A sociolinguistic survey report of the Zay people in Ethiopia

DigitalResources Electronic Survey Report 2011-046 ® A Sociolinguistic Survey Report of the Zay People in Ethiopia Linda Jordan Jillian Netzley Hussein Mohammed !! " " #$%% SIL Electronic Survey Report 2011-046, November 2011 © 2011 and SIL International® All rights reserved 2 Abstract 1. Introduction 1.1 Geography 1.2 People and language 1.3 Other previous research 2. Goals of the research 3. Methodology 3.1 Procedures 3.2 Data sources 3.3 Analysis 4. Results 4.1 Sociolinguistic questionnaires 4.2 Community ownership and support questionnaire 4.3 Oromo SRT 4.4 Amharic RTT 5. Data analysis 5.1 Sociolinguistic questionnaire 5.2 Community ownership and support questionnaire 5.3 Mean SRT scores for men and women 5.4 Mean RTT scores for men and women 6. Conclusions and recommendations Appendix A. Sociolinguistic questionnaire Appendix B. B.1. Descriptive statistics - SRT B.2 Two sample T-test and confidence interval B D B References 3 Abstract The optimal language for literature and educational materials is not the same for all Zay areas. The data gathered during the current study points to Zay as optimal for the islands on Lake Ziway and Oromo as optimal for the lakeshores. However, the Zay people living on the islands would probably be well served by Amharic literature and educational materials until most of them immigrate to the shore or the Oromo educational system causes a shift in preference to Oromo. Zay’s case is one of an endangered language that could prove to be a development success story, but only if the level of motivation for a language development project is high enough to initiate and sustain the effort. 1. Introduction The Zay have a very strong Orthodox Christian tradition, inhabiting islands that each host a church or monastery, one which was thought to have been a resting place for the Ark of the Covenant. Amhara priests were brought in to serve and subsequently brought the Amharic language to the Zay people, who use it as their foremost language of religious activities. In addition to this, they live in a strongly Oromo area, and schooling is conducted in the Oromo language. The Zay language, still widely spoken on the islands, is classified among the HarariEast Gurage languages of the Ethio-Semitic subfamily (Gordon 2005). 1.1 Geography The Zay people live on the islands of and areas surrounding Ethiopia’s Lake Ziway (see figure: 1), located approximately 160 kilometers south of Addis Ababa. According to the current administrative divisions, most Zay people now live in the East Shewa and Arsi zones of Oromiya Region. The initial survey of Zay, conducted by SIL Ethiopia and the Institute of Ethiopian Studies (IES) (Wedekind and Wedekind 2002), found a Zay population in the following areas: Herera, Mek’i, Boch’eessa (near Bojji), Ziway (Zway Town), Fundurro Island, Tsedecha Island, and Debra Tsiyon Island (Tullu Guddo). The second joint survey (Gardner and Siebert 2002) discovered that a few Zay were also living on Gelilla Island and Debre Sina Island. 4 Figure 1. Rthiopia’s Lake Ziway and surrounding area. 1.2 People and language The 1994 Population and Housing Census of Ethiopia contains no information with specific reference to the Zay people. During the two previously-mentioned surveys, there were estimated to be 4,880 mother-tongue speakers of Zay. The neighboring Oromo know the people as Lak'i (Laqi). The language has also been known as Ziway or Gelilla (Gordon 2005). According to Bender’s (1971) lexicostatistical analysis, Zay has 61 percent lexical similarity with Harari and 70 percent with Silt’e. The report of the first SIL-IES survey also contains Zay and Silt’e wordlists as well as information on the population of Zay speakers in different Zay areas, including a map (Wedekind and Wedekind 2002). The second SIL-IES trip to the Zay area had the purpose of gathering sociolinguistic data on bilingualism, language use, and language attitudes (Gardner and Siebert 2002). Sociolinguistic questionnaires were administered to 14 men of varying ages and educational backgrounds, all of whom were native Zay speakers from the areas surrounding the lake. The interviewees were found to have a positive attitude toward their mother tongue, stating that their language will be 5 used for many generations. However, only one subject disputed the idea that young people now speak Zay less frequently, while seven confirmed that some children understand Zay but will not speak it. The Zay people were also found to have positive attitudes toward Oromo and Amharic, which was the language of instruction in their schools at that time. According to the interviews, the Zay people believed Oromo and Zay to be equally important languages to pass along to their children. Leslau (1999) gives a detailed description of the Zay language, based on data collected during his trip to the Ziway area in 1950. His topics include phonology, pronouns, nouns, adjectives, numerals, verbs, positional relations, conjunctions, subordinate clauses, adverbs, enclitics, the position of Zay within East Gurage, and an English-Zay, Zay-English dictionary. 1.3 Other previous research During the First International Conference on Endangered Languages of Ethiopia (Addis Ababa, 27–30 April 2005), Endeshaw Woldemariam (2005) of Addis Ababa University presented a paper about the impact of migration on language shift in the Zay area. He observed that the same economic pressures that force Zay islanders to move to the shores of Lake Ziway also push the Zay language toward extinction. While Zay is preserved on the islands, migration brings the people into greater contact with surrounding languages (particularly Oromo), which have begun to replace Zay as their mother tongue. 2. Goals of the research The main research goal of the survey was to determine the optimal language or languages for literature and educational materials in different ethnically Zay areas. The main concepts involved in addressing this question are language attitude, bilingualism and language vitality. The objectives pertinent to these concepts include assessing the people’s attitudes toward Zay, Oromo and Amharic, testing their proficiency in Oromo and Amharic and investigating the vernacular language vitality in ethnically Zay areas. Additional research questions were investigated by: • determining the likelihood that a Zay language development project would be valued by the following: 6 • investigating the level of interest and motivation for a language-development project in Zay communities and • investigating the likelihood that significant portions of the Zay population will learn to read and write the vernacular. 3. Methodology Four different methods were used to accomplish the research goals. The procedures, data sources, and analysis techniques for each are described in the following paragraphs. 3.1 Procedures The survey team members used Amharic when administering the questionnaires and tests described in the following paragraphs. The level of Amharic proficiency encountered in the study area was usually high enough for the participants to comprehend and respond to interview questions in that language. In general, they were also able to understand the test procedures as described in Amharic. When necessary, local interpreters were available if clarification was needed in either Zay or Oromo. 3.1.1 Sociolinguistic questionnaires In each community visited during this survey, a group sociolinguistic interview was conducted in order to get an overall picture of multilingualism and language attitudes among the Zay (see appendix A). The completion of each interview took several hours and covered the areas of multilingualism, language use, language attitudes, dialect attitudes, social interaction patterns, language vitality, and language development. Group interviews were followed by individual interviews with a cross section of the population. These focused more on individual attitudes toward languages of wider communication, covering the issues of language attitudes, social interaction, language vitality, and language development. 3.1.2 Community ownership and support questionnaire 7 A support evaluation questionnaire was used in order to investigate the level of interest, motivation and support for a potential language-development project in Zay communities. This questionnaire was adapted from a community ownership continuum developed in the Philippines (Benn. 2004). Community leaders were asked approximately 15 variables that cover a wide range of factors affecting community support for language development. These include the following: • value placed by the community on using Zay in different domains, • commitment to the production of Zay literature, and • financial aspects of literature production and use. For each variable, four possible choices were given. The first choice, worth one point, represented the lowest level of support. The fourth choice, representing the highest level of support, was awarded four points. The second and third choices reflected moderate to good levels of support. The point total for the choices given by each community leader was calculated, in addition to average scores for each location and for all the leaders as a group. 3.1.3 Sentence Repetition Test (SRT) The Oromo SRT was developed and administered as described by Radloff (1991). Test subjects listened to 15 Oromo sentences of increasing difficulty and tried to mimic each one in turn. Each sentence was scored on a three-point scale according to how many errors were committed in the repetition. The points for each individual sentence were added to get the total score, which is an indicator of Oromo proficiency. 3.1.4 Recorded Text Test (RTT) The RTT protocol for this survey was a modified form of the standard RTT as described by Casad (1974). The initial 30-question pilot test was administered to ten native speakers of Amharic from the Amhara areas of Gonder and Gojjam. If anyone missed a question, that question was discarded. Of the remaining questions, the ten that 8 represented the greatest variety of semantic categories were selected for the Amharic RTT. Comprehension was tested by playing the Amharic text to individual Zay subjects and pausing at intervals to ask questions in Amharic, for a total of ten questions asked during the text test. The questions were asked orally instead of editing them into the recorded text, because test subjects in rural areas often have difficulty distinguishing recorded questions from the rest of the text. 3.2 Data sources 3.2.1 Sociolinguistic questionnaires Each of the two interview groups was to be comprised of at least ten participants age 15 and older, preferably both men and women from a range of ages. The survey team waited to begin the interview until at least ten participants were present. Since such an interview is often held in an informal setting, participants come and go as they see fit. The information is taken as being from the group as a whole, instead of from individuals within that group. A total of 19 individual sociolinguistic interviews were also conducted, ten on Tullu Guddo Island and nine in Boch’eessa. The sample at each location was comprised of at least two of the following: 3.2.2 • males / females, • with / without some formal education, and • below / above 25 years of age. Community ownership and support questionnaire The community support evaluation questionnaire was first administered to ten traditional, political or religious leaders in the villages of Durumejja and Masno on Tullu Guddo Island. After that, ten more leaders were interviewed in the villages of Koka and Bojji in the Boch'eessa area, for a total of 20 interviewees in two different areas. 9 3.2.3 SRT protocol The SRT protocol involved testing a sample of the residents of each area. The original goal was to test 30 people at each location, ideally with five men and five women being tested in each of three age categories (10–24, 25–34 and 35+). A full sample of 30 was obtained on the island, but only 16 people were tested on the shore, because it was apparent that Oromo had become that community’s first language. The sampling method used was stratified judgment (quota) sampling, using the community’s social networks to find test subjects as described by Radloff (1991). Age, sex and education were judged to be the variables impacting Oromo bilingualism in this area, as Oromo is the language of education in the region where this survey was conducted. Although the researchers did not sample for education level specifically, the highest level of education obtained by each subject was noted for later analysis. 3.2.4 Sampling for the RTT Not much effort was spent on sampling for the RTT because it is not as powerful a tool for testing bilingualism as the SRT. During the development of the Amharic RTT, it was not calibrated to an independent measure of second language proficiency, as was the Oromo SRT (Radloff. 1991). Therefore, it was used only to get a general idea of Amharic ability relative to other Zay test subjects and mother-tongue speakers of Amharic. The same basic rules for sampling, as described for the SRT (see section 3.2.3), were still used as a guide. 3.3 Analysis 3.3.1 Sociolinguistic questionnaires The answers to interview questions in the two different locations were compared and evaluated in light of other observations and survey results. This was done for both the group and individual interviews. 10 3.3.2 Community ownership and support questionnaire After the point totals and averages were calculated, the result for each location was grouped into one of four categories for purposes of evaluation. These levels of support were roughly defined, according to point averages, as follows: • low: 15–25, • moderate: 25–35, • good: 35–45, and • excellent: 45+. More importantly, the actual averages for the locations were compared, giving some indication of the relative amount of support for language development in the communities that were visited. The results were also analyzed in light of the information gathered during the sociolinguistic interviews. 3.3.3 SRTanalysis The SRT scores were compiled and compared in the following two different ways. First, the mean and standard deviation were first calculated for men, women, each age group and each location, as well as for different groupings according to educational level and for the entire test population (see 4.3). The means were then compared, using the two-sample t-test to see if the differences between them are statistically significant. Secondly, the percentage of test subjects who scored 25 or lower was calculated for the age-sex categories within the entire population, each test location, and each education grouping. A score of 25 represents a proficiency level of 2+ according to the scale of the Reported Proficiency Evaluation (RPE) to which the SRT was calibrated (Radloff 1991). In other words, the percentage of test subjects who are likely at less than level 3 (good, general proficiency) was calculated. 11 3.3.4 RTT analysis As for the SRT scores, the mean and standard deviation were first calculated for all the different groupings (see section 4.4). The means were also compared with the twosample t-test as previously mentioned. 4. Results 4.1 Sociolinguistic questionnaires 4.1.1 Multilingualism The people call themselves Zay and their language Zayigna. Others call them Lak'i, a name that the Tullu Guddo interviewees do not like. In Boch'eessa, where there is much stronger Oromo influence, both Zay and Lak'i are neutral names. All the Tullu Guddo group interviewees, their spouses, and most of their parents can speak Zay, Oromo, and Amharic. The Boch’eessa interviewees and most of their spouses and parents can also speak all three languages. On Tullu Guddo Island, most of the interviewees and their parents speak Zay as their first language, whereas it is the first language of all their spouses and children (see table 1). In Boch’eessa, most of the interviewees, their spouses, and parents speak Zay as their first language, but the children’s first language is Oromo (see table 2). The language of education in both areas is currently Oromo. Table 1. First language of groups interviewed on Tullu Guddo Island Group Zay Amharic Oromo Kistane Total Respondents 9 2 1 ---- 12 Fathers 10 ---- 1 1 12 Mothers 10 ---- 1 1 12 Spouses 12 ---- ---- ---- 12 Children 12 ---- ---- --- 12 12 Table 2. First language of groups interviewed in Boch’eessa 4.1.2 Group Zay Oromo Kistane Silt'e Total Respondents 10 5 ---- ---- 15 Fathers 14 1 ---- 15 Mothers 12 1 2 ---- 15 Spouses 8 4 2 1 15 Children ---- 15 ---- --- 15 Language use On Tullu Guddo Island, Zay is the only language used within the family, while Oromo is the only language used when communicating with siblings and children in Boch’eessa. Zay, Oromo, or some Amharic is used when communicating with parents or spouses in Boch’eessa. Zay is also used when communicating with friends or elders on Tullu Guddo Island but, in Boch’eessa, Oromo is the only language used in such situations. In town, where matters of administration are concerned or at markets, the Tullu Guddo and Boch’eessa interviewees all use Oromo or Amharic. Amharic is used for public worship in both places. Zay is used when the Tullu Guddo interviewees pray at home, and either Zay or Oromo is used for prayer at home in Boch’eessa. Amharic is also used for counting money or other things on Tullu Guddo Island, and either Amharic or Oromo is used for these purposes in Boch’eessa. 4.1.3 Language attitudes All the Tullu Guddo interviewees have a positive attitude toward Zay, as well as Amharic. In addition, three of the individual interviewees said that Oromo would be their third choice, while Gurage is the third choice for one interviewee. The Boch’eessa group stated that they have a positive attitude toward Oromod; all the individual interviewees there have a positive attitude toward Amharic as well as Oromo. However, for six 13 individuals, Amharic would be their first choice, while Oromo is first choice for three of them. All interviewees on Tullu Guddo Island said that they would like their children to learn Amharic. The Boch’eessa group interviewees would like their children to learn Amharic as well as Oromo. Four of the individual interviewees there would like their children to learn Amharic, two would like English, one would like Amharic and Oromo, one would like Oromo and Zay, and one would like Zay. According to the Tullu Guddo group, marrying an Oromo, Amhara, or Gurage is not forbidden for a Zay person. Only one individual interviewee there does not approve of marriage outside the ethnic group, but the other interviewees have a variety of ethnic preferences1 for intermarriage (see table 3). The Boch’eessa group interviewees stated that marrying an Oromo or Gurage is not forbidden, and all of the individual interviewees there would accept outside marriage (see table 4). Table 3. Intermarriage preferences of individual interviewees on Tullu Guddo Island 1 Preferences Number No intermarriage 1 Amhara or Gurage 3 Amhara or Oromo 2 Oromo or Gurage 1 Amhara 1 Tigre 1 Any Christian 1 Total 10 The first choice is always to marry a Zay. 14 Table 4. Intermarriage preferences of individual interviewees in Boch’eessa Preferences Number No intermarriage ---- Oromo 3 Gurage or Oromo 3 Oromo or Amhara 2 Gurage 1 Total 9 On Tullu Guddo Island, the attitude toward intermarriage seems quite different when questioned from another angle. Six of the individual interviewees would not allow their children to marry outside the ethnic group. Two people indicated they would accept it if their children married Amhara; one man said that it would be acceptable as long as his children married Christians. Another man said that he would like his children to marry educated people. In Boch’eessa, all the individual interviewees would allow their children to marry outsiders; however, their preferences differ. Two people would accept it if their children married Amhara, and one said that it would be all right if his children married Oromo. Three interviewees would like their children to marry Amhara, Gurage, or Tigre. For three other interviewees, it would be acceptable if their children married Christians of any ethnic group. 4.1.4 Attitudes to dialects Dialect variations in Zay are insignificant. According to the Tullu Guddo group interviewees, Zay is closer to Silt’e than to any other language. They mentioned that Masno (a village on Tullu Guddo Island) is the best place to learn Zay. The Boch’eessa group members agree that dialect variation is not an important issue. They believe that the islands in Lake Ziway are the best places to learn the Zay language. 15 4.1.5 Social interaction patterns The Zay of Tullu Guddo Island trade with Ziway Town, Mek’i, K’idame Gebeya, Asella, and Ch’effee Jila. Those who live in Boch’eessa trade with Ziway Town, Bulbula, Mek’i, Asella, and Adami Tullu. In both places, they use Amharic or Oromo for communication when they go to the villages to trade, depending on whom they meet. Nine of ten individual interviewees on Tullu Guddo Island have some sort of traditional (i.e. godparent-godchild), marriage or blood relationship with Gurage, Amhara, or Tigre people. Three of nine interviewees in Boch’eessa are in some way related to Amhara and three to Oromo. Two interviewees there have Silt’e or Gurage relatives by marriage, and one interviewee has Oromo and Silt’e relatives. The Zay of Boch’eessa also live together and interact with Oromo friends in different ceremonies, like weddings and funerals. 4.1.6 Language vitality The Zay language is alive and still spoken by children on the islands, but the children who attend school learn in Oromo. The Tullu Guddo Island group mentioned that, in other places like Boch’eessa and Mak’dela, every child now has Oromo as their first language. The Boch’eessa group and individual interviewees confirmed this. Only older people use the Zay language there. The group interviewees on Tullu Guddo Island doubt that their language will be used for many generations. The individual interviewees there have different perceptions of the language’s vitality. Only four people believe that the next generation will use it, while another four doubt that it will be used. Two interviewees have no idea whether their language will continue to be used or not. Because people are leaving the islands from time to time and Oromo is highly influencing Zay, many are afraid that their language might die out. The Boch’eessa group interviewees believe that, in the near future no one is going to use Zay, and the individual interviewees agree that the generation to come will not use it in their area. 4.1.7 Language development The Tullu Guddo Island individual interviewees expressed their interest in the development of the language and said that they would be happy if something on Zay 16 history or culture would be published. The Boch’eessa interviewees also expressed interest, saying that they would be happy if Zay were written and preserved. Tullu Guddo Island residents have a positive attitude toward publications in Amharic. They also said that they would buy publications in Zay if they were available. The Zay of Boch’eessa have a positive attitude toward publications in Oromo. Group interviewees at both Tullu Guddo and Boch’eessa said that, if there were schools to teach them to read and write in the Zay language they would go to them, but they have not yet seen anything written in Zay. Both communities would also be happy to have a Zay radio program. 4.2 Community ownership and support questionnaire On 19 and 20 April 2005, ten people were interviewed in the villages of Durumejja and Masno on Tullu Guddo Island. On 21 April 2005, ten more people were interviewed in the villages of Koka and Bojji in the Boch’eessa area. The purpose of these interviews was to evaluate the level of interest in developing Zay. The average scores obtained are summarized in table 5. Table 5. Summary of average scores Number of Average scores Evaluation interviewees Tullu Guddo 10 38 Good Boch’eessa 10 30 Moderate Total 20 34 Moderate 4.3 Oromo SRT The Oromo SRT was administered in the communities on Tullu Guddo Island, one of the largest islands of Lake Ziway, and in the village of Boch’eessa outside of Ziway Town, on the shore surrounding Lake Ziway. Table 6 lists the mean SRT scores and the standard deviations within each category. 17 Table 6. Oromo mean SRT scores and standard deviations within each category Category Mean Standard deviation Population 27.15 10.20 Male 26.04 10.76 Female 28.48 9.58 Ages 10–24 29.15 11.25 Ages 25–34 29.92 8.64 Ages 35+ 24.33 10.08 Island 24.93 10.43 Shore 31.31 8.58 No formal education 22.08 9.36 29.53 8.66 28.75 11.42 28.19 10.37 32.71 7.74 Attended grade 2, 3, 4, or 5 Attended grade 6 or higher Attended grade 2, 3, 4, 5, or 6 Attended grade 7 or higher 4.4 Amharic RTT The Amharic RTT was administered in the same two places as the SRT. Table 7 lists the mean RTT scores and the standard deviations within each category. 18 Table 7. Amharic mean SRT scores and standard deviations within each category Category Mean Standard deviation Population 81.74 16.37 Male 82.69 15.89 Female 80.50 17.31 Ages 10–24 78.89 16.41 Ages 25–34 88.18 12.50 Ages 35+ 80.59 18.19 Island 80.62 16.64 Shore 84.29 16.04 No formal education 73.57 18.23 86.00 11.74 85.00 15.66 82.63 14.85 89.23 13.20 Attended grade 2, 3, 4, or 5 Attended grade 6 or higher Attended grade 2, 3, 4, 5, or 6 Attended grade 7 or higher 19 5. Data analysis 5.1 Sociolinguistic questionnaire Determining the optimal language for literature and educational materials for the Zay people was the main research goal of this survey. This section tries to answer this question from the sociolinguistic aspect, specifically in relation to language vitality, bilingualism, and attitudes. Generally speaking, the Zay people use three languages: Zay, Oromo, and Amharic. The use of Zay on the lakeshore is decreasing, if not completely dying out. A definite split can be seen between the generation of the Boch’eessa interviewees and that of their children (see table 2). Oromo is the mother tongue of the younger people in that area and no doubt will increasingly replace Zay over the years, becoming the first language of the Zay community on the shores of Lake Ziway. It is interesting to note that Amharic is used as the language of public worship in both locations, probably because the Zay are predominantly Orthodox Christians. This may also partially explain why such a positive attitude toward Amharic was found among the Zay of Boch’eessa, in a mostly Oromo-speaking area. Six out of nine individual interviewees there actually listed Amharic as their first choice of language. Although the attitude toward Amharic seems to be uniformly high among the Zay, there is a sharp distinction between the attitudes toward Zay and Oromo on the island and on the shore. On Tullu Guddo Island, Oromo was largely ignored during the discussion of language attitudes, with only three individual interviewees listing it as their third choice. Likewise, there was no mention of Zay among the Boch’eessa interviewees during the same discussion. The islanders’ Zay identity still appears to be quite strong, but the Boch’eessa residents may be identifying with Amharic in order to distinguish themselves from their Oromo-speaking neighbors. Only two individual interviewees in Boch’eessa said that they would like their children to learn Oromo. This does not necessarily indicate a negative attitude toward the language, but rather the practical choice of parents who know their children will grow up speaking Oromo regardless of whether they study it in school. Indeed, early primary education, both on Tullu Guddo Island and in Boch’eessa is now in Oromo, so their children’s acquisition of Oromo would not be a felt need for most parents in either location. This also explains why many parents in both locations 20 mentioned that they want their children to learn Amharic, since it is neither the language of instruction nor used widely in that area and might not be picked up as easily as Oromo. It is particularly informative that only two individual interviewees in Boch’eessa said that they would like their children to learn Zay. As shown by the results of the community ownership and support questionnaire (see section 4.2), the Zay people of the lakeshores are not necessarily against the development of their traditional language, but neither are they very supportive of it. They do not seem to be greatly concerned that their children do not speak Zay. The Boch’eessa community’s closer identification with their Oromo neighbors is shown by their interethnic marriage preferences. Eight of nine individual interviewees there (see table 4) mentioned Oromo among their marriage preferences, but only two mentioned Amhara. The results on Tullu Guddo Island were quite different; only three of ten considered Oromo to be among their preferences, while six listed Amhara. The interviewees’ attitudes about intermarriage appear to change when they are questioned from the standpoint of what they would allow their own children to do. The Boch’eessa interviewees are open to the idea of intermarriage, as all of them stated that they would let their children marry outside the ethnic group. However, their preferences for their children shift away from the Oromo, with only one of nine saying they would accept an Oromo spouse for their child. Surprisingly, five mentioned that they would allow their children to marry Amhara, which also seems to contradict the results shown in table 4. Perhaps the Zay of the lakeshores are experiencing some tension between their desire to accept and be accepted by their Oromo neighbors on one hand and to maintain their distinct identity as Semitic language speakers on the other hand.2 There were also three individual interviewees in Boch’eessa who said it would be acceptable if their children married Christians of any ethnic group, so it may be an issue of religious identity, instead of or in addition to ethnolinguistic identity.3 On Tullu Guddo Island, the interviewees’ attitudes toward intermarriage are much more conservative when it comes to their children. Six of ten would not allow their children to marry outside the ethnic group whereas, previously, only one had stated opposition to intermarriage. 2 3 Both Zay and Amharic are Ethio-Semitic languages; Oromo is Cushitic. The surrounding Oromo speakers are predominantly Muslim. 21 The only other ethnic group mentioned in regard to acceptable marriage partners for their children was Amhara. Considering the previously mentioned points, it seems that the acceptability of literature in any of these three languages would depend on the purpose of that literature, as well as exactly who would be using it. For example, literature meant to inform the Zay community about issues, such as health and agricultural practices, may be most clearly understood in Zay on the islands and in Oromo on the shores. At this point, educational materials might be more appealing to the entire Zay population if they were in Amharic, but this is not likely to be feasible within the current Oromo educational system of the area. Literature on cultural and religious matters would probably be more acceptable in either Zay or Amharic on the islands and in Amharic on the shores. Practically speaking, for most purposes, Oromo would be the best choice for the shores, and Amharic would be the best choice for the islands. This takes into consideration the self-stated positive attitudes toward Oromo publications in Boch’eessa and Amharic publications on Tullu Guddo Island, as well as the Zay language’s slide toward extinction. 5.2 Community ownership and support questionnaire For the purpose of evaluating the scores (see table 5) as a measure of possible support and interest, the range of 15–25 is considered to be low, 25–35 is moderate, 35–45 is good and 45+ is excellent. According to these scores, the interest level of the Tullu Guddo interviewees appears to be higher than that of the Boch'eessa interviewees; this is not surprising, in light of Zay’s greater vitality on the islands. Even the individual interviewees in Boch’eessa, who expressed an interest in developing Zay, said that they “would be happy if Zay is written and preserved.” This is a further indication that they are aware of Zay’s endangerment and probably do not see the language’s development as an urgent personal need. Its use is decreasing drastically on the lakeshores and, as noted in section 5.1, the people there do not seem to be too concerned about their children learning Zay. These results show some level of personal commitment to contribute to the development of the language, but overall they are fairly moderate. Even the Tullu Guddo average (38) is on the low side of “good,” as it is only 8 points higher than the very moderate Boch’eessa average. 22 Although the group and individual interviewees in both locations said they were interested in the development of Zay, it is doubtful that the interest is very strong in either area. With the widespread multilingualism and positive attitude toward literature in languages of wider communication, the motivation toward language development is low. Also, the identity issue may not be a strong motivator. It is possible for the Zay population to set themselves apart as Orthodox Christians from a Semitic background by using Amharic. This should be kept in mind when interpreting the interest in language development shown during the sociolinguistic interviews. 5.3 Mean SRT scores for men and women A significant difference could not be proven between the mean SRT scores either for men and women or for any of the age categories (see appendix B for descriptive statistics and the twosample t-test). A significant difference was found between the speakers tested on the island (24.93) and those on the lakeshores (31.31), which was to be expected, based on the finding that Oromo is replacing Zay in Boch’eessa. Also significant was the difference between those who had attended school (28.19 to 32.71) and those who had not4 (22.08), although the number of years of education made no significant difference. Since Oromo is currently the language of instruction in the Zay area, this result was also expected. Table 8 notes the percentage of people tested in each category who scored 25 or lower on the Oromo SRT. A 25 or lower represents a proficiency level of 2+ or less. 4 It should also be noted that all the participants with no education, except one, belong to the oldest age grouping. 23 Table 8. Percentage of people tested in each age category who scored 25 or lower on the Oromo SRT Population 10–24 25–34 35+ All Male 42.9% 28.6% 27.3% 32.0% Female 0.0% 0.0% 60.0% 28.6% All 23.1% 16.7% 42.9% 30.4% Male 60.0% 50.0% 33.3% 46.7% Female 0.0% 0.0% 71.4% 33.3% All 30.0% 28.6% 53.8% 40.0% Male 0.0% 0.0% 20.0% 10.0% Female 0.0% 0.0% 33.3% 16.7% All 0.0% 0.0% 25.5% 12.5% Male NA NA 50.0% 50.0% Female NA 0.0% 62.5% 55.6% All NA 0.0% 58.3% 53.8% Male 50.0% 0.0% 0.0% 16.7% Female 0.0% 0.0% 50.0% 20.0% All 40.0% 0.0% 28.6 17.6% Island Shore No education Grade 2, 3, 4, or 5 24 Table 8. Percentage of people tested in each age category who scored 25 or lower on the Oromo SRT (continued) 10–24 25–34 35+ All Male 33.3% 33.3% 66.7% 44.4% Female 0.0% 0.0% NA 0.0% All 12.5% 20.0% 66.7% 25.0% Male 60.0% 25.0% 16.7% 33.3% Female 0.0% 0.0 50.0% 9.1% All 27.3% 14.3% 25% 23.1% 0.0% 0.0% 50.0% 16.7% NA 0.0% NA 0.0% 0.0% 0.0% 50.0% 14.3% Population Grade 6 or higher Grade 2, 3, 4, 5, or 6 Grade 7 or higher Male Female All Forty percent of all people tested on the island scored 25 or lower, meaning that those forty percent have an Oromo proficiency of 2+ or less. On the shore, a mere 12.5 percent scored a 25 or below, showing widespread high proficiency. The only category of people who had an overwhelmingly large number of speakers with an Oromo proficiency of 2+ or less was that of older women on the island. 5.4 Mean RTT scores for men and women A significant difference could not be proven between the mean RTT scores either for men and women or for the speakers tested on the island and those on the lakeshores (see appendix B). It is important to note that the standard deviation is quite high for most of the categories, especially for the test subjects with no formal education and those aged 35 and over. A possible explanation is that severable variables that were not considered in this study could have a major effect on the 25 Amharic fluency of test subjects with no formal education. Among these variables are the amount of travel, interaction with outsiders, and even the level of participation in the Orthodox Church. The high standard deviation for the 35+ age category is probably because most of the participants with no education belong to that group. The women’s scores also have a higher standard deviation than the men’s scores, perhaps for a similar reason. Just as schooling seems to act as an equalizer for RTT scores among those with some formal education, travel, and exposure to outsiders is a likely equalizer for the men’s scores. Fewer women have the opportunity to travel, causing a greater difference between the scores of the educated and uneducated women. In general, a significant difference was found between those who had attended school (82.63 to 89.23) and those who had not (73.57), although, at the 6th grade level, there were some low scores that pulled down the average and increased the standard deviation. The mean score for the 10–24 age category was also significantly lower than for the 25–34 year-olds. Both education and exposure to outsiders could have had an effect on these scores. Not only are younger test subjects less likely to have traveled, but they are also less likely to have attended school for an extended amount of time while Amharic was still the language of instruction.5 6. Conclusions and recommendations The optimal language for literature and educational materials is not the same for all Zay areas. The attitude toward Zay was more strongly positive on Tullu Guddo Island, while the attitude toward Oromo was more strongly positive in Boch’eessa. The attitude toward Amharic was found to be uniformly positive on both the island and the lakeshore. As for language proficiency, the Oromo SRT scores were significantly higher in Boch’eessa than they were on Tullu Guddo Island, although there was no significant difference in Amharic scores according to location (for both Oromo and Amharic, education had a significant impact on the scores). Finally, Zay was found to be a vital language on the islands, where the children grow up speaking it although they do attend school in Oromo. Nevertheless, it will soon be extinct on the shores of the lake, where only adults speak it and the mother tongue of all the children is Oromo. When combined with what is known about social interaction patterns, this information points to Zay as the optimal language for the islands and Oromo as the optimal language for the lakeshores. 5 The language of instruction in the Zay area changed from Amharic to Oromo within the decade prior to this study. 26 However, the situation of the islands must be examined more carefully. Zay is endangered because of the high rate of emigration off the islands, the only place where it is still vital. Amharic is the more acceptable second language on the islands, and the people's proficiency is quite high. The Zay also have a reputation for being a very well-educated group, which, as previously noted, has a significant impact on both Amharic and Oromo proficiency. The Zay living on the islands would probably be well served by Amharic literature until most of them have immigrated to the shore or the Oromo educational system causes a shift to Oromo. The islanders' interest in language development was rated as "Good" on a scale of Low Moderate - Good - Excellent (the Zay living on shore had only a moderate interest). The language is quite closely related to Silt'e, so adaptation from existing Silt’e educational materials may be possible, if desired. The Zay people’s relatively high level of education also indicates the likelihood that significant portions of the population would learn to read and write the language, if it were developed. Zay’s case is one of an endangered language that could prove to be a development success story, but only if the level of motivation for a language-development project is indeed high enough to initiate and sustain the effort. 27 Appendices Appendix A. Sociolinguistic questionnaire A.1 Identification of respondent 1. 2. 3. 4. 5. 6. 7. 8. A.2 Multilingualism 9. 10. 11. 12. 12a. 12b. 12c. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. A.3 Name Sex Age Occupation Religion Education Place of birth Place of residence What is your first language? Which other languages do you speak and understand? Do you speak one better than the others? Rank them. Which of these can you read and write? Apart from your own village, where have you lived at least for 1 year of your life? How long have you lived there? What languages did you speak there? Did the people there understand you well? What was the first language your father learned as a child? Which other languages does he speak and understand? Does he speak one better than the others? Rank them. Can he read and write one of these languages? What was the first language your mother learned? Which other languages does she speak and understand? Does she speak one better than the others? Rank them. Can she read and write any of these? Which languages do your parents speak to each other? Which languages do your brothers and sisters speak and understand? Can they read and write any of these? What was the first language your husband/wife learned? Which other languages does he/she speak and understand? Does he/she speak one better than the other? Rank them. Can he/she read and write one of these languages? What is the first language of your children? Which languages do your children speak and understand? Do they speak one better than the other? Rank them. Can they read and write one of these languages? What language do children in this village learn first? Do many children learn another language before they start school? Which? Do young people in your village speak their mother tongue (MT) well, the way it ought to be spoken? Language use 31. 32. Which language do you speak most often…with your father? With your mother? 28 A.3 Language use (continued) 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. A.4 Attitudes to languages 49. 50. 51. 52. 53. 54. 55. A.5 Which villages speak MT exactly like you? (List them.) Which villages speak your language differently- but you can still understand them? Which speak it so differently that you don’t understand? Which is the best village for an outsider to live in to learn your language? Are there MT people who speak it poorly? Where do they live? Social interaction patterns 61. 62. 63. 64. A.7 Is it good to allow a young man or woman (MT speaker) to marry a woman or man who is not a (MT) speaker? Does this happen very often? Which language is best for a teacher to use in school? Why? Which languages should be taught in school? If a young person speaks (L2 / trade language) at home, would an old person be unhappy about it? What is the most useful language to know around here? Is it OK for your child to marry a non-MT speaking person? Attitudes to dialects 56. 57. 58. 59. 60. A.6 With your brothers and sisters? With your husband/wife? With your children? With your friends? In your village? At the local market? With the elders of your village? In the fields / at work? At the big market? At the clinic? In church / mosque / traditional religious ceremonies? With the administrators of the district? When you are dreaming? When you are praying at home? When you are angry? When you are counting money or things? Which villages do most of your wives come from? Which villages invite you for feasts and dances? Which villages do you trade with? Which language(s) is/are used for communication when you go to the villages mentioned above? Language vitality 65. 66. 67. Do you think that your people are in the process of changing? Do they adopt the customs of (an)other group(s)? Do you know any MT people who do not speak MT anymore? Are there very many? Where do they live? Do you think that young MT people speak MT less and less? 29 A.7 Language vitality (continued) 68. A.8 When the children of this village grow up and have children of their own, do you think those children will speak your language? Development of the language 69. 70. 70a. 71. 72. 73. 74. 75. 76. 77. Which language do you think would be best to choose for making books and newspapers? Why? Do you think it would be good to have something published in your language? What would you like most? Would it be good to have other written MT materials (books, magazines, or newspapers)? If there were schools to teach you how to read and write in your language, would you come to them? Would you like your children to learn to read and write the mother tongue? If there were books in your language, would you be willing to pay for them? Say 2 Birr? [About the equivalent of a quarter (US$ 0.25); insert an appropriate amount in local currency here.] Have you ever seen anything written in your language? What? Have you ever tried to write in your language? Is there a program on radio in your language? Do you listen to it? Would you like to hear your language on the radio? 30 Appendix B. B.1. Descriptive statistics - SRT B.1.1 Population Variable Age Edu SRT N 46 46 46 Mean 33.61 4.043 27.15 Median 33.00 5.000 28.50 TrMean 32.81 3.905 27.60 Variable Age Edu SRT Minimum 10.00 0.000 2.00 Maximum 75.00 12.000 43.00 Q1 18.50 0.000 23.25 Q3 44.25 6.000 35.00 Variable Age Edu SRT N 25 25 25 Mean 35.76 4.920 26.04 Median 34.00 5.000 28.00 TrMean 35.13 4.826 26.30 Variable Age Edu SRT Minimum 11.00 0.000 3.00 Maximum 75.00 12.000 43.00 Q1 19.50 3.000 24.00 Q3 44.50 6.500 32.00 StDev 16.83 3.245 10.20 SEMean 2.48 0.478 1.50 StDev 19.16 3.239 10.76 SEMean 3.83 0.648 2.15 StDev 13.56 3.000 9.58 SEMean 2.96 0.655 2.09 B.1.2 Male B.1.3 Female Variable Age Edu SRT N 21 21 21 Mean 31.05 3.000 28.48 Median 32.00 3.000 31.00 TrMean 30.89 2.842 29.32 Variable Age Edu SRT Minimum 10.00 0.000 2.00 Maximum 55.00 9.000 39.00 Q1 15.50 0.000 20.00 Q3 42.50 6.000 36.00 31 B.1.4 Ages 10–24 Variable Age Edu SRT N 13 13 13 Mean 14.462 5.615 29.15 Median 14.000 6.000 31.00 TrMean 14.364 5.455 30.36 Variable Age Edu SRT Minimum 10.000 3.000 3.00 Maximum 20.000 10.000 42.00 Q1 12.500 4.000 25.50 Q3 16.500 6.000 36.50 StDev 2.933 1.981 11.25 SEMean 0.813 0.549 3.12 StDev 3.288 3.380 8.64 SEMean 0.949 0.976 2.49 B.1.5 Ages 25–34 Variable Age Edu SRT N 12 12 12 Mean 28.417 5.833 29.92 Median 27.000 5.500 29.50 TrMean 28.200 5.800 30.70 Variable Age Edu SRT Minimum 25.000 0.000 9.00 Maximum 34.000 12.000 43.00 Q1 26.000 3.500 26.25 Q3 31.500 8.500 36.00 B.1.6 Ages 35+ Variable Age Edu SRT N 21 21 21 Mean 48.43 2.048 24.33 Median 45.00 0.000 26.00 TrMean 47.74 1.789 24.74 Variable Age Edu SRT Minimum 35.00 0.000 2.00 Maximum 75.00 9.000 39.00 Q1 40.00 0.000 18.50 Q3 53.50 4.500 33.00 N 30 30 30 Mean 30.83 4.233 24.93 Median 30.00 5.000 27.00 TrMean 30.04 4.038 25.58 StDev 11.86 2.711 10.08 SEMean 2.59 0.592 2.20 B.1.7 Island Variable Age Edu SRT StDev 15.10 3.159 10.43 SEMean 2.76 0.577 1.90 32 B.1.7 Island (continued) Variable Age Edu SRT Minimum 10.00 0.000 2.00 Maximum 75.00 12.000 39.00 Q1 15.75 1.500 18.75 Q3 42.50 6.000 31.75 Variable Age Edu SRT N 16 16 16 Mean 38.81 3.687 31.31 Median 34.50 4.000 34.00 TrMean 38.29 3.500 32.07 Variable Age Edu SRT Minimum 13.00 0.000 9.00 Maximum 72.00 10.000 43.00 Q1 25.25 0.000 26.00 Q3 54.25 5.750 36.75 B.1.8 Shore StDev 19.08 3.478 8.58 SEMean 4.77 0.869 2.14 StDev 14.40 0.00000 9.36 SEMean 3.99 0.00000 2.60 StDev 12.80 1.088 8.66 SEMean 3.10 0.264 2.10 B.1.9 No formal education Variable Age Edu SRT N 13 13 13 Mean 51.38 0.00000 22.08 Median 45.00 0.00000 25.00 TrMean 51.18 0.00000 22.73 Variable Age Edu SRT Minimum 30.00 0.00000 2.00 Maximum 75.00 0.00000 35.00 Q1 40.00 0.00000 17.50 Q3 67.00 0.00000 27.00 B.1.10 Attended school until grade 2, 3, 4, or 5 Variable Age Edu SRT N 17 17 17 Mean 29.47 3.941 29.53 Median 28.00 4.000 30.00 TrMean 29.27 4.000 30.67 Variable Age Edu SRT Minimum 10.00 2.000 3.00 Maximum 52.00 5.000 39.00 Q1 17.50 3.000 27.00 Q3 39.50 5.000 35.50 33 B.1.11 Attended grade 6 or higher Variable Age Edu SRT N 16 16 16 Mean 23.56 7.438 28.75 Variable Age Edu SRT Minimum 13.00 6.000 7.00 Maximum 44.00 12.000 43.00 Median 22.00 6.000 31.00 Q1 14.00 6.000 24.50 TrMean 22.86 7.214 29.29 StDev 10.38 1.965 11.42 SEMean 2.59 0.491 2.86 StDev 12.35 1.325 10.37 SEMean 2.42 0.260 2.03 StDev 10.85 1.604 7.74 SEMean 4.10 0.606 2.93 Q3 32.25 9.000 36.75 B.1.12 Attended school until grade 2, 3, 4, 5, or 6 Variable Age Edu SRT N 26 26 26 Mean 26.35 4.654 28.19 Median 26.00 5.000 30.50 Variable Age Edu SRT Minimum 10.00 2.000 3.00 Maximum 52.00 6.000 39.00 Q1 14.00 3.000 25.50 TrMean 25.96 4.708 28.79 Q3 38.25 6.000 36.00 B.1.13 Attended grade 7 or higher Variable Age Edu SRT N 7 7 7 Mean 27.57 9.286 32.71 Median 27.00 9.000 31.00 TrMean 27.57 9.286 32.71 Variable Age Edu SRT Minimum 13.00 7.000 24.00 Maximum 44.00 12.000 43.00 Q1 16.00 8.000 26.00 Q3 34.00 10.000 42.00 B.2 Two sample T-test and confidence interval B.2.1 Two sample T for male vs female Male Female N 25 21 Mean 26.0 28.48 StDev 10.8 9.58 SEMean 2.2 2.1 95% CI for mu male - mu female: ( -8.5, 3.6) T-test mu male = mu female (vs <): T = -0.81 P = 0.21 DF = 43 34 B.2.2 Two sample T for ages 10-24 vs ages 25-34 Ages 10Ages 25- N 13 12 Mean 29.2 29.92 StDev 11.2 8.64 SEMean 3.1 2.5 95% CI for mu ages 10- - mu ages 25-: ( -9.0, 7.5) T-test mu ages 10- = mu ages 25- (vs <): T = -0.19 P = 0.43 DF = 22 B.2.3 Two sample T for ages 35+ vs ages 10-24 Ages 35+ Ages 10- N 21 13 Mean 24.3 29.2 StDev 10.1 11.2 SEMean 2.2 3.1 95% CI for mu ages 35+ - mu ages 10-: ( -12.7, 3.1) T-test mu ages 35+ = mu Ages 10- (vs <): T = -1.26 P = 0.11 DF = 23 B.2.4 Two sample T for ages 35+ vs ages 25-34 Ages 35+ Ages 25- N 21 12 Mean 24.3 29.92 StDev 10.1 8.64 SEMean 2.2 2.5 95% CI for mu ages 35+ - mu ages 25-: ( -12.4, 1.3) T-test mu ages 35+ = mu ages 25- (vs <): T = -1.68 P = 0.053 DF = 26 B.2.5 Two sample T for island vs shore Island Shore N 30 16 Mean 24.9 31.31 StDev 10.4 8.58 SEMean 1.9 2.1 95% CI for mu island - mu shore: ( -12.2, -0.6) T-test mu island = mu shore (vs <): T = -2.22 P = 0.016 B.2.6 Two sample T for no edu vs 2, 3, 4, 5 No edu 2,3,4,5 N 13 17 Mean 22.08 29.53 StDev 9.36 8.66 SEMean 2.6 2.1 DF = 36 35 B.2.6 Two sample T for no edu vs 2, 3, 4, 5 (continued) 95% CI for mu no edu - mu 2, 3, 4, 5: ( -14.3, -0.6) T-test mu no edu = mu 2, 3, 4, 5 (vs <): T = -2.23 P = 0.018 DF = 24 B.2.7 Two sample T for no edu vs 6+ No edu 6+ N 13 16 Mean 22.08 28.8 StDev 9.36 11.4 SEMean 2.6 2.9 95% CI for mu no edu - mu 6+: ( -14.6, 1.3) T-test mu no edu = mu 6+ (vs <): T = -1.73 P = 0.048 DF = 26 B.2.8 Two sample T for 6+ vs 2, 3, 4, 5 6+ 2,3,4,5 N 16 17 Mean 28.8 29.53 StDev 11.4 8.66 SEMean 2.9 2.1 95% CI for mu 6+ - mu 2, 3, 4, 5: ( -8.1, 6.5) T-test mu 6+ = mu 2, 3, 4, 5 (vs <): T = -0.22 P = 0.41 DF = 27 B.2.9 Two sample T for no edu vs 2, 3, 4, 5, 6 No edu 2,3,4,5, N 13 26 Mean 22.08 28.2 StDev 9.36 10.4 SEMean 2.6 2.0 95% CI for mu no edu - mu 2, 3, 4, 5,: ( -12.9, 0.7) T-test mu no edu = mu 2, 3, 4, 5, (vs <): T = -1.85 P = 0.038 DF = 26 B.2.10 Two sample T for no edu vs 7+ No edu 7+ N 13 7 Mean 22.08 32.71 StDev 9.36 7.74 SEMean 2.6 2.9 95% CI for mu no edu - mu 7+: ( -19.0, -2.2) T-test mu no edu = mu 7+ (vs <): T = -2.72 P = 0.0083 DF = 14 36 B.2.11 Two sample T for 2, 3, 4, 5, 6 vs 7+ 2,3,4,5, 7+ N 26 7 Mean 28.2 32.71 StDev 10.4 7.74 SEMean 2.0 2.9 95% CI for mu 2, 3, 4, 5, - mu 7+: ( -12.3, 3.2) T-Test mu 2, 3, 4, 5, = mu 7+ (vs <): T = -1.27 P = 0.11 DF = 12 B.3 Descriptive statistics - RTT B.3.1 Population Variable Age Edu RTT N 46 46 46 Mean 32.26 4.543 81.74 Median 29.00 5.000 85.00 TrMean 31.40 4.452 82.62 Variable Age Edu RTT Minimum 10.00 0.000 40.00 Maximum 75.00 12.000 100.00 Q1 20.00 0.000 70.00 Q3 40.50 7.000 92.50 Variable Age Edu RTT N 26 26 26 Mean 34.69 5.154 82.69 Median 30.00 5.500 80.00 TrMean 33.87 5.083 83.33 Variable Age Edu RTT Minimum 14.00 0.000 50.00 Maximum 75.00 12.000 100.00 Q1 22.00 2.250 80.00 Q3 42.75 8.250 100.00 Mean 29.10 3.750 80.50 Median 27.50 4.500 90.00 TrMean 28.72 3.667 81.67 StDev 15.41 3.607 16.37 SEMean 2.27 0.532 2.41 StDev 16.47 3.619 15.89 SEMean 3.23 0.710 3.12 StDev 13.68 3.522 17.31 SEMean 3.06 0.788 3.87 B.3.2 Male B.3.3 Female Variable Age Edu RTT N 20 20 20 37 B.3.3 Female (continued) Variable Age Edu RTT Minimum 10.00 0.000 40.00 Maximum 55.00 9.000 100.00 Q1 17.00 0.000 62.50 Q3 40.00 6.000 90.00 B.3.4 Ages 10–24 Variable Age Edu RTT N 18 18 18 Mean 18.33 6.111 78.89 Median 18.50 6.000 80.00 TrMean 18.50 6.250 79.37 Variable Age Edu RTT Minimum 10.00 0.000 50.00 Maximum 24.00 10.000 100.00 Q1 14.75 4.750 67.50 Q3 22.25 8.250 92.50 StDev 4.35 2.447 16.41 SEMean 1.03 0.577 3.87 StDev 3.110 3.90 12.50 SEMean 0.938 1.18 3.77 StDev 11.26 3.276 18.19 SEMean 2.73 0.795 4.41 B.3.5 Ages 25–34 Variable Age Edu RTT N 11 11 11 Mean 29.455 5.73 88.18 Median 30.000 6.00 90.00 TrMean 29.444 5.67 90.00 Variable Age Edu RTT Minimum 25.000 0.00 60.00 Maximum 34.000 12.00 100.00 Q1 27.000 2.00 90.00 Q3 32.000 9.00 100.00 B.3.6 Age 35+ Variable Age Edu RTT N 17 17 17 Mean 48.82 2.118 80.59 Median 45.00 0.000 80.00 TrMean 47.80 1.800 82.00 Variable Age Edu RTT Minimum 38.00 0.000 40.00 Maximum 75.00 9.000 100.00 Q1 40.00 0.000 70.00 Q3 55.00 4.500 95.00 38 B.3.7 Island Variable Age Edu RTT N 32 32 32 Mean 30.09 4.625 80.62 Median 27.50 5.000 85.00 TrMean 29.64 4.536 81.79 Variable Age Edu RTT Minimum 10.00 0.000 40.00 Maximum 55.00 12.000 100.00 Q1 18.25 0.500 70.00 Variable Age Edu RTT N 14 14 14 Mean 37.21 4.36 84.29 Median 30.00 5.50 85.00 TrMean 36.08 4.25 85.83 Variable Age Edu RTT Minimum 13.00 0.00 50.00 Maximum 75.00 10.00 100.00 Q1 24.00 0.00 77.50 Q3 51.25 9.00 100.00 StDev 13.09 3.386 16.64 SEMean 2.31 0.599 2.94 StDev 19.38 4.20 16.04 SEMean 5.18 1.12 4.29 StDev 15.65 0.00000 18.23 SEMean 4.18 0.00000 4.87 StDev 12.34 1.054 11.74 SEMean 3.90 0.333 3.71 Q3 40.00 7.000 90.00 B.3.8 Shore B.3.9 No formal education Variable Age Edu RTT N 14 14 14 Mean 47.07 0.00000 73.57 Median 45.00 0.00000 80.00 TrMean 47.00 0.00000 74.17 Variable Age Edu RTT Minimum 20.00 0.00000 40.00 Maximum 75.00 0.00000 100.00 Q1 37.50 0.00000 57.50 Q3 58.00 0.00000 90.00 B.3.10 Attended school until grade 2, 3, 4, or 5 Variable Age Edu RTT N 10 10 10 Mean 27.70 4.000 86.00 Median 30.00 4.000 90.00 TrMean 27.75 4.125 87.50 39 B.3.10 Attended school until grade 2, 3, 4, or 5 (continued) Variable Age Edu RTT Minimum 10.00 2.000 60.00 Maximum 45.00 5.000 100.00 Q1 14.75 3.000 80.00 Q3 40.00 5.000 92.50 B.3.11 Attended grade 6 or higher Variable Age Edu RTT N 22 22 22 Mean 24.91 7.682 85.00 Median 24.00 7.000 90.00 TrMean 24.45 7.550 86.00 Variable Age Edu RTT Minimum 13.00 6.000 50.00 Maximum 46.00 12.000 100.00 Q1 17.75 6.000 70.00 Q3 29.00 9.000 100.00 StDev 8.91 1.783 15.66 SEMean 1.90 0.380 3.34 StDev 11.38 1.268 14.85 SEMean 2.61 0.291 3.41 StDev 7.65 1.405 13.20 SEMean 2.12 0.390 3.66 B.3.12 Attended school until grade 2, 3, 4, 5, or 6 Variable Age Edu RTT N 19 19 19 Mean 24.68 4.947 82.63 Median 20.00 5.000 90.00 TrMean 24.35 5.059 83.53 Variable Age Edu RTT Minimum 10.00 2.000 50.00 Maximum 45.00 6.000 100.00 Q1 15.00 4.000 70.00 Q3 33.00 6.000 90.00 B.3.13 Attended grade 7 or higher Variable Age Edu RTT N 13 13 13 Mean 27.38 8.846 89.23 Median 25.00 9.000 90.00 TrMean 26.55 8.727 90.91 Variable Age Edu RTT Minimum 18.00 7.000 60.00 Maximum 46.00 12.000 100.00 Q1 22.50 7.500 80.00 Q3 31.00 9.500 100.00 40 B.4 Two sample T-test and confidence interval B.4.1 Two sample T for female vs male Female Male N 20 26 Mean 80.5 82.7 StDev 17.3 15.9 SEMean 3.9 3.1 95% CI for mu female - mu male: ( -12.2, 7.9) T-test mu female = mu male (vs <): T = -0.44 P = 0.33 DF = 39 B.4.2 Two sample T for island vs shore Island Shore N 32 14 Mean 80.6 84.3 StDev 16.6 16.0 SEMean 2.9 4.3 95% CI for mu island - mu shore: ( -14.4, 7.0) T-test mu island = mu shore (vs <): T = -0.70 P = 0.24 DF = 25 B.4.3 Two sample T for 10-24 vs 25-34 10-24 25-34 N 18 11 Mean 78.9 88.2 StDev 16.4 12.5 SEMean 3.9 3.8 95% CI for mu 10-24 - mu 25-34: ( -20.4, 1.8) T-test mu 10-24 = mu 25-34 (vs <): T = -1.72 P = 0.049 DF = 25 B.4.4 Two sample T for 10-24 vs 35+ 10-24 35+ N 18 17 Mean 78.9 80.6 StDev 16.4 18.2 SEMean 3.9 4.4 95% CI for mu 10-24 - mu 35+: ( -13.7, 10.3) T-test mu 10-24 = mu 35+ (vs <): T = -0.29 P = 0.39 DF = 32 41 B.4.5 Two sample T for 35+ vs 25-34 35+ 25-34 N 17 11 Mean 80.6 88.2 StDev 18.2 12.5 SEMean 4.4 3.8 95% CI for mu 35+ - mu 25-34: ( -19.5, 4.4) T-test mu 35+ = mu 25-34 (vs <): T = -1.31 P = 0.10 DF = 25 B.4.6 Two sample T for no edu vs 2-6 No edu 2-6 N 14 19 Mean 73.6 82.6 StDev 18.2 14.8 SEMean 4.9 3.4 95% CI for mu no edu - mu 2-6: ( -21.3, 3.2) T-test mu no edu = mu 2-6 (vs <): T = -1.52 P = 0.070 DF = 24 B.4.7 Two sample T for no edu vs 7+ No edu 7+ N 14 13 Mean 73.6 89.2 StDev 18.2 13.2 SEMean 4.9 3.7 95% CI for mu no edu - mu 7+: ( -28.3, -3.0) T-test mu no edu = mu 7+ (vs <): T = -2.57 P = 0.0086 DF = 23 B.4.8 Two sample T for 2-6 vs 7+ 2-6 7+ N 19 13 Mean 82.6 89.2 StDev 14.8 13.2 SEMean 3.4 3.7 95% CI for mu 2-6 - mu 7+: ( -16.9, 3.7) T-test mu 2-6 = mu 7+ (vs <): T = -1.32 P = 0.099 DF = 27 B.4.9 Two sample T for no edu vs 2-5 No edu 2-5 N 14 10 Mean 73.6 86.0 StDev 18.2 11.7 SEMean 4.9 3.7 95% CI for mu no edu - mu 2-5: ( -25.2, 0.3) T-test mu no edu = mu 2-5 (vs <): T = -2.03 P = 0.028 DF = 21 42 B.4.10 Two sample T for no edu vs 6+ No edu 6+ N 14 22 Mean 73.6 85.0 StDev 18.2 15.7 SEMean 4.9 3.3 95% CI for mu no edu - mu 6+: ( -23.6, 0.8) T-test mu no edu = mu 6+ (vs <): T = -1.93 P = 0.032 DF = 24 B.4.11 Two sample T for 2-5 vs 6+ 2-5 6+ N 10 22 Mean 86.0 85.0 StDev 11.7 15.7 SEMean 3.7 3.3 95% CI for mu 2-5 - mu 6+: ( -9.3, 11.3) T-test mu 2-5 = mu 6+ (vs <): T = 0.20 P = 0.58 DF = 23 43 References Bender, Marvin L. 1971. The languages of Ethiopia: a new lexicostatistic classification and some problems of diffusion. Anthropological Linguistics 13 (5): 165–288. Benn, Keith. 2004. Evaluating community support: an attempt at objectivity. Unpublished ms. Casad, Eugene H. 1974. Dialect intelligibility testing. Norman, Oklahoma: Summer Institute of Linguistics of the University of Oklahoma. Gardner, Simon and Ralph Siebert. 2002. Sociolinguistic survey report of the Zay language area. SIL Electronic Survey Reports 2002–024. Dallas: SIL International. Online. URL: http://www.sil.org/silesr/2002/024/SILESR2002-024.pdf. Accessed on 23 January 2008. Gordon, Raymond G., Jr. (ed.). 2005. Ethnologue: Languages of the World, Fifteenth Edition. Dallas, TX: SIL International. Online version: http://www.ethnologue.com/. Leslau, Wolf. 1999. Ziway Etiopic documents: grammar & dictionary. Radloff, Carla F. 1991. Sentence repetition testing for studies in community bilingualism. Dallas: The Summer Institute of Linguistics and the University of Texas at Arlington. Wedekind, Klaus and Charlotte Wedekind. 2002. Initial SLLE survey of the Ziway area. SIL Electronic Survey Reports 2002-063. Dallas: SIL International. Online. URL: http://www.sil.org/silesr/2002/SILESR2002-063.pdf. Accessed on 23 January 2008. Woldemariam, Endeshaw. 2005. Migration and language shift: the case of Zay. International Conference on Endangered Languages of Ethiopia. Addis Ababa, 27–30 April 2005.