An Introduction to Applied Linguistics

particular those who have already developed their word recognition skills.
Language Threshold
A major research topic for L2 reading is the extent to which L2 language proficiency 
is needed as a support for L2 reading before L1 reading strategies and skills can 
be used effectively in an L2 context. Alderson (1984) posed the question most 
cogently in a book chapter entitled ‘Reading in a foreign language: a reading 
problem or a language problem?’. Research results at that time pointed in both 
directions, and led to the formulation of two apparently contradictory positions: 
the so-called ‘language threshold’ or ‘short-circuit hypothesis’ and the ‘linguistic 
interdependence hypothesis’. The language threshold hypothesis maintained 
that some minimal threshold of proficiency in the L2 must be attained in order 
for the reader’s first language reading skills to transfer to reading in the second 
language. The linguistic interdependence hypothesis maintained that reading or 
learning to read is accomplished only once, and that once learners have matured 
in their ability to read in the first language, the awareness of the reading process 
transfers to the second language and does not need to be relearned. Thus, reading 
performance in the second language was claimed to share a common underlying 
proficiency with reading ability in the first language (Cummins, 1979).

---

224 An Introduction to Applied Linguistics
In the first widely available empirical study to use a cross-linguistic research 
design with learners of varying L1 reading ability, L2 language proficiency and 
L2 reading ability, and utilizing multiple regression analyses, Carrell (1991) 
investigated two groups of second language learners in the USA: native speakers of 
Spanish learning English, and native speakers of English learning Spanish. Results 
showed that both independent variables (L1 reading ability and L2 proficiency), 
when taken together, were statistically significant predictors of second language 
reading ability, together accounting for 35 per cent (for the native Spanish 
group) and 53 per cent (for the native English group) of the variance in second 
language reading. However, in the native Spanish group (whose L2 proficiency 
was higher than the native English group), L1 reading ability appeared to be the 
more important predictor of L2 reading. Conversely, in the native English group 
(with lower overall L2 proficiency), second language proficiency appeared to be 
the more important predictor of L2 reading.
Bernhardt and Kamil (1995) further tested the language threshold and language 
interdependence hypotheses with adult native English speakers learning Spanish 
as the L2 at university level in the USA. Proficiency levels consisted of beginning 
freshmen, intermediate juniors and seniors who had had up to five semesters 
of Spanish study, and advanced learners who had had up to seven semesters of 
Spanish. Bernhardt and Kamil (1995) were able to account for 48 per cent of the 
variance in L2 reading by both L1 reading and L2 proficiency. Between 10 per cent 
and 16 per cent of the 48 per cent was due to L1 reading; between 32 per cent and 
38 per cent was due to L2 proficiency. For these learners, as with the similar group 
in Carrell’s (1991) study, second language proficiency was a stronger predictor of 
second language reading than was first language reading ability.
Lee and Schallert (1997) also tested the language threshold hypothesis directly, 
and did so in an EFL context, with a large sample (n = 809) of Korean middle and 
high school students exhibiting a wide range of abilities in both their L1 and 
L2 English reading, and in their L2 proficiency. Basic results yielded a squared 
multiple correlation coefficient indicating 62 per cent of the variance in L2 
reading due to the two independent variables. Approximately twice as much of 
the variance in L2 reading was due to L2 proficiency as was due to L1 reading (57 
per cent versus 30 per cent). More recently Yamashita (2002) studied 241 Japanese 
university students, assessing their L1 reading abilities, L2 language proficiency 
and L2 reading abilities. The total shared variance with the two predictor variables 
(L1 reading, L2 proficiency) was 40 per cent. The L2 language proficiency variable 
was the much stronger predictor of the two.
Although all the findings of the studies summarized above are consistent with 
the existence of a language threshold, the evidence is complicated and is also 
interpretable in terms of a continuously changing relationship as L2 proficiency 
increases, and not necessarily in terms of the existence of a specific ‘threshold’. 
Moreover, assuming that a threshold exists, it is not likely that it could be 
determined in absolute terms, even for a given population of learners.
Implications: It seems that a certain level of L2 proficiency is necessary before L1 
reading strategies and skills can be utilized effectively in L2 reading. Therefore, 
L2 reading development must take place in a learning context that also promotes 
overall L2 language proficiency, at least for lower-level students.

---

225
Reading
The Role of Background Knowledge in Reading
Work done in the 1970s and 1980s (Steffensen, Joag-dev and Anderson, 1979; 
Johnson, 1981) clearly established the role of background knowledge in second 
language reading. Further training studies showed that for students who lacked 
appropriate cultural background knowledge (or ‘content schemata’) for particular 
texts, explicit teaching of appropriate background information could facilitate 
second language reading (Floyd and Carrell, 1987).
Bernhardt (1991) was one of the first to caution against a predictive relationship 
between background knowledge and second or foreign language reading 
comprehension. Whilst finding that the effects of background knowledge were 
statistically significantly correlated with recall protocol scores on the topic 
(Pearson’s r = 0.27; p<0.05), Bernhardt (1991) pointed out the weak nature of the 
correlation. Moreover, when results were broken out by individual texts, which 
had been controlled for similarity in style and text-readability, correlations ranged 
from 0.11 to 0.59, all weak to moderate correlations. Thus, there were definite text 
content effects above and beyond prior knowledge effects.
More recent research has continued to show strong effects for background 
knowledge, but has also shown that there are complex interactions between 
background knowledge and other factors in second or foreign language reading. 
For example, Pritchard (1990) demonstrated the interaction of cultural content 
schemata and reading strategies, with students using different sets of strategies for 
culturally familiar than for culturally unfamiliar passages. Carrell and Wise (1998), 
exploring the relationship between background knowledge and topic interest, 
found a significant interaction between the two. If either prior knowledge or topic 
interest is high, students perform better than if both prior knowledge and topic 
interest are low.
Implications: Appropriate background knowledge about the topic being read 
helps learners understand the reading better. It is an important element in reading 
comprehension, but only one of many.
Knowledge of Text Structure and Discourse Cues
Beyond background knowledge of the content domain of a text, empirical 
research has confirmed that texts have particular rhetorical organizational 
patterns and that readers’ background knowledge of text structure and discourse 
cues significantly affect their reading in a second or foreign language (Carrell, 
1984a, 1984b). Moreover, training studies have also been conducted which show 
the facilitating effects on foreign or second language reading of teaching students 
to recognize and use text mapping strategies to represent the rhetorical structure 
of texts (Carrell, 1985; Carrell, Pharis and Liberto, 1989; Raymond, 1993; Tang, 
1992).
Carrell (1992), in a study of university-level ESL students’ awareness (recognition 
and use) of text structure and reading comprehension, found that those students 
who used the structure of the original passages to organize their written recalls 
remembered significantly more total ideas from the original passage than did 
those who did not. Thus, this study shows that students who possess a specialized 

---

226 An Introduction to Applied Linguistics
kind of background knowledge – awareness of different patterns used by authors 
to organize expository texts – are more likely to use a structure strategy when they 
read and, therefore, are also more likely to understand and remember more of 
what they read.
While there are relatively few additional studies of reading and discourse 
structure awareness in the past ten years, Jiang and Grabe (2007) highlighted the 
positive influence of discourse structure awareness on reading abilities, providing 
a comprehensive review of research on visual representations of text structure 
on reading comprehension. Their review showed that training with graphic 
representations which explicitly showed how the text information is organized 
(for example, cause–effect, comparison–contrast, problem–solution) improved 
students’ reading comprehension abilities.
Implications: L2 readers can benefit from an understanding of the text structures 
which organize L2 texts, and can profit from making those structures explicit. 
Training in awareness of text structure, and specifically how it organizes information 
in texts, will improve students reading comprehension over time.
Meta-cognition and Reading Strategies
In the 1980s, researchers pointed out the importance of meta-cognition as a 
factor that influences students’ reading abilities (Brown, Armbruster and Baker, 
1986). They asserted that ‘meta-cognition plays a vital role in reading’. One’s 
‘knowledge’ (for example, of strategies for learning from texts, of the differing 
demands of various reading tasks, of text structures and of one’s own strengths 
and weaknesses as a reader and learner) as well as ‘control’ or ‘regulation’ of one’s 
own actions while reading for different purposes are two different aspects of 
meta-cognition. Successful readers demonstrate higher levels of meta-cognitive 
knowledge as well as control of their reading; less successful and novice readers 
show less sophistication in meta-cognition (Baker, 2008; Baker and Beall, 2009).
One important aspect of meta-cognition is controlling one’s reading process 
through the use of strategies (see Chapter 10, Focus on the Language Learner: Styles,