JALT Conference Teaching Listening presentation

Teaching Listening Using Bottom-Up Approaches

Andrew Blyth    Room 901, 11 to 11.25am, Saturday 24th Nov

Thanks for your interest in this topic and for coming to my presentation. Here are some resources:

This presentation is also supported by HelloSpace.Me, websites for teachers and academics. If you like this website, build your own with content tailored to you and your classes. Contact hey@hellospace.me.


There have been few options for teachers to really teach listening. There has been no credible evidence in support of listening strategies, and extensive listening is not enough. Recent research from cognitive science and pedagogy indicates that teaching listening with bottom-up approaches is effective, and is easy to do. This workshop will introduce some simple activities you can use on Monday morning to get your students listening better using proven methodology.

Does anyone know how to do extensive listening?

Recently, my colleague asked me for my collated list of Extensive Listening (EL) links, which prompted a short discussion on EL, and how to do it. This quickly highlighted the fact that we had two different ways to think about how to do EL. Unlike Extensive Reading (ER) we are all on the same page (pun intended), but the opposite is true for EL.

The problem with EL is that nobody ever figured out an elegant way to do it. I’ve written about EL before, after someone asked me how to do it (Blyth, 2015), which followed on from my ELT Journal article (Blyth, 2012), but I could only give generic advice. The big problem that listening teachers have, ever since the audio-lingual era, is that we’ve been too busy to do anything but borrow ideas from reading pedagogy (Field, 2008). EL is probably no different. We just uncritically adopted ER principles.

With ER, I provide my students with the following information, which is based on Day & Bamford (2002):

Rules: Extensive reading must be 1. Fun or interesting; 2. Easy; 3. You choose your own book, magazine, blog, anything; 4. Read lots; 5. Talk about it. Research suggests that you should know about 98% of words to have effective comprehension of the text. Choosing: for every 100 words, you should know 98 so it’s ok if you don’t know 2 words, but if more than three or four unknown words you should choose a different book…

So which of the above criteria can be applied to EL?

  1. Fun or interesting
  2. Easy
  3. Choose your own
  4. Listen lots
  5. Talk about it

But not “6. How to choose an EL text your level”. Listening texts are also invisible. With an ER book, you can look at it briefly and decide if the book is worth further investigation or not. Vision provides information up to 1000 times faster than any other sense. In contrast, students cannot know if the EL text is at, above, or below their level until they had already committed to putting on the headphones, pushing the play button, and waited to get a handle on the contents. With a book students can quickly see the level information, the number of headwords, the word count, size of font, number of lines per page, the thickness of the book, the type of pictures, and the usage of images too, among many other cues. In contrast, a listening text is a mystery bag, or like a box of chocolates, “you never know what you’re going to get”, until you start listening. This mystery element is the part that has always troubled me, and perhaps my students.

So far, there are a number of assumptions to get out of the way before we can get to the final question, “Is there a better way to do EL?”.

  1. Is extensive listening even needed?
  2. Is it effective?
  3. What is the current system of EL?
  4. Is the current system working?
  5. Is there a better way to do it?

The above questions deserve far more attention than what I can afford here, but brief answers are attempted just the same. Firstly, EL is indeed needed. For instance, Field (2008) argued that students on study abroad and similar programmes often struggle. They might be tested on their grammar, and placed into a class level based on their test-taking competency and grammar knowledge, but not their listening skills. Consequently, these students, especially those from the Far East, tend to do badly in class because they cannot access their grammar teacher’s verbal instruction, and so they blame themselves rather than the level placement apparatus. Another way to answer the first question is from a Vygotskian perspective: practice makes perfect. Research in brain science, suggests the more a certain set of neural pathways are used, the more connections are made, which results information being able to flow more efficiently  (Burton, 2013). That is, the more practice students get in listening, the better they will become.

Secondly, is EL effective? Masters and PhD students of Renandya & Farrell (2011) clearly demonstrate this. Students who did extensive listening outperformed students who received only listening strategy instruction. I argued in support of this point that strategy instruction is merely teaching students to compensate for a lack of skill, rather than teaching the skill to plug the gap (Blyth, 2012). Chang and Millett (2014) provided additional evidence that practising EL results in improved listening performance than no EL practice (or at least ER alone).

Thirdly, what is the current system of EL? On this point, I cannot give a clear answer, to which Chang & Millett say, “The practice of EL in L2 teaching is a relatively novel idea and its theoretical framework is undeveloped, so not many language teachers have the knowledge to carry it out” (2016, p351). Everyone seems to have their own idea on what and how to define EL, and how to implement it. So, to define what “4. Listen lots” means, as in, how much is “a lot”, cannot be done. What texts can be used? For Chang and Millett (2016) EL is listening to graded reader audio CDs. For Rob Warring (What is EL?), it is listening to a variety of (ungraded) web based texts like those listed on my Listening page.

Fourthly, despite the lack of clarity for the third question, EL is still more effective than teaching listening strategies. It still seems that any kind of practice is better than no practice.

Fifth and finally, is there a better way to do EL? I would like to begin discussion on this point. To me, the mystery bag element has to be dealt with. To begin with, how can we define the “mystery bag”? Firstly, grading; how can a student know if a particular or group of similar listening texts are level appropriate for him or herself. Secondly, what does level appropriate mean? What are the key features that make a listening text appropriate, whilst others inappropriate? Secondly, how much listening is needed on a weekly basis; assuming that a week is an appropriate increment of time to manage listening practice. On this point, my boss had his idea of the minimum time for students to listen, whilst I had mine. Mine demanded that students do at least thirty minutes a week, but they could choose a five minute text, and listen to it six times; or a fifteen minute text, and listen to it twice; or two or three different texts, but still total listening experience must be at least thirty minutes. Obviously, the number of times of practice matters, as argued earlier. However, what is the optimal length and number of times to listen?

Establishing standardised metadata

Publishers of graded readers have, in their various ways, admitted that certain metadata are needed. As mentioned earlier, information about the number of headwords, word count, CEF level, genre, and more all help the student decide if the book might be level appropriate for them. At the moment, most EL texts lack such information, except where it might be listed as “easy”, “medium”, or “difficult”, based on the curator’s judgement. Consequently, in the interests of making things easy for students, I would propose a similar system. I would hope that these criteria I’m suggesting not be the new standard, but be discussed, tested, and modified based on an informed position. Then, after rigorous testing, will the most appropriate ones be used. The metadata to consider are:

  1. CEF level
  2. Genre & topic
  3. Nature of text: authentic, genuine, or contrived.
  4. Length of text
  5. Words per minute
  6. Number of speakers
  7. Number of words outside of the 2,000 most common word range
  8. Number of academic and technical words from the AWL
  9. Number of idioms used
  10. Script availability: Yes / No
  11. What native-like phonological features are present?


It is hoped that future research on EL can begin to address a number of key issues. Firstly, what is an optimum length of listening time per week, and optimum number of times to listen to each text. Secondly, what is the best practice for doing EL, specially for an education environment where accountability (and grades) are concerned? Thirdly, what metadata are most useful for students to help them decide if a text is appropriate for their current listening level. Which also implies another question, what is a “listening level”?


Blyth, A. (2012) Extensive listening vs. listening strategies: Response to Seigel. ELT Journal, 66 (2), pp. 236-239.

Burton, R. (2013) A Skeptics Guide to the Mind. St Martin’s Press.

Chang, A. and Millett, S. (2014) The effect of extensive listening on developing L2 listening fluency+ some hard evidence. ELT Journal, 68 (1), pp 31-40.

Chang, A., and Millett, S. (2016) Developing L2 listening fluency through extended listening-focused activities in an extensive listening programme. RELC Journal, 47 (3), pp. 349-362.

Day, R., and Bamford, J. (2002) Top Ten Principles for Teaching Extensive Reading. Reading in a Foreign Language. 14 (2), pp. 136-141.

Field, J. (2008) Listening in the Language Classroom. Cambridge, UK: Cambridge University Press.

Renandya, W., and Farrell, T. (2011). ‘Teacher, the tape is too fast!’ Extensive listening in ELT. ELT Journal, 65(1), 52-59.