On the first exposure of to a new word, all that is likely to be picked is some

Chapter Two Literature Review

2.1 Nature of L2 Vocabulary Learning

Many scholars (Brown, Sagers & La Porte, 1999; Ellis & He, 1999) share the view of the second language vocabulary learning process being “gradual, incremental, and requiring repeated exposures to the same words in various collocations and in various situations” (Gass, 1999, p. 319). This incremental nature of vocabulary

acquisition manifests itself, firstly, through the non-instantaneous learning of a word’s knowledge elements, which involve knowledge of its orthographic and phonological form, meanings, grammatical behavior, associations, collocations, frequency and register (Nation, 1999; Richards, 1976). Some of these word knowledge aspects, as Schmitt (2000) pointed out, may develop later or earlier than others and at different rates. Another dimension of vocabulary development being incremental in nature can be accounted for by different degrees of knowing a word. Language learners, at any point in time, before attaining the level of full mastery, are likely to know more about certain word knowledge aspects than others (Schmitt, 2000). Schmitt (2000) gave the following account he assumed as typical of how different word knowledge aspects are acquired (p. 117):

On the first exposure of to a new word, all that is likely to be picked is some

sense of word and meaning…. As the person gains a few more exposures

these features will be consolidated, and perhaps some other meaning senses

will be encountered. But it will probably be very later in the acquisition

process before a person develops intuitions about the word’s frequency,

register constraints, and collocational behavior, simply because these

features require a great deal of examples to the appropriate values.

Such portrayal underscores once again the point that learning vocabulary is

“multistaged, with different vocabulary items” and their components “situated at different points of several related continua at any given time.”(Gass, 1999, p. 327)

To sum up, mastery of all the lexical knowledge aspects of a given word does not occur either instantaneously or simultaneously; rather, learning a vocabulary item is an incremental and probably recursive process beginning with the first encounter with a previously unknown word and culminating with integration of new linguistic

knowledge into the current mental lexicon.

2.2 Construct of Incidental Vocabulary Learning

It is customarily acknowledged in L1 and L2 vocabulary research that much－if not most－vocabulary items are acquired “incidentally,” that is, as a by-product, not the target, of cognitive activities, such listening and reading, where learners focus their attention on comprehending meaning while picking up new word knowledge or competencies without any conscious intention of doing so (Ellis, 1999; Gass, 1999;

Huckin & Goady, 1999; Hulstijn, 2000; Nation, 2001; Paribakht & Wesche, 1999a;

Paribakht & Wesche, 1999b; Schmitt, 2000). However, attention, as it has been convincingly argued, does not suffice to clarify such type of learning, for at least some degree of consciousness is involved in incidental learning when learners exposed to new lexical items in the input (Ellis, 1999; Paribakht & Wesche, 1999a).

What distinguishes incidental and intentional vocabulary learning, then, lies in the

distinction between “focal” and “peripheral” attention; “whereas intentional learning

requires focal attention to be placed deliberately on the linguistic code (on form or

for-meaning connections), incidental learning requires focal attention to be placed on

meaning (i.e. message content) but allows peripheral attention to be directed at

form.”(Ellis, 1999, p. 35)

Contrasted with the pedagogical sense implicated above, incidental and

intentional learning, in operational terms, “can simply be distinguished in terms of the use of pre-learning instructions that either do, or do not, forewarn subjects about the existence of a subsequent retention test.” (Eysenck, 1982, p. 198, cited from Hulstijn, 2000) In other words, telling or not telling students that they will be tested on their newly acquired lexical knowledge after they complete a reading or listening task is the critical feature distinguishing incidental from intentional learning of vocabulary.

In a nutshell, incidental vocabulary learning occurs when one’s primary attention is given to the use of language (be it to comprehend meaning or to communicate) rather than to the learning of language itself.

2.3 Incidental Vocabulary Acquisition through Reading

A substantial body of research in the last two decades has confirmed the conventional belief that second language learners can acquire their vocabulary

knowledge through reading in an incidental fashion (Day, Omura & Hiramatsu, 1991;

Dupuy & Krashen, 1993; Huckin & Goady, 1999; Hulstijn, 1992; Hulstijn, Hollander,

& Greidanus, 1996; Joe, 1995; Knight, 1994; Krashen, 1989; Paribakht & Wesche, 1996). Krashen (1989), a leading proponent of extensive reading, maintains that language learners, by receiving comprehensible input while reading, acquire vocabulary and spelling most efficiently. He credits this to the Input Hypothesis, which states that successful language learning results from comprehensible input as the essential external ingredient coupled with a powerful internal language acquisition device. In contrast with Krashen’s strong assertion of learning from reading

accounting for most vocabulary acquisition, most researchers seem to take a less strong stance (Day, Omura & Hiramatsu, 1991; Dupuy & Krashen, 1993; Horst, Cobb,

& Meara, 1998; Hulstijn, 1992; Hulstijn, Hollander, & Greidanus, 1996; Paribakht &

Wesche, 1996; Zahar, Cobb, & Spada, 2001). They all point to the role for reading in vocabulary acquisition as being unpredictable, slow and not necessarily the most effective and, hence, have raised questions about the suitability of the approach for second language learners (Paribakht, & Wesche, 1997; Zimmermen, 1997). Besides, it is suggested that the amount of incidental lexical gains from reading for global

comprehension has been found to be quite small, yet statistically significant.

The following four representative, oft-cited studies all set learners a reading task without telling them to pay attention to vocabulary, and established reading as a source of increase in lexical knowledge by comparing the word knowledge of the learners with exposures to the chosen target words while reading to that of a control group that had no encounters with the target words in the text. In the Pitts, White, and Krashen (1989) study, ESL intermediate adult language learners (N = 35) were required to read the first two chapters (6700 words) of "A Clockwork Orange," a novel containing a number of slang words of Russian origin. The participants answered multiple-choice vocabulary questions 10 minutes after 60 minutes reading the text. A control group, which did not read the text, was also tested on the target vocabulary items. A modest, but significantly amount of vocabulary gain (6.4%) was revealed in the experiment as compared to the control group. Day, Omura &

Hiramatsu (1991) conducted a similar study where Japanese high school EFL learners (N = 92) and undergraduates read silently an adapted version of a story (1032 words) and then took a multiple-choice question vocabulary test. The findings showed that the subjects acquired significantly (5.8%) more than the control group, which did not read the text.

Dupuy & Krashen (1993) had ESL third-semester university students of French

watch 5 scenes of a play on film, and then read the following five scenes (15 pages

long). The subjects were then given a surprise multiple-choice vocabulary test on the

colloquial words from the text. Their performance, compared to a control group with no intervention, was found to be significantly better. Horst, Cobb & Meara (1998) replicated Pitts, White, & Krashen (1989)’s study with Jordanian EFL college students (N = 34). The participants read a full native speaker novel (21232 words) in a

comprehension-focused extensive reading program. At the end of the reading treatment, they were tested with multiple-choice questions on word meaning and a word association test. The results indicated a significant vocabulary pickup rate (20%

of the multiple-choice question items and 16 % of the word associations).

Time-consuming and seemingly haphazard as the aforementioned studies proved learning vocabulary incidentally while reading to be, they were not without issues and faults in research design, some of which might possibly account in part for the low lexical acquisition rates in them.

First of all, none of the research works ensured that the subjects were unfamiliar with the target words by administering a pretest. The use of a pre-test enables

researchers to select target words that are truly unfamiliar to the subjects, and even if the participants turn out to have some degrees of knowledge of the items, the test results can serve as a baseline against which the posttest scores can be assessed.

Secondly, the target words differed in their frequency of occurrences in text, which may have contributed to the widely varied learning rates, ranging from as high as 20%

to as low as 5.8%. It is generally recognized that the words that are encountered by learners for more times are more likely to be learned, compared to the ones met only once or twice (Nation, 2001). However, as to how many and what types of exposures are needed for a word to be acquired, there is still no consensus so far (Huckin &

Goady, 1999; Zahar, Cobb, & Spada, 2001), in that “there are so many variables

involved in learning a word that it is impossible to determine any one threshold for

number of exposures.”(Huckin & Goady, 1999, p185) Rott (1999) summarized

literature on vocabulary learning from context and concluded four major factors that can influence the outcome of inferring word meaning from context: (a) learners’

linguistic knowledge about the linguistic properties of an unknown word, (b) context properties in which the unknown word appears, (c) the approach taken by the

language learner to infer meaning, and (d) cognitive processes that influence L2 reader’s awareness if and attention to unfamiliar words.

A third critical factor for the variations in learning uptake rates would be the density of unknown words in text. Although the proportions of the unknown words in the four studies were not identified, there was a large discrepancy of text difficulty among the reading materials, with two (Pitts, White, and Krashen, 1989; Horst, Cobb

& Meara, 1998) involving longer, more authentic articles, and the other two (Day, Omura & Hiramatsu, 1991; Dupuy & Krashen, 1993) using shorter, less challenging stories. Pitts et al. (1989) reported in their study that more than half of the participants failed to finish reading the assign chapters, which was probably a result of lack of textual control. According to Nation (1990) and Laufer (1997a & 1997b), it is likely that at least 95% of the running words in a text need to be already familiar to the learner for general text comprehension to happen. Nation (2001) even suggests that a coverage rate of 98% or higher is optimal for second language readers to gain

adequate comprehension and acquire vocabulary incidentally in a text. The known word coverage in the present study is above 99%.

Moreover, the subjects’ lexical improvement in each of the four studies was measured merely by multiple choice tasks, a kind of measurement tool that have been documented to show some possible complications (Anderson & Freebody, 1981;

Meara & Buxton, 1987; Wesche & Paribakht, 1996; Waring, 1999, cited from Waring

& Takaki, 2003). First, multiple-choice tests have the risk of the subjects guessing

randomly. Secondly, multiple-choice tests only assess and focus on form-meaning

recognition, understating the importance of other types of word knowledge and thus neglecting to give credit for partial, small amount of learning. As a result, the findings of the four studies above might have been underscored. In an attempt to overcome this instrumentation limitation, the present study employed Vocabulary Knowledge Scale (VKS), a vocabulary knowledge instrument which has been proven to be sensitive to gains in lexical knowledge resulting from reading (Wesche & Paribakht, 1996) and the purpose of which “is not to estimate general vocabulary knowledge, but rather to track the early development of specific words in an instructional or experimental situation.” (Wesche & Paribakht, 1996, p. 33)

Another variable, overlooked by the foregoing four studies, but probably an important factor in the low acquisition percentages, is the role context played in facilitating acquisition of the new words. The contextual clues for inferring meaning of the target words in the studies were not examined, so it is not known whether the vocabulary items tested were adequately cued or not by the surrounding text for successful lexical inferencing. It is very likely quite a number of the new words lacking in contextual information were included in the vocabulary gain tests. The same experiment design flaw was also identified in Chang (2002)’s study, which aimed to investigate the effects of marginal glosses and pocket electronic dictionaries on incidental vocabulary learning. The subjects (N = 92), divided into three groups, 1) with marginal gloss support 2) with pocket electronic dictionary support, and 3) with no external support, all read a text containing 16 words, examined subjectively by the participants’ instructor to be unknown to the subjects without reference to the

presence of contextual clues. The result of the vocabulary gain evaluation showed that

the control group, which read for meaning without support, had a low lexical learning

rate (5%). Even though there is no agreement pertaining to the amount or types of

contextual clues required for a word’s meaning to be inferred or acquired successfully

from context (De Bot, Paribakht & Wesche, 1997; Fraser, 1999; Huckin & Bloch, 1993; Morrison, 1996), with no clues there to be exploited, it is impossible for any learning to occur (Laufer, 1997a). It is also worth noting that, nevertheless, there seems to be a trade-off between richness of contextual information and success of incidental lexical inferencing: “the very redundancy of information in a given context which enables a reader to guess an unknown word successfully could also predict that the same reader is less likely to learn the word because he or she was able to

comprehend the text without knowing the word.”(Nation & Coady, 1988, p. 101) The factors involved in guessing word meaning from context are so many that there is practically no way to locate on the contextual information continuum the middle ground where cues provided by context are sufficient enough for a given vocabulary item to be acquired by any individual learner, but not so redundant as to cause it to be ignored. Even so, to overcome the limitation of the previous research on the

contextual clues surrounding the target words, the present study made an attempt to ensure the presence of contextual cue(s) for every target word tested.

While it appears that, despite the experiment design issues mentioned above, vocabulary uptake from reading incidentally may be really rather small and

unpredictable, and inferring meaning is not always an easy or efficient strategy for second language learners to use (Hulstijn, Hollander & Greidanus, 1996; Huckin &

Coady, 1999; Laufer, 2003), its cumulative effects on substantial lexical growth can never be valued enough (Nagy, 1997; Gass, 1999; Krashen, 1989; Nation, 2001;

Waring, & Takaki, 2003). Except for the 2000 high-frequency core vocabulary items,

most scholars seem to concur with the view that vocabulary learning predominately

occurs through extensive incidental reading (Hulstijn, 2000; Nation, 2001; Schmitt,

2000). Plus, it is estimated that if a learner reads one million running words of text a

year, and if the pickup rate 5%, the annual gain would roughly be more than 1000

words (Nagy, 1997, cited from Nation, 2001). On the other hand, Nation (2001) suggests that this low incidental lexical learning rate could be “a benefit rather than a cause for concern,” (Nation, 2001, p. 238) in that “the small, gradual increments of learning a word from context under normal conditions of incidental learning

encourage a flexible approach to finally determining the meaning and make it unlikely that an initial, strong but wrong interpretation will be made and maintained.”(Nation, 2001, p. 239) In addition, it is revealed that some aspects of word knowledge are more amenable to incidental learning than others. The repeated encounters and varied contexts can lead to more reliable intuitions of word collocations, register constraints, frequency, additional meaning senses, and consolidation of the known lexical aspects (Fraser, 1999; Gass, 1999; Nation, 2001; Schmitt, 2000; Waring & Takaki, 2003).

To conclude, incidental lexical acquisition from context can be construed as a reflection of the incremental nature of vocabulary learning － a gradual process where lexical representations in mental lexicon, instead of building up as a result of learning from concise, neatly contained dictionary definitions, accumulate as

“conglomerates of association with other words and collocation of words in context.“(Gass, 1999, p. 238)

2.4 Incidental Vocabulary Acquisition through Listening

Although it is generally acknowledged that in both L1 and L2 lexical

development can take place when learners attempt to comprehend new words in both spoken and written contexts (Day et al., 1991; Ellis, 1994; Ellis, 1999; Gass, 1999;

Huckin & Goady, 1999; Hulstijn, 2000; Krashen, 1989; Nation, 2001; Paribakht &

Wesche, 1999a; Paribakht & Wesche, 1999b; Schmitt, 2000; Sternberg, 1987), the

role of oral input in promoting L2 incidental vocabulary acquisition, as compared to

that of written input, has been under-researched and received relatively less attention

(Ellis, 1999; McCarthy & Carter, 1997; Paribakht & Wesche, 1999b). It is suggested that reading dominates as a source for processing new words, particularly for learners in a foreign language setting, because the main access to target language sources is often limited to print material (Krashen, 1989; Rott, 1999). It is not until recent decades that it is evidenced by a growing body of empirical research that picking up news words incidentally in oral contexts is vital and the amount is considerable as well (Nation, 2001). Ellis (1999) even claims that oral discourse is more conducive to incidental lexical learning than written one by stating:

“Inferencing (learning word meaning from context) may be a more effective mechanism of acquisition when the input is oral rather than it is written. In the case of written text the only cues available to learners for inferencing the meaning of the new words are those contained in the text itself (although there may also be extra-linguistic cues in the form of pictures and diagrams)….Oral text….In addition to the cues provided by the verbal text itself, there are other cues available from the intonation and gestures used by the speaker and from the situational context.” (p.37)

Incidental learning of new vocabulary items from oral input has been reported mostly in the studies where L1 learners listened to stories read aloud. Senechal, Thomas and Monker (1995) tried to assess how children having different levels of vocabulary knowledge learn new vocabulary items incidentally from listening to stories read aloud. Eighty 4-year old children listened to reading s of a book. Findings show that all the participants acquired some, though not much vocabulary knowledge.

Robbins and Ehri (1994) conducted a study where 51 kindergartners listened to the same storybook read two times, 2 to 4 days apart, and then completed a

multiple-choice test on knowledge of unfamiliar words. The results confirm that

reading stories aloud does contribute modestly to vocabulary growth. In another study

conducted with kindergarten children, Eller, Pappas and Brown (1988) also found that learners acquired vocabulary knowledge from listening to stories read aloud to them, without formal instruction. The subjects listened to two different stories and then were asked to read them by themselves. The findings show that the children used the target words with higher frequency and more accuracy. Labonty (1988) conducted a study examining whether third-graders (N = 128) increased their vocabulary knowledge as a result of listening to unfamiliar words within a story. Following the reading, an oral posttest on vocabulary and sentences was given. Results indicate that the children who heard the story from which the words and sentences were taken had a better understanding of those words and recognized sentences from that story better than the children who had heard the other story. In a study involving 175 fourth graders from 6 different classes in two elementary schools, Brett, Rothlein, and Michael (1996) explored the effects of three conditions on fourth-grade students' vocabulary acquisition: (1) listening to stories, with brief explanation of unfamiliar words; (2) listening to stories with no explanation of unfamiliar words; and (3) having no exposure to stories or to word explanations. The participants listened to each of the two stories for once, over a period of 5 days, and then were given a pretest and posttest on 10 target words. The results indicate that the two groups with oral input both had vocabulary gains, and the group that listened to two stories along with brief explanation of the target learned significantly more new words than the students who heard the stories without any explanation of the words. It seems evident from the aforementioned studies that increase in vocabulary knowledge can occur incidentally, without explicit formal instruction for L1 children.

In second or foreign language, incidental vocabulary acquisition through

listening to stories, by comparison, has not been valued and researched as much, with

only a couple of empirical studies being available. In Elley’s experiments (1989), ESL

elementary students in New Zealand demonstrated significant and long-term vocabulary acquisition after listening to a story read aloud to them by teachers.

Results of pretests and posttests show that 1) students exposed to a story three times without explanation showed a mean vocabulary gain of 19 %; 2) students exposed to once with brief explanation produced a gain of 20%; 3) students exposed with three times plus explanation achieved an increase of 33%. In a study done with ESL first graders, Fondas (1992) examined the effects of reading aloud on students’ vocabulary learning, a teaching method described by her as “a naturalistic approach” which allows the students to acquire new vocabulary from a naturalistic conversational setting. The experimentation lasted for a period of 10 weeks, during which five books were read orally to the participants once each day, five days a week, combined with guided discussions focusing on developing critical thinking skills, before, during, and after the reading. The results of multiple-choice pretests and posttests on 25 target vocabulary items show that 80% of the subjects demonstrated enhanced knowledge of the selected vocabulary words.

Ulanoff and Pucci (1999) examined two bilingual methodologies: concurrent translation and preview-review, by comparing the gains made by third-grade ESL students of Spanish, who were randomly assigned to one of the three groups: the control (no treatment) (N = 16), concurrent translation (N = 21), where the teacher translated the story from English to Spanish as the subjects listened to it, and

preview-review (N = 23), where important points and difficult vocabulary in Spanish

were previewed before listening and also reviewed after listening. All the students

were given a multiple-choice test to assess their knowledge of the 20 vocabulary

items selected from the text before listening to a story in English, and the same test

was administered after the treatment to examine gains in scores. The results indicate

that there was an overall gain of 19% for the group without any instruction, 12% for

the concurrent translation group, and 57% for the preview-review group.

As evidenced in the three foregoing studies, second or foreign language learners can actually learn incidentally from oral contexts and achieve gains in vocabulary when engaged in listening to stories, and the increase of target word knowledge could be even greater when oral reading scaffolded by subsequent mediation, such as teacher explanation for the target words; however, the subjects involved were limited to only elementary graders. To generalize the results to a wider population, the present study experimented with a totally different age group, EFL senior high school

teenagers.

Positive research findings on lexical gains through listening were also obtained

with ESL learners listening to academic lectures. In an empirical study exploring the

effects of academic lecture listening on vocabulary acquisition, Vidal (2003) provided

evidence that listening to academic lectures in an EFL setting can be considered a

source of vocabulary acquisition. During a four-week period and on different days,

the EFL university freshmen participants (N = 122) viewed three video-taped lectures

on the economic, social-cultural, and environmental impacts of tourism, each of

which was of approximately equal length (14-15 minutes). Immediately after each

viewing, their level of knowledge for each target vocabulary item (36 items in total,

12 items for each lecture) was measured on a modified version of Vocabulary

Knowledge Scale, and their lecture comprehension was evaluated by three ten-item

true-false and three twenty-item listening cloze tests. A significant increase in

vocabulary knowledge from the mean score of the pretest 1.5 to posttest 30.4 was

found (based on estimated marginal means). Smidt and Hegelheimer (2004) explored

the effects of online authentic academic lectures embedded in a CALL activity on

listening comprehension, incidental vocabulary acquisition, and strategy use. A total

of 24 adult EFL learners between the ages of 18-42 participated. During a 25-min

duration, through a self-paced, autonomous learning mode, they first watched a 15-min digitalized online video lecture on horticulture, and then completed 10 multiple-choice questions on comprehension and three partial dictation passages on knowledge of 20 of the most difficult target vocabulary. The mean performance of all the learners increased from 3.8 on the pretest to 7.0 on the posttest, indicating that incidental vocabulary acquisition did occurred as a by-product of the academic lecture CALL activity.

Considering the primacy of oral input in facilitating vocabulary learning, it is surprising that there has only been noticeably sparse research devoted to L2

vocabulary acquisition through listening. Therefore, there is clearly a need for further experimentation in this respect.

2.5 Differences and Similarities between Reading and Listening

It has long been assumed that reading and listening are two very similar and highly correlated language modalities, partly because traditionally they are both considered “receptive skills” ,and also partly because comprehension is a general construct that applies to both reading and listening (Lund, 1991; Osada, 2004),

typically involving three stages: perception, parsing, and utilization (Anderson, 1990).

In addition, it is suggested that reading and listening comprehension are both dynamic

and complex processes that “draw on knowledge of the linguistic code, cognitive

processing skills, schema-based understanding, contextual cues both within and

outside the text” and “both skills can be characterized as problem-solving activities

involving formation of hypothesis, the drawing of references, and the resolution of

ambiguities and uncertainties in the input in order to assign meaning.”(Omaggio,

1993, p.165) However, it is now realized that reading and listening are separate sets of

skills in their own right and that research from either may not automatically transfer to

the other (Lund, 1991; Osada, 2004). In fact, the nature of the input (speech or writing) and the input processing processes are quite different.

Considering text structure and interaction with audience, spoken text is described as more “fragmented (loosely structured)” and “involved (interactive with the

listener)”, whereas written text is more “integrated (densely structured)” and

“detached (lacking in interaction with the listener).”(Flowerdew & Miller, 2005, p.48) Buck (2001) summarizes major linguistic differences between spoken and written languages as follows:

z

In spoken language idea units tend to be shorter, with simpler syntax, whereas written idea units tend to be more dense, often using complex syntax, such as dependent and subordinate clauses, to convey more information.

z

In spoken language idea units tens to be strung together by coordinating conjunction (and, or, but etc.), whereas written idea units tend to be joined in more complex ways.

z

Spoken language usually has hesitations: pauses, fillers and repetitions that give the speakers more thinking time, as well as repairs, such as false starts,

correction in grammar or vocabulary, and afterthoughts.

z

Spoken language tends to have more non-standard features such as dialect, slang, and colloquialisms, whereas written language tends to be more formal,

conservative, and “correct”.

z

Spoken language tends to be far more personal, with more emotional

involvement and much less precision. Speakers tend to indicate their feelings more, with expressions such as “I think” or “I mean”, or by making direct reference to the listener. They also tend to be content with gross approximations, or use overstatements and exaggerations (p. 10-11).

It is worth noting that spoken and written texts do not belong to two distinctive

categories, however. Spoken and written texts can be positioned along an oral-literate continuum exhibiting more or less the “spokenness” and “writtenness” features, “with oral texts at one end, having more characteristics of spoken language that are typically associated with casual conversation, and literate texts at the other end, having more characteristics of written language that are especially associated with expository written prose.” (Buck, 2001, p. 11)

An oral text is ephemeral, characterized as “here one minute, gone the next”

(Nunan, 1989). “It exits in time, rather than space” (Lund, 1991, p. 201), so that it must be processed as it is uttered. When the input being processed, immediate

perception, understanding, retention, integration, and evaluation all need to be at work at the same time. What makes it even more challenging is that listeners have no control over the rate of input and that oral text is not available for rehearing. As a result, any inattention to what is being said at the moment may easily cause him or her to lose an important part of the message, or even all of it. Oral text processing,

moving along a time axis, being real-time in nature, thus, imposes a heavy cognitive load on listeners. In contrast, a written text is permanent, with its overall duration and organization presented to readers at a glance. Also, it can be “received, stored and referred back to any time.” (Nunan, 1989, p. 36) A reader, with considerable control over a text, can dwell on or skip over a part of it, or easily go back and forward for rereading it.

Despite the fact that visual and aural texts vary in nature and in the way the input is processed, the general comprehension processing in the two modalities, as proposed by L1 and L2 researchers and theorists (Brown, 1998; Brown, 2001; Buck, 2001;

Flowerdew, & Miller, 2005; Omaggio, 1993; Oxford, 1993; Richards & Schmidt,

2002), can be explained by the three following models－bottom-up, top-down, and

interactive processing. Bottom-up refers to language processing starting at the lowest

levels, i.e., linguistic information such as letter recognition, word recognition, proceeding to sentence level syntax, and then to discourse level in a serial manner.

Top-down processing, on the other hand, is theorized to be conceptually-driven, where

mental relations of text meaning to our existing knowledge are formed in order to predict, interpret, or compensate for lower level information. Interactive process model, also known as parallel processing, the most widely recognized theory of reading and listening comprehension, holds that language comprehension is the outcome of micro-and macro-linguistic decoding processes interacting with one another simultaneously in a cyclic fashion.

Given the distinguishing features between spoken and written languages

discussed above, it is not unreasonable to suggest the possibility that the two different language modalities may result in differences in vocabulary learning outcomes.

However, to date, essentially, this is more of a theoretical view than a fact, due to the

lack of literature reporting empirical support; available are only studies concerning

analysis of second language reading and listening comprehension (Lin, 2003; Lund,

1991; Park, 2004) and correlations between English spoken and written vocabulary

sizes and EFL listening comprehension competence (Tsai, 2005). In light of this gap

in research, one of the goals of this research is to gain a better understanding of the

effects of spoken and written input on vocabulary acquisition.

2.6 Research Hypotheses

Hypothesis 1: The incidental vocabulary gain and retention obtained from written input are significantly greater than those from oral input.

As English teaching in Taiwan focuses predominantly on reading and writing instructions and listening proficiency development tends to be trivialized in the high school English classroom (Yen, 1987, cited from Lee, 2006), it is hypothesized that vocabulary knowledge gained and retained through reading would be significantly greater than that through listening.

Hypothesis 2: Text comprehension correlates significantly and positively with incidental vocabulary acquisition.

Previous research has observed correlating relationship between text

comprehension and vocabulary knowledge existing in both L1 and L2 (Daneman, 1996; Kintsch, 1998, Koda, 1996, Laufer, 1997, cited from Pulido, 2004). It is assumed that the greater level of text comprehension, the greater success in

constructing a mental representation from understanding a message (be it written or aural) and in turn, the greater “the likelihood of successful inferencing to solve the meaning of any unfamiliar words deemed relevant or important” (p.502, Pulido, 2004).

The above hypothesis is based on this assumption and the findings in Pulido’s (2004) study that levels of text comprehension was found to be directly proportionate to success in vocabulary acquisition.

Hypothesis 3: Proficiency correlates significantly and positively with incidental vocabulary acquisition.

It is generally agreed that stronger learners have larger vocabulary and are better

at decoding messages, constructing and integrating ideas from context than weaker

learners, so as a result, higher-ability learners tend to be more successful in resolving the meaning of unfamiliar words than lower-ability ones (Ferrel & Daloğlu, 2006;

Nation, 2001; Pulido, 2004; Vidal, 2003). The findings confirming this view were found in the studies where EFL learners acquired vocabulary while listening to academic lectures (Vidal, 2003) and the studies where subjects learned words incidentally while reading for overall text comprehension (Ferrel & Daloğlu, 2006;

Pulido, 2004). With research substantiating this observation, it is logical to hypothesize that learners’ proficiency would correlate significantly and positively with incidental vocabulary acquisition.

Hypothesis 4: Exposure frequency of vocabulary has an effect on incidental vocabulary acquisition and retention.

Nation (2001) indicated that repetition of a word in a text is generally in

moderate positive correlation with success in learning it. In Rott’s study, learners were exposed to unfamiliar words either two, four, or six times during reading, and

significant effects of frequency on vocabulary learning were observed. Based on the

observation, it is hypothesized that exposure frequency of vocabulary would exert an

effect on incidental vocabulary acquisition and retention