Hello all you helpful spelling folk.  I have incorporated lots of the word sum strategies and tips from Pete and Gina into my lessons, but there is one thing that comes up occasionally with my one-on-one dyslexic tutorees that for which I know it not the answer.  I understand that some of the choices for spelling with ea  vs.  ee  can be explained because the vowel sound changes in the base depending on affixes added on.....for instance, we spell "heal" with an ea so that the sound of ea can change to be in words such as "health".   One cannot follow this logic, however, with the words "keep", "sleep", and "weep" which, my very smart students note, should have their past tense spelled "keeped", "sleeped", and "weeped"  because, we all know, that suffix ed can say /ed/, /d/ or /t/ and "vowel teams" sometimes have more than one sound.  Why do these words change completely for their past tense when others strictly follow the word sum/matrix word-building guidelines?  I hope I have explained this understandably.  I'd sure love to hear back about the reasoning, and how to present this to my students.  Thank you,  Donna

Comments (2)

  1. Natalie

Pete, I love your clear and concise explanations. So valuable!

Thank you for your efforts to teach!


  1. Peter Bowers

Hello Donna!

Great question. You've given the evidence for your thinking and the source of your confusion. You also give no sense in your question that because you don't understand these spellings, that this must be evidence of a spelling exception. Instead you do what a scientist should do -- identify the edge of their understanding, and seek guidance.

I'll offer an initial response to see if I can help point you and your students in a generative direction.

You write:

I understand that some of the choices for spelling with ea vs. ee can be explained because the vowel sound changes in the base depending on affixes added on.....for instance, we spell "heal" with an ea so that the sound of ea can change to be in words such as "health".

This is a perfect foundational understanding on which to explain your surprise at the spellings of <kept>, <slept>, <wept>.

Before we get to the content of your question, I just want to model the use of linguistic symbols, and terminology that will help bring precision to your question. I think that you will find that if you take on the challenge of learning to use these scientific conventions in your correspondences, and with your students, I think you will find that it forces a precision in your own thinking -- and thus that of your students. I'll re-write that section using those conventions and terminology. Consider the contrast, and then I'll add a note about it:

I understand that some of the choices for spelling with <ea> vs. <ee> digraphs can often be explained because the vowel phoneme differs in the base depending on affixes added on..... for instance, we spell <heal> with an <ea> digraph so that the <ea> digraph that represents the phoneme /iː/ ("long e") that we have in the word "heal" and the /ɛ/ ("short e") in words such as "health".

Note how the use of angle brackets <> for orthographic information makes the orthographic structures you are pointing to stand out to the reader from the rest of the text. When I see angle brackets, I spell out what is inside rather than try to pronounce it. This means I can use the quotation marks when I want to cite a word that I intend the reader to pronounce. I've also used the slash brackets "/ /" to mark phonological information. I've used IPA symbols there, but since I know many are just learning those, I've put in parentheses the terminology familiar in schools for "long and short vowels".

The other subtle point I changed here is the language about how the phonemes "change". The way you have written it, I get the sense that you look at the way we say the word "heal" as English speakers is somehow the "natural state" of the pronunciation of the base <heal>. Instead, I try to remember that no morpheme (base or affix) actually has a pronunciation until it is in a word. I've tried to word my version of your statement so that there is no sense that one pronunciation is the default -- just that the <ea> digraph fits the spelling of this base best because it is able to represent any of its phonological representations.

There is a lot in there to think about. Don't feel you have to do all of that right away, but I'm hoping you see the precision in thinking that comes with precision in language. I also know that in posting on Real Spellers, there is the extra issue that to write anything within angle brackets, you need to bold the text within the angle brackets. That's a bit of a hassle, but Matt Berman had to work hard to find any platform that would let us use angle brackets at all!

Now for the next part of your message...

"One cannot follow this logic, however, with the words "keep", "sleep", and "weep" which, my very smart students note, should have their past tense spelled "keeped", "sleeped", and "weeped" because, we all know, that suffix ed can say /ed/, /d/ or /t/..."

Good attempt at the IPA! The only bit that needs revision is for the syllabic pronunciation of the <-ed> suffix. That one looks like this: /ɪd/. The way you have written it, many might think of it as being pronounced like we pronounce the final <ed> in a word like "bed".

More importantly, your students are sharp noticers. But I would caution them on their sense of what "should" be. Instead, I might express what they are thinking in this way:

We are surprised by the spelling of the words <slept>, <kept> and <wept> because they are clearly how we write the past tenses of <sleep>, <keep> and <weep>, and usually we expect to mark the past tense of a word with the <-ed> suffix. As well, we wonder why we don't spell the base with an <ea> digraph that would allow for the different pronunciations of that base in the present and past.

Does that re-writing represent you and your students question? The difference I'm trying to signal is that actually with these words you have found evidence that what you thought "should" be the case actually is not. That is not spelling being weird, it is evidence your students noticed that there is clearly more than one way to mark the past tense in English. You've collected evidence that while adding the <-ed> suffix to a base or stem is the default way to mark the past tense, there are clearly other ways too. Your real question is about understanding these other ways of marking past tense.

I'll share a few ideas about this.

One thing is that, as far as I know, the <-ed> suffix is now the default way of marking past tense. Whenever we make up a new word in English that can be inflected into a past tense, the only way we do it now is adding the <-ed> suffix.

However, as your word scientists have noted, there have been other ways of marking past tense. You might want to have your crew go on a treasure hunt of words that mark past tenses in ways other than using <-ed>. Some interesting ones that are not the same as your three examples include <do> / <did>; <run> / <ran>, and even special ones where there is no spelling or pronunciation connection between the present and past tense forms: <go>/ <went>.

As you collect such examples, there are some characteristics to note. One is that any of these words will be very old. I don't know for sure, but I expect these forms to be of Old English origin.

The examples you found are particularly interesting because as you note, currently it seems like the combination of the <ea> digraph in the base and the standard <-ed> suffix would be the best fit. However, another way to look at this is that the fact that these words do not follow that expected convention is offering evidence that these words are very old. Think of it like finding an extremely old coin in your change one day. The fact that a different design than you are used to is on your coin is not evidence that the coin is "wrong" just that you happened to have encountered a very old one that still marks its history.

There is also an orthographic concept that is appropriate to address here. Linguists do talk about "non-productive" suffixes and even "fossilized suffixes". It turns out that there was a time when we did add <-t> as a past tense suffix, but we do not anymore. That's why it is non-productive. We don't use this suffix to produce new words.

But you may notice another issue. If we think of a word like <slept> as having a non-productive <-t> suffix, then what is the base it is being fixed to? I don't know a current base <slep>. So to conceive of this word as having an old <-t> suffix, we have to also conceive of it having an old base <slep> for the idea of "sleep".

We can hypothesize that at one point this is how that word was analyzed, but whatever our analysis is, can't say that the words <sleep> and <slept> share a base element. There is no analysis that leaves them with the same base. So how do we explain the orthographic connection between these words?

You were asking this question when you wrote this: "Why do these words change completely for their past tense when others strictly follow the word sum/matrix word-building guidelines? "

Well, just because two words don't share a base element doesn't mean that they are not related. (Note, I used the word "element" in "base element" to signal a written base.) What other kind of meaning-spelling relationship could they have?

I am sure that these words ARE etymological relatives. So I am quite happy to treat words like <slept>, <wept>, and <kept> as free bases in current English. It may well be that historically they were analyzable. But that doesn't mean I have to treat them as analyzable today. If I found current words with the base <kep>, <slep> and <wep> with suffixes other that <-t>, then I would be happy analyzing them. For now, I am happy with the understanding that you've found words that are etymologically related to the words you thought should have been the base. But since there is no evidence of a word sum linking the base you expected with the past tense spelling we have, we just have to reject that initial hypothesis. Excellent!

Finally, while many describe such words as "irregular verbs" I reject that kind of terminology that has the connotation that there is something wrong with this way of marking past tense. Linguists have a non-pejorative term for such verbs. They can be called "strong verbs". They are conceived of as "strong" because they do not just do the default thing that most verbs do. Verbs that just follow the standard conventions are referred to as "weak verbs".

So hats off to you and your student spelling scientists. You've noticed an interesting set of words that don't do what you expect. By asking the question, you've helped identify a whole class of verbs called "strong verbs" that explain many words beyond the ones you asked about.

You've also given yourself a key reminder that when your word sum fails to resolve in the way you expect it to -- recognize that as evidence that your hypothesized morphological structure is rejected! That means it's time to look for another way to understand the relationships of the words you encountered. An extremely common explanation will be that the words you thought were in the same morphological family actually are not. Instead the connection you perceive is likely to be the result of an etymological relationship. Your words do share a root, they just don't share a base!

I recommend that you (or any readers) take a look at THIS LINK on morphological and etymological families. (http://www.wordworkskingston.com/WordWorks/Morphological_vs._Etymological_Links.html)

And if you have a copy of the RS Tool Box, there is a great theme 4G on "When to use the suffix <-t> instead of <-ed>. Email me if you are curious about that (peterbowers1@mac.com).

Cheers, and feel free to follow up with more questions!


There are no comments posted here yet