2.1.2 Dialect intelligibility
Another goal of the dialect study is to investigate the degree of inherent intelligibility between the different speech varieties spoken in the Western Indo-Nepal Tarai. Inherent intelligibility is
understanding that is the result of two speech varieties being closely related to one another, as opposed to understanding that is acquired by exposure. Intelligibility testing, used in tandem with lexical
similarity, helps determine the most meaningful grouping of speech varieties. Inherent intelligibility is inferred from a sample’s average understanding of a recorded text. The results are expressed as a
percentage, based on the average of scores on a recorded text test. Sample size and standard deviation are also given.
2.2 Bilingualism
The third goal, concerning community bilingualism in Hindi, was formulated because of the important role that Hindi plays in education, commerce, media, and religion. Extensive and high-level bilingualism
would suggest that local communities can effectively use media in these languages. In this study, bilingualism is investigated using both recorded text tests and a sentence repetition test. The results for
recorded text tests are expressed as in §2.1.2. The results for sentence repetition testing are expressed as the percentage of the population at a particular level of fluency.
2.3 Language use and language attitudes
The fourth goal concerning language attitudes towards Tharu was formulated because of the powerful effect that attitude has on the acceptance and use of literature, and because present language use
patterns provide an indication of the future of a language. Strong attitudes against Hindi, or in favor of the development of Tharu, could significantly restrict the understanding and acceptance of written
materials in Hindi, even if research suggests it is adequately understood by Tharu speakers. This investigation uses formal and informal questionnaires, in addition to observation. Results are expressed
as the percentage of subjects giving a particular response; they are also documented with quotes from individuals interviewed and with observations by the researchers.
2.4 Tharu culture
The final goal, that of investigating the Tharu culture, was formulated because of the need to understand important cultural differences before initiating cross-cultural work. In this study, culture is investigated
by observation and informal interviews, using a list of topics and questions to guide the investigation. Results are presented in this report as a summary of general information about the Tharu, included
mostly in §1. More specific details are given in “A Summary of Tharu Culture” Appendix C.
3 Summary of findings
3.1 Tharu speech varieties
There is considerable diversity among the Tharu varieties encountered in this survey. Language differences seem to result more from clan differences than from geographic separation, though the effect
of both is evident.
The different varieties share as little as 58 percent lexical similarity. Among these, there are four subgroups that arise from the study of the wordlists: 1 Buksa, 2 Rana, 3 Kathoriya and Sunha, and 4
Dangora from west to east. These four subgroups form a dialect chain in which similarity decreases with distance. From the wordlists, Kathoriya Tharu KkP
4
appears to be a potential center point in the chain, a variety which even the ends of the chain might understand.
Recorded text testing confirms the suspicion that Kathoriya Tharu is a center point in a dialect chain. Of those varieties tested on their understanding of Kathoriya Tharu KkP, subjects averaged 90
percent or better. At the ends of the dialect chain, however, Rana Tharu RNs understood the Dang Tharu DDK text at an average of only 51 percent DDK understood the RNs text at an average of 68
percent.
Besides Kathoriya KkP, no other reference point was understood well at all of the other test points. Interestingly, not only is KkP understood well at the other test points, but also subjects in KkP
understood the other reference points better than any other test point. Only four of six possible test points were tested on the Kathoriya recorded text test RTT. Neither
Rana RNs nor Buksa BNM both in Nainital District were tested on the Kathoriya RTT. Because RNs, BNM, and RKB had such uniformly high scores among themselves, it is expected that results from RKB
would be fairly representative of results from RNs and BNM. Final verification of the suspected high intelligibility at these two points needs further testing.
From lexical similarity results, it appears that Chitwan Tharu, to the east of this survey area, is quite different from the varieties studied here, and should be thought of as a separate language.
3.2 Tharu’s relationship to Hindi