The Influence of Visual Stimuli in Ethnobotanical Data Collection Using the Listing Task Method
TATIANA MOTA MIRANDA
Universidade Federal de Santa Catarina–Campus Trindade
MARIA CHRISTINA DE MELLO AMOROZO
JOSÉ SÍLVIO GOVONE
Universidade Estadual Paulista–Campus Rio Claro
DANIELA MOTA MIRANDA
The listing task, a method used in social and behavioral sciences, is frequently used in ethnobotanical research to construct folk taxonomies and select relevant items for subsequent research. The objective of the present study was to determine whether visual stimuli are associated with responses to the theme “plants” or if context influences the answers. Interviews were conducted with 400 women in Rio Claro, São Paulo, Brazil, in four different locations: three with a visible presence of plants (a plant store, a supermarket, and a public plaza) and one with no plants (a street corner in the center of the city). The women were asked to name plants. Analysis indicates that visual stimuli influenced responses and that this is more marked in the plant store than in the other locations. The plants cited most often—roses, orchids, ferns, violets, and daisies—were, with little variation, the same in all the locales studied.
Keywords: ethnobotany; field methods; listing tasks; visual stimuli
Free listing is widely used in cognitive social and behavioral science as well as in ethnobotany and ethnozoology (Bernard 1988) to determine the contents of a cultural domain (Weller and Romney 1988; Borgatti 1994). The frequency of items mentioned and the order in which they are used measure the items’ cultural relevance or salience (Weller and Romney 1988; Martin 1995; Cotton 1996; Sutrop 2001).
Listing tasks have generally been used in combination with other methods in ethnobotanical research, designed not only for folk classification (e.g., Boster 1984; Berlin 1992) but also for understanding the content of traditional knowledge about the uses and care of plants and animals and how that knowledge changes over time (Nolan 1998; Benz et al. 2000). We know from Brewer’s work (2002; Brewer, Garrett, and Rinaldi 2002) that the list length is substantially influenced by probing. The content of lists, however, may be influenced by the physical context in which they are gathered. We report here on a test of this potential response effect. There are threats to the validity of our findings, but the overall pattern suggests strongly that environmental cues in the direct or peripheral vision of informants do have an effect on the content of their lists. Specifically, in ethnobotanical studies, it is not unusual (perhaps not even unexpected) to conduct interviews in situations in which plants are within sight of the informant. We investigate here the extent to which this may influence the results of a listing task.
The study was carried out in the municipality of Rio Claro in São Paulo State in southeastern Brazil. The municipality has a population of 168,218 (IBGE 2005) and is located in a region that has a well-developed, expanding economy.
The study comprised interviews with 400 women between the ages of 20 and 65. Women were selected to eliminate potential gender effects and because they more commonly frequent two of the research locales (plant store and supermarket). The data collection was done in the following distinct locations in Rio Claro:
1. the public garden (GAR), a plaza in the center of the city surrounded by commercial businesses and banks, where the vegetation is composed of trees, shrubs, and herbaceous plants as well as epiphytes and vines;
2. a large plant store (FLO) with a diverse number of plants, many of them in the flowering stage;
3. the fruit and vegetable section of a supermarket (SUP); and
4. a street corner in the center of town (control), where no plants were within view (CON).
The sites selected are frequented daily by people of varying backgrounds; patronage in the plant store is more restricted but is nonetheless diverse. Respondents were selected randomly; the fourth person passing by was chosen. The term selected for elicitation was plants as it is a term that is used naturally by the population in general to designate plant species. One hundred interviews were conducted in each location. In the first three locations, fifty respondents were asked to name five different plants (the short list) and fifty were asked to name ten plants (the long list). This allowed us to test the influence of visual stimuli on the length of the list. Immediately after each interview, the interviewer noted the plants cited that were within the interviewee’s field of vision (everything that could be seen clearly from the point where she stood) and (from prior work) in the surroundings (in the rest of the locale, not within clear view). In the control location, only the long lists were used. The tabulation of the data was done with the assistance of the software Tabulação de Pesquisa—Listagem Livre (Research tabulation—free listing) developed by one of the authors (Daniela Mota Miranda).
The comparisons were done between locales and types of list:
• To verify the influence of visual stimuli, a comparison was made among the plants cited and of the proportion found in the environment.
• To verify the association between locales and plant types, the plants were grouped into three categories of use (ornamental, medicinal, and food) and one category of life form (arboreal). These groupings are based on previous research (Silva-Almeida and Amorozo 1998; Fox 1999; Zar 1999; Savassi 2001; Eichemberg 2003).
• The number of citations per location of the plants most cited was also compared to verify variations in the plants’ popularity. The chi-square and the chi-square test of comparison of two proportions were used for all the above comparisons. Pearson’s correlation was used to test the association between the proportion of plants cited in the long lists that were present in the environment and the order of citation (Zar 1999; Ayres et al. 2000).
Of the 400 women interviewed, approximately 55% were 20 to 39 years old, with others ranging from 40 to 65. Approximately 30% had completed eighth grade; 38.5% had graduated from high school; and 31% had more than 12 years of schooling. Only two interviewees were illiterate. Sixty-five percent of respondents reported never having lived in a rural area, while 35% reported having had that life experience.
The Influence of Visual Stimuli
1. Comparison between locales. The mean number of citations per interview in each location was 4.92 plants in the plant store, 4.9 in the supermarket, and 4.76 in the public garden for the short lists (five citations). For the long lists (ten citations), the mean was 9.16 plants in the plant store, 9.46 in the supermarket, 8.92 in the public garden, and 8.87 in the control location. Fourteen percent of informants failed to cite the total number of names requested, claiming they were unable to remember, did not know, or were in a hurry.
The percentages of plants cited that were present within the interviewee’s field of vision or the surroundings where the interview was being conducted were highest in the plant store: 91% and 95% of the short and the long lists, respectively. In the other locations, the indices were lower, reaching 36% and 26% in the public garden and 58% and 52% in the supermarket (Table 1). In general, there is a significant difference between the locations studied in the number of plants cited that are within the field of vision, the surroundings, or both for both the long and the short list. In relation to the control location, the plant store presented the largest difference, followed by the supermarket and public garden. The comparison between the plant store, supermarket, and public garden also showed greater influence of the local environment in the plant store. The differences between the interview locales are, in general, more accentuated for the long lists than for the short lists (Table 1).
Regarding the four categories of plants considered, a significant association was found for the short lists (χ2 = 82.10; p
The apparently greater influence of visual stimuli in the plant store (on naming ornamental plants) and in the supermarket (on naming food plants) may be explained by the predominance of ornamental and food plants in the two locales. However, we cannot disregard the possible influence of the objective of a respondent’s visit to a plant store or a supermarket, nor can we rule out the effect of some plants’ being better known by name than others. In the first instance, having certain things in mind during the shopping experience must also be controlled for in further experiments. In the second, previous knowledge of names of plants would have to be controlled.
2. The plants cited most often. The ten most cited plants on the short and long lists, with the respective number of citations in each location, are shown in Figures 1 and 2. All are considered ornamental plants. Among them, the rose, orchid, fern, violet, and daisy are most consistently remembered by respondents in all locations and for both lists. We conclude that these are the most culturally important for the population interviewed. Looking at just these five plants, there were no significant differences in numbers of citations of ferns between the locations for either list. In the short list, the only significant difference was for the violet, cited more often in the plant store than in the public garden (χ2 = 4.04; p
As already mentioned, ornamental plants appear to be more relevant in the urban culture. Studies in the same region, using other methods, have produced findings that support those reported here. For example, the predominance of ornamental plants compared to edible and medicinal plants was reported by Eichemberg (2003) in a study about the characterization of old urban yards in Rio Claro, with 63% of the plants found being ornamental. In another study about diversity and use of plants in yards in Rio Claro (Fox 1999), the predominance of ornamental plants was also observed. There are various explanations for the preference for ornamental plants, such as the reduction in space of urban residences (which restricts the cultivation and presence of plants), the ready availability of commercial food and medicinal products, and the insertion of women into the workforce.
3. List size and order of citation of the items. A comparison of short and long lists in each locale provided no conclusive results. In the plant store, there is a tendency toward a significantly greater influence of the environment on the responses to the long lists. In the public garden and the supermarket, however, there is the opposite tendency, with responses to short lists appearing to be more influenced by the surroundings (Table 4). Nor was there a correlation, in any of the three locales with plants, between the proportion of plants cited that were present in the environment and the order in which they were cited (Pearson correlation coefficient p > .05).
The comparisons and analyses presented indicate that visual stimuli and the context of the situation in which the interviewees find themselves influence responses to a listing task about plants. This influence varies by locale of the research. To be included in the responses, the plants present in the environment have to be known and recognized as pertaining to the domain elicited. The analyses of the data collected reveal a more marked influence in the plant store compared to the other locations. In this case, the existing tendency toward cultural emphasis on ornamental plants in the study population is reinforced. Our results also suggest that there may be an indirect influence of the environment. The type of plant or products present can influence the type of plant recalled. Thus, edible plants were recalled more frequently in the supermarket, whereas flowers were remembered more often in the plant store. Thus, the type of plant present may trigger the memory of similar plants, by whatever characteristic—utility, general appearance, plant habit, and so on— acting as a “semantic cue.” A parallel can be drawn with the findings of Brewer (2002) and Brewer, Garrett, and Rinaldi (2002), who found that items elicited in free-listing tasks are very efficient semantic cues for later elicitation of similar items in successive listing tasks. Visual stimuli and the context of the interview show association with features of our listing tasks and thus appear to influence responses to those tasks. However, our data are not sufficient to establish the mechanism for this association. In the future, free lists rather than numerically constrained lists should be used, and research is needed to control the exposure of each interviewee to standardized visual stimuli. Listing tasks allow us to obtain a set of culturally relevant terms and are powerful tools for ethnobotanical research. Nevertheless, those who use these techniques must remain attentive to the environment in which interviews are conducted to minimize external influences.
Ayres, M., M. Ayres Jr., D. L. Ayres, and A. S. Santos. 2000. BioEstat 2.0—Aplicações estatísticas nas áreas das ciências biológicas e médicas. Belém, Brazil: Sociedade Civil Mamirauá, MCT/CNPq.
Benz, B. F., J. E. Cevallos, F. M. Santana, J. A. Rosales, and S. M. Graf. 2000. Losing knowledge about plant use in the Sierra de Manantlan Biosphere Reserve, Mexico. Economic Botany 54 (2): 183–91.
Berlin, B. 1992. Ethnobiological classification—Principles of categorization of plants and animals in traditional societies. Princeton, NJ: Princeton University Press.
Bernard, H. R. 1988. Research methods in cultural anthropology. Newbury Park, CA: Sage.
Borgatti, S. P. 1994. Cultural domain analysis. Journal of Quantitative Anthropology 4:261–78.
Boster, J. 1984. Inferring decision making from preferences and behavior: An analysis of Aguaruna Jívaro manioc selection. Human Ecology 12 (4): 343–58.
Brewer, D. 2002. Supplementary interviewing techniques to maximize output in free listing tasks. Field Methods 14 (1): 108–18.
Brewer, D., S. B. Garrett, and G. Rinaldi. 2002. Free-listed items are effective cues for eliciting additional items in semantic domains. Applied Cognitive Psychology 16:343–58.
Cotton, C. M. 1996. Ethnobotany—Principles and applications. Chichester, UK: John Wiley.
Eichemberg, M. T. 2003. Os quintais antigos na área urbana de Rio Claro–SP: Um enfoque etnobotânico. MSc dissertation, Instituto de Biociências, Universidade Estadual Paulista, Rio Claro, Brazil.
Fox,V. C. 1999. Diversidade e uso de plantas nos quintais de Ferraz, município de Rio Claro–SP. Monograph, Instituto de Biociências, Universidade Estadual Paulista, Rio Claro, Brazil. Instituto Brasileiro de Geografia e Estatística (IBGE). 2005. Available at http://www.ibge .gov.br. Accessed 20 June 2005.
Martin, G. 1995. Ethnobotany—A methods manual. London: Chapman & Hall.
Nolan, J. M. 1998. The roots of tradition: Social ecology, cultural geography and medicinal plant knowledge in the Ozarks–Quachita Highlands. Journal of Ethnobiology 18 (2): 249–69.
Savassi, A. P. G. 2001. Levantamento do conhecimento popular sobre plantas medicinais em Ajapí, município de Rio Claro, SP, Brasil. Monograph, Instituto de Biociências, Universidade Estadual Paulista, Rio Claro, Brazil.
Silva-Almeida, M. F., and M. C. M. Amorozo. 1998. Medicina popular no distrito de Ferraz, Município de Rio Claro, Estado de São Paulo. Brazilian Journal of Ecology 2:36–46.
Sutrop, U. 2001. List task and a cognitive salience index. Field Methods 13 (3): 263–76.
Weller, S. C., and A. K. Romney.1988. Systematic data collection. Newbury Park, CA: Sage.
Zar, J. H. 1999. Biostatistical analysis. 4th ed. Upper Saddle River, NJ: Prentice Hall.