English spelling has five kinds of logic

This Wednesday at 6pm AEST I’m presenting Learning Difficulties Australia’s free Weekly Wednesday Webinar (yikes, but I’m back on LDA’s Council so thought I’d better contribute. Also, here in Melbourne we’re going into Stage 4 lockdown, our Jacinda The Warrior Princess hammers ready to smash the virus, so I don’t have an awful lot else on).

Here’s a bit of a preview, to help you decide whether to attend live and ask me Hard Questions in the chat (register here) , or perhaps watch it later on Youtube.

The title of the webinar is “English spelling has five kinds of logic”, and I’m hoping it helps teachers and parents better understand and explain our superficially bizarre spelling system. Much of what I’ll say won’t be news to LDAers or other experts, but I hope they find some useful ideas, angles and resources, and maybe a few laughs, and will want to send the link to others.

English spelling makes total sense once you know where it comes from and what it’s for. English is a mishmash of languages, squashed together and warped over time. You can hear some of the historical changes in this punny video, in which (interestingly, for word nerds) Elizabeth I is played by Ben Crystal, son of brilliant UK linguist David Crystal.

Our spelling system is a code that represents speech sounds (phonemes) with letters and letter combinations (graphemes), with special spellings reserved for many meaningful word parts (morphemes).

Sometimes our spellings represent the way words used to be pronounced, in which case I usually tell kids “that’s how people said it in the Olden Days”. I’m not aware of any evidence that lots of etymological detail helps kids much with spelling, but if they’re keen, we look words up in this etymology dictionary.

The phonemes in one’s speech vary slightly depending on accent, and many sounds (ch, j, ay, ie, oe, ue, ou, oy, air and ear) are actually two sounds (I can’t explain this well in print, but will do so in the webinar) but for making-sense-of-spelling purposes in my accent, it’s easiest to divide them into 44 phonemes. Click here for my daggy video listing them.

Once you’re a proficient reader and speller, the idea that words are made of sounds seems obvious. It’s easy to forget that this is not at all obvious to beginners, and thus fail to teach it well (this is called the Curse of Knowledge).

US Professor Mark Seidenberg’s website has videos showing how sounds shmoosh together in words making them hard to discern (plus lots of other interesting stuff), and how each phoneme (an idea in our heads) is produced in more than one way in the mouth (allophones), depending on which other sounds are shmooshing up against it.

It’s also easy to forget how confusable many of our speech sounds are in connected speech. The reason many preschoolers cutely say /t/ instead of /k/ is because they’re both voiceless sounds produced in the same manner (stops), but /t/ is further forward in the mouth. In connected speech, they sound almost identical, as this Great Mask Debate video shows:

Discerning sounds is not just an auditory Thing, it’s also affected by visual information, because when our brains receive conflicting information they try to make sense of it:

Speech sounds must be the starting point for any logical discussion of spelling. The relationships they have to letters I plan to discuss in the webinar are based on the ideas of Dr Diane McGuinness, author of Why Children Can’t Read and What We Can Do About It, a ye oldie bookie (1999) but a goodie. She wrote (and many people now agree, see for example this Spell Links video) that most “sounds of letters” phonics programs are teaching the code backwards, which makes it hard for many children to identify and remember the patterns.

Writing was invented as a way of transcribing talk, so we should be teaching children that “sounds have or represent letters/spellings” (not “letters have sounds”). It’s a seemingly subtle distinction, till you try to explain to someone what the “sound of the letter e” is based on its appearance in the following common words: the, went, we, there, like, one, are, they, have, her, because.

Can’t do it? Me neither, unless you have half an hour. The letter “e” is used in the spellings of many different sounds, but when you start from the letter and a bunch of high-frequency words (as many children are encouraged to do), you end up totally confused.

Sounds are the backbone of our spelling system, but as they are fast, invisible and ephemeral, they too often get short shrift. Five kinds of mapping logic relate sounds to letters:

  • One sound can be written with one letter (one-to-one).
  • One sound can be written with two letters (one-to-one[two]) e.g. sh, ou, or sometimes even three or four letters e.g. dge, augh (one-to-one[three or four]).
  • Most sounds can be written several ways e.g. out, cow, drought (one-to-many).
  • Some spellings represent more than one sound e.g. out, you, touch, cough, soul (many-to-one).

In addition, meaningful parts used to build longer words often have specific spellings. The suffixes -y, -ee and -ie are all pronounced /ee/ in the words bossy, runny, classy, employee, devotee, refugee, Aussie, duckie and cutie, but the first suffix creates adjectives, the second creates recipient nouns (the employee gets employed, the refugee gets refuge etc) and the third creates diminutives. These spellings help skilled readers extract meaning efficiently.

As Mark Seidenberg and many others have shown, the best way to become a good reader is to have clear, explicit, early instruction on how sounds relate to letters, and then do a large amount of unsupervised, implicit learning, allowing one’s brain to gather data about the statistical regularities of the language. But anyone who’s ever gathered data knows that if you don’t have a clear idea of what you’re gathering, it can all be a bit garbage-in, garbage-out.

Diane McGuinness was onto what we now call Statistical Learning over 20 years ago, and wrote:

Perceptual learning or perceptual memory occurs by a process called ‘probability matching’. The brain is specially adapted to learn visual and auditory patterns automatically by seeing or hearing them a few times. If, and only if, the eye is trained to look at the letter(s) representing one sound, these patterns will be coded without effort. The brain searches for ‘reoccurring regularities’. What is frequently encountered will be remembered. What is very strange or discrepant will be noticed.

“When a system, such as our spelling system, cannot be easily classified, the solution is to organise it according to its probability structure, and teach it by systematic exposure to this structure. This probability structure must be based on how the code was written.” (Why Children Can’t Read and What We Can Do About It, p102-3).

I hope that gives you enough information to decide whether to attend my Wednesday Webinar, without too many spoilers. I promise there’ll be some practical teaching/intervention ideas, including things that work online, in the mix.

17 thoughts on “English spelling has five kinds of logic

    1. alison Post author

      Hi Frank, I think the argument that morphology should be taught BEFORE or INSTEAD OF phonics is ludicrous, but that the word sums approach is a useful way to show kids who are already able to read and spell lots of short words how to build longer words. It’s not actually that different from what many phonics programs already do, for example if you look at the Phonic Books Amber Guardians workbook, there are lots of activities that involve pulling words apart into prefixes, base words/roots and suffixes, and learning what the spellings of common morphemes are, and what they mean. The idea that inquiry is the only valid, child-friendly way to teach is also I think rather absurd. I could teach a room full of adults about the phonemes of English by helping them to inquire into the operations of their own mouths and ears, and discover what linguists all know anew, which would take a few hours. I’m yet to find a room full of adults who are interested in doing that. They’d much rather I just taught them the main things linguists know about the sounds, rapidly, directly and explicitly, so they can get on with using that knowledge. All the best, Alison

      1. Lisa

        What fun clips Alison! Isn’t David Crystal’s son a character!! Interesting to touch on statistical learning again after the recent article in the Speech Pathology Australia clinical practice journal. I think we are going to be hearing more about SL in the future. In adults who become competent at their occupation, we call the same thing ‘Pattern Recognition ‘, its an implicit learning that develops to a sophisticated yet almost subconscious level in those with lots of experience.

      2. Frank

        Thanks for your thoughts. This is some what I have been surmising, I just wonder why they are making such a big deal of what they are doing – when it is just proper spelling instruction at the end of the day, with a little bit of a different emphasis against the grain of mountains of research.

    2. LEX

      I don’t understand why anyone asks Allison, who has zero experience with SWI, what her opinion is about SWI. True to form, she miss apprehends SWI in her response. No one but no one who actually works with SWI claims that morphology must be taught prior to phonology. This is a constant false refrain in the anti-SWI phonics crowd.

      If you want to know about SWI, it’s best to learn about it from people who understand it. Allison is not among them.

      1. alison Post author

        I’m so glad to hear that. Perhaps they could have worded it better when in Beyond Phonics: The Case for Teaching Children the Logic of the English Spelling System” Jeffrey S. Bowers and Peter N. Bowers wrote: “Indeed, it is often claimed that morphological instruction should occur only after phonics… However, these claims and findings provide no evidence against the claim that SWI should be adopted from the very beginning.” I start teaching morphemes when introducing CVCC words, with regular plurals, but I’d be very interested to know how morphology could be taught from earlier than that at a code level. At an oral language level, of course, that’s what speech pathologists do, and I’ve been doing it for 30 years.

        1. LEX

          Your comment is awaiting moderation.
          “However, these claims and findings provide no evidence against the claim that SWI should be adopted from the very beginning.”

          Yes. And SWI is not “morphology.” SWI studies the *interrelation* of morphology, etymology, and phonology, from the beginning. Phonics teaches “phonology” (actually, it’s a mish-mash of phonology and phonetics) using disembodied “phonemes” (actually phones) in drills, detached syllables, and nonsense words. SWI studies orthographic phonology IN MORPHEMES from the get-go, because that’s how phonemes are actually organized.

          It’s not poorly worded; you understand it poorly.

          When you say, “I start teaching morphemes when introducing CVCC words, with regular plurals,” you sound like every other phonics practitioner on planet Earth. That’s not unique. And there is no researched evidence that that sequence is empirically The Correct One or even the best of several possibilities. It’s just what you were taught to do.

          I am so glad to know that you’d “be very interested to know how morphology could be taught from earlier than that at a code level.” You’re in luck! Pete Bowers has online classes that are friendly to Australian time zones. So do I. In fact, I’m about to start a new round of Introduction to Structured Word Inquiry. So instead of imagining (or failing to imagine) how not just “morphology” — but the INTERRELATION OF MORPHOLOGY, ETYMOLOGY, and PHONOLOGY — can be taught from the get-go, people like you can actually study it!

          So again, please stop mischaracterizing SWI as just morphology, or as morphology first. That’s a false representation.I get that you can’t imagine what it looks like; likewise, I also cannot imagine what driving a team of sled dogs looks like, because I have never studied it have I even tried. So I don’t pretend to be an expert about it, and don’t invite people to ask me questions on the subject.

          My apologies for the voice-based typos in my original post.

          1. alison Post author

            Now you’re making generalisations about phonics, but there are good and awful phonics programs, I agree that some of the things done in the name of phonics need to stop. I’m personally not a fan of teaching nonsense words, though I use them for assessment, because the kids I work with are having a hard enough time learning real words, and the triangle/connectionist model which makes the most sense to me of the research I have read contains a semantic processor (for morphemes, both bound and free). However, I ddon’t work with mainstream learners, and take the point that many nonsense words are syllables in long words and that skilled readers process multiple layers/patterns in words – phoneme, grapheme, morpheme, onset-rime and syllable. I wasn’t taught to teach plurals first out of all the morphemes, they are just developmentally early and orthographically simple, so it makes sense to teach them before the rest. I don’t think there is such a thing as The Correct sequence to teach, but there are some key principles (simple to complex, developmentally easy to less easy, coverage of all the phonemes starting with the most common spellings, coverage of the most common morphemes starting with the inflectional ones etc) that apply to designing any teaching sequence. I’ve got about a million things on but I will have a look at the courses you describe, perhaps this is all in fact a big misunderstanding, and we can all stop having what might be just a tiresome chicken-and-egg argument, and turn our attention to making sure teachers are equipped with the knowledge of phonology, orthography, morphology and etymology they need to do a good job of teaching children to read. This knowledge is so sadly lacking in initial teacher education, and poorly reflected in the way many or perhaps most children are currently taught. I hope one day you get to drive a dog sled, it looks like great fun, if we had snow here I’d be lining up to try it (and probably falling over and getting laughed at). Don’t worry about the typos in the last post! I can fix them if you like. Alison

      2. Frank

        I’ve also had discussions with people heavily involved with SWI and get all sorts of interesting responses – like questioning well established scientific findings in the field of reading. Then there are people like yourself who says Louisa Moats “lies to children” and chooses to make fun of people by calling them “phombies”, so I’ll ask whoever I want what I want to get a well rounded view of the situation. This is probably a healthy thing to do when studying something that has cult like tendencies.

        1. LEX

          Gee, Frank, which of these two things do you imagine happened first:

          1. I called people Phombies

          2. Proponents of phonics continually and often willfully mischaracterize my work and the work of my colleagues, misrepresent the writing system, repeatedly publish a bunch of etymological falsehoods, and accuse my colleagues and me of being in a cult.

          I’ll give you a hint: It was humber 2.

          I’m going to ask all my friends what they think of you, too, just to get a well-rounded perspective.

          I’d also like to remind you that “questioning well established scientific findings” IS WHAT SCIENCE DOES.

          1. LEX

            I remember you, Frank. You’re the one who said, “the biting and overzealous nature of that Gina woman over at LeX” that gave me my current Twitter handle. Thanks!

          2. alison Post author

            Hi Frank and Lex, just letting you know that I am coming perilously close to turning off both your blog comments for fear that one of you will say something defamatory which I’ll then get the blame for even though I didn’t write it, as the publisher of this blog. My website is not a place to have a slanging match, please go and do that on Twitter like everyone else. Thanks and have a nice day! Alison

  1. Wendy Savaris

    Excellent as always Alison. Thank you so much for all you do!
    You will smash the Webinar and we will smash the virus. xo

  2. Jenny Bode

    Dear Alison, I’m sure I speak on behalf of millions of your ‘faceless/nameless’ followers, you are a LEGEND! Thank you for sharing your knowledge. How awesome that Pam Kastner lists your one videos in her top 10! I wouldn’t miss your webinar for anything! Jenny

  3. Lindsay

    Many thanks for a very useful webinar Alison – I always learn something new when I listen to you, which is great – I always to add to my therapy skills. Please can you let me know if we can use those wonderful visuals you used to explain the first 3 “logics” – i.e the s+h with house and shapes analogies, and most sounds are spelled a few ways (houses)? Do we need to purchase them from somewhere?


Leave a Reply

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