Sounds and orthography

 

The Greenlandic morphemes are written with digits for syntactic class, and with some other signs which indicate how the morphemes change when they are concatenated to form words. The rules for concatenation may be formulated in a simpler way if we start by realising what letter sequences a Greenlandic word in general may (and may not) contain.

 

The Greenlandic orthography reflects the language spoken in Central West Greenland, from Paamiut (Frederikshåb) in the south to Sisimiut (Holsteinsborg) in the north. Samuel Kleinschmidt created a consistent orthography for this language and described it in his Greenlandic grammar (1851) and in his Greenlandic and Danish dictionary (1871, K here). The new Greenlandic orthography introduced in 1973 is a simplification of Kleinschmidt's which reflects the modern Greenlandic pronounciation better.

 

Greenlandic has only three vowels: a, i and u. i,u are written with the letters e,o when a q or r follows in the same word. The vowels also occur as long; this is indicated by writing the letter double: "aa", "ii" ("ee"), "uu" ("oo"). A diphtong "ai" (i.e. long a with a final component of i) occurs as the last sound in some inflected words.

 

Two vowels may be immediate neighbours in a Greenlandic word; but not if they are forms of the same fundamental vowel (a/i/u) or if the first is a form of a . Two similar vowels fuse into one long, and "a" + "i,u" -> "aa" (but "a" + word final "i" -> "ai"). If the first vowel is long, a separating "v" or "j" is inserted between them in most cases; "v" is the most common, but "j" is inserted before u and after "aa" < "a" + "i", and also between the prefix aa9 and the demonstrative stem av6: "aajanna". Between i,u and another vowel a weak j,v is spoken but never written.

 

There are also a few cases of "i" + "u" -> "ii". In East Greenlandic, southern Greenlandic and Upernavik in the north it is a general rule that u -> i when the vowel of the preceding syllable is not u (nor aa < a+u) and there is no labial consonant (p,m,v) between them. A consonant with labial component, i.e. from "p" or "m" + another consonant (see below), will also block this vowel transition. In southernmost West Greenland (Kap Farvel) there can be no such transition in two consecutive syllables. ittoorpoq (old orthography igtôrpoq) -> itseerpoq and ittortoq -> itseertoq, not itseerteq. The rules have been formulated by Jørgen Rischel (1975, 2009), see [Grammar]Greenlandic Grammars.

 

Greenlandic had originally two vowels i, let us call them i.1 and i.2 . i.1 always occurs as i, but i.2 (written I or E in the morphs) may disappear between consonants or occur as "a" before a vowel and in a few other cases. A letter i(,e) when neighbour to a vowel letter is never from i.2 . Several Eskimo dialects in Alaska have neutral "e" (so-called 'shwa') where Greenlandic has i,e < i.2 .

 

Greenlandic has 14 consonants:

 

            labial    alveolar         velar

                                   front   back

 

stop        p         t            k       q

nasal       m         n            ng      r(ng) [nr]

lateral               l

fricative   v         s j          g       r

 

The stop consonants are pronounced with a complete blocking of the air stream from the lungs. The nasals have air passage through the nose, and the lateral has opening by the sides of the tongue. Nasals are always voiced, stops always unvoiced. Fricatives are spoken with a restriction in the mouth; s is always unvoiced. j (as in qajaq) would be written y in English (kayak).

 

A glottal unvoiced fricative "h" may occur initially in a few interjections. Its place in the table would be to the right of r in a new column.

 

Most consonants also occur long. Length is in general indicated by writing the letter double:

 

             labial    alveolar         velar

                                    front   back

 

stop         pp rp     tt rt ts     kk      qq

nasal        mm rm     nn rn        nng     rng [nnr]

lateral                ll rl

fricative    ff rf     ss rs        gg      rr

 

But long ng is written "nng", and long v is written with the letter f (reflecting the pronounciation, see below). The labial and alveolar consonants also occur long with an initial component r; this initial component is heard only as a retracted pronounciation (so-called 'uvularisation') of the preceding vowel. (Remember that the vowels i,u are written with the letters e,o when a q or r follows in the same word.) There is no long j, but instead a long t with affrication (written ts). t is pronounced with affrication before i, and tt -> ts here always. The lateral and the fricatives are unvoiced and are spoken with a strong hissing sound (like s) when long.

 

[nr] and [nnr] are just my proposed signs for the back velar nasal. Written "rng" usually denotes (long) nnr, which is the more frequent sound. (Short) nr is written "q-" or "r" or "rng": "sooq-una = soor(ng)una" 'why (ask about) that = of course'. Some speakers use a nasalized (short) r instead of short nr.

 

(Genuine) Greenlandic words end in a vowel or a short stop consonant, and they begin with one of the same sounds or with a short m/n/s. Two consonants cannot be neighbours in a genuine Greenlandic word; it is avoided, or they fuse into one long. In general the result will be a long version of the last consonant: k + rp = k + (p)p -> "pp". But if only one of the consonants is p/k/q, you get a long version of the other: k + l = l + k -> "ll". The result of the fusion will have an initial component r if the first consonant has it or if one of the consonants is q: q + l = l + q -> "rl". But m + q and n + q often -> "rng" [my nnr] instead of "rm" resp. "rn", probably through a stage m,n + [nr].

 

The language avoids (1) long consonant + consonant, and also (2) k/q + velar consonant (in any order). Thus we get (1) "ernerup" ernEq4.#°p0 (e,g), not "ernup" (rn+q) or "errup" (rn+r) with drop of E = i.2 before the vowel of the inflexion. And (2) "karra" kangEq4.a0 (e,3ie,n), probably through a stage kangEra; ng + q would give a long ng with initial component r, but there is no such sound (note that "rng" denotes [nnr] or [nr]). A stem final k/q is simply deleted (2) before any suffix or inflexion whith initial velar consonant (but q + initial g -> "r", though). It is therefore unnecessary to write "-" between "." and q or k. And neither between "." and ng if you note that -k (or possibly -t) + deverbal .ngavoq in some cases gives -nngavoq: tunngavoq, nipinngavoq.

 

Some speakers of Central West Greenlandic still distinguish two consonants s. We may call them s.1 and s.2, as they are written "s" resp. "ss" in the old orthography (see below). s.2 is a retroflex s.1 when they are pronounced differently. Other Eskimo dialects have a voiced fricative (j or retroflex z) where Greenlandic has s.2 . It corresponds to apical r in some old loan words, e.g. "musaq" 'carrot' (compare old Norwegian mura 'Potentilla anserina'). And we have s.2 in some frequent morphemes where the voiced fricative corresponding to t (i.e. ð = the first sound in English 'the') would be expected: "igasoq" iga2.Toq40 (e,n) 'a cook'. "s" as first sound in a Greenlandic word is s.1 except in the onomatopoeic verbal stem seeq2 'hiss'. "ts" has s.1 always.

 

A t in contact with i.1 often becomes s.1 . But tt after i.1 may -> "ts" instead, and tt before any i -> "ts" as said above. In North West Greenland (Aasiaat to Uummannaq) rt doesn't -> rs(.1) however, and before a,u they have tt instead of "ts". Another characteristic of North West Greenlandic is that "g" and "ng" both are pronounced as a nasalized (voiced fricative) g.

 

The two s-consonants behave differently under gemination, see [Grammar]Concatenation of morphemes (4). "s" = s.1 (and "j") in general -> ts when geminated: "natsat" nasaq4.#°t0 'caps'. "s" = s.2 is not usually changed by gemination: "igasut" iga2.Toq4.#°t0 (f,) 'cooks'. But in a few cases "s" = s.2 -> "ss" (long s.2): "qissivoq" 'finds a piece of driftwood' ("qisuk" '[drift]wood'). The stems in files i really consist of a constant part and a variable part: na saq4, qi zuk4.

 

When you give a word through the push-button [+-], Ordbogeeraq = (Os)P is shown in Samuel Kleinschmidt's orthography, see [Guide]The dictionaries. As said above, it uses the sign /ss/ for s.2 . Length may be indicated with an accent over a vowel letter: /â/ = "aa", /ák/ = "akk", /àk/ (or really with a tilde: /ãk/) = "aakk". The old orthography has not fully effected the fusion of adjacent vowels or consonants, so you will meet combinations like /ai (even inside a word), ae, ao, au/ = "aa", /gk, vk/ = "kk", /ngm, vm/ = "mm". q,Q are written as /k with short vertical, K'/. Long l is written with the sign dl: /tdl, gdl, vdl/ = "ll", /rdl/ = "rl". Word final i,u (but not "i" in final "ai") is written with the letters e,o as before the letters q and r in the same word. In some cases a "v" is written between u and a or i/e.

 

In Ordbogeeraq /l'/ denotes a retroflex l (compare s.2) from Norwegian apical r in loan words: /pal'áma/ = "pal'amma" 'lighter' (< pram), see "usingiaassuaq".