• Nie Znaleziono Wyników

FL/L2 reading – a language problem or a reading problem?

PART I: ESSENTIAL COMPONENTS OF READING –

1. FL/L2 reading – a language problem or a reading problem?

Th e discussion begins with a presentation of two confl icting hypotheses about FL/

L2 reading: the Linguistic Th reshold Hypothesis (Clarke 1980; Yorio 1971) and the Linguistic Interdependence Hypothesis (Cummins 1979a, 1986, 1991). Th e short circuit hypothesis, recently referred to as the Linguistic Th reshold Hypothesis (LTH), claims that in order to read in FL/L2, a learner must reach a certain level of FL/L2 linguistic ability. Th e Common Underlying Profi ciency (CUP) Hypothesis, recently referred to as the Linguistic Interdependence Hypothesis, states that FL/

L2 reading performance largely depends on L1 reading abilities. Let us focus on the two hypotheses and the research evidence supporting their presumptions.

1.1. The Linguistic Threshold Hypothesis

FL/L2 profi ciency plays a considerable role in FL/L2 reading. Th e Linguistic Th reshold Hypothesis claims that in order to read in FL/L2, the learner must reach a certain level of target language linguistic ability. Th e results of reading studies,

e.g., Devine (1993) and Kusiak (2000), indicate that there is an interdependence between FL/L2 language profi ciency and FL/L2 reading abilities.

Th e Linguistic Th reshold Hypothesis also concerns the question as to whether limited profi ciency in FL/L2 restricts readers in using very specifi c types of textual information, such as the discourse constraints of a text. Cziko’s (1978) studies indicated that the reading performance of advanced English profi ciency French students resembled that of native English speakers; they were more sensitive to syntactic, semantic and discourse constraints in a text and applied more nontextual information. Low language profi ciency readers, on the other hand, based their read-ing on bottom-up strategies, e.g., sensitivity to the graphic information of a text.

On the basis of these results, Cziko suggested that FL reader strategies are related to the level of FL competence.

Alderson (1984) suggested the modifi cation of the Linguistic Th reshold Hypoth-esis. He assumed that ineffi cient FL/L2 reading is due to L1 reading strategies not being activated in FL/L2 reading because of poor FL/L2 knowledge. Clarke’s (1993) study indicated that there is some transfer of skills but the language of the text “short circuits” the good reading ability, reverting the reader to poor reading strategies.

Researchers have attempted to answer the question: What is the linguistic threshold? Cummins (1979b, cited in Devine 1993) claims that the threshold can-not be defi ned in absolute terms. It will depend on the kind of reading task and the reader’s level of available and relevant background knowledge. Laufer and Sim (1985) attempted to express the qualities of the linguistic threshold in a numerical manner. In their study, they used the reading section of the Cambridge First Certifi -cate of English and their own reading strategy examination. Th ey concluded that the linguistic threshold necessary to read English for Academic Purposes successfully corresponds to a 65%–70% score on the Cambridge FCE test. Th e FL/L2 threshold was characterised by the researchers as knowledge of vocabulary, subject matter, discourse markers and syntactic structure, in decreasing order of importance.

1.2. The Linguistic Interdependence Hypothesis

Th e Linguistic Interdependence Hypothesis states that FL/L2 reading depends upon the reading abilities in one’s L1 rather than upon the learner’s level of FL/L2 competence. Coady (1979) asserts that FL/L2 reading is a reading problem, not a lan-guage problem. Th erefore, diffi culties in FL/L2 reading would be due to lack of the

“old” L1 reading skills or failure in transferring L1 reading skills into FL/L2 reading.

Cowan (1976, cited in Alderson 1984) posited a parallel processing theory of reading, which claims that the strategies readers use are language-specifi c. Read-ing diffi culties in a FL/L2 result from the reader’s predictions about the syntactic clues which are related to the strategies used in reading in a native language. Cowan gives an example of English readers reading German, who being used to subject-verb-object order will be confused by German sentences with object-verb-subject

order. Alderson (1984: 11) concludes that the “corollary of the parallel processing theory is that the knowledge of the foreign language will aff ect the development of the foreign language strategies: the less of the foreign language you know, the more likely you are to read as in your fi rst language.” Th e issue of cross-linguistic variations in sentence processing is discussed in Section 2.4. below.

Similarly, Grabe and Stoller (2002) consider a level of L1 reading abilities crucial.

Th ey claim that

students who are weak in L1 literacy abilities cannot be expected to transfer many sup-porting resources to L2 reading contexts. Th e types of abilities that students use in their L1 reading represent the upper limit of what can be expected for linguistic transfer, stra-tegic practices, problem-solving experiences, task completion skills and metacognitive awareness of reading processes (Grabe and Stoller 2002: 56).

1.3. Consolidating the Linguistic Threshold

and the Linguistic Interdependence Hypotheses

Bernhardt and Kamil (1995) reviewed the studies investigating the contribution of L1 reading and FL/L2 profi ciency to FL/L2 reading, and observed considerable con-sistency in the amount of variance accounted for by L1 reading (from 20% upwards) and FL/L2 profi ciency (from 30% upwards), with FL/L2 profi ciency consistently a more powerful predictor. Th ey pointed out that most studies leave 50% of the variance in FL/L2 reading ability unexplained.

Reading studies conducted at the end of the 20th century indicate that both fi rst language reading ability and foreign/second language profi ciency have signifi -cant eff ects on foreign/second language reading ability, e.g., Bossers (1992, cited in Alderson 2001), Carrell (1991). An interesting connection between the level of FL/L2 knowledge and FL/L2 reading was observed by Bossers, who found that FL/L2 knowledge (especially vocabulary knowledge) was strongly related to FL/L2 reading comprehension at a lower level of FL/L2 profi ciency. Th e researcher found that only “at relatively advanced levels of L2 profi ciency did fi rst-language reading ability prove to be the sole predictor of second-language reading” (Bossers 1992, cited in Alderson 2001: 39). Similarly, Block (1986) observed that profi cient FL/L2 readers performed similarly to profi cient L1 readers; less profi cient FL/L2 readers performed similarly to less profi cient L1 readers. Block (1986: 336) concluded that

“strategic resources, thus, seem more important than specifi c linguistic knowledge for these readers.”

In a recent study, van Gelderen et al. (2007) contributed considerably to the results discussed above. In a longitudinal study, the researchers investigated the eff ects of metacognitive knowledge, language knowledge and processing skills on L1 (Dutch) and FL (English) reading comprehension. Th e researchers looked at the development of 13-year-old adolescents over a period of two years (from grade 8 to grade 10). Th e fi ndings indicated high and growing correlations between L1 and FL reading comprehension. L1 and FL reading comprehension proved to be highly related to each other at the beginning of grade 8; this relationship became

even stronger in the following two years. Th e results indicated that the interconnec-tion between L1 and L2 reading comprehension would rise to 70% of the explained variance by L1 reading comprehension alone, much higher than the other studies, e.g., those quoted above, suggest. Th e researchers discuss the implications of their study in relation to the Linguistic Th reshold and the Linguistic Interdependence Hypotheses. Th e fi nding that L1 reading comprehension seemed to bear many similarities to FL reading comprehension, even in grade 8, did not support the threshold hypothesis. It was clear that at the beginning of the study the learners did not present a high level of FL competence – a language threshold assumed to enable FL learners to transfer strategies from L1 reading. Th e results showed that good L1 readers were also good FL readers, which was not due to very advanced FL compe-tence. Th e researchers underline the fact that the interconnection between L1 and FL reading comprehension increased in the following two years, which shows that FL reading comprehension becomes even more similar to L1 reading compre-hension “subsuming more and more the skills that are also involved in L1 reading comprehension” (van Gelderen et al. 2007: 487). Th e investigators concluded that their study seems to support a cross-language transfer of reading skills.

To sum up, it seems that most research points to a more signifi cant impact of FL/L2 competence than L1 reading on FL/L2 reading. However, the most recent studies suggest the possibility of an inter-language transfer of reading strategies.

More research is needed into the contribution of FL/L2 competence and L1 read-ing to FL/L2 readread-ing, preferably in relation to other factors, such as background knowledge.

2. Lower-level language processing –