A comparative study of tag questions in four Asian Englishes from a corpus-based approach

Abstract

The purpose of this study was to describe and compare the features and use of tag questions in four Outer Circle Asian Englishes in order to investigate how a variant tag questions system of English has been modified in the process of indigenization. Four Asian Englishes were analyzed using the International Corpus of English, namely, Hong Kong English (ICE-HK), Philippine English (ICE-PHI), Indian English (ICE-IND), and Singapore English (ICE-SIN). The Canadian corpus of ICE was also analyzed in order to enable comparison with an Inner Circle English. A search using concordance software yielded 181 tag questions in ICE-HK, 54 in ICE-PHI, 125 in ICE-IND, 284 in ICE-SIN, and 167 in ICE-CAN. The features of tag questions were statistically analyzed and compared from the following points: frequency of occurrence, text types, polarity, features of tags, and grammatical agreement between anchors and tags. The use of tag questions was then analyzed by classifying tag questions into seven functional categories: informational, confirmatory, facilitative, attitudinal, peremptory, aggressive, and others. Relationships between polarity types and functions were investigated as well. The results indicated that Asian Englishes have characteristic features especially in terms of the polarity and the main use of tag questions. Two potential explanations, namely, the influence of substrate languages and the developmental phases of Englishes, are discussed.

Keywords:

• tag questions
• text types
• polarity
• tags
• functions
• the International Corpus of English
• Asian Englishes

Notes

The results of this study were presented at the 18th Annual Conference of the International Association for World Englishes in Guangzhou in December 2012.

¹ See, for example, Kachru (1992) for a detailed explanation of the Inner Circle, the Outer Circle, and the Expanding Circle of English.

² In this paper, each variety is referred to as Hong Kong English, Philippine English, Indian English, Singapore English, and Canadian English, respectively, following the terminology used in the International Corpus of English.

³ Singapore, for example, is said to be in a diglossic situation of Colloquial Singapore English and Standard Singapore English (e.g. Zhiming and Huaqing, 2006). This has led some researchers to use only the spoken component of the corpus for the analysis of English used in Singapore (e.g. Sheng, 2007). However, the present study uses the entire corpus for the analysis due to the reason mentioned above. This is in line with studies such as Mukherjee and Gries (2009) and Nesselhauf (2009).

⁴ ‘Constraints’ here include the elimination of extra-corpus texts. Some examples of other constraints are ‘eliminate potential tags that follow interrogative words (e.g. what, why)’, ‘eliminate potential tags that are followed by a present participle’, and so on.

⁵ Columbus (2010) treats ‘isn’t it’ and ‘is it’ as invariant tags where ‘the subject and number of the verb in the main clause did not agree with the tag’ (p. 296). In the present study, however, ‘isn’t it’ and ‘is it’ will be treated as tag questions regardless of the agreement as the definition in this study demonstrates. This treatment of ‘isn’t it’ and ‘is it’ is in line with other studies on tag questions such as Wong (2007) and Borlongan (2008). This does not exclude the possibility that tag questions could be used invariantly. For example, Cheng and Warren (2001) took an intermediate position by pointing out that ‘isn’t it’ and ‘is it’ are invariant forms of tag questions. That is, they analyzed these in the framework of tag questions, at the same time noting that they might be used invariantly.

⁶ The labels were adopted from Tottie and Hoffmann (2006), except for the use of the label ‘facilitative’ instead of ‘facilitating’.

⁷ The number of texts in each text type varies, yet the proportions of text types in the corpora are assumed to reflect those in the English varieties they represent (see Nelson, 1996). Accordingly, the original number of texts in each category is retained as it is when conducting chi-square tests and calculating the proportions expressed as percentage.

⁸ See note 5.

⁹ Only the necessary markup symbols for the contextual analysis are retained in the examples: ‘<,> short pause’, ‘<,,> long pause’, ‘<[>…</[> overlapping string’, ‘<{>…</{> overlapping string set’, and ‘<,> short pause’. Tag questions in the examples are italicized.

¹⁰ As there were only a few instances of informational tag questions, the relationships between polarity types and informational tag questions remain to be investigated.