INTERNATIONAL JOURNAL OF TOURISM RESEARCH Int. J. Tourism Res. 3, 301312 (2001) DOI: 10.1002/jtr.

332

Ethnography: A Research Method in Practice

Catherine Palmer* University of Brighton, School of Service Management, 49 Darley Road, Eastbourne, East Sussex, UK ABSTRACT This article discusses the ethnographic methods employed for a doctoral study focusing upon heritage tourism and English national identity. It explains the fundamental principles behind the research and how the eld work was conducted. In addition, it examines the issues of practicality which emerged during the research process and discusses how these were resolved. The article concludes by highlighting the importance of ethnography as a tool for understanding and interpreting visitor behaviour. Copyright # 2001 John Wiley & Sons, Ltd. Keywords: ethnography; anthropology; heritage; national identity; behaviour INTRODUCTION nthropological eld research is traditionally associated with the practice of participant observation, pioneered by such authors as Malinowski (1922), where the anthropologist lives with the local community and compiles detailed accounts of the life, traditions and cultural practices of the local people. Depending upon the purposes of the research, information is obtained from a range of individuals, such as local inhabitants, employees of particular organisations and members of groups or associations. More specialised data also can be obtained from `key' sources. For example, government departments and local institutions. Overall, Hammersley and Atkinson (1983, p. 2) state that the ethnographer is one who:
* Correspondence to: Catherine Palmer, University of Brighton, School of Service Management, 49 Darley Road, Eastbourne, East Sussex, UK.

participates, overtly or covertly, in people's daily lives watching listening. collecting whatever data are available to throw light on the issues with which he or she is concerned. it (ethnography) also bears a close resemblance to the routine ways in which people make sense of the world in everyday life. These writings or diaries, referred to as ethnographic accounts, provide the basis for the interpretation of a culture and in the case such as Levi-Strauss, a forum for philosophical discourse and debate (see Levi-Strauss, 1972, 1986; Geertz, 1973; Clifford, 1988). This concept of the researcher observing, or living as a part of the community remains a feature of contemporary anthropological research (see Rose, 1990). Ethnography is thus grounded in eld work that enables the researcher to study the activities of people in their everyday settings. Hence the ethnographer is concerned with understanding the world from the point of view of those studied. This requires a stance that is uncritical of the behaviour and activities observed and moreover, one which remains open to elements that cannot be codied at the time of the study (see Baszanger and Dodier, 1997). According to Nunez (1989) this ethnographic routine of participant observation combined with written diaries or eld notes and interviews also can be applied to the study of tourism. Tourism and anthropology. Nash (1996) states that the rst serious study of tourism by an anthropologist was that by Nunez of week-end tourism in a Mexican village, back in the 1960s. Since that time tourism has evolved from a frivolous activity into a serious business, both commercially and academically. Furthermore, the legitimation of tourism as a suitable topic for anthropological enquiry has grown steaCopyright # 2001 John Wiley & Sons, Ltd.

302 dily since Nunez's pioneering study (see Nash, 1981; Nash and Smith, 1991). This has resulted in the emergence of several key academic debates about the subject, in both journal articles and text book form (Graburn, 1983; Crick, 1989; Selwyn, 1994; Nash, 1996; Burns, 1999). According to Graburn (1983) the anthropology of tourism has concentrated on two main areas. First, the study of tourism and the nature of tourism itself and second, the study of the social, cultural and economic impact of tourism on host societies and populations; with the relationship between cultural contact and cultural change being the primary focus for attention. Indeed, Valene Smith's seminal text Hosts and Guests (1989, originally published in 1977), contains a variety of anthropological case studies concerning the impact of tourism on the host community. Prior to this work one of the rst major text books to deal with the question of tourism within the social sciences was Dean MacCannell's (1976) The Tourist: a New Theory of the Leisure Class. This set out to present a critique of tourism as a feature of modernity. This critique involved a new kind of ethnographic report, one that clearly demonstrated that ethnography could be redirected away from peasant societies and applied to the modern world. As MacCannell (1989, p. 5) states `by following the tourists, we may be able to arrive at a better understanding of ourselves'. To this end he undertook to study groups of tourists in Paris, to follow them around on their sightseeing trips, sometimes to join in, sometimes just to observe, keeping detailed notes of his experiences. MacCannell is acknowledged as having made a signicant contribution to the study of tourism. Although not without his critics at the time (see Schudson, 1979) he was, along with Nunez, one of the rst social scientists to apply anthropological research techniques to the eld of tourism. Whereas the process of participant observation within anthropology may have at one time concentrated on an `expose' of the primitive, the work of MacCannell (1992) illustrates its relevance and applicability to the modern world and to tourism. Indeed, ethnography and participant observation do not have to entail long periods
Copyright # 2001 John Wiley & Sons, Ltd.

C. Palmer of time living with another community, it all depends on the nature of the research topic and the amount of detail required. Taking a leaf out of MacCannell's book the focus of attention can be the tourists themselves, their patterns of behaviour, responses and interaction with both themselves and the local community. Furthermore, tourist behaviour does not only have to be studied in relation to the inhabitants of a destination. It also can be viewed in terms of the actual attraction, such as a monument, natural scenic phenomenon, museums, historic sites and so on (see McManus, 1987; Walter, 1996). The study underpinning the research methods discussed here aimed to draw together such research possibilities by examining the relationship between tourists and specic heritage attractions. So as to arrive at a better understanding of what MacCannell (1992) refers to above as `ourselves'. Specically in this case `ourselves' as a nation and `ourselves' as an identity. Purpose of the research. The overall purpose of the study was an investigation into the relationship between heritage tourism and English national identity. The general theoretical framework for this lay primarily within that concerned with the birth of nations, nationalism and the emergence of national consciousness (see Kedourie, 1960; Gellner, 1983; Smith, 1983, 1986, 1991; Anderson, 1991; Connor, 1993). The specic focus of the study was an examination of both the components of Englishness and the social processes by which identity and belonging were communicated and maintained. It concentrated on one particular articulation of nation-ness, that of heritage tourism and was centred around three tourist sites. Battle Abbey in Sussex (the location for the Battle of Hastings of 1066), Hever Castle in Kent (the former home of Anne Boleyn, one of the six wives of Henry VIII) and Chartwell (the home of Sir Winston Churchill), also in Kent. Battle Abbey was chosen because the Battle of Hastings represents the last time England was invaded. The date 1066 and the battle associated with it are consistently referred to as the origin of a nation, in books, documentaries (see Phillips, 1989; Starkey, 1998) and tourist promotions (see Hastings Borough
Int. J. Tourism Res. 3, 301312 (2001)

Ethnography: A Research Method in Practice Council, 1997). Hever Castle's inclusion was based upon its association with the Tudor period generally, and with the romance between Anne Boleyn and Henry VIII in particular. Indeed, the Tudor period is often taken to be the start of modern England (Elton, 1974). Thus notions of what it means to be English, conceptions of a national identity, are often rooted in Tudor times (Storry and Childs, 1997). Finally, Chartwell was chosen because of its association with recent history, with World War Two and Winston Churchill, an event and a person intimately bound up with notions of patriotism and the foundations upon which such a sentiment is based. If men and women were prepared to die for their country then there must have been some understanding of what it was they were trying to protect. There were several issues deemed central to the study, and it is important to detail these here as they directly underpin the research methods under discussion. Firstly, the tourism industry's reliance upon the material and symbolic resources of the nation's cultural heritage as a means of both dening and promoting identity and belonging. Secondly, how individuals, as tourists and employees, related to and understood the sites. In other words, how the sites communicated a sense of Englishness the social mechanisms by which identity was conveyed and if they did, then what aspects were being displayed. The third key issue, and that which signicantly informed the other two, concerned the symbolic transmission of the nation's cultural heritage. As McCrone (1992, p. 195) states `the question to ask is not how best do cultural forms reect an essential national identity, but how do cultural forms actually help to construct and shape identity'. Any study of the ways in which a nation's cultural heritage underpins national identity cannot ignore the fact that cultural icons, practices and artefacts have a symbolic, as well as a physical function (see Malinowski, 1944). As such, the cultural framework within which this study was set is best illustrated by Geertz's (1973, p. 5) understanding of the concept of culture: the concept of culture I espouse is essentially a semiotic one. Believing with
Copyright # 2001 John Wiley & Sons, Ltd.

303 Max Weber, that man is an animal suspended in webs of signicance he himself has spun, I take culture to be these webs, and the analysis of it to be therefore not an experimental science in search of law but an interpretive one in search of meaning. The above provides a brief overview of the issues supporting the selection of the ethnographic research methods. The discussion now turns to an examination of these methods. The research methods Given that the primary focus of the study was the cultural representation of nation-ness, the symbolic display of `ourselves', this involved analysing the ways in which individuals both related to and with the cultural locations, practices and artefacts on view. As Geertz (1973, p. 17) states `behaviour must be attended to, and with some exactness, because it is through the ow of behaviour or, more precisely, social action that cultural forms nd articulation'. The overall aim, then, was to use a range of methods that enabled the researcher to reconstruct patterns of common-sense thinking by searching for underlying themes. Hence the methods were not designed to reveal divisions between different groups of respondents but rather to uncover the thought processes behind visitor reactions to each site. A similar strategy was used by Billig (1992, p. 19) in his study of attitudes towards the British Royal Family, where he argues that `ideology is to be found in those patterns of common-sense thinking which cut across class, age and gender distinctions. This means looking for commonalities in what is said (and) in what is not said'. Indeed, Geertz (1973, p. 24) argues that, `the whole point of a semiotic approach to culture is to aid us in gaining access to the conceptual world in which our subjects live '. As such, a key feature of the research methods was the identication and analysis of the structures of signication inherent in the three sites under investigation. Battle Abbey, Hever Castle and Chartwell represent specic social settings within which individuals interact with themselves and with others. It was through the observation and recording of such
Int. J. Tourism Res. 3, 301312 (2001)

304 interactions that the visitors `conceptual world' was to be accessed. Accessing the conceptual world. Although the participation in, and the recording of, the social settings in which people live is a key principle of ethnographic research, it was important not to lose sight of the need to remain objective throughout the period of investigation. Bearing this in mind triangulation, or multiple methods (see Denzin, 1970, 1999) were necessary if reliable information was to be gathered for analysis. As Hammersley and Atkinson (1983, p. 198) illustrate: data-source triangulation involves the comparison of data relating to the same phenomenon but deriving from different phases of the eldwork. This is very time consuming but, besides providing a validity check, it also gives added depth to the description of the social meanings involved in a setting. In addition to this Baszanger and Dodier (1997) have argued that a study becomes ethnographic when the researcher is careful to connect the facts that s/he observes with the specic features of the backdrop against which these facts occur. It is thus important not to ignore the general historic period in which each site is located, as this affects the analysis and interpretation of the ndings. Bearing all these points in mind the following methods were used. (1) The review and analysis of books, documents, guides and other relevant published and unpublished material relating to the three sites some of this was obtained from the library of the Sussex Archaeological Society. (2) Tape recorded interviews with key people associated with the sites. For example, employees, managers, members of related associations and relevant individuals from the organisations that owned the sites. For example, English Heritage and the National Trust, (3) Participant, covert, observation of visitors on a number of different occasions/times with written notes being made of behaviour and conversations.
Copyright # 2001 John Wiley & Sons, Ltd.

C. Palmer (4) Tape recorded interviews with a random selection of visitors at each site; every tenth person or group was approached. A total of 200 people were interviewed. The total number of visitor interviews was based upon similar studies, such as that of Billig (1992) who interviewed 175 people on their feelings about the British Royal Family. Here Billig (1992, p. 16) stated that his survey size represented, ` a comparatively large sample for a qualitative study: i.e. a study which seeks to analyse the meanings of what people say, as opposed to analysing statistically the results from standard questionnaires'. Likewise, Ireland's (1990) study of visitor experiences at Land's End in Cornwall was based upon administered questionnaire interviews with 200 people. Both these studies relied upon one research method, whereas this study was designed around four interdependent methods of data collection. This illicited a rich and varied amount of information, thus enabling a more in-depth analysis to be carried out than would have been the case had fewer methods been used. The interviews with key people and tourists were based upon set themes that had emerged from the review and analysis stage. Thus, the same questions were not asked at each site. Moreover, the interviews with the tourists, and where appropriate with the key people, were conducted along the line of conversations. The intention here was to put people at their ease and so increase the possibility of obtaining information that may more readily indicate underlying feelings, assumptions and beliefs. Cohen (1984, p. 226) provides an apt illustration of the way in which these interviews were conducted: the proper ethnographic interview is a conversation in which ethnographers risk the appearance of naivety and ignorance in order continually to satisfy themselves that they have understood what is being said the conversations are instruments for stripping away the ballasts of expectation and assumption The primary research was started in the summer of 1995. As this work was undertaken on a part-time basis there were restrictions in
Int. J. Tourism Res. 3, 301312 (2001)

Ethnography: A Research Method in Practice terms of when this took place and how long could be spent `in the eld'. The nal stage of the eldwork was completed in April 1997. The following breakdown outlines the key stages of the research. (1) The review and analysis of books, documents, guides and other relevant published and unpublished material relating to the three sites. Preliminary assessment of each site. (2) The interviews with key people, tourists and the visitor observations. Although these are presented as separate stages there was obviously overlap between them in terms of the writing up of information and its preliminary analysis. The only site where employee interviews did not take place was Hever Castle. This was at the express wish of the Managing Director, who felt that staff were too busy and that such a process would interfere with and detract from their duties. The administrators of both Battle Abbey and Chartwell did give their consent for the researcher to talk to the guides. This was an interesting issue as it highlighted the different approaches adopted by the sites depending on whether, or not, they were privately owned. Both Battle Abbey and Chartwell are `publicly' owned by English Heritage and the National Trust respectively. English Heritage is sponsored by the Department for Culture and the National Trust, although a private charity, operates under the mandate of preserving historic places for the benet of the nation. Hever Castle, meanwhile, is owned by a private property company specialising in the buying and selling of large estates. The interviews with site employees took various forms, but all were in part naturalistic in that they took place on the job, so that people were not taken away from their duties for too long a period. All interviews with managers or administrators took place in their respective ofces. As Battle Abbey is an outdoor site discussions with custodians were conducted in portacabins or out in the open. Most interviews were carried out individually although one, between three custodians, took place during a tea break. At Chartwell the pattern was similar. Interviews were conCopyright # 2001 John Wiley & Sons, Ltd.

305 ducted inside the house, either just prior to opening or in `gaps' between visitors coming and going. Useful recordings were also made of communal tea-break conversations in the staff rest area. As far as the interviews with tourists are concerned these were undertaken according to the characteristics of each site. No interviews were carried out inside Chartwell or Hever Castle. This was for two reasons. Firstly, the respective managers did not wish visitors to be approached as this may have detracted from their visit, and secondly, it was felt that this would have affected researcher anonymity and compromised the covert observation of visitor behaviour. At Hever Castle visitor interviews were restricted to the area around the shops, restaurant and the Guthrie exhibition of model houses. However, this did not pose a problem as this area encompasses the main thoroughfare for visitors coming from and going to the castle and gardens. Battle Abbey, however, is an outdoor site with what can be described as a migrant population of tourist visitors. The difculties of sampling such a group of people were discussed by Ireland in his study of Land's End. Here Ireland (1990, p. 35) divided his site into four areas so as ` to replicate as far as possible, the exploratory behaviour of the visitor after leaving the car park'. The main difference between Battle Abbey and Land's End is that the `migrant' visitors at Battle are guided around the site, to a degree, by the recorded tour on hand-held interpreters and by information boards located at various places around the site. This is not to imply that all visitors follow the same route, as the interpreters allow them to pick and choose what information they listen to and the sequence in which it is heard. Furthermore, there are two possible tours of the site, one around the main abbey ruins and one around the perimeter of the battle scene. This latter tour is a mile long and mostly across elds and around woods and as such is not usually taken by families with young children or the elderly. It is also less frequented during the autumn and winter months for obvious reasons. However, whichever tour is taken all visitors at some point pass close by the stone marking the spot where King Harold is reputed to have
Int. J. Tourism Res. 3, 301312 (2001)

306 fallen. Consequently, the two locations chosen for the visitor interviews included the area on and around this particular spot and outside the gatehouse exhibition, which leads to the shop and the exit. A similar strategy was adopted at Chartwell, where the visitor interviews were carried out in the gardens surrounding the house. Here two sites were selected: the area around the car park, adjoining the shop and restaurant, and the Terrace Lawn. This latter area is located just beyond the door by which visitors exit the house and leads into other parts of the garden and down towards Churchill's painting studio. The shop/restaurant area enabled access to those who were not intending to visit the house this time, but who were visiting the shop and/or the restaurant. Theory into practice The sociologist Norman Denzin (1970) stated that, in practice, the sociological enterprise of theory and research is far from an idealised process, immaculately conceived and elegantly executed. In reality concepts do not automatically generate operational denitions and theories do not fall into place once all the data are in. `This is not the way research gets done. What typically occurs is a complex form of interaction between the demands subjects place upon observers and the actual research carried out' (Denzin, 1970, p. 315). This problem of rationalising theory and practice is somewhat inherent in the title of Hammersley and Atkinson's (1983) book Ethnography: Principles in Practice. This illustrates that doing ethnography is not always as easy as reading about it. `At rst blush, the conduct of ethnography is deceptively simple: anyone can do it, apparently' (1983, p. 27). The seeming simplicity of ethnography is based upon the premise that, because it consists of open-ended observation and description it cannot be predetermined. However, Hammersley and Atkinson (1983) state that this does not eliminate the need for pre-eldwork preparation, nor does it mean that researcher behaviour in the eld need be haphazard. What is needed, they argue, is a research design that is reexive, that allows the researcher to review progress and to assess how
Copyright # 2001 John Wiley & Sons, Ltd.

C. Palmer what has been learnt may affect the future direction of the research. Thus, an interview with the UK President of the International Churchill Society helped to inform the questions asked of the stewards at Chartwell. Similarly, some stewards were interviewed more than once if it was felt that clarication was required, or that further follow-up questions were needed. Finally, the researcher's own interviews with, and observations of visitors were utilized in the interviews with site employees. A reexive process thus requires a researcher to be an active listener and to respond to both the answers and the behaviour of the interviewee. It also may lead a researcher to change the questions asked and their sequence depending upon the situation: ethnographers do not decide beforehand the questions they want to ask, though they may enter the interview with a list of issues to be covered. Nor do ethnographers restrict themselves to a single mode of questioning. On different occasions, or at different points in the same interview, the approach may be non-directive or directive depending on the function that the questioning is intended to serve (Hammersley and Atkinson, 1983, p. 113). Thus, open-ended questions/phrases were used, such as `what do you think/feel about ?, `what particular memories will you take home with you today?', and `can you explain what you mean by that comment?'. Likewise, some issues that had arisen out of the analysis of the published and unpublished material sometimes led to specic, directive, questions being asked. Although these can be leading, the important point: is to assess the likely direction of bias that the question will introduce a useful tactic is to make the question `lead' in a direction opposite to that in which one expects the answer to lie and thus avoid the danger of simply misleadingly conrming one's expectations (Hammersley and Atkinson, 1983, pp. 115116). Similarly: if conversation disappears it can someInt. J. Tourism Res. 3, 301312 (2001)

Ethnography: A Research Method in Practice times be rekindled by indicating that you know something already, by proffering a different opinion or by providing a calculatedly wrong assertion (Kemp and Ellen, 1984, p. 234). At times such tactics were used, for example, `Churchill is/is not really a relevant/important gure today'. At other times specic directive questions were used to test out people's reactions to certain key phrases that kept reappearing in the material about the three sites; that the Battle of Hastings was the origin of a nation, that Hever Castle was a treasured part of the nation's heritage and that Churchill was regarded as the greatest living Englishman. In order to examine people's reactions to these statements, whether they agreed with them or not and why, the following questions were sometimes used. `Do you think Churchill is the greatest living Englishman?', `Do you think Hever Castle is a treasured part of the nation's heritage?' and `why is 1066 referred to as the origin of a nation?'. Such questions enabled visitors to agree or disagree and to give their reasons for doing so. Thus, not all visitors agreed that Churchill was `great' or that Hever Castle was anymore treasured than anywhere else. In terms of the visitor interviews, care had to be exercised with regard to bias so that, as far as possible, a representative cross-section of visitors was interviewed. This was addressed by approaching every tenth person or group. As all the conversations were recorded the tapes had to be transcribed. This was not only a lengthy process but also a very complex one. The conversations with visitors displayed the lack of logical `ow' that characterizes natural speech, as opposed to that which indicates a well prepared and thought out response. Moreover, visitors tended to `talk-over' each other, so that two conversations could be taking place at the same time. This often resulted in sentences that just `ended', followed by another persons input that may not have related to what had gone before. Such `difculties' did, however, tend to add value to the ndings, rather more than they detracted from it. This was primarily because, when confronted with questions they have not expected or prepared for, people tend to
Copyright # 2001 John Wiley & Sons, Ltd.

307 respond more instinctively than they may do if given advanced warning of what is to come. In such situations the responses are arguably more `honest' as the visitor does not have time to think of what might be considered a `suitable' reply, or of what the researcher may be `looking for' (see Billig (1992) on this point). With both the visitor and the employee interviews, care was also required with regard to `leading' questions. Thus, when introducing herself to tourists the researcher would not specify the exact nature of her research as this would have enabled them, even if unconsciously, to perhaps `prepare' for the questions to follow. However, during the course of the conversation the purpose of the research became obvious. Techniques used to deal with this included that previously indicated by Cohen (1984), namely feigning naivety and ignorance about the meaning behind a particular statement. This technique was also used when talking to the site employees and to such as, the Regional Marketing Manager for English Heritage, the Tourism Marketing Manager for Hastings Borough Council and the UK President of the International Churchill Societies. Issues of presentation and practicality. The interviews and observations in the research were reproduced directly from the transcripts. Although the conventions associated with conversation and discourse analysis were reviewed (see Heritage, 1997; Potter, 1997) it was felt that these would not have added signicantly to the analysis. Here, as with Billig's (1992) study, the analytic focus was upon what was said and its general context, rather than upon the minutiae of the interaction itself. Overall, grammatical ambiguities were amended, for example punctuation was added to aid clarity. Minor hesitations were omitted, except when central to the analysis. For example, pauses indicating a persons inability to answer a question or expand upon a point they had made. As regards the visitor observations these were conducted on a covert, participatory basis whereby the researcher observed visitor behaviour and noted down comments that were made by them as they went around each
Int. J. Tourism Res. 3, 301312 (2001)

308 site. Although such a process can be seen as an example of non-participatory observation it was felt that this did not accurately depict what actually occurred. Despite the fact that the researcher was not taking part in formalized guided tours, she was participating in the experience of touring the sites and, indeed, had to be seen to do so in order to avoid detection. Consequently, she was also aware of her own reactions to the sites, not merely those of the visitors. Thus, the research was neither wholly participatory, nor wholly non-participatory. (See the recently published text by Denzin and Lincoln (1999a) and, in particular, the chapter by Adler and Adler). Although covert research is an accepted part of ethnography there is a need to address any ethical concerns that may arise during the course of the research process. Indeed, some authors have stated that this method is the least satisfactory on ethical grounds because it can court accusations of `spying' (Clark et al., 1998). There are alternative views, however, `I take the stance that justies unannounced, disguised research methods. If we are not permitted to study things that people wish hidden then sociology will remain a science of public conduct based on evidence and data given us by volunteers' (Denzin 1970, p. x111). A similar point has been made more recently by Punch (1999, p. 172), who further adds, `academics in weighing up the balancing edge between overtcovert, and between openessless than open should take into account the consequences for the subjects, the profession, and, not least, for themselves'. Indeed, almost every research method has its own attendant problems and difculties that have to be considered and worked through (see Punch, 1999). For example, the `open' ethnographer runs the risk of affecting or changing the behaviour of those s/he observes. Similarly, interviews and questionnaires create as well as measure attitudes. Likewise, Mars and Nicod's (1984, p. 11) anthropological study of the occupational culture of waiting staff entailed Nicod becoming a covert participant observer, accepted as part of the group and thus privy to its cultural codes. This, however, led to him becoming involved in acts of petty theft so as not to `blow' his cover. Similarly, Dan Rose's covert
Copyright # 2001 John Wiley & Sons, Ltd.

C. Palmer study of black-American street life entailed him living with the community for two years and taking a job as a car mechanic. During this time he was challenged to a ght and had to `pay the consequences' for trying to outhustle someone (Rose, 1990, p. 11). It is clear, therefore, that the characteristics, requirements and implications of a particular method cannot be ignored, whether these are ethical in nature or merely practical. (For a recent detailed discussion of such issues see Denzin and Lincoln (1999b)). As regards this study the researcher dened the ethical boundaries as follows. (1) At all times the researcher had to be sensitive to what was going on around her and to the reasons why people were visiting the sites in the rst place. (2) Maintaining a discrete distance from those being observed at any one time was thus felt to be important. Although this had to be balanced with the need to `hear' what was being said. (3) Ensuring that the notes made of visitor comments and behaviour were utilized solely in accordance with the academic requirements of the research; the ultimate aim being the advancement of knowledge, not an expose of individuals or of information. (4) It was felt that written notes were acceptable ethically, but that the covert taperecording of comments was not, even though this meant that not everything could be written down. Two methods were considered for recording the visitor comments and observations, manual note-taking and the use of a tape-recorder. The covert tape-recording of conversations has been used previously by researchers at tourist sites. For example, in McManus' (1987) study of communication with and between visitors at the British museum. Here a microphone was placed near the museum exhibits so that only conversations relating to the exhibits would be recorded. In the initial stages of this research the use of a tape-recorder was tested for its applicability and some conversations were recorded. However, on balance this method was not taken further for two reasons. Firstly, the complexity of each site was such that it
Int. J. Tourism Res. 3, 301312 (2001)

Ethnography: A Research Method in Practice would have been impossible to locate microphones at an appropriate number of places to generate enough worthwhile material. Secondly, a hand-held tape-recorder was not sensitive enough to pick up conversations unless held within sight of the visitors, which again made the collection of material almost impossible. Consequently, the manual notetaking of comments and observations was found to be the best method of data collection in this instance. Although this study did not involve the researcher living, or working, with a particular community or occupational group, as was the case for both Nicod and Rose, it still presented certain practical difculties that needed to be addressed. Such difculties did, at times, cause the researcher to suffer from what Rose (1990) has described as the maddening frustration of ethnographic eldwork. Indeed, the researcher was constantly assessing and reassessing her role in the light of the practical difculties thrown up in the eld. Although the issues involved in carrying out the interviews have been discussed, there were particular difculties attached to the covert observation of visitor behaviour and the noting of visitor comments. These difculties are detailed as follows. (1) Objectivity/bias. As already noted, the researcher was participating in the experience of touring the sites and was thus aware of her own reactions, emotions and thoughts. However, Rose (1990) stresses that the author's voice and emotional reactions should not be excluded, as they too require consideration and analysis. Likewise, Hammersley and Atkinson (1983, pp. 166167) have also stated that the researcher's own personal reactions can be used to alert him/her to possible issues: there is a constant interplay between the personal and the emotional on the one hand, and the intellectual on the other. Private response is thus transformed, by reexive analysis, into potential public knowledge the carefully made eldwork journal will enable the conscientious ethnographer painstakingly to retrace and explicate the development of the research
Copyright # 2001 John Wiley & Sons, Ltd.

309 design, the emergence of analytic themes, and the systematic collection of data. Nevertheless, objectivity had to be constantly monitored in terms of what was being recorded so that it did not merely `replicate' the researcher's own experiences. Being aware of the need to remain detached and objective resulted in a continual questioning of the process by the researcher. As Mars and Nicod (1984, pp. 1415) argue, those who adopt a covert stance are often more rigorous as regards the collection and use of material. This is because they have a greater obligation to meet considerations such as objectivity, trust and privacy. (2) Selectivity. It was impossible to follow every visitor or group of visitors, or to record every conversation. Likewise, it was not possible to ask visitors their age, therefore an `educated' guess had to made. Nevertheless, the issue of selectivity had to be addressed. This entailed a constant review of who was being observed with regard to age and gender. This was particularly important at Chartwell with a primary appeal to a specic age range, those aged 45 and above. It was thus necessary to observe those who did not t this typical visitor prole to see what effect, if any, Chartwell had on a younger generation. However, as it was not possible to ask the observed visitors their specic ages, an `educated' guess had to be made with regard to such as children, adolescents and people who `appeared' to the researcher to be under the age of 45. Although the gender of the speaker could be noted, the nationality of the speaker again had to be `presumed', except for those who appeared `obviously' British. (3) Maintaining anonymity. This entailed the adoption of various strategies, for example the use of a small notebook that could be `hidden' beneath the main guide book. When comments were being noted down the researcher would often ensure she was looking `elsewhere' rather than directly at those who were being observed. This `elsewhere' could be at particular artefacts, out of a nearby window, or by appearing to be writing down the details from an
Int. J. Tourism Res. 3, 301312 (2001)

310 information board. It was also necessary not to appear to be `following' particular people/groups. Indeed, two visitors at Chartwell, on being approached later in the grounds, did state that they had noticed the researcher in the house and had wondered what she was doing. (4) Space to write. This was a recurrent problem at both Hever Castle and Chartwell, but for slightly different reasons. As Hever is not only a very popular attraction, but also one that is open on a seasonal basis, the observations took place during the busy summer period. The rooms were thus very congested and the researcher often had to write in small spaces crammed against walls or windows. At Chartwell the rooms are not overly large, but are more the proportion of an `ordinary' house. When coupled with the use of ropes to control visitor ows this often left little space in which to write. (5) Writing up notes. As stated previously it was impossible to follow everyone and to write down everything that was said. As the researcher did not possess the ability to write shorthand she devised her own method of `fast' note taking combined with frequent short periods `of rest' to ensure legibility later on. It was, therefore, important to write-up such notes quite soon afterwards and this was done either on the same day, or by the following day at the latest. (6) The need to write/listen/think/observe all at once. This took some practice to achieve and in reality it was always a compromise between what was possible given the practical constraints listed above. Despite this, natural speech is not usually a sustained monologue, but rather a few words here, then perhaps a pause while individuals read from guide books, information boards or just study a particular room/ornament, before more words are spoken. Likewise, as stated with the visitor interviews, people do not always converse logically and their expressions often resemble `thoughts out loud' rather than actual conversations. Consequently, comments may appear disjointed, conversations can end abruptly and even move to
Copyright # 2001 John Wiley & Sons, Ltd.

C. Palmer different and often seemingly unrelated topics. So although covertly observed conversations could not be recorded in total, verbatim comments and responses were certainly possible. Although all these factors highlight the difculties inherent in research that is both covert and semi-participatory, in some sense they are also advantageous, because they highlight the naturalistic setting in which the research was carried out. This was not a controlled laboratory experiment, but rather a sustained investigation into the activities, motivations and feelings of individuals within specic social settings. As Geertz (1973) illustrates: ethnography is thick description Doing ethnography is like trying to read (in the sense of construct a reading of) a manuscript foreign, faded, full of ellipses, incoherencies, suspicious emendations, and tendentious commentaries, but written not in conventionalized graphs of sound but in transient examples of shaped behaviour.

Conclusion The above discussion has outlined the methodological principles underpinning the study outlined in the introduction, the basis of what Geertz (1973) refers to above as `thick description'. It has addressed some of the problems inherent in qualitative research of this kind and considered the practical limitations of the methods used. More generally, it highlights the relevance and applicability of ethnography, not only in terms of tourism research, but also in relation to the analysis and interpretation of human behaviour in specic social settings. Such analysis is necessary if social scientists are to expand their knowledge of the individual and collective world and to formulate theoretical approaches and hypotheses as a result. In terms of tourism there are a wide, and often complex, array of social settings and interactions to be studied. For example, between the tourist and the local population, between an individual tourist and groups of other tourists and between the tourist and the
Int. J. Tourism Res. 3, 301312 (2001)

Ethnography: A Research Method in Practice environment in which he/she nds themself. Indeed, a series of television programmes in early 1998 covertly observed different nationalities of tourists with the aim of uncovering the `truth' behind the national stereotype (The Tourist Trap, 1998). These programmes claimed to be anthropologically based and despite the many methodological questions they generated, the series did uncover some interesting insights into group behaviour. All of the above research potentialities provide opportunities to analyse the underlying meanings behind how and why people behave as they do. A better understanding of such issues can only contribute towards a more thorough appreciation of the differences between ourselves as individuals and ourselves as groups of people or nations. The ethnographic principles and practices outlined here thus offer a rich and diverse method of investigation. REFERENCES
Adler, PA, Adler, P. 1999. Observational techniques. In Collecting and Interpreting Qualitative Materials, Denzin, NK, Lincoln, YS (eds). Sage: London; 79109. Anderson, B. 1991. Imagined Communities: Reections on the Origin and Spread of Nationalism, 2nd edn. Verso: London. Atkinson, P. 1990. The Ethnographic Imagination. Routledge: London. Baszanger, I, Dodier, N. 1997. Ethnography: relating the part to the whole. In Qualitative Research. Theory, Method and Practice, Silverman, D (ed.). Sage: London; 823. Billig, M. 1992. Talking of the Royal Family. Routledge: London. Burns, PM. 1999. An Introduction to Tourism and Anthropology. Routledge: London. Clark, M, Riley, M, Wilkie, E, Wood, R. 1998. Researching and Writing Dissertations in Hospitality and Tourism. Thomson: London. Clifford, J. 1988. The Predicament of Culture. Harvard University Press: London. Cohen, AP. 1984. Informants. In Ethnographic Research: A Guide to General Conduct, Ellen, RF (ed.). Research Methods in Social Anthropology, vol. 1, Academic Press: London; 223229. Connor, W. 1993. Beyond reason: the nature of the ethnonational bond. Ethnic and Racial Studies 16(3): 373389. Crick, M. 1989. Representatives of sun, sex, savings and servility: International Tourism in the social
Copyright # 2001 John Wiley & Sons, Ltd.

311
sciences. Annual Review of Anthropology 18: 307 344. Denzin, NK. 1970. The Research Art in Sociology: a Theoretical Introduction to Sociological Methods. Butterworth: London. Denzin, NK. 1999. Introduction. In The Landscape of Qualitative Research. Theories and Issues, Denzin, NK, Lincoln, YS (eds). Sage: London; 134. Denzin, NK, Lincoln, YS (eds). 1999a. Collecting and Interpreting Qualitative Materials. Sage: London. Denzin, NK, Lincoln, YS (eds). 1999b. The Landscape of Qualitative Research. Theories and Issues. Sage: London. Elton, GR. 1974. England under the Tudors, 2nd edn. Methuen: London. Gellner, E. 1983. Nations and Nationalism. Basil Blackwell: Oxford. Geertz, C. 1973. The Interpretation of Cultures. Basic Books: New York. Graburn, NHH. 1983. The anthropology of tourism. Annals of Tourism Research 10(1): 935. Hammersley, M, Atkinson, P. 1983. Ethnography. Principles in Practice. Routledge: London. Hastings Borough Council. 1997. Short Breaks and Holidays in Battle, Bexhill, Hastings, Pevensey and Rye. Promotional booklet, Hastings. Heritage, J. 1997. Conversation analysis and institutional talk: analysing data. In Qualitative Research. Theory, Method and Practice, Silverman, D (ed.). Sage: London; 161182. Ireland, MJ. 1990. Come to Cornwall, come to Land's End. A study of visitor experience at a touristic sight. Problems of Tourism 8(34): 3353. Kedourie, E. 1960. Nationalism. Hutchinson: London. Kemp, JH, Ellen, RF. 1984. Informal interviewing. In Ethnographic Research: A Guide to General Conduct, Ellen, RF (ed). Research Methods in Social Anthropology, vol. 1, Academic Press: London; 229236. Levi-Strauss, C. 1972. The Savage Mind. Weidenfeld and Nicolson: London. Levi-Strauss, C. 1986. The Raw and the Cooked. Penguin: London. MacCannell, D. 1989. The Tourist: a New Theory of the Leisure Class, 2nd edn. Schocken: New York. MacCannell, D. 1992. Empty Meeting Grounds. The Tourist Papers. Routledge: London. Malinowski, B. 1922. Argonauts of the Western Pacic. Routledge and Kegan Paul: London. Malinowski, B. 1944. A Scientic Theory of Culture and Other Essays. Oxford University Press: Oxford. Mars, G, Nicod, M. 1984. The World of Waiters. Allen & Unwin: London. McCrone, D. 1992. Understanding Scotland: the Sociology of a Stateless Nation. Routledge: London.
Int. J. Tourism Res. 3, 301312 (2001)

312
McManus, PM. 1987. It's the company you keep. The social determination of learning-related behaviour in a science museum. The International Journal of Museum Management and Curatorship 6: 263270. Nash, D. 1981. Tourism as an anthropological subject. Current Anthropology 25(5): 461481. Nash, D, Smith, V. 1991. Anthropology and Tourism. Annals of Tourism Research 18(1): 1226. Nash, D. 1996. The Anthropology of Tourism. Pergamon: Oxford. Nunez, T. 1989. Touristic studies in anthropological perspective. In Hosts and Guests. The Anthropology of Tourism, 2nd edn, Smith, V (ed.). University of Pennsylvania Press: Philadelphia; 265279. Phillips, M. 1989. 1066 Origin of a Nation, 3rd edn. Michael Phillips: Bexhill on Sea. Potter, J. 1997. Discourse analysis as a way of analysing naturally occurring talk. In Qualitative Research. Theory, Method and Practice, Silverman, D (ed.). Sage: London; 144160. Punch, M. 1999. Politics and ethics in qualitative research. In The Landscape of Qualitative Research. Theories and Issues, Denzin, NK, Lincoln, YS (ed). Sage: London; 156184. Rose, D. 1990. Living the Ethnographic Life. Sage: London.

C. Palmer
Schudson, MS. 1979. Review essay: on tourism and modern culture. American Journal of Sociology 84(5): 12491258. Selwyn, T. 1994. The anthropology of tourism: reections on the state of the art. In Tourism the State of the Art, Seaton, AV (ed.). Wiley: Chichester; 729736. Smith, AD. 1983. Theories of Nationalism, 2nd edn. Duckworth: London. Smith, AD. 1986. Ethnic Origin of Nations. Blackwell: Oxford. Smith, AD. 1991. National Identity. Penguin: London. Smith, V (ed.). 1989. Hosts and Guests. The Anthropology of Tourism, 2nd edn. University of Pennsylvania: Philadelphia. Starkey, D. 1998. Henry VIII. Channel 4 Television, 29 March. Storry, M, Childs, P. 1997. Introduction. In British Cultural Identities, Storry, M, Childs, P (eds). Routledge: London; 140. The Tourist Trap. 1998. Channel 4 Television, 5 May. (A total of six programmes transmitted.) Walter, T. 1996. From museum to morgue? Electronic guides in Roman Bath. Tourism Management 17(4): 241245.

Copyright # 2001 John Wiley & Sons, Ltd.

Int. J. Tourism Res. 3, 301312 (2001)