This past week, I had the pleasure of working with the staff of the Lower School at the University School of Nashville. Every workshop, seminar, and discussion about orthography leads to new learning for me; nothing motivates like understanding, and I am always moved when I find my orthographic foothold more secure than it was before.
I’ve been giving public and private workshops on English spelling for the better part of a decade. Not only do I learn more about language, but I also learn more patience, empathy, flexibility, and diplomacy. I’ve grown in my capacity to respond to questions that destabilize me or make me uncomfortable. There are still a million things I’m not good at as a scholar and a teacher, but the organic and unplanned trajectory of my work continues to bear fruit.
One of the things I’ve gotten better at understanding over the years is the etymological governance of grapheme choice. It’s captivating and revelatory in a way that morphology alone cannot be. Etymology is, of course, an unignorable factor in the writing system. As I told teachers, you don’t have to like it, but hey, I don’t like the San Andreas fault and so far it’s still there.
Now, etymology is not the only factor that governs grapheme choice. Phonology can and does play a role: the <k> in <provoke> does not hearken back to an ancient root (compare <provocation> and the Latin vocare), but maintains the phonology of the /k/ before the <e> necessary to mark the phonology of the <o>. Place value — the position of a grapheme within a morpheme — also can optimize one grapheme over another. The famously “final” <ck>, <tch>, and <dg(e)> graphemes are, of course, base-final, not just word-final.
Many of my posts on this site deal with grapheme choice, and my LEX Grapheme Deck is a tool designed to help clarify the processes involved in orthographic phonology. One grapheme choice, however, that has always troubled me is the <j> in words like major and majesty. The base element is demonstrably <maj>, but the grapheme <j> is typically not final. The default final spelling for /ʤ/ is <ge> (see large, cage, magic, and the like). The <dg(e)> spelling follows a single vowel letter (see bridge and edge etc.).
Historically, <j> and <i> were the same letter. The non-finalness of <j> is part and parcel of the non-finalness of <i>. Likewise, the letters <v> and <u> have a common ancestor, and neither is final in English either. We can explain the <j> etymologically, because the root of this word family, the Latin maior (adj. “great”), has an <i> in it. However, it’s not unheard of for <g> and <i> to mark a relationship, as evidenced by such etymological cousins as tail~tag, paint~pigment, frail~fragile, chain~chignon, complain~plangent~plague, strain~stringent, rail~regulate, maybe rain~irrigate, and many, many more.
So why not spell this word family with a <g> instead of a <j>?
In Nashville, I mentioned major and majesty as examples of a base element that ends in a <j>. I said I hadn’t figured out why yet, but I wasn’t about to call it an exception. As soon as I said that, I realized that only a <j> would work in the words <major> and <majuscule>; because the <j> is followed by an <o> in one case and a <u> in the other, a <g> would not work as we’d be left with spellings unsuited to the /ʤ/ phoneme: *<magor> and *<maguscule>. While this would work phonologically for *<magesty> (a frequently unnoticed misspelling, by the way), it wouldn’t work for the whole word family. Etymological relatives like <mega> and <magnate> use a <g> as a <j> would not work. These words all point back to a PIE root *meg(h), denoting “great.”
Moreover, words that do have an <mage> base in English, like <magic>, <mage>, <magi>, and <magus>, have a totally distinct etymological origin. Their PIE root is *magh (1), denoting “to be able, to have power.” Relatives include may, might, machine, and main. The <maj> offers us a differentiation from this <mage>. Since the meanings “great” and “having power” could easily be conflated, careful study is needed to peel apart these two historical families.
I’ve wondered about that for years. Now I get it.
How great is that?