Learning island foodways: tasting ethnographic methods
Michael R. McDonald
Florida Gulf Coast University, mmcdonal@fgcu.edu
ABSTRACT
Roatan, largest of the Bay Islands off the northern coast of Honduras has in recent years, been faced with broad-based social, environmental and cultural change. Various ethnographic studies assert that in-migration of Spanish speakers from the mainland and the growing class of tourists and expatriates have amplified existing differences among island residents, and influenced change in cultural characteristics including foodways. Using data from systematic elicitation in three ethnic and linguistically distinct communities, collected by students from Florida Gulf Coast University's Ethnographic Research Expedition Roatan 05, this article abstracts and interprets variation in cultural domains surrounding common foods. A greater goal of the project is to immerse students in the entire ethnographic experience. The project was highly successful as the location, topic and iterative methodology brought students into the ethnographic research enterprise in an effective and rewarding manner. The implications of this learning-centered approach and subsequent findings are discussed and indicate further direction for research on this topic.
KEYWORDS: ethnographic field-school, foodways, Bay Islands, Honduras Introduction
The beliefs and practices surrounding the production, preparation, presentation and consumption of food have come to be known as foodways in the anthropological and allied literatures. Within the concept is a tacit recognition that food provides more than mere calories and nutrients for human consumers. Foodways form a powerful component of culture that affects a range of behaviors and interactions among people. As a force of culture, food and foodways, like other significant features of the natural or social environment, need to be organized into meaningful categories by the members of society who employ them in daily life. The ethnographic literature abounds with examples of how hot-cold dichotomies are employed by cultural groups from Southeast Asia to Central America to organize thinking and regulate consumption of foods (Messer 1987; Manderson 1986). Other ethnographic studies have illustrated how foodways provide structure, order and a sense of identity in social relationships ranging from the local to national (Weismantel 1988;
ANTHROPOLOGICAL NOTEBOOKS 12 (1): 35-48. ISSN 1408-032X
© Slovene Anthropological Society 2006
Wilk1999). As an individual learns what to eat, he or she also learns what to call food items, and the rules, norms and customs dictating when, how and with whom to eat them. In so doing he or she becomes party to information that members of the cultural group recognize as being related or associated within a 'cultural domain' (Borgatti 1999). An empirical task in this research project includes determining the extent and shape of cultural domains surrounding foodways in spatially adjacent yet culturally distinct communities.
Roatan Island is home to no fewer than three ethnically and linguistically distinct communities. While people from these communities interact to a high degree and their diets appear largely similar, the people maintain strong senses of separate identity -supported by spatial and social segregation. If the cultural boundaries separating the various ethnic groups on Roatan are as significant emically as suggested in ethnographic literature (Evans 1971, 1986; Davidson 1974; Stonich 2000), then ethnographic investigation could make explicit the cultural domains relevant to foodways. While cultural domains are generally conservative, they are not static. The global processes of market capitalism, delocalization (Pelto and Pelto 1983) and tourism have contributed to a changed environment on Roatan altering both diet and food practices for all groups. A secondary interest in this research is to determine how and if the study of foodways domains might be used more fully in the study of culture change.
As the research for this article has been conducted under the auspices of an ethnographic field school, an overarching interest in this work is to evaluate its contributions in light of the scholarship of teaching and learning anthropology in the field context. This asserts that the topic of foodways is especially effective for immersing student ethnographers in another culture and guiding their development through methodologies and important principles of the discipline toward meaningful understanding of ethnographic accounts.
Setting
Roatan is the largest of the Bay Islands, an archipelago of islands and cays in the Western Caribbean off the North coast of Honduras, Central America. Extensive fringing reefs provide the island with beautiful beaches and vistas and a protective habitat supporting variety of marine life which once formed the backbone of the local economy. Owing to the relative isolation of life on all islands and a fascinating colonial and post-colonial history, Bay Islanders tend to regard themselves as a people apart, highlighting their linguistic and cultural differences from the mainland of Honduras (Evans 1986). Recent waves of immigration from the mainland and a booming tourist industry have amplified the social complexity of the islands which is today, stratified, multiethnic and multilingual. While etic attempts to categorize Bay Islanders would fail to capture the rich internal diversity, there are relevant emic categories employed widely on Roatan. The 'English', include both 'White' and 'Black' Creoles, descendents from 19th century immigrants to Roatan from the Cayman Islands and elsewhere in the English-speaking Caribbean and beyond. Both Anglophone groups claim 'English' ethnicity but they are by and large segregated socially, economically and spatially from each other and other groups on the islands (Evans 1971, 1986; Stonich 2000).
Another group with longstanding claim to Roatan is the Garifuna. Known formerly in the anthropological literature as the 'Black Carib' (Gonzalez 1988), the Garifuna offer a textbook example of a syncretic and reconstituted culture. Their own origin legends and historic accounts concur citing their ethnogenesis with the intermarriage of Carib Indians and ship-wrecked Africans on the islands of the Eastern Caribbean in the mid 18th century. Forcibly displaced to Roatan in 1797 from St. Vincent, the Garifuna inhabited and thrived on a nearly deserted Roatan and have since dispersed their members to settlements along the coast of Honduras and adjacent countries maintaining their culture, customs and their own Garifuna or Garinagu language in a Diaspora of transnational communities (Gonzalez 1988). Throughout the 19th and much ofthe 20th century the Garifuna subsisted through fishing and agriculture (Stansbury and Sierra 2004: 459). In more recent times they have turned to wage labor including wage labor, commercial fishing and employment in the merchant marine fleets of various nations.
Most ethnographic research accounts of the island have noted the longstanding divisions on the island of Roatan between White and Black Islanders. Stonich writes, 'ethnic and cultural distinctions between these groups have played important roles in social and economic relations, the two groups also hold and exercise significantly different degrees of power' (2000: 49). On Roatan, segregation does not fallout neatly along racial lines, the Creole Black Islanders and Garifuna are also acutely aware of their separate identities, with Garifuna maintaining that they are, 'more authentic than the Creoles' who 'are generally considered by the Garifuna to be the offspring of slaves' (Kirtsoglou and Theodossopoulos 2004: 141). For their part, the English hold themselves in higher regard as well, employing language as an important cultural marker. One of the student ethnographers witnessed a child in Barrio Ingles, Punta Gorda correcting another for speaking in Spanish, stating, ' We speak English here\ Minimally these groups maintain a rhetorical social distance, resist homogenizing generalizations and in more extreme cases regard each other with disdain or simply avoid interaction. All three groups: Black and White English speakers and the Garifuna do however regard themselves as 'Islanders' and have a keen sense of the 'Spanish other' on Roatan. Referred to as 'ladinos' in the ethnographic literature; and 'Spaniards', 'Hondurans' or 'mainlanders' by many Islanders, the Spanish speaking residents of Roatan have increased from a mere 7% of the population in the 1970s (Davidson 1974: 130) to represent at least 50% of the island population in some communities currently (Stonich 2000: 90 and 97). The assertion of identity works in both directions, Stonich notes, 'mainlanders in turn refer to 'white' islanders as 'caracoles' (conch or snails) and to 'colored' islanders as 'negritos'(ibid: 52). The dramatic increase in migration from the mainland has been matched by government efforts to Hispanicize the island. Schools, government, banks and businesses operate primarily in the national language. Various revivalist efforts both formal and informal have emerged to sustain English language use in churches, schools and social-civic organizations (Stonich 2000: 53,176). This linguistic, cultural and geographic segregation is an important factor setting the context for this work.
To the already complex social scene, more people have entered and change has followed. In recent years, Roatan has become a popular tourist destination drawing divers,
cruise ships and a class of expatriates who have taken up seasonal or year-round residence. The remarkable growth in the tourist sector over the past decade now allows nearly 100,000 people to visit the Bay Islands every year (Stonich 2000: 90). Stonich has noted further a corresponding rise in the number hotels and restaurants on Roatan over the same time period. As population shifts to include more expatriates for seasonal or year- round residency, and more migrants from the mainland attracted to the booming service economy, the type of foods available on Roatan has also witnessed change. Foods that are now readily available on Roatan include fresh milk, imported fruits and vegetables from North and South America. Food imports from the mainland and points beyond have kept pace with the rising immigrant population. According to informants consulted in this project, garden produce including lettuce, carrots, onions and tomatoes, now widely available throughout Roatan, were rare or unheard of two decades ago. Likewise some formerly common insular foods including conch, lobster and iguana are increasingly scarce, owing to various pressures on the environment and or changes in laws governing harvest and consumption.
To better understand how people on Roatan conceptualize and employ foods amid this flux was the project assigned to students from Florida Gulf Coast University's Ethnographic Research Expedition, Roatan 05. Pairs of students were assigned one of three communities on the eastern end of the island:
Camp Bay a small beachfront community of approximately 20 households is rather remote, former fishing village was our first research site. Some fishing and subsistence gardening still go on in Camp Bay; however most people are now supported by wage occupations outside of the community and by remittances from family members working abroad. The village is as yet removed from the main road and has thus far been isolated from much of the tourism and migration affecting other parts of the island. English remains the primary language and with the exception of a few expatriates, the residents are all Black Creole Islanders.
The town of Punta Gorda (Barrio Ingles) became our second research site. It is a larger community of over 100 households. Like Camp Bay it is also a fishing community and stretches along the North coast of Roatan. Punta Gorda is the primary settlement for the Garifuna population on Roatan and the town has a store, several churches a school, a community center, several beach-front bars and restaurants. Included in the town's seven neighborhoods or barrios is Barrio Ingles where a community of Black Creole islanders also reside. Spanish, English and Garifuna are all spoken within Punta Gorda. Our hope was that this site would provide for a nice blend of cultural and linguistic influences reflected in foodways. Traditional enterprises such as small-scale fishing and conch diving are pursued by a few individuals. Some people are also engaged in small-scale food production and wage labor within and outside of Punta Gorda. Many residents in Punta Gorda have worked abroad or have family members who are currently abroad or working on ships.
Juticalpa-is one of the older ladino communities on Roatan. Unlike the other two communities in the study, Juticalpa is located within an interior valley on the east end of Roatan. A town of the same name in the Department of Olancho on the mainland of Honduras is the home from which migrants to Roatan have been coming for the past 30 years or so. The island locale of Juticalpa is a small settlement of about 25 homes with a
school, two churches and several pulperias (small local stores). Spanish is the principal language spoken and the people make their living primarily from wage work in the area. Many households produce from some to much of their own foods in subsistence gardens located near the town. Juticalpans also collect wild foods from the mangrove and wooded areas surrounding the town. Even long-time residents still identify strongly with their hometowns on the mainland and consider themselves to be separate from other island residents even identifying themselves as ' Indio' to sharpen the contrast with 'Islenos' (islanders), in conversation with the ethnographers.
Methods
These present data are drawn from the second year of this multi-year ethnographic research expedition-field school on Roatan. The curriculum provided students with a one-week refresher course in Spanish language and a developmental sequence of eight data collection activities that tapered students into and out of the ethnographic enterprise including: unobtrusive observation, participant observation, map-making, time allocation studies, 24 hour recall interviews, market survey, free listing and pile sorting. In addition to these formal ethnographic tasks, students were also involved in various less structured interactions with island foodways. They took part in fishing trips, collected crabs in the mangrove and wild foods in the forest. One student learned how to render coconut oil and another prepared cashews; both learning foodways as participant observers under the supervision of key informants. Students were further required to write up their field notes, and discuss the day-to-day progress with the instructors and each other. Ethnography was at once the objective of the course and a central pedagogical strategy. On the surface, all tasks were meant to collect relevant data but they unfolded in a manner that kept pace with the developmental, pragmatic and meaning-making needs of students learning a new culture. Sequencing the tasks prepared students for better results and more synthetic thinking. For instance, as the students built rapport with people in their communities, they were required to explain the project and obtain written consent. Toward the end of our stay on Roatan they were required to reciprocate the accommodation shown them by contributing at least 6 hours of community service to their hosts. This last activity in particular became more meaningful personally and pedagogically than would have been the case had it occurred at the outset of the project.
Free lists and key lists
Shortly after assignment to the project sites, students began by collecting free lists from people to generate a comprehensive list of common foods known in the three villages. This minimally intrusive activity helped students break into the ethnographic enterprise and laid the foundation for further activities in the project. Sixty two respondents provided a composite list of 239 food items for the free list task. As a preliminary analysis activity, the students determined the rank order and salience of these items to form a key list of 30 items detailed in Table 1 below. The use of salience ranking (Smith and Borgatti 1997: 208) assured that these foods were known by informants from all three communities thereby making the comparisons meaningful.
Table 1
KEY LIST - ROATÁN 05
RICE (1)
COCONUT MILK (7) YUCA (13)
BANANAS (2) MANGO (8)	EGGS (14)
PLANTAINS (3) CRAB (9)	CONCH (15)
CHICKEN (4) SHRIMP ( 10)	LIME ( 16)
FISH (5)	APPLE (11)	CARROT (17)
BEANS (6) SPAGHETTI (12)	MACHUCA (18)
CANDY (19) LETTUCE (25) SUGAR (20) CAKE (26) SUGAR CANE (21) ONION (27) TURTLE (22) COFFEE (28) OATMEAL (23) CASHEW (29) CASABE (24) MASECA (30)
This key list became essential in subsequent tasks including the market survey, the 24 hour diet recall interview and ultimately the pile sort task, which will be considered in the balance of this . An important note here is that informants who supplied the initial free list responses were also included in the subsequent tasks. This is one of the unique and promising characteristics of the iterative type of ethnographic research possible in a teaching format. Preliminary findings, assumptions and methods can be refined through repeated interaction between student ethnographers and the informants over the course of data collection.
As one of the interests of this project is to study culture change, two important items were included in the key list despite their relatively low salience, to track change in island foodways. Lettuce (item no.25), is emblematic of new foods on Roatan. Various informants reported that in times past, lettuce along with other salad vegetables including carrots and tomatoes could not be grown on the island. Heads of iceberg-type lettuce imported from the mainland are now widely available on the island, distributed by itinerant grocers and sold at pulperias as well as the major grocery stores. At over five times the cost, locally grown hydroponic lettuce has recently become available on the island as well. Sold primarily to restaurants, hotels and the two main grocery stores, this high-end product appeals primarily to the expatriate and tourist sector but foretells the wider food list that economic change will bring to Roatan.
By contrast, turtle (item no. 22), was included in the key list as a formerly common food. Decline in turtle population and a concomitant conservation effort throughout the Bay Islands and the region, has cast turtle into an ambiguous status as a food, nearing taboo for many. If caught unintentionally in fishing nets, turtle is still consumed although people are well aware that it is forbidden and are generally reluctant to discuss it. Turtle thus represents food from the past and how people conceptualize it also indicates their changing cultural domains. Lettuce and turtle will be referred to as tracking foods in the remainder of this .
Pile sorts and verbalization
After the key list was formalized, a chained referral or snowball sampling method (Bernard 1995: 97) was used to identify thirty one individuals from across the three communities to participate in the Pile Sorting Task. Pile sorting is a categorization task in long use by cognitive anthropologists (Perchonock and Werner 1968: 300) to uncover how individu-
als think about and understand those learned rules that govern food preparation, consumption, and edibility. We presume a level of consensus and cultural integration exists within the communities.
All informants were presented with a single pile sort task. Cards prepared for the task that included hand-drawn images of the key list foods with Spanish and English names on the front and a corresponding number on the back. The informants were asked to sort cards into piles that, 'made sense to them' and then were asked to describe each pile and how the foods in each pile were similar. Two conditions were placed on the task. The first was that there must be more than one pile: thereby disallowing extreme 'lumpers'. The obverse problem of extreme 'splitters' or placing every item in a pile by itself was disallowed as the second condition. The online tools, CardZort® and CardCluster ® (Toro 2004)1, were employed to abstract and assess the cultural domains. To facilitate the hypothesis testing of cultural dissimilarity set out in this project, an aggregate of the answers of all respondents from each community were represented in tree diagrams, reproduced for comparison in Figures 1, 2, and 3.
Data and analysis
The findings are discussed in two sections, the first is a summary and interpretation of the results from the pile sorting tasks. These data allow readily for a quantitative and etic interpretation that differs from the qualitative and emic explanations offered by informants in their verbal explanations accompanying the piles in the second part.
The respondents (n=12) from Camp Bay provided the most easily recognizable pattern of food groupings among the three communities and construct the food domain in two dimensions: provenience and taste, and the former tended to be more prominent. A seafood group included: shrimp (10), conch (15), fish (5), and crab (9). The most dissimilar food in this group is turtle (22); where six informants indicated that they did not eat turtle or placed it within a group of foods they did not eat. A second definitive cluster of plant foods includes garden produce: yuca (13), onion (27), carrots (17) lettuce (25), and plantain (3) and another includes these fruits: mango (8), apple (11), cashew (29), and banana
The CardZort program essentially creates a similarity matrix of the 30 food items for each respondent wherein each term is arrayed in relation to every other term. When two terms appear in the same pile, there is a '1' placed in that cell of the matrix, indicating similarity. When two terms do not appear in the same pile, there is a '0' placed in that cell of the matrix, indicating dissimilarity. These matrices are then averaged over all respondents, resulting in the percentage of times each term appears with each other term. The more often the foods appear together, the more similar they are thought to be by informants. The similarities and differences among items in the domain can then be examined by subjecting the aggregate similarity matrix to hierarchical clustering. The Card Cluster ® program uses the complete linkage method for hierarchical clustering (also called furthest neighbor) an agglomerative clustering strategy that computes and displays similarity of 2 clusters as the most dissimilar pair of objects, where one occurs in each cluster (Romensburg 1984). The program displays the responses organized in the matrices in hierarchical tree diagrams with more similar foods appearing close together and dissimilar ones spaced farther apart.
Cam p Bsy -Complex	Simil.rily
(2). Note that lettuce, while not grown in the saline soils around Camp Bay, is organized cognitively with other garden foods, suggesting that this new food has been made meaningful by association in a known category and thus accommodated in the cultural domain without ambiguity.
The respondents at Camp Bay grouped the remaining items in two clusters on what appears to be the basis of taste. The first includes a cluster of sweet foods, most of which are coincidently purchased at the store: coconut milk (7), sugar (20), oatmeal (23), lime (16), sugarcane (21), candy (19), cake (26) and coffee (28). Limes and coffee may appear to be out of place, but they are both consumed by locals as beverages with copious amounts of sugar and would therefore be associated with other sweet foods. The second category of prepared foods are neutral in taste and include: cassabe (24), (cassava bread), maseca (30) (a fine cornmeal flour), machuca (18) (a common Garifuna dish of smashed plantains), spaghetti (12), eggs (14), rice (1), beans (6) and chicken (4), these foods share principally a need to be cooked before they can be consumed.
Barrio Ingles Com plete
Percent Similarity
Fig. 2 - Barrio Ingles
Barrio Ingles provided a pattern of responses that was similar to Camp Bay in some ways and similar to the Juticalpa response in others. It was less complex overall than both with only three clusters emerging at the level of 50-60% similarity. The largest and most significant grouping of foods included:
coconut milk (7), yuca (13), machuca (18), fish (5), cassabe (24), banana (2), plantain (3), beans (6), lime (16), mango (8), cashew (29), sugarcane (21), turtle (22), shrimp (10), conch (15) and crab (9).
Another group included:
carrots (17), lettuce (25), apples (11), spaghetti (12), candy (19),
oatmeal (23), rice (1), chicken (4), and onions (27).
Finally, the third group included:
eggs (14), maseca (30), cake (26), sugar (20) and coffee (28).
The most significant sorting mechanism in operation among the nine (9) Barrio Ingles informants appears to be provenience. The items in the first group are or were produced locally on the island, both on land and in the sea. One exception is coconut milk. While it is a traditional food, most coconut milk consumed today is imported in cans from Asia as the local coconut crop has been devastated by 'lethal yellowing' in recent years. The retention of coconut milk in this category appears to be evidence of cultural lag. People remain inclined to think about coconut milk as a local food despite its current provenience. This stands in interesting contrast with the tracking item turtle. While turtle is thought of as local, it is no more tightly associated with seafood than with foods produced in the garden. The other tracking food, lettuce, is tightly associated with carrots and to a lesser extent, apples, but not with other locally grown garden foods. All three items are imported from the mainland and purchased at either the grocery stores or from the itinerant produce vendors. From this it seems lettuce is clearly a known food and thought of in the same manner as other plant foods brought into Barrio Ingles from off the island.
Fig. 3 - Juticalpa
The third grouping of foods includes items commonly consumed together as breakfast foods eggs and maseca. Foods taken together as a snack: coffee, sugar and cake also appear together (cakes are quite popular on Roatan informants from all locations spoke about them). This pattern of grouping by meal menus anticipates the clustering strategy employed in the village of Juticalpa and may indicate some diffusion of both the custom of cake and coffee and the cognitive inclination to organize them together as a cluster.
The responses from Juticalpa provide a different pattern of food categories than those employed in the other two communities. At the level of 50 % similarity, these respondents grouped foods into seven groups. Eggs (14) and maseca (30) are tightly grouped as a typical breakfast meal and were reported by informants as foods consumed daily or nearly so. Rice (1) and shrimp (10) are linked and join lettuce (25), carrots (17) and lime (16) in what appears to be a two course meal. As it is more common on the mainland and consumed frequently by informants, the tracking food, lettuce appears unremarkable as a food for the Juticalpans and it appears to continue in the same domain role. One female informant from Juticalpa commented that they (the mainlanders) had taught the people of Roatan to eat salads. The 'Latin American Trinity' of coffee (26), cake (26) and sugar (20) form another cluster suggesting cultural continuity in foodways for these respondents. A seafood soup grouping with plantains (3), conch (15), fish (5) and coconut milk (7) appears to constitute a recipe adopted in entirety from island foodways without modification- a strong indication of cultural assimilation toward island foodways. The most ambiguous cluster for the Juticalpa sample included: machuca (18), turtle (22), onions (27), candy (19), oatmeal (23) and cassabe (24). In addition to turtle, Cassabe and Machuca could be considered 'island' or Garifuna foods. Their association in this grouping indicates a perception of otherness by the mainlanders and constitutes a group of foods that are not eaten or are not yet fully assimilated.
Verbalization
In the three dendrograms depicted above various foods are placed in different categories, indicating perhaps that the principles employed for creating piles are not necessarily the same for the different groups. An important and complementary part of the pile sorting task was to collect the informants' verbal explanations that guided their pile making choices. Of the thirty one informants, the number of piles sorted ranged from 2 to 14. After eliminating expressions of personal preference and redundancies from the ninety seven total verbal responses, the remaining labels fell into two main types. Eight of these were taxo-nomic in nature- indicating morphological similarity or common provenience. The most common, used three or more times included 'same color', 'comes from the sea', 'sweets', 'grown on Roatan', 'fruits', 'Garifuna foods' and 'no come' (not eaten). Second, twenty-nine verbalizations were categories that referred to the situation or context in which the food was consumed such as breakfast foods: maseca and eggs or party foods: cake and candy. These are called script categories (sensu Ross and Murphy 1999), 'because the usually indicate a time or situation in which the food is eaten' (504).
Twenty five of the thirty one informants employed one or more script categories in describing their piles including all of the Juticalpa informants. Clearly there are alternate
organizational schemas for the classification of foods within a domain and the preference for script categories over the taxonomic may bring into clearer focus the cultural differences between Spanish and English speakers on Roatan.
Discussion
To review, the expectation was that through cognitive elicitation methods, informants from the three communities would reveal differences in the cultural domains relating to foods. For one, it was expected that the cultural domain of foodways for immigrants from the mainland now residing in Juticalpa, would reflect the linguistic and cultural differences thought to separate them from longer-term residents of the island. The responses from Juticalpa revealed noticeable differences primarily in how informants responded to the pile sort task. All chose to organize the items into meals or foods otherwise consumed together in script categories, the singular taxonomic category included a group of tree fruits: apple, mango and a close analog, banana.
The responses from the Camp Bay sample are thought to best contrast with Juticalpa and reflect the most traditional island thoughts and customs with respect to food. As an English speaking community they preserve a culturally legacy dating back two hundred and fifty years. The dendrogram appears to employ a taxonomy-based on provenience with food clusters representing longstanding adaptation to an island environment complemented by categories of purchased and prepared foods. However, most Camp Bay informants also employed script categories in explaining their pile sorts.
Two alternative organizations can operate at the same time; a food can be both part of a typical meal and come from the garden. The pile sort technique alone presumes a single hierarchical logic and cannot readily capture cross classification. This case has illustrated the necessity of eliciting emic explanation to accompany pile sorts to guard against a hasty or biased interpretation of pile sort data.
The Barrio Ingles sample included a high proportion of English speakers reflecting a sampling convenience for the student researchers. While respondents had varying degrees of association with the Garifuna people, the sample is at best a mixed representation of Garifuna and Black Creole Islanders. That noted, for these respondents there remains a clear distinction in the domain between traditionally-produced foods versus those items purchased at the store, a distinction made by the respondents in Camp Bay but not by those in Juticalpa. These taxonomic categories are complemented by one that appears to be a script category defining a breakfast menu similar in form to those employed in Juticalpa.
With respect to the second aspiration of the project, foodways domains do hold promise for measuring culture change as well. Several indicators of resistance, lag or identity retention were described. For instance the tracking food, turtle was placed into an out-group category of island food or non-food by Juticalpans. Indicators of assimilation included the wholesale adoption of a typical island meal-seafood soup- by the residents of Juticalpa and the incorporation of lettuce into a garden taxonomy by Camp Bay residents. Other facets of change were indicated in Barrio Ingles where turtle appears to be more visibly in decline as a food item than in the neighboring Camp Bay community. These
are indications of change in foodways across the island, with Camp Bay being the most conservative in this study and Juticalpa the least so. A way to further quantify the extent and rate of assimilation of Juticalpans on Roatan would be to compare their responses against a sample of mainlanders. Longitudinal change among islander groups could be approached through other techniques as well including ethno-historic and archival studies.
The data for this project were collected by students under the direction of the author in the context of a four week ethnographic field school. What is reasonably sure from the collective efforts of this team is that informants in each community share in a domain set of ideas about foods. Directions clearly indicated for further research include employing the same pile sort technique in adjacent communities of Black Creole Islanders and Ladinos, measuring levels of cultural consensus within the presumed cultural domains and searching for the foods or practices that form the edges of the domains. A more tightly controlled sample of Garifuna speakers within Punta Gorda would also provide better data for inferences and conclusions about the extent of cultural consensus and differences within cognitive domains. Other dimensions of the domain including the norms, customs and lore attached to particular foods would enhance our understanding of island foodways and these could be had from a range of techniques not employed in this study.
The brevity and the methodology of the project makes the data and findings subject to an array of criticism leveled against other short-duration ethnographic studies (Utarini,Winkvist & Pelto 2001). For example, rigorous methods could be employed to more accurately sample the populations and their attendant domains. The purposive sampling employed here prioritizes individuals who are knowledgeable, accessible and amenable to participation in the project. This represents a trade off for the goals of the project which prioritized student participation and experiential learning in ethnographic research methodology. In this respect, the project was highly successful. Each student was immersed in a method of ethnographic learning that challenged their intellects and nurtured their individual development. Due to the setting and various pressures, no one could drop a class as they might back home and importantly no one asked to go home. To succeed in the course meant making meaning of the place and people. Students were required to negotiate entry and exit to and from a field site, struggle with communication, logistics and norms of behavior, and acquire information that contributed to the larger, known ethnographic objective. In later evaluations students wrote that the project: 'was thrilling', 'changed my life', 'most interesting thing I've done in college'. This brief study reinforces the author's commitment to teaching ethnographic methods through an experiential ethnographic pedagogy and to the idea that transformation is an equally laudable outcome for any ethnographic enterprise.
Acknowledgments
The research reported here was conducted under the auspices of Florida Gulf Coast University's (FGCU) Ethnographic Research Expedition - Roatan 05 with grant support from the Meftah International Internship Foundation, and administrative support from the College of Arts and Sciences at FGCU. wish to thank Noemi Creagan and acknowledge her valuable contributions to the collection of data, instruction of courses, and overall contributions that made the project possible and rewarding. Finally I wish to acknowledge the important contribution to the collection of ethnographic data by students enrolled in the 2005 program: Tyler Armstrong, Grant Bailey, Alyssa Butler, Ashley Dumm, Bridget Regan and Michael Wylde. Thank you all for your good humor and tireless effort.
References
Bernard, H. Russell 1995. Research Methods in Anthropology. Walnut Creek, CA: Altamira Press.
Borgatti, Stephen P. 1999. Elicitation Techniques for Cultural Domain Analysis. In: Jean J. Schensul, Margaret D. Le Compte, Bonnie K. Nastasi and Stephen P. Borgatti (eds.), Enhanced Ethnographic Methods: Audiovisual Techniques, Focused group, Interviews and Elicitation Techniques. (Ethnographer s Toolkit Vol. 3). Walnut Creek, CA: Altamira Press, pp. 115-151. Davidson, William V 1974. Historical Geographyofthe Bay Islands, Honduras. Birmingham: Southern University Press. Evans, David. K. 1971. Obeah and Witchcraft on Roatan Island. In: D.E. Walker Jr.(ed.), Systems of North American Witchcraft and Sorcery. Moscow, ID: Anthropological Monographs of the University of Idaho, No1., pp.109-124. Evans, David K. 1986. Roatan Island: A Quarter Century of Demographic and Nutritional
Change. Collegium Anthropologica 10: 81-89. Gonzalez, Nancy L. 1988. Sojourners of the Caribbean: Ethnogenesis andEthnohistory
ofthe Garifuna. Urbana: University of Illinois Press. Kirtsoglou, Elizabeth and Dimitrios Theodossopoulos 2004. 'They are Taking Our Culture Away': Tourism and Culture Commodification in the Garifuna Community of Roatan. Critique ofAnthropology24(2): 135-157. Manderson, Lenore. 1986. Food Classification and Restriction in Peninsular Malaysia: Nature, Culture, Hot and Cold. In: Lenore Manderson (ed.), Shared Wealth and Symbol: Food, Culture and Society in Oceania and Southeast Asia. New York: Cambridge University Press, pp. 127-143. Messer, Ellen. 1987. The Hot and cold in Mesoamerican Indigenous and Hispanicized
Thought. Social Science and Medicine 25(4): 339-46. Pelto, Gretel H.and Pertti J. Pelto. 1983. Diet and Delocalization: Dietary Change Since
1750. Journal ofInterdisciplinary History 14(2): 507-28. Perchonock, Norma and Oswald Werner. 1968. Navajo Systems of Classification: Some Implications of Food. Ethnology8: 299-342.
Romesburg, H. Charles. 1984. Cluster Analysis for Researchers. Belmont, CA: Lifetime Learning Publishers.
Ross, Brian H. and Gregory L. Murphy. 1999. Food for Thought: Cross-classification and Category Organization in a Complex Real-world Domain. Cognitive Psychology 38: 495-553.
Smith, J. Jerome and Stepher P. Borgatti. 1997. Salience Counts—And So Does Accuracy: Correcting and Updating a Measure for Free-List-Item Salience. Journal ofLin-guistic Anthropology 7(2): 208-209. Online publication date: 1-Dec-1997.
Stansbury, James P. and Manuel Sierra. 2004. Risks, Stigma and Honduran Garifuna conceptions of HIV/AIDS. Social Science and Medicine 59: 457-471.
Stonich, Susan C. 2000. The Other Side ofParadise: Tourism, Conservation and Development in the Bay Islands. New York: Cognizant Communication Corporation.
Toro, Jorge A. 2004. CardZort and CardCluster. Depaul University, School of CTI.
Utarini, Adi, Anna Winkvist and Gretel H. Pelto. 2001. Appraising Studies in Health Using Rapid Assessment Procedures (RAP): Eleven Critical Criteria. Human Organization 60 (4): 390-400.
Weismantel, Mary J. 1988. Food, Gender and Poverty in the Ecuadorian Andes. Philadelphia: University of Pennsylvania Press.
Wilk, Richard J. 1999. 'Real Belizean Food': Building Local Identity in the Transitional Caribbean. American Anthropologist 101(2): 244-255.
POVZETEK
Roatan, največji otok v otočju Bay, ki leži ob severni obali Hondurasa, je v zadnjih letih doživel številne družbene, okoljske in kulturne spremembe. Različne etnografske študije so pokazale, da so priseljevanje špansko govoreče populacije s celine in naraščajoča skupnost stalno naseljenih tujcev ter številni turisti, povečali že obstoječe družbene razlike med otočani in vplivali na kulturne spremembe na otoku in s tem tudi na prehranjevalne navade. Na podlagi podatkov zbranih z metodo sistematičnega izvabljanja v treh etnično in jezikovno različnih skupnostih, ki so jih zbrali študenti Florida Gulf Coast University na etnografski ekspediciji Roatan 2005, ta članek analizira razlike v kulturnih kontekstih glavnih otoških živil. Širši cilj projekta je bila vključitev študentov v celotno etnografsko izkušnjo. Projekt je bil zelo uspešen, saj so lokacija, tema in metoda uspešno uvedle študente v etnografsko raziskovalno dejavnost. Članek obravnava učinke takega učnega pristopa in nakaže bodoče smernice za raziskovanje kulturnih kontekstov prehranjevanja na otoku.
KLJUČNE BESEDE: etnografska terenska šola, prehranska kultura, Bay otočje, Honduras