Of Mice and Keyboards

is for Hamlet


Every now and then as I spell my way through syllabic text, I find a character that looks remarkably like—by golly, it is an H. But what’s it doing there? How did get into the syllabary without any accom­panying vowel?

Turns out there are at least three ways of representing the [h] sound in UCAS. Along with the maverick ᕼ, we’ve got the series ᕵ, ᕹ, ᕷ and . . . ᓯ, ᓴ, ᓱ. The first set is called “Nunavik H” by the Unicode folks. I will take their word for it. If the second set looks suspiciously familiar, it should: it’s simply the s series. Some dialects pronounce it h. But it’s “really” an s, so that’s how it is written. Unless you live in Nunavik, I guess. It’s the same principle as r in Europe. The letter is pronounced in several different ways, but no single language—

r r r

Oh, all right. No single dialect uses more than one. One of the Great Divides in Norwegian is between the “front” r (the alveolar trill, as in Spanish and Italian), used in Eastern dialects, and the “back” r (the uvular trill, as in French, German and, yup, Inuktitut), used in Western dialects. In some places you will also hear the English-style “fuzzy” r sound—but it isn’t an r, it’s a variant of l, so it doesn’t count. Like Inuinnaqtun, where that same not-really-an-r sound crops up in place of j. Or, if you insist, y. Given the choice, I’d let the Moravians pick my orthography rather than the English. But nobody asked me.


The ᕼ is a different critter entirely. It’s not a variant of something else—an allophone, if you want to be show-offy about it—it’s an unequi­vocal H. In Inuktitut you meet it only in loanwords from English: ᕼᐋᑭᖅ (haakiq), ᕼᐊᕗᑯᕌᕝ (havukuraav), ᕼᐊᒪᓚᑦ (hamalat) and so on. Attaching a vowel is obviously impossible. The H letterform can only point in two directions: up-and-down, or sideways. You need at least three positions, preferably four, to fit into the UCAS syllabary.

Incidentally, that ᕼᐊᒪᓚᑦ (Hamalat) has a firm place on the list of terms that the Hansard’s trans­lators need to reach consensus on. The full range of spellings requires a Regular Expression: Ha*m+a?l, with final (?!e) if you need to weed out English forms. That’s:

H: always capital­ized

a: one, two—or none at all, because some writers can’t stop themselves from treatingas “ha” with built-in vowel

m: one or two

a: optional

. . . and from there on it’s an inflectional free-for-all.

b is for Aivilik


To keep the ᕼ company, there is what Unicode calls the “Aivilik B” or rather ᖯ*, as in—I assume—ᖃᖯᓗᓈᖅ (qablunaaq). The ᕼ occurs only word-initially, the ᖯ only medially, so they’re capital­ized to fit.

I do not get along well with the Aivilik B. In the first place, I never use it; in fact I’ve never even seen it outside a font chart. In the second place, it looks exactly like ᑲ (ka). And finally, I’m constantly typing it by mistake for ᑉ in words like ᑯᑉᓗ (kublu).

For those who are into Terminology: the [b] and [p] sounds occur in complementary distribution, becoming voiced before voiced consonants. So there is no need for separate graphemes. The same goes for [d] and [t].

Obvious solution: edit my keyboard to get rid of it. Type b, and my computer obligingly behaves as if I’d typed the p I really meant to use. (I am a typographic wimp. I use the Inuktitut-QWERTY key layout. Also Hebrew-QWERTY, Devanagari-QWERTY, and so on.) Some more editing and I may be able to say goodbye to persistent typos like ᓐᒋ for ᖏ and ᖅᖃ for ᖅᑲ. But for now, I’ve settled for the most important addition: typing our jolly little inuksuk ᐀ directly from the keyboard.

If I could solve all problems by editing a keyboard layout, I would be a happy woman.

* If your only syllabic font is Pigiarniq, you can’t see this character. It looks like . . . drumroll . . . a lower-case sans-serif b.

Conversely, if you don’t have Pigiarniq—or its elusive cousin Uqammaq—you can’t see the inuksuk. No other UCAS fonts have this character. But that’s another story.

S is for Syllabary


Right about now, some know-it-all is pointing out that we wouldn’t have these problems if the language switched over to the Roman alphabet. Well, it is obviously important to be able to write “hockey” with as little difficulty as possible. Cursory research suggests that the convenience of computer manu­fac­turers—specifically, that computer manufacturer—is another prime consi­deration. There is also, of course, the desire to become more like the English, because they’ve always been so nice to us, and respectful of our language and culture, and . . . Uhm. Something like that, anyway.

Once upon a time, it was generally believed that the alphabet represented the highest achieve­ment in human writing systems. Well, of course it did. It’s what the Europeans used. It was also believed that all human languages really wanted to be Latin. Several centuries of missionary linguistics illustrate the point in a pretty dramatic—or, if you prefer, horrifying or hilarious—way. How many pages can you fill with “conjugations” of nagligijanga* that are all exactly the same, over and over again, except for a change in affix? How long can you do this before someone starts complaining about the waste of paper, and wondering what they’re not learning? Answers, unless you are a Moravian: (a) Infinite and (b) Forever.

Anyway, ’tain’t so. (Dual. ᐱᙱᑦᑑᒃ.) Ask the nearest passing linguist. Most languages—including, as it happens, English—have no interest whatsoever in being Latin. You knew that. But also: any language that can be written in a syllabary should be. (English can’t. It’s got too many consonant clusters and closed syllables. The same applied to Greek, earlier in the same historical sequence.) Rough numbers: If you use a logographic writing system such as Chinese, it takes at least ten years to attain functional literacy. If you use an alphabet, about four years. If you use a syllabary: months. Korean can be learned in an afternoon. The entire Cherokee people became literate within a few years. All those medieval Japanese novels were written by women—in the Hiragana syllabary—because the men were, among other things, still struggling to learn Chinese characters.

It would be nice to think that the missionaries were ahead of their time, or exceptionally enlightened. Nice but not likely. The only ones who really understood the language were the Moravians—and they stuck with Roman type. (Count your blessings: at home they used fraktur. Roman was reserved for non-Germanic languages—or books that were likely to be read by foreign academics, like Kleinschmidt’s classic Grammatik der Grönländischen Sprache .) I’ve got a nasty suspicion that the missionaries figured a primitive writing system was about all the primitive Eskimo intellect could handle.

Sometimes people really do do the right thing for the wrong reason.

* Or, if you prefer, nagligivaa. A glorious detail of missionary linguistics is the unwritten rule—it may even have been a written rule—that the Canonical Transitive Verb, used in all printed conjuga­tions, is . . . love. Just try to ignore that pesky gi and gusuk business. If your friendly local missionary doesn’t understand it, it can’t possibly mean anything important.

Unlike other missionaries, the Moravians had no interest in teaching “civilization”. Their only aim was to spread the Word. It is no coincidence that the most useful missionary grammars were produced by Moravians.

. . . mit theilweisem einschluss des Labradordialects, if you want to be precise about it.

ᖃ is for Qallunaaq


If you look closely at the UCAS syllabary, you’ll see that the little doodads attached to ᕿ, ᖏ and ᖠ are flags that say “I was designed by a European.”* In fact the whole rotational system may be telling us the same thing. “Natural” syllabaries use a set of 80-200 entirely different characters. But if your writing system was developed by people who really don’t get the syllabary concept, you’re liable to get stuck with characters that have a “consonant” piece and a “vowel” piece. So a pointy element

p = ᐱᐯᐸᐳ

plus a downward-facing element

i = ᐃᐱᑎᕕ

gives you a pointy downward-facing character

pi = ᐱ.

Put a dot on top, like the familiar European macron, and it becomes long:

pī = pii = ᐲ.

You’ll even find people calling the syllabary an abugida—a system like Devanagari where vowel signs are added to the base consonant. But that’s heading in the wrong direction. A true abugida can always be broken into physically separate pieces:

p(a) + ि -i = पि pi

and so on. You can’t do that with ᐱ.

All of that is secondary, anyway. So’s the overall layout—the “alphabet”, if you want to call it that. It’s a pretty linear term, and sylla­baries aren’t linear. If the missionaries had been thinking, they would have laid out the rows in pi, pa, pu order. Makes the mirror-imaging a lot easier to learn. No problem. Write up your own chart and you can arrange them any way you like.

For me, every­thing fell into place as soon as I’d internalized the pattern ᐊᐃᐅ = aiu. That gives you your starter set:

ᐸᕙᑕᕋ   ᐱᕕᑎᕆ   ᐳᕗᑐᕈ

all facing the same way.

The second pattern is ᑭᑲᑯ = kikaku, which gives you almost everything else:

ᑭᕿᒋᖏᒥ   ᑲᖃᒐᖓᒪ   ᑯᖁᒍᖑᒧ . . .

including, though you may not immediately realize it,

ᓯᔨ   ᓴᔭ   ᓱᔪ

That leaves only

ᓂᓕ   ᓇᓚ   ᓄᓗ

I struggled and suffered with those. Or at least ᓇᓗᖓᓚᐅᖅᓯᒪᔪᖓ (nalungalauqsimajunga).

OK. I’m done. Just needed to get that out of my system.

* Or at least fine-tuned by a European working on a foundation laid by the, ahem, Cree. We don’t have to go there.

The sequence you would get in ilanna + it + u if you were allowed to have three consecutive vowels in Inuktitut.