The perception of native vs. non-native Danish speech: Bent and Bradlow’s matched interlanguage speech intelligibility benefit revisited

Размещено на http://allbest.ru

THE PERCEPTION OF NATIVE VS. NON-NATIVE DANISH SPEECH: BENT AND BRADLOW'S MATCHED INTERLANGUAGE SPEECH INTELLIGIBILITY BENEFIT REVISITED

Pawel Korpal

Adam Mickiewicz University, Poland

Mikolaj Sobkowiak

Adam Mickiewicz University, Poland

The main objective of the study was to test the applicability of Bent and Bradlow's matched interlanguage speech intelligibility benefit to the Danish-Polish language pair. We aimed to verify whether it was easier for Polish students of Danish to understand a Danish native speaker or a Polish speaker with a proficient command of Danish. Sixteen Polish students, divided into two groups of eight, listened to two recordings of two Danish texts: one recorded by a native speaker of Danish and the other one -- by a native speaker of Polish who is a graduate of Danish philology from a Polish university. Before the experiment, all of the recordings were evaluated in terms of traces of foreign accent using a 7-point Likert scale, the experts being native speakers of Danish. The evaluators assessed the Polish native speaker's pronunciation as proficient, but they identified certain segmental and suprasegmental features in his speech that are common indicators of a foreign accent in Danish. During the experiment, participants were asked to fill in each recording transcript with twenty missing words. The analysis of the results revealed that the participants scored higher when listening to the text recorded by the Polish speaker. Hence, the matched interlanguage speech intelligibility benefit was observed in a study using Polish as L1 (native language) and Danish as a foreign language. The study may provide a valuable insight into the question of non-native speech perception, foreign-accented speech and the veracity of the matched interlanguage speech intelligibility benefit for the Polish-Danish language pair.

Keywords: matched interlanguage speech intelligibility benefit; Danish speech; native and foreign-accented speech.

ВОСПРИЯТИЕ РОДНОЙ И НЕРОДНОЙ ДАТСКОЙ РЕЧИ: ГИПОТЕЗА БЕНТА И БРЕДЛОУ О ПРЕИМУЩЕСТВАХ ВЗАИМОПОНЯТНОЙ МЕЖЪЯЗЫКОВОЙ УСТНОЙ РЕЧИ

Павел Корпал

Университет им. Адама Мицкевича, Польша

Миколай Собковяк

Университет им. Адама Мицкевича, Польша

Основная цель исследования состояла в том, чтобы проверить гипотезу Бента и Бредлоу о преимуществах внятности межъязыковой устной речи по отношению к датско-польской языковой паре. Мы стремились проверить, легче ли польским студентам, изучающим датский язык, понять носителя датского языка или носителя польского языка, владеющего датским языком в совершенстве. Шестнадцать польских студентов-данистов, разделенные на две группы по восемь человек, прослушали две записи двух датских текстов: одну записал носитель датского языка, а другую -- носитель польского языка, выпускник датского отделения филологического факультета польского университета. Перед экспериментом все записи оценивались с точки зрения следов иностранного акцента по 7-балльной шкале Лайкерта, причем оценивающими экспертами были носители датского языка. Эксперты оценили произношение носителя польского языка как компетентное, но они выявили в его речи определенные сегментальные и супрасегментальные особенности, которые являются общими индикаторами иностранного акцента в датском языке. В ходе эксперимента участникам предлагалось заполнить каждый протокол записи пропущенными словами -- всего двадцать слов. Анализ результатов показал, что участники набрали больше баллов при прослушивании текста, записанного польским спикером. Таким образом, преимущество внятности межъязыковой устной речи наблюдалось в исследовании с использованием польского языка в качестве L1 (родного языка) и датского языка в качестве иностранного. Это исследование может дать ценное представление о восприятии неродной речи, речи с иностранным акцентом и о пользе взаимопонятной межъязыковой устной речи польско-датской языковой пары.

Ключевые слова: гипотеза о преимуществах взаимопонятной межъязыковой устной речи; устная речь датского языка; устная речь с родным и иностранным акцентом

interlanguage speech polish native danish speaker

1. INTRODUCTION

In a globalised world, where people learn foreign languages in order to be able to communicate with people all over the world, the question of the perception of foreign-accented speech appears to be of the crucial importance. The question of foreign accent has been widely discussed in phonetics and phonology [Flege, 1988; Bongaerts et al., 1997; Flege and Liu, 2001; Flege and Fletcher, 1992; Piske, MacKay and Flege, 2001; Weil, 2001; Ioup, 2008; Munro, 2008; Sebastian-Galles, 2008]. Researchers have tried do define the notion of foreign-accented speech and enumerate its features [Ioup, 2008; Munro, 2008]. At times, foreign-accented speech has been perceived as a deficiency. For instance, Greene and Wells claimed that “foreign accent, being of the nature of imperfect or defective speech, is the result of incorrect articulation and enunciation and is therefore classified, from our therapeutic viewpoint, as stammering speech” [Greene, Wells, 1927, p. 24, as quoted in Munro 2008, p. 193]. This notwithstanding, apart from being treated as a speech deviation, foreign accent may also be perceived as an inherent part of second language learning, i. e. as a feature of interlanguage not fully devoid of L1 interference. In addition to being a natural stage in the process of learning a language, foreign-accented speech may also facilitate speech comprehension, which has been shown by Bent and Bradlow (2003).

Bent and Bradlow [Bent, Bradlow, 2003, p. 1600] came up with the idea of the `matched interlanguage speech intelligibility benefit' according to which “a native language match between a non-native talker and a non-native listener facilitates speech intelligibility” [Bent, Bradlow, 2003, p.1606]. The researchers carried out an experiment in which native speakers of Chinese, Korean and English were asked to read aloud several English sentences. Then, their task was to recognise some words from the previously recorded sentences. It was observed that native English speakers were the most intelligible for native English listeners. However, when the listener was a non-native English language user, they understood non-native English speakers equally well as native ones, or even better. This phenomenon was referred to as the interlanguage speech intelligibility benefit (ISIB) [Bent, Bradlow, 2003, p. 1600]. Interestingly, “this interlanguage intelligibility benefit extended to the situation where the non-native talker and listeners came from different language backgrounds, giving rise to the mismatched interlanguage speech intelligibility benefit” [Bent, Bradlow, 2003, p. 1600].

Similar to Bent and Bradlow (2003), as early as in the 1970s, Smith and Rafiqzad (1979) noticed that foreign-accented speech could be more intelligible to non-native listeners than native speech. Nevertheless, no talker-listener native language match was identified in the study [Smith, Rafiqzad, 1979]. Elsewhere, Imai, Flege and Walley (2003) tested the intelligibility benefit in a word recognition task. The hypothesis that native Spanish listeners would recognise more English words produced with a Spanish accent than without accent was corroborated in the study. It was also confirmed that native English listeners recognised fewer English words produced with a Spanish accent than without traces of foreign accent [Imai, Flege, Walley, 2003, p. 846].

Bent and Bradlow's (2003) interlanguage speech intelligibility benefit was also tested in more recent research. Hayes-Harb et al. (2008) examined the intelligibility of both native and Mandarin-accented English utterances for native Mandarin and native English listeners. The authors found that when listening to Mandarin-accented English speech, native Mandarin participants outperformed native English participants in identifying traces of foreign accentedness in English words. Nevertheless, Mandarin-accented English speech was not considered to be more comprehensible than native English speech by native Mandarin participants (2008: 664). Hence, the ISIB for non-native listeners was observed, but not for talkers. Similar results in support of the ISIB for listeners were obtained by Xie and Fowler (2013). Elsewhere, Podlipsky, Simackova and Petraz (2016) compared the notion of intelligibility of foreign-accented speech with its credibility. The authors found evidence in support of the ISIB for both the matched and mismatched condition. However, Podlipsky, Simackova and Petraz found that “matched and mismatched non-native listeners tend to trust native statements more” [Podlipsky, Simackova, Petraz, 2016, p. 30]. Hence, even though foreign-accented speech may be more comprehensible to both matched and mismatched non-native listeners, the intelligibility benefit shall not be extended to include credibility ratings of native and non-native speech.

Importantly, in some other studies, no evidence of the ISIB was found [e. g. Munro, Derwing, Morton, 2006; Wang, Van Heuven, 2015]. This raises the question as to whether the intelligibility benefit may be modulated by a given language pair, the phonological systems of the languages involved, or the language proficiency of a non-native speaker. Thus, the present study was conducted to test the perception of native and non-native speech in the context of the Danish-Polish language pair.

2. AIM

The main purpose of the experiment was to investigate whether it was easier for Polish students of Danish to understand a Danish native speaker or a Polish speaker with a proficient command of Danish. In other words, the experimental study was designed to test the applicability of the matched interlanguage speech intelligibility benefit proposed by Bent and Bradlow (2003) to the Danish-Polish language pair. It was examined whether there was a statistically significant difference in comprehension of a Danish native speaker and a non-native speaker who has achieved a proficient command of Danish as a foreign language in the group of Polish students of Danish.

3. PARTICIPANTS

Participants recruited for the study were sixteen first-year students of Danish language and culture at the Faculty of Modern Languages and Literatures at Adam Mickiewicz University, Poznan. Four of them were male and twelve of them were female. Their age ranged from 20 to 26 years old (M=20.9; SD = 0.94). All of them had completed 8 months of intensive language training in Danish at the university (approximately 240 hours) prior to the experiment. None of the participants suffered from a hearing impairment that could compromise their ability to comprehend texts presented in the form of audio recordings.

4. MATERIALS

The materials used in the study comprised four recordings of two Danish texts (see Appendix). The recordings were of equal length and delivery rate, as well as lexical and syntactic complexity (LIX readability formula developed by Carl-Hugo Bjornsson -- Text 1 = 27; Text 2 = 29, indicating equal text difficulty). The LIX index uses both a word factor and a sentence factor, and takes into account the number of long words and average sentence length. Texts with a score not exceeding 30 are interpreted as very easy. The texts were prepared in such a way that they would be adjusted to the participants' command of Danish. Both texts were stories about people who met in Denmark and decided to get married. They were adapted from the original texts featured as extra reading materials for the Danish language coursebook, Det kommer! [Jeppesen, Maribo, 2008] к Two parallel versions of each text were prepared: one recorded by a native speaker of Danish and the other -- by a native speaker of Polish with a proficient command of Danish. The Polish speaker was a graduate of Danish as a foreign language from Adam Mickiewicz University, Poznan.

The length of each recording and the number of words in both texts are provided in Table 1.

Table 1. The length of recordings and the number of words

5. PROCEDURE

Before the experiment the participants were informed about the procedure of the experiment and asked to sign the informed consent form. In the course of the experiment, each participant listened to two recordings of speeches in Danish, i. e. one delivered by a native speaker of Danish, while the other one -- by a native speaker of Polish with a proficient command of Danish. Right before listening to each recording the participants were provided with a transcript, each time with twenty words missing. The participants were not allowed to read the text before listening to the audio recording. The words missing comprised mainly short words of high frequency. The participants were informed that they would listen to both recordings only once. Their task was to insert the missing words as they listened to the recording. Such a procedure was designed to investigate whether it was easier for the students to recognise the Danish words uttered by the native speaker of Danish or Polish. The experiment was conducted in two groups of eight students. In order to minimise the effect The texts are available through the publisher's website at https://ny.alfabetadig- ital.dk/wp-content/uploads/2017/06/Supplerende_l%C3%A6setekster_Det_kom- mer_2udg.pdf (p. 44 and 46). of confounding variables, the order of recordings was counterbalanced across the participants, i. e. one of the groups listened to the native speaker recording first and then the text recorded by the non-native speaker, while the other group listened to the recordings in the reverse order.

After the experiment, the authors collected and coded the scripts filled in by a given participant. A participant's result was based on the number of correctly recognised Danish words. For each correct answer one point was given. Thus, the maximum score for each recording was 20 points. IBM SPSS Statistics 21 software was used to perform a statistical analysis of the results.

It was hypothesised that the matched interlanguage speech intelligibility benefit [Bent, Bradlow, 2003] would apply to the Danish-Pol- ish language pair. Hence, the participants would score equally or even higher when listening to the text recorded by the Pole with a proficient command of Danish, compared to the native speaker of Danish.

6. NATIVE AND NON-NATIVE SPEAKER EVALUATION

Before the experiment was conducted, all four recordings were evaluated in terms of traces of foreign accent by native speakers of Danish. Text 1 recorded by the native speaker and Text 2 recorded by the non-native speaker were rated by three native speakers of Danish. Text 2 recorded by the native speaker and Text 1 recorded by the non-native speaker were evaluated by three other native speakers of Danish. The raters' task was to provide answers on a 7-point Likert scale indicating foreign accentedness, with 1 indicating highly accented speech and 7 indicating native-like speech. The raters were also asked to briefly justify their answers. They were not informed about the language background of either speaker. The raters' assessment is provided in Table 2.

With regard to the text recorded by the native speaker of Danish, the evaluators gave a unanimous answer (M=7; SD = 0). On the other hand, scores given to the native speaker of Polish ranged from 3 to 6 (M = 5; SD = 1.26). They assessed his pronunciation as proficient but enumerated certain features which might suggest that he is not a native speaker of Danish, i. e. incorrect word stress in a couple of examples, incorrect pronunciation of back vowels. However, the evaluators' general impression was that the speaker's accent is of very good quality with minor disfluencies which might give away his language background.

7. RESULTS

A statistical analysis was performed to test whether there was a difference in comprehension of native vs. non-native Danish speech, i. e. whether the participants would score higher when filling in the transcript of the text recorded by the non-native speaker of Danish. Out of the maximum of 20 points the participants scored on average 11.19 points for the text recorded by the native speaker and 13.13 points for the text recorded by the non-native speaker of Danish (i. e. the native speaker of Polish). A paired-samples t-test showed that that there was a statistically significant difference in the participants' scores for the native (M = 11.19; SD = 2.20) and the non-native Danish speech (M = 13.13; SD = 2.28); t(15) = -3.564, p = 0.003, which might point to the fact that it was easier for the participants to comprehend nonnative Danish speech, in comparison to the native speaker of Danish. In this way the authors' hypothesis was confirmed in the course of data analysis. Thus, evidence of Bent and Bradlow's (2003) ISIB was found in the present study. It appears that traces of foreign accent may indeed facilitate speech intelligibility when listeners share the native language of the speaker.

8. DISCUSSION AND FURTHER RESEARCH

The obtained results in support of the matched interlanguage speech intelligibility benefit proposed by Bent and Bradlow (2003) are in line with previous research [Imai, Flege, Walley, 2003; Hayes-Harb,Smith, Bent, Bradlow, 2008; Xie, Fowler, 2013; Podlipsky, Simackova, Petraz, 2016]. The results of the present study showed that the ISIB may also apply to the Danish-Polish language pair. On a local level, this might be explained by the fact that Polish speakers of Danish do not assimilate sounds as much as native speakers of Danish would do it. This, in turn, might facilitate comprehension of Danish speech. Even if a non-native speaker of Danish with a proficient command of the language makes a mistake, Polish students may not find it hard to understand what has been said because such mistakes often stem from the differences between the phonological systems of Danish and Polish and, hence, there is a great probability that they would make the same mistake. To give an example, Polish students of Danish may find it difficult to distinguish between Danish front vowels. The Danish language is believed to pose a challenge to learners as it has as many as ten front vowels [Gronnum 2007, p. 140], compared with Polish three front vowels [Nagorko, 1996, p. 34]. To summarise, features of a phonological system of one's L1 may potentially compromise the perception of sounds in a foreign language.

From the didactic point of view, the obtained results might suggest that understanding a native speaker of Danish poses a challenge to Polish students of Danish. As observed in the present study, the overall intelligibility score for native Danish speech was 2 points lower than for non-native Danish speech. This discrepancy might point to the importance of foreign language courses offered by native speakers. Listening to Danish native speech may help Polish students to effectively communicate with Danes in the natural language environment.

As for further research, the study may be extended to test Bent and Bradlow's mismatched interlanguage speech intelligibility benefit. In such a study, the perception of Danish used by a native speaker of a language other than Polish and Danish would be investigated. Furthermore, the effect of the listeners' L2 (Danish) proficiency on the emergence of the ISIB could be tested in a study with two experimental groups, i. e. first-year as opposed to third-year students of Danish. Similarly, the speaker's L2 proficiency could be tested as an independent variable in further research on the ISIB in the Polish-Danish language pair. In the study presented in this paper, the mean score of the Polish speaker was 5 out of 7 points. A native speaker of Polish with a near-native command of Danish and a mean score exceeding 6 points could be recruited for another study. Furthermore, data from more participants could be analysed to give the study more external validity and, thus, to obtain more reliable results. For the sake of group homogeneity, only first-year students of Danish were tested in the study. However, the experiment could be conducted in two or three consecutive years in order to collect a more representative sample of first-year students of Danish. Despite these limitations, nevertheless, the study might hopefully provide a valuable insight into the question of non-native speech perception, foreign-accented speech and the veracity of the matched interlanguage speech intelligibility benefit for the Polish-Danish language pair. Further research is, however, needed to understand how the ISIB may be modulated by the speaker's and listeners' L2 proficiency.

REFERENCES

Bent T., Bradlow A. R. The interlanguage speech intelligibility benefit. Acoustical Society of America 114 (3), 2003. P. 1600-1610.

Bongaerts T., Summeren Van C., Planken B., Schils E. Age and ultimate attainment in the pronunciation of a foreign language. Studies in Second Language Acquisition 19, 1997. P. 447-465.

Flege J. E. Factors affecting degree of perceived foreign accent in English sentences. Journal of the Acoustical Society of America 84, 1988. P. 70-79.

Flege J. E., Fletcher K. L. Talker and listener effects on degree of perceived foreign accent. Journal of the Acoustical Society of America 91, 1992. P. 370-389.

Flege J. E., Liu S. The effect of experience on adults' acquisition of a second language. Studies in Second Language Acquisition 23, 2001. P. 527-552.

Greene J. S., Wells E. J. The cause and cure of speech disorders: A textbook for students and teachers on stuttering, stammering and voice conditions. New York: The Macmillan Company, 1927. 458 p.

Grannum N. Rodgrod med flode -- En lille bog om dansk fonetik. Kobenhavn: Akademisk Forlag, 2007. 212 p.

Hayes-Harb R., Smith B. L., Bent T., Bradlow A. R. The interlanguage speech intelligibility benefit for native speakers of Mandarin: Production and perception of English wordfinal voicing contrasts. Journal of Phonetics 36 (4), 2008. P. 664-679.

Imai S., Flege J. E., Walley A. Spoken word recognition of accented and unaccented speech: Lexical factors affecting native and non-native listeners. Proceedings of the International Congress on Phonetic Science, 2003. P. 845-848.

Ioup G. Exploring the role of age in the acquisition of a second language phonology. Phonology and second language acquisition. Eds J. G. Hansen Edwards, M. L. Zam- pini. Amsterdam: John Benjamins Publishing Company, 2008. P. 41-62.

Jeppesen B., Maribo G. Det kommer!: Dansk som andetsprog pa mellemtrin.

Grundbog. K0benhavn: Alfabeta, 2008, 128 p.

Munro M. J. Foreign accent and speech intelligibility. Phonology and second language acquisition. J. G. Hansen Edwards, M. L. Zampini (eds). Amsterdam: John Benjamins Publishing Company, 2008. P. 193-218.

Munro M. J., Derwing T. M., Morton S. L. The mutual intelligibility of L2 speech. Studies in Second Language Acquisition 28 (1), 2006. P 111-131.

Nagorko A. Zarys gramatyki polskiej. Warszawa: Wydawnictwo Naukowe PWN, 1996. 235 p.

Piske T., MacKay I., Flege J. E. Factors affecting degree of foreign accent in an L2: A review. Journal of Phonetics 29, 2001. P. 191-215.

Podlipsky V. J., Simackova S., Petraz D. Is there an interlanguage speech credibility benefit? Topics in Linguistics 17 (1), 2016. P. 30-44.

Sebastian-Galles N. Cross-language speech perception. The handbook of speech perception. Eds D. B. Pisoni, R. E. Remez. Oxford: Blackwell Publishing, 2008. P. 546-566.

Smith L. E., Rafiqzad K. English for cross-cultural communication: The question of intelligibility. TESOL Quarterly 13, 1979. P. 371-380.

Wang H., Heuven Van V. J. The Interlanguage Speech Intelligibility Benefit as bias toward native-language phonology. i-Perception 6 (6), 2015. P. 1-13.

Weil S. A. Foreign accented speech: Adaptation and generalization. 2001. Available at: http://lpl.psy.ohio-state.edu/documents/swthesis-21_000.pdf (accessed: 15.08.2017).

Xie X., Fowler C. A. Listening with a foreign-accent: The interlanguage speech intelligibility benefit in Mandarin speakers of English. Journal of Phonetics 41 (5), 2013. P. 369-378.

APPENDIX

Text 1