Reading in a second language

Author: H. Catherine Walter


Reading in a second language calls for fast, automatic word decoding and access to the mental lexicon (dictionary); this means working on building speed and fluency and on learning to recognise at least 10,000 words in the new language. Learners can build speed and fluency by learning vocabulary systematically and by doing lots of easy (‘extensive’) reading. Learners will also read better in their second language if they learn about text characteristics, and if they know how to handle a variety of strategies for getting meaning from texts. Background knowledge about the second-language culture will make comprehension easier as well.

Table of contents

1. Introduction

In order to understand how L2 learners handle reading, one must first establish how they may be considered to process text at a cognitive level. After establishing that reading depends on parallel 'top-down' and 'bottom-up' processing, we go on to consider the main factors that contribute to reading fluency and vocabulary acquisition; and we also cover the related aspects of teaching and learning practice that may contribute to developing these aspects of L2 reading skills. We then consider the role played by top-down factors: self-monitoring, awareness of text types and background knowledge.

2. Basic cognitive issues

'Top-down' models of reading were popular from the 1970s, the most frequently cited example of this approach being Goodman (1967): 'Reading: A Psycholinguistic Guessing Game'. In classroom practice, this widely replaced the 'bottom-up' models that had been in use, and in which the reader deals with letters, words, and sentences in rank order, each step depending on the preceding one (see for example Gough 1972). Top-down models assume that the reader interrogates the text rather than processing it completely, getting meaning by comparing expectations to a sample of information from the text.

Exclusively top-down models have been definitively falsified by empirical studies showing that skilled readers do not sample portions of the text, but rather process the entire text, rapidly and automatically. In currently accepted evidence-based models of reading, bottom-up processes (like word recognition and lexical access) and top-down processes (like integrating background knowledge) proceed in parallel (Rumelhart 1977, Stanovich 1980). Both sorts of processes are vital to skilled reading. The degree to which lower- and higher-level capacities may come into play in this interactive process may well differ, according to the purpose for reading. Looking quickly over a text, either to decide whether to read it or for gist (skimming) or looking for specific information in a text (scanning), will not require the same mix as reading for general understanding or pleasure (q.v. below). Intensive reading (i.e. critical reading, or reading to learn) will require yet a different balance between the two processes (Urquhart & Weir 1998:46). Higher-level processes may be best developed through intensive reading, for this focuses on aspects of the text and on the conscious use of strategies. The typical intensive reading text will be just above the level at which the reader can easily read.

However, for all types of reading, the reader needs automaticity both of word recognition and of lexical access (recognising the word so as to find its meaning in memory, and silently activating its pronunciation). When a word is recognised, it enters the phonological loop of working memory, becoming available for consultation and integration into a mental representation of the text (Gathercole & Baddeley 1995). The faster a reader's speed of retrieval from the mental lexicon, which is linked to their level of vocabulary knowledge, the more proficient their reading comprehension will be. These features are treated below in terms of fluency and vocabulary.

3. Reading fluency and vocabulary acquisition


Many students of languages at university level will have years of experience of L2 reading, but some will be starting ab initio; and some of these will be learning a non-Roman writing system (alphabetic or logographic - to be treated elsewhere in this Guide). However, even most students doing single honours courses in European languages will probably not have had enough exposure to L2 text to have acquired sufficient fluency in basic decoding and word recognition.

Optimal rates for processing prose are around 300 words per minute; and for fluent adult readers this is constant, regardless of the difficulty of the text (Carver, 1982, 1983, 1990). Even advanced bilinguals (see Segalowitz, Poulsen & Komoda 1991) read as much as 30% slower than this in L2, employing cognitive resources that would otherwise be used for higher-level processes.

Effective means of building reading fluency are timed- and paced-reading activities, word recognition exercises, read-aloud group and pair work, re-reading activities, and extensive reading in and out of the classroom (see Grabe & Stoller 2001 for clear descriptions of these activity types).

Extensive reading - abundant exposure to printed materials at or just below a comfortable level of comprehension - is vital for the development of automaticity in low-level processing, providing as it does repeated exposure to frequent vocabulary items. Extensive reading means reading unproblematic self-chosen materials, for information and enjoyment; reading texts easy enough for dictionaries to be unnecessary; and giving feedback to the teacher only about how much was read, about general meaning and about enjoyment, not about the structure or language of the texts. Ideally, this will include sustained silent reading in class. See Day & Bamford (1998) for a clear rationale and procedures for extensive reading, and a list (page 34) of studies showing some of its benefits, both focus-oriented and collateral.

In several studies, L2 students who engaged in extensive reading over a period of time showed significantly more improvement in L2 writing skills than control groups who did not practise extensive reading (see for example Hafiz & Tudor 1990; Tudor & Hafiz 1989; Mason & Krashen 1997). A speculative explanation for this effect relates to the hypothesized unconscious acquisition of the rhetorical conventions of L2 written genres, but this has not been tested.


In order to read comfortably, skilled readers need to have receptive mastery of 95% or more of the words in a text, recognising them rapidly (Grabe & Stoller 2002:186). (Receptive mastery does not require that learners be able to use all of these words productively in their speaking and writing.) Contrary to previous assumptions, good readers do not use context to infer meaning so often as less-skilled readers do; they do not need to, because they know the words (Juel 1999). Moreover, guessing word meanings is a low-yield strategy: in Parry's (1991) study, L2 university students only guessed meanings correctly 50% of the time.

It has been shown that L2 vocabulary difficulty cannot be compensated for by topic familiarity (Freebody & Anderson 1983). So how many words do L2 readers need to know? Hazenburg & Hulstijn (1996), in a careful study of L2 Dutch university students and their reading needs, demonstrated that a minimum of 10,000 headwords was necessary to understand 90% of the vocabulary needed in university study in Dutch L2, proper names accounting for another 5%. (A headword, e.g. communicate, is assumed to give access to derivatives, e.g. communication, uncommunicative.)

How can we help university L2 learners to gain mastery of the 10,000 or so words they will need to become fluent L2 readers? One way is by overt targeted teaching. Grabe & Stoller (2002:79) recommend focusing on the 2,000 to 3,000 most common words in a language 'as an essential foundation for word-recognition automaticity, and then [focusing] on vocabulary that is appropriate to specific topics and fields of study'. To determine which are the most common words, computerised corpora are now available for a number of languages; the European Corpus Initiative's Multilingual Corpus 1 (1994) contains 27 languages, mostly European; and an email query to the LINGUIST Discussion List ( will yield information about corpora of other languages. (For further discussion see the article Linguistics via the Web in this Guide.) Cooper (1984) found that what distinguished skilled from less-skilled L2 academic readers was vocabulary, and notably their level of knowledge of (1) sub-technical words used in academic discourse, and (2) common sentence connectors. These therefore merit overt teaching, as do other devices that signal the structure of a text - teaching of conjunctions and other discourse markers has been shown to be productive (Grabe & Stoller 2002, 80-81).
(On targeted vocabulary teaching in generic or specialised language programmes, see the article on IWLPs in this Guide)

There is an argument for spending some class time in every lesson on vocabulary activities, e.g. making students aware of the value of learning the most common and useful words of the language and helping them to plan how to do so; exploring various ways of recording vocabulary (in lists or networks; with translations, antonyms and synonyms, contextualised in sentences, in sense and/or function categories); studying patterns of word formation such as using common suffixes and prefixes; teaching effective use of monolingual and bilingual dictionaries (see also the articles Dictionaries and Vocabulary Acquisition in this Guide).

4. Higher-level skills

Schoonen, Hulstijn & Bosser (1998) studied Dutch learners of English: increasingly, as proficiency grew, metacognitive knowledge made a stronger contribution to L2 reading comprehension skill. '[K]nowledge of text characteristics, knowledge of reading strategies, and, to a lesser extent, knowledge of reading goals [are] important domains' (page 98). On this basis Grabe & Stoller (2002:148) assert that 'metacognitive instruction in text structure and reading strategies is likely to ... support more advanced reading'.

Less-skilled L2 readers do not necessarily have fewer strategies than skilled readers, but they are less able to choose the appropriate strategy for the problem at hand. Teachers can help learners to become aware of what strategies they use when reading, of what other strategies are available to them, and of how to use strategies selectively (Anderson 1991). Kucan & Beck (1997) successfully helped L1 readers work with 'inconsiderate texts' (texts that were difficult for any reason, from poor organisation and difficult vocabulary to unfamiliar cultural assumptions) by giving them tools for addressing the text and the author's purpose critically.

A 'think-aloud' technique, in which the teacher goes through the text demonstrating the dialogue between the critical reader and the text, is an excellent means of introducing students to new strategies. Students can then do think-aloud demonstrations for one another in pairs or groups. (See Brown et al. (1996) for an account of think-aloud techniques for developing strategy use.)

The L2 reading comprehension threshold and metacognition

In fluent L1 reading the metacognitive ability to monitor one's own comprehension is somewhat automatic, though skilled readers do tend to monitor reading consciously more often than less-skilled readers (Block 1992). Even skilled L1 readers, however, fail to monitor their reading in L2 successfully until they have crossed the 'L2 reading comprehension threshold'. This is a well-documented phenomenon in which literate L1 readers do not transfer their higher-level reading comprehension skills to L2 until they reach a certain threshold of proficiency in the new language (see a meta-analysis of findings in Bernhardt & Kamil 1995). The threshold, which has been described for several pairs of languages, occurs somewhere in the intermediate proficiency range. What seems to get transferred when the threshold is crossed is the ability to build reliable mental representations of text (Walter forthcoming).

The value of translation

Translation, previously seen as an unhelpful strategy, has been revalidated: Kern (1994) showed that, when reading difficult texts, skilled L2 readers used mental translation into L1 to help maintain concentration and to keep information active while addressing a textual problem.

Background knowledge

Background knowledge has an effect on comprehension. The idea that this knowledge is codified into mental schemata (as proposed by e.g. Carrell 1983) lacks empirical confirmation. Nonetheless, 'it has been shown that even across passages on the same general theme, which had identical structure and syntax and very similar vocabulary, the more familiar version is better recalled' (Alderson 2000:43). This is true a fortiori when the text is situated in an unfamiliar culture: Steffensen et al. (1979) gave texts about weddings to L1 readers from India and North America and observed that cultural familiarity or the lack thereof led the readers to make numerous inferences about the events and situations in the text. Giving L2 readers access to information about the L2 culture can be an important way of helping them with reading comprehension.

Text types

Knowledge about the rhetorical organisation of various L2 text types will also influence comprehension. Different text types are organised differently, and fluent L1 readers know, often unconsciously, what sort of information to expect, and where it will occur, in each type. The rhetorical conventions governing the organisation of a text type can be very different from one language to another, and while these might be acquired unconsciously by voracious L2 readers, being taught the new conventions explicitly can facilitate L2 comprehension for all readers (Carrell 1985, 1992).

5. Concluding observations

This article has described aspects of L2 reading for which classroom practice can be informed by empirical evidence: for example, vocabulary and fluency work, including extensive reading, to promote automaticity of word recognition and lexical access; instruction in deployment of strategies; provision of background knowledge and of the conventions of L2 text types. However, many questions remain to be answered. For example, the language teaching community would benefit from knowing in detail which aspects of skilled reading can be taught explicitly and which are most efficiently learnt through extensive exposure to L2 texts.

Much valuable information about L2 reading has come from the language assessment research community. One interesting discussion in that field is whether to base reading assessment on the operationalisation of a construct of reading; or on a target language use perspective (Bachman 1999; see the discussion in Alderson 2000:138-201).

Further reading

The literature on reading is vast, and this article has given only a brief overview of main areas of possible interest. For clear and more extensive accounts of the state of the art, see Grabe & Stoller 2002, Alderson 2000, and Urquhart & Weir 1998. For a well developed account of good instructional practices in L2 reading (focused on English L2 but easily transferable), see Grabe & Stoller 2001.


Alderson, J. C. (2000). Assessing Reading. Cambridge: Cambridge University Press.

Anderson, N. J. (1991). Individual Differences in Strategy Use in Second Language Reading and Testing. The Modern Language Journal, 75:460-72.

Bernhardt, E. B. & M. L. Kamil (1995). Interpreting Relationships between First Language and Second Language Reading: Consolidating the Linguistic Threshold and the Linguistic Interdependence Hypotheses. Applied Linguistics 16, 1:15-34.

Block, E. L. (1992). See how they Read: Comprehension Monitoring of L1 and L2 Readers. TESOL Quarterly 26, 2:319-43.

Brown, R., M. Pressley, P. Van Meter & T. Schuder (1996). A Quasi-experimental Validation of Transactional Strategy Instruction with Low-achieving Second-grade Students. Journal of Educational Psychology 88:18-37.

Carrell, P. L. (1983). Some Issues in Studying the Role of Schemata, or Background Knowledge, in Second Language Comprehension. Reading in a Foreign Language 1, 2:81-92.

Carrell, P. L. (1985). Facilitating ESL Reading by Teaching Text Structure. TESOL Quarterly 19:727-52.

Carrell, P. L. (1992). Awareness of Text Structure: Effects on Recall. Language Learning 42:1-20.

Carver, R. P. (1982). Optimal Rate of Reading Prose. Reading Research Quarterly 17, 1:56-88.

Carver, R. P. (1983). Is Reading Rate Constant or Flexible? Reading Research Quarterly 18, 2:190-215.

Carver, R. P. (1990). Reading Rate: A Review of Research and Theory. New York: Academic Press.

Cooper, M. (1984). Linguistic Competence of Practised and Unpractised Non-native Readers of English. In J. C. Alderson & A. H. Urquhart (eds), Reading in a Foreign Language, 122-35. Harlow: Longman Group Ltd.

Day, R. R. & J. Bamford (1998). Extensive Reading in the Second Language Classroom. Cambridge: Cambridge University Press.

Freebody, P. & R. C. Anderson (1983). Effects of Vocabulary Difficulty, Text Cohesion and Schema Availability on Reading Comprehension. Reading Research Quarterly 18, 3:277-94.

Gathercole, S. E. & A. D. Baddeley (1995). Working Memory and Language. Hove and Hillsdale NJ: Lawrence Erlbaum Associates.

Goodman, K. S. (1967). Reading: A Psycholinguistic Guessing Game. Journal of the Reading Specialist 6:126-35.

Gough, P. B. (1972). One Second of Reading. In F. J. Kavanagh & G. Mattingly (eds), Language by Ear and by Eye, 331-58. Cambridge, MA: MIT Press.

Grabe, W. & F. L. Stoller (2001). Reading for Academic Purposes: Guidelines for the ESL/EFL Teacher. In M. Celce-Murcia (ed.), Teaching English as a Second or Foreign Language (3rd edition), 187-203. Boston: Heinle & Heinle.

Grabe, W. & F. L. Stoller (2002). Teaching and Researching Reading. Harlow: Pearson Education Ltd.

Hafiz, F. M. and I. Tudor (1990). Graded Readers as an Input Medium in L2 Learning. System 18, 1:31-42.

Hazenburg, S. & J. H. Hulstijn (1996). Defining a Minimal Receptive Second-language Vocabulary for Non-native University Students: An Empirical Investigation. Applied Linguistics 17:145-63.

Janopoulos, M. (1986). The Relationship of Pleasure Reading and Second Language Writing Proficiency. TESOL Quarterly 20, 4:763-68.

Juel, C. (1999). The Messenger may be Wrong, but the Message may be Right. In J. Oakhill & S. Beard (eds), Reading Development and the Teaching of Reading, 201-12. Malden, MA: Blackwell.

Kern, R. G. (1994). The Role of Mental Translation in Second Language Reading. Studies in Second Language Acquisition 16:441-61.

Kucan, L. & I. Beck (1997). Thinking Aloud and Reading Comprehension Research: Inquiry, Instruction, and Social Interaction. Review of Educational Research 67:271-99.

Mason, B. & S. Krashen (1997). Extensive Reading in English as a Foreign Language. System 25, 1:91-102.

Parry, K. (1991). Building a Vocabulary through Academic Reading. TESOL Quarterly 25:629-53.

Rumelhart, D. E. (1977). Toward an Interactive Model of Reading. In S. Dornic (ed.), Attention and Performance 6:573-603. New York: Academic Press.

Schoonen, R., J. Hulstijn & B. Bossers (1998). Metacognitive and Language-specific Knowledge in Native and Foreign Language Reading Comprehension: An Empirical Study among Dutch Students in Grades 6, 8 and 10. Language Learning 48:71-106.

Segalowitz, N., C. Poulsen, & M. Komoda (1991). Lower Level Components of Reading Skill in Higher Level Bilinguals: Implications for Reading Instruction. In J. H. Hulstijn & J. F. Matter (eds), Reading in Two Languages. AILA Review, 8:15-30. Amsterdam: Free University Press.

Stanovich, K. E. (1980). Toward an interactive compensatory model of individual differences in the development of reading fluency. Reading Research Quarterly 16:32-71.

Steffensen, M. S., C. Joag-Dev & R. C. Anderson (1979). A Cross-cultural Perspective on Reading Comprehension. Reading Research Quarterly XV, 1:10-29.

Tudor, I. & F. Hafiz (1989). Extensive Reading as a means of Input to L2 Learning. Journal of Research in Reading 12, 2:164-78.

Urquhart, S. & C. Weir (1998). Reading in a Second Language: Process, Product and Practice. London: Addison Wesley Longman Ltd.

Walter, H. C. (forthcoming). The L2 reading comprehension threshold is linked to mental representations of text and to L2 working memory. Forthcoming in Applied Linguistics.

Related links


European Corpus Initiative (1994). Multilingual Corpus 1.

Referencing this article

Below are the possible formats for citing Good Practice Guide articles. If you are writing for a journal, please check the author instructions for full details before submitting your article.

  • MLA style:
    Canning, John. "Disability and Residence Abroad". Southampton, 2004. Subject Centre for Languages, Linguistics and Area Studies Guide to Good Practice. 7 October 2008.
  • Author (Date) style:
    Canning, J. (2004). "Disability and residence abroad." Subject Centre for Languages, Linguistics and Area Studies Good Practice Guide. Retrieved 7 October 2008, from