처리유창성에 대한 사회심리학적 실험 논문

Observational study on the relationship between accented speech and prosocial behavior

Abstract

Accent is one of the major factors that influences the impression of a speech. To see whether there is a meaningful relationship between lectures given in a non-Standard American English (nSAE) accent and applause at the end of each lectures, randomly-assigned 112 lecturers at the University of Toronto were observed. As predicted, lectures given in the nSAE accent did not in general receive applause regardless of the class size, class duration, etc. whereas those given in the SAE accent either received or not received applause, suggesting that other independent variables are at work in this case. While this result indirectly supports previous studies that confirmed a positive relationship between accented speech and its negative evaluation, it further illustrates that a speech given in the nSAE accent may subtly influence the social behavior of the perceiver toward the speech in question.

Keywords: perceptual fluency, nSAE accent, SAE accent, prosocial behaviour

Introduction

In what way does accented public speech influence the social behavior of the perceiver? Perceptual fluency is the level of ease in processing certain stimuli or information. Low perceptual fluency, for instance, implies that a stimulus in question is relatively difficult to process and thus requires more cognitive resources whereas high perceptual fluency refers to the opposite. A number of social psychological and psychophysical studies have shown that speech with high perceptual fluency is not only judged more positively and truthfully (Reber, Winkielman, & Schwarz, 1998; Reber & Schwarz, 1999), but also judged as more pleasant and familiar (Zajonc, 1968). It thus comes as no surprise that a speech given in the nSAE accent is more difficult to process compared to that given in the SAE accent (Munro & Derwing, 1995), suggesting that a public speech in a foreign accent would likely be judged as less positive and less credible (Lev-Ari & Keysar, 2010). Indeed, as Shuck (2006) points out in his study, people often rationalize their negative evaluations and biased attitudes toward those with the nSAE accent precisely because their foreign accents decrease the perceptual fluency of the message being spoken.

Given that comprehending the nSAE-accented speech demands more cognitive resources and is more likely to receive negative evaluation, my study aimed to explore the relationship between the low perceptual fluency caused by the nSAE accent of lecturers and its influence on the social behaviour (applause) of the perceiver—in this case the undergraduate students at the University of Toronto. Similar to the aforementioned study on the perceptual fluency of accented speech and its affective evaluation, namely that of Lev-Ari and Keysar (2010) in which a series of speeches in the nSAE accent was rated as less credible and more difficult to understand by the participants, I investigated whether nSAE-accented speech is correlated to its negative evaluation. The crucial difference between my study and this research lies in the fact that I observed the actual social behavior of the participants with regard to the speech in question. Thus in this study I am presupposing the preexisting experimental results that nSAE-accented speech is evaluated more negatively most of the time due to its low perceptual fluency and thereupon speculating that, because this is so, such negative evaluations of the nSAE-accented speech would also impact upon the likelihood of the prosocial behavior of the perceiver. On the assumption that applause is a prosocial behavior that functions as a sign of approval, I hypothesize that university lectures given in the nSAE accent would be less likely to receive applause while those given in the SAE accent would be more likely to receive applause.

Method

A.Design and Participants

One hundred twelve lectures at the University of Toronto were randomly selected from the faculty of Arts and Science by using the official University of Toronto Course Finder which contains all of the available list of undergraduate courses at the University of Toronto. A sample selection was done by first dividing all courses into three subgroups: Humanities (1416), Science (897), and Social science (873). Since I had to attend each lecture to collect data, each class had to be large enough to minimize my interference; yet a cutoff for the minimum class size had to be small enough so that I could collect qualitatively meaningful data. Thus small classes, arbitrarily defined as classes with n<80 students, were excluded and only those with course descriptions that contain both specific classroom locations and lecture schedules were considered in the first round of the selection process. Consequently, (1) all 300+ level courses, (2) all Research Opportunity programs, (3) all of The One Programs (small classes designed for freshmen), and (4) Fall classes were excluded, as observation took place on the winter semester. Then every 3rd course from the remaining list, which was alphabetized, was selected and every 4th course from the remaining list was selected in reverse alphabetic order, from which I additionally excluded those that overlap with my lecture schedule. A total of 112 lectures (48 Humanities, 39 Science, 25 Social science class) were selected as a result.

The participants that I observed were 112 lecturers at the University of Toronto and the attendees of their lectures that I randomly selected. Although the attendees of each lecture had a distinctive sex ratio and so did the age demographic, there were approximately equivalent numbers of males and females, and the age demographic of the participants ranged from 18 to 24 in the case of the attendees. On the other hand, lecturers were mostly males and their age demographic ranged from early 30s to late 50s. Participants varied greatly in their ethnic backgrounds.

B.Procedure

Since all lectures had their own schedules and many of them overlapped with each other, observation continued for 4 full weeks: from January 30th to March 3rd, 2017, excluding Reading Week. I attended about 30 lectures per week, collecting 4~5 observational data per day. Some lectures were given in the same classroom but they were mostly given in distinctive lecture halls. All lectures were given on weekdays and class duration was at least 50 minutes and at most 3 hours. Class size ranged from 80 to 1500 students and average class size was 240 students. To maximize the efficiency of data collection, I attended each lecture approximately 15~20 minutes before the termination and observed (1) the presence of the nSAE accent of each lecturer which was determined based on how close his/her accent was to the SAE accent and (2) whether attendees applauded at the end. To minimize the interruption, I always used backdoors or side doors when possible. I recorded the observations via my Pocket Electronic Dictionary (PED) with a word memo function, pretending that I am texting someone else.

The first categorical variable, the presence of the nSAE accent in each lecture, served as a stimulus with either low/high perceptual fluency. Because the SAE accent is admittedly a loose concept that refers to the North American accent without explicit pronunciational indicator of speaker’s ethnicity and regional origin, the lecturer’s accent was marked as the nSAE accent if and only if it was either ethnically/regionally identifiable or it was, at least, not the SAE accent. Using this definition, the SAE accent was operationalized as an accent that is ethnically/regionally unidentifiable and that is not the nSAE accent. Each lecture was therefore categorized as either SAE-accented or nSAE-accented.

The second categorical variable, the presence of applause, was the dependent variable of my observational study which was categorized in terms of number of people applauding. Although not applauding does not necessarily mean that the lecture in question is per se being negatively evaluated, given that absence of applause is neither a sign of approval nor disapproval, this is at least a more negative condition compared to the applauding condition. For this reason, the applauding condition was operationalized as the positive condition in which noise is created by one or more people clapping their hands as a sign of approval, and absence of applause was operationalized as a negative condition in which no one applauds at the end.

Accordingly, observations were classified in one of the four categories: (1) SAE-accented lecture and applauding, (2) SAE-accented lecture and no applauding, (3) nSAE–accented lecture and applauding, and finally (4) nSAE-accented lecture and no applauding.

Results

Table 1: Accented speech versus the presence of applause (Total=112)

I initially hypothesized that the nSAE–accented lectures would be less likely to receive applause while the SAE-accented lectures would be more likely to receive applause. A chi-square test of independence was performed to examine the relationship between the presence of the nSAE accent and no applause. The relationship between these variables was significant, χ²(1, N=112) = 11.849, p = .001. As indicated in Table 1, perceivers were less likely to applaud at the end of the nSAE-accented lecture. The data analysis therefore supports my hypothesis.

Discussion

Approximately 56% of the lecturer (42 out of 75) with the SAE accent received applause while about 22% of the lecturer (8 out of 37) with the nSAE accent received applause. Thus obtained results support my hypothesis since the attendees of each lecture were less likely to applaud at the end of the nSAE-accented lecture compared to those of the SAE-accented lecture. Given that applauding is a prosocial behavior and not applauding is, as previously assumed, not a prosocial behavior, this result is consistent with the previous research in that speech tends to be evaluated more negatively if it is given in the nSAE accent (Winkielman & Cacioppo, 2001; Lev-Ari & Keysar, 2010).

While previous research was mostly focused on perceiver’s subjective evaluation of the accented speech (Reber, Winkielman, & Schwarz, 2004; Reber & Schwarz, 1999), the key difference of my observational study from the hitherto accumulated perceptual fluency literature is that my study concerned the impact of the accented speech upon the social behavior of the perceiver. Thus, the results that I have found may not only be an additional support for the positive correlation between the accented speech and the perceiver’s subjective evaluation, but also for the positive relationship between the accented speech and the perceiver's affective social behaviour.

There are, of course, some limitations in this study that must be addressed. Because applause begins almost always with a few people followed by a large-scale conformity, I could not rule out the possibility that applause, at least in some cases, is merely a consequence of conformity rather than a direct result of the accent of a particular lecturer. Although I attempted to solve this problem by recognizing that, in the applauding condition, at least one person decided to clap anyways while no one decided to clap in the no applauding condition, I could not completely solve this problem since this approach merely delineates how a few random attendees evaluate the lecture in question, not how the attendees in general evaluate the lecture in question.

Other confounding variables that could have contaminated my result include class size, class duration, attractiveness of the lecturer, difficulty of lecture contents, time of day and many more. However, the most serious confounding variable seems to be what the lecturer says at the end of the class. As I attend 112 lectures, I repeatedly encountered the lecturers who continually commented on various topics (ex. homework, assignments, topics for the next class) versus those who simply announced that the class is over. The shorter the ending comment, the more likely it seemed that the attendees clap. While my observational result explicitly demonstrates that there is in fact a positive relationship between accented speech and applause, this potential third variable may explain the fact that there were exceptions in my result, as can be observed from the Table 1─in some cases the attendees decided to applaud even in the nSAE-accented lectures and in other cases the attendees decided not to applaud even in the SAE-accented lectures.

This illustrates that low perceptual fluency of a speech caused by the nSAE accent may not be the only factor that hinders people from applauding. Hence more rigorous correlational research is needed to explicate exactly to what degree a speech given in the nSAE accent, that is, a speech with low perceptual fluency, influences the perceiver’s social behavior toward the speech in question. Because there were a number of potential confounds in my study, verifying the impact of one variable at a time is crucial, which can only be done through experimental research.

In conclusion, since nSAE-accented speech is, either consciously noticeable or not, relatively difficult to comprehend, it decreases the overall perceptual fluency which not only can induce the perceiver to devalue it (Munro & Derwing, 1995; Whittlesea, Jacoby, & Girard, 1990), but can also negatively influence the attitude and social behavior of the perceiver toward it (Shuck, 2006). Considering today’s multicultural and polyethnic society, the importance of the implications of such experimental results cannot be overstated, for it can direct our attitude toward foreigners with the nSAE accent from minor impact like our evaluations of a particular communicator all the way to our decisions with regard to the presidential/provincial elections─not to mention that it may also promote rationalization of and even intensification of antisocial attitudes toward speakers with the nSAE accent. With this in mind, future research topics concerning perceptual fluency should not be delimited to affective judgements on accented speech, but also deal with an affective social behavior as a reaction to the accented speech in question.

References

Lev-Ari, S., & Keysar, B. (2010). Why don’t we believe non-native speakers? The influence of accent on credibility. Journal of Experimental Social Psychology, 46, 1093–1096.

Munro, M. J., & Derwing, T. M. (1995). Processing time, accent, and comprehensibility in the perception of native and foreign-accented speech. Language and Speech, 38, 289−306.

Reber, R., & Schwarz, N. (1999). Effects of perceptual fluency on judgments of truth. Consciousness and Cognition, 8, 338–342.

Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8, 364–382.

Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective judgments. Psychological Science, 9, 45–48.

Shuck, G. (2006). Racializing the nonnative English speaker. Journal of Language, Identity, and Education, 5, 259–276.

Whittlesea, B.W. A., Jacoby, L. L., & Girard, K. (1990). Illusions of immediate memory: Evidence of an attributional basis for feelings of familiarity and perceptual quality. Journal of Memory and Language, 29, 716–732.

Winkielman, P., & Cacioppo, J. (2001). Mind at ease puts a smile on the face: psychophysiological evidence that processing facilitation elicits positive affect. Journal of personality and social psychology, 81(6), 989.

Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology. Monograph Supplement, 9, 1−27.