Scheffler on The Lexical Approach

In January 2015, ELTJ published a commentary with the title Lexical priming and explicit grammar in foreign language instruction, which provoked a reply and counter-reply in subsequent issues. Scheffler (2015) argues that, pace Hoey’s theory of lexical priming, “lexis should be subordinated to grammar in FL teaching.”  Scheffler reminds us that Hoey sees lexical priming as “the mechanism that drives language acquisition”; that the successful language learner recognises, understands and produces lexical phrases as ready-made chunks; and that, consequently, teachers should concentrate on vocabulary in context and particularly on fixed expressions in speech. Scheffler’s reply is that mastery of lexical associations takes too long to be a viable objective for classroom-based foreign language learning and that grammar-based teaching is more efficacious.

According to Scheffler, in order to reach proficiency through learning lexical chunks, EFL learners have two options: either they use the same subconscious mechanism that operates in L1 acquisition, or they consciously apply themselves to the study of appropriate language material. As to the first option, studies of first language acquisition cited by Scheffler show that by the time they’re five years old children have encountered more than twelve million meaningful utterances in communicative context.  Scheffler comments: “No classroom input can come close to this amount of linguistic data”. Regarding the second option, mastering lexical associations through the conscious study of chunks,  collocations, etc., Scheffler cites Pawley and Syder’s (1983) claim that native speakers know ‘hundreds of thousands’ of memorized sequences, and argues that “in the classroom setting, committing to memory even a small subset of these would be a daunting task.”

So, says Scheffler, given that the subconscious or conscious learning of lexical chunks is not a viable option, classroom time should be spent focusing more on grammatical systems than on lists of lexical phrases. He cites Spada and Tomita’s (2010) meta-analysis as convincing evidence that explicit grammar instruction is more effective than implicit instruction for both simple and complex English grammar structures. Scheffler concludes by suggesting that classroom EFL teaching should provide “a combination of grammatically oriented presentation and lexically oriented practice” involving explicit grammar explanations followed by practice to encourage lexical priming. Schleffer offers this example: the teacher presents and explains the present perfect as a grammatical category and then provides practice in the form of drills. The drills allow links between the most frequent lexical instantiations of the present perfect  to be established, and these links are further strengthened in communicative activities and in exposure outside the classroom.  Such procedures offer learners both explicit and implicit instruction. Explicit instruction aims at linguistic awareness, proceduralization of explicit knowledge, and lexical priming, while implicit instruction reinforce primings established in class and gradually create new ones.

I’d say that Scheffler’s suggestion is a bit of a dog’s dinner, mutton dressed as lamb, old wine in new bottles, a botched attempt to have your cake and eat it. Rather than effortlessly trot out more examples of my store of food-and-drink-related-put-downs, let me be more specific:

First, no attempt is made to evaluate Hoey’s theory of language or of SLA. What are the strengths and weaknesses of a theory which makes collocational priming the key construct in both a description of language and an explanation of how it is learned?

Next, Spada and Tomita’s (2010) meta-analysis in no way supports the view that the presentation and practice of grammatical categories is the best, or even a good, way to organise classroom-based ELT. The authors limit themselves to the claim that adult learners sometimes benefit from explicit attention to form and from opportunities to practise explicit knowledge.  Nothing in Spada and Tomita’s review suggests that organising a syllabus around the presentation and practice of pre-selected bits of grammar like the present perfect is recommendable; nothing in the review challenges the findings of studies which support the view that most explicit grammar teaching falls on deaf ears most of the time, and that teaching can affect the rate but not the route of interlanguage development.

Finally, Scheffler begins by saying that basing ELT on the implicit or explicit learning of lexical chunks is unrealistic because there’s too much to learn. That’s a good point, but why then does he end up trying to somehow cram all this lexical chunk learning into the last part of the lesson? The claim is that in a grammar-based PPP approach where communicative activities are included, the explicit grammar teaching will help lexical priming, while the communicative activities will reinforce primings and create new ones. I can see no reason for thinking that this would work. And, of course, it contradicts Hoey’s theory. Grammar views lexical items as isolated elements organised by syntax; and this is exactly the view that Hoey wants to challenge. What sense does it make to expect the explicit teaching of the forms of the present perfect to facilitate fabulous amounts of lexical priming?

Despite its shortcoming, Scheffler’s article does draw attention to the fact that no proper account has ever been given by proponents of the lexical approach of how exposure to massive amounts of what Lewis (1993) calls “suitable input” should be organised into a syllabus.  Walkley and Dellar have published coursebooks which they claim exemplify the lexical approach, but I’ve never managed to get review copies and I can’t bring myself to buy them. As far as I can gather from Dellar’s public pronouncements , he thinks teaching should concentrate on giving learners repeated exposure to the most frequent words in English in context, but important questions remain unanswered.

  1. How should the repeated exposure to massive numbers of lexical chunks be organised? Is frequency of occurrence in the biggest corpora the only criterion for the selection and presentation of lexical items, or are there others?
  2. How do teachers make classroom sessions dedicated to “repeated exposure to the most frequent words in English” interesting and motivating?
  3. How much input do learners need? How do Walkley and Dellar respond to the research findings which suggest that it’s unreasonable to expect FL classroom learners to remember even a small subset of what native speakers know? Sinclair (2004: 282) warns of “the risk of a combinatorial explosion, leading to an unmanageable number of lexical items” and Harwood (2002: 142) warns against “learner overload”, insisting that “implementing a lexical approach requires a delicate balancing act” between exploiting the richness of fine-grained corpus-derived descriptions and keeping the learning load at a manageable level.
  4. How do teachers help learners notice and store the thousands of lexical chunks which are required for a minimum level of proficiency? Put another way, how do teachers help learners turn massive loads of input into an ability to use the language for effective communication?

The fact is that, pace Hoey and Lewis, L1 acquisition is not the same as the acquisition which takes place in FL classrooms. As Granger (2011) points out, Lewis claims that “phrases acquired as wholes are the primary resource by which the syntactic system is mastered” (Lewis (1993: 95). This assertion, frequently found in the Lexical Approach literature, is based on L1 acquisition studies which demonstrate that children first acquire chunks and then progressively analyse the underlying patterns and generalize them into regular syntactic rules (Wray 2002).  But (as Wray points out in her overview of findings on formulaicity in SLA), in classroom-based L2 acquisition, learners don’t get enough exposure for the ‘unpacking’ process to take place, and as a result formulaic sequences don’t contribute, as they do in L1 acquisition, to the mastery of grammatical forms. Granger concludes that “while lexical phrases are likely to have some generative role in L2 learning, it would be a foolhardy gamble to rely primarily on the generative power of lexical phrases.” She goes on to cite Pulverness (2007: 182-183) who points to the risk of the ‘phrasebook effect’, whereby lexical items accumulate in an arbitrary way as learners get presented with an ever-expanding lexicon without being given a structural framework within which to make use of all the lexis.

Scheffler’s main objective is to defend the grammar-based syllabus against Hoey’s  suggestion that lexis should be subordinated to grammar in FL teaching. I think he does a poor job of it, but at least he gives voice once again to some unanswered questions. We’ve been waiting for an answer for a while now, and I doubt that we’ll get one any time soon. Meanwhile, perhaps we should turn our attention to the newer but somehow more interesting question of what we are to make of Hoey’s collocational priming: Can this construct lead to a satisfactory explanation of both language and of SLA?


Granger, S. (2011) From phraseology to pedagogy: Challenges and prospects. In: Uhrig, P., Chunks in the Description of Language. A tribute to John Sinclair.  Mouton de Gruyter : Berlin and New York.

Harwood, N. (2002) Taking a lexical approach to teaching: principles and problems. International Journal of Applied Linguistics 12/2: 139-155.

Lewis, M. (1993). The Lexical Approach. The State of ELT and a Way Forward. Hove: Language Teaching Publications.

Pawley, A. and Syder, F. (1983) Two puzzles for linguistic theory: nativelike selection and nativelike fluency. in J. C. Richards and R. Schmidt (eds.) Language and Communication. London: Longman.

Pulverness, A. (2007) Review of McCarthy, M. & F. O’Dell, English Collocations in Use, ELT Journal 61: 182-185.

Scheffler, P. (2015) Lexical priming and explicit grammar in foreign language instruction. ELT  Journal, 69,1, 93-96.

Sinclair, J. (ed.) (2004) How to use corpora in language teaching. Benjamin : Amsterdam.

Spada, N. and Tomita, Y. (2010) Interactions between type of instruction and type of language feature: a meta-analysis. Language Learning 60/2: 263–308.

Wray, A. (2002) Formulaic Language and the Lexicon. Cambridge: Cambridge University Press.


5 thoughts on “Scheffler on The Lexical Approach

  1. Dear Geoff,
    Thank you for giving a somewhat lame day an intellectual kick. It seems that after Scheffler discovered that the world is not flat, he thinks it hollow; perceiving some truth, he replies with more superstition.
    (I haven’t read the original article by Scheffler. My reply will take what is stated in your post.)
    1. We have an idea of what proficient speakers can do with a language.
    2. We can trace aspects of development for L1 with some accuracy. We have an idea, for example, how much input children receive.

    Given 1 and 2, L2 instruction must show significant economy if it wants to equal L1 outcomes.

    This brings up the popular question whether explicit or implicit instruction is to be preferred in language teaching.

    I am not familiar with the Spada analysis, but I remember Katherine Doughty’s review of Norris and Ortega (2000) in Handbook of Second Language Acquisition (2003), and Krashen’s reply in Explorations in Language Acquisition and Use (2003). Krashen’s review of studies that hope to provide positive evidence for implicit grammar instruction effectiveness does not end with a surprise: e.g. “this study only confirms that we can make small improvements in processes that the brain does poorly in the first place” (36).

    I found Doughty’s conclusion after reviewing the evidence for explicit instruction revealing: “…, to date little is known concerning SLA processes in instructed settings. With respect to research on type of instruction, taking together biases revealed in approach to, and duration of, L2 instruction, and the demonstrated biases in measurement, we have more properly interpreted the apparent advantage for explicit instruction as an artifact of cumulative bias” (274). Somewhat disturbing.

    Why should evidence that confirms that we can repeat rule governed behavior in a controlled environment more reliable if we have been taught the rules explicitly beforehand be used to support explicit instruction on a grand scale? If the grand scale is communicative competence, performing some learned behavior seems to be far off the mark.

    I think the dichotomy (a poor one!), explicit/implicit, is pitched against the wrong learning task:
    If we could agree that the object of language learning is the ability to put together strings of words in accordance with some explicit rule, then it seems fair to say that teaching those rules and having students practice the process is more efficient than hoping that students will “come up with” the expected outcome by mere exposure to multiple samples. But we can only settle this once we agree on the initial “if”.

    The question is not if grammar rules are learned more efficiently explicitly or implicitly. (Why would I try to elicit the present perfect instead of just showing it? I have come across this idea of “figuring out the grammar behind it” and find it puts an extra and unnecessary cognitive load on students.) The question is, I’d suggest, if lasting language competence in a second language is more successfully achieved through explicit or implicit strategies. This competence, it seems, would imply the appropriate and unrehearsed use of language in order to achieve intended purposes.

    Here we are back at chasing the rabbit. What is language? “Grammar views lexical items as isolated elements organised by syntax; and this is exactly the view that Hoey wants to challenge.” Exactly, that is the problem. Hoey and Lewis follow Sinclair’s criticism of the two pronged view of language. How grammar and lexis can be perceived as one is as challenging having a cookie and eating it. But the question is important because it is the underpinning assumption that informs method.

    Liked by 1 person

    • Hi Thom,

      Thanks for this. Two quick comments.

      Krashen’s 2003 reply to Doughty, which you refer to, is certainly worth reading. While Krashen’s theory suffers from hopelessly ill-defined constructs, I think he’s basically right: SLA is mostly an unconscious process.

      How to combine lexis and grammar in ELT is obviously a vital question. I think Sinclair was closer to the mark than Hoey and his new disciples Dellar, Walkley, Selivan,


      • Hi Geoff,

        What makes you think “SLA is mostly an unconscious process”? This is not my experience since I’ve been thinking about it and observing my own internal behaviour.

        Not only for learning languages, but for learning anything. I have to put my presence on whatever is to be learned. It may be only for a fraction of a second and I may forget about the moment very quickly, being often more interested in the result than in the process.

        As a teacher I made the same assumption about my students’ learning. It can only be an assumption because other people’s minds can’t be observed directly, but students frequently said “But you can read my mind!”

        My classes were ones where, as Thom says, “the appropriate and unrehearsed use of language in order to achieve intended purposes” was how students spent 90% of their time. That is, not learning lexical chunks nor implicit and/or explicit grammar rules.`



      • Hi Glenys,

        Thanks for this; it’s always a pleasure for me to see a comment from you.

        You ask what makes me think that SLA is mostly an unconscious process. I’m only repeating what the leading SLA scholars say, for example as summarised by Doughty and Long in both the 2003 and 2009 volumes of the Handbook of SLA. Research into SLA suggests that most of the time, learners are not consciously aware of learning formal aspects of the code, be that phonology, word families, grammar or formulaic chunks. The primary requirements for SLA are comprehensible input and engagement in communication. Having said that, SLA research also indicates that some attention to form can significantly speed up the rate of acquisition and also help learners get beyond an intermediate plateau. This amounts to a “weak interface” position regarding explicit and implicit learning, and seems to me to be consistent with what you say here about your experiences as a learner and teacher.

        I’d be very interested to hear your response to this.

        Best wishes.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s