Friday, January 15, 2016

What are the problems with Whole Language and why doesn't it work?

The allure of using Whole Language to teach children to read lies mainly in the fact that, as you'd expect, humans are heavily biased towards meaning and a whole word approach has an immediate appeal because, at the beginning, it seems so easy. On the other side of the methodological divide, learning how to recognise letter shapes as representations of sounds is hard work from the start even though it gets easier as learning progresses.

So, what is wrong with a Whole Language approach?

Well, firstly, whole word instruction isn’t generative! Each word has to be learned from someone who already knows that word. Thus, every unknown word constitutes potentially a barrier to meaning and understanding.

Secondly, and this follows on from the previous point, whole language is hugely time-consuming. Each new word requires oodles of practice and teachers need to provide multiple opportunities for practice, which is why books with a Whole Language orientation are filled with endless repetition of ‘key’ words. 

If on the other hand we teach children even a small number of sound-spelling correspondences and we get them to practise the skills of segmenting and blending, they can quickly read (and spell) dozens and dozens of words.

Here's an example of what I mean: if we teach children just 12 sound-spelling correspondences ([ a ], [ i ], [ m ], [ s ], [ t ], [ n ], [ o ], [ p ], [ b ], [ c ], [ g ] and [ h ]) and we teach them to segment and blend, they will be able to read and spell around 70+ words, as well as to be able to read and write sentences in which those words appear.

Thirdly, by demanding that pupils attend to whole words, there is no reason why they should give attention to the detail of words. In fact, research shows that children are likely to attend to the outer segments of words but not to the internal details. As there are so many words in the English language that look very similar, a Whole Language approach always makes children susceptible to mis-reading and mis-spelling even very common words. And this is exactly what we see in practice.

Finally, a Whole Language approach is very frustrating for children who simply cannot remember words as wholes from the start and anybody who teaches young children knows and can identify such pupils within weeks of them beginning school. However, it isn’t just such children who have trouble from the outset. In fact, pupils with lots of prior learning and very good visual memories are likely to be even more at risk in the long term because, inevitably, as they are required to learn more and more words, they begin to run out of visual memory and begin to guess wildly or to try and work out what words are from other, contextual clues, such as syntax, illustrations and discourse. Furthermore, once these bad habits have become ingrained, as with any habit that is practised, the longer it goes on, the harder it is to correct.

And all of this is compounded by teachers who haven’t been properly trained to teach the sensible and only alternative to Whole Language/Look and Say/Mixed methods.

So, I say again: train the teachers!


  1. Re-posted here, John:

  2. Thanks for re-posting, Debbie.

  3. So, I say again: train the teachers!

    The thing is, teachers have already been trained, and few Head Teachers believe that their teachers need more training.

    Moreover, the results of studies that have been done of "high quality professional development" indicate that although the talk of teachers changes, their teaching doesn't.

    More moreover, it really doesn't take much "training" to reliably teach little kids how to handle the Alphabetic Code and the other conventions of English text. Any child who can talk in full sentences and participate in everyday conversation has the necessary, phonological, syntactic, and semantic prerequisites. And the success of homeschooling Mums with little or no professional training teaching their kids to read is evidence to the point.

    The best "data" on the matter have yet to be analyzed. Although not a whole lot of schools took advantage of the UK government's funding match for synthetic phonics training, there are enough to make a Natural Experiment game of it. The numbers could be run, but I don't see anything to support "it's the training."

    Meanwhile, the beat goes on.

  4. Sorry, but I simply don't agree with you on this, Dick.
    Many teachers haven't been trained or the training they've received or are receiving from many training institutions is so dreadful as not to count as training.
    It's certainly true to say that when teachers have been trained, they do often status quo, or find it difficult to resist the temptation to incorporate stuff that they have always done but which runs counter to their new training. Even when a clear rationale for doing something new has been given, old habits are hard to break.
    Again, on the point about teaching little kids to read (and spell): when the code is simple - one letter-one sound, it's hard to go wrong, unless children are being asked to memorise whole words. However, even at this level, orientation is immensely important. Is it being made clear to children that the squiggles represent the sounds in the English language because this is crucial when they come to the complexities of the code: more than one way of spelling sounds, spellings representing more than one sound, and how to read and spell polysyllabic words.
    In addition, the potential pitfalls along the way - 'letters make or say sounds', teaching 'onset and rime', teaching letter names as well as sounds from the start, etc, etc. - combine to increase confusion and failure.
    Where we have schools implementing Sounds-Write with fidelity, 100% or close to 100% of children pass the phonics screening check and then go on to be able to read and spell to a very high level of proficiency.
    How do we get the message across? Well, if government would support some serious, long-term research to compare approaches, we might learn something.
    I'm game!


  5. I concur completely with all these points. The thing is, we're stuck with the kids, teachers, Head Teachers and Governments we have, not those we would like to have. "Train the teachers" isn't the route to get the job done of teaching all little kids how to read (or know the reason why.) The fact that some schools are not "implementing Sounds-Write with fidelity" after training proves the point, doesn't it?

    The schools and teachers that believe they are "improving Sounds-Write" or are "doing Sounds-Write right, by doing what they are doing attribute their instructional failure to the kids, parents, "poverty" and so on--anything but the instruction they've provided. They could be right. However. The fact that schools and teachers with the "same" [kids, parents and poverty] are getting the job done, proves that they are not right. It's in the instruction.

    Government-supported "serious, long-term research to compare approaches" just isn't in the cards. The good news is while "more research" is always a good thing, in this case it isn't necessary. The bulk of the data needed for a Natural Schooling Experiment have already been collected or are being collected routinely in England. Preliminary results can be obtained at negligible time and cost. But the game for this "simple research" isn't yet in play.