國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
If a phoneme fits the meaning of the word in which it occurs, it reinforces the meaning, but if it does not fit the meaning, then their occurrence in that context is of the common arbitrary type, and no question of correlation arises. (p.46) As for the source of this type of sound symbolism, Bloomfield (1895) proposed a perspective that as long as a word is semantically expressive, it may establish a link between its meaning and any of its sounds, and then “send forth this sound ( or sounds) upon predatory expeditions into domains where the sound is at first a stranger and parasite” (p.409-410). Thus, the meaning of the word is arbitrarily decided by any sound-element contained in it. In other words, at first sound symbolism happened accidentally in a language; hence it is arbitrary. After the sound is associated with a specific meaning and accepted by the language users, new words are created by this principle — using the same or similar sound to coin the new word to meet the need of that specific meaning.
Research on Sound Symbolism
Although the concept of sound symbolism is not new, research on this topic has
been intermittent. Research on sound symbolism in English can be traced back to Sapir (1929). As mentioned earlier, he invented word pairs with C-V-C syllable structure (e.g. mal and mil) and offered the participants arbitrarily selected meanings to match with these words. For instance, the participants were asked to identify which word they thought represented a small or large table. His findings showed that the participants felt the vowel /a/ was suggestive of greater magnitude than the vowel /i/for the participants in his experiment more frequently matched the invented word mil to the meaning of the small table and mal to the large table. Based on a series of psycholinguistic experiments, Sapir found a strong relationship between English vowels and word meanings that vowels and their associations with magnitude
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
correspond with the natural parameters of the vocal tract during the articulation. In other words, the fact that the vowel /a/ is uttered with an open and expanded vocal cavity appears to designate greater magnitude than the vowel /i/, which is pronounced with a relatively smaller and constricted vocal tract. This classical experiment was replicated by the follow-up research (Newman, 1933; Tarte & Barrit, 1971; Ultan, 1978) and the findings conformed to Sapir’s. Ultan (1978) examined the diminutive marking in 136 languages in his research and found that smallness was signified by a high front vowel like /i/. Thus, Ultan extended Sapir’s finding from English to numerous other languages. Other research findings (Ciccotosto, 1991; Hinton et al., 1994; Mok, 2001) also indicated that the phenomenon of sound symbolism is universal.
Considerable amount of research (Hinton et al., 1994; Tarte & Barrit, 1971;
Ultan, 1978) has confirmed the existence of the phenomenon. In short, cross-linguistic as well as language specific research on sound symbolism has presented an
accumulation of evidence, indicating that sound symbolism is more than just an exception to the rule that the connection between sound and meaning is arbitrary (Hinton et al., 1994.). As there are no exhaustive lists of English sound symbols so far, it is hard to determine the extent to which sound symbolism will affect word learning in English. Despite this, Parault and Schwanenflugel (2006) estimated the effect of sound symbolism in English. Their findings showed that 24% of English words examined in their research were sound symbolic in nature and they concluded that sound symbolism could be an important factor in word learning. Resulting from this conclusion, Parault and Parkinson (2008) suggested that although these sound symbols will not always lead to the correct definition for an unknown word, “when faced with innumerable possible meanings for a word, a strategy that works even just 24% of the time is better than no strategy at all”(p.669).
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Sound Symbolism and English Vocabulary Teaching
In literature, related research concerning sound symbolism has been focused on two dimensions: the exploration of sound symbolism to prove the psychological reality and the search for universality of the phenomenon among the world languages.
Numerous experiments (Ciccotosto, 1991; Hinton et al., 1994; Sapir, 1929; Tarte &
Barrit, 1971; Ultan, 1978) were conducted and have proved that sound symbolism is a universal phenomenon.
Not until recent years, was the theory of sound symbolism incorporated into vocabulary teaching from the viewpoint of phonetics. Some research has applied sound symbolism to vocabulary learning (Berko-Gleason, 2005; Yashida, 2003), and suggested that it is easier for children to learn sound-symbolic words because the connection between sound and meaning is obvious, as in the example of
onomatopoeias. As a result, sound symbolism plays a role in word learning.
Nonetheless, within the EFL framework, the present study only focuses on literature related to English sound symbolism. Parault and Schwanenflugel’s (2006) research compared the correctness of word meanings English-speaking adults inferred from obsolete English sound symbolic and non-sound symbolic words. Their finding showed that the participants were 85% better at generating correct meanings of sound symbolic versus non-sound symbolic words. Parault’s (2006) follow-up study
confirmed the finding and indicated that sound symbols were useful for learning new words. In a recent study, Parault and Parkinson (2008) concluded that sound
symbolism is a word property which influences the learning of unknown words. To date, however, few studies have combined sound symbolism theory with English classroom vocabulary teaching. Chen’s (2000) research discussed the relationship between sound and meaning of English words from the perspective of articulatory phonetics and acoustic phonetics. Afterwards, Mo (2005) systematically introduced
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the relationship between English sounds and vocabulary teaching with the theory of sound symbolism. These studies are the initial step toward the combination of theory and practice.
In addition to the theories of sound symbolism, several researchers further provided some useful pedagogical implications from sound symbolism in literature.
Mo (2005) believes that sound symbolism and sound switching are predictable. If an English teacher understands the phenomena and the related theories, she will be able to facilitate students’ mental activities during the vocabulary instruction. Other pedagogical implications were described as follows. (1) Sound symbolism instruction reinforces phonics instruction; (2) sound symbolism instruction corresponds to
cognitive learning principles; (3) sound symbolism assists the training of phonological awareness; and (4) sound symbolism instruction is complementary to other
vocabulary teaching approaches (Chen, 2000). Furthermore, Parault and Parkinson (2008) argued that the contribution to education sound symbolism made was to help students narrow down the possible meanings for unknown words.
Previous studies on sound symbolism showed the connection between the theory and the application to vocabulary instruction (See Table 2.1). However, there is an absence of empirical research to confirm the effectiveness of incorporating sound symbolism into English vocabulary teaching. The present study attempts to fill this gap by designing an experimental research to investigate the effect of the instruction of sound symbolism in the English teaching classroom and the target words designed in this present study included the last two categories of sound symbolism reviewed in this chapter, i.e. synesthetic and conventional sound symbolism.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 2.1 Related Studies on Sound Symbolism
Language Focus Previous studies
English Others Existence Vocabulary Teaching
Empirical Study
Sapir (1929); Tarte &
Barrit (1971)
Ultan (1978);
Ciccotosto (1991);
Hinton et al. (1994)
Chen (2000);
Mo (2005); Parault &
Parkinson (2008)
Mok (2001)
Yashida (2003)
Parault &
Schwanenflugel (2006)
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
CHAPTER THREE METHODOLOGY
This empirical study aims to investigate the effectiveness of sound symbolism instruction in vocabulary teaching. This chapter describes the methodology of the research, including participants of the present research, instruments employed by the researcher/teacher5, vocabulary instruction, procedure throughout the study and data analysis of the present research.
Participants
The participants for the study consisted of two ninth-grade classes of 72 students at a public junior high school in Hsin-chu City. They were chosen because they could be considered homogeneous for two reasons. First, all ninth-graders at this school had been placed in a normal s-type distribution on the basis of their performance in an entrance IQ test. Second, all of the participants were of similar English proficiency.
The participants were given the 2009 2nd English Test of the Basic Competence Test (EBCT) to ensure they were homogeneous in English proficiency. Their scores on EBCT were collected and compared through the independent samples t-test. As shown in Table 3.1, the mean score of Class A was 37.25 (N=36) with a standard deviation of 19.53, and that of Class B was 38.03 (N=36) with a standard deviation of 20.686. According to the statistics in Table 3.2, these two classes passed the Levene’s test (F=0.169, p=0.683>0.05), indicating that the two classes were homogeneous. The t-test for equality of means also showed that there was no significant difference between the two classes (t=0.164, p=0.87>0.05).
5 In this study, the researcher is also the teacher of the experimental group and the control group.
6 The total score of EBCT is 80 points.
‧
Table 3.1 Statistics of Participants’ English Proficiency Test Scores
Group Number Mean S.D.
Class A 36 37.25 19.53
Class B 36 38.03 20.68
Table 3.2 Independent Samples t-test on Participants’ English Proficiency Test Levene’s Test for
Equality of Variances
t-test for Equality of Means
With the homogeneity, one of the classes was randomly assigned as the
experimental group, and the other as the control group. In addition to the homogeneity, these two classes were selected for the reason that they were the researcher’s own classes, so the experimental research was able to be conveniently conducted in practice. In the experimental group, the participants were stratified into two general proficiency groups based on the EBCT scores, which were labeled as High group and Low group. The cut-off point was the mean of all the participants’ EBCT scores (M=37.25). The High group included students with scores of 38 and above and the Low group with EBCT scores of 37.25 and below. The numbers of the participants in High and Low proficiency groups were 16 and 20 respectively.
Instruments
The instruments designed in this study included the 2009 2nd EBCT (See Appendix A), a test for word selection (See Appendix B), a vocabulary pre- and
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
post-test (See Appendix C), the modified think-aloud protocols sheet (See Appendix D), and interviews (See Appendix E).