The Corpuswocky

Or maybe “The Jabbercorpus” has a better ring to it.

It’s well-known amongst teachers of FCE classes that learners, having made it to B2 level (and if they haven’t, they’re really in the wrong class, as evidenced by the plethora of inexperienced teachers desperately trying to work out how to teach all forms and uses of, say, ‘used to’ from scratch) and sitting in their FCE lessons, will be expected to take the language apart and put it back together again. That means filling in the grammar words, manipulating the morphemes and starting to master both the structure and content of English.

All those years of “What class/type/kind/etc of word is it? (Blank stares) Noun? Verb? Adjective?”. Those poor so and sos thought that we, as L1 English speakers (please don’t lynch me if I inadvertently refer to them as native speakers) had to learn these word classes to speak our own language. Truth be told, we were taught them…sort of…apparently, but I was well into my 20s before I could confidently state the difference between a noun and a verb, and I’m fairly confident I was using both of them correctly before that.(Since then, I have even learned the verb of ‘noun’. It’s nominise. And shame on you if you went to look that up!)

No, no, the whole ‘noun’ or ‘verb’ malarkey was so,when they got to the FCE exam, they could apply exam strategies to Reading part 3 and determine if the missing word was a noun, verb, etc.  Of course, you’ll see this strategy in classes everywhere, and why not, if there’s nothing to say it doesn’t work? Maybe it does work, maybe it doesn’t. Even if it has no effect on language acquisition or exam mastery, there are grounds for using it (lowering affective filter by providing learners with a familiar task, seemingly justifiable link between language development and test proficiency, etc). The problem is, when getting learners to apply these distinctions as part of their exam training, it’s not at all clear whether they put the right word in the gap because they already knew which word to put in their, or whether they applied their knowledge of word families to determine

i.) there is a 1st person singular subject pronoun before the gap

ii.) The word after the gap ends in ‘-ly’, suggesting it is an adverb

iii.) The missing word must be a verb

Does it matter if the student only wrote the missing word because they knew the word?


The whole point of the FCE is for the candidate to demonstrate how close they are to language proficiency. As L1 speakers of English, we don’t need to fart about analysing the content, form and structure of our own utterances, lexical items, clauses, etc, so it makes perfect sense that a sound pedagogical EFL system would mimic this as closely as possible. The candidate knows the correct word, so obvious they know the language, at least well enough to pass the exam.


The FCE is designed to test how well candidates understand the language. Unfortunately, since we are not mind readers (presumably: there might be a chakra waving witch out there typing up the end-of-course reports before the course gets underway) we have limited means of checking understanding. It’s no secret that many activities designed to ‘check understanding’ are, in reality, activities designed to get the learner to do some sort of productive skills activity that, in some roundabout fashion, reflects or indicates that they understand the content of whatever it is they were doing. Case in point, I had a B1 student with terrible speaking skills and, seemingly, terrible reading skills. His reading was poor, so the logic of EFL goes, because he couldn’t explain the text, or answer questions about the text, in the L2.

Then he read the entire text out in Spanish without missing a beat. The text, by the way, was from Unit 2 of English File 3rd Ed.

Anyway, back to the point, we don’t check understanding of receptive skills, we check how well learners can use their productive skills to show their understanding of their receptive skills. And there’s a similar idea coursing through the veins of the FCE. Candidates aren’t tested on their understanding of English. They’re tested on their ability to use their (arguably declarative) knowledge of English to show their (arguably procedural) knowledge of English.

Aside from that, the idea that L1 and L2 acquisition are congruent is, to say the least, controversial. It is not obvious that we would learn an L2 in the same way as an L1. Some people do, but I’m always happy to allow for exceptions (so we get more cases of “Now, I know that your flask of water boiled at 102C, but in your report, write 100C” in high-school chemistry classrooms, and fewer cases of rewritten rulebooks with every inexplicable result). Hence, L2 learners probably can’t rely on a strong communicative approach to learn the language. They need rules, patterns, etc. Research indicates that learning type-patterns has more learning gains than learning token items. That is, if learners understand the rule or logic behind the item they’re grappling with, they’ll pick up related items much quicker.

It follows, then, that knowing why the noun goes in that gap, and therefore knowing the syntax, the structure, of the sentence, is a great deal better for the learner than knowing which word in particular fits in that gap in particular.

Evidently, then, we need, as teachers, to be able to distinguish between learners who just know which item is correct, and those who know why it is correct. In my initial year of teaching, I was introduced to the Jabberwocky as a means of teaching FCE. A few years after that, I was introduced to the idea of using corpora in the EFL classroom. Needless to say, the two go brilliantly hand-in-hand.

Traditionally, learners read the poem and determine what word class each of the nonsense words are, using the structure of the text to help them. An evident issue, here, is that the language is already quite difficult, with some archaic items, without throwing in a load of nonsense items. So, the teacher has to find a way to preteach or check vocabulary, which takes up lesson time, and risks the usual pitfalls of preteaching vocab (the learners might already know it, they might know words that seem more difficult, they might not know words that seem less difficult, etc).

Having trained the learners to use the BNC/COCA corpus (or any one that you feel happy to use), learners read through the poem and identify words they think are nonsense words. As a bonus vocab-building exercise, they can identify any words they don’t know (which would, by design, include nonsense words), and enter them into the corpus. Any that bring up zero or very few results are likely to be nonsense words. If doing the bonus-vocab malarkey, learners could spend some time trying to work out how to use the unknown-yet-not-nonsense words by looking at the contexts in the corpus. It makes for a fun-filled (e.g.) 10 minutes of class time.

Once learners have identified the nonsense words, they can then use the structure of the text to work out the word class of each word. As a follow-up, you could give them an FCE Reading part 3 test containing nonsense words. Easy to set up – Just take one already made, and replace the stems to the right of the text with words you invented.

This approach is more learner-centred than the traditional one, and has the added bonus of providing corpus training to the learners. The benefits, then, extend beyond the lesson, and beyond FCE.

It is, of course, is far from sufficient to getting them to pass the exam, but it does allow you, as teacher, to check their understanding of the language, and for them, as learners, to demonstrate that they do know how to use a noun, a verb, etc.


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