them. The reasons for the central role of paraphrase lie in two aspects:first, the fact
that any utterance can be paraphrased in an indefinite number of different versions
proves the semiformal nature of the creation of meaning; second, the availability of
paraphrase allows the semantics of a language to remain within the language and
avoids unsuitable ways of organizing the meaning (Sinclare 2004: 123). Simply
put, meaning is organized in the form of language and can be organized in various
forms of language. So, if we compare the meaning to the content and language
forms to containers, the more containers we provide, the more content we can
accommodate.
Using familiar ways to express the same meaning is a very effective way of
mentally representing new information. Qualitative data analysis in this research
has proved paraphrasing assists students’ automatic building of meaning.
Paraphrase skill training can also help students to expand their vocabulary and
enrich their linguistic knowledge. With a bigger repertoire of linguistic resources as
a hospitable host in mind, more new and unfamiliar information can be treated as
welcome guests. Without a due amount of the linguistic repertoire, L2 listeners
mightfind it difficult to represent new pieces of information in mind and retain
them.
Based on verbal protocols, there are two types of paraphrase: lexical paraphrase
that uses synonyms to substitute the original words and syntactic paraphrase that
changes the word order or change the original sentence into a different syntactic
pattern. Rost (1994) concluded the effectiveness of using online-summary as a way
to assess listeners’mental representation of lectures and also suggested that online
summary should be used as an approach of academic listening instruction. If stu-
dents are encouraged to do online retelling or summarizing while listening to an
academic lecture, their paraphrase skills will be improved and meanwhile their
mental representation of the lecture content will also be strengthened.
9.7 Future Directions
More comprehensive needs analysis of academic lecture comprehension is needed,
especially in the contexts of China’s tertiary education so that task demands can be
more authentic compared with the real language use domain.
Cognitive processes and test scores can be associated with due reliability and
validity if a larger sample size is attainable. Based on a larger sample of TAP data,
trends and correlations might be explored and the argument of cognitive processes
as intrinsic indicators of test-takers’competence can be further proved with sta-
tistical significance.
Comparison of test types or test measures is also one of the key issues to be
addressed in the future. Currently, indirect test types such as MCQ, gap-filling or
multiple matching are still the main stream test measures of listening. How direct
test measures in terms of academic listening work is worth further investigation,
e.g., summarization in the oral or written form.
9.6 Recommendations for Pedagogy 167