Attitudes to speech varieties of English

The scales which were intended to measure various aspects of learners‟ attitudes towards speech varieties also revealed regularities along the lines of native and non-native speaker accents. The following table shows the differences in mean scores of the constructs related to stereotypes towards the different speakers in the audio samples. As it was mentioned in Section 3.2.2, the components included Likert scales which measured the degree to which learners attributed certain personal qualities to the speakers based on his English pronunciation such as being reliable, cool, kind, intelligent, educated or working in a high position. The data showed that the native (Inner Circle) accents triggered more positive stereotypes from the learners, while the speakers of non-native (non-Inner Circle) varieties of English were judged less favourably. The most positive stereotypes were attributed to the speaker with the American accent and the least favourable to the French and Egyptian Arabic speaker. The difference between the latter two was not significant statistically, which implies that similar stereotypes were evoked by learners by the two unfamiliar non-Inner Circle .speech varieties of English.

125

Table 10. The differences in the mean scores of the scales related to stereotypes towards the British, American, Arabic and French speaker

RP American Arabic French

RP x -4.138** 10.604** 11.534**

American x 13.561** 14.549**

Arabic x .349

(sig= .728)

French x

** differences significant at the <0.0001 level (2-tailed)

The Bogardus scales, which measured social distance or proximity towards the speakers, also exhibited as similar pattern as shown in the following table showing the T values of a paired sample T-test showing the differences between the speakers. The two unfamiliar non-Inner Circle varieties, Arabic and French English showed the smallest albeit still significant difference between the mean scores, while American and RP English were also relatively close to each other. It can also be observed in the data that while learners showed a clear preference to engage in various activities with the two native speakers, they showed an inclination towards the American English (mean=4.16) speaker over the RP speaker (mean=3.78). The mean score for the American Speaker (4.16) would fall on the “strongly agree”, while the score of the RP speaker on the “somewhat agree” section of the Likert scale.

126

Table 11. The degree of difference between the mean scores of the social proximity related to the speakers of the speech varieties shown by the T-value

RP GA Egyptian Arabic French

RP x

GA -4.13** x

Egyptian Arabic 20.63** 20.71** x

French 20.39** 23.07** 2.05** x

** differences are significant at the <0.0001 level (2-tailed)

There are also significant correlations between the four Bogardus scales as shown in the table below, suggesting that there was an interaction between the underlying factors that influence the learners‟ judgements of the different speakers. According to this data, those learners who showed more acceptance towards the RP speaker were also more accepting towards the American and French speaker. A weaker and less significant relationship of the same type can be seen between the American and the French speaker, while there is a stronger and more significant connection between the Arab and the French speaker.

Table 12. Pearson‟s Correlations between the Bogardus scales

RP GA Egyptian Arabic French

RP x .414** .359** .065

American x .163 .208*

Egyptian Arabic x .563**

French x

*Correlation is significant at the 0.01 level (2-tailed)

**Correlation is significant at the 0.05 level (2-tailed)

127

It can be assumed that the background of these relationships is that there was an interplay between the underlying factors which influenced the learners‟ judgements about the speakers.

Learners might have categorized different speakers in a similar way at a certain level, for instance, the two native speakers who might have represent Anglo-Saxon culture to learners, or they might have felt the same solidarity towards the two non-native speakers. On the other hand, the background factors may have been related to the learners‟ personality traits, such as open-mindedness, or their geo-political views on the countries of the speakers, which could explain why they related similarly to different speakers. Nevertheless, based on this data, it can only be speculated what these underlying factors might be exactly.

There seemed to be a relationship between the learners‟ willingness to engage with the speakers and the stereotypes they had towards them based on the speaekers‟ accents, as it can be observed in the correlations between the Bogardus scales and the scales which measured the learners‟ judgements about the speakers‟ personal qualities and socio-educational background.

Table 13. Pearson‟s Correlations between the Bogardus scales related to each speech variety and the corresponding scale on the judgements of the speakers

RP American Arabic French

RP .626** .302* .170 .166

GA .350** .617** .121 .225*

Egyptian Arabic .233* .141 .560** .376**

French .097 .173 .298** .571**

**Correlation is significant at the 0.01 level (2-tailed)

*Correlation is significant at the 0.05 level (2-tailed)

128

This data shows that the learners‟ willingness to engage with certain speakers can be predicted to a certain extent from the stereotypical personality traits and socio-educational features that learners associate with speakers of English based on their pronunciation. The data also suggests that these relationships are not entirely unique to each speaker, but judgements of the accent of one speaker can also be connected to willingness to engage with another type of speaker with a different accent. This can be seen from the correlations between the Bogardus scale of the American speaker and the personal judgement scale of the RP speaker or between the Bogardus scale of the French speaker and the personality scale of the Arab speaker. Similarly to the correlations between the Bogardus scales, there can only be assumptions about the exact relationship between the Bogardus scales and the personality scales related to different speech varieties.

Last but not least, there was a significant correlation between the scales related to the acceptance of different pronunciations in English and in Hungarian (.694** sig<0.0001). This indicated that those learners who were comfortable with hearing different accents in their mother tongue were also likely to be more open and accepting towards different speech varieties of English. The underlying reason for this correlation might have been general beliefs about languages as such, whether learners believed that speakers of a language should conform to certain standards or languages are diverse and variable by nature.

In conclusion, it can be said that learners showed characteristic attitudes towards the pronunciation speakers of different English speech varieties mainly along the lines of whether the speaker had an ENL accent or a non-ENL pronunciation which was unfamiliar to learners.

This seemed to what stereotypes learners attached to the speaker based on his accent, which showed a relationship with learners‟ symbolic social distance from the speaker. This appeared to

129

corroborate the findings of Study One according to which learners had more favourable attitudes towards standard speech varieties of ENL speakers of English from the Inner Circle, contrary to speech varieties of non ENL speakers from the Outer and the Expanding circle. These results are discussed further in Sections 4.4 and 4.5 of the following chapter.

130

4 Discussion