Philip Newton's journal

Fun when languages interfere… while I was practising Klingon flashcards, the word 'IrneH (maternal uncle—mother’s brother) came up, and my initial instinct was to read it [ɪɴnɛχ]: interpreting the sequence rn as in Inuktitut! (Where the r here stands for a uvular nasal.)

For what it’s worth, I think that in Greenlandic rn is [ɴː]: it may be the only case where the first consonant did not assimilate completely to the second one, but instead the uvular-ness of the /ʁ/ “survived” and was carried over onto the nasal.

The confusion was aided, no doubt, by my knowledge of the Inuktitut word irniq “son”, which starts very similarly to the Klingon word. (Greenlandic spelling would be erneq.)

In other Inuktitut-related news, it’s always fun to see a bit of Greenlandic and understand it based on the bit of Inuktitut I know so far: recognising the cognates and undoing the sound changes.

Latest example: the Greenland Language Secretariat Oqaasileriffik, which I understood as Uqausilirivik (uqausilirivvik, uqausiliribvik, depending on dialect): uqaq- “speak” + -siq (something like “abstract quality of; -ness”, I think) = “speech” + -liri (something like “to deal with something professionally”, I think) = “linguistics”(?) + -(b/v)vik “place where something is done”—so something along the lines of “linguistics institute”?

Hm, looking it up on Uqailaut, I see that it’s probably -usiq “custom, way, habit” in the middle, not just -siq. And uqausiq is indeed “speech” or “language”.

Anyway, the Greenlandic derives from this straightforwardly by converting the diphthong au into the long vowel aa, spelling u and i as o and e, respectively, before the uvular consonants q and r, and turning geminative (voiced) fricatives into voiceless geminates, which in the case of *vv --> ff is marked in the spelling. (I suppose because the letter f happened to there in the Latin alphabet.)

I was reading a web page somewhere (probably an article on Nunatsiaq News Online) when an advertisement in syllabics caught my eye.

I had a closer look and saw that it was a quadrilingual (English—French—Inuktitut—Inuinnaqtun) advertisement asking for comments on the Proposed Uqausivut Comprehensive Plan.

I had a bit of a look through the materials there and the strategies and rationales outlined in the plan, and parallels struck me between Inuktitut and Romansh.

The strongest was definitely the bit where they said that one challenge was the fact that different communities spoke different dialects and that they wanted to work towards a written standard to enable Inuit to communicate more easily with one another. I suppose that’s one area where Graubünden is about 25 years ahead of them: they already had people working on standardised written forms and are introducing one in more and more spheres.

But there are other parallels:

• Both languages are spoken in an area where the traditional language is under strong attack from a foreign language, due not only to amount of media available but also because it is the language of education and work (English in Canada, German in Switzerland)
• Especially if you only consider Graubünden north of the mountain range: both languages are spoken in an area where there are two predominant languages, yet three official languages; all three are equal according to the law, but two of them are more equal than the others (the third being French in Nunavut, Italian in Graubünden)
• Both languages are spoken in the high Arctic by a people traditionally relying on hunting… oh, wait. Though Graubünden was predominantly agrarian, I suppose, and the move to an industrial or service-oriented society (with the societal and linguistic changes that entails) probably comes with challenges.

One area where Graubünden has a huge advantage, though, in my opinion: when it comes to new technology, it has the common European stock of Greek and Latin to draw on. And not only that, it has models to follow: for example, French and Italian are not only both Ausbausprachen with highly-developed vocabulary, but are also linguistically fairly closely related to Romansh, making it easire to use them as models, and German, the other Ausbausprache relevant, tends to draw word stock for “advanced” vocabulary from the same sources.

So if you want to translate, say, “market capitalisation”, then you just twiddle the endings a bit and come up with chapitalisaziun dal martgà: essentially what happens if you take CAPITALISATIONEM and MERCATUM and apply the appropriate sound changes.

Inuktitut doesn’t have this readily available: not only is there no classical language that plays the role of Latin for French (or Sanskrit for Hindi, etc.), but cognate languages have similar situations, so you can’t generally just look to (say) Aleut or Iñupiaq to see how they word this or that new concept.

True, Inuktitut has a wide array of suffixes letting it express many new concepts, but for a combination of precision and concision, it’s surely hard to beat borrowings: especially since you can borrow what’s essentially a synonym and use it in only one of its senses. (The first that came to me was English sheath vs Latin vagina: which has the same array of meanings in Latin as the English word does in English AFAIK, but as a loan word in English is used for a restricted subset, leading to more precision with a good measure of consision. FWIW, German uses the same word for both.)

Anyway.

Other things also rang bells, such as Inuktitut/Romansh education in the lower grades (scheduled to be extended to K–12 in Nunavut by 2019).

In some ways, Nunavut has a leg up on Graubünden by the fact that a higher percentage speaks Inuktitut compared to Romansh, and that you have ethnic differences that make it easier to combine “being an Inuk” with “being able to speak Inuktitut” which is harder in Graubünden: I don’t think there’s any such thing as an “ethnic Romansh speaker”.

(I also wonder whether a bigger push for Inuktitut will lead to the equivalent of Rumantschuns: something like “professional Inuktitut speakers” who speak it because it gives them certain privileges or access to funding or something due to the status of “speaker of endangered language”, rather than because they feel a deep relationship to the language. Though I’m not quite sure of the exact sphere of meaning of Rumantschun so I might be missing the point.)

And finally, I was amused to read that the French-speaking radio station in Iqaluit is CFRT-FM, which apparently (cf. page 52 of the Uqausivut plan document) stands for cé frette! :)

I've finished reading through Inuktitut Essentials: A Phrasebook now. (I haven't committed it to memory, of course, but I think I've read every example sentence and every page except for the glossary at the end.)

I have mixed feelings about it. I could probably sum them up something like this: if you want to say something that has a phrase for it, you're fine; if you want to say anything not explicitly mentioned, you may be stuck.

Granted, Inuktitut grammar is rather different from English and fairly complex, with a multitude of suffixes for various things, and even the "basic" possession-number-case inflections are pretty many. (Thank goodness you don't have gender or different declensions, at least! Unless you count "nouns ending in a vowel / -k / -t / -q" as separate declensions.) Also, the high average number of morphemes per word makes it hard to take a given example and adapt it to other situations without knowing which bit means what.

Still. It seems as if some of the sentences are overly precise: they'll let you say exactly that, but are not necessarily friendly to expressing other things.

You have "My tent site has many spiders" (but no way of saying "your tent site", let alone, "Where is your tent site?"), "Where are the band from?" (but no way of saying, "From Igloolik/Iqaluit/Cambridge Bay/Ottawa/...."), "I would like butter" (but there's no plain "butter" in the glossary), ᖃᖓᑦᑕᐅᑎᒐ ᐊᑭᖃᓚᐅᖅᑐᖅ $1,200-ᓂᒃ Qangattautiga akiqalauqtuq $1,200-nik "My airplane ticket cost $1,200" (without having any idea how to pronounce "$1,200-nik"; I'm guessing something along the lines of ᐅᐊᓐ ᑕᐅᓴᓐ ᑑ ᓴᓐᓂᑦ ᑖᓚᓂᒃ uan tausan tuu hannit taalanik) and so on. There's also no mention of "nasal r": r stands for two different sounds, depending on its environment, and this text, like many others, only talks about the "default", stand-alone sound.

Also, I think there's at least one typo (ᐱᙳᐊᕕᒻᒥᙶᖅᑐᒍᑦ Pinnguavimminngaaqtugut "We are coming from the rec centre" should be, I think, ᐱᙳᐊᕐᕕᒻᒥᙶᖅᑐᒍᑦ Pinnguarvimminngaaqtugut, given ᐱᙳᐊᕐᕕᒃ pinnguarvik "recreational centre"). No other ones that were immediately obvious to me, which is good, though I think I was suspicious a couple more times. Though I suppose that given the number of pages in the book, one (or three) typos is not that bad.

Basically, I think it could have benefitted from having a much more extensive section on grammar and inflection, and a bit more breakdown of how the bits combine to reach the meaning in the smooth/idiomatic translation.

Essentially, I guess I wanted a Kauderwelsch Sprachführer: Inuktitut Wort für Wort :) (Unfortunately, their section on Canada has English, French, [Anglo-]Canadian Slang, Québécois Slang, and Sioux/Lakota, but no Inuktitut. They do have one for West Greenlandic, though. Oddly enough, it's not listed under "Denmark" [though Faroese is, for example], and "Greenland" isn't on their list of countries.)

Also, I'm beginning to understand why Mick Mallon wrote in Elementary Inuktitut Dialogues:

We get lots of requests for "conversational" courses in Inuktitut, courses that are not as demanding or rigorously logical as our normal ones, which require a lot of patient plodding and analytical thinking on the part of students. People say, "Can't you just give me a few phrases to use in common situations around town? That's all I need."

Let us try to explain why we have resisted these pleas until now. Inuktitut is not like most European languages that you can "pick up" bit by bit, such as Spanish or Italian. Those languages have an underlying structure similar to English, so that you can slot in the words you learn into the familiar patterns of your own speech, and get away with it most of the time. That won't work in Inuktitut, where the structure is completely different.

There is probably something to be said for learning grammar before (or along with) phrases, if you want to be able to vary the phrases at all.

(For what it's worth, the Elementary Inuktitut Dialogues seem to do more taking-things-apart-into-chunks and explaining. On the other hand, they also cover less ground than Inuktitut Essentials; it's a trade-off.)

I’m listening to CFFB right now (CBC North Radio One, Iqaluit), in the hope of hearing some Inuktitut.

And indeed I already heard a bit; a mixture of central and regional programming, the former in Canadian English, the latter in a mixture of Inuktitut and (slightly-accented) English.

Two things I’ve noticed are that the -q- in Iqaluit seems to be a bit fricative, something like Iχaluit (presumably because of its intervocalic position?); and the pronunciation of -ou- in house and about, which sounds distinctively different to what I’m used to but which I’m not sure how to notate. (The first half of the diphthong sounds closer to schwa, though.)

…and I just noticed some French influence: French-accented English from some speakers, and French names pronounced in French rather than with an anglicised pronunciation.

As I read about Inuktitut and look over materials, little words and suffixes stick in my memory. And then it sometimes happens I look at a word (especially a place name) and suddenly realise how it’s made up—always an interesting epiphany. (Sort of like suddenly realising that “Cambridge” means “bridge over the river Cam”, perhaps. Not sure that really does it justice, though.)

Anyway, two recent words were Natsilingmiut ᓇᑦᓯᓕᖕᒥᐅᑦ and Ulukhaktok ᐅᓗᒃᓴᖅᑑᖅ.

Natsilingmiut came first; I realised it presumably came from natsik “seal” + -lik “place with something” + -miut “inhabitants of” = people from Natsilik, from the place with seals.

Later came Ulukhaktok: I thought, hm /h/ in the west corresponds to /s/ in the east, so that’d be Uluksaktok. And then I realised that the morphemes are presumably not Uluk-hak-tok but rather Ulu-khak-tok, and that it’s probably Uluksaqtuuq / ulu-ksaq-tuuq in eastern Inuktitut (Wikipedia confirms the final long vowel and the two qs): ulu “ulu (traditional Inuit knife)” + -ksaq “potential” + -tuuq “place with lots of something” = “place with lots of ‘potential ulus’”, which matches the translations “the place where ulu parts are found” and “a large bluff where we used to collect raw material to make ulus” which the Wikipedia article gives. I suppose the literal “potential ulus” means “raw materials for ulus; parts for ulus” in this case.

And given what else I’ve read about western dialects, I can imagine that the /kh/ in Ulukhaqtuuq is pronounced as [x].

Another thing that comes to mind with -ksaq is what I read somewhere… something along the lines of how items on shelves in a shop are called something with -ksaq in it because they are potentially for sale—and items in the storeroom of the shop would have two -ksaqs in them because they are only potentially “potential-items-for-sale”!

Like many other parliaments in the Westminster system of government, the Legislative Assembly of Nunavut publishes a Hansard containing transcripts of the debates.

These Hansards are available not only in English, but also in Inuktitut.

This availability of the same material in two separate languages seems like a valuable resource for Machine Translation research and corpus research as well as for the development of language processing tools for Inuktitut, much as the Canadian Hansards—published in both French and English—are for that language pair.

And indeed, InuktitutComputing.ca makes use of those parallel English and Inuktitut Hansards to try to extract parallel sentences automatically and to use those to refine their electronic morphological analysers and dictionaries.

However, I read in a paper by the Institute for Information Technology, National Research Council Canada on this research that

The Canadian Hansard is transcribed in both languages so what was said in English is transcribed in English and then translated into French and vice versa. For the Nunavut Hansard, in contrast, a complete English version of the proceedings is prepared and then this is translated into Inuktitut, even when the original proceedings were spoken in Inuktitut.

And that surprises me a bit.

Looking through a couple of issues at random, I see “(interpretation)” in the English version and “(tusaajitigut)”[1] in the Inuktitut version every now and then; I presume those are cases where the words transcribed were actually spoken in the other language.

But still, I wonder why the Inuktitut words aren’t transcribed as well and then inserted verbatim into the Inuktitut version, rather than relying on a double translation (spoken Inuktitut to written English to written Inuktitut).

Also, a PowerPoint presentation describing the process of extracting data from this parallel corpus mentions as one of the difficulties in matching English and Inuktitut words, phrases, and morphemes up that the Inuktitut edition of the Hansards is written in many dialects, depending on the translator, and gives as example translations of “school”: ilinniarvik, ilisavik, ilinniaqvik, ilitarvik, ilinniavik. (Some of those—in particular “ilinniarvik” vs “ilinniaqvik”—may not be dialect differences but merely spelling mistakes and/or adherence vs non-adherence to a particular spelling reform/standardisation [I read that the use of -q- or -r- in the middle of words is a particularly frequent spelling mistake due to a decision at a standardisation meeting that was, in retrospect, a mistake]; the presentation also notes that no spelling checkers were used and that this is part of the difficulty.)

So this adds to the problem: while the Legislative Assembly page states that The policy of Nunavut’s Hansard editors is to provide a verbatim transcript with a minimum of editing and without any alteration of the meaning of a Member's speech, the wording (as opposed to the meaning) may well be modified if the English translation of a speech given in Inuktitut is translated into Inuktitut by someone speaking a different dialect than the original speaker.

I don’t know, it all seems a bit unsatisfactory to me, and as if Inuktitut is a bit of a second-class language for the Assembly.

I imagine that practical considerations led to those decisions, and that it may be difficult to do it “properly”.

[1] “tusaajitigut” seems to mean “through interpreters”; for the curious, it breaks down into tusaaji “interpreter (one who listens with intent to understand)” [from tusaa- “to listen to something (with intent to understand or catch the meaning of the sound)” {itself from tusa- “to hear” + -a- “(action done several times, or continuously)”} + ji “-er; one who does”] + -tigut, mark of the vialis case (“through”) and plural number.