The Writing is on the Wall — of China

I received an inquiry online from a follower in China — I’ll call him Earl — that’s a fairly common question about morphology. I was impressed with how closely Earl has worked to read and understand my work and the understanding it offers; that’s no mean feat even for a native English speaker.

Earl’s English is excellent, and his drive to understand a writing system so different from that of his mother tongue is admirable. His question was about the suffix addition patterns — specifically consonant doubling, when it does and doesn’t happen, and why. Here’s what he asked:

I have some questions about this rule for a long time, because I find some cases don’t follow that rule. For example, “write+ing–>writing” is ok, but does “write+en–>written” still follow the rule? According to the rule it should be rewritten to “write+en–>writen”, because there is a silent “e” in word write, namely, non-syllabic vowel, so, why is not that case? Another counter-example is ladle, the word ladle sum is “lad+le–>ladle”, well, lad is a base accorded with the condition of base part of the rule of consonant doubling, because suffix “le” in word ladle is syllabic equivalent to vowel suffix, so the word ladle should be laddle in accordance with that rule, but in fact, laddle as the written form of ladle is wrong. Wow, I’m still confused with this so far. Could you explain it to me for this exception? I’m looking forward to your response. Thanks a bunch. Best regards!

Clearly, Earl already has a pretty deep understanding of how this suffix addition pattern ought to be working, and he did a great job of articulating that understanding. Of course, I don’t traffic in confusion or exceptions, so it was my hope to re-frame the question in a way that makes more sense, not just provide an answer.

So here’s my explanation:

Hello, Earl,

A word sum — or lexical algorithm — is a self-checking mechanism. You are correct that you cannot synthesize an <-en> suffix to the base <write> and get <written>:

< write + en → *writen >

Therefore, < write + en > cannot be the deep structure of that word. We can also use the lexical algorithm analytically, in the other direction, to determine the deep structure of a written word:

< written → writ + en >

That’s a coherent word sum. The final <t> in the base element, <writ>, doubles when we add the vowel suffix, <-en>.

In an orthographic word sum, we’re not adding a suffix to a “word,” but to a base element, which may be free (able to stand on its own as a word), or bound. Both <write> and <writ> are free forms; they can be realized on their own as words. They can also each take a set of prefixes and suffixes and other bases for compounding, like in < type + writ + en > or < writ + s > or < hand + write + ing >.

It is clear that the two base elements, <write> and <writ> are related, but they are etymologically related — like word first cousins — not morphologically related. Words like writing and writes share the <write> base, and are like siblings. But <write> and <writ> both have their own morphological families.

The same thing is true with wise & wisdom: you cannot add <dom> to <wise> and replace the final <e>; rather, both <wise> and <wis> are base elements. One is free and the other is bound. They share a common ancestor, but the final <e> marks them as distinct written forms.

The same thing is also true with do & did — you cannot add a suffix to <do> and realize <did>. Clearly they are related, but they are word cousins rather than word siblings. Likewise foot and feet, or mouse and mice.

The reasons for having distinct but related forms in the same paradigm are historical: plural nouns like men, feet, mice are the result of a historical process called i-umlaut, while past-tense and past-participle verbs marked with a vowel change (like write~wrote~written or wear~wore~worn) are strong verbs, a kind of verbal paradigm common in Germanic languages.

Now, on to <ladle>: Again, the word sum is a self-checking mechanism, and you are correct that

< lad + le → *laddle >

Likewise, < lad + en → *ladden >, and < lad + ing → *ladding >.

But remember, we can use word sums analytically to reveal the real base element:

< ladle → lade + le >

The <lade> can be considered a free base element, but it’s archaic, so for present-day, current English, one might consider it a bound base. It is cousin to <load>, and that relationship explains the <oa> spelling in load. A “bill of lading” is like a packing slip for a large freight or commercial shipment. It the document that lists what is in the load.

Your assessment of the syllabic <l> in the <le> suffix as being vocalic and thus replacing a final <e> is accurate. We see that in < sidle → side + le > as well.

Your English is excellent, and you’ve clearly done a lot of study of my understanding! Keep up the great work!

Earl was really grateful for that response, and he articulated that his gratitude was motivated by his ability to understand:

Thanks a billion! I think I got it, wow, your excellent explanations make me so comfortable that I am touched so much. Self-checking mechanism is a kind of hint of morphological boundary. Thanks again for your work and works.

I really appreciate Earl allowing me to share his question and his understanding with everyone. I wouldn’t trade this kind of dialogue for all the tea in China.