The value of reading as a way to develop vocabulary is well-recognized in L1 development. Nagy, Herman, and Anderson (1985) conducted a research on how
children learn vocabulary in their native language. They involved elementary school students as subjects, used passages from school textbooks as texts, and concluded from their data that when children see unfamiliar words in print, a small but statistically significant increase of word knowledge typically occurs. Thus it is proposed that for L1 learners the vast majority of vocabulary words are learned gradually through repeated exposures in various discourse contexts, in other words, that most native speakers’ vocabulary learning occurs through reading and inferring from context. To increase vocabulary growth, the most effective way, hence, is to engage children in lots of reading.
This assumption that sees reading as the major vehicle for continued vocabulary acquisition, which might hold true for L1 learners, however, might not be equally well applied to L2 learners. Krashen, a leading proponent of extensive reading, has aroused some suspicion and doubt on his allegation. He drew on the Input Hypothesis that more comprehensible input results in more language acquisition, and then proposed that more reading involved is associated with greater competence in vocabulary (Krashen, 1989). He even claimed that for L2 learners reading alone is a powerful means and a principle source of developing vocabulary, and he advocated maximized quantities of reading instead of any direct instruction (Krashen, 1993). In spite of Krashen’s vigorous advocacy of reading for L2 vocabulary acquisition, Coady (1997)
pointed out that among the 144 studies Krashen (1989) analyzed in the attempt to provide evidence for the superiority of the Input Hypothesis, an overwhelming majority of them involved actually native speakers rather than L2 learners. In those few studies that actually involved L2 learners (Pitts, White, and Krashen, 1989; Day, Omura, and Hiramatsu, 1991; Dupuy and Krashen, 1993), the control groups were not given any exposure to the target words prior to the test; it is thus hardly surprising that the experimental groups demonstrated better vocabulary knowledge of the target words. As Coady (1997) mentioned, “research that positively supports Krashen’s claims with regards to L2 vocabulary acquisition is still very limited” (p, 226).
Judging from the abovementioned comment of Coady, the value of reading for L2 vocabulary growth to occur incidentally still remains uncertain. It seems a normal, unenhanced reading scheme for L2 vocabulary acquisition still has some limitations.
The first challenge is on its efficiency (how many exposures are required to ensure successful vocabulary acquisition and retention) (Min, 2008), especially in the drastic different EFL learning environments where exposures to target vocabulary are often more limited. As Laufer (2003) pointed out, experiments (Day, Omura, & Hiramatsu, 1991; Hulstijn, 1992; Knight, 1994; Paribakht & Wesche, 1997; Pitt, White, &
Krashen, 1992) all report a very small amount of L2 incidental lexical growth through reading practice, only 1-5 words per text of up to 7,000 words. Paribakht and Wesche
(1997) also indicated that the progress of L2 vocabulary gain through the increasing of L2 reading practice is “slow” and “laborious” (p. 175). Hulstijn (1992) stated in an even more explicit way: “we would like to draw the reader’s attention to a plain but important fact: the retention of word meanings in a true incidental learning task is very low indeed” (p. 122). He considered it very unlikely that the meaning of a word can be learned fully upon only few occurrences. Since the chance is very small that L2 learners remember the meaning of a word occurring only once in a text being read for its content, the whole vocabulary gain through reading practice depends critically on repeated exposures. And in fact, for the effect of repeated exposures to take place, or in other words, for the cumulative effect to take place, what really matters is not just to encounter a new word again, but to encounter it before it is forgotten (Laufer, 2003). In a normal reading condition, learners sometimes do not keep meeting words that they have just learned in the following context with an ideal interval. In such case, the small amount of learning of a new word cannot be reinforced soon by another meeting, and then “that learning will be lost”, making the L2 vocabulary acquisition
“very fragile” (Nation, 2001, p. 155), which cast more doubt on the efficiency of normal reading condition for vocabulary acquisition.
Besides the efficiency issue mentioned above, there are other concerns regarding the effectiveness (whether learners can correctly infer the meaning of an unknown
word and retain it) (Min, 2008). Laufer (2003) questioned the assumption that, on encountering and noticing an unfamiliar word, learners will use contextual clues to successfully infer its meaning. According to Laufer, not all contexts provide clues for unknown words; and when they do, clues are sometimes ignored when learners think the message is already understood, when the correct meaning contradicts with learners’ world knowledge, or when the clues themselves are unknown words and therefore unusable to the learner. Even when the clues are used and the word is correctly inferred, it doesn’t necessarily lead to retention of the word. Words that can be successfully guessed quickly and easily, Laufer argued, demand less processing effort, which work against their retention.
Lyster (1990) also challenged the effectiveness of incidental lexical growth through a normal reading-for-meaning scheme, especially when the goal is for an accurate productive use of vocabulary to take place. He discussed on Krashen’s (1982) SLA “theory” that acquisition doesn’t come from conscious learning but takes place only when learners are exposed to comprehensible input and focus on meaning, and contended that the theory is untenable. Lyster concluded that exclusive meaning-based language-use activities are “not a sufficient condition” (p. 168) for successful L2 learning, and neither is it enough for the internalization of the intricacy of all the lexico-semantic-syntactical features that rule and underlie accurate
production of L2 vocabulary.
In addition to efficiency and effectiveness issues, pedagogically, the normal unenhanced reading scheme for L2 learners’ vocabulary development to occur incidentally also causes some uncertainty. Lexical growth from this reading scheme is usually described as a “by-product” of reading, or as the accidentally learned information without learners’ deliberate intention to learn that information. In so saying, it is implicated that vocabulary gain through reading is an unpredictable process, and the unpredictability can make the learning process difficult to be influenced either through teachers’ instruction and selection of materials or through learners’ approach to the reading task (Rott, 1999).
In consideration of the above limitations, the normal unenhanced reading scheme for vocabulary acquisition, though likely accounting for a large proportion of L1 vocabulary development, may not be as applicable for L2 learners. As Paribakht and Wesche (1997) indicated, “if systematic development of L2 vocabulary is desired, it cannot be left to the students themselves. They cannot be expected to ‘pick up’
substantial or specific vocabulary knowledge through reading exposure without guidance” (p. 177). Thus, it seems some intervention needs to be implemented to accompany reading for better vocabulary acquisition to take place.