Mala Pandaram Procedures silesr2015 029.

other hand, the data collected from seven Mala Veda settlements in Kollam district by Hyrunnisa Beegam 1991 shows that the speech of Mala Vedan people in this area has only marginal differences from Malayalam. She suggested that the speech and culture of the people known under the name VedaMalaveda has to be intensely investigated due to their inter-group relation.

2.10 Mala Pandaram

The Mala Pandaram are also known as Hill Pandaram. The majority of them live in the forest tracts of Kollam and Pathanamthitta districts of Kerala and the rest live in Tamil Nadu. According to the 1991 census, the population of Mala Pandaram is 2839 in Kerala and 1930 in Tamil Nadu. The 1999 ITDP report claims that there are only six Mala Pandaram families in Idukki district. In Idukki they live in Pirmed and Peruvanthanam panchayats. The people claim that they came to Kerala from Madurai and Thirunelveli districts of Tamil Nadu. They live in groups of three or four families for a while, and then eventually move on to another site. These sites are generally in the deep interior forests, away from other people. They have the institution of territorial chieftainship, and they remain some of the poorest people in Kerala. They are still semi- nomadic and in the hunting stage of economic development Menon 1996:212. They speak a dialect of Malayalam, locally termed Pandaram Bhasha. The Mala Pandaram converse with others in Malayalam and use Malayalam script Singh 1994:733. In Menon 1996:212, Luiz reports that they “speak a poor dialect with many Tamil and Malayalam words and phrases.” Their religion is a mix of Hinduism and their traditional faith. According to the 1991 census, 37 of the Mala Pandaram are literate 44 of males and 31 of the females. 3 Dialect areas One of the primary goals of this survey was to find out the degree of relationship that exists between Malayalam, Tamil and the speech varieties of Idukki district. Various tools were utilised to accomplish this objective, one of which was lexical comparison of wordlists. Another method of assessing relationships among various speech varieties was dialect comprehension testing using Recorded Text Testing RTT. Finally, formal questionnaires helped to make conclusions about these relationships. These methods and their findings will be discussed in detail in this chapter.

3.1 Lexical similarity

One method of assessing the relationship among speech varieties is to compare the degree of similarity in their vocabularies. This is referred to as lexical similarity. Speakers of varieties that have more terms in common thus a higher percentage of lexical similarity generally, though not always, understand one another better than do speakers of varieties that have fewer terms in common. Since only elicited words and simple verb constructions are analysed by this method, lexical similarity comparisons alone cannot indicate how well certain speech communities understand one another. It can, however, assist in obtaining a broad perspective of the relationships among speech varieties and give support for results using more sophisticated testing methods, such as comprehension studies.

3.1.1 Procedures

The tool used for determining lexical similarity in this survey was a 210-item wordlist, consisting of items of basic vocabulary, which has been standardised and contextualised for use in sociolinguistic surveys of this type in South Asia. The elicited words were transcribed using the International Phonetic Alphabet IPA shown in appendix A. To provide maximum reliability, some of the wordlists were checked with a second mother tongue speaker at the same site. Each wordlist was compared with every other wordlist, item by item, to determine to what degree they were phonetically similar. Those words that were judged similar were grouped together. Once the entire wordlist was evaluated, the total number of word pair similarities was tallied. This amount was then divided by the total number of items compared and multiplied by 100, giving what is called the lexical similarity percentage. This process of evaluation was carried out according to standards set forth in Blair 1990:30–33 and facilitated using a computer program called WordSurv Wimbish 1989. This program is designed to quickly perform the counting of word pair similarities and to calculate the lexical similarity percentage between each pair of wordlists. For a more complete description of counting procedures used in determining lexical similarity, refer to Appendix B.

3.1.2 Site selection

Twenty-one wordlists were compared in this lexical similarity study. Seventeen of the wordlists were collected during this survey in Muthuvan, Mannan, Paliyan and Mala Pulayan Karavazhi Pulayan villages. More focus was given to Muthuvan and Mannan speech varieties. Fifteen wordlists were collected from them, representing eight and seven locations respectively. It was reported that Paliyan and Mala Pulayan people speak a form of Tamil. One wordlist was collected from each of these varieties to verify the language situation. Two wordlists from Urali from a previous survey are also included in the lexical similarity comparison. Finally, one standard wordlist in Malayalam, the official language of the state covered in this survey, and one wordlist in Tamil, the neighbouring state language, were included in the lexical similarity comparison. Eight wordlists were collected from Muthuvan. Five of these were from the Tamil Muthuvan variety and three were from the Malayalam Muthuvan variety. These sites were selected based on people group division as Tamil Muthuvan and Malayalam Muthuvan, geographical distribution interiorexterior and reported variation in dialect areas. Travel facility and permission to visit the area also influenced the site selection. Mannan wordlists were collected from seven sites based on geographical distribution interiorexterior, reported dialect variation and cultural importance. In addition, importance also was given to visit major settlements of the area for wordlist collection. Wordlists from Mala Vedan, Mala Arayan and Ulladan were not collected from the survey area since it is believed that the language varieties are no longer in use and have been replaced by Malayalam. Surveyors could not visit Mala Pandaram villages, as there was no information about any settlement that consists of more than three families in the district. Map 2 shows the locations of the wordlist sites. Table 5 gives the speech variety, location and origin of the wordlists utilised in this project. Map 2. Wordlist sites © NLCI 2015 Source: New Life Computer Institute, India. Used with permission. 14 Table 5. Wordlist sites Language Speech Variety Village Settlement Interior Exterior Block Tahsil District State Origin Tamil Muthuvan AnachalItticity Exterior Adimali Devikulam Idukki Kerala This Survey Tamil Muthuvan Chempakathozhu Exterior Devikulam Devikulam Idukki Kerala This Survey Tamil Muthuvan Kavakudi Interior Devikulam Devikulam Idukki Kerala This Survey Tamil Muthuvan Kozhiyala Interior Devikulam Devikulam Idukki Kerala This Survey Tamil Muthuvan Valsapetti Interior Devikulam Devikulam Idukki Kerala This Survey Malayalam Muthuvan Kunchipara Interior Neriyamangalam Kothamangalam Ernakulam Kerala This Survey Malayalam Muthuvan Thalayirappan Exterior Adimali Devikulam Idukki Kerala This Survey Malayalam Muthuvan Kurathikudi Interior Adimali Devikulam Idukki Kerala This Survey Mannan Vattamedu Exterior Idukki Udumpanchola Idukki Kerala This Survey Mannan Veliyampara Interior Devikulam Devikulam Idukki Kerala This Survey Mannan Kumily Exterior Azhutha Pirmed Idukki Kerala This Survey Mannan Kovilmala Exterior Kattappana Udumpanchola Idukki Kerala This Survey Mannan Kodakallu Interior Adimali Devikulam Idukki Kerala This Survey Mannan Chinnapara Exterior Adimali Devikulam Idukki Kerala This Survey Mannan Thinkalkadu Interior Nedumkandam Udumpanchola Idukki Kerala This Survey Urali Poovanthikudi Interior Kattappana Udumpanchola Idukki Kerala Betta Kurumba Urali Vanchivayal Interior Pirmed Pirmed Idukki Kerala Betta Kurumba Mala Pulayan Dendukombu Exterior Devikulam Devikulam Idukki Kerala This Survey Paliyan Lebbakandam Exterior Azhutha Pirmed Idukki Kerala This Survey Tamil Pudukottai and Tuticorin Tamil Nadu Betta Kurumba Malayalam Ernakulam Kerala Betta Kurumba Table 6. Lexical similarity percentages I Muthuvan, Itticity 94 B Muthuvan, Chempakathozhu 93 93 W Muthuvan, Kavakudi 93 92 91 O Muthuvan, Kozhiyala Cluster-A 93 91 91 92 S Muthuvan, Valsapetti 88 84 83 85 85 J Muthuvan, Kunchipara 84 80 83 80 80 86 R Muthuvan, Thalayirappan 81 78 77 78 79 88 86 K Muthuvan, Kurathikudi 67 64 66 65 67 71 66 63 D Mannan, Vattamedu 68 65 66 66 68 71 67 67 96 N Mannan, Veliyampara 65 66 65 64 67 69 66 64 96 93 M Mannan, Kumily 65 65 65 64 67 67 66 64 94 95 96 Z Mannan, Kovilmala Cluster-B 67 65 65 64 67 71 66 65 94 93 92 92 A Mannan, Kodakallu 63 62 63 63 64 68 66 63 91 94 90 92 90 C Mannan, Chinnapara 65 63 63 65 65 68 67 67 89 88 88 90 86 86 T Mannan, Thinkalkadu 55 54 52 54 55 58 61 64 51 53 52 53 51 50 52 P Urali, Poovanthikudi Cluster-C 55 54 53 52 55 58 59 64 52 54 53 52 52 50 53 83 V Urali, Vanchivayal 70 67 67 66 71 66 65 64 67 67 64 66 66 63 66 53 54 U Mala Pulayan, Dendukombu Cluster-D 70 67 66 66 71 67 65 66 64 65 66 67 64 62 65 56 57 85 L Paliyan, Lebbakandam 67 64 62 64 65 65 62 64 57 60 58 61 58 57 60 54 54 79 75 Y Tamil 63 60 58 62 60 65 63 68 57 61 59 59 57 56 64 59 61 61 62 69 F Malayalam 16

3.1.3 Results and analysis

Table 6 shows the lexical similarity percentage of wordlists compared in this study. According to Blair 1990:24, it can typically be concluded that two-speech varieties that have less than 60 lexical similarity are different languages. For speech varieties that have greater than 60 lexical similarity, intelligibility testing should be done to determine their relationship. In the following analysis and discussion, wordlists are grouped in different clusters based on their highest lexical similarity. The study gives more emphasis to the lexical relationships within each tribal variety and with the state languages Malayalam and Tamil, rather than the relationships of different tribal groups’ varieties with each other. Table 7 shows the ranges of lexical similarity within each cluster. Table 7. Ranges of lexical similarity within each cluster Cluster Speech varieties Range of lexical similarity A Muthuvan 77–94 B Mannan 86–96 C Urali 83 D Mala Pulayan and Paliyan 85 3.1.3.1 Cluster-A Muthuvan There are two types of Muthuvan wordlists, namely Tamil Muthuvan and Malayalam Muthuvan. All of the wordlists from Muthuvan varieties show a range from 77 to 94 of lexical similarity with each other. Within Tamil Muthuvan varieties, the percentages vary from 91 to 94 and make it clear that they are varieties of the same speech form. Malayalam Muthuvan varieties share a range of 86 to 88 lexical similarity with each other. The distances between the Malayalam Muthuvan sites are great and there does not seem to be much contact between these villages. This may be the reason for the lower degree of lexical similarity among the wordlists from Malayalam Muthuvan as opposed to Tamil Muthuvan villages. The Tamil Muthuvan wordlists share a range of 77 to 88 of lexical similarity with the Malayalam Muthuvan wordlists. This may be taken as a rough indication that they may be different speech varieties of the same language. Intelligibility testing may help clarify the situation. The wordlists collected from the Tamil Muthuvan variety show slightly greater similarity with Tamil from 62 to 67 than Malayalam 58 to 63, the state language of Kerala. On the other hand, Malayalam Muthuvan varieties show about the same resemblance to Malayalam 63 to 68 and Tamil 62 to 65. The study shows that the influence of Malayalam and Tamil can be seen in Muthuvan varieties. 3.1.3.2 Cluster-B Mannan The wordlists collected from Mannan varieties appear to be quite similar. The overall range of lexical similarity among these wordlists is from 86 to 96. It can be observed that the wordlist from Thinkalkadu shows comparatively less similarity with other Mannan varieties. The overall percentage of similarity among Mannan wordlists is higher when the Thinkalkadu wordlist is not included in the comparison ranging from 90 to 96. The Thinkalkadu wordlist shows slightly greater similarity with Malayalam 64 than other Mannan wordlists do. The influence of Malayalam might explain this variation from other Mannan wordlists. The Mannan wordlists show a range of 57 to 61 similarity with the Tamil wordlist. Although it was reported that Mannan is a dialect of Tamil, the lexical study shows that they are different languages. Likewise, the Mannan wordlists share only 56 to 64 similarity with Malayalam. These percentages are on the borderline of what is considered to be different languages 3.1.3.3 Cluster-C Urali The Urali wordlists share 83 lexical similarity with each other. They share a range of 59 to 61 of similarity with Malayalam and 54 of similarity with Tamil and show that Urali is probably a different language from Malayalam and Tamil. Compared to other groups, Urali shows the most dissimilarity with the wordlists from other groups, as almost all percentages are in the 50s. 3.1.3.4 Cluster-D Mala Pulayan and Paliyan The wordlists from Mala Pulayan and Paliyan shows more similarity to Tamil than other wordlists in this survey, with 79 and 75 respectively. This gives the impression that they are related varieties to Tamil. Although Mala Pulayan and Paliyan are different tribal communities, their wordlists show 85 lexical similarity to each other. Lexical comparison shows that the speech forms of Mala Pulayan and Paliyan are different from Malayalam 61 and 62 similarity respectively.

3.1.4 Conclusion

Based on lexical similarity, wordlists were grouped into different clusters, each of which represented the people groups studied. Within each cluster, lexical similarity was fairly high, indicating there was not much dialectal variation. Muthuvan showed more variation, particularly between Tamil Muthuvan and Malayalam Muthuvan. In general, the relationship between the tribal varieties and Malayalam and Tamil is such that they are on the border of being considered different languages. Mala Pulayan and Paliyan, however show a higher similarity with Tamil.

3.2 Intelligibility testing

Maggard 1998:14 has noted, “The definition of a language and a dialect is not always clear. The two terms have been used in many different ways. Common usage often applies the term language to the large, prestigious languages, which have an established written literature. The term dialect is then used for all other speech varieties. Some linguists use language to refer to speech varieties that share similar vocabularies, phonological andor grammatical systems. Many times, the sense in which the two terms are used can vary.” The researchers believe that an important factor in determining the distinction between a language and a dialect is how well speech communities can understand one another. Low intelligibility between two speech varieties, even if one has been classified as a dialect of the other, impedes the ability of one group to understand the other Grimes 2000:vii. Thus comprehension testing, which allows a look into the approximate understanding of natural speech, was an important component of this research.

3.2.1 Procedures

Recorded Text Testing RTT is one tool that helps assess the degree to which speakers of related linguistic varieties understand one another. A three to five minute natural, personal-experience narrative is recorded on cassette from a mother tongue speaker of the speech variety in question. It is then evaluated with a group of mother tongue speakers from the same region by a procedure called Hometown Testing HTT. This ensures that the story is representative of the speech variety in that area and is suitable to be used for testing in other sites. Mother tongue speakers from other locations and differing speech varieties then listen to the recorded stories and are asked questions, interspersed in the story, to test their comprehension. Subjects are permitted to take tests of stories from other locations only if they perform well on a hometown test. This ensures that the test-taking procedure is understood. Ten is considered the minimum number of subjects to be given this test, and subjects’ responses to the story questions are noted down and scored. A person’s score is considered a reflection of his comprehension of the text, and the average score of all subjects at a test point is indicative of the community’s intelligibility of the speech variety of the story’s origin. Included with the test point’s average score is a calculation for the variation between individual subjects’ scores, known as standard deviation, which helps in interpreting how representative those scores are. After each story, subjects are asked questions such as how different they felt the speech was from their own and how much they could understand. These subjective post-RTT responses give an additional perspective for interpreting the objective test data. If a subject’s answers to these questions are comparable with his or her score, it gives more certainty to the results. If, however, the post-RTT responses and test score show some dissimilarity, then this discrepancy can be investigated. Refer to appendix C1 for a more complete description of Recorded Text Testing, as well as to Casad 1974. The stories and questions used in the testing also appear in appendix C2. In appendix C3, the demographic profiles of the subjects at each test site, the test scores and the post-HTTRTT responses are given.

3.2.2 Site selection

The primary objective of this study was to investigate the comprehension of speakers of Tamil Muthuvan and Malayalam Muthuvan of each other’s variety. There is only occasional contact between these two Muthuvan groups. In this study, two stories were developed one in a Tamil Muthuvan settlement and the other in a Malayalam Muthuvan settlement and tested. Another objective was to investigate the reported variation among Mannan speech varieties. For this purpose, one story was developed from a northern part of the Mannan-speaking area and tested in a southern part. Table 8 summarises the information about the stories and map 3 shows the test sites. Table 8. Stories utilised in this project Languagespeech variety Location Tahsil District Tamil Muthuvan Kozhiyala Devikulam Idukki Malayalam Muthuvan Thalayirappan Devikulam Idukki Mannan Kodakallu Devikulam Idukki Map 3. RTT and LUAV sites. © NLCI 2015 Source: NLCI. 2015. Used with permission. 3.2.2.1 Kozhiyala Tamil Muthuvan A Tamil Muthuvan story was recorded in Kozhiyala, a settlement of Devikulam Tahsil. Kozhiyala is an interior settlement that is situated in the forest. This would represent a central area of the Tamil Muthuvan variety. Idamalakudi is also a major settlement of Muthuvan and may be the most geographically central place for Tamil Muthuvan. However, the researchers could not get permission to enter that area. 3.2.2.2 Thalayirappan Malayalam Muthuvan Malayalam Muthuvan is spoken in Adimali block of Devikulam tahsil. Most of the settlements are in very interior locations. A Malayalam Muthuvan story was recorded in Thalayirappan settlement. The settlement is six kilometres away from Adimali town and more easily accessible than other Malayalam Muthuvan settlements. This ideal location both represented the Malayalam Muthuvan variety and allowed the surveyors to have good contact with the people. 3.2.2.3 Kavakudi Tamil Muthuvan Kavakudi is a Tamil Muthuvan settlement. It represents the north-eastern concentration of Tamil Muthuvan settlements. This site was selected for gauging comprehension of Tamil Muthuvan people of the Malayalam Muthuvan text from Thalayirappan. Another reason for selecting this site was to investigate the acceptability of the language used in the Kozhiyala story that comes under the central area of Tamil Muthuvan variety. 3.2.2.4 Kodakallu Mannan A Mannan story was collected from the settlement of Kodakallu in Devikulam tahsil to investigate whether there is any variation in the Mannan variety, as had been reported by this people group. It was believed that recording a story in one extreme end of the area and administering comprehension testing with it in another extreme end of the area would help clarify this situation. It was reported that people in Kodakallu speak a pure variety of the Mannan language, as compared with other Mannan settlements in that region. This is an interior village and has little direct contact with outsiders. The researchers had already made good contacts with the residents, which also supported the selection of this site for story collection. 3.2.2.5 Kumily Mannan Kumily is located at an extreme southern end of the Mannan area. It was reported that their speech form has a unique style. Thus, it appeared to be a good location in which to check the acceptability and understanding of the speech form used in the Kodakallu story.

3.2.3 Intelligibility testing results and analysis

The results of intelligibility testing are shown in table 9. The rows of the table list the villages where the stories were tested and the columns list each story used for testing. Hometown and control tests are given in shaded boxes. Table 9. Results of Recorded Text Testing Stories Played Communities Tested Snake Story Tamil Muthuvan Our Lifestyle Story Malayalam Muthuvan Elephant Story Mannan Kozhiyala Tamil Muthuvan Avg SD N= 95 6.5 11 Thalayirappan Malayalam Muthuvan Avg SD N= 87 13.1 11 96 4.2 11 Kavakudi Tamil Muthuvan Avg SD N= 90 7.6 11 82 15.8 11 Kodakallu Mannan Avg SD N= 91 5.7 10 Kumily Mannan Avg SD N= 92 7.7 20 In interpreting RTT results, three pieces of information are necessary. The first is average percentage shown as Avg , which is the mean or average of all the participants’ individual scores on a particular story at a particular test site. Also necessary is a measure of how much individuals’ scores varied from the community average, called standard deviation shown as SD. The third important piece of information is the size of the sample, that is, the number of people that were tested at each site shown as N=. In addition, to be truly representative, a sample should include people from significant demographic categories, such as both men and women, young and old, and educated and uneducated. The relationship between test averages and their standard deviation has been summarised by Blair 1990:25 and can be seen in table 10. Table 10. Relationship between test averages and standard deviation Standard Deviation High Low Average Score High Situation 1 Many people understand the story well, but some have difficulty. Situation 2 Most people understand the story. Low Situation 3 Many people cannot understand the story, but a few are able to answer correctly. Situation 4 Few people are able to understand the story. Since results of field-administered methods such as Recorded Text Testing cannot be completely isolated from potential biases, O’Leary 1994 recommends that results from RTTs not be interpreted in terms of fixed numerical thresholds, but rather be evaluated in light of other indicators of intelligibility such as lexical similarity, dialect opinions, and reported patterns of contact and communication. In general, however, RTT mean scores of around 80 or higher with accompanying low standard deviations usually ten and below; high standard deviations are about 15 and above are usually taken to indicate that representatives of the test point dialect display adequate understanding of the variety represented by the recording. Conversely, RTT average scores below 60 are interpreted to indicate inadequate comprehension. The following section highlights the results of comprehension testing, discussed in terms of the understanding of each story. 3.2.3.1 Tamil Muthuvan story The test subjects of Kozhiyala scored well on their HTT with a low standard deviation, indicating that the story adequately represents their speech variety. Post-HTT responses indicate that the subjects believe that the language in the story is good, pure and represents their area. The Tamil Muthuvan text from Kozhiyala was tested among Malayalam Muthuvan subjects at Thalayirappan for comprehension. The Thalayirappan subjects averaged 87 on the Kozhiyala story with a standard deviation of 13.1 which is neither high nor low. It is thus believed that most Thalayirappan Malayalam Muthuvan subjects understand the Kozhiyala Tamil Muthuvan story, but some have difficulty. Five out of the 11 Thalayirappan subjects mentioned that the Muthuvan language of the Kozhiyala text was mixed with Tamil. Three of the subjects stated that the language is Muthuvan, whereas two said that the language is not theirs. Most of the subjects mentioned the name of Tamil Muthuvan settlements as being the origin of the text. The subjects identified the language used in the story according to the language variety Tamil mixing, the people Tamil Muthuvan and previous contact having lived in that area. Four out of the five subjects that were asked the question “Was the text pure?” reported that “Yes, it is pure and good.” Most of the subjects mentioned that the language variety used in the text was somewhat different either a little different or very different from their local language variety. Seven out of 11 subjects whom the researchers asked this question reported that they fully understood the story, whereas three said they did not. The comprehension testing among Malayalam Muthuvan subjects showed that many of them understood the Tamil Muthuvan variety used in the story. It seems that the Malayalam Muthuvan subjects accepted the Tamil Muthuvan variety as a good variety. Most of the subjects reported that Tamil has influenced the language used in the story. The Tamil Muthuvan subjects from Kavakudi scored an average of 90 with a standard deviation of 7.6 on Kozhiyala story, indicating that the subjects understood the story well. The language of the story was identified as their village variety or other Tamil Muthuvan village variety. The Post RTT responses reveal that the subjects understood the story fully and commented that the language was pure. 3.2.3.2 Malayalam Muthuvan story Thalayirappan subjects scored an average of 96 on their HTT with a low standard deviation of 4.2, indicating a valid HTT and an understandable text. The text was understood and identified by all the subjects as being from their own variety. They also reported that the language of the story was pure and good. The story from Thalayirappan was played for subjects in Kavakudi, a Tamil Muthuvan settlement. This was done in order to test comprehension and to learn about attitudes towards the Malayalam Muthuvan language variety. The subjects from Kavakudi scored an average of 82 with a standard deviation of 15.8. Though it can be interpreted that the Tamil Muthuvan subjects understood the story, the high standard deviation indicates that some of them had problems understanding the text. The Kavakudi subjects identified the Thalayirappan text as belonging to Muthuvan, by stating Malayalam Muthuvan settlement names as the origin of the story, such as Kurathikudi and Padikappu. Almost all of them pointed out that the text was mixed with Malayalam and that helped them in identifying the text as being spoken in the Malayalam Muthuvan variety. Only four subjects were asked whether the text was pure – to which three of them replied ‘yes’. All but one subject reported that the Thalayirappan text was only a little different from their speech. 3.2.3.3 Mannan story The subjects for the Kodakallu HTT averaged a score of 91 with a standard deviation of 5.7 on the ‘elephant story’. The subjects identified the language of the story as being the same as they speak in their village and said that the speech variety is good and free from any mixing, except for one subject who reported that there is one Malayalam word in the text. The 20 Mannan subjects in Kumily had an average of 92 on the story from Kodakallu, with a low standard deviation of 7.7. Various responses were given concerning the origin of the story. Four mentioned that the story was from their settlement, whereas five realised that the story was not from their place. Three subjects said that it belonged to Kovilmala, the king’s place. Some subjects mentioned other places by name. Finally, another five subjects said that they did not know where the story was from. The responses, taken as a whole, give the impression that a majority of the Kumily subjects did not think that the story was from their own village. Most of the subjects reported that, because of the slight difference in style, wording and tone from that of their own speech, the story was not from their own area. All but five of the subjects considered the speech good and pure. Two subjects said that it was not pure because they believe that Malayalam language was mixed into the text. Half of the subjects stated that there is no difference between their speech and the speech of the story. And about half of the subjects reported that the ‘tone and wording’ of the story is a little different from how they speak. All of the subjects claimed to have fully understood the story. Some RTT subjects suggested that there may be minor variation among the speech varieties of Mannan in their style, wording and tone. Some influence of Malayalam vocabulary can be observed. Despite these observations, most subjects understood the story.

3.2.4 Conclusion

The comprehension testing among the speakers of Malayalam Muthuvan showed that many of them understood the Tamil Muthuvan variety. Many of them found no complexity in understanding this related speech variety. In general, there were no strong negative attitudes expressed by either group on the stories they listened to. There were comments from both groups about the amount of mixing of Tamil and Malayalam in the Tamil Muthuvan and Malayalam Muthuvan respectively. Because of this, many people identified the origin of the speech variety they listened to. The intelligibility study among Mannan gives the impression that there is little language variation among Mannan speech forms. Moreover, some RTT subjects suggested that there may be minor variation among the speech varieties of Mannan in their style, wording and tone. Despite these observations, most subjects understood the story. 4 Language use, attitudes and vitality A study of language use patterns attempts to describe which languages or speech varieties members of a community use in different social situations. These situations, called domains, are contexts in which the use of one language variety is considered more appropriate than another Fasold 1984:183. A study of language attitudes generally attempts to describe people’s feelings and preferences towards their own language and other speech varieties around them, and what value they place on those languages. Ultimately, these views, whether explicit or unexpressed, will influence the results of efforts towards literacy and the acceptability of literature development. Language vitality is another key concept in sociolinguistic research. It refers to the prospects for a language to continue to be spoken by mother tongue speakers and passed on to future generations. Language maintenance refers to a stage in which a section of the people use the second language more, but most continue to speak the mother tongue. The total replacement of a mother tongue with a second language can be referred to as language shift. Many variables have been said to contribute to vitality, such as social status of the language, the number of speakers and whether it has a writing system.

4.1 Procedures

Orally administered questionnaires were the primary method for assessing patterns of language use, attitudes and vitality in this survey. In addition to these questionnaires, observations and informal interviews were also made. The questionnaires were asked in Malayalam or the subject’s mother tongue if they did not speak the state language. The Language Use, Attitudes and Vitality LUAV questionnaire can be found in Appendix D. Questionnaire responses were collected from six sites shown in map 3, three each from Muthuvan and Mannan communities. The following codes and abbreviations are used in this chapter: MT means ‘mother tongue’, LWC means ‘language of wider communication’ Malayalam and “Both” when used in tables means ‘mother tongue and language of wider communication Malayalam’. Table 11 shows the questionnaire sites. Table 11. Questionnaire sites People Site Block District Muthuvan Kozhiyala Devikulam Idukki Muthuvan Kavakudi Devikulam Idukki Muthuvan Thalayirappan Devikulam Idukki Mannan Kodakallu Devikulam Idukki Mannan Kovilmala Kattappana Idukki Mannan Kumily Azhutha Idukki

4.2 Sampling distribution for questionnaire subjects