Assessing knowledge of English intonation patterns by L2 speakers Peggy Mok 1 , Yanjun Yin 1 , Jane Setter 2 , Noor Mat Nayan 2 1 The Chinese University of Hong Kong 2 University of Reading peggymok@cuhk.edu.hk, j.e.setter@reading.ac.uk Abstract English intonation can be difficult for L2 speakers to learn, particularly for those whose L1 intonation system works differently from English. This study investigates whether Hong Kong English (HKE) speakers whose L1 is Cantonese have knowledge of the appropriate English intonation patterns in specific contexts. Results from an intonation pattern selection tasks indicate that HKE speakers (n = 40) performed worse than British English speakers (n = 25) in general. For some sentence types, selection patterns of HKE speakers and native English speakers were quite similar, while HKE speakers had particular difficulty with tag questions. The lack of equivalent structures in L1 may explain their difficulties. Interestingly, native English speakers also showed much variation in their intonation selection for some sentence types. The results suggest that HKE speakers have partial knowledge of English intonation patterns. The lack of sufficient knowledge, in addition to L1 influence, can explain the differences between native and L2 English intonation choices. Index Terms: intonation, English, L2 acquisition, Cantonese 1. Introduction Intonation, acquired very early by first language (L1) speakers [1], is thought to be one of the most difficult aspects of a second language (L2) to learn. Chun [2] claimed that it is ‘seemingly impossible’ to learn English intonation, and Taylor [3] claimed that teachers believe it is ‘not teachable, and not learnable either’. While there are published works which indicate intonation patterns for English, e.g., on British English, [4, 5, 6], it can be difficult to pin down the meaning of certain contours in use. One only has to compare the descriptions of the ten tunes given in [5] to realise it is a rather complex area. There is ample evidence that the intonation patterns of L2 speakers differ substantially from those of native speakers. As intonation carries different functions (e.g., attitudinal, grammatical, discoursal and focusing), and in that intonation is highly contextual, it can be particularly difficult for L2 speakers to master. Previous studies on L2 English intonation usually focus on production differences or meaning interpretation by L2 speakers [e.g., 7, 8]. However, a fundamental question remains largely unanswered: do L2 speakers actually know what the appropriate intonation patterns should be in particular contexts? Are the different intonation patterns produced by L2 speakers simply due to L1 influence, or does the lack of understanding of appropriate English patterns also have a role to play? This study investigates Hong Kong English (HKE) speaking listeners’ ability to select a suitable intonation pattern in a given narrative context. HKE speakers were chosen for this study for several reasons. First, HKE is an emergent variety of English [9, 10], but not much research has been done on the intonation of HKE. Second, speakers of HKE start to learn English from an early age, but the curriculum has little emphasis on English prosody. There are native English teachers from, e.g., Britain, the United States and Australia in many schools and kindergartens, so some learners have exposure to different English intonation patterns in actual use, but little formal instruction means they may not develop knowledge of how they are used in various contexts. Third, the intonation system works very differently in the first language (Cantonese) of HKE speakers. There is intonation in Cantonese, but intonational pitch variation is constrained to preserve lexical tones. Instead, attitudinal and discoursal meanings are conveyed using a rich inventory of sentence- final particles [11]. It is, therefore, of interest to see if Cantonese speakers notice which English intonation patterns are appropriate in different sentential contexts. Previous studies have observed that there is an over- generalization of nuclear tone in HKE. Bolton and Kwok [12] reported that HKE speakers would apply the falling tone to all statements. However, Setter et al. [10] observed that there was an increase in the use of the rising tone in statements (uptalk). Bolton and Kwok [12] also suggested that all questions in HKE have a rising tone, regardless of question types. Cheng et al. [13] and Lin [14] instead argued that while most questions have a rising tone in HKE, wh-questions are produced with a falling tone. The frequency of use of different tone types in HKE is different from native English speakers as well. In British English, the falling tone is the most common tone (50%), followed by Rise and Fall-Rise (40%) [4]. [15] also reported 75% of Fall, 13% of Rise and 2% of Fall-Rise nuclear tones in British TV commercials. In HKE, both [10, 13] found that the Level tone was the most common tone instead (49.3% and 43.3% respectively). [13] reported that the second and third most common tones are Fall [37.1%] and Rise [5.7%], whereas [10] found that the second and third most common tones are Rise (24.4%) and Fall (23.0%) with very similar occurrence. Although the Rise-Fall tone is also uncommon in British English (3% in [15]), it is hardly ever used in HKE: only 0.016% and 0.59% as reported by [13] and [10] respectively. Setter et al. [10] noted that, unlike British English, the use of Rise-Fall in HKE has nothing to do with indignation, sarcasm, being surprised or impressed, so it may mean that HKE speakers simply have no knowledge of the functions of such a tone at all. The lack of knowledge of intonation functions may also partly explain other differences between the HKE and native speakers mentioned above, but so far, few studies have Speech Prosody 2016, May 31 - June 3, 2016, Boston, MA, USA 543