The setting
The study took place from April 2007-April 2008 in Tabi, a small Yucatec Maya rural village located in the central part of the state of Yucatan, Mexico. The economy of the community is mostly based on subsistence agriculture, although increasingly the youth are migrating to the regional cities to perform low-skilled jobs. The communally held land (known as
ejido) surrounding the village is characterized by low deciduous forest [
42], which is utilized in the process of swidden agriculture practiced by most farmers. There is a public kindergarten, elementary school (grades 1–6), and middle school (grades 7–9) located in Tabi. The nearest high school (grades 10–12), which is also state-run, is located 11 kilometers from Tabi in Sotuta, the municipal center. Bilingualism is common in the community (89%) with only 9.5% the population monolingual in Yucatec Maya, and 1.5% in Spanish [
43]. The massive Catholic Church that dominates the main plaza dates back to the 1700s. Smaller Protestant churches have been built in Tabi over the last several decades as an increasing number of people are moving away from the Catholic Church. Access to modern medicine has also increased during the same time period. In the late 1980s, a small, state-run health care clinic with a visiting physician was established in the community, and then in 2002, a yearly rotating medical resident from one of the medical schools was permanently installed in the clinic. Prescription pharmaceuticals are available at the state-run clinic, and over-the-counter pharmaceuticals at one of the six privately-owned corner stores in the community.
Type of participants involved
The village is comprised of 698 people living in 122 households. The study design was based on households, and one adult from almost every household in Tabi participated in the study (n = 116). The adult selected was the most knowledgeable about medicinal plants in the household as determined by members of the household. In households where no one was considered knowledgeable, the participant was randomly chosen. Slightly more than half of the participants were female (54.3%), and ranged in age from 16 to 87, with a mean age of 46.4 (SD = 17.2) (Table
1). Although few participants received no formal education (15.5%), only 6.8% went on to middle school, and 30.2% were illiterate.
Table 1
Participant demographics (n = 116)
Age in years [Mean (SD)] | 46.4 (17.2) |
Gender [Percent] | |
Male | 45.7 |
Female | 54.3 |
Literate in Spanish [Percent] | |
Yes | 69.8 |
No | 30.2 |
Education [Percent] | |
No formal schooling | 15.5 |
Completed grades 1-3 | 46.5 |
Completed grades 4-6 | 31.1 |
Completed grades 7-9 | 6.8 |
Religion [Percent] | |
Agnostic/Atheist | 20.7 |
Catholic | 45.7 |
Protestant | 33.6 |
Livelihood [Percent] | |
Homemaker | 54.3 |
Full-Time Subsistence Farmers | 31.9 |
Subsistence Farmers and Temporary Wage Laborers | 12.1 |
Full-Time Wage laborers | 1.7 |
Places of residence [Percent] | |
Only Tabi | 81 |
Other regional small towns | 6 |
Regional cities (Merida or Cancun) | 13 |
Farthest Distance Travelled in Kilometers as the Crow Flies [Mean (SD)] | 254.5 (570.9) |
Relative lifestyle scale [Percent] | |
0 (Very traditional lifestyle) | 1.7 |
1 | 3.4 |
2 | 26.7 |
3 | 22.4 |
4 | 24.1 |
5 | 19.8 |
6 (Very modern lifestyle) | 1.7 |
Relative economic prosperity scale [Percent] | |
1 (Very low relative economic prosperity) | 8.6 |
2 | 31.9 |
3 | 34.5 |
4 | 20.7 |
5 (Very high relative economic prosperity) | 4.3 |
Interest in medicinal plants [Percent] | |
Very interested and interested | 93.9 |
Neutral | 0.9 |
Very uninterested and uninterested | 5.2 |
Used most to treat common illness [Percent] | |
Medicinal Plants | 69 |
Pharmaceuticals | 30.2 |
God’s Will | 0.9 |
General course of action for common illness [Percent] | |
Treat the illness at home | 66.4 |
Visit a family member or acquaintance to treat the illness | 3.4 |
Visit a traditional healer to treat the illness | 0.9 |
Visit a conventional medical doctor to treat the illness | 28.4 |
Have faith in God’s will | 0.9 |
The most common religion in Tabi continues to be Catholicism (45.7%); however, many people had joined one of the Protestant churches in the community (33.6%) or had no religious affiliation (20.7%). All of the women were homemakers who also care for the small livestock raised and horticultural crops grown within the area surrounding their houses. The majority of the men (69.8%) dedicated all their work time to tending the agricultural fields and large livestock on ejido land away from the home. Some men (26.4%) primarily performed temporary wage labor, and tended their fields the rest of the time, and only a very few (3.8%) were employed as full-time wage laborers. Thus, most of the participants spent much of their day outside in nature.
The majority of participants (81%) had lived in Tabi their entire lives. Of the people who lived elsewhere, most lived in Merida or Cancun (13%), the two largest regional cities; the remaining few (6%) lived in other small towns on the Yucatan Peninsula. The average distance people from Tabi had travelled as the crow flies is 254.5 kilometers. The farthest the majority of people travelled were to Merida (40.5%) or Cancun (27.6%). Of those who had travelled farther, the most common destination was Mexico City (5%). Two participants travelled to the United States and Canada as guest farm workers.
Almost three-quarters of the participants practiced a mix of traditional and modern lifestyles as determined by a series of items assessed through participant observation, and identified by participants as key indicators of a more modern lifestyle, including: consumption of leavened bread; occasional or frequent purchasing of machine-made tortillas; preference for speaking Spanish with children; preference for speaking Spanish with other adults; preference for wearing Western clothes; and ownership of a bed (see [
44]). Based on a series of economic indicators assessed through participant observation, and identified by participants (including ownership of a hammock, television, stereo, refrigerator, and/or stove), the majority of participants (87.1%) had moderate economic prosperity. Almost all participants reported an interest in medicinal plants (94%). Over three-quarters of participants (88.8%) had used medicinal plants to treat their children, and over two-thirds of participants (69%) had used medicinal plants more often than pharmaceuticals (30.2%) to treat common illnesses. Additionally, most respondents (66.4%) reported treating common illnesses at home.
Data collection and analysis
This was a cross-sectional study, meaning the variables were measured at one point in time among one group of people [
45]. This research project was approved by the University of Florida Institutional Review Board on April 3, 2007 (IRB protocol number 2007-U-0259), and verbal informed consent was received from all participants prior to participating in the study per IRB approval. Multiple data collection and analysis techniques were utilized in this study to address our five objectives, including cultural consensus analysis (CCA), social network analysis (SNA), and surveys and statistical analyses [
39].
Our first objective was to assess variation in herbal remedy knowledge in Tabi. First, we administered an herbal remedy knowledge questionnaire, and then we ran CCA. Typically CCA is a two part process. The first is a free-listing task asking respondents in the cultural context to list items, in this case herbal remedies consisting of a plant and an illness the plant can help treat. We did this task with 40 respondents in Tabi. The second part of CCA is a systematics survey asking respondents about a subset of remedies listed by the first group. For our study the survey developed in the second step consisted of 43 questions in the format “Can
plant x cure
illness y?”. We created the questionnaire by substituting
plant x and
illness y in approximately half of the questions in plant and illness combinations that had been free listed by at least 2 out of 40 people from the first step. The other half of the questions were plant and illness combinations that were not known remedies in Tabi or in the Yucatec Maya ethnobotanical literature. These items were added because CCA is sensitive to greatly unequal proportions of yes and no responses [
46]. For each question, the response choices were “yes”, “no”, or “uncertain”. In cases where the individual was uncertain, the response was imputed by a simulated coin toss since CCA operates under the assumption that respondents will guess when uncertain [
46,
47].
We analyzed the data by running a CCA in UCINET, which measured individual variation in agreement of botanical remedies among households within Tabi and determined if there was one common culture of herbal remedy knowledge in Tabi [
47,
48]. The data from Tabi did not violate any of the model’s assumptions, thus there is one common culture of herbal remedy knowledge, and the first factor values represent individual competence scores for that knowledge [
39]. A more detailed description of this methodology and some of the results have been previously published [
39,
40]. In summary, the average competence scores was 0.64 (SD = 0.20) with a range from a low level of competence at 0.08 to a high level of competence at 0.95. To put this in perspective, a competence level of 1.0 would mean that a participant’s responses to herbal remedy survey matched the group consensus in responses. A score of zero means they never agreed with the group. The relatively large standard deviation and the wide range in competence scores indicate that across households in Tabi, there is a great deal of variation in agreement about botanical remedies.
Our second objective was to describe the herbal remedy knowledge network in Tabi, and determine participants’ positions within that network. First, we asked each participant: “Have you asked
person x about herbal remedies?” substituting the names of each of the other 115 participants for
person x. Then, we analyzed the relational data using UCINET and NetDraw [
48]. UCINET produced several overall measures of network structure and individual position within the network (centrality measures), and NetDraw provided a visualization of the whole network. Pearson’s correlations were performed to assess the relationship between competence scores and relational measures. We have reported on this methodology and some of the results in a previous publication [
39]. In this paper we present new data and a novel focus specifically on in-degree, the only centrality measure that was correlated with competence scores (
r = 0.28,
p < 0.01,
N = 116). In-degree represents the number of individuals who asked a participant about medicinal plant remedies.
Our third objective was to measure individual attributes of the participants. Attributes of network members can be used to try to explain their network position. These attributes were collected by administering a survey. The questions in the survey were informed by previous research on intra-cultural variation in traditional ecological knowledge, and ethnographic information obtained during the first six months of fieldwork. The questions focused on age [
4-
8], gender [
7,
9-
15], livelihood strategies [
7,
12,
16,
17], level of formal education [
8,
18-
20], religious affiliations [
8,
9], economic prosperity [
8,
12,
17,
22,
23], lifestyle [
8,
9,
24-
27], range and migration [
8,
21], treatment preferences and perceived interest in herbal medicine [
7,
24,
28,
29] since these are factors that have been associated with knowledge variation. Descriptive statistics were performed and the results are presented in Table
1. Pearson’s correlations and ANOVAs were performed to assess the relationship between competence scores, and continuous and categorical attribute variables, respectively. A general description of the results and a discussion of the inferential findings have been previously reported [
44].
Our fourth objective was to test the hypothesis that greater individual herbal remedy knowledge is positively associated with the position of individuals within the botanical remedy knowledge network; that is, the more people report talking to someone about remedies the more that person tends to agree with the group. The alternative, a lack of association, would suggest that people report talking about remedies with those who are not experts. We did this test by running regressions using permutation tests with competence scores as the outcome variable, and individual centrality measures and the attribute variables as the explanatory variables [
49]. Description of these analyses and results are previously published [
39], but in summary the model predicted 26.1% of the variation in competence scores (
F = 3.34,
p = 0.04,
N = 116) and age masked the association between competence scores and all other variables including in-degree (
B = 0.01,
p < .01,
N = 116). We then ran a Pearson’s correlation between age and in-degree and determined that they are positively and strongly associated (
r = 0.48,
p < 0.01), suggesting that not only are older individuals more knowledgeable about herbal remedies, but they are also a source of information about herbal remedies for more people.
Our fifth objective, and the primary focus of this paper, was to determine if the relationship between age and knowledge identified in objective four was consistent across age groups. We found that overall competence increases with age, but this pattern may not be the same for all age cohorts. If the pattern differs by age group and is associated with an event or major change in the community, then it provides evidence for changes in acquisition patterns which may be resulting in knowledge loss [
25,
41]. We divided the participants into two different age cohorts, and ran Pearson’s correlations between competence scores, in-degree and age variables. The age cohorts were 16–45 and 46–87 for age and competence, and 16–50 and 51–87 for age and in-degree. We also conducted ethnographic interviews with a subsample of 20 participants related to the findings to foster interpretation. The interviews were focused on community history, and patterns in acquisition and transmission.