Reading in a Foreign Language just published my critique of an article from the previous issue (Macalister and Webb, 2019a) on the topic of children’s literature and adult ESL readers. There are also two rebuttals to my piece (Webb & Macalister, 2019b; Macalister, 2019)).
First, a TL;DR summary on the exchange.
Macalister and Webb (2019a) claimed that in order to read “children’s literature” with good comprehension, you would need to know at least 8,000 of the most commonly occurring words in English. (In a previous study (Webb & Macalister, 2013) they claimed the figure was 10,000 words.) They based their findings on a set of stories written for children by the New Zealand Ministry of Education.
Why is this important? Because the main source of good reading material for English beginners and low intermediates are graded readers (stories written for second language acquirers). But graded readers in English currently only take you up to the 4,000 word family level. After that, you have to find texts written for native speakers.
If Macalister and Webb are correct, there’s a four to six thousand word gap between graded readers and what would be presumed to be the next easiest kind of book, children’s stories.
So the researchers conclude that children’s books are too hard for intermediate English speakers. The gap is too large. To bridge this supposed gap, Macalister and Webb propose explicit vocabulary instruction of 200+ high-frequency words that appear in children’s stories. This, they argue, would put children’s literature within reach of intermediate English readers.
Now, we know from previous research that you also need to know 8,000 to 10,000 words to read classic novels and your morning newspaper (Nation, 2006). So consider what Macalister and Webb are really saying: children’s literature is as hard, in terms of vocabulary difficulty, as Charles Dickens and the New York Times.
Also for my response to Macalister and Webb, just for fun, I analyzed the vocabulary difficulty of their own academic article. I found that it, too, required being at about the 8,000 word family level in order to read it with good comprehension (98% vocabulary coverage).
From this we can also conclude that, if Macalister and Webb are correct, children’s literature is as hard as a scholarly article about children’s literature.
If that conclusion sounds like it might be a bit off, then you’re probably not an academic.
In my brief critique (about one page), I point out where Macalister and Webb went wrong. The children’s stories they used in their study are clearly not typical of the books children actually read for pleasure.
We know this in part because three years ago I published – in the very same journal! – an analysis of books ranging from the Box Car Kids to Twilight. I determined that there were lots of series books that were much easier, in the 3,000 to 7,000 word level range (McQuillan, 2016a).
In other words, once you graduate from graded readers, you can safely move into a lot of children’s and young adult novels. There is no gap.
There was one rebuttal signed by both authors, and a separate rebuttal from the second author only.
Webb & Macalister (2019b)
Before responding to my critique, Webb and Macalister (2019b) give a long review of the arguments for why children’s literature isn’t appropriate for beginning and low-intermediate adult ESL readers, those below the 3,000 to 4,000 word family level. I agree. That’s why I said that children’s literature can be appropriate for readers who are now beyond graded readers, meaning at or above 3,000 to 4,000 words.
The rest of their response can be summarized in five points:
(1) Our clam that you need an average of 8,000 to 10,000 words to get to 98% vocabulary coverage for children’s literature does not mean that each text would require 8,000 words.
Well, yes, averages are calculated from a range of numbers. If you analyze 500 stories and determine the “average” vocabulary coverage is 8,000 words, it does not mean that every single story in that corpus would require that. But that’s true of all averages. There will be some texts above and some below the mean.
The whole point of reporting averages is to give the “central tendency” of a set of data. Webb and Macalister concluded in their original article that children’s lit was too hard based on this average. They can’t now go back and say, “You know, actually, averages can be deceiving!”
Note that I avoided this issue in my own analysis of series books by reporting vocabulary difficulty by author or book series. I did checks to make sure that within a given series, the vocabulary levels were relatively (within 1,000 word families) stable.
(2) Yes, “children’s literature” really is as hard as scholarly literature.
Webb and Macalister double down on their claim that both “children’s literature” and scholarly articles really do have the same vocabulary difficulty. As noted in my critique, this rather odd conclusion is defensible only if we assume their corpus is typical of the things children actually read. I’ve already shown this to be false (McQuillan, 2016a).
(3) Knowing the words on the page isn’t the same as being able to comprehend the text.
This is true. You also need some background knowledge to understand a text. But the basic working premise of all corpus studies on vocabulary (including their own) is that, all things being equal, you need a certain number of word families to understand the texts in a given corpus. No one would argue that knowing the words alone is sufficient. This really is a strawman argument.
(4) The fact that case studies show some readers can move from graded readers to children’s and young adult books does not mean that “all learners will be able to do this.”
Again, I agree, although I suspect that pleasure reading is crucial for the large majority of those who reach advanced levels in a second language. That’s an empirical question that all of us agree requires more study. However, the question isn’t whether ALL learners can do something (a nearly impossible claim to support), but that many or most will be able to.
(5) McQuillan is proposing a “single approach” for acquiring vocabulary, but some students may benefit from other approaches.
I did not claim that reading is the ONLY way to improve vocabulary. Obviously we also acquire vocabulary through the other form of comprehensible input, listening.
What I did claim is that free reading is, for intermediate English readers, the most efficient way to acquire new words, and it is much more efficient than Webb and Macalister’s solution of explicit vocabulary instruction (citations below).
Macalister (2019)
In a separate rebuttal, Macalister responds to my critique by recounting his own personal history of how he tried reading Harry Potter in a second language (French).
You can read that history yourself, but his argument basically comes down to this:
- I am a self-proclaimed intermediate reader in French.
- I found reading a single children’s series (Harry Potter) difficult.
- Therefore, children’s literature is too difficult for most intermediates.
A few reactions to Macalister’s rebuttal:
First, he seems to forget one of the underlying premises of his and Webb’s own analysis, which is that texts for which you have at least 98% vocabulary coverage can be read with good comprehension.
If Macalister found Harry Potter too difficult, then he’s probably not at 98% for those books.
The solution to Macalister’s difficulties is simple: Find an easier book!
Second, I cited a few articles in my critique that give “case studies” of adults who moved from graded readers to young adult fiction (Mason & Krashen, 2017; Uden et al, 2014). Macalister dismisses these as involving only “highly motivated” readers, and therefore not typical of most second language students.
While this may have been true for the subjects in Uden et al., there’s no indication of this in the other group of case studies I cited. More importantly, as I noted in my critique, “high motivation” may not be the key ingredient.
In the cases I discussed, readers choose their own books, thus ensuring that they will be both interesting and comprehensible (and if they’re not, they’d put them down and read something else).
Third, Macalister thinks explicit instruction on a high-frequency word list for the Potter novels would have been useful to him. The evidence suggests that this would not be true for most acquirers. As I explained in another article that also appeared in Reading in a Foreign Language, traditional vocabulary learning methods are among the least efficient and effective ways to improve vocabulary and comprehension.
And if motivation is a concern to Macalister, it’s hard to argue that memorizing vocabulary words is any student’s idea of a good time (a point I discussed here). If explicit vocabulary and grammar instruction were the key to helping unmotivated students, then our classrooms would already be filled with the most eager and excited language acquirers you can imagine.
Reading is the most efficient, motivating, and pleasant way to improve your vocabulary, provided you don’t pick books that are too difficult for you, as Krashen (2004) points out. I’ve written a series of studies to show this (McQuillan, 2016b; McQuillan, 2019a; McQuillan, 2019b, McQuillan, 2019c), as have Beniko Mason and many others.
Follow Us!