論文題目：LANGUAGE CHOICE AND SOCIAL MEDIA IN UKRAINE
著者：ﾊﾟﾌﾞﾘｰ ﾎﾞｸﾞﾀﾞﾝ （Pavliy Bogdan）
論文審査委員：ジョナサン ルイス、足羽興志子、五十嵐 陽介、Milen Martchev
1. Outline of the Thesis
This thesis uses data from the microblogging service Twitter and social
networking service Facebook to analyze the language preferences of online social
media users in Ukraine. It describes the linguistic choices of social media users
in Ukraine in relation to their gender, age and geographical location. The main
dataset used is a body of geotagged tweets sent from the territory of Ukraine in
2015, and a supplementary study is made of posts and comments published on
the Facebook pages of 24 local governments between 2010 and 2016.
The findings suggest that in relation to language use on social media, overall
Russian is used much more than Ukrainian: the ratio of Russian to Ukrainian in
the Twitter dataset is approximately 6:1. It also finds that the borderline
between greater and more moderate use of Russian language lies not in the
countryʼs far east, where the majority of those surveyed by the 2001 census
declared Russian to be their mother tongue, but rather more centrally, either
following or even veering to the west of the electoral border that has been drawn
by the results of national elections since 2004.
The thesis also finds a stark difference in language behavior between those
who tweet in Ukrainian and those who tweet in Russian. Whereas more than half
of those using Ukrainian also tweeted in Russian, fewer than one in ten of those
using Russian also tweeted in Ukrainian. Use of both languages was higher in
urban areas for both groups.
Finally, after developing an algorithm to detect the gender of users from the
text of Russian and Ukrainian tweets, the author finds that female users in the
dataset outnumber male users by nearly two to one. He also finds that female
users are statistically significantly more likely than men to write fewer than the
overall average (11.2%) of their tweets in Ukrainian.
The analysis of Facebook pages suggests that local governments in some
regions ignore the status of Ukrainian as the only language for official use and
adapt their language use to that of their citizens. The use of Russian and
Ukrainian in page updates by local governments tends to reflect the language
use of citizens in their areas as reported in the 2001 census.
2. Evaluation of the Thesis
The research presented in this thesis represents probably the first study to use
social media to investigate the geographical and demographic aspects of
language use in Ukraine. This research is particularly significant given the
politicized nature of language use and language policy in Ukraine, as well as the
lack of nationwide surveys̶the last census was carried out in 2001. Several
studies have investigated the use of social media during the countryʼs recent
political turmoil, but this thesis is unique in taking a broader approach and
analyzing all geotagged Twitter traffic, most of which is not political, to gain
insights into the daily language use of social media users across Ukraine.
This thesis also breaks new ground in its analysis of the relation between
gender and language choice. While a large number of studies have investigated
why, how and how much women and men use social media, surprisingly few
have focused on which language men and women use in their daily online
communications. This thesis confronts and deals to an adequate extent with the
technical problem of how to identify the gender of Twitter users about whom we
lack basic demographic information.
This thesis is also somewhat unusual in using data from more than one social
networking service. While the thesis centers on the analysis of tweets for reasons
of data availability, a limited amount of data from Facebook is used to triangulate
the findings of the Twitter analysis to the extent possible.
The thesis does however have some weak points. Chapter 2 devotes a lot of
space to previous research on bilingualism, but bilingualism is not a central part
of the empirical study. In addition, more use could have been made of language
prestige in explaining the results. For example, the thesis is unable to offer an
explanation for its finding that female Twitter users in Ukraine have a stronger
preference for Russian than their male counterparts. This misses the sociolinguistic
research showing that women tend to use more prestigious languages
than men. Similarly, the breakdown of language preferences by region and
gender shows that women do not prefer Russian over Ukrainian in the west of
the country, which may indicate that Ukrainian is not less prestigious than
Russian there. Finally, the rather low success rate of identifying male users
requires us to be cautious in evaluating the results.
On 12 January 2018 we examined Mr. Bogdan Pavliy regarding his PhD thesis “Language Choice and Social Media in Ukraine.” Mr. Pavliy satisfactorily answered all our questions regarding his thesis. We therefore conclude that Mr. Pavliy has achieved the requisite level of academic achievement and ability to be awarded the degree of PhD in Social Sciences from this University.