穨Review. Pdf

participants’ speaking ability by asking them to describe pictures, there were 
limitations of this type of oral test. According to Chiu (1997), describing pictures 
could test certain organizations such as description or narration, or certain verb tenses, 
but “in general, this [describing pictures] barely reflects the testees’ communicative 
competence (p.73).” McNamara (1996) also implied that “the weakness of current 
models [of speaking test] is that they focus too much on the individual candidate 
rather than the candidate in interaction (p. 86).” For the reasons stated above, two 
interaction-based oral tasks were designed to investigate the EFL learners’ verbal and 
non-verbal features of communicative competence. The design of the oral task in 
this study complied with Browns’ (2001) notion of best tests of oral proficiency, 
which included “live performance (as opposed to taped), a careful specification of 
tasks to be accomplished during the test, and a scoring rubric that was truly 
descriptive of ability (p.395).”
According to Savignon (1983), communicative competence was relative and 
context specific. That is, the definition of communicative competence would vary 
according to the language proficiency of the speakers involved. With this notion in 
mind, the model of communicative competence proposed by Canale’s (1983a) was 

---

49
adapted and altered to fit the language proficiency of the participants involved in this 
study. The sociolinguistic competence included in Canale’s (1983a) model was 
replaced with the non-verbal features of communicative competence. Rationale of 
such replacement would be shortly discussed in later paragraphs.
2.8.2.1 Verbal Features of Communicative Competence 
The verbal aspects of communicative competence under investigation in this 
study then consisted of (1) linguistic competence, (2) discourse competence, and (3) 
strategic competence. 
The linguistic competence, or the grammatical competence, referred to the 
knowledge of lexical items and of rules of morphology, syntax, sentence-grammar 
semantics, and phonology (Canale, 1983a). In assessing the students’ linguistic 
competence through the oral performance, the rating scale was of great importance.
As Chiu (1997) noted, the use of an appropriate rating scale and sufficient training for 
raters appeared to be the key points in reducing the rater effect when assessing oral 
performance. In order to measure more integrated language use, the scoring rubric 
developed by Weir (1990) was adopted as the grading criteria, which included (1) 
appropriateness, (2) grammatical accuracy, (3) intelligibility, (4) fluency, and (5) the 
adequacy of vocabulary for purpose (Weir, 1990). Such grading criteria could 
possibly achieve the maximal balance between grammaticality and appropriateness, as 
well as accuracy and fluency, as discussed in section 2.1.2.1 and 2.1.2.2. 
As for the discourse competence, the most common definition was the ability to 
produce unified spoken discourse that showed coherence and cohesion (Canale, 1980).
However, with such entry level as the participants in this study, the scope of the 
discourse competence was altered to be (1) the knowledge and ability to utilize 
discourse markers of opening, pre-closing, and closing in conversation to achieve the 

---

50
coherence as well as cohesion and (2) the length of pause between turn-takings.
How the students opened their conversation, what kind of pre-closing signals, and 
what kind of closing strategies they used to end their conversation were closely 
related to the cohesion and coherence of their discourse. And the pause between 
turns was also related to the cohesion and coherence in speaking. 
From the observation of most EFL learners’ oral performance in the pilot study 
done by the researcher, the most common pause between turn takings was less than 
three seconds. Therefore, pausing less than three seconds would be considered 
normal or acceptable in this study. With limited language proficiency, these students 
tended to pause longer than native speakers in normal conversation
3
. If they paused 
too long, the effect of coherence and cohesion would certainly be reduced.
Accordingly, Newman (1982) also found that inserting pauses of four to seven 
seconds into natural conversations resulted in higher discomfort ratings.
Therefore, any pause longer than seven seconds would be considered silence, one of 
the major signs of communication breakdown, in the present study.
How the students repaired their own silence as well as how they fixed others’ 
silence would be reckoned as the strategic competence, which was referred to the 
possession of coping strategies in actual performance. Strategic competence, as 
defined by Canale (1983a), was composed of verbal and non-verbal communication 
strategies that might be called into action (1) to compensate for communication 
breakdowns due to limiting conditions in actual communication or insufficient 
competence in one or more of the other areas of communicative competence; and (2) 
to enhance the effectiveness of communication (Canale, 1983a).
3
According to the researcher’s observation of native speaker’s conversation, the most common pause 
between turn takings would be 0.5 seconds. 

---

51
To most EFL learners, especially learners with limited proficiency, 
communication breakdown might happen very often due to the lack of repertoire of 
vocabulary or cultural miscues. Therefore, the ability to fix or repair when 
communication breakdown inevitably occurred was particularly important in terms of 
strategic competence. Silence without a justified cause could be annoying and 
irritating during face-to-face communication and was considered one of the most 
common phenomena of communication breakdown. Without appropriate dealing, 
the effects of communication could be reduced to a great extent. How one 
responded to communication breakdown either caused by the addressee or the speaker 
was an important sign of communicative competence.