Key lesson from my recent debate on phonics: Don’t Talk about the evidence!

My recent YouTube/podcast debate with Kathyn Garforth and the subsequent twitter exchange are two more striking examples of proponents of phonics failing to address my critique of the evidence.  I thought I would summarize some of the highlights and propose another debate with a researcher who claims that the science of reading supports phonics.  Any takers?  Kathy Rastle?  Pamela Snow?  Mark Seidenberg? Jennifer Buckingham? Timothy Shanahan?

Regarding the main debate (see this link), Garforth claimed that there is good empirical evidence for systematic phonics but was not able to cite any evidence or address any of the criticisms I made regarding the evidence (at least that is my take, listen to the debate and judge for yourself).  But it started getting a bit absurd and unpleasant on twitter when Garforth accused me of changing the rules of the debate just a few days earlier:

 

Then the debate moderator quickly came to Garforth’s defense, writing:

 

When I posted on twitter the offending email that was taken to narrow the scope of the debate, Nate Joseph clarified.

Indeed, quite a nuanced difference.  And in any case:

A sad end to an otherwise enjoyable debate.

But then there was a second, and in many ways more interesting follow-up twitter debate, with many dozens if not 100s of postings.  What is again so striking is how no one addresses my main point – the topic of the debate – namely, is there any good evidence for phonics?  It all starts here: https://twitter.com/jeffrey_bowers/status/1372655334079225860

For example, Pamala Snow comments:

I responded:

The continuing debate regarding whether phonics can explain the spelling of <dogs> was extensive, but Dr. Snow (again) does not engage with the question of whether there is strong empirical evidence for phonics.  Helpfully Greg Ashman joins in with one tweet:

But I refrained from responding.

Kathy Rastle later joined in the debate as to whether phonics can explain the spelling of <dogs> writing:

Oddly, Snow “liked” this comment despite the fact that it provides a different explanation of the spelling of <dogs>. As far as I can tell, according to Snow, phonics instructors think that the z-sound in in the spoken word “dogs” is actually the phoneme /s/ that only sounds like a /z/ “because of the influence of voicing in the preceding consonant /g/”.  Because the z-sound in <dogs> is actually an /s/ phoneme, phonics can explain the spelling of <dogs> and this can be explained to children.  By contrast, Rastle is not showing how phonics is explaining the spelling of <dogs>, she is pointing out that the English curriculum includes some morphological instruction.  Indeed, it is an example of explaining how morphology and phonology interact.  I think Kathy’s account is easier to explain to children.  And indeed, it raises the question as to whether instruction should more often highlight the many interactions between morphology and phonology (just don’t mention SWI).

But again, Rastle avoids my claim that there is little or no empirical evidence for phonics. I responded:

I did not get an answer, so the next day I wrote:

Let’s see if Rastle responds.  Or is there another pro-phonics researcher happy to have a debate that focuses on the evidence?

For the more neutral observer, who has been told over and over that the science of reading strongly supports systematic phonics, do you find this all a bit strange?  And for anyone who thinks it is just unimaginable that so many experts could be wrong, the unfortunate fact is that it is all too common in science.  Have a look at this excellent talk by Dorothy Bishop entitled “Publication bias and citation bias: A major reason why you can’t believe much of what you read” (https://www.youtube.com/watch?v=cQGD_Uw-Bj8&t=2s Go to 8:00 into talk if you don’t have time to watch to entire talk).  Bishop summarizes research documenting strong publication bias (selectively publishing studies that obtain positive outcomes), outcome reporting bias (highlighting positive effects and ignoring the nonsignificant results within a study), spin (where positive findings are highlighted/exaggerated in the abstracts and summaries, with clarifications and disclaimers buried deep in the paper), and citation bias (where the articles that challenge the positive outcomes, if published, are largely ignored).  Bishop is not discussing phonics in her talk, but all these sources of bias are central to building and maintaining the illusion that there is strong evidence for systematic phonics.

But the most common approach to supporting phonics in the face of (very) limited evidence was not considered by Bishop, namely, “whataboutism”.  SWI does not teach GPCs! SWI is just a form of phonics!  How do you teach SWI on day one of instruction! What about the limited evidence for SWI!  The English curriculum includes instruction in morphology! Phonics can explain the spelling of <dogs>!  Don’t let this distract you from an important question:  What is the basis for claiming that the evidence for phonics is strong?

 

2 thoughts on “Key lesson from my recent debate on phonics: Don’t Talk about the evidence!

  1. Dear Researchers
    I am a parent of a dyslexic, mildly dyslexic myself, and had a dyslexic parent. I just say, at least you are talking about phonics. And spelling. Isn’t this what it is about? In 2003 my first child entered school. I did the usual parent thing, and sent them to school. In second grade the teacher used Spalding’s version of phonics to teach spelling to the kids. I saw great progress in my child’s ability. But 3 years later, with a new Principal, spelling tests were forbidden and teachers were threatened with punishment if they did give tests. Fast forward to 2016 when my youngest is in 1st. No spelling is taught in the public school. The teacher selected a few words child misspelled on a paper, and 5 became spelling words. Sonic( as in the hedgehog comic book) was one. Maybe 15 words for the whole year. Now, in 2021,I am arguing with the school’s special education teacher over if my child could receive some spelling instruction . Of any kind. I requested a spelling goal, I am told, and I quote, “No, because I want him to use the tools the computer offers to help with spelling. At this age, it’s time to start using the technology that is available for writing because that is what his Core teachers will want him to use as well.” My child can not spell the word BREAD. And spell check on the computer only helps if you are close to the word. And not teaching my dyslexia child except by osmosis is not teaching. Yes, I am upset. Argue all you like about phonics and SWI–I say USE them Both. They both have their place. But can we agree that spelling needs to be taught somehow and not just left to chance? Especially for the child with dyslexia. PS. I have taken classes in phonics and am learning SWI. I also have been learning grammar. Learning how to divide the words into syllables, and learning digraphs like oi and oy , with one being in a word and one being at the end of the word, or ea can make three sounds, and which I learned in the phonics instruction class, was huge for me. Now seeing SWI, and breaking off the prefixes and suffixes makes so much sense. But first I need to see the affixes to know they are there. Both explicit phonics and SWI have helped me on my journey. But sounding it out,,,,well, last week it took me 6 tries using spellcheck to get a word I knew how to say but didn’t know spell to be the right one. And my 6th grader is just not going to try that hard to find a word.

  2. Whether the [z] in dogs is the phoneme /z/ or an allophone of the phoneme /s/ is purely academic. Phonemes are abstract, mental categories, not single specific pronunciations. Because phonemes are all in your head, you can argue until the cows come home about what phoneme something is. It’s the orthography that makes that human thought visible as text. Phonics doesn’t even begin to try to understand that.

Leave a Reply

Your email address will not be published. Required fields are marked *