Index of latest magazines

Index of selected articles

Subscribe Now

Guidelines for publication

Home What and Who we are The Conference Training Courses The Magazine Members

La Revista GRETA

Selected Interview from Volume 9, nº 1

Words, Words and Knowledge of Words

An interview with Norbert Schmitt

Bryan Robinson

American-born Norbert Schmitt teaches at the University of Nottingham in the UK. In recent years he has published a number of research articles and books on topics in the field of vocabulary learning and teaching. These include Vocabulary: Description, Acquisition, and Pedagogy (1999) co-edited with Michael McCarthy, and Vocabulary in Language Teaching (2000) both from Cambridge University Press. His partner, Diane Schmitt, is a leading member of IATEFL and coordinates the Research SIG. Norbert and Diane presented sessions at the XVI Jornadas in September 2000, and after the conference was over, Norbert kindly spent some time talking to Greta about his EFL career, his work on vocabulary, and applications of this in the classroom.

Greta: How did you get involved in the area of vocabulary?

Norbert Schmitt: I don’t know why I got into vocabulary - who knows why people specialize in what they do? But at some point during my MA in ELT, I found myself thinking more and more about lexical issues.

My PhD was about vocabulary acquisition. I was trying to look at how individual words are learned. There’s been a lot of research on vocabulary size in terms of how many words are learned in a classroom, or how many are learned in a year, or how many words you know: for example, native speakers add around 1000-2000 words a year to their lexicon when they are growing up. But there hasn’t been very much research on tracking individual words, and how words go from not being known at all, all the way through partial knowledge ‘I can use/understand the word receptively’, to full control where ‘I can use it in any situation I want’ and ‘I can use it appropriately’.

Greta: So that’s where the different types of word knowledge come in?

N.S.: Yes, that was the particular perspective I used in this research. I drew on Paul Nation’s 1990 taxonomy of eight different types of word knowledge: the written form, spoken form, collocation knowledge, grammatical knowledge, meaning knowledge, association knowledge, frequency and appropriateness.

I tracked three advanced learners who were PhD students at the University of Nottingham. I gave them eleven words: some of the words were totally new to them; some were words they’d seen before and maybe used a little bit; and other words were relatively common and they knew them pretty well. In that way I could track words at different degrees of knowledge, and see how the learners’ knowledge of these words changed over time. They usually learned a little bit about the unknown words by the end of one year, but from the words that were partially known, I could measure which of the different types of word knowledge they actually improved on.

I looked at meaning, collocations, associations, spelling, grammatical parts of speech, and the derivations within the word family. For example, one thing I explored was whether the learners could use a target word such as ‘stimulate’ as a noun ‘stimulation’, could they use it in the verb form ‘stimulate’, could they use it as an adjective ‘stimulative’? This was one area they didn’t really improve on much at all.

Greta: Is that because they were using them in oral contexts?

N.S.: Well, the research was designed to track the changes in word knowledge, so I can’t say much

about the underlying causes. But one could speculate that maybe their written control was better than their oral control. I do know that the learning was more or less incidental: only one of the three people said that they actually looked up the words in a dictionary; the others said that they just read books and that they thought about the words only when we met every six months.

In any case, learning the full word knowledge specification for a word takes a long time, and for most words, learners only possess partial knowledge. I’ve just finished a project with a colleague in California, Cheryl Zimmerman, and what we found was that even advanced students in English-medium higher education have gaps in their knowledge of the derivative forms of a word family. They know one or two word forms – usually the verb and the noun forms - but the adverb and the adjective forms are harder and it’s actually rare for a learner even at that level of proficiency to know all of the four forms. What does that mean in a practical sense? Well, learners usually have a partial knowledge of words, which means that they have a limited range that they can use them in, so maybe they know a word like ‘stimulate’, and maybe they can use it as a verb, maybe as a noun, but if they need to use it in an adjective sense they might not be able to; so they’ll either use a wrong word form, or they’ll avoid using it altogether, or maybe the word is not even an option in their minds.

Greta: From the point of view of the classroom teacher, what is the way in which those gaps can be filled? Is it just a question of exposure to word forms in different contexts? Is it a matter of using manipulative exercises?

N.S.: That’s a good question! It seems to me that the word knowledge framework can be useful in developing reasonable answers to questions like these. First, you ask what learners know about a word, and what you want them to know. Can they spell the word? Do they know the appropriate register? Do they know the grammatical constraints? Then you match the knowledge gaps with an appropriate pedagogical technique. For example, some of the types of word knowledge, like meaning, are amenable to conscious learning. You can actually sit down, try to memorize meaning, try to match the word form with the meaning, try to make mental links in your mind. You can also work on spelling consciously, because once you know the spelling of some words, you can use that knowledge by analogy to learn other similarly-spelled words. So some kinds of word knowledge are more amenable to a conscious ‘Let’s learn this’ approach. But there are other kinds, like collocations, that are not. We can teach a few collocations to certain words, but it’s always going to be limited; there are millions of collocations. Teaching all of the collocation partners for a word is just not feasible, and so this type of knowledge is probably best learned from exposure.

Greta: But collocation is vital because it’s what makes texts sound native-like and natural.

N.S.: A certain word has certain partners, but that doesn’t necessarily help you when you take a different word. There’s no obvious system and each collocation has to be learned individually because you can’t depend on rules to help you. Collocation, the frequency of a word, the register of a word are all use-based types of knowledge which are hard to teach. You can help your students to a certain extent, but for them to really get a sense of how and where to use a word, they basically have to learn from exposure and by building their intuitions.

Diane Schmitt: But I still think there’s a role for teaching in that. I largely teach English for Academic Purposes, but I do bring up the different types of word knowledge and my

students are not aware of many of them. If I ask ‘What does it mean to know a word?’, when we get down to it, the students normally say something about pronunciation, meaning, ...

Greta: Dictionary meanings?

D.S.: ...dictionary meanings, and maybe formality. So I think there is a teaching role for continual awareness raising of the other things, in that students need to be aware that they should be looking out for patterns and collocations.

N.S.: We have these different kinds of word knowledge, so teachers need to be aware of all of them, and make their students aware that they exist. Obviously, students won’t learn all of the word knowledge information for each word instantly. Collocations will take a long time to learn, but students have to be aware that collocations are there, and that they are required to really master a word.

Greta: But it’s also important for teachers to help learners become more aware of the need to be systematic in the way they learn, in the way they record collocations, and that they recycle them.

N.S.: Yes, and that brings us back to the idea of intentional versus incidental learning. I think the consensus nowadays is that we need both. Before, it was `We have to teach it explicitly if it’s going to be learned’, then we got away from that and it became top-down, communicative language teaching: `It’ll just be learned incidentally by using the language.’

Greta: …that would be the ‘intuition’ you were talking about…

N.S.: Right, but now people realize that we need both approaches because they do different things. We need explicit instruction to teach at least the meanings and the starting partial knowledge of the most frequent words in English. You need to use the most frequent 2000 words of English to do virtually anything in the language, so there’s a role for explicit instruction in teaching them, but there are limitations in the number of lower-frequency words which can be taught. But it’s not just that limitation, although it’s the thing that’s usually been cited. Explicit teaching also has limitations in the types of word knowledge it can actually address, as I mentioned before. So we need both the explicit instruction approach and the exposure approach because they address different kinds of word knowledge. Also, we need to have the exposure approach because it helps to recycle the words that are already partially learned, perhaps from explicit teaching. Without this consolidation, the words initially learned through explicit teaching may be forgotten and the effort wasted. Reading is an obvious source of exposure. but you need to have a lot of it just to recycle and consolidate the partially-learned words, let alone to be introduced to new ones.

Greta: Vocabulary is a key element in testing performance at different levels of language learning.

N.S.: Yes, but here the question is what type of vocabulary test is used. Traditional vocabulary tests typically measure written form and meaning knowledge of pre-selected words. These features give the examiner control over the words to be tested and conveniently produce answers which can be marked as correct/incorrect. But work is also being done on test formats where learners produce a piece of text and the words produced are analyzed. The advantage here is that we get a better idea of a learner’s productive knowledge, and might even get some idea of the learner’s knowledge of appropriateness, grammatical control, and so on. The problem is that it is very difficult to predict which words will show up in any particular learner’s examination output.

Greta: But students don’t necessarily show their best in exam contexts...

N.S.: No, so perhaps a greater weighting should be given to the day-to-day formative assessment which teachers do in the classroom as part of their normal teaching. But I do think there’s a case for vocabulary testing which can help us to judge a learner’s vocabulary size. In classroom situations, a vocabulary size score can be very informative for teachers, particularly for placement or diagnostic purposes. Beyond this, I think it is useful to have a certain amount of vocabulary testing simply for the washback effect. If we as teachers think vocabulary is important, then we should be testing it at least to some degree so that learners realize that it is a learning priority. It goes back to the old truism: If you want learners to think something is important, put it in a test.

Greta: When you’re measuring over a longer period, intermediate students are the most frustrated because they are the ones who don’t see their progress. Teaching intermediate students, one of the most difficult things is to encourage them to read. Even a test is sometimes not enough because it doesn’t motivate sufficiently to counter the frustration factor. People will spend hours with a graded reader and a dictionary and this makes them more and more frustrated.

D.S.: One thing is that for intermediate students, the issue is not necessarily learning more words, but learning more about the words that they know. Most students measure themselves against their peers, and unconsciously against their Ll reading. If you just constantly teach only new words, they’re probably being forgotten nearly as quickly as they’re taught. It’s more important to go back and consolidate these words than to teach another new word. It’s worth the investment to go back, and that is something that most teachers might find that they don’t do.

Greta: Coursebooks don’t let you do that, and except in a few contexts there isn’t any framework that helps teachers to review and to work backwards. In ESP, there are publications such as Peter Wilburg’s filofax-like Business English were he offers mind maps for learners to go back and to draw and redraw the mind map.

N.S.: A lot of people criticize word lists. But I think the problem lies in the way teachers have used them, rather than in any inherent

shortcomings. Teachers might ask their students to memorize some words for homework and then test them the next day. But often that is the only time the teachers work with those words. No wonder word lists don’t work if used like this. But words lists make a lot of sense if we conceive of vocabulary learning as an incremental process. Learners can study the lists at home on their own time and get an initial exposure to the new words. The teacher can the use the precious classroom time to consolidate the words, and expand the learner’s knowledge of them. This doesn’t necessarily mean separate vocabulary exercises. Teachers have to make up examples to teach grammar or any other linguistic skill, so why not use the vocabulary which students were exposed to the night before? Many language games are also great for recycling vocabulary. So although coursebooks might be restrictive in the way they present vocabulary, I think there is a lot of scope to include vocabulary recycling in the supplementary materials. In any case, most course books set vocabulary goals that are way too low.

D.S.: In Japan, there’s a list of 580 words to be learnt after three years of junior high school, which is unrealistically low.

Greta: 2000 words would be about three years work in Spain.

N.S.: If we know that 2000 words forms the lexical basis for general oral communicative competency, I would advocate priming the vocabulary pump early. Instead of teaching grammar at the beginning, I would make the first year primarily a vocabulary-learning year. In essence, I think it makes sense to put the lexical resources in place before we go on to work on the other aspects of linguistic knowledge. For example, think of how much easier it would be to understand grammar if you could read a wide variety of sentences illustrating the grammar point, and knew all of the words and how their meanings related to each other. My hunch is that the time ‘lost’ initially by focusing primarily on vocabulary would be rapidly made up because grammatical knowledge and pragmatic learning would advance so much more quickly once they were focused on.

Greta: So you’re talking about a beginner’s course, where would primary school learners start?

N.S: Let’s be radical, let’s say we’ve got two years, about 80 weeks’ schooling, with two sessions a week. Now, if we divide 2000 words by eighty sessions a year that gives us 12 new words a session over two years. The students would initially learn those words as homework, and then the teachers would recycle, consolidate, and expand upon the words in the classroom. The teachers could use receptive tasks for recently-learned words, and move to productive tasks for words which had been recycled a number of times. The focus would be on the most frequent words, but the words wouldn’t be selected solely on a frequency basis. Words necessary for classroom management, like page. pencil, and eraser would be taught early so that classroom instructions could be given in English as soon as possible. The very frequent grammatical words wouldn’t be taught explicitly but the students could start learning them incidentally from the teacher’s language input. Make it fun, make it engaging, make it non-threatening.

I think such an approach would be motivating because 1) learners expect to have to learn vocabulary and 2) they would be able to start communicating in a meaningful way sooner. If you teach the first 2000 words in two years, then in three years students would be able to communicate in L2 environments. These primary school learners would be able to walk into a restaurant and buy a hamburger. They wouldn’t be perfectly correct grammatically, but they would get their message across.

Having this vocabulary in place would also accelerate the development of the four skills. The more vocabulary you have and the more automatic it is, the more mental resources are freed up for other elements of language processing, such as listening or reading.

Greta: So when can we expect this revolutionary primary course to appear?

N.S.: When the Spanish ministry reads this interview...!


Nation, I.SP. 1990. Teaching and Learning Vocabulary. Massachusetts: Newbury House.

Wilberg, Peter



C/Natalio Rivas nº 1, 1º izq Oficina 2

Teléfono: 958 20 20 11
Fax: 958 28 32 46

Página principal

Designer & Webmaster
Isabel Pérez Torres