ or more ‘spontaneous’ than the
written.5
We have discussed a range of documentary sources, but we have not yet paid attention
to the investigation of social activities that directly involve generating documents.
Fieldwork in literate societies – especially in formal organizations – is likely to
encompass the production and use of documents of various kinds. In the following
section we turn to such activities and their documentary products.
Documents in context
In some settings it would be hard to conceive of anything approaching an ethnographic
account without some attention to documentary material in use. For instance, in his
study of locomotive engineers, Gamst (1980) drew on a range of documentary sources:
Some documents are published, for example: rule books, timetables, technical manuals
for use of equipment, and instructional, regulating, and investigating publications
of many kinds used by railroads, trade unions, government, and other firms.
Unpublished documents include: official correspondence, reports in mimeographed
and other forms, railroad operating bulletins and circulars, train orders, operating
messages, and sundry other items.
(Gamst 1980: viii)
Whether or not one would draw on all such sources, one would certainly expect an
ethnography of work on the railway to make full reference to such features as operating
schedules and timetables (whatever disgruntled passengers might feel). A similar
instance is provided by Zerubavel (1979) in his formal analysis of time in hospitals;
he necessarily draws on such sources as timetables, work rosters and clinical rotations,
as embodied in organizational documents. In many organizational settings the use and
production of such documents are an integral part of everyday life (Prior 2003).
Similarly, the ethnographic study of scientific work – especially the genre of
‘laboratory studies’ – cannot proceed adequately without acknowledgement of the
work of writing. For instance, Latour and Woolgar (1979), in their classic study of a
biomedical laboratory, document the centrality of written outputs. The scientific
laboratory is fundamentally preoccupied with what these authors call ‘inscriptions’:
that is, representations of natural phenomena, and the texts that are the products of the
laboratory. Scientific papers are the currency that circulates within and between scientific
research groups. One cannot address the complex social realities of scientific work,
and the production of scientific knowledge, without paying serious attention to how
and why scientific papers are written. The sociology of scientific knowledge is now
replete with studies of written texts and other forms of representation (for example,
Lynch and Woolgar 1990). And the same approach may be extended to all organizational
and professional settings.
Douglas, writing in 1967, commented on the increasing importance of ‘official’
numerical data in contemporary society (see also Porter 1995; Power 1997), while
128 Documents and other artefacts, real and virtual
5 Derrida (1976) has identified this as a key theme in Western thought.
simultaneously regretting the relative failure of sociological commentators to address
this as a topic:
Throughout the Western world today there exists a general belief that one
knows something only when it has been counted. . . . Considering the importance
of such statistics for the formation and testing of all kinds of common-sense and
scientific theories of human action, it is a remarkable fact that there is at present
very little systematic knowledge of the functioning of official statistics-keeping
organizations.
(Douglas 1967: 163)
Since Douglas made those observations, there has been an increasing amount of work
along the lines suggested (see Prior 2003, 2004). However, in comparison with the
sheer volume of ‘literate’ record-keeping and documentation in contemporary society,
the coverage remains at best patchy. There is still, apparently, a tacit assumption that
ethnographic research can appropriately represent contemporary social worlds as
essentially oral cultures. Many studies of medical settings, for instance, focus exclusively
on spoken interaction between medical practitioners and their patients, or between
health professionals, with relatively little attention to activities of reading and writing.
As Rees remarks: ‘Both medicine and medical sociology have to a large extent neglected
the record. Indeed, so rarely is it mentioned that one could be forgiven for thinking
that medicine is a purely oral discipline’ (Rees 1981: 55).
Pettinari (1988) demonstrates the value of close attention to ‘writing’ in a medical
setting. Here is provided a detailed account of how surgeons write their reports on
operations, and in particular of how junior surgeons learn such occupational skills.
There are ways in which the operation is represented competently in surgeons’ reports,
and the appropriate formulations are acquired over time with professional experience.
The written account is a fundamental element in the everyday organization of surgical
work. Its production and use are an important focus for an ethnographic account of
surgery in general.
In a similar vein is Coffey’s (1993) ethnography of accountants in training. Based
on fieldwork in an office of an international accounting firm, she documented aspects
of trainees’ acquisition of accountancy expertise. She studied bookkeeping skills together
with the trainees, and describes how they acquired skill and judgement in reading
documentary sources such as balance sheets. It would clearly be absurd to represent
the world of the corporate accountant as non-literate or as non-numerate – and a
comprehensive ethnographic account must include reference to how organizational
documents are read, interpreted, and used.
Because of the critique of ‘official statistics’, stemming largely from the ethnomethodological
movement, some contemporary ethnographers may feel reluctant to
engage in the systematic use of documentary data. We believe that they are right to
treat seriously objections against employing ‘official’ data simply as a resource not a
topic, but that they would be wrong to ignore such materials. The point of departure
for critics of ‘data from official sources’ was the contention that, traditionally, the
tendency had been for sociologists to treat such information at face value, and not to
pay adequate attention to its character as a social product. This opens up an important
area of investigation, rather than implying that official data should be ignored.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 129
It is, of course, a long-standing concern of sociologists that data derived from
official sources may be inadequate in some way: that they may be subject to bias or
distortion, or that bureaucracies’ practical concerns may mean that data are not
formulated in accordance with sociologists’ interests. The ethnomethodologists, on the
other hand, proposed more radical problems. Cicourel (1976) remarks, for instance:
for years sociologists have complained about ‘bad’ statistics and distorted
bureaucratic record-keeping but have not made the procedures producing the ‘bad’
materials we label ‘data’ a subject of study. The basic assumption of conventional
research on crime, delinquency and law is to view compliance and deviance as
having their own ontological significance, and the measuring rod is some set of
presumably ‘clear’ rules whose meaning is also ontologically and epistemologically
‘clear’.
(Cicourel 1976: 331)
The argument is that rather than being viewed as more or less biased sources of data,
official documents and enumerations should be treated as social products: they must
be examined, not relied on uncritically as a research resource.
In this way attention is diverted towards the investigation of the socially organized
practices whereby rates, categories, and statistics are produced by those whose job it
is to generate and interpret such phenomena. An early example of work in this vein
was that of Sudnow (1965) on the production of ‘normal crimes’ in a Public Defender’s
office. Sudnow details the practical reasoning that informs how particular crimes or
misdemeanours become categorized in the course of organized activities such as plea
bargaining. Thus, he looks ‘behind’ the categories of official designations and crime
rates – based on convictions – to the work of interpretation and negotiation that generates
such statistics. In addition to Sudnow’s ethnographic study of crime rates, other studies
of the same period included those of Cicourel (1967) on juvenile justice, and of Cicourel
and Kitsuse (1963) on the organization of educational decision-making, the categorization
of students, and their official biographies. More recent research in a similar
vein includes a welter of constructionist accounts of social problems (see, for example,
Holstein and Miller 1989, 1993). Similar in focus are Prior’s studies of the social
organization of death, with particular emphasis on the classification of causes, and
of mental illness (Prior 1985, 1989, 1993). In this context one should also note the
observations of Prior and Bloor (1993) on the life-table as a cultural and historical
artefact.
The origins of the ‘official statistics’ debate in sociology were potentially misleading,
important though the general perspective was. Issues became polarized quite
unnecessarily. The problems associated with data from official sources were important,
and they related directly to classic problems of sociological analysis, such as the
explanation of suicide (Douglas 1967; Atkinson 1978); but they were by no means
unique. The careful ethnographer will be aware that all classes of data have their problems,
and all are produced socially; none can be treated as ‘transparent’ representations
of ‘reality’. The recognition of reflexivity in social research entails such an awareness
(Holstein and Miller 1993). As a result, there is no logical reason to regard documents
or similar information as especially problematic or totally vitiated. As Bulmer (1980)
remarks in this context:
130 Documents and other artefacts, real and virtual
Firstly, there is no logical reason why awareness of possible serious sources of
error in official data should lead to their rejection for research purposes. It could
as well point to the need for methodological work to secure their improvement.
Secondly, a great many of the more thorough-going critiques of official statistics
relate to statistics of suicide, crime, and delinquency, areas in which there are
special problems of reliable and valid measurement, notoriously so. The specific
problems encountered in these fields are not, ipso facto, generalizable to all official
statistics whatever their content. Thirdly, cases of the extensive use of official data
– for example, by demographers – do not suggest that those who use them are
unaware of the possible pitfalls in doing so. The world is not made up just of
knowledgeable sceptics and naive hard-line positivists.
(Bulmer 1980: 508)
In other words, then, while drawing some inspiration from the ethnomethodological
critique of ‘official statistics’ and similar documentary sources, we by no means endorse
a radical view which suggests that such sources are of no value as a resource. Data
of this kind raise problems, to be sure, but they provide information as well as opening
up a range of analytic problems that are worth investigation. The ethnographer, like
any other social scientist, may well draw on such documents and representations.
Furthermore, he or she may be particularly well placed to engage in principled and
systematic research bearing on their validity and reliability as data, through first-hand
investigation of the contexts of their production and use.
Rees (1981), in his work on medical records, indicates the situated significance of
readership and authorship within a professional setting:
What the House Officer writes, and the way in which he goes about constructing
the history and examination, is one way his seniors can make inferences about the
standard of his other activities. The supposition others make is that a House Officer
who writes an organized and clearly thought out account of his work will be well
organized in the way he carries out those activities. By paying attention to the
construction of the account, and by ensuring that it conforms to the accepted
model, the House Officer is able to influence one of the ways in which he will be
judged by his seniors.
(Rees 1981: 58–9)
This reflects Garfinkel’s (1967) remarks on records, where he suggests that they should
be thought of as ‘contractual’ rather than as ‘actuarial’. That is, they are not literal
accounts of ‘what happened’ but are tokens of the fact that the relevant personnel went
about their business competently and reasonably (see also Prior 2003). This is something
that was taken up by Dingwall (1977b) in his study of the education of health visitors.
He writes about the students’ production of records of their visits to clients, and notes
that since the actual conduct of the work is invisible to the supervisor, the record is
the main focus of administrative control. Likewise, the record constitutes a major means
of self-defence for these ‘face-workers’. And, of course, the role of documents, of
various kinds, in regimes of ‘transparent’ accountability has increased substantially in
recent decades, with the rise of the ‘audit society’ (Power 1997).
In various ways, then, records have considerable importance in certain social settings.
In some, the production of ‘paperwork’ is a major preoccupation. Even in organizations
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 131
that have people-processing functions, this usually involves the translation of events
into records of those events which can be filed, stored, and manipulated. Such files
are a primary resource for members of the organization in getting through their everyday
work. Often, the exigencies of record-making can play an important part in organizing
the work that gets done, and the routines used to accomplish it. Records of previous
encounters with clients can be used to formulate appropriate objectives and activities
for a current consultation. As Dingwall (1977b) writes of his student health visitors:
The good health visitor can derive sufficient data from the face sheet to identify
the relevant areas of her knowledge about clients and the tasks she should be
accomplishing in a visit. Unusual events are flagged in various ways. Thus, a child
who is at risk may be marked by a red star on the card. Particular social problems
may be pencilled on the cover.
(Dingwall 1977b: 112)
Heath (1981) has also commented on this sort of use of medical records in the
context of doctor–patient encounters. He explains how general practitioners use their
record cards to open the encounter with the patient:
It is often through the elaboration of the appropriate record’s contents, prior to
the initiation of first topic, that the doctor is able to render the relevant characteristics
of the patient, and thereby design a ‘successful’ first topic initiator.
(Heath 1981: 85)
Records, then, are used to establish actors as ‘cases’ with situated identities, which
conform to ‘normal’ categories or deviate from them in identifiable ways. Records are
made and used in accordance with organizational routines, and depend for their
intelligibility on shared cultural assumptions. Records construct a ‘documentary reality’
that, by virtue of its very documentation, is often granted a sort of privilege. Although
their production is a socially organized activity, official records have a certain anonymity
which leads to their treatment by members as objective, factual statements rather than
as mere personal belief, opinion, or guesswork. (It is, of course, the case that some
records may contain specific entries, such as differential medical or psychiatric diagnoses,
that are explicitly flagged as tentative or contested.)
It should be apparent from what we have outlined already that there are many locales
where literate social activity is of some social significance, and may indeed be of major
importance. Modern industrial and administrative bureaucracies, and professional or
educational settings, are obvious cases in point. It requires little reflection to remind
oneself of how pervasive are the activities of writing and reading written documents.
And even in the case of settings where documents are not a central feature there is
often an enormous amount of written material available that can be an invaluable
research resource.
The presence and significance of documentary products provide the ethnographer
with a rich vein of analytic topics, as well as valuable sources of data and information.
Such topics include: How are documents written? How are they read? Who writes
them? Who reads them? For what purposes? On what occasions? With what outcomes?
What is recorded, and how? What is omitted? What does the writer seem to take for
granted about the reader(s)? What do readers need to know in order to make sense of
132 Documents and other artefacts, real and virtual
them? The list can be extended readily, and the exploration of such questions would
lead the ethnographer inexorably towards a systematic examination of each and every
aspect of everyday life in the setting in question.
The ethnographer who takes no account of such matters, on the other hand, ignores
at his or her peril these features of a literate culture. There is nothing to be gained,
and much to be lost, by representing such a culture as if it were an essentially oral
tradition. In the scrutiny of documentary sources, the ethnographer thus recognizes
and builds on his or her socialized competence as a member of a literate culture. Not
only does the researcher read and write, but also he or she reflects on the very activities
of reading and writing in social settings. Thus, such everyday activities are incorporated
into the ethnographer’s topics of inquiry as well as furnishing analytic and interpretative
resources.
We have concentrated here for the most part on written documents, but it is perhaps
worth stressing that materials containing images of various kinds, whether plans,
drawings, photographs, or sounds are also often available. The mass media are a major
source here, but by no means the only one. Images of various kinds are often part of
the documents that are generated via organizational work, for example prospectuses
for schools and universities, brochures of various kinds, the pictures (in print and
online) that play a key role in the buying and selling of houses, etc. Many of the same
questions need to be asked about these materials as about written documents; though
the kinds of analysis used are likely to be distinctive (Emmison and Smith 2000; Pink
2004b).
Heath’s work on technologically mediated action provides a series of key examples.
Heath and Luff, for instance, analyse collaborative work and interaction in control
rooms of the London Underground system, where visual data on screen is a key feature
of everyday work (Heath and Luff 2000). In settings of scientific and medical work,
visual culture may also be of fundamental importance, with data themselves being
constructed in visual terms. Dumit’s ethnography of brain-imaging is but one recent
example: the striking images of brain scans he reproduces do not just illustrate the
text. The visualization of brain function is what the bioscientists do, and their scientific
work is intensely visual in character. Consequently, what is important here is not just
the ‘visual analysis’ of scientific work, but the analysis of scientific visual work. There
is, in other words, a direct homology between the social action under investigation
and its ethnographic representation (Dumit 2004).
Artefacts
Just as ethnographers can sometimes overlook the literate quality of many social settings,
they often seem to ignore the role of material artefacts. While the rise in studies of
material culture has brought some of the relevant issues to the fore, there remains a
tendency for material goods and objects to be neglected in ethnographic work, being
left within a circumscribed boundary of specialist interests. Our (all too brief) account
here is intended merely to remind ethnographers that the ‘fields’ in which they conduct
fieldwork are populated not only with social actors, but with ‘things’ of many sorts.
Of course, the ethnographic gaze has encompassed material culture since the earliest
days. Malinowski’s classic anthropological research among the Trobrianders included
a close reading of the building of a canoe, and how the practical construction work
simultaneously evoked and aligned cooperative social relations. Likewise, his account
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 133
of reciprocity and exchange relationships in the kula system necessarily involved some
attention to the material goods that are circulated among exchange-partners across the
archipelago. However, this focus on material artefacts is not very common in presentday
ethnographic studies.
Rather than treating the analysis of such artefacts as a separate domain, we wish
to stress that material goods, objects and traces need to be analysed in their broad
ethnographic contexts. Moreover, the ethnography of everyday life demands attention
to its material features, and how social actors engage with physical things. This is
much more than just applying attention to physical surroundings and material context.
There are many social phenomena that are impossible without the use of material
goods. There are many social relationships that are crystallized and embodied in material
objects. There are many forms of work that necessarily imply the competent and skilled
manipulation of physical resources, whether as an official part of the job or in the
production of ‘homers’, artefacts for personal use (Anteby 2003).
For instance, Atkinson’s publications on haematologists highlight some of the material
work that goes into the production of medical facts and opinions. Haematologists spend
a lot of time talking about their patients, and transforming those patients into ‘cases’,
‘opinions’ and ‘diagnoses’ (Atkinson 1995). But their work is not just accomplished
through talk. They also have to manipulate the physical traces of the body: the blood
and bone-marrow have to be collected and read for signs of disease. This involves
their physical manipulation, through techniques such as staining, in order to render
them visible and legible. The preparation of a slide for microscopic inspection demands
a certain physical and perceptual dexterity, as does use of the microscope itself. The
skilful activities of inspecting physical traces of peripheral blood or bone-marrow are
embedded in the exchange of talk between professional colleagues, or in teaching
encounters. Objects, traces, skills and talk are mutually implicated, and the ethnography
of professional knowledge-production has to take account of them.
These observations in a medical setting parallel a more general interest in ethnographic
studies of scientific sites, such as laboratories and scientific networks. While a great
deal of the sociology of scientific knowledge has been disproportionately concerned
with the cultural-cognitive aspects of scientific controversy and discovery, some studies
have explicitly incorporated an interest in the material circumstances of scientific work.
Actor-network theory takes such an interest to an extreme (occasionally absurd) length.
It makes no principled distinction between human and material actants within a complex
of interrelationships. Hence, technical equipment can be analysed in terms of the theories
of knowledge it embodies, and the active part it plays in shaping scientific knowledge.
However, it is not necessary to engage in fully fledged actor-network theory in order
to recognize the significance of materials in the production of scientific and technological
knowledge. The ethnographer in this field needs to acquire some degree of technical
expertise and an understanding of how material artefacts are made and used. This
includes an appreciation of the physical qualities of things. A contemporary oral history
of the Moog synthesizer, which played a significant part in the development of rock
music, includes an appreciation of its technical qualities as well as the biographical
relations in which it was embedded (Pinch and Trocco 2002). Ethnography and the
archaeology of the present also converge in the analysis of computing and information
technology (Finn 2001; English-Lueck 2002).
The ‘thing-ness’ of things in their material and social contexts needs to be understood
by ethnographers. We are attentive to the nuances of social interaction, and we devote
134 Documents and other artefacts, real and virtual
a considerable amount of time and effort to the analysis of social action. We need to
give equally appropriate attention to things. They have material qualities, based on
their physical composition, they have surface texture, shape and colour. Ethnographers
have not been particularly good at exploring the significance of appearance in the
material world about them. The social worlds that many of them describe have been
oddly empty spaces, with little or no attention paid to the physical surroundings. Those
ethnographic worlds are often flat and monochrome, in that aesthetic qualities, such
as colour, are ignored. Yet, if we want to make sense of many social worlds, we ought
to take account of how they are physically constituted. It would, for instance, seem
odd to describe work in a modern office setting without paying heed to its physical
layout (walled rooms or open plan, number of floors, connections to other buildings,
etc.), its colour schemes, its furnishings and the like. Such places are designed and
their design embodies corporate interests and implicit values. These features both
constrain the social relations of work in distinctive ways and are resources that are
used and even ‘redesigned’ by workers.
Atkinson’s (2006a) ethnography of the Welsh National Opera Company gives central
attention to several aspects of its material context. The production and performance
of an opera depend not just on singing and acting, and the key features of an opera
go beyond the work of the director and the conductor. The realization of music theatre
involves the creation of material circumstances, such as the design and construction
of the set. Moreover, once the set has been created, it in turn poses physical constraints
on but also provides material opportunities to producers and performers. Atkinson’s
monograph includes graphic accounts of how performers had to grapple (literally as
well as metaphorically) with recalcitrant aspects of the set: walls that move unpredictably
(Simon Boccanegra), or boats that are difficult to manhandle (Peter Grimes). Atkinson
(n.d.) writes about how the props department in the opera company used their craft
skills to engage in ‘bricolage’ to create the physical objects required for a new production.
This department was also a physical archive of past productions, and a source of
situated ‘ethno-archaeology’ through which the past achievements of the opera company
could be traced. Individual pieces that had required particular ingenuity for their
construction, or that have particular aesthetic value, are treasured by the props department
as trophies of their own past ‘operatic’ triumphs.
Any theatrical performance, not least that of the opera, depends on material objects.
However, as Erving Goffman showed many years ago, we all depend upon ‘props’ of
various kinds in more mundane settings. Furthermore, there are ethnographic accounts
that are explicitly focused on the collection and display of material artefacts. Macdonald’s
(2002) ethnography of London’s Science Museum is an excellent case in point. She
narrates the processes that went into the creation of a major museum exhibit on food.
Such ethnographic examinations of the curating and display of material goods do more
than documenting what happens in such esoteric places as museums and art galleries
– important though those topics are in their own right. They also draw attention to
more mundane aspects of material culture. The collection and display of objects is an
important feature of mundane home cultures, for instance. Homes can be thought
of as having ‘museum-like’ qualities, in that individual memorabilia or collections of
objects can be self-consciously displayed. Hurdley’s (2006, 2007) multimedia ethnography
of mantelpieces in British homes is a useful case in point. The mantelpiece –
the traditional structure over an open fireplace, still incorporated in modern homes that
have central heating – is a key place for the display of ornaments, gifts, photographs
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 135
and the like. There is a situated aesthetic that informs many of these displays, and the
mantelpiece arrangement is often a focus of the living room. Moreover, the objects
themselves embody their owners’ memories: memory is not a mental state from this
perspective, but is inscribed in material objects, and in the autobiographical narratives
that they evoke. Mantelpiece objects can also embody the personal ties and mutual
obligations of gift relationships. A deceptively trivial space such as the mantelpiece
thus provides a microcosm of everyday domestic, aesthetic and interpersonal arrangements.
The homespun example of the mantelpiece also directs ethnographic attention
towards the more self-conscious cultural phenomenon of ‘collecting’ art and other
material artefacts (see Painter 2002; Belk 1995; Pearce 1994).
The physicality of social phenomena can be extended to incorporate an appreciation
of the built environment and physical space. While it is by no means a universal failing,
it is undeniable that many ethnographic accounts seem to lack a sense of places and
spaces. This is not just a matter of putting things into a ‘context’. Rather, we ought
to pay serious attention to the material circumstances that constrain social activity,
how a sense of place is reflected in individual and collective identities, and how places
are used by social actors, just as they use any material and symbolic resources. There
are, of course, some clear examples that can be drawn on for inspiration. Pierre
Bourdieu’s (1979) classic analysis of the Kabyle house is a case in point. He analyses
the house not merely as a physical locus of everyday life, but also as a site of symbolic
ordering.6
In a similar vein, anthropological accounts of Mediterranean societies stress
the significance of the house, and its relationship to the street, as a physical coding of
respectability – especially as regards the women of the household – and as an affirmation
of strongly differentiated gender roles.
The house as a physical embodiment of identity, and its material decoration, are
explored by Gregory (2003) in her exploration of householders’ aesthetic work on late
Victorian houses. These include the house-owners’ need to purify the house of traces
of previous occupants (carpets and soft furnishings, décor and colour schemes, bathroom
suites and fittings etc.), often expressed in quite violent terms (‘ripping out’ fireplaces,
for instance). The material decoration and furnishing of the house are not simply
backdrops to the everyday performance of identity and biography. They are deeply
implicated in the joint construction of a household by (for instance) couples who
transform an older house into their own domestic space. These issues of personal
identity are often informed by a sense of the identity of the house itself: the ‘authenticity’
of the material fabric of an older house is an important aspect of such reconstruction.
Fireplaces, doors, windows and other ‘features’ may – for some house-owners at
least – need to be in period if their personal, emotional investment in the house is to
be adequately rewarding. These are partly aesthetic judgements, of course. They are
also partly reflections of collective taste: in previous generations, the ‘modernization’
of the home would have been a far more common goal than its restoration. But
they are profoundly dependent on the ethno-aesthetic of the social actors themselves,
and depend upon the values attached to material goods (see Reid and Crowley 2000;
Valis 2003).
The ethnographic appreciation of things should also extend to their creation and
manufacture. As we have already noted, the rise of cultural studies and the associated
emphasis on material culture have given rise to a rich vein of sources and analytic
136 Documents and other artefacts, real and virtual
6 For a critical assessment of his argument, see Silverstein (2004).
perspectives that can be drawn into the ethnographic imagination (see Miller 1998,
2001a, 2001b; Buchli 2002). They include the ethnographic analysis of processes
of design, an aspect of contemporary material culture that has attracted considerable
interest on the part of social scientists.7
This includes ethnographic analysis of the
design process (e.g Henderson 1998; Salvador et al. 1999) and the creation of visual
culture itself (Frosh 2003). The same is true of architecture, urban design or the
analysis of spaces and places (e.g. Dodds and Tavernor 2001; Borden 2002; Podalsky
2002; Crowley and Reid 2002; Butler and Robson 2003). We are now more than ever
aware of the significance of the ‘environment’, as simultaneously shaped and shaping
in the performance of individual and collective identities.
We repeat here a caveat. We are not referring to these issues simply in the interests
of documenting the ‘contexts’ of social activity. These are not epiphenomenal issues
either. Ethnographic research needs to pay close and serious attention to the material
goods and circumstances that are integral to the organization of everyday social life.
People do not act in a vacuum. Not only do they do things with words, but also they
do things with things. The sort of issues that researchers sometimes lump together
as ‘experience’, ‘biography’ or ‘memory’ are often embodied in material goods and
personal possessions. The performance of work involves a sustained engagement with
material means. The enactment of ritual normally involves the manipulation of objects
charged with special significance. These and similar kinds of social action all call for
a systematic ethnographic attention to the material world. We do not need a separate
specialism or subdiscipline of ‘material culture’ in order to address these issues. On
the contrary, they should be incorporated into the fabric of ethnographic inquiry, just
as they contribute to the fabric of social life.
Finally, we should note that while physical space and objects have always been
important for human behaviour, in recent times the virtual spaces created through fixed
and mobile communication devices and the internet have gained increasing significance.
While some ethnographers have given much attention to ‘online cultures’, there has
perhaps been rather less awareness of how these virtual spaces shape more ordinary
forms of social interaction, how virtual interaction intertwines with that which is faceto-face,
and so on (Franklin and Lowry 2001). Here, too, there is a danger involved
in specialization, that these issues will be ignored where they are not the main focus
of inquiry. Moreover, of course, this is a field of expanding significance.
Ethnography in digital spaces
Digital technology has expanded our very notion of what constitutes a ‘field’. Virtual
fields and virtual fieldwork are now possible, and are assuming increasing significance
in a social world that is simultaneously global and digital for some populations. Virtual
communities and networks present particular challenges and opportunities for ethnographic
research. Likewise, ethnographies of digital life itself are important aspects of
contemporary social research. In addition, there are many opportunities to exploit digital
technologies to reach particular research populations.
There are many kinds of informant who can be usefully accessed via the internet.
For example, Stewart’s research on people with irritable bowel disease included online
access to sufferers. This is a very good example of how a potentially embarrassing or
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 137
7 See Julier (2000) for a useful introduction.
stigmatizing condition can be studied ‘at a distance’ when face-to-face interviewing
and personal recruitment to a study might prove more troublesome (Stewart and Williams
2005). In much the same way Susie Scott conducted an interactionist project on shy
people and constructions of shyness. She was able to contact and interact with shy
people through internet access. Shyness is another prime example of the sort of
personal quality or characteristic that lends itself especially well to internet research.
Contrary to what one might initially and naively believe, shy people can be very willing
to talk at length about themselves, and they are extremely reflective about their own
interactional style. Indeed, their behaviour is often shy precisely because they have a
heightened sense of self, are alert to the interactional niceties of dialogic interaction
and group dynamics. Scott found that informants could and would produce very full,
self-aware accounts of shyness, and that they could find the impersonal channel of
communication of the internet quite amenable (Scott 2004).
Such research does not necessarily study ‘naturally occurring’ communities that
exist in cyberspace, although in the course of setting up noticeboards and chat rooms
researchers can create temporary, research-situated groupings that approximate to online
focus groups. Their advantage over face-to-face focus groups is that they can persist
more or less indefinitely, and actors can be enrolled on a rolling basis, while participation
is not restricted by constraints of time and travel. Equally, however, we recognize that,
increasingly, there are ‘naturally occurring’ communities in virtual, digital space.
Researchers who use online means of accessing informants and participants are often
tapping into existing networks of social actors who maintain ‘virtual’ communities
through sustained interaction by digital means. It is clear that, for some participants
at least, this is a prime social reality, and a major source of identity (Gatson and
Zweerink 2004; Hammersley and Treseder 2007).
Our historical preferences for face-to-face communities and intense, local sites of
interaction should not blind us to the fact that contemporary forms of communication
can transform our sense of what is ‘local’ into widely distributed networks, and that
‘communities’ can (and do) exist in many different forms. We should also not forget
that, as we noted earlier, one of the earliest inspirations for Chicago sociology was
The Polish Peasant. The starting point of this was the exchange of communications,
by letter, between Polish settlers in Chicago and relatives and friends living in Poland.
There is no intrinsic difference between the slow exchange of written communication
and the instant exchange of electronic messages. The density and speed of the latter,
however, can create a more intense sense of shared experience and of a shared social
world.
There do exist, therefore, ‘virtual communities’ that operate only in cyberspace, and
can only be studied ethnographically through the same medium. Cyber-ethnography
deals with social action and social organization in such virtual settings. Matthew
Williams (2006), for instance, has undertaken an ethnographic study of multi-user
domains (MUDs) in which participants can take on virtual identities (‘avatars’), and
interact with one another in virtual worlds. Williams paid particular attention to
mechanisms of social control within such domains, which are explicitly ‘policed’ and
are also subject to less formal forms of social control. Williams acted as a virtual
participant observer within these virtual worlds. His research amply demonstrates that
the proper preoccupations of ethnographic fieldwork can be brought to bear in such
settings. There are issues of identity, of community and of social control that are
vibrantly relevant to virtual worlds as much as they are to any other kind. The processes
138 Documents and other artefacts, real and virtual
of negotiation, interaction and identity-formation are just as ‘real’ also. There is no
fundamental distinction between ‘virtual’; and ‘real’ environments in social terms, and
research in the digital age needs to take account of that.8
Conclusion
In this chapter we have focused on the role that the study of documents, physical
objects and virtual communication can play in ethnographic work. These have not
always been given the attention they deserve within ethnographic work, the primary
emphasis usually being on data from interviews and participant observation in faceto-face
rather than virtual settings. We have tried to show the potential offered by
these sources of data in the context of work that also draws on the more usual sources.
In the next chapter we will look at issues surrounding the recording and preparation
of data for analysis.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Documents and other artefacts, real and virtual 139
8 For useful discussions of virtual methods, see Markham (2004) and Hine (2005).
7 Recording and organizing data
In previous chapters we have examined the main sources of data that ethnographers
use: observation and participation; participants’ oral accounts, both those that are
‘naturally occurring’ and those elicited in interviews; plus documents and artefacts of
various kinds. In this chapter we want to look at how these data may need to be
preserved or recorded, and organized, in preparation for the process of analysis. What
is involved here varies considerably depending upon the nature of the data. In the case
of documents and material objects, relatively little processing may sometimes be
required, though these data will need to be catalogued in ways that make them available
for analysis. By contrast, in the case of observing patterns of social interaction and
collecting participant accounts, the task of recording the data is much more significant
and time-consuming.
Technological developments have had a major impact upon the sort of data available
to ethnographers, on how it can be recorded, and on the task of organizing it so as to
facilitate analysis. CDs, DVDs and the internet represent relatively new sources
of data, and ethnographers have made increasing use of them. Going a little further
back into the past, the production of relatively cheap and portable audio-, and then
video-, recorders transformed much ethnographic data collection. And, in relation to
organizing data, the availability of cheap but very powerful personal computers, and
of word-processing programs and specialized software for handling textual data, has
been very significant. In recent years, the rise of digital audio-recorders and cameras,
and the increased capacity of computers to handle pictures and sound as well as text,
have opened up new possibilities.
We will begin by discussing what may be involved in processing documentary and
physical data, then examine the methods available for recording observational
and interview material, including transcription. Attention will be given to how the
analytic process must carry on alongside data processing. Finally, we will move on to
strategies for data organization.
Documents and other materials
We often need to collect and use documentary material from the research setting, as
well as employing that from elsewhere. As we have noted, this kind of data can take
a wide variety of forms (see Chapter 6). Some documents or material objects are freely
available and can be retained by the ethnographer for later use. This is often true, for
example, of such items as promotional material, guides, and circulars. Other documents
can be bought or otherwise acquired. For example, material on the internet can often
be downloaded and printed out.
Even when documentary sources are not produced in large numbers and are
unavailable in electronic form, the researcher may be able to produce copies for retention
(though here, and elsewhere, copyright issues need to be borne in mind). Photocopiers
are available in some settings where research is carried out, and the ethnographer may
be allowed access to them. The availability of relatively cheap and portable scanners
also eases the problem considerably. However, what is produced by this latter means
where text has been inputted into a word-processor, will almost always require
considerable tidying up, correction, and filling in; much depends here upon the nature
of the texts being scanned. It may also be possible to take digital photographs of posters
and of other short texts.
Sometimes, there is no alternative to note-taking in recording the data from
documents, or that pertaining to material objects. Here copying documents in toto may
be necessary, but it is not always the most effective recording strategy. While it may
reduce the danger of omitting something important, or losing the context, those
advantages have to be set against the cost in time. It may be necessary to make
judgements about which documents, and what parts of documents, are and are not to
be copied. One can index a document so that the relevant sections can be consulted,
as appropriate, at later stages of the research. This can be done relatively quickly, but
it may only be of value if there will be easy future return access to the documentary
sources concerned. One may also summarize relevant sections or copy them out by
hand: by summarizing one can cover much more material, thus releasing scarce time
for work of other kinds, but it will involve some loss of information and introduces
an additional layer of interpretation.
These three modes of note-taking – copying by hand, indexing, and summarizing
– are not mutually exclusive, of course, and each should be used according to the
likely future accessibility of the documents and the anticipated use to which the notes
will be put. Of particular importance here is whether they relate to central themes of
the study or provide background information. These considerations may vary across
different documents or even sections of documents, and judgements about them may
change over time. It should also be remembered that written notes need not necessarily
be made on the spot: where access to documents is restricted it may be more efficient
to read indexes, summaries, or relevant sections into a portable audio-recorder, where
this is possible, and then transcribe the recording later.
In the case of material objects, as with documents, copies or pictures may be readily
available. Alternatively, it may be necessary to take photographs or make videorecordings,
or perhaps to produce hand-drawn sketches. The use of photography and
video-recording is discussed below.
Recording observations and interviews: fieldnotes
Fieldnotes are the traditional means in ethnography for recording observational and
interview data. Originally, these were handwritten, but now they can sometimes be
inputted directly into a handheld or laptop. Whether fieldnotes can be written at all,
how, and covering what issues, depends on the nature of the research, the setting(s)
in which fieldwork occurs, and the role(s) taken on by the ethnographer.1
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 141
1 For detailed guidance about the production of fieldnotes, see Emerson et al. (1995). Sanjek (1990)
offers illuminating discussions of various aspects of fieldnotes, in the context of anthropology.
The writing of fieldnotes is not something that is (or should be) shrouded in mystery.
It is not an especially esoteric activity. On the other hand, it does often constitute a
central research activity, and it should be carried out with as much care and selfconscious
awareness as possible. A research project can be as well organized and as
theoretically sophisticated as you like, but with inadequate note-taking the exercise
could be like using an expensive camera with poor quality film. In both cases, the
resolution will prove unsatisfactory, and the results will be poor. Only foggy pictures
result.
Fieldnotes are always selective: it is not possible to capture everything. And there
is a trade-off between breadth of focus and detail. What is recorded will depend on
one’s general sense of what is relevant to the foreshadowed research problems, as well
as on background expectations (Wolfinger 2002). Moreover, as we shall see, the
character of fieldnotes may well change, in terms of focus and detail, over the course
of the research. The completion of fieldnotes is not an entirely straightforward matter,
then. Like most aspects of intellectual craft, some care and attention to detail are
prerequisites: satisfactory note-taking needs to be worked at. It is a skill demanding
repeated assessment of purposes and priorities, and of the costs and benefits of different
strategies. Thus, the standard injunction, ‘write down what you see and hear’, glosses
over a number of important issues. Among other things, the fieldworker will want to
ask what to write down, how to write it down, and when to write it down.
The making of fieldnotes has been part of the invisible oral tradition of craft
knowledge, and many who embark on their first project have to find their own way
of doing things. So let us try to deal with some of the practical questions raised above.
First, when to write notes? In principle, one should aim to make notes as soon as
possible after the observed action. Most fieldworkers report that, while they can train
themselves to improve recall, the quality of their notes diminishes rapidly with the
passage of time; the detail is quickly lost, and whole episodes can be forgotten or
become irreparably muddled. The ideal would be to make notes during actual participant
observation. But this is not always possible, and even when it is possible the opportunities
may be very limited. There are often restrictions arising from the social characteristics
of the research setting, as well as from the ethnographer’s role(s) in the field.
If the research is covert, then note-taking in the course of participation will often
be practically impossible. In many settings, participants do not carry around notebooks
(paper or electronic), and are not visibly engaged in a continual process of jotting
down notes. In many circumstances, such activity would prove very disruptive either
preventing ‘natural’ participation or generating distraction and distrust. It is unlikely,
for example, that Calvey (2000) could take copious notes while acting as a ‘bouncer’
on the door of a club.
In a few contexts, of course, writing may be such an unremarkable activity that
covert note-taking is possible. Rather surprisingly, Graham (1995) was able to take
fieldnotes while working on a car assembly line. She writes:
Luckily, the station that I worked during my last few months required that I carry
a clipboard for noting damage to the car bodies as they entered our area; hence it
was not unusual to see me jotting down notes. However, even before I was
assigned to that particular station, I was able to stand next to one of my parts
racks and take down notes. Each team member had paper and pencil for noting
142 Recording and organizing data
parts shortages and for keeping daily records on car numbers and various other
assigned duties. It was not out of the ordinary to observe people writing.
(Graham 1995: 16–17)
And observers in a covert study of patient life in mental hospitals found that they
could take notes, since staff simply took this as a further sign of their mental illness
(Rosenhahn 1973)!
Of course, note-taking is not always possible or easy even in overt research. To
some extent our comments concerning covert participation apply here as well. The
conduct of note-taking must be broadly congruent with the social setting under scrutiny.
In some contexts, however ‘well socialized’ the hosts, open and continuous note-taking
will be perceived as inappropriate or threatening, and will prove disruptive. In other
contexts fairly extensive notes can be recorded without undue disruption. A classic
example is Whyte (1981), who took on the role of secretary to the Italian Community
Club because it enabled him to take notes unobtrusively in their meetings.
However, even in situations where note-taking is a ‘normal’ kind of activity, such
as educational settings, care must be exercised if disruption is to be avoided. Olesen
and Whittaker’s (1968) research on student nurses is a case in point:
I feel it much easier to write when the students write, and listen when they do; I
have noticed that when I attempt to write when the students are not, I attract [the
tutor’s] attention and on a few occasions she seems to falter in what she is saying.
. . . Similarly when all the students are writing and I am not, but rather looking
at her, I again seem to ‘put her off’. And so it is that I’ve become a student,
sometimes slightly at the loss of my self-esteem when I find myself lazily inserting
a pencil in my mouth. (Fieldnotes: February, third year.)
(Olesen and Whittaker 1968: 28)
For a variety of reasons, many of the initial fieldnotes that ethnographers take are
jottings, snatched in the course of observed action. A common joke about ethnographers
relates to their frequent trips to the lavatory where such hasty notes can be scribbled
in private. Even the briefest of notes can be valuable aids in the construction of a more
detailed account. As Schatzman and Strauss suggest: ‘A single word, even one merely
descriptive of the dress of a person, or a particular word uttered by someone usually
is enough to “trip off” a string of images that afford substantial reconstruction of the
observed scene’ (Schatzman and Strauss 1973: 95). Moreover, it is important to record
even things that one does not immediately understand, because these may turn out to
be important later.
Even if it proves possible to make fairly extensive notes in the field, they – like
brief jottings – will need to be worked up, expanded on, and developed. Many social
activities have a timetable of their own, and it may prove possible within these to
match phases of observation with periods of writing up fieldnotes. For instance,
Atkinson’s work on haematologists in American and British hospitals was structured
by regular schedules of clinical ‘rounds’, ‘grand rounds’, ‘conferences’, ‘mortality and
morbidity reviews’, and similar occasions for medical talk. The pattern of data collection
was fitted into the rhythm of the hospital, which allowed for periods of time in the
canteen or the library, or back at the university, or at home, when detailed notes could
be constructed (Atkinson 1992a, 1995).
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 143
In other settings, the phasing of observation and writing will be much less straightforward
to organize, but there are usually times when participants are engaged in
activities that are not relevant to the research. At the very least they will usually sleep
at regular times and, at the risk of fatigue, notes can be written up then. Carey (1972)
reports a rare exception: that of ‘speed freaks’ (amphetamine users) who, under heavy
doses, stay awake for several days in a hyperactive state:
The peculiar round of life wherein people stay up for three, four or five days at
a time and then sleep for several days posed enormous practical difficulties for
the research. Our conventional commitments (family, friends, teaching
responsibilities) had to be put aside for a time so that we could adapt ourselves
more realistically to this youthful scene. As we became more familiar with this
particular universe, we developed a crude sampling plan that called for observations
at a number of different gathering spots, and this relieved us somewhat from a
very exacting round of life. If we were interested, however, in what happened
during the course of a run when a small group of people started shooting speed
intravenously, it meant that one or two fieldworkers had to be present at the
beginning and be relieved periodically by other members of the team until the run
was over. Fatigue was a constant problem and suggests that more than one
fieldworker is required in this type of research.
(Carey 1972: 82)
Clearly, in such cases, finding time to write up fieldnotes poses particularly severe
problems. The problem remains serious, however, even with less exhausting schedules.
But some time for writing up fieldnotes must always be set aside, especially where
this is the main means of data recording. There is no advantage in observing social
action over extended periods if inadequate time is allowed for the preparation of notes.
The information will quickly trickle away, and the effort will be wasted. There is
always the temptation to try to observe everything, and the consequent fear that in
withdrawing from the field one will miss some vital incident. Understandable though
such feelings are, they must, in most circumstances, be suppressed in the interests of
producing good-quality notes. Nevertheless, the trade-off between data collection and
data recording must be recognized and resolved, in accordance with the overall research
strategy and purpose. Thus, for example, the organization of periods of observation,
with alternating periods of writing and other work, must be done with a view to the
systematic sampling of action and actors (see Chapter 2).
It is difficult to overemphasize the importance of meticulous note-taking where this
is the main means of data recording. The memory should never be relied on entirely,
and a good maxim is ‘if in doubt, write it down’. Furthermore, it is absolutely essential
that one keep up to date in processing notes. Without the discipline of daily writing,
the observations will fade from memory, and the ethnography will all too easily
become incoherent and muddled.
What of the form and content of fieldnotes? As noted earlier, one can never record
everything; social scenes are truly inexhaustible in this sense. Some selection has to
be made. However, the nature of this is likely to change over time. During the early
days of a research project, the scope of the notes is likely to be fairly wide, and one
will probably be reluctant to emphasize any particular aspects. Indeed, one will probably
not be in a position to make such a selection of topics. As the research progresses,
144 Recording and organizing data
and emergent issues are identified, the notes will become more restricted and more
focused in subject matter. Moreover, features that previously seemed insignificant may
come to take on new meaning, a point that Johnson (1975) illustrates from his research
on social workers:
Gradually I began to ‘hear different things said’ in the setting. This happened
through a shift in attention from what was said or done to how it was said or done.
The following excerpts from the fieldnotes illustrate several instances of my
changing awareness. From the notes near the end of the sixth month of the
observations:
Another thing that happened today. I was standing by Bill’s desk when
Art passed by and asked Bill to cover the phone for a couple of minutes
while he walked through a request for County Supp over to Bess Lanston,
an EW supervisor. Now I don’t know how many times I’ve heard a
comment like that; so many times that it’s not even problematic any more.
In fact, it’s so routine that I’m surprised that I even made a note to
remember it. The striking feature about this is that in my first days at
Metro [the social work agency] I would have wanted to know all about
what kind of form he was taking over there, what County Supp was, why
and how one used it, got it, didn’t get it, or whatever, who and where
Bess Lanston was, what she did and so on. But all the time I’ve missed
what was crucial about such a comment, the fact that he was walking it
through. Before I would have only heard what he was doing or why, but
today, instead, I began to hear the how.
(Johnson 1975: 197)
So, as analytical ideas develop and change, what is ‘significant’ and what must be
included in the fieldnotes change. Over time, notes can also change in character, in
particular becoming more concrete and detailed. Indeed, the preservation of concreteness
is an important consideration in fieldnote writing. For most analytic purposes,
compressed summary accounts will prove inadequate for the detailed and systematic
comparison or aggregation of information across contexts or across occasions. As far
as possible, therefore, speech should be rendered in a manner that approximates to a
verbatim report and represents nonverbal behaviour in relatively concrete terms; this
minimizes the level of inference and thus facilitates the construction and reconstruction
of the analysis.
The actual words people use can be of considerable analytic importance. The ‘situated
vocabularies’ employed provide us with valuable information about the ways in which
members of a particular culture organize their perceptions of the world, and so engage
in the ‘social construction of reality’. Situated vocabularies and folk taxonomies
incorporate the typifications and recipes for action that constitute the stock-of-knowledge
and practical reasoning of the members. Arensberg and Kimball (1968) provide an
example from their study of interpersonal relations among family members in rural
Ireland around the turn of the twentieth century. They note how the terms ‘boy’ and
‘girl’ are used by parents to refer to their children irrespective of age, indicating the
sharply hierarchical character of parent–child relations: ‘Sociological adulthood has
little to do with physiological adulthood. Age brings little change of modes of address
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 145
and ways of treating and regarding one another in the relationships within the farm
family’ (Arensberg and Kimball 1968: 59).
The potential significance and detail of the connotations of such members’ terms
apply equally to the use of argot. American hospital speech includes the term ‘gomer’,
which is part of the rich and colourful vocabulary characteristic of most medical settings.
George and Dundes (1978) summarize its use:
What precisely is a ‘gomer’? He is typically an older man who is both dirty and
debilitated. He has extremely poor personal hygiene and he is often a chronic
alcoholic. A derelict or down-and-outer, the gomer is normally on welfare. He has
an extensive listing of multiple admissions to the hospital. From the gomer’s
standpoint, life inside the hospital is so much better than the miserable existence
he endures outside that he exerts every effort to gain admission, or rather
readmission, to the hospital. Moreover, once admitted, the gomer attempts to remain
there as long as possible Because of the gomer’s desire to stay in the hospital he
frequently pretends to be ill or he lacks interest in getting well on those occasions
when he is really sick.
(George and Dundes 1978: 570)
Of course, this brief account glosses over a wide range of uses and connotations
associated with this one folk term. In practice, the research worker will not be content
to generate such a composite or summary definition. The important task is to be able
to document and retrieve the actual contexts of use for such folk terms.
Kondo’s (1990) ethnography of the production of identities in Japan provides an
exemplary documentation of the terms and idioms of identity in various social contexts.
She examines, for instance, the idiomatic use of Shitamachi and Yamanote: literally,
different parts of Tokyo, used to convey different orientations, lifestyles and identities.
Likewise, she explores the subtle usages and connotations of ‘ie’ and ‘uchi’. Both
terms have flexible, context-dependent meanings. The former refers to the intergenerational
continuity of the group, the latter to the ‘ingroup’ as defined on any
particular occasion, whether ‘company, school, club, or nation’ (Kondo 1990: 141).
The ability to trace the social contexts of such idioms is dependent on the delicacy of
one’s ethnographic data: usage and social context must be identified with precision.
Making fieldnotes as concrete and descriptive as possible is not without its cost,
however. Generally, the more closely this ideal is approximated, the more restricted
the scope of the notes. Unless the focus of the research is extremely narrow, some
concreteness and detail will have to be sacrificed for increased scope. Whatever the
level of concreteness of fieldnotes, however, it is essential that direct quotations are
clearly distinguished from summaries in the researcher’s words, and that gaps and
uncertainties in the record are clearly indicated. If speakers’ original words cannot be
reconstructed adequately, then indirect speech may be used to indicate the style and
content. When we refer back to the notes there should be no ambiguity concerning
the ‘voices’ that are represented there. One should not have to puzzle over ‘Is that
what they themselves said?’ The observer’s own descriptive glosses should be kept
clearly distinct.
It is equally important that records of speech and action should be located in relation
to who was present, where, at what time, and under what circumstances. When it comes
to the analysis stage, when one will be gathering together, categorizing, comparing,
146 Recording and organizing data
and contrasting instances, it may be crucial that ‘context’ (participants, audience, setting,
etc.) can be identified.
Fieldnotes cannot possibly provide a comprehensive record of the research setting.
The ethnographer acquires a great deal more tacit knowledge than is ever contained
in the written record. He or she necessarily uses ‘head notes’ or memory to fill in and
recontextualize recorded events and utterances. One should not become totally wedded
to the fieldnotes, as if they were the sum total of available information. Despite the
scepticism of some commentators (for example, Agar 1980), however, the collection
and maintenence of fieldnotes remain a major method of ethnographic recording.
Up to now, we have discussed fieldnotes in relation to observation, but they may
also be used to record data from interviews. Sometimes, interviewees will refuse to
allow the discussion to be audio-recorded; sometimes the ethnographer may judge that
such recording will dissuade frankness or increase nervousness to an unacceptable
level. Where fieldnotes are relied on in interviews, much the same considerations apply
as in observation: decisions must be made about when they are to be written, what is
to be noted, and how. It is important, incidentally, to have some record of the
interviewer’s questions, as well as of the answers. Once again, reliance will most likely
have to be placed on jotted notes, and the dilemma of summarizing versus verbatim
reporting is just as acute. Similarly, note-taking in interviews can prove disruptive,
much as in the tutorial cited by Olesen and Whittaker (1968), with the interviewee
becoming self-conscious about what is being written down. Furthermore, the need to
take notes makes very difficult the kind of interviewing we advocated in Chapter 5.
Much of the interviewer’s attention will be taken up with recording what has been
said rather than thinking about it. Nevertheless, combining interviewing and note-taking
is a skill that can be learned with practice.
In light of the problems associated with writing fieldnotes, particularly in the context
of interviews, the advantages of audio-recording, and perhaps even video-recording,
are obvious. However, valuable though these are, they too involve difficulties and
limitations.
Digitally recording observations and interviews
It is very easy to demonstrate the major differences – in volume and detail – between
a permanent recording and an observer’s reconstruction of a strip of spoken action in
fieldnotes. Since the technology of permanent recording is now readily available, in
small and reliable formats, there are many possibilities for better quality data in key
respects. The uses of video or film, still photography, and audio-recording offer various
options for data collection and storage. At the same time, their use is not always
possible, may sometimes have unwelcome disadvantages, and can shape the process
of ethnographic work in ways that are undesirable. No means of data recording should
be simply adopted as a matter of routine: reflexive awareness is required here as much
as anywhere else.
Audio-recording
For the reasons we have suggested, if possible the ethnographer will usually wish to
audio-record interviews rather than to rely on fieldnotes. And, occasionally, the use
of video-recording may even be necessary for this purpose. There are also often
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 147
considerable advantages in using these methods in recording observational data; though
the practicalities and reactive effects must be remembered.
We should also note that, despite its advantages over fieldnotes, the use of audiorecording
does not provide a perfect and comprehensive record; and audio-recording
processes of social interaction that are ‘on the move’ can be challenging, to say the
least. In some cases background noise may make the recording virtually unintelligible.
Also, any recording will be highly selective. Decisions have to be made about whether
to use a single microphone or more than one, and where to place them; and these
decisions are likely to determine what talk is and is not recorded. Furthermore, not
only is non-verbal behaviour not captured in audio-recordings, but even such matters
as who is being addressed, or how the talk relates to any material objects being used,
are often not preserved. So, as we have already noted, the availability of audio-recording
does not remove the necessity for the writing of fieldnotes. Indeed, an overemphasis
on audio-recording can distort one’s sense of ‘the field’, by focusing data collection
on places where social processes can be recorded by this means, and concentrating
attention on the analysis of spoken action. It may be possible to overcome some of
these limitations by using visual recording devices, but these too have limitations and
drawbacks.
Photography and video-recording
Photography has long been used in anthropology; by contrast, it has only been employed
more recently, for the most part, in sociology and some other disciplines (Collier and
Collier 1986; Ball and Smith 1992, 2001; Harper 1993, 2000, 2006). Ethnographic
film also has a long history (see Marks 1995). Video-recording is more recent but is
now widely used. Despite this, there is still a tendency to think of written language
as the privileged medium of scholarly communication, and there are some tensions in
the use of visual materials in ‘a discipline of words’ (Ball and Smith 1992: 5).
The collection and use of visual materials has gained particular currency from the
increasing influence of cultural analysis in the social sciences. Visual materials, and
representations of material culture, are key aspects of such research. This includes the
semiotic analysis of photographs, films, and television programmes; the analysis of
architectural space; the ethnography of design, and the analysis of domestic goods and
spaces (see Pink 2004a). There is clearly opportunity for more traditional ethnographic
research to draw on and incorporate such analytic strategies. At the same time, decisions
have to be taken about whether visual data and the analytic strategies they make possible
are appropriate for the purposes of a particular inquiry, and whether use of the recording
devices concerned is viable in the field being investigated.
We are used to thinking of photography, film, and video-recording as producing
faithful, realistic images of the world about us; but such habits of our own culture
should not blind us to the fact that these forms of representation are partial and conventional.
Decisions have to be made about from where shots are to be taken, whether
the camera should be fixed or mobile, whether a single focus is to be adopted or
whether the focus should vary; and if so when and how. Where the position and
focus are not fixed, operation of the camera is likely to be full-time, so that it will be
difficult if not impossible to observe and take notes simultaneously. Yet complementary
observation and note-taking will almost certainly be necessary. Here, as elsewhere,
148 Recording and organizing data
contextual features will need to be documented, since by no means everything will be
‘in shot’, and some of what is in shot will be obscured. A team approach may well
be required, then.
Of course, ‘visual data’ do not have to be created ex nihilo by ethnographers. As
we noted in Chapter 6, there are many situations where visual materials are integral
to social action and social organization. Furthermore, reflexivity in the visual domain
can also be extended to providing one’s research hosts and informants with the means
to record their own visual materials. They can thus become co-producers of ‘data’.
The availability of cheap and easy-to-use cameras not only means that ethnographers
can use them extensively, but also means that participants can be encouraged to use
them as well.2
This may be a useful means for understanding how participants –
literally and figuratively – view their world (see, for example, Allan 2005; Darbyshire
et al. 2005). Such data can also be highly illuminating in making sense of how participants
consume spaces and objects, or how they move in given environments. Participantproduced
images can also be used as stimuli in the conduct of interviews.
Recent decades have transformed the use of visual materials: the advent of newer
digital technologies have created the opportunity for ethnographic work that genuinely
incorporates visual materials into an integrated ethnographic enterprise. In particular,
digital cameras and digital camcorders, together with the opportunity to scan visual
and documentary sources into digital databases, have transformed the opportunities
available to ethnographers, both ‘in the field’ and beyond. While we do not wish to
imply that ‘anyone can do it’ without training, or without an understanding of what
they are undertaking, it is clearly possible for visual materials to be gathered and
analysed in a way that was rarely possible for earlier generations.
Despite these developments in digital technology, transcription is still usually required
for the purposes of analysis where audio- or video-recording is employed; even though
the development of this technology now means that sounds and images from digital
audio-recorders, cameras, and camcorders can be incorporated directly into fieldnotes,
or even into research reports (see Chapter 9).
Transcription
Transcription is a time-consuming business, and this must be allowed for in planning
research. There are no hard-and-fast rules, but the ratio of transcribing time to recorded
time is always high: at least five to one. If the talk is multi-party, if there is background
noise, and if video-recordings are being transcribed, then it may take a great deal
longer. Also relevant is the kind of transcription required: the detailed formats often
used by discourse analysts are a great deal more time-consuming to produce than those
which are more common in ethnographic work.
We do not intend to provide detailed instructions as to the preparation of transcripts,
but a number of general precepts can be noted. In the first place, a decision needs to
be made about whether full transcription is necessary. An alternative is to treat the
audio- or video-recording as a document, indexing and summarizing much of it,
transcribing only what seems essential. This may save considerable time, but it risks
relevant material being overlooked – especially since what will be judged relevant will
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 149
2 Disposable film cameras are a good means for generating images here, though what they produce is
less easy to store and manage than digital photographs.
change over time, and what is on the recordings that has not been transcribed will
probably be forgotten.
Even where full transcription is to be carried out, as already indicated a decision
must be made about how detailed this should be. There are well-established conventions
for the preparation of transcripts. Some of these have been developed for the purposes
of conversation or discourse analysis. They use the typographical characters of the
standard keyboard and printer to represent some basic features of speech – such as
pauses, overlaps, and interruptions – as well as how words are pronounced, when the
speaker speeds up or slows down, where emphasis is placed, and so on. These will
be essential for some research purposes, less important for others. Decisions must be
made about what is to be represented in the transcription. Obviously, the more detailed
the transcription, the more time it will take to produce; furthermore, very detailed transcriptions
are by no means easy to read. At the same time, it is important that relevant
features of the talk are included; and even then recourse may well sometimes need to
be made to the original recording for purposes of analysis. The planning and conduct
of research using audio- and video-recorded data must therefore involve strategic
decisions about the kind of data to be collected, and the degree of detail to be preserved
in the transcription.3
Of course, fixed decisions do not have to be made at the start,
and initial judgements may need to be reviewed.
Transcribing the video part of video-records is even more challenging, and timeconsuming
than dealing with audio-recordings. Here, conventions have been developed
for indicating direction of gaze, so as to coordinate this with the pattern of talk. Sketches
and screenshots of physical actions taking place alongside talk have also been used.
Other researchers have provided written commentaries on what is happening instead
of, or in addition to, these devices. Once again, including information of this kind
multiplies the time required to transcribe data. However, it allows non-verbal aspects
of social interaction and situations to be taken into account in a way that would otherwise
be difficult. Here, as elsewhere, what kind of transcription is required depends upon
the phenomena being studied and the purposes of inquiry. At the same time, the costs
in terms of time and resources must be taken into account.4
Analytic notes, memos, and fieldwork journals
While one is reading documents, writing up fieldnotes, or transcribing audiovisual
materials, promising analytic ideas often arise. It is important to make note of these,
as they may prove useful later when analysing the data. It is essential, though, to do
this in a way that retains a clear distinction between analytic notes, on the one hand,
and both the accounts provided by participants and the researcher’s own descriptions
of actions and situations, on the other. For example, analytic notes may be placed in
square brackets or double parentheses; or, as we shall see, they may be put in a fieldwork
journal.
It is also important to engage in the regular review and development of analytic
ideas through writing analytic memos. These are not fully developed working papers
150 Recording and organizing data
3 For further discussion of transcription, see Mishler (1991); Atkinson (1992b); West (1996); Lapadat
(1999); Poland (2002).
4 For discussions of transcription issues relating to video data, see Goodwin (1981, 2001); Heath (1997,
2004) and Heath and Hindmarsh (2002).
but occasional written notes whereby progress is assessed, emergent ideas are identified,
research strategy is sketched out, and so on. We quoted extracts from an analytic memo
in Chapter 2 (pp. 27–8). It is all too easy simply to allow one’s fieldnotes and other
types of data to pile up day-by-day and week-by-week. The very accumulation of
material usually imparts a satisfying sense of progress, which can be measured in
physical terms as notebooks are filled, interviews completed, periods of observation
ticked off, or different research settings investigated. But it is a grave error to let this
work build up without regular reflection and review. Under such circumstances the
sense of progress may prove illusory, and a good deal of the data collection could be
unnecessarily aimless.
As we have emphasized, the formulation of precise problems, hypotheses, and an
appropriate research strategy is an emergent feature of ethnography. This process of
progressive focusing means that the collection of data must be guided by the developing
clarification of topics for inquiry. The regular production of analytic memoranda will
force the ethnographer to go through such a process of explication. Ideally, every
period of observation should result in processed notes and the reflexive monitoring of
the research process. As the memoranda accumulate, they will constitute preliminary
analyses, providing the researcher with guidance through the corpus of data. If this is
done there is no danger of being confronted at the end of the fieldwork with an
undifferentiated collection of material, with only one’s memory to guide analysis.
The construction of analytic notes and memos therefore constitutes precisely the
sort of internal dialogue, or thinking aloud, that is the essence of reflexive ethnography.
Such activity should help one avoid lapsing into the ‘natural attitude’ and ‘thinking
as usual’ in the field. Rather than coming to take one’s understanding on trust, one is
forced to question what one knows, how such knowledge has been acquired, the degree
of certainty of such knowledge, and what further lines of inquiry are implied.
As already noted, analytic notes and memoranda may be appended to the daily
fieldnotes, or they can be incorporated into yet another form of written account, the
fieldwork journal. Such a journal or diary provides a running account of the conduct
of the research. This includes a record not only of the fieldwork, but also of the
ethnographer’s own personal feelings and involvement (Coffey 1999). The latter are
not simply the basis for gratuitous introspection or narcissistic self-absorption. As we
point out elsewhere in this book, feelings of personal comfort, anxiety, surprise, shock,
or revulsion are of analytic significance. In the first place, our feelings enter into and
colour the social relationships we engage in during fieldwork. Second, such personal
and subjective responses will inevitably influence one’s choice of what is noteworthy,
what is regarded as strange and problematic, and what appears to be mundane and
obvious. One often relies implicitly on such feelings; therefore, their existence and
possible influence must be acknowledged and, if possible, explicated in written form.
Similarly, feelings of anxiety can pose limitations on data collection, leading to a
restricting tunnel vision. One of us (Atkinson 1992a) has reflected on how his personal
feelings about general medicine and surgery clearly influenced the nature and balance
of his published research on medical education.
There is a constant interplay between the personal and emotional on the one hand,
and the intellectual on the other. Private response should be transformed, by reflection
and analysis, into potential public knowledge. The fieldwork journal is the vehicle for
this kind of transformation. At a more mundane level, perhaps, the carefully made
fieldwork journal will enable the conscientious ethnographer painstakingly to retrace
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 151
and explicate the development of the research design, the emergence of analytic themes,
and the systematic collection of data. The provision of such a ‘natural history’ of the
research can be a crucial component of the complete ethnography.
Up to now, we have focused on data processing, emphasizing the decisions involved,
and how this should be integrally related to the process of data analysis. In the remainder
of the chapter we will examine the means that can be employed for the storage, indexing,
and retrieval of ethnographic data.
Data storage, indexing and retrieval
Ethnographers usually store observational data records chronologically. Likewise,
interview transcripts and the like are normally kept as complete records of the individual
interview, and stored in order. However, the process of analysis will often require
active reorganization of the data into themes and categories – often breaking the texts
up into discrete chunks or segments and identifying them in accordance with an indexing
or ‘coding’ system. (This is less common in conversation and discourse analysis, where
the focus is often on local patterns.)
The coding of the data in terms of categories provides an important infrastructure
for later searching and retrieval. It can also play an active role in the process of
discovery, as the Webbs noted in one of the earliest methodological texts:
It enables the scientific worker to break up his subject-matter, so as to isolate and
examine at his leisure its various component parts, and to recombine the facts
when they have been thus released from all accustomed categories, in new and
experimental groupings.
(Webb and Webb 1932: 83)
Moreover, the selection of categories as a basis for organizing the data can be of some
significance, as a later sociologist notes, referring to what became a classic ethnographic
study:
As I gathered my early research data, I had to decide how I was to organize the
written notes. In the very early stage of exploration, I simply put all the notes, in
chronological order, in a single folder. As I was to go on to study a number of
different groups and problems, it was obvious that this was no solution at all.
I had to subdivide the notes. There seemed to be two main possibilities. I could
organize the notes topically, with folders for politics, rackets, the church, the
family, and so on. Or I could organize the notes in terms of the groups on
which they were based, which would mean having folders on the Nortons, the
Italian Community Club, and so on. Without really thinking the problem through,
I began filing material on the group basis, reasoning that I could later redivide
it on a topical basis when I had a better knowledge of what the relevant topics
should be.
As the material in the folders piled up, I came to realize that the organization
of notes by social groups fitted in with the way in which my study was developing.
For example, we have a college-boy member of the Italian Community Club saying:
‘These racketeers give our district a bad name. They should really be cleaned out
of here.’ And we have a member of the Nortons saying: ‘These racketeers are
152 Recording and organizing data
really all right. When you need help, they’ll give it to you. The legitimate
businessman – he won’t even give you the time of day.’ Should these notes be
filed under ‘Racketeers, attitudes toward’? If so, they would only show that there
are conflicting attitudes toward racketeers in Cornerville. Only a questionnaire
(which is hardly feasible for such a topic) would show the distribution of attitudes
in the district. Furthermore, how important would it be to know how many people
felt one way or another on this topic? It seemed to me of much greater scientific
interest to be able to relate the attitude to the group in which the individual
participated. This shows why two individuals could be expected to have quite
different attitudes on a given topic.
(Whyte 1981: 309)
Whyte’s comments here emphasize the importance of anticipating how the data might
be used. Equally important, no system of filing or coding and retrieval can ever remove
the necessity to remain sensitive to the immediate social context of speech and action.
The allocation of data to categories in ethnography has usually differed from the
kind of coding typical in quantitative research, including content analysis (Krippendorff
1980; however see also Franzosi 2004). In ethnographic coding, there is no requirement
that items of data be assigned to one and only one category, or that there be explicit
rules for assigning them:
We code [the fieldnotes] inclusively, that is to say if we have any reason to think
that anything might go under the heading, we will put it in. We do not lose anything.
We also code them in multiple categories, under anything that might be felt to be
cogent. As a general rule, we want to get back anything that could conceivably
bear on a given interest. . . . It is a search procedure for getting all of the material
that is pertinent.
(Becker 1968: 245)
Indeed, Lofland (1971) argues that in the case of analytic categories it pays to be
‘wild’, to include anything, however long a shot.
The identification of categories is central to the process of analysis (although it should
not be confused with analysis per se). As a result, the list of categories in terms of
which the data are organized generally undergoes considerable change over the course
of the research. In particular, there is typically a shift towards more abstract categories
as the work develops (see Chapter 8).
In the past, ethnographers manipulated their data by means of the physical indexing
and sorting of precious manuscript and typescript texts. Nowadays, many ethnographers
use personal computers to facilitate the storage, indexing and retrieval of textual data
for analytic purposes. To a considerable extent, the computer software for ethnographic
data storage and retrieval recapitulates the procedures associated with earlier, manual
approaches. We shall comment on manual techniques before going on to discuss
computer-based applications. It is important not to assume that all ethnographic data
must now be stored and searched on computer. For many researchers there will still
be a place for simple manual procedures.
Organizing and reorganizing the data in terms of categories can be done in a number
of ways. The simplest is ‘coding the record’. Here data are coded, that is, assigned to
categories, on the running record itself (or, better, a copy of it). Comments relating
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 153
the data to descriptive or analytic categories are written in the margin, on the reverse,
or on interleaved pages, depending on the format of the data themselves. This is quick,
and preserves the sense of ‘reading’ the data. It is not, however, well adapted to
subsequent procedures of searching and retrieving data segments. In a more sophisticated
version of this strategy, an analytic index is produced. Here each data segment is
indexed under a developing set of headings, stored on index cards or in a simple
database program. Identically or similarly coded segments can thus be found in the
original hard copy of the data when required. However, this does not make their
comparison very easy.
An alternative method of data organization, used by many ethnographers in the past,
is physical sorting. Multiple copies of the data are made, and each segment of the data
is stored in folders representing all the categories to which it is deemed relevant. With
this approach, ethnographers can find all the data collected together when they come
to analyse and write up a particular theme. At the same time, the physical storage of
multiple copies has limitations: not least the time taken to produce copies and the sheer
space requirements of a large and complex data set.
There is little doubt that when handling small data sets, such simple procedures can
serve users just as well as more complex data-handling software. However, the capacities
of word-processing have become so commonplace that most of us now take for granted
the remarkable flexibility that it provides in storing, manipulating, sharing and searching
textual materials. We can input and store vast amounts of written and visual materials.
Moreover, it is possible to use wordprocessing programs to copy segments of data
and paste them into folders relevant to different analytic themes. However, ethnographers
now have available to them specialized software for the management of texts
and pictures. Initially, this was restricted to following more or less the same pattern
as the coding and retrieval involved in physical sorting. In other words, it relied on
the procedure of marking segments of data with codes, and then using the capacity
of the software rapidly to search the data set(s) for all segments tagged with the same
code. In this way, instances in the data can be collated and aggregated under thematic,
analytic headings.
However, over the course of recent years, the software has evolved to develop
functions that reflect more fully the array of opportunities that are presented by digital
technology. Over the same period, digital technologies themselves have become more
widespread, more affordable, and better adapted to the needs of field researchers. There
is now a specialist literature that deals directly with the use of CAQDAS (Computer
Aided Qualitative Data Analysis Software), and that provides practical advice on how
to use it. In the course of this chapter, we shall discuss the use of such analytic software
in general terms, rather than attempting to offer the kind of user’s guide that other
publications readily provide (see, for instance, Lee and Fielding 2004).
One advantage of coding software over manual methods was the capacity to search
the data comprehensively, avoiding partial or improperly selective sampling of events
and examples. The fact that such searches were virtually instantaneous, encouraged
more thorough examination of the data. The advantages of CAQDAS coding were
more far-reaching, however. Manual searching is not an effective way of combining
multiple codes. The great virtue of code-and-retrieve software capabilities lay in the
fact that one could attach several codes to particular stretches of data, with segments
overlapping one another, and codes nesting within one another. This facilitates more
complex forms of analysis.
154 Recording and organizing data
In this way, the complexity of coding could reflect the sort of complexity that one
would hope to find in qualitative data. After all, interview informants do not talk about
just one topic at a time. Equally, observed social action does not present itself to the
ethnographer one theme at a time. The social world – and therefore our data – is
complex in its enactments. Data, therefore, need to be coded densely, and to be faithful
to the many possible topics and themes that are discoverable. It has been the great
strength of CAQDAS not only that data can be coded in complex ways, but also that
the data can then be searched and retrieved in ways that respect that complexity. Even
elementary Boolean logic (X and Y, X not Y, etc.) can permit the analyst to conduct
sophisticated searches. Equally, the capacity to treat code words as synonyms, and
hence to combine codings, can help the analyst build up from detailed, concrete
codings to more generic, analytically productive concepts.
The basic ideas and procedures of CAQDAS embodied the ideas about ethnographic
analysis that were reflected in grounded theorizing – or at least the versions of grounded
theorizing that were widely current at the time it was originally developed (Glaser and
Strauss 1967; Strauss 1987).5
Indeed, the authors of the software were in some cases
quite explicit about the conceptual relationship between the two. Although it is always
unwise to imply that any computer software exercises agency in predetermining how
analysts conceptualize their work, and although senior commentators in the field stress
the independence of software from any single analytic paradigm, there is little doubt
that the parallel evolution of the software with other methodological writing has meant
that there is a mutual influence. For many researchers and their students, the development
of CAQDAS has meant the implementation of a ‘grounded theory’ strategy, where
grounded theorizing itself has been largely equated to an analysis based on data-coding.6
The availability of software has also allowed the searching of large data sets for
‘indigenous’ terms in the texts of fieldnotes or transcripts. In other words, it is possible
to find every example of the use of a given word by participants or informants, or the
presence of such a term in one’s own (or another researcher’s) fieldnotes. Again, the
use of Boolean algebra allows the grouping of terms together thereby creating complex
search strategies. The use, therefore, of bulk-indexing software, that had originally
been developed for tasks like indexing legal and other official documents, was a
potentially useful resource for ethnographic and other qualitative analysts. While the
commercial qualitative analysis software often has such a facility incorporated within
it, specialized text-retrieval software – such as Wordcruncher or ZiINDEX – should
remain part of the digital resource-base of the contemporary ethnographer.
One key development in CAQDAS software has been an increasing emphasis on
so-called ‘theory-building’. Whereas code-and-retrieve is predicated on establishing
the relationship between codes and data, theory-building goes beyond this, the software
being designed to show relationships among codes. Even so, as Atkinson, among others,
has pointed out, a great deal of what passes for ‘analysis’ in current ethnographic
research is often based on a ‘culture of fragmentation’. That is, it is assumed that the
basis of analysis rests in the work of segmenting and disaggregating data. It has been
described in terms of decontextualizing data from its original location (e.g. in the
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 155
5 Grounded theorizing is discussed in Chapter 8.
6 This is not the place for arcane disputes, but we should note that Lee and Fielding (1996) have challenged
this observation. However, their grounds for doing so are not watertight, since they rely for evidence
only on explicit references to the use of grounded theorizing.
fieldnotes) and recontextualizing them into analytically driven categories. Now this is
unexceptional in and of itself, and it would be hard to quarrel with as one productive
way of managing data. But this is not the only way of thinking about data and it is
certainly not the only way of thinking about analysis and the generation of theory.
Approaches that are more sensitive to the immediate contexts of action, and that resist
the treatment of instances as being susceptible to simple aggregation, are characteristic
of many forms of discourse and narrative analysis. While code-based analyses do not
necessarily violate those analytic considerations, they can readily lead to an overemphasis
on decontextualized instances.
Kelle (2004: 485) has pointed out that many of the functions of contemporary
CAQDAS are under-used by most researchers. ‘In the ongoing discussions of the 1990s
it became apparent that enhanced strategies for hypothesis testing, quantitatively oriented
typology construction and “qualitative comparative analysis” were only seldom applied
by CAQDAS users’ (see also Dotzler 1995). Consequently, in evaluating the actual
impact of CAQDAS on research practice, it is important to distinguish between the
claims entered on its behalf by developers and advocates, the functional possibilities
inscribed in the software, and the actual uses to which it is put. We should not be
surprised to discover that, in practice, software has not led inexorably to more
sophisticated and complex analytic work with ethnographic data. Our own experience
suggests that many students and researchers are in practice using software still to
perform the sort of sorting and retrieval tasks that were done manually in previous
generations. This is not a problem in itself, but it detracts from overheated claims to
the effect that the software is encouraging ‘theory-building’.
It is important to realize that CAQDAS is by no means confined to the management
of textual materials like transcripts and fieldnotes. It is often also possible to include
and code visual and sound data. Indeed, there is no reason in principle to exclude any
data that can be rendered digitally from inclusion in the standard software strategies.
Furthermore, some packages also provide for combining quantitative and qualitative
data.
Conclusion
While it is impossible to render explicit all the data acquired in fieldwork, every effort
must be made to record it. Memory is an inadequate basis for subsequent analysis. Of
course, most kinds of data recording are necessarily selective and involve some
interpretation, however minimal. There is no set of basic, indubitable data available
from which all else may be inferred. What is collected or recorded, and how, will
depend in large part on the purposes and priorities of the research, and the conditions
in which it is carried out. Moreover, in using various recording techniques we must
remain aware of the effects their use may be having, both on participants and on our
analytic interpretations; and be prepared to modify the strategy in light of this. Similarly,
there is no finally correct way to store information or to retrieve it for analysis. The
various systems – including currently available software – differ in appropriateness
according to one’s purposes, the nature of the data collected, the facilities and finance
available, the size and scope of the research project, as well as personal convenience.
As with other aspects of ethnographic research, then, recording, storing, and retrieving
data must be viewed as part of the reflexive process. Decisions are to be made, monitored,
and – if necessary – remade in the light of methodological, practical and ethical
156 Recording and organizing data
considerations. At the same time, however, these techniques play an important role
in promoting the quality of ethnographic research. They provide a crucial resource in
assessing typicality of examples, checking construct-indicator linkages, searching for
negative cases, triangulating across different data sources and stages of the fieldwork,
and assessing the role of the researcher in shaping the nature of the data and the
findings. In short, they facilitate – but should not determine – the process of analysis,
a topic to which we turn in the next chapter.
1EEE
2
3
4
5
6
7
8
9
10
1
2
EEE3
4
5
6
7
8
9
20
1
2
3
4
5EEE
6
7
8
9
30
1
2
3
4
5
6
7
8
9
40
1
2
3
4
5
6
7
41EEE
Recording and organizing data 157