kitchen table math, the sequel: "Why English-Speaking Children Can't Read"

Saturday, November 8, 2008

"Why English-Speaking Children Can't Read"

Why English-Speaking Children Can’t Read

As the universal-education movement began gathering momentum, educators broke ranks with nineteenth-century traditions. Reading instruction got so far off track that the twentieth century will go down in history as the century of the demise of the English alphabet code. The final reckoning of an unceasing attempt on its life came in the 1990s. For the first time, properly conducted national testing, international reading surveys, cross-cultural studies, and classroom research pointed to the inescapable conclusion hat reading instruction in English-speaking countries is a disaster. The functional illiteracy rate for American 9-year-olds is 43 percent (Mullis, Campbell, and Farstrup 1993; Campbell et al. 1996).

International reading surveys carried out by Statistics Canada brought dismal news (Organization for Economic Cooperation and Development, 1995, 1997). In six English-speaking nations, the proportion of functionally illiterate/very poor readers among 16- to 65-year olds ranged from a low of 42 percent in Canada to a high of 52 percent in the United Kingdom. These figures were in stark contrast to those of many European nations. The comparable figure for Sweden was 28 percent. Sweden’s functional illiteracy rate for 16- to 25-year-olds (level 1 of 5 levels) is 3.8 percent. This rate is nearly three times higher in Canada (10.7 percent), and six times higher in the United States (23.5 percent).

In 1993, an astonishing report came in from Austria. Heinz Wimmer set out to study poor readers and initiated a citywide search. He asked 60 second-to fourth-grade teachers in Salzburg to refer their worst readers for special testing. They identified 120 children, about 7-8 percent of the school population. Imagine Wimmer’s surprise when the worst readers in the city scored close to 100 percent correct on a test of reading accuracy and did nearly as well in spelling. Clearly, none of these children had any difficulty with the German alphabet code. It turned out their problem was reading too slowly. But slow is a relative term. How slow is slow?

To find out, Wimmer collaborated with an English researcher (Wimmer and Goswami 1994) to compare normal 7- and 9-year olds from Salzburg and London. The results were startling. The Austrian 7-year olds read comparable material as rapidly and fluently as the English 9-year-olds, while making half as many errors. Yet the Austrian 7-year olds had had 1 year of reading instruction, while the English 9-year olds had been learning to read for 4 or 5 years. Equal speed and half the errors in one-quarter of the learning time is an eightfold increase in efficiency!

Wimmer and his colleagues (Lander, Wimmer, and Frith 1997) got the same extraordinary results when they compared their worst readers (incredibly slow) with English children identified as “dyslexic” (incredibly inaccurate). The children were asked to read text consisting of nonsense words. The so-called Austrian slow readers were no only more accurate than the English “dyslexics,” but they read twice as fast. The average Austrian “slow reader” would be able to read a 500-word passage in about 10 minutes, misreading only 7 percent of the words. The average English “dyslexic” would read only 260 words in this time, and misread 40 percent of the words. It seems the expression “worst reader” is relative as well.

An even more dramatic study was reported from Italyy. Cossu, Rossini, and Marshall (1993) tested Down’s syndrome children with IQs in the 40s (100 is average) on three difficult reading tests. They scored around 90 (100 is average) on three difficult reading tests. They scored around 90 percent correct, breezing through Italian words like shaliare and funebre. However, they could not comprehend what they read, and they failed miserably on tests of phoneme awareness, the skill that is supposed to be essential to decoding.

What is going on?

The answer is simple. European countries with high literacy rates have a twofold advantage. First, they have a transparent alphabet code, a nearly perfect one-to-one correspondence between each individual sound (phoneme) in the language and a visual symbol—a letter or letter pair (digraph). For languages with more sounds than letters in the alphabet (English has 40+ sounds), this problem was handled sensibly. When a letter or digraph is reused to represent more than one sound, it is marked by a special symbol (a diacritic) to signal a different pronunciation. In German, an umlaut distinguishes the vowel sound in Baume (boimeh). And while a sound can occasionally be spelled more than one way, there is never more than one way to read a letter or digraph. The English spelling system suffers from both afflictions: multiple spellings for the same phoneme, and multiple ways to decode letters and letter sequences. This is the definition of an “opaque” writing system.

Reading instruction is the second part of the equation. To a great extent, reading instruction is a function of the complexity of the spelling code. Teaching a transparent writing system is far easier than teaching an opaque one, because it is obvious (transparent) how it works. Teaching can be streamlined and proceeds at a rapid pace. In Austria, children are taught the sounds of the German language and which letter(s) represents each sound. Reading and spelling are integrated at every step, which reinforces the code nature of a writing system—that is, the fact that the operations are reversible, involving both encoding and decoding. No clutter or noise clogs the process, such as teaching letter names or lots of sight words. Because basic reading instruction is fast and pretty well guaranteed, it can begin late – at 6 in most countries (age 7 in Scandinavian countries) –and early (after 1 year or less) Parents sleep soundly in their beds, safe in the knowledge that their child will be reading and spelling by the of the first year of school. (This is not to say that inappropriate teaching methods cannot mollify the advantages of a transparent alphabet.)

The cross-cultural comparisons reveal that the source of English-speaking children’s difficulties in learning to read and spell is the English spelling system and the way it is taught. These comparisons provide irrefutable evidence that a biological theory of “dyslexia,” a deficit presumed to be a property of the child, is untenable, ruling out the popular “phonological-deficit theory” of dyslexia. For a biological theory to be accurate, dyslexia would have to occur at the same rate in all populations. Otherwise, some type of genetic abnormality would be specific to people who learn an English alphabet code and be absent in people who live in countries with a transparent alphabet, where poor readers are rare. A disorder entirely tied to a particular alphabetic writing system is patently absurd and has no scientific basis. English-speaking children have trouble learning to read and spell because of our complex spelling code and because of current teaching methods, not because of aberrant genes.

Early Reading Instruction: What Science Really Tells Us about How to Teach Reading
by Diane McGuinness, pages 1-3
It's always worse than you think.

(I've asked Liz Ditz to weigh in on the question of what biologically-based reading and/or learning disabilities are.)

Le scandale de l'illetrrisme (nouvel obs: the scandal of illiteracy)
dyslexie, vraiment? ) (nouvel obs: true dyslexia? - whole language in France)
Comment en est-on arrivé là? (nouvel obs: How did we get here?)
French spelling

Why English speaking children can't read

Lucy Calkins on teaching children to write
Becky C on starting at the top

instructional casualties in America
curriculum casualties: figures
forcing hearing children to learn as deaf children must
Rory: I frickin' hate whole language!

thank you, whole language


ElizabethB said...

I have remediated a bunch of kids and have a current class of 9 with the help of volunteers from my church.

I believe that about 99% of my students are suffering from sight words, not biological based dyslexia.

I currently have one student whose father is a slow reader despite a very good phonics program in the lower grades (he remembers it, and knows all the rules. His wife doesn't know all the rules, she is a better reader than him although she got whole word methods. She's helping out in the class of 9 students and is learning the rules!

At any rate, I tested my UPP, a specially marked print I developed for remedial students, with these students. The student who I suspect has some form of biologically based dyslexia did better with a version of my UPP without hyphens, and said the hyphens were distracting. All of the other students preferred the hyphen version, and about half of them read markedly better with the hyphens.

I had a version of the UPP with blue vowels, but ALL of these 9 students were slowed down to some degree by the blue vowels and found them distracting.

I was surprised, the blue vowel version seemed much clearer to me. I'm working through my phonics lesson movies, one or two a night, removing blue vowels.

Here's my UPP page for those that are interested:

The student I suspect of biological dyslexia has different error patterns than my normal students with sight word based problems. However, some of the error patterns are the same.

But, the remediation is the same regardless of the cause--a lot of phonics, no sight words, nonsense words, spelling, and syllables.

Rudbeckia Hirta said...

I'd love to see a study comparing English-speaking children to French-speaking children, as French also has very strange spelling rules (especially the silent -ent on the 3rd person plural of -er verbs!).

palisadesk said...

This is usually a discussion that goes around in circles, because there is no agreed-upon definition of "dyslexia." Thus, arguments about what percent of what population where has "it" tend to be muddled and contradictory.

What we do know:

-a large number of individuals with "dyslexia-like" symptoms (poor decoding, scrambling words, reversing words, etc. ) are remediable with systematic instruction, and do not show phonological deficits -- problems blending, segmenting or manipulating sounds in words. You can call them "acquired dyslexics" or (as Zig Engelmann called them), "dysteachics." Sally Shaywitz identified them as "persistently poor readers" (with no apparent neurological disability) whose faulty reading practices she posits as environmental in origin.

- a small number of individuals have great difficulty learning to read, even with the best instruction. It is hypothesized that these difficulties are neurological in origin, but in most cases there is no clear evidence one way or another. It is an empirical question. These individuals may differ greatly from each other in the types of problems they have – lexical retrieval, working memory, visual and auditory integration, etc., but they all can be taught with explicit methods which may need to be adjusted to the individual.

-readers in other countries and languages also experience the types of problems commonly lumped under the category "dyslexia," that is, difficulty blending sounds, discriminating words and phonemes, retrieving symbols, words or meanings, etc. A literature search on the phenomenon in languages such as Spanish and Italian will yield many studies of "dyslexia" in those languages. Decoding is more easily taught, but learners with the problems blending sounds etc. will exhibit the same behaviors as "dyslexics" in English.

-early and systematic teaching of phoneme-grapheme relationships (many engaging, fun and multisensory programs exist for doing this with children as young as 3) tend to prevent most reading failure. Children with specific difficulties will still require additional instruction and support. However, if we could improve the instruction offered to the majority of children, the resources could be made available for the truly needy.

In most cases, the "cause" of the "reading problem" doesn't matter – the course of action is to teach the individual the needed skills.

Catherine Johnson said...

Hi Rudbeckia -- French kids are apparently now doing very poorly, thanks to the introduction of whole language teaching there. I still haven't gotten around to getting Ed to translate the article on constructivism in French schools, but he did translate the article on rising dyslexia there.

Comment en est-on arrive la?

(not sure that has the sidebar about dyslexia - Ed is translating the passage on orthography now)

Catherine Johnson said...

Here's the sidebar on dyslexia in France:

Is it really dyslexia?

Ed has translated that one.

He'll get the orthography passage translated shortly.

Linda Seebach said...

You quote McGuiness as saying, "For a biological theory to be accurate, dyslexia would have to occur at the same rate in all populations." (I assume the highlighting is yours.) While it is likely true that biology is not responsible for all the differences in dyslexia between populations speaking different languages of varying degrees of transparency, her claim about biology is simply false. Many biological traits manifest at different rates in different populations, either because the populations have on average different alleles governing these traits, or because the populations live in different environments where the alleles have different effects.

Anonymous said...

I decided to make a whole post in rebuttal to McGuinness's mistaken assertion, rather than a comment. It's here:

Why McGuinness is Mistaken: Some Reading Problems ARE Biologically Based

Catherine Johnson said...

Thanks, Liz!

Catherine Johnson said...

off-topic: what has become of the little trash can below each comment - the one that allows you to delete comments if you wish?

Catherine Johnson said...

Linda - Obviously you're right, and I was trying to work out whether McGuinness' argument is wrong as a result.

I'm pretty sure her observation is "biologically correct."

As I understand it, she isn't saying that ethnic groups & subpopulations, etc. can't have genetic differences. Her argument is that English speaking peoples as a whole, throughout the world, can't have a systematic genetic difference from non-English speaking peoples as a whole.

In other words, if you were to do a separated-at-birth twin study looking at identical twins who were sent to homes speaking different native languages, you would find that twins sent to a family whose native language is English would have a higher rate of reading disability than twins sent to Spanish- or German-speaking homes.

I'm not that far into her book, but my sense is that she's criticizing the "phonemic awareness" theory of dyslexia specifically -- not the idea that reading problems would have a biological basis.

Catherine Johnson said...

Apparently you have to click on the link for the individual post to find the trash cans...

concernedCTparent said...

I've been wondering the same thing. What happened to the "delete forever" option?

Anonymous said...

I don't think she understands gene-environment interactions. You can have a genetic susceptibility for something that is unmasked or made worse by an environmental stimulus. So the underlying problem may be there, but is only manifested under certain conditions.

This is one reason why cancers occur at different rates in different populations.