Comparing retelling in Phase 2 and the gap-filling task in Phase 1, we may arrive
at a conclusion that retelling is easier because the immediate information for
retelling is stored in the working memory while for Phase 1, the answer would
evade the participants if they fail to select the relevant propositional units and write
them down in their notes. Hence, the gap-filling task after the mini-lecture is more
cognitively demanding, which involves a more complicated construct than online
retelling. Nevertheless, the consistent protocols across two phases related to all the
four gap types present participants’ different cognitive processes involved in
decoding, selective attention, meaning and discourse construction. I would still
argue retelling can be treated as a way to assess academic listening construct. For
lower cognitive processes targeted by key point gaps and details gaps, the Phase 1
TAPs and Phase 2 retelling protocols look consistent, which emphasizes the
argument that decoding ability is thefirst watershed dividing successful listeners
and less successful listeners. For inference and summary gaps, the Phase 1 TAPs
and Phase 2 retelling protocols are partially consistent, for retelling is not as cog-
nitively demanding as the gap-filling task. The construct of gap-filling task also
encompasses selective note-taking that helps participants with later information
retrieval and the notes are actually the written form of their discourse
representation.
8.7.2 Parsing Complicated Sentences
The definition of parsing has not reached consensus yet. In Buck’sdefinition,
parsing means“establishing the relationship between the meaning of individual
words and the meaning of whole utterances”(Buck 2001: 16) while in Field’s
definition, parsing means“tracing the grammatical structure that links a group of
words”(Field 2008: 349). Here, parsing is more related to the way on how par-
ticipants are processing syntactic structures of complicated sentences. Gernsbacher
(1990, 1997) provided the structure building framework in reading which encom-
passed syntactic paring and propositional integration as essential components of
reading comprehension and proposed three components in sentence processing:
laying a foundation structure, mapping new information onto an existing structure
or opening a new structure. We need to see if sentence parsing in listening follows
the same pattern. The following retelling protocols concern how participants parsed
a long and complicated sentence:
Fan: For instance, you maymake connection between what you are reading and what you
have known before and you mayidentify the concepts you do not understand very well, and
evaluate the importance ofwhat?
Tan: For example, you maymake connections between what you’ve already known and
what you’ve not and make, what’s the second point, and toevaluate the information that
you don’t know and that’s the third point. To evaluate, that’sabreakdown.
8.7 Other Aspects 149