Findings from the sociolinguistic survey of the Lolo people
===
1 Introduction
1.1 Motivation for research
Late in 1998, we were asked by our supervisors in SIL–Mozambique to consider a work assignment among the Lolo of Zambezia Province, Mozambique. According to their latest official research (Shrum 1998), this language group had the most obvious need in Mozambique for language development by SIL.
However, further research—which we will specify and develop in this document—gave evidence contrary to this statement of need. As a result, we were asked to take a closer look at the linguistic
Figure 1
situation of the Lolo. This paper will demonstrate from field findings the necessity for a separate
language development program for the speakers of Lolo.
Niger-Congo
Atlantic-Congo
1.2 The Lolo
Volta-Congo
The Lolo are a people group of Western Zambezia Province, living
chiefly in the District of Morrumbala, and concentrated in the Administrative Post of Derre. Their population was estimated at 92,400 during the 1997 census (INE
(New) Benue-Congo
Bantoid
2001). The speech variety is classified as Central Narrow Bantu (Grimes 2000), with linguistic lineage as indicated in Figure 1 * , in Guthrie’s zone P (SIL 2000). Its
Southern
linguistic neighbors include Chuwabu speakers to the south, Sena and Valade
speakers to the Southwest, Nyanja speakers to the West and Northwest (near the Malawi Border), Takwane speakers to the East, and Maoni, Marenje, and Kokola
Narrow Bantu
speakers to the Northeast. (See Map Set 1 in appendix A).
Other (non-Northwest)
The Lolo are largely horticulturists, relying almost totally on their machambas (gardens) for sustenance, while also raising chickens, pigeons, goats,
Central
Lolo
pigs, and occasionally guinea pigs for consumption. Some will augment these
staples by hunting small game. Nearly all large game was exhausted during the
East
War for Independence and the Civil Conflict. Many families grow some cotton for
N or thw e st
a cash crop, which they sell to Agrimo, a state run cooperative.
N o n-N a rro w Ba n tu
N o rth er n
Other sources of income include selling excess grain to markets, usually
in Malawi (to the north). For some, the Mozambican markets of Morrumbala,
Oko
Mocuba, Alto Benfica, and Quelimane are more accessible. When asked, those
U ka an /Akp es
who felt they could choose market venue said they chose to sell in Malawi, since
D efo id
Malawi has a record of better prices.
Edo id
N up oid
1.3 The linguistic context
Ido mo id
Igb o id
With the commercial importance of Malawi comes exposure to its official
C ross R ive r
African language, Nyanja (also called Chichewa). Many reported ability to use
Kainji
Nyanja effectively in the markets, and that the Chichewa Bible is commonly
Pla to id
(though not universally) used in the churches.
Kru
As to other linguistic contexts, many gain exposure to Sena through trips
(N e w ) Kw a
to the district capitol, Morrumbala. Many vendors in the local markets are
N o rth Vo lta -C o ng o
nonresidents from Chuwabu-speaking areas like Nicoadala and Quelimane.
D o go n(? )
Toward the edges of Lolo territory, there is casual contact with Takwane speakers,
Atla ntic(? )
as well as reason to travel through Takwane-speaking areas, though there are no
markets of attraction within them. There is no known influence from the other
Ijo id (?)
neighboring speech communities due to markets or worship; these have influence
Kor do fa nian
only through proximity, and possibly through regional schools, for example in Alto Benfica or Mocuba.
M an de
a composite of charts by Williamson and Watters, in Bendor-Samuel (1989)
1.4 Area description and route followed
The majority of Lolo territory is low-lying hills, covered with moderate to dense woodlands. Water availability does not seem to be a problem, as the sizeable Luala River flows right through Derre Center. It and the Tambisa River seem to have many tributary creeks that drain most of the Lolo area.
The roads are all dirt, with the larger ones being of solid clay/sand construction. All these were in very good condition, with decent bridges everywhere except for two under construction, just east of Derre Center. The lesser roads we traveled were also passable, but could be problematic during the rainy season. One of our hosts said the bridge just before his property (in Dula) would be underwater by mid-season, and impassable until May. During our travels, we had no troubles with mobility. We hardly ever engaged four-wheel drive and never needed to lock the differential. Bicycles were brought on the first trip, and proved quite useful.
Two trips into the Lolo speech community were necessary. The first trip (21–29 July 1999) we gathered word lists, tested intelligibility with Chuwabu, and researched sociolinguistic opinions of the populace. When we later analyzed this data we discovered the need for intelligibility testing with Takwane, and so we planned a second trip (3–6 January 2000) specifically for this purpose. Both trips used essentially the same route. (See Map Set 2 in appendix A.)
A leading objective was to extend testing for inherent intelligibility. Therefore, a chief goal for our first trip was to locate populations most isolated from other language contexts, especially the Chuwabu context. A basic piece of history shaped this path. During Mozambique’s Civil Conflict, Renamo forces passed through Morrumbala District. They did this in such a way that many of the Lolo to the south of Derre Center fled to Chuwabu-speaking areas near Quelimane, while most of the Lolo to the north stayed put. If we wanted to test for inherent intelligibility, we would need to avoid the southern portion of the Lolo community, where there was great potential for learned intelligibility and/or bilingual overlay. This could easily cloud the results. We settled on Dula and Nyanzaza—two sites to the north—and Derre Center itself, as the population center. (See Map 2a in appendix A.)
Our second route was planned along similar lines—to find communities with little potential contact with Takwane speakers. Dula was again a prime example. We started there, the most remote location, and worked our way back to Mocuba. (See Map 2b in appendix A.)
Mocuba was the base of operations for our surveys—Lolo and otherwise—though well outside of the Lolo 1 speech community. At the start of our surveys, we traveled west to Alto Benfica (shown on some maps as Macatanja), then turned south to Derre Center. We continued on the main road past Derre Center for 3.5 kilometers, then turned northwest on a secondary road. This secondary road traveled north for approximately 20 km to Muaherua then about 5 km more to Dula, our most remote location. After gathering data there, we backtracked to Derre Center and, on the first trip, passed through and turned north 0.5 km east of there, continuing 20–25 km to our final area of Terla and Nyanzaza.
On our second trip we collected data in Dula as well, again backtracking through Derre Center afterward. This time, however, we continued on the main road toward Alto Benfica, stopping at Nacuela and Matecula, just west of the Manguze River. Our guide said this river was the accepted demarcation for the end of Lolo territory, as well as the border between Morrumbala and Mocuba districts. (See Map 2b in appendix A.) We then returned to Mocuba.
1.5 Significant materials by other writers
One paper of relevance to our subject is the survey report by Jeffery and Margaret Shrum, Western Zambezia Language Survey in Mozambique (1998). The rapid assessment given by the Shrum survey gives us quality information about the Lolo speech variety and the speech varieties that surround this group.
Their work addressed five undeveloped speech varieties in this area—Marenje, Kokola, Valade, Lolo, and Maoni. The following items gave us strong suggestion that Lolo was the most viable variety for SIL development:
• The Marenje community (approx. 75,000 persons) is much smaller than the Lolo community (approx. 200,000 persons). Additionally, several of those the Shrums interviewed had difficulty deciding whether they were Marenje or Kokola, showing an eroding sense of cultural identity (section 3.4.3). It was the Shrums’ opinion
1 Although some maps show a secondary road that travels directly west of Mocuba for Derre, it is not passable by car or truck, even 4x4.
that “pressure from the larger and more influential groups in the area may make Marenje nonviable [for their own language development program]” (section 5.3.1).
• The Shrums state that the Kokola community is also much smaller than the Lolo community (50,000–75,000
Mozambican Kokolani). Many of those interviewed preferred to speak other languages over Kokola. Nyanja was highly valued by those interviewed. Moreover, Nyanja is undoubtedly highly used by the Malawian Kokolani (75,000 persons), as it is an official language-of-wider-communication there. (section 3.4.4)
• The Maoni community was described by several sources as a variant of Marenje. All five of the Shrums’ Maoni
participants identified their group as being related to Marenje. (sections 3.2.2, 3.5) • Valade was elusive to the Shrums; they found no one at first who claimed to be Valade. In their observation log
(1997), they record meeting a man 12km from Mepinha on the way to Quelimane who at first claimed to be Lolo. When the Shrums mentioned Valade, he admitted that he was Valade. In addition, several Lolo informants on the author’s survey trips stated that the Lolo word list the Shrums’ collected is actually Valade. This evidence seems to suggest that Valade speakers view their language as a variant of Lolo.
• The Lolo view and esteem their language highly. Nearly all Lolo speakers prefer to use it over all other languages, and want their children to learn it first. (1998:section 3.4.5)
Additionally, the Shrums’ report gave evidence through statistical means that Lolo is lexically distinct from neighboring languages that were being actively developed (Sena, Chuwabu, Lomwe, and Nyanja). Because of these distinctions, the Shrums recommended Lolo for development.
Another paper of relevance is “The Chuwabu Language Cluster” (1999) by James and Virginia Vinton. This report on the Vintons’ survey work in southern Zambezia relevantly drew into question this recommendation. Their survey of the Chuwabu language and its variants had as one focus intelligibility testing, which was beyond the scope of the Shrum survey. In one area of the Vintons’ survey, just south and east of Derre Center, Lolo speakers 2 showed a remarkably high understanding of Chuwabu (Vinton 1999:15, 19). An explanation of this apparent inconsistency would be necessary before any action could be taken on Lolo language development.
2 Purpose and aims
2.1 Purpose of the survey
We propose to resolve, through this survey, the differences in the results of the aforementioned surveys. We will extend, through more data over a greater geographic area, the understanding of how distinct Lolo is from its neighbors. We will also tell why these differences in assessment occurred. Ultimately, we will show conclusively that the Lolo speech community is a candidate for the assistance of SIL-Moçambique as a distinct language- development program.
2.2 Aims of the survey
The aims of our research were threefold: • Develop a broader base of word list findings through several lists in diverse areas of the Lolo-speaking
community. • Research through a larger area the language views and cultural situation of the Lolo. • Extend intelligibility research throughout the speech community, focusing on the more linguistically isolated
zones.
2 average RTT score, 93%, standard deviation = 11%
3 Procedures
3.1 Word lists
3.1.1 Reason for use
One way of quickly assessing similarity/dissimilarity between speech varieties is by comparing the words they use for the same ideas. If, after following a specific, rigorous word list analysis, a certain level of (theoretical) shared language history—percent of cognates—is not found, the capacity for inherent intelligibility is very slim. Thus, word list analysis is a first step toward assessing intelligibility between speech varieties.
3.1.2 Design
We used the SIL/NELIMO 200-word list (NELIMO 1990) to obtain data for researching lexical similarity. It consists of 200 words of several grammatical categories. These are grouped first by grammatical category, then by semantic domain, for ease of recognition by the respondents. See appendix A for these lists.
3.1.3 Administration
The word lists were gathered from a panel of three to eight mother tongue Lolo speakers, appointed by the local governing official. He was asked to provide persons who were known as first-language Lolo speakers, and known to speak it well. All list elicitation was administered by the author. Following the procedure indicated by Wimbish (1989:80), suspect entries were checked and verified against word lists already obtained for the neighboring speech varieties, as responses were given. The panel was permitted to discuss the word in question, but in general one spokesman, chosen by the group beforehand, gave the answer of consensus.
3.1.4 Analysis
Analysis was handled largely through the lexico-statistical computer application WORDSURV (v. 2.5.12). Word lists for Chuwabu, Takwane, Nyanja, Marenje, Kokola, Sena, Manyawa, and Lomwe speech varieties were jointly analyzed for cognate statistics. This was done after word-by-word inspection for (a) groupings of apparent
cognates, and (b) the removal of incongruent data 3 from the analysis. The results and their implications are detailed below.
3.2 Recorded Text Testing (RTT)
3.2.1 Reason for use
Shared cognate percentages between word lists of speech varieties have a threshold value for usefulness. Studies of this type use an upper confidence limit of 70% (cited by Bergman 1990:8.1.6). When the value is less than threshold, we can safely term these speech varieties as “mutually unintelligible”. When the value surpasses this threshold, further testing is necessary to determine intelligibility. Recorded Text Testing (hereafter “RTT”) is the next step recommended by our sources (Grimes 1989:4.1). RTT use was indicated by Lolo’s high cognate percentage with Takwane (lowest upper confidence limit = 85.8%), and by the prior RTT results for Chuwabu gathered by the Vintons (1999) (avg. score, 93%, st. dev. = 11%).
3.2.2 Design
The two RTTs utilized on this survey had slightly differing design. These differences are due to different lines of development. We will explain the distinctions in the following two sections.
3 Incongruent data usually occurred because the five or more teams involved in collecting these lists over the years encountered different sub-senses of the concept in question (e.g. some receiving “push along” and others the word for “shove”
for the list word ‘push’). Although this reduced the size of the word samples, better than 170 words for each speech variety remained for analysis, a quantity well above the 100 often utilized for studies of this type.
3.2.2.1 Design of the RTT for intelligibility of Takwane
Takwane RTT design followed the guidelines set forth in Casad (1974), Blair (1990), and Bergman (1990:4.2). Two short texts (1–5 minutes each) are recorded from native speakers, one text of the speech variety under study (Lolo), the other text of the speech variety of comparison (Takwane in this instance). The former (called the control text) establishes the test-taking capacity of the participant, while the latter examines the participant’s understanding of the neighboring speech variety. Each text is listened to twice; once in its entirety, and again with intermittent pauses for context-relevant questions, used for evaluating participant understanding.
To establish the quality of the texts themselves, each is tested on speakers of that particular speech variety (called home testing) before it is used to test a speaker of another speech variety. This was done for both texts, and the questions that native speakers could not answer correctly were eliminated from the exam. In addition, the Takwane text had further quality confirmation in the high scores achieved on other surveys (91% average score during the Manyawa survey (conducted by the author, report forthcoming)).
3.2.2.2 Design of RTT for intelligibility of Chuwabu
As one aim of this survey was to extend the intelligibility research of the Vintons, we decided to use the RTT they developed for their survey. This would make the RTT results of the two surveys comparable. This RTT had no control test in the Lolo speech variety, but rather two Chuwabu texts, the official text for intelligibility evaluation, preceded by a shorter, simpler text to aid in evaluating each participant’s test-taking ability. Although we proceeded without a Lolo control test, we feel confident that post-test interviews brought to light any participants who had problems in test taking. As well, the adequacy of the test design was supported by the score differences between our three test locations. We will bring further light to this in our explanation of RTT results (see section 4.2.1.1).
Standard RTT design has the context-relevant questions recorded directly onto the test recording. In their report, as well as personal conversation, the Vintons detailed difficulties with this methodology (1999:14). Many test participants who understood the recorded text had great difficulty in responding well to the pre-recorded questions. Due to this, they opted to forgo the questions altogether, instead asking for a summary at each pause in the text. In an effort to maintain comparability of results (and avoid the same pitfalls), our testing also forwent the prerecording of the questions. Still, to achieve a comfortable level of objectivity, we decided to utilize questions
read by a mother-tongue speaker, 4 trained for this work. The same mother tongue speaker was used throughout both the Takwane and Chuwabu testing.
3.2.3 Administration
At the start of each test administration, our host (government administrator, local chief, or host compound owner) would begin gathering exam participants. We requested an equal number of young vs. old, and male vs. female. This was not always possible, and our ratios were sometimes 6 to 4. We would administer a brief pre-test questionnaire to each participant. This was to record basic demographics as well as ensuring the participant was from the location and speech community of focus.
The participant would then listen to each text twice, once in its entirety, and once with several interruptions 4 for contextually based questions (given in Lolo by our interpreter). Records were kept of right and wrong responses, as well as any answers that might reveal information useful to the survey. See appendix B.
The test was administered in three different locations for each RTT. Our aim was to collect data from monolingual or near-monolingual situations, as well as geographically diverse locations.
3.2.4 Analysis
Percentages of correct responses were recorded for each test administered. Average scores were calculated, as were statistical standard deviation, for specific demographic profiles, as well as for overall numbers. Each demographic was statistically analyzed to determine if there exists a significant difference between the two groups in question. Demographic profiles include age, sex, residence, outside language exposure, and long-term relocation outside of the Lolo speech community.
4 Pastor Jonas Paive Martins, our indispensable interpreter, guide, and friend, as well as Derre-born technical consultant for Lolo.
3.3 Sociolinguistic questionnaires
Questions fell into two basic categories: background demographics, and language opinion. The first was to offer profiles with which to examine our data, while the second was to give light to native opinions about Lolo in the context of the surrounding speech varieties. See sample questionnaire in appendix C.
The questionnaires were administered to either one or two persons at a time, through an interpreter when Portuguese was not understood. They were customarily completed just prior to RTT administration.
4 Tool results and findings
4.1 Word lists—lexico-statistical findings
Before we compare Lolo’s word list data with the other speech varieties, we would like to point out in background the very high percentages found between the Lolo variants themselves (94, 95, and 96 percent; potential maximum 96.8 percent lexical agreement between all three). 5 This confirms the quality of choices for research
locations, unifies the Lolo lists for comparison with the other speech varieties, and will give us a solid backdrop for the comparison of results from the Recorded Text Tests.
Since lexico-statistics is a tool with limitations, we hesitate to publish such a complete table as table 1. There is often a temptation to make more of the numbers than is reasonable or justified. As we have said above, less than a 70% upper confidence limit indicates that the speech varieties are inherently unintelligible. Greater values tell us the speech varieties could possess noteworthy intelligibility, but need further testing.
That being said, let us note the dividing line in table 1. Those speech varieties below the line share, by the methodology, no practical inherent intelligibility with those above the line. Hence, the development programs existent in Nyanja, Sena, and Lomwe should not be applicable to the other speech varieties (except for the dubious borderline case of Lomwe and Takwane). The values found above the line express the need for further investigation of these speech varieties, with other tools.
Table 1: Lexico-statistical Results—Percentage of Shared Cognates, with upper/lower confidence levels. Lolo of Derre Center 94±2.8
Lolo of Dula 95±2.7
Lolo of Nyanzaza
83±4.5 7 81±4.8 81±4.8 77±6.5 74±6.7 72±5.5 Takwane 70±5.6 7 71±5.6 69±5.7 66±7.2 69±7.1 66±5.8 71±5.5 Chuwabu 56±6.1 7 56±6.1 55±6.1 57±7.6 52±7.6 55±6.1 65±5.8 59±6.0 Lomwe 52±6.1 7 54±6.1 54±6.1 52±7.6 57±7.6 52±6.1 48±6.1 57±6.0 Sena
5 Upper confidence limit of the lowest scoring pair. See table 1. 6 “Valade” is written within quotation marks because of the uncertainty of the name. The word list was gathered by colleagues in a place called Muandiua, and assumed to be Lolo. As our team was driving to our first survey destination, our guide
(a Lolo mother-tongue speaker and native to the region) pointed out that Valade was spoken there. As we gathered my word lists and noted discrepancies, we asked several informants about them. At times they would state, “That’s another language”, but usually, “That’s Valade”. With the similarity shown here between the “Valade” list and the lists this survey gathered, we can take “Valade” to be at least a mixture of Lolo and another speech variety, and very possibly Valade itself.
7 Speech varieties with a Bible or current development program.
Of the speech varieties above the threshold, Valade, Marenje, and Kokola do not have language development programs. Since they had no such resources to offer Lolo development, they were beyond the scope of our research. We did not test Lolo speakers for intelligibility with these three speech varieties.
Only Chuwabu and Takwane have current language development programs. Hence, we must look at other test results to make decisions about the intelligibility of these two speech varieties by the Lolo. See section 4.2.
4.2 RTT results—Intelligibility of Chuwabu and Takwane by Lolo speakers
In general, the most valuable results from RTT studies are the overall averages—51.5% for Chuwabu, 56.67% for Takwane, with high variation in the scores of both (see table 2). Hence, intelligibility levels are well within the range to show them as separate “languages”, and variation indicates that much of the intelligibility is learned. Still, some demographic groups showed details of interest. These will be handled separately for Chuwabu and Takwane in table 2.
Table 2 – Overall
#of Participants
Average
Standard
RTT scores
Score
Deviation
Takwane RTT
Chuwabu RTT
4.2.1 Chuwabu intelligibility evaluation
4.2.1.1 Background and Overall Score Evaluation
We started our testing in Derre Center, our southernmost and most densely populated point of contact. It is also essential to note that these participants largely fled south to Chuwabu speaking territories during the Civil Conflict and were the most broadly traveled of the three groups. With this increased exposure to Chuwabu came the expected high scores for intelligibility with Chuwabu (Derre Center had an average score of 74.2%—at least 35% higher than the other two locations. See table 3)
Table 3 – Chuwabu RTT scores divided by location
Standard Deviation Overall Score
# of Participants
Average Score
36.7% Location A Derre Center
22.0% Location B Dula
36.0% Location C Terla/Nyanzaza
This fits quite well with the Vintons’ findings. They showed very high scores from their Lolo participants, all of whom were tested to the south of Derre Center, and nearly all of whom fled the conflict to Chuwabu speaking areas (Vinton 1999:18).
Although many participants had high scores, the average score was still below the accepted inherent intelligibility threshold (75%). Variation was quite high as well, indicating much of the existent intelligibility is learned.
The high scores of Derre Center in general provided an excellent backdrop to the lower scores in the next two more linguistically isolated areas. Certainly without this high-score backdrop, the extremely low scores in Dula and Nyanzaza would have shed doubt on test design and test administration.
4.2.1.2 Evaluation of Different Demographic Groups
Our most important demographic feature concerned participants’ former residence. As mentioned, many moved to safer places during war-years, including into Chuwabu-speaking zones. The scores of such refugees were significantly higher (43.7% higher, see table 4) than the others. This again contradicts inherent intelligibility of Chuwabu by Lolo mother-tongue speakers.
Table 4 – Chuwabu scores divided by residence patterns
# of Participants Average Standard
Deviation Overall Scores: Never Lived in Chuwabu-speaking zone
Score
13.6% Location A – Derre:
Lived several years in a Chuwabu-speaking zone
Never Lived in Chuwabu-speaking zone
11.79% Location B – Dula:
Lived several years in a Chuwabu-speaking zone
Never Lived in Chuwabu-speaking zone
10 39.2% 35.58% Lived several years in a Chuwabu-speaking zone
N/A Location C – Terla/Nyanzaza:
0 N/A
Never Lived in Chuwabu-speaking zone
Lived several years in a Chuwabu-speaking zone
Data having figures above threshold levels for intelligibility indicated in bold
The reader will also note that older participants scored significantly higher than the younger participants (see table 5). This leads one to believe that increased life experience increases Chuwabu comprehension, and points once again to learned intelligibility. The lower scores of the younger participants allow us to also surmise that the Chuwabo-speaking area is probably not expanding through the current Lolo-speaking area.
Table 5 – Chuwabu scores divided by
#of Participants
Average
Standard
Deviation Overall Scores: Older than 30
age
Score
34.2% Younger than 31
36.15% Location A – Derre:
Older than 30
15.52% Younger than 31
28.3679% Location B – Dula:
39.7% Younger than 31
Older than 30
35.06% Location C – Terla/Nyanzaza:
Older than 30
36.12% Younger than 31
Data having figures above threshold levels for intelligibility indicated in bold
4.2.2 Takwane intelligibility evaluation
On our trip to evaluate Lolo intelligibility of Takwane, we started in the more linguistically isolated zones—Dula, then Muaherua (see Map 2b in appendix A.). If, here, we received low RTT scores, it would be a clear statement on intelligibility—we could then be quite certain the Lolo’s understanding of Takwane is not inherent. On the other hand, if the RTT scores were not low, we would have no such guarantee, and would need to labor further into the demographic and geographic dynamics of these intelligibility levels.
A clear statement was received. The ten participants in this first test area averaged 35.45%, almost forty percent lower than threshold for dialect or bilingual intelligibility. The large amount of variation (st. dev. = 33%) also tells us this level of intelligibility is largely learned, not inherent (see table 6).
Table 6 – Takwane RTT scores
# of
Average
Standard Deviation
divided by location
Participants
Score
Overall Score
29.0% Location 1 Dula
33.0% Location 2 S. Muaherua
25.0% Location 3 Nacuela
This being established, we headed out of “deep Lolo”, towards areas closer to lines of commerce, in the direction of Takwane communities (see Map 2b in appendix A). Still, throughout our travels we were always testing Lolo mother-tongue speakers. As we progressed through Muaherua to Nacuela, the average scores increased (to 64.55% and 70.00%, respectively). Also the variance decreased (to 25% and 13.6, respectively), demonstrating more homogeneity in these groups toward Takwane bilingualism (see table 6).
There is no statistical difference 8 between men and women in regard to these scores. The same is true between older (>30 years) and younger (<31 years) participants. Our only demographic figure that showed a
difference in scores was exposure to other speech variety contexts. This figure was the culmination of several questions. Did the participant travel much? What were the mother tongues of the parents? Did he/she live outside of the Lolo speech community for extended periods? When these showed a broader exposure to other speech variety contexts, we studied their scores under the heading “High Exposure to Other Speech Varieties”. Overall, this division’s average scores fell above the threshold figure for intelligibility of Takwane (78.41%, threshold 75%. See table 7).
Table 7 – Takwane RTT scores divided by
# of Participants
Average
Standard
Deviation Low exposure, all locations
exposure to other speech varieties .
High exposure, all locations
17.5% Low Exposure, Location 1 (Dula)
High Exposure, Location 1 (Dula)
0.0% Low Exposure, Location 2 (S. Muaherua)
High Exposure, Location 2 (S. Muaherua)
10.5% Low Exposure, Location 3 (Nacuela) 9 8 no statistical difference
High Exposure, Location 3 (Nacuela)
Data having figures above threshold levels for intelligibility indicated in bold . We should note that in some areas, persons with such high exposure were not in abundance. Note that in
Dula (Location 1), our test sample included only one person that met such criteria. We had low participation at South Muaherua (Location 2) as well. Although this is just a reflection of the linguistic isolation of the test locations, it still softens this conclusion.
We should also note that the figures from Nacuela (Location 3), our location closest to the Takwane community, did not demonstrate a statistical difference between the low exposure and high exposure groups. We might expect this, due to general proximity and hence high casual exposure to Takwane. Still, the general indication of these statistics (that linguistic exposure augments intelligibility) is not negated by the figures from Location 3; the figures are simply unable to support the indication.
Regardless of these weaknesses, we can be fairly certain our greatest contributor to score variation is exposure to Takwane.
4.3 Results from sociolinguistic questionnaires
Through lexico-statistical analysis and intelligibility testing, we have established from a linguistic point of view how the neighboring language development programs could not easily be applied to the Lolo community. The
9 Statistical validity for comparing these pairs of data was found using the Mann-Whitney U test. Low exposure average score: 71.2%, St. Dev.: 10.6%.
High exposure average score: 68.2%, St. Dev.: 18.9% High exposure average score: 68.2%, St. Dev.: 18.9%
The responses to the sociolinguistic questionnaires give us an overwhelming “yes”. No one in the Lolo speech community thought Lolo development should not happen. There is not even a hint of the sentiment that Sena, Chuwabu, Nyanja, or Takwane are superior speech varieties. Nor was it felt that what is developed by and for these others should be used for the Lolo.
The thirty participants described Lolo as nearly universal in its use at home, in the marketplace, and at church. Only one participant said they used Portuguese as the primary language at home, and three participants (all from Dula) said they used Nyanja in their market contacts. These three also described the use of Nyanja at church, but only with translation into Lolo. Various participants also mentioned Chuwabu, Portuguese, and Lomwe at times, but always as a secondary possibility for use in these contexts.
Seventeen responded to our question as to which language their children should learn first. All indicated Lolo except for one who thought it should be Portuguese.
5 Conclusion and recommendations
Of the six speech varieties bordering the Lolo community, that have or had a language development program, 10 only Chuwabu and Takwane share sufficient cognates to possibly include the Lolo community in their
scope of service. Of these two, Lolo speakers did not demonstrate sufficient intelligibility as to support inclusion in these language programs.
Regarding its comparison to the two other surveys, our work comes to the same basic conclusion found in the Shrums’ report (1998). Still, our results do not differ greatly with the Vintons’ results (1999). Rather, our results add focus to their results concerning Lolo. They found their test location to show a high level of intelligibility with Chuwabu. We found this area to have the highest level of such intelligibility throughout the Lolo community. Our research adds that the mean level of intelligibility does not reach a functional level, and is in fact learned.
Since there are no sociolinguistic opinions against such a course, the data recommends a separate development program for the Lolo speech variety. Further, we recommend additional research on the applicability of
a Lolo development program to the Marenje, Kokola, Valade, and Maoni. The data shows them as more lexically similar to Lolo than any other speech variety. The Lolo program could possibly be these communities’ best road toward language development.
Appendices
10 Portuguese, Lomwe, Sena, Nyanja, Chuwabu, and Takwane
Appendix A. Map set 1
Map 1c: Speech Varieties of District of Morrumbala.
Map 1a: Mozambique, Province of
Differing administrative
Zambezia shaded.
posts shown by shading. Inset indicates area of Map
1d below.
Map 1b: Zambezia, District of Morrumbala shaded.
Map 1d: General research area and neighboring speech varieties.
Map set 2
Map 2a Route 1: Chuwabu RTT
July 1999
Map 2b Route 2: Takwane RTT
January 2000
Appendix B Word list data
This should probably be titled “Morpheme List Data”, as there are few entire words listed here. All affixes, as best as possible, have been removed, so as not to skew the results. Each word asked for is listed in English and Portuguese, and is followed by the responses for the speech variety being investigated. Every response listed is preceded by a letter which indicates cognate grouping; all those responses utiliizing the same letter are judged to be cognates. Preceding this letter is the name of the speech variety.
1 arm
braço
4 blood
sangue
Lolo of Derre Center
a| sa 0giri Lolo of Dula
a| m uono***
Lolo of Derre Center
a| sa 0giri Lolo of Nyanzaza
a| wono***
Lolo of Dula
a| sa 0giri Takwane
a| m uono***
Lolo of Nyanzaza
a| sa 0g'ri Chwabo
a| m ono***
Takwane
a| sa 0g'ri Lolo(Valade?)
a| m ono***
Chwabo
a| sa 0gri Manyawa Munguluni
a|
Lolo(Valade?)
b| mwasi Nyanja/Chichewa
a| m uono***
Manyawa Munguluni
c| g ¡zi Marenji
a| nk nnn***
Nyanja/Chichewa
d| k ame Cocola
a|
Marenji
c| gazi Sena
a| wono***
Cocola
e| ropa Lomwe
a| nk nno***
Sena
a| m ono***
Lomwe
f| pome
a| ni 0gn Lolo of Dula
Lolo of Derre Center
c|
Lolo of Derre Center
a| lu 0gu Lolo of Nyanzaza
d| tana
Lolo of Dula
a| ni 0gn Takwane
d| tana
Lolo of Nyanzaza
a| ni 0go Chwabo
c| duli*****
Takwane
a| ni 0go Lolo(Valade?)
c| Çuni*****
Chwabo
a| nu 0gu Manyawa Munguluni
b| mafudi
Lolo(Valade?)
b| mwi:li Nyanja/Chichewa
c| duli*****
Manyawa Munguluni
d| s ¡n¡
Nyanja/Chichewa
d| t
Marenji
c| t *ut*u Cocola
c| dulinduli
Marenji
c| t *ut*u Sena
c| duli*****
Cocola
Sena a| nu 0gu Lomwe
d| sana
c| Çuli*****
Lolo of Derre Center
a| s is i Lolo of Dula
a| i 0gulu
Lolo of Derre Center
a| s is i Lolo of Nyanzaza
a| i 0gulu
Lolo of Dula
a| s is i Takwane
a| i 0gulu
Lolo of Nyanzaza
a| t it i Chwabo
a| e gulu
Takwane
a| ru gulu
Chwabo
a| t
Lolo(Valade?)
a| s is i Manyawa Munguluni
a| !i0gudu
Lolo(Valade?)
a| t it i Nyanja/Chichewa
a| ru gulu
Manyawa Munguluni
a| t 5*is i Marenji
c| mimb ¡
Nyanja/Chichewa
a| s is i Cocola
a| kulu
Marenji
a| s is i Sena
a| i 0gulu
Cocola
a| s is i Lomwe i a| ru kulu Lomwe b| ma hi
c| mimba
Sena
Lomwe
a| jaru
7 bone
osso
olho Lolo of Derre Center
b| k *uva
11 eye
Lolo of Dula
b| so Lolo of Nyanzaza
b| k *uva
Lolo of Derre Center
b| so Takwane
b| k *uva
Lolo of Dula
b| so Chwabo
b| k uva
Lolo of Nyanzaza
a| nt o Lolo(Valade?)
a| gogo Ço
Takwane
a| nt Manyawa Munguluni
b| k u $ag
Chwabo
a| Ü*o Nyanja/Chichewa
b| k uva
Lolo(Valade?)
a| nt o Marenji
c| fup ¡
Manyawa Munguluni
b| su Cocola
b| k *uva
Nyanja/Chichewa
b| s n! Sena
b| k uva
Marenji
b| s n Lomwe
dedo Lolo of Derre Center
a| b 'le
12 finger
Lolo of Dula
a| buno Lolo of Nyanzaza
a| b 'le
Lolo of Derre Center
a| buno Takwane
a| b 'le
Lolo of Dula
a| buno Chwabo
a| b 'le
Lolo of Nyanzaza
a| buno Lolo(Valade?)
a| bele
Takwane
a| buno Manyawa Munguluni
a| b 'le
Chwabo
a| buno Nyanja/Chichewa
a| b 'le
Lolo(Valade?)
a| buno Marenji
a| ere
Manyawa Munguluni
b| ¡l¡ Cocola
a| b 'le
Nyanja/Chichewa
a| bun n Sena
a| b 'li
Marenji
a| bunu Lomwe
a| 'ri
c| kata
9 buttocks
rabo
unha Lolo of Derre Center
13 fingernail
a| t au
Lolo of Dula
a| jala Lolo of Nyanzaza
a| t au
Lolo of Derre Center
a| jala Takwane
a| t au
Lolo of Dula
a| jala Chwabo
a| t *au
Lolo of Nyanzaza
a| jala Lolo(Valade?)
a| r ao
Takwane
a| jala Manyawa Munguluni
a| Ü au
Chwabo
a| jala Nyanja/Chichewa
a| r ao
Lolo(Valade?)
b| ek warukwa Marenji
==no entry===
Manyawa Munguluni
d| k *¡&¡o Cocola
c| ubu
Nyanja/Chichewa
a| Õal¡ Sena
a| t *au
Marenji
a| jala Lomwe
b| taku
c| ngoli
Lomwe
b| k * aru***
10 ear
orelha
Lolo of Derre Center
a| aru
Lolo of Dula
a| aru
Lolo of Nyanzaza
Lolo(Valade?)
a| ar Â
Manyawa Munguluni
a| aro
Nyanja/Chichewa
d| k *utu
Marenji
a| ar Â
Cocola
a| aru
Sena
d| k utu
Lomwe
a| rima
14 foot
pé
intestinos Lolo of Derre Center
18 intestines
a| jalo
Lolo of Dula
Lolo of Derre Center b| njelu Lolo of Nyanzaza
a| jalo
b| njelu Takwane
a| jalo
Lolo of Dula
Lolo of Nyanzaza b| njelu Chwabo
a| jalu
c| ru bo Lolo(Valade?)
a| Õalo
Takwane
c| ru bo Manyawa Munguluni
a| jalu
Chwabo
b| njelo Nyanja/Chichewa
a| jalu
Lolo(Valade?)
Manyawa Munguluni c| ru bu Marenji
c| p *¡zi
Nyanja/Chichewa c| tumbo Cocola
a| alo
c| rum n Sena
a| jalu
Marenji
c| ru bo Lomwe
a| jalo
Cocola
b| na:ni
Sena
c| tumbu
joelho Lolo of Derre Center
19 knee
a| d a d a
Lolo of Derre Center a| bo Lolo of Nyanzaza
Lolo of Dula
a| d and a
a| bondo Takwane
a| d and a
Lolo of Dula
Lolo of Nyanzaza a| bondo Chwabo
a| d a d a
b| kut a Lolo(Valade?)
a| l a d a
Takwane
a| bo Ço Manyawa Munguluni
a| d and a
Chwabo
a| bondo Nyanja/Chichewa
a| l a d a
Lolo(Valade?)
Manyawa Munguluni b| kut a Marenji
a| dz ¡n <a
Nyanja/Chichewa a| nndn Cocola
b| ma ta
a| bo Sena
a| d and a
Marenji
a| bondo Lomwe
a| zan ja
Cocola
b| na:ta
Sena
a| bondo
cabeça
fígado Lolo of Derre Center
20 liver
b| muru
Lolo of Dula
Lolo of Derre Center b| lop *a Lolo of Nyanzaza
b| muru
b| lop *a Takwane
b| muru
Lolo of Dula
Lolo of Nyanzaza b| lop *a Chwabo
b| muru
a| !aba Lolo(Valade?)
a| so Îo
Takwane
c| fafa Manyawa Munguluni
b| muru
Chwabo
b| lop ¡ Nyanja/Chichewa
b| muru
Lolo(Valade?)
Manyawa Munguluni a| aba Marenji
b| mutu
Nyanja/Chichewa d| windi Cocola
b| muru
b| lop *a Sena
b| muru
Marenji
b| lop *a Lomwe
a| s nlo
Cocola
b| moro
Sena
c| fafa
coração
Lolo of Derre Center
a| rima
Lolo of Dula
a| rima
Lolo of Nyanzaza
a| rima
Takwane
a| rima
Chwabo
a| rim
Lolo(Valade?)
a| ri
Manyawa Munguluni
a| rima
Nyanja/Chichewa
a| tim ¡
Marenji
a| rima
Cocola
a| rima
Sena
a| tima a| tima
25 skin
a| lomu
Lolo of Dula
Lolo of Derre Center a| t *emb' Lolo of Nyanzaza
a| lomu
a| t *emb' Takwane
a| lomo
Lolo of Dula
Lolo of Nyanzaza a| t *emb' Chwabo
a| lomo
e| kup *a Lolo(Valade?)
a| lomu
Takwane
a| Ü e be Manyawa Munguluni
a| lomo
Chwabo
b| k u 0gu Nyanja/Chichewa
b| jano
Lolo(Valade?)
Manyawa Munguluni c| dabara Marenji
c| p ¡k¡mw¡
Nyanja/Chichewa b| k *ungu Cocola
a| lomo
d| kwi Sena
a| lomo
Marenji
a| t 'mbe Lomwe
a| l nmo
Cocola
b| jano
Sena
g| kanda
lágrimas Lolo of Derre Center
b| t 5umbu
26 tears
Lolo of Derre Center a| !odi Lolo of Nyanzaza
Lolo of Dula
b| t 5umbu
a| !odi Takwane
b| t 5ombo
Lolo of Dula
Lolo of Nyanzaza a| sodi Chwabo
a| toku
b| t ori Lolo(Valade?)
a| t
Takwane
b| nt Manyawa Munguluni
b| t 5ombo
Chwabo
a| !odi Nyanja/Chichewa
a| tukwa
Lolo(Valade?)
Manyawa Munguluni b| t *ori Marenji
b| t 5nmbn
Nyanja/Chichewa a| s nzi Cocola
b| t
a| sodi Sena
b| t 5ombo
Marenji
a| sodi Lomwe
b| t 5nmbu
Cocola
a| teku
Sena
a| sozi
pescoço
língua Lolo of Derre Center
a| k o !i
27 tongue
Lolo of Dula
Lolo of Derre Center a| lu**mi Lolo of Nyanzaza
a| k o !i
a| lu**mi Takwane
a| k o !i
Lolo of Dula
Lolo of Nyanzaza a| lu**mi Chwabo
a| k o !i
a| lu**mi Lolo(Valade?)
a| k o Ü* i
Takwane
a| lu**mi Manyawa Munguluni
a| k o ! oi
Chwabo
a| lu**mi Nyanja/Chichewa
b| sigo
Lolo(Valade?)
Manyawa Munguluni a| lu**mi Marenji
a| kh ns i
Nyanja/Chichewa a| lirimi Cocola
a| k o i
a| lu**mi Sena
a| k o !i
Marenji
a| lu**mi Lomwe
a| k ns i
Cocola
b| siko
Sena
a| li**me
Lomwe
a| lu**mi
24 nose
nariz
Lolo of Derre Center
a| p *uno
Lolo of Dula
a| p *uno
Lolo of Nyanzaza
a| p *uno
Takwane
a| p *una
Chwabo
a| p uno
Lolo(Valade?)
a| p *uno
Manyawa Munguluni
a| p *una
Nyanja/Chichewa
a| p *uno
Sena
a| p *uno a| p *uno
32 human being
a| no
Lolo of Dula
a| t *o Lolo of Nyanzaza
a| no
Lolo of Derre Center
a| u t *u Takwane
a| no
Lolo of Dula
a| t *o Chwabo
a| no
Lolo of Nyanzaza
a| t u Lolo(Valade?)
a| no
Takwane
a| u t *u Manyawa Munguluni
a| nu
Chwabo
a| t u Nyanja/Chichewa
a| no
Lolo(Valade?)
a| t u Marenji
a| nu
Manyawa Munguluni
a| nt Cocola
a| n n
Nyanja/Chichewa
a| t u Sena
a| nu
Marenji
a| u t *u Lomwe
homem Lolo of Derre Center
33 man
c| o so
Lolo of Dula
a| lombwana Lolo of Nyanzaza
c| o so
Lolo of Derre Center
a| lombwana Takwane
c| o so
Lolo of Dula
a| lombwana Chwabo
b| jedo
Lolo of Nyanzaza
a| lombwana Lolo(Valade?)
b| ri Ço
Takwane
a| lo bwana Manyawa Munguluni
b| odo
Chwabo
b| ma Nyanja/Chichewa
b| rudo
Lolo(Valade?)
a| lombwana Marenji
c| k ndzn
Manyawa Munguluni
d| mp nngn Cocola
c| o zn
Nyanja/Chichewa
a| lombwana Sena
c| o so
Marenji
a| lombwana Lomwe
e| tindo
Cocola
b| jodo
Sena
c| muna
Lomwe
a| lo bwana
30 child
criança
filho Lolo of Derre Center
34 son
a| ana
Lolo of Dula
a| ana*** Lolo of Nyanzaza
a| ana
Lolo of Derre Center
a| ana*** Takwane
a| ana
Lolo of Dula
a| ana*** Chwabo
a| ana
Lolo of Nyanzaza
a| ana*** Lolo(Valade?)
a| ana
Takwane
a| ana*** Manyawa Munguluni
a| ana
Chwabo
a| ana 0ga Nyanja/Chichewa
a| ana
Lolo(Valade?)
a| ana*** Marenji
a| ana
Manyawa Munguluni
0 ==no entry=== Cocola
a| a
Nyanja/Chichewa
a| a Sena
a| ana
Marenji
a| ana*** Lomwe
a| ana***
Lomwe
a| ana*ga
31 father
pai
Lolo of Derre Center
a| ba ba
Lolo of Dula
a| ba ba
Lolo of Nyanzaza
Lolo(Valade?)
a| ba ba
Manyawa Munguluni
a| aba ba
Nyanja/Chichewa
a| ambo
Marenji
a| ba ba
Cocola
a| ba ba
Sena
a| ba ba
Lomwe
a| balame
35 twins
gémios
galinha Lolo of Derre Center
39 chicken
b| anasambili
Lolo of Dula
Lolo of Derre Center a| k u ** Lolo of Nyanzaza
b| anasambili
a| k u Takwane
b| anasambini
Lolo of Dula
a| k u Chwabo
c| anapat *a
Lolo of Nyanzaza
a| k u ** Lolo(Valade?)
a| anakono
akwane
a| k u ** Manyawa Munguluni
b| anasambili
Chwabo
a| k *uy** Nyanja/Chichewa
c| anapat a
Lolo(Valade?)
Manyawa Munguluni a| k u ** Marenji
e| map ¡s¡
a| k *u ku Cocola
b| anasa
Nyanja/Chichewa
b| ko o Sena
d| kamwini
Marenji
a| k u ** Lomwe
e| m p *asa
Cocola
f| p *opo
vaca Lolo of Derre Center
a| !i ana
40 cow
Lolo of Dula
Lolo of Derre Center a| 0ombe Lolo of Nyanzaza
a| !i ana
a| 0ombe Takwane
a| !i ana
Lolo of Dula
a| 0ombe Chwabo
a| i !i ana
Lolo of Nyanzaza
a| 0ombe Lolo(Valade?)
a| i !i ana
Takwane
a| 0ombe Manyawa Munguluni
a| i !i ana
Chwabo
a| 0ombe Nyanja/Chichewa
a| !i ana
Lolo(Valade?)
Manyawa Munguluni a| 0ombe Marenji
b| k ¡zi
a| Ônmbe Cocola
a| i ana
Nyanja/Chichewa
a| Sena
a| !i ana
Marenji
a| 0ombe Lomwe
b| kazi
Cocola
a| tijana
Sena
a| Ônmbe
cão Lolo of Derre Center
a| Õjama
41 dog
Lolo of Derre Center a| nambw a Lolo of Nyanzaza
Lolo of Dula
a| Õjama
a| nambw a Takwane j a| n ama
a| Õjama
Lolo of Dula
a| nambw a Chwabo
Lolo of Nyanzaza
a| na bw a Lolo(Valade?)
a| n ama
Takwane
a| na bw a Manyawa Munguluni
a| Õjama
Chwabo
a| na Nyanja/Chichewa
a| n ama
Lolo(Valade?)
Manyawa Munguluni a| na bw 'a Marenji
a| Õ ¡m¡
b| galu Cocola a| Õjama
a| Õjama
Nyanja/Chichewa
a| na Sena
Marenji
a| nambw a Lomwe
a| Õjama
a| nambw a
pássaro
Lolo of Derre Center
a| balam
Lolo of Dula
a| balame
Lolo of Nyanzaza m a| balam ' Takwane
a| balame
Chwabo
a| balame
Lolo(Valade?)
a| balame
Manyawa Munguluni
a| balame
Nyanja/Chichewa
a| mb ¡l¡m'
Marenji
a| bala
Cocola m a| balame Sena
a| balami
Lomwe
c| mwaco
42 elephant
elefante
piolho Lolo of Derre Center
46 louse
b| t embo
Lolo of Dula
b| t *umbo Lolo of Nyanzaza
b| t embo
Lolo of Derre Center
b| t *umbo Takwane
b| t embo
Lolo of Dula
b| t *umbo Chwabo
b| t *e bo
Lolo of Nyanzaza
b| t *u bo Lolo(Valade?)
b| d e o
Takwane
a| talaw Manyawa Munguluni
a|
Chwabo
b| t *umbo Nyanja/Chichewa
b| t *e bo
Lolo(Valade?)
Manyawa Munguluni b| t *u bo Marenji
a| nd <nvu
c| sa w' Cocola
b| t 'mbo
Nyanja/Chichewa
b| t *umbo Sena
b| t *'mbo
Marenji
b| t *ombo Lomwe
a| sawawa
Lomwe
a| talawa
43 fish
peixe
cobra Lolo of Derre Center
a| somb
47 snake
Lolo of Dula
a| Õjowa Lolo of Nyanzaza
a| somb
Lolo of Derre Center
a| Õjowa Takwane
a| somb
Lolo of Dula
a| njo ; a Chwabo
a| somba
Lolo of Nyanzaza
a| Õjo a Lolo(Valade?)
a| o ba
Takwane
a| n o Manyawa Munguluni
a| somb ¡
Chwabo
a| ; a Nyanja/Chichewa
a| o ba
Lolo(Valade?)
Manyawa Munguluni a| n o a Marenji
a| somba
a| nz nka Cocola
a| s nmba
Nyanja/Chichewa
a| Õjowa Sena b| ¤jama za mazi
a| somba
Marenji
a| Õjowa Lomwe
Cocola
c| hopa
cágado Lolo of Derre Ctr.
48 tortoise
a| mbuzi******
Lolo of Dula
a| w amba Lolo of Nyanzaza
a| mbudi******
Lolo of Derre Center
a| w amba Takwane m a| buzi******
a| mbuzi******
Lolo of Dula
a| o ; amba Chwabo
Lolo of Nyanzaza
a| k *a ba Lolo(Valade?)
a| mbuzi******
Takwane
a| k a ba Manyawa Munguluni
a| mbudi !iiana
Chwabo
a| w amba Nyanja/Chichewa
a| mbusi******
Lolo(Valade?)
Manyawa Munguluni a| k a ba Marenji
a| mbuzi******
a| k ¡mb¡ Cocola
a| mbudi******
Nyanja/Chichewa
a| amba Sena
a| mbudi******
Marenji
a| w amba Lomwe
a| mbuzi******
Cocola
b| puri
Sena
a| k amba
minhoca Lolo of Derre Ctr.
a| p *o do goma
49 worm
DISQUALIFIED: most entries were similar to Lolo of Dula
a| p ondo 0goma
“minhoca”, the Portuguese for worm. Lolo of Nyanzaza
a| p *o do goma
Takwane
a| p *o do goma
Chwabo
a| p o Ço gom
Lolo(Valade?)
b| k *alamu
Manyawa Munguluni
a| p o do goma
Nyanja/Chichewa
d| kango
Marenji
a| p o do goma
b| k alamu
Manyawa Munguluni
Nyanja/Chichewa
a| 5ira
a| ila Lolo of Derre Center
Marenji
a| zai
a| ila Lolo of Dula
b| ond <e
Cocola
a| ida Lolo of Nyanzaza
Sena
a| zai
a| ila Takwane
asa Lolo(Valade?)
b| ond <¡
55 wing
Manyawa Munguluni
b| p ep elo Nyanja/Chichewa
a| sai
Lolo of Derre Center
b| p ep elo Marenji
c| zira
Lolo of Dula
b| p ep elo Cocola
b| o t 5e Lolo of Nyanzaza
b| ***p elo Sena
b| ond <e
Takwane
a| kapwa Lomwe
Lolo(Valade?)
b| p *ep*edu
Manyawa Munguluni
b| ***p elo
51 fat
c| p *ikn DISQUALIFIED: multiple senses between lists: fat,
gordura
Nyanja/Chichewa
b| p ep elo oil.
Marenji
Cocola
b| p *ep*elo
Sena
b| p ap idu
*elo Lolo of Derre Center
Lomwe
b| ***p
52 feather of a bird
pena
lago Lolo of Dula
a| t *e0ga
56 lake
DISQUALIFIED: multiple senses between lists: Lolo of Nyanzaza
a| t *e0ga
“lake”, “puddle”, and “calm area in a river”. Takwane
c| tete
c| t 'te
Chwabo
a| t 'ng
montanha Lolo(Valade?)
a| t *'0g¡
57 mountain
Manyawa Munguluni
a| a 0go Nyanja/Chichewa
c| tet '
Lolo of Derre Center
a| a 0go Marenji
a| t *'0g¡
Lolo of Dula
a| a 0go Cocola
a| t e 0ga
Lolo of Nyanzaza
a| a Sena
a| t *e0ga
Takwane
a| a 0go Lomwe
a| t inga
Chwabo
d| purur
Lolo(Valade?)