ELT teachers shouldn’t prefer blondes!

Posted on: 9 October 2012
By:
14 Comments »
Filed under: Lexis and frequency, Materials writing

Thinking, fast and slow

I recently asked a colleague, which word was more frequent arise or blonde and perhaps some of you have been drawn to this post by the same question on Twitter. The immediate answer from my colleague was blonde. They were wrong! There is some evidence to suggest that my colleague – and perhaps you – are not alone: teachers (and students) are not very good at recognising frequency of words (see McCrostie 2007).

Examples you can think of, rather than frequency

One of the reasons for this – particularly in the heat of the classroom where teachers may be under time pressure – can be found in Daniel Kahnamen’s book Thinking, fast and slow. Kahneman’s work focuses on the limits of intuition, quick thought and expertise. He highlights a number of ‘heuristic’ biases associated with ‘quick’ thought and one of these is a so-called availability bias. This is when people under- or overestimate the frequency of an event or thing, because rather than step back and compare events rationally using a variety of tools, people replace the issue with an ‘availabilty heuristic’, which is the ease with which examples come to mind. For example, because we may remember more examples of say terrorist incidents compared to the electrocutions – terrorist attacks being much more widely reported than people electrocuting themselves – people may greatly overestimate the risk of death or injury from a terrorist attack compared the same risk from electrocution.

 

Language Teachers’ availabilty bias

This is an issue for teachers (and students) because availability or ease of recall will often be connected to categories, pictures in our minds, vivid ideas, personal experiences etc. and perhaps with words that are spoken more than those which are read or written. As a result, we are likely to overestimate the frequency that we see or hear words like blonde, pear, ski, microwave, french, crowded, purple, jumper, beard or sociable and underestimate the frequency of words such as mood, economy, provide, policy, arise, adequate, data, deposit, discipline, extent, grant. All the first set are over the level of top 5000 most frequent words  according the Macmillan Advanced Dictionary (pear, ski, sociable and French (!) are not even in the top 7500); all the second set are in the top 2500 most common.

 

As an example of this bias you might try thinking of how many blonde people you can think of compared to things that are adequate. A second related issue for teachers and students is that when we come to think of examples, the ease with which we can think of the example is likely to also be an influence.

 

Basic grammar and more complex examples and priming

Words such as those in the first group can be exemplified with the most basic representative structure of English:

            She’s blonde.

            It’s a pear.

            He’s French

            I like skiing.

            I have a jumper.

etc.

 

A lot of the words in the second set often require either additional information or more complex sentence structure that go beyond this basic sentence pattern. This is not just problematic in terms of choosing words over others with greater frequency, it also means we tend to fail in providing a better range of true collocations and grammar, which students need above these ‘representations’ of usage.

 

Finally, some of these words in the first set may also be more available to teachers of English and Language learners because of another psychological phenomena, that we have been primed by our previous teaching and learning to expect most of these words at lower levels in lexical sets on appearance or food etc. We assume that this is a legitimate choice without thinking. See future blog postings on lexical sets.

 

Getting better at frequency and giving examples

We can get better at recognising frequent words. One way is to test yourself on guessing frequency. That’s part of what we will do in other postings in this category. You can also test yourself on the website which goes with the McCrostie article above. However, simply knowing frequency is only part of the issue. Teaching less frequent words can be of use at times. It may be difficult to avoid teaching at least some nationalities such as French when this is the first kind of conversation many foreign speakers will have. Furthermore, if we can give good examples, then this often makes the teaching and time spent worthwhile because it will often involve using a lot of more frequent language. It is the combination of low frequency and single word lists which we would see as inefficient. So, unlike the compleat lexical tutor website above, our aim would be to think not only which is more frequent but also how it might be used.

 

Arise the blonde!

So as an example, let’s take blonde and arise. Why is arise more common, and potentially more useful. Secondly how could we make better use of the less frequent blonde.

Here are some possible examples for arise.

If any problems arise, or you need help, please call. You have my mobile.

If anything arises while your staying, just email me. I check it regularly.

I don’t think we need any help now, but we can always ask, if the need arises.

I think I have enough information. I can always call you, if the need arises.

It is easy to imagine a variety of situations and contexts for these sentences. You might not want to teach them to an Elementary class, because of the complexity, but there’s no reason it couldn’t appear in a text. And we might teach it productively at Pre-int or Int. Have you seen it taught?

As for blonde, when might we actually say it? It is clearly rare for us to describe people in the way we do in EFL classes (He’s tall. He has long blonde hair and a beard. He wears jeans and brown shoes, etc.). Could we at least show some alternative patterns:

Who’s the blonde man? I haven’t seen him before.

My Dad’s got blonde hair like me.

These are not really more challenging sentences that descriptions (especially if translated). However, I think these sentences are open to being changed and personalised in more interesting ways than a description such as she’s blonde. You might also imagine a fuller conversation around the sentences. They have, I think, more communicative value. Any other examples?

 

 

 

14 Responses to “ELT teachers shouldn’t prefer blondes!”

  1. Great article! I’m planning on writing something about this soon and I’ll probably link to this piece!

  2. Thanks for a really interesting blog post.

    As a corpus researcher and lexicographer, I’m well aware of the gap between intuition and frequency. I’ve also read research into the usefulness (or lack of it!) of teaching lexical sets (esp. when they’re just lists of single words). But I still think there’s an innate appeal of teaching and learning some of these basic sets, even if we don’t actually use them that much. Perhaps it’s that we feel that we have certain linguistic slots in our brain that just need filling. It would feel somehow odd to learn a language without knowing the words for certain concrete groups of things (colours, foods, nationalities, furniture, etc.).

    I know as a lexicographer that we always include certain lexical sets in ELT dictionaries even when they don’t merit being there on frequency terms, just because people expect to see them or like them. For example, even in an intermediate level learner’s dictionary you’ll usually find a set of animals like elephant, giraffe, lion and tiger, not because they crop up in everyday conversation (unless perhaps you live in certain parts of the world!) but because young learners have fun learning them :)

  3. Thanks for this post – really interesting. I agree about heuristic bias, for example, people thinking that a terrorist attack is likely, or that their child is likely to be abducted because of there being so much about these (very rare) things in the media. However, I would say that for word frequency, it’s worth thinking about how Macmillan got their statistics on words like ‘blonde’ and ‘arise’. ‘Blonde’ is definitely more frequent in our house than ‘arise’. My younger daughter is blonde, and my other daughter isn’t, and this, and other people’s degree of blondeness is often discussed. I think it really *is* very commonly used in everyday informal conversation. Macmillan’s corpus is news/politics/business-heavy, and the frequency stats reflect this.

  4. Thanks for spending the time to reply. I guess I understand why lexical sets are there – because they were in previous courses (students / teachers therefore expect them) and I guess because in our mother tongues we have these words apparently stored together. However, I think both are somewhat flawed arguments. The first argument is just a problem of life and changing attitudes: we’ve always done it like that / it works, but excludes the possibility of something that might be better. The second argument is understanding things the wrong way round. I know lots of words for animals in both English and Spanish but I think to a large extent I learned the words separately from each other. The fact people generally keep shoes in the same place, doesn’t mean they bought them all at the same time – I’d suggest that would be a sign of addiction / madness! The same (up to a point!) applies to lexical sets. As for fun, I agree with you that there is IS something very satisfying about learning lists of vocabulary, because it’s so achievable. However, is that fun and sense of achievement only apply to a lexical set? Having said that, I doubt we’ll do away with lexical sets altogether, but I think we can change our approach to them in other ways. You might like to look at the post: the problem with lexical sets and a comment by leoxicon.

  5. Glad you enjoyed the post and thanks for spreading the word on your post.

  6. Glad you liked the post and appreciate you taking time to comment. I don’t know if you’ve read Kahneman’s book, but one of the points he makes is that people are very resistant to the idea that they are biased!;-) We do obviously use blonde quite frequently and you are right to imply that in speaking, blonde is far more common than it is in writing. However, if you look at the Longman Contemporary English Dictionary, arise is still as common in spoken English as blonde is, (top 3000 most common words). However, arise is also in the top 2000 words in written English – so of much wider use. While I certainly believe that students come to class to learn and practice speaking more than most skills, working in EAP over recent years I’ve become increasingly aware of how many students require English for study and work and for reading and writing. In this and future posts, I would not like to suggest we stop teaching a word like blonde, more that we should be aware of the frequency of words like arise and start teaching them.

  7. Just posted a new thing on my blog and linked to this. People not thinking they are biased is called bias blind spot I think. http://en.wikipedia.org/wiki/Bias_blind_spot

  8. Thank you for the reference to my blog post. Of course, I agree with you! The only thing I would say though is that what comes out of corpora are not necessarily the best examples for our students as they tend to be fragments (see concordances). However, thinking of good examples is also difficult and can’t be left to spontaneity in most cases but need to be thought about and learnt. Using a good dictionary (usually based on corpora), testing ideas with google searches or simply thinking about when I might say it and to whom; asking yourself what would go before my sentence and what would go after, will often lead to better examples.

  9. [...] small amounts. Late on, @theteacherjames retweeted a couple of more links c/o @EBEFL - a piece here by Andrew Walkely and another by @EBEFL here - as to why corpora can be useful and which both [...]

  10. [...]   [...]

  11. What maddens me is that the course books that most of us EFL teachers rely on are failing us as teachers and our students by not providing the kind of language that needs to be learnt. And as so many of us have the constraint of having to follow the textbook to fullfil the ‘syllabus’ (def product approach – schools/companies want tangible ‘results’) that we have no alternative but to teach the usual ‘He’s tall and blond’ descriptions, even while instinctively knowing that this is not what the student needs. Any more radical, process approaches are often frowned on. ho hum…. We’re faced with quite a dilemma.

  12. [...] … teachers (and students) are not very good at recognising frequency of words …  [...]

  13. I think there are some books which are beginning to move away form this – we certainly tried to do it with Innovations and Outcomes. But it’s not easy and it’s not just publishers who are to blame. I will never forget a teacher opening Elementary innovations and seeing Unit 4 ‘going to’ – oh too difficult. Maybe this was our fault for labelling what was taught as ‘grammar’. In fact the students learn What are you doing later / tonight / at the weekend? and I’m / We’re going to + verb (most of those verb phrases being based around go). Lexically, it’s a very simple conversation, but teachers want to see and publishers want to highlight the grammar and it has to be in a particular order.

    I think you are right to pinpoint the issue of syllabus and testing. When you start thinking about a lexical syllabus, it’s a bit messy. Which words should be taught in which combinations and chunks. Once you start thinking in this way there can be many possible syllabuses, but schools and national curricula want a simple solution that can be written on a few pages – which ‘grammar’ often provides.

    I guess though that there are moments and examples in any coursebook that can be exploited – check out the follow-up to this post – http://blog.westminster.ac.uk/celt/2012/10/11/the-problem-of-lexical-sets-and-some-alternative-approaches/

  14. […] http://blog.westminster.ac.uk/celt/2012/10/09/elt-teachers-shouldnt-prefer-blondes/ […]

Leave a Reply