The effects of context on incidental vocabulary learning, Teaching, learning etc (thanx silvermoth)
[ Pobierz całość w formacie PDF ]
Reading in a Foreign Language
October 2008, Volume 20, No. 2
ISSN 1539-0578 pp. 232–245
The effects of context on incidental vocabulary learning
Stuart Webb
Victoria University of Wellington
New Zealand
Abstract
Japanese university students learning English as a foreign language (EFL) encountered
10 target words in 3 sets of 10 short contexts that were rated on the amount of
information available to infer the target words’ meanings. One group of learners met the
target words in contexts rated more highly than the contexts read by the other group. A
surprise vocabulary test that measured recall of form, recognition of form, recall of
meaning, and recognition of meaning was administered after the treatments. The results
showed that the group that read the contexts containing more contextual clues had
significantly higher scores on both tests of meaning. The findings indicate that the quality
of the context rather than the number of encounters with target words may have a greater
effect on gaining knowledge of meaning. Conversely, it is the number of encounters that
will have a greater effect on knowledge of form.
Keywords
: incidental learning, context, vocabulary knowledge, meaning, form, orthography,
repetition
Research has shown that both first language (L1) learners (Jenkins, Stein, & Wysocki, 1984;
Nagy, Anderson, & Herman, 1987; Nagy, Herman, & Anderson, 1985; Shu, Anderson, & Zhang,
1995) and second language (L2) learners (Day, Omura, & Hiramatsu, 1991; Dupuy & Krashen,
1993; Hulstijn, 1992; Pitts, White, & Krashen, 1989; Waring & Takaki, 2003) may incidentally
gain knowledge of meaning through reading. While researchers tend to agree that incidental
learning is responsible for the vast majority of L1 vocabulary learning (Nation, 2001; Schmitt,
2000), there is some suggestion that explicit learning of vocabulary may be responsible for most
L2 vocabulary learning (Laufer, 1991, 2001; Laufer & Paribakht, 1998; Webb, 2008). However,
researchers agree that incidental vocabulary learning should be encouraged and incorporated into
L2 learning (see for example, Hunt & Beglar, 2005; Nation, 2001; Schmitt, 2000; Waring &
Takaki, 2003). Because learners incidentally gain knowledge of words in small increments,
building upon their previous gains through repeated encounters until a word is known, incidental
vocabulary learning can be a relatively slow process when there are long gaps between
encounters. Currently it is not clear how many encounters are needed to learn an unknown L2
word. Hulstijn, Hollander, and Greidanus (1996) found that there was little difference between
encountering target words once or three times. Rott (1999) suggested that six encounters may be
enough to learn a word. Horst, Cobb, and Meara (1998) suggested eight encounters are needed,
Saragi, Nation, and Meister (1978) suggested 10 encounters, Webb (2007a) suggested that more
Webb: The effects of context on incidental vocabulary learning 233
than 10 encounters are needed, and Waring and Takaki (2003) reported that it may take more
than 20 encounters to incidentally learn the meaning of a word. Moreover, in some of the studies
the number of encounters needed to learn the meaning of a word varied considerably between the
target words (Horst, et al., 1998; Saragi et al., 1978).
Context may be one reason the number of repetitions needed to learn individual words varies. In
some sentences the meaning of an unknown word might be transparent but in others it may be
opaque. Beck, McKeown, and McCaslin (1983) suggested that many contexts may be deceptive,
leading learners to infer an incorrect meaning. In studies that involve reading books, there may
be too many encounters for researchers to take each context into account. However, it could be
expected that target words that are met repeatedly in sentences that offer some information about
the meaning of a word are going to be learned before those that appear in less informative or
misleading sentences. Unless context is taken into consideration, it may be difficult to make an
accurate assessment of incidental vocabulary learning. This may be particularly true in L2
learning where the number and frequency of encounters with unknown words are likely to be
less than in L1 learning.
Context may have been the reason for the contrasting results in Rott (1999) and Hulstijn et al.
(1996). In the Rott (1999) study, the texts were created with enough contextual clues for the
learners to be able to infer the meaning of the target words. Whereas, in the study by Hulstijn et
al., the text was authentic, and the researchers had determined that it was “extremely difficult to
infer the exact meaning” of the target words from the context (p. 330).
Differences between the types of contexts used in research may often account for conflicting
results. This may be the case in many studies on incidental learning because a large variety of
contexts have been used. For example, Herman, Anderson, Nagy, and Pearson (1987) used four
different types of context in a study of incidental vocabulary learning. They used a 1,230-word
passage from a text and three edited versions with varying degrees of implicit and explicit clues
about the target words. Other contexts used in incidental learning of vocabulary experiments
were a novel (Saragi et al., 1978), a graded reader (Horst et al., 1998), specially constructed
paragraphs (Jenkins, Stein, & Wysocki, 1984), and narrative and expository texts of about 1,000
words (Nagy, Herman, & Anderson, 1985). Studies on learning from context have used single
sentences (Dempster, 1987; Laufer & Shmueli, 1997), multiple sentences (Dempster, 1987), L2
sentences together with their L1 translations (Griffin, 1992), L1 sentences containing the L2
target vocabulary (Pickering, 1982), three sentences one of which was a definition (Gipe &
Arnold, 1979), and L1 and L2 glossed passages (Laufer & Shmueli, 1997). The wide range of
contexts may lead to misinterpretation of results. Would students learning vocabulary from a
specially constructed passage containing many clues about a target word’s meaning learn as
easily from an authentic text? Probably not (Beck et al., 1983; Herman et al. 1987). Without a
clear definition of context, it seems likely that results will continue to vary between studies.
The present study was carefully designed to gain insight into the effects of context on incidental
vocabulary learning. Perhaps due to the length of the texts and the large number of encounters
with target words, previous L2 research has provided little information about the contexts in
which target words were met. Short contexts each containing a single target word were used in
this study. Each context was rated on the information that could be used to infer the meanings of
Reading in a Foreign Language
20(2)
Webb: The effects of context on incidental vocabulary learning 234
the target words. By rating the contexts in which target words were encountered, the effects of
more and less informative contexts could be compared.
Method
Participants
The participants in this study were 50 Japanese native speakers learning English as a foreign
language in four 2
nd
-year university classes in Fukuoka, Japan. All of the participants had studied
English for a minimum of 7 years and had scored 80 percent or higher at the second 1,000-word
level of Version 1 of the Vocabulary Levels Test (Schmitt, 2000). Their average raw score was
27.7/30, indicating that they had mastered that level (Schmitt, Schmitt, & Clapham, 2001) and
should have little difficulty understanding all of the running words in the treatments.
Design
The experiment was conducted within one 90-minute class period. The participants were
randomly assigned to two groups: experimental and comparison. Each group completed a
reading comprehension task. The task involved reading three sets of 10 sentences, each sentence
containing 1 of 10 target words. The target words were disguised forms that had replaced 10 L2
words. The use of disguised forms ensured that the participants would have no prior knowledge
of those items. The amount of information presented in the sentences that could be used to learn
the meanings of the target words varied between the two tasks. Each set of 10 sentences was
presented on one page. The first page was the same for both groups. However, the second and
third pages that were presented to the experimental group were rated as being more informative
than those sentences presented to the comparison group. Each group was given 12 minutes to
complete their treatment. Four minutes was provided to read each page. In pilot tests, 4 minutes
was found to be enough time for the participants to read all of the sentences on a page. A surprise
vocabulary test measuring knowledge of form and meaning was administered after the treatments.
The results of the groups were then compared to determine whether the amount of information
presented in the contexts had a significant effect on vocabulary learning.
Target Words
Six nouns and four verbs were selected as target words because nouns and verbs are the most
common parts of speech found in natural text, and the 6:4 ratio approximates their proportional
frequency of occurrence in language use (Kucera & Francis, 1967). Target words were chosen
based on their frequency of occurrence in a selection of graded readers. The number of target
words was determined during pilot studies. Factors, which were considered when establishing
the number of target words, were subject fatigue and the time needed to complete the treatment
and tests.
To ensure that the participants had no prior knowledge of the target words, they were replaced
with disguised forms. The disguised forms and their L1 meanings were as follows:
ancon
(hospital),
cader
(lunch),
dangy
(street),
denent
(remember),
faddam
(write),
hodet
(face),
masco
Reading in a Foreign Language
20(2)
Webb: The effects of context on incidental vocabulary learning 235
(train),
pacon
(wear),
sagod
(visit), and
tasper
(evening). To avoid confusion with partially
known or known L2 items, the spellings of the disguised forms did not always conform to
common L2 spellings. While there was a small possibility that this may have affected incidental
learning of form, it should not have had any influence on the learning of meaning. In interviews
with learners who took part in pilot studies, they reported that they believed all of the disguised
forms to be authentic L2 words despite the fact that they were likely to have had L2 knowledge
of form for the original target words. This was possible because the target words represented
meanings that had low-frequency L2 synonyms. For example, although the participants may
have known the words
hospital
and
write
, they were unlikely to know synonyms such as
sanatorium
,
infirmary
,
inscribe
, and
scribble
. Although the learners may have known the most
frequent L2 forms, they may have assumed that there were less frequent synonyms that they did
not know. Because the participants in this study did not know that disguised forms had been used,
the experiment should have simulated authentic vocabulary learning.
1
Contexts
Participants encountered target words in short contexts (see Appendix A). A context was either
one or two sentences long and averaged 14 words. The contexts were presented in random order
on each page. Contexts were taken from the following graded readers from the Oxford
Bookworm series:
The Elephant Man
,
Lord Jim
,
Agatha Christie
,
Chemical Secret
,
The Garden
Party and Other Stories
,
and
Decline and Fall
. Only one sense of a target word’s meaning
appeared in the context and the part of speech of each target word never changed from context to
context. Extensive pilot testing was used to ascertain that participants were familiar with all of
the running words in the contexts. Contexts that contained words that were unknown to any
learner in the pilot tests were not used in the study. However, proper names that may not have
been known to the participants such as
King’s Cross
and
Harrogate
were included in the
contexts because this is typical in graded readers. In studies where it is particularly difficult to
determine the effects of a variable, controlling for unknown proper nouns may provide a more
accurate assessment of its effects.
Order and Rating of Contexts
Each sentence in the treatments was rated on the amount of information it provided about a target
word’s meaning. Researchers need to take into account the fact that the meaning of unknown
words may be relatively clear in some contexts while in others it may be opaque or misleading.
By rating each context on the information that could be used to learn the target words, the effects
of context could be isolated. One weakness of previous L2 research is that little or no
information has been provided about individual contexts. This is likely due to the large number
of occurrences of target words and much longer texts in those studies.
The contexts were rated by two native speakers. Only contexts that were given the same rating
by both native speakers were included in the study. The contexts were rated according to the
following scale (adapted from Webb, 2007a):
1. Extremely unlikely that the target word can be guessed correctly. The text contains no
contextual clues and may be misleading.
Reading in a Foreign Language
20(2)
Webb: The effects of context on incidental vocabulary learning 236
2. It is unlikely that the exact meaning of the target word can be inferred. However,
information in the context may lead to partial knowledge of the target word’s meaning.
3. Information in the context may make it possible to infer the meaning of the target word.
However, there are a number of choices. Participants may gain partial knowledge.
4. Participants have a good chance of inferring the meaning correctly. There are few
meanings that are logical apart from the correct meaning. Participants should gain at least
partial knowledge.
The following examples for the target words
sagod
(visit) and
ancon
(hospital) were rated from 1
(
least informative
) to 4 (
most informative
) on the scale above.
1. Her brother Edward was always with her when the King came to sagod.
2. He could read about things, and talk to his visitors, but he could not go out of the ancon
by himself. He thought and played like a child.
3. “She’s ill, that’s all,” I said quickly. “We’re going to the ancon.”
4. He was not ill, and of course the beds in the ancon are for ill people.
The order of the sentences was determined by their ratings. For both groups, the most
informative context for each target word was the first presentation. However, the second and
third encounters with the target words were different. The experimental group met the target
words in the second most informative context in the second presentation and the third most
informative context in the third presentation. For the comparison group, the most informative
context was followed by the two least informative contexts in the second and third encounters
with the target words. The context ratings for each sentence are shown in Table 1.
Table 1.
Ratings for each context
More informative
Less informative
Target word
Encounter
Encounter
1
st
2
nd
3
rd
1
st
2
nd
3
rd
ancon (hopital)
4
4
4
4
2
2
faddam (write)
4
4
3
4
3
3
masco (train)
4
3
3
4
3
2
pacon (wear)
4
3
3
4
3
3
sagod (visit)
4
3
3
4
1
1
denent (remember)
4
4
4
4
3
2
dangy (street)
3
2
2
3
2
1
cader (lunch)
3
3
3
3
2
2
hodet (face)
4
4
3
4
2
2
tasper (evening)
4
4
3
4
2
2
Reading in a Foreign Language
20(2)
[ Pobierz całość w formacie PDF ]