Why MFL teachers may have to rethink their approach to foreign language reading instruction



In a previous post I already concerned myself with reading instruction; more specifically, I advocated that much reading skills instruction in UK classrooms tends to revolve around comprehension tasks. I also pointed out how such practice is detrimental to the development of reading proficiency, as it does little more than testing students on their ability to find details in a text rarely engaging students in real life tasks. As I advocated in that post, reading instruction should do much more than that: should inspire L2 learners to read independently and equip them with effective reading skills and learning strategies (e.g. using online dictionaries or knowing how to exploit the full learning potential of an online article). Moreover students should be given a degree of choice in terms of the to-be-read-texts.

Thus, foreign language teachers need first and foremost to develop their students’ self-efficacy as readers, i.e. the belief that they can read, they can comprehend L2 texts whilst making reading an enjoyable experience. To believe that s/he can be effective at reading an L2 text a learner will need more than a growth mindset; s/he will have to experience repeated episodes of success in reading tasks and a feeling of progression.

For teachers to enable students to experience such success they need to be able to understand the cognitive processes involved in the L2-learner comprehension of L2 written text. Most importantly, they need to be conversant with some recent research findings which are somewhat counter-intuitive and may force them to reconsider the way they teach not just reading, but all the other three macro-skills too.

Thus, in this post I set out to concisely outline how the human brain processes foreign language written text and explain why UK teachers may have to change their current instructional approach to reading.

Top-down processing accounts of L2 reading

Before the early 70’s the dominant theory of reading comprehension was that the reconstruction of the intended meaning of a text proceeds ‘bottom-up’, i.e. from the decoding of the smallest units (letters) to incrementally larger units, i.e.: words, then clause, then sentences, etc. Reading was seen as a linear process of recognizing one word after the other until the entire meaning of a sentence is grasped. This view of the reading process was discounted by subsequent research.

Seminal work by Goodman (1972) and subsequently Rumelhart (1980) turned reading theory upside-down (literally!). Rumelhart’s model of reading proposed that the human brain processes written text using top-down rather than bottom-up processing. His theory, rooted in cognitive psychology and still widely accepted by many scholars and researchers, posited the existence of cognitive structures called schemata which encapsulate all of our background knowledge with regards to specific life situations or concepts and consist of elaborate frameworks of objects and relations which we use to make sense of the world. For instance, when we think about a ‘haunted house’ we will activate schemata which contain all the experiences we had stored over the years about haunted houses – whether mediated by fiction or in real life. Schemata being also culturally situated, a Chinese learner’s schemata about a haunted house may be different from an Italian or a Maori’s. Schemata are, in this sense, the building blocks of cognition and “reflect the experiences, conceptual understanding, attitudes, values, skills, and strategies …[we] bring to a text situation” (Vacca & Vacca, 1999, p. 15).

To go back to the haunted house example. Imagine one is reading a short story about a haunted house (in their native language). Top down processing theory posits that that learner, in order to comprehend the text will apply his/her knowledge about the topic (content schemata) and of the genre-specific features of short-story texts (discourse schemata). Content and discourse schemata will be activated by cues in the text and applied to reconstruct the intended meaning; so, for instance the sentence ‘she saw a ghost’ will activate a range of expectations about the consequences of seeing a ghost (e.g  ‘she screamed’, ‘she ran’); one of them may match automatically what comes next or may not. Failure to understand, then, may mean that (a) the cue in the text is ineffective or (b) there is no schema in the brain which matches that cue and that text. In this sense, reading is not just a receptive skill, but require construction of meaning and cognition, in that, if we find in the text information we do not have a schema for, that information may result in the creation of a new one.

This model has been applied by L2 theorists and researchers to L2 reading, too: L2 learners would apply their content and formal schemata to makes sense of L2 text. Consistent with this theory, schemata application would not require the reader to recognize every single lexical item and morpheme. This psycholinguistic framework, viewing reading as a game of guessing, sampling, predicting, and verifying top-down hypotheses, emphasizes the role of higher level syntactic and semantic processes and minimizes the role of component and bottom-up processes.The application of schemata entailing that one does not need to decode every single word in the text, you may now understand why on your PGCE you were told that you should teach students to look for key words to enable them to understand texts. And I am sure quite a lot of you still model this strategy with GCSE, IB or A level groups.

However, just as I did, you too will have found that this inferential approach does not work all the time. In the absence of a solid and wide-ranging vocabulary repertoire, this inferential approach often leads students to making wrong assumptions about the intended meaning of L2 text. So, when someone uses the typical UK textbook with simple and predictable texts packed with known word and cognates about very familiar topics like daily routine, free time and hobbies, etc… this approach may work. With more complex and less predictable texts (e.g. authentic texts) however, this is often not the case.

Interactive models of L2 reading

Hence, more recently, the pendulum has swung back again: in recent years, scholars and researchers have re-discovered the importance of bottom-up processing in reading comprehension. New cognitive accounts of the reading process have been proposed which are widely accepted by the academic community: interactive models which recognize the synergy of top-down and bottom-up processing in reconstructing the intended meaning of L2-texts. From this perspective, it is claimed that information processing of text is driven by both bottom up and top-down information’ i.e. the processing of the physical stimuli (bottom-up processing) and the context provided by expectation and previous knowledge (top-down processing) (Carrell et al., 1998).  Prior knowledge with the help of accelerated bottom-up processes influences the perception, speed, and conceptual framework in reading processes. This view proposes multiple, independent, parallel routes simultaneously processing information with a cross-checking mechanism. Active routes are contingent upon the information presented, the individual’s knowledge and the task demands (Grabe, 2004).

In conclusion, whilst we read in the L2 our working memory activates different systems simultaneously to process the different levels of the text in an attempt to comprehend the author’s intended meaning: higher order skills (e.g. content schemata) and lower order skills (e.g. letter and word recognition).

The role of phonological processing and oral fluency in reading proficiency

One specific set of lower order skills has received particular attention in recent years: lower level verbal processing in working memory and, in particular phonological processing. There is a vast body of research evidence indicating that poor readers exhibits deficits in phonological processing and ability in general.

There are a number of reasons as to why efficient phonological processing correlates with high level of reading proficiency. Firstly, as discussed in previous posts (e.g. ‘Words in the minds’) the establishment of a complete and solid phonological representation for a word appears to be the first and the most important requisite for success in early L2 vocabulary acquisition for a young L2 learner (Segalowitz et al, 1991)

Secondly, there is clear evidence that meaning activation in Working Memory is mediated through phonology (e.g. Metsala & Ehri, 1998; Proctor, Carlo, August, and Snow, 2005). This is because when we learn a word, we encode it through its phonological representation (see my description of the role of articulatory loop in ‘Eight important facts about working memory’ for more info on this point); hence, when we identify a word, its phonological representation is automatically and very rapidly activated and precedes the retrieval of its meaning from Long-term memory. In other words, meaning activation is mediated by phonology.

Thirdly, rapid lower level verbal processing means that the brain can free up cognitive space in Working Memory during reading; this means that there is more space available for higher level cognitive processing, from the application of formal schemata (e.g. the analysis of grammar/syntax) to the application of content schemata.

Another important set of evidence points to a strong correlation between oral fluency and reading proficiency (e.g. Geva and Ryan, 1993 and Droop and Verhoeven, 2004). Droop and Verhoeven’s study is particularly interesting as the two groups they compared were equivalent at pre-test in terms of knowledge of vocabulary but not in terms of oral fluency; the group with higher levels of oral fluency were the more proficient readers. Hence better oracy skills correlated with more effective reading skills.

General implications for reading instruction in the foreign language classrooms

Effective reading comprehension being dependent on how effective top-down and bottom-up processing are performed, L2 reading instruction must concern itself with, on the one hand, training students in the skillful application of schemata; on the other, it must provide learners with masses of instruction in (a) topic-specific vocabulary and word-recognition skills; (b) metalinguistic knowledge (the ability to recognize parts of speech, noun/verb/adjective inflections, syntactic order, etc.) ; (c) discourse markers (connectives) and their function as text organizers and, much more than it is usually done, (d) phonological awareness.

It should be pointed out that of the four elements just listed, two, range of vocabulary and phonological awareness are the most widely acknowledged by research as effective enhancers of reading proficiency. Hence I strongly recommend these should take priority in our teaching of reading skills.

Practical implications

The obvious corollary of the above discussion for the foreign language classroom is that a sound approach to reading instruction must include fairly traditionally activities such as:

  • Schemata activation activities – These should include; (1) pre-reading activities activating the background knowledge students have about the topic(s) dealt with in the to-be-read text (e.g. brainstorming student assumptions as to why people smoke before reading an article on the causes of smoking); (2) activities which require students to predict / infer what comes next in a text based on their knowledge of the world, e.g. jigsaw reading or ‘guess what comes next’ tasks (whereby a very short story where only the opening line is visible to start with is displayed on the classroom screen and the students have to infer what the next line is about); (3) before reading a challenging L2 text students may be asked to read similar texts in the L1- an idea originated with Krashen;
  • Vocabulary building activities of the likes found at language-gym.com (work-outs section). These should be carried out routinely prior to engaging students in any reading task and should focus on the words included in the to-be-read texts in order to lessen cognitive load during reading; they should also be carried out after each reading task for consolidation purposes;
  • Metalinguistic tasks engaging students in contextualized structural analysis of the target text (e.g. whereby students are asked to identify to what part-of-speech category words belong to)
  • Extensive practice in the recognition of discourse markers (e.g. gap-fill or translation tasks);
  • Narrow reading tasks – these kill a lot of birds with one stone as narrow reading helps enhancing vocabulary by constantly recycling it from text to text (five or six texts should be used) and by requiring the application of the same schemata set;
  • Metacognitive retrospective tasks – students are asked to reflect on two or three main issues that impeded their understanding of the target text and what they could do to overcome them.

The most important implications for L2 reading instruction, however, refer to oral fluency and phonological skills and their link with reading proficiency. Teachers may have to focus much more than they currently do, in my experience on enhancing phonological awareness. In a previous post on ‘Listening micro-skills enhancers’ (https://gianfrancoconti.wordpress.com/2015/06/16/seven-micro-listening-enhancers-you-may-not-be-using-often-enough-in-your-lessons/)  I indicated several examples of very-easy-to-set-up activities that students enjoy, which focus on this hyper-neglected dimension of oracy.

Moreover, the evidence that higher levels of oral fluency correlate with higher levels of reading proficiency even when vocabulary range is equivalent entails that ways must be find to integrate lots of oral practice in the foreign language classroom – a pedagogical recommendation that I have often made in my posts, advocating that a substantial chunk of each lesson should be devoted to learner-to-learner oral interaction tasks (of the communicative sort).


Based on the above discussion teachers may have to rethink the way they teach reading skills. Firstly, their approach to reading skill instruction should focus systematically on both top-down and bottom-up processing skills. The two skills can be taught separately, obviously; they do not have to be explicitly integrated in every single lesson. Oral fluency, vocabulary building and phonological awareness must be focused on much more than it is currently done in foreign language lessons as they are pivotal to the development of reading proficiency.