Select Page

This is a sort of stop-gap post, while I compose another one that will summarise a session about lexical sets I saw at IATEFL a couple of weeks ago.

In a happily serendipitous, synergetic way this post is also a bit about lexis and is also something that I’ve wanted to write about for a while.

According to Thornbury, a gap fill is an exercise in which words are removed from a text and replaced with blank spaces. They look similar to cloze tests, although they lack the random choice of missing lexis that cloze tests have because, in a gap fill, the words are removed for a particular reason, e.g. because they belong to a particular word class that the person who wrote the exercise wants to focus on (2006: 30). At IATEFL in Harrogate, Andrew Walkley talked about particular features of lexical sets and how the teaching and testing of these sometimes missed a lot of potential for language work and development (and actually that certain features of lexical sets haven’t really seen the light of day in ELT – but more on that in a later post).

So what, I hear you say. ‘Gap fills are nothing to get too heated up about’. Well, yes, this is true. But I have been thinking a bit about one thing that Andrew suggested doing in his talk relating to lexical sets and activities that you could set up in class after having investigated a particular set of words and phrases. In fact it’s lead me to think of two ways you could approach a fairly standard ‘gap fill’ classroom activity scenario and (potentially) reap a lot more from the language being investigated than a simple ‘fill in the blanks so I can see what you know’ type of thing. What’s more, if you attempt this kind of thing with published ELT materials, let’s just say that it might be a little boring…

I’m far more interested in real-life language, what the methodology books tell me is authentic. What makes material authentic? And how do you make use of it as a teacher? Click here for a slightly comedic definition from English Droid. For me, it’s not necessarily taking something I rave about into the classroom, as this can kind of suck out the joy of something for me! Rather, I take authentic from a sort of ‘that’s what I’d say’ or ‘that’s what I’ve heard’ view about language.

Both in terms of this type of authenticity, tweets are often a great resource to use for language work. What’s more, the bite-size nature of a tweet (i.e. it’s not more than 140 characters) means it’s not going to flood my classroom like a published text or page from a course book  sometimes can. However, how do you go about finding juicy tweets that will serve as interesting, unusual (maybe a little bit rude!) fodder for the language classroom. Tricky. But lo! There is something that I’ve stumbled across recently, and you’ll find the link to it at the bottom of this post. I don’t want to give away the source to you just yet, as it may deny me the opportunity to show how to milk the resource for its interestingness potential.

This is also where we bring in the approach to lexis that Andrew described in his talk, by deconstructing our material (the tweets) and looking for associations with particular words  and phrases, and then doing the same, but within a sort of gap-fill framework. What you would do with your learners is this:

Ask them for associations with these words and phrases (to be altered depending on your source tweets, cultural or other knowledge your students have*, how risqué you’re feeling, etc.):

  • new dog
  • cat
  • chair
  • blog*
  • something constructive
  • PlayStation*
  • Gary Barlow*
  • David Hasselhoff*
  • angel

and so on…

You might choose to do this on the whiteboard (interactive or plain), or you could you David Warr’s Language Plant Maker thing (click on Have a go) to make something pretty like this:

my new dog

You might even like to contrast a few of them – for example, are there any marked differences in words and phrases associated with Gary Barlow and David Hasselhoff, or what about if you change new dog for old dog or old cat (this was one of the most interesting points of Andrew Walkley’s talk, blog post on it coming up, promise!)

The other angle of attack would be to take the tweets that some of the nouns above were removed from, and using them as substitution frames to be filled in with any possible language (no wrong answers at this stage – only language that can be reworked, refined, and improved). You might end up with something like these:

  • I could … with my bare hands!
  • Just because he … doesn’t make him a bad person.
  • I need to tell you: I’ve got a …
  • How is your … ?

  • I can’t relate to him. He doesn’t even …
  • Today I will … unless I get bored.
  • If you don’t find me … in two minutes, you’re dead.
  • I’ve not been able to look at … since that dream.

You can focus on the language that’s generated, ask more questions using the language that comes up, e.g. Can you relate to people who don’t X? Would he be more or less of a bad person if …? and so on (also another interesting idea Andrew presented in his IATEFL session). Then, if you really want, you can show the original tweets (NOTE – some questionable language is used, and you may be offended – I apologise if this is the case and advise you to use tweets like this in your classroom only if you feel comfortable doing so. I accept NO RESPONSIBILITY for what might happen!)

You can see the source material tweets for the language activities above here: Overheard in London on the Time Out London blog


  • Thornbury, S. (2006) An A to Z of ELT, Macmillan: Oxford
  • Walkley, A. (2014) Lexical sets, texts and vocabulary choice, IATEFL Harrogate 2014
  • Note – Andrew writes about this sort of deconstruction of language in this post on the CELT @ Westminster blog: ELT teachers shouldn’t prefer blondes!