Page 6 of 15

Re: Odd natlang features thread

Posted: Tue Jul 24, 2012 8:48 am
by Gulliver
Magb wrote:The lack of dialectal variation in Iceland is fascinating to me too. I think it has to be at least partly a result of the relationship Icelanders have with their own language. This is of course a generalization, but I would say that Icelanders in general are very proud of their language, and very protective of it. As an example, there used to be a low status sociolect pejoratively called flámæli, which IIRC featured among other things some vowel mergers, but the sociolect is now completely gone. Another example is Icelandic's almost total resistance to loanwords. It seems to me that a society with an above average sense of pride in its language is likely to display less internal linguistic variation -- though I'm sure you can find countless counter examples to this as well.
I read a really good paper about speech communities (or social networks or the other one) that gave a really convincing theory as to why Icelandic has always been a conservative language, and showed other similarly conservative speech communities (or social networks or the other one) with similar features in their social history. But I can't remember who by. Milroy & Milroy were cited a lot. And Labov. But as it's impossible to write about sociolinguistics without mentioning them. I'll have a look when I get home, if you like?

Re: Odd natlang features thread

Posted: Tue Jul 24, 2012 1:59 pm
by Magb
Gulliver wrote:I'll have a look when I get home, if you like?
That would be great!
merijn wrote:Now that we are discussing loans in Icelandic, I have a question about Icelandic that has been bugging me, but that doesn't warrant its own thread. Other western European languages have loaned some derivational suffixes that is used to form new learned words. For instance, Latin -tio became -tion in English, -tie in Dutch and -cion in Spanish, Latin -tas became -ty in English, -teit in Dutch and -dad in Spanish, Greek -ismos became -ism in English, -isme in Dutch and -ismo in Spanish, and Greek -izo became -ize in English, -iseren in Dutch and -izar in Spanish. Has Icelandic loaned any derivational suffix from Latin or Greek to form new words?
There's -ismi for the second one (e.g. femínismi). Can't think of anything for the other two, but my Icelandic vocabulary is no great shakes.

Re: Odd natlang features thread

Posted: Tue Jul 24, 2012 2:33 pm
by Wattmann
Magb wrote:
Gulliver wrote:I'll have a look when I get home, if you like?
That would be great!
I second this!
Magb wrote: There's -ismi for the second one (e.g. femínismi). Can't think of anything for the other two, but my Icelandic vocabulary is no great shakes.
Aren't femínismi, marxismi etc. wholesome loans? I've gotten that idea now that I went around looking for them.

Re: Odd natlang features thread

Posted: Tue Jul 24, 2012 2:37 pm
by merijn
Thanks. I found this list of suffixes loaned from Danish (at least I think that it is) which includes some Latinate suffixes.

Re: Odd natlang features thread

Posted: Tue Jul 24, 2012 3:04 pm
by Gulliver
Magb wrote:
Gulliver wrote:I'll have a look when I get home, if you like?
That would be great!
Here's the JSTOR link. You probably have access through your local academic library.

Incidentally, here is a link to an example of a PDF in a zip file. I thought those of you interested in file compression might be interested.

EDIT: The bit about Icelandic is very near the end.

Re: Odd natlang features thread

Posted: Wed Jul 25, 2012 11:57 pm
by Kereb
Wattmann wrote:[...]and "völva" [(female) seer].
:D

Re: Odd natlang features thread

Posted: Thu Jul 26, 2012 1:03 am
by Shrdlu
Piitish is great when it comes to stuff like this, so here's an other one I found: In some speakers speech the past participle, the neuter nouns definite article and the verbaliser has become the same, -e, and is only differentiated by the words themselves or the word order.

Also, where Swedish uses an additional definite article, Piitish uses what I believe would be called noun incorporation, with the root unchanged and the adjective simply tacked on before it and the whole word spoken slightly faster.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 1:59 am
by Nortaneous
The verbs for 'and' in Walman.
In Walman, a language in the Torricelli family spoken in Papua New Guinea, there are two words which have the function of conjoining noun phrases but which have the morphology of transitive verbs, exhibiting subject agreement with the first conjunct and object agreement with the second conjunct.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 3:25 am
by Xephyr
treegod wrote:
patiku wrote:If you say "beer can" with a British accent, it sounds like you're saying "bacon" with a Jamaican accent.
You mean, there's just one British Accent???

I they it, and I can sort of see what you mean, it resembles a Jamaican accent (just one?), but to say it's the same is pushing it. Imagine the Queen saying "beer can" and then trying to convince us she's saying "bacon" with a Jamaican accent (my Southern accent is actually far less posh).
Just like there's just one American Accent, apparently.
jmcd wrote:
Travis B. wrote:But by context you can generally tell what people mean by "British accent", if they use that term at all; e.g. most Americans, e.g. patiku, in saying that would really mean either one of Received Pronunciation or Estuary English, especially since these are probably the only English English varieties most Americans have any awareness of in the first place (except for maybe Cockney, but that is always thought of as some historical thing from back in the days of Dickens and like, not as something anyone actually speaks today).
So basically because they don't know much about it.
And hey, look, apparently everybody in Scotland speaks exactly the same, too!

Yeah I know I'm digging up old shit, I don't care.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 3:32 am
by Nortaneous
ugh fiiiine i'll do it

04:18 < Cev> there's a nominative and an accusative word for "and" in Hopi
04:18 < Cev> depending on whether the noun phrase is nom or acc
04:18 < Cev> the two nouns also get case marked

page 4: http://dickgrune.com/Summaries/Languages/Hopi.pdf

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 9:22 pm
by Chagen
Nortaneous wrote:The verbs for 'and' in Walman.
In Walman, a language in the Torricelli family spoken in Papua New Guinea, there are two words which have the function of conjoining noun phrases but which have the morphology of transitive verbs, exhibiting subject agreement with the first conjunct and object agreement with the second conjunct.
Oh man that sounds cool. I gotta steal that for a conlang later down the line.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 9:34 pm
by ----
Something kind of similar is the word for 'because' in Navajo, which can be inflected for person; when it appears in the form 'because X happened' it's always X áhóót’įįd biniinaa, but you could say something like Shiniinaa sits’il 'It broke into pieces because of me' where the only indication of the first person being involved is the inflection on 'because'.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 10:07 pm
by Chagen
Navajo's absurd compactness always amazes me.

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 11:08 pm
by Nortaneous
this about the language that has chidí naaʼnaʼí beeʼeldǫǫhtsoh bikááʼ dah naaznilígíí for "tank"

Re: Odd natlang features thread

Posted: Mon Jul 30, 2012 11:31 pm
by Chagen
Okay, absurd compactness when it comes to native words and not loans.

What does that word for tank literally translate to?

Re: Odd natlang features thread

Posted: Tue Jul 31, 2012 12:34 am
by clawgrip
Wiktionary wrote:Composed of chidí naaʼnaʼí (“caterpillar tractor”), from chidí (“car”) + naaʼnaʼ (“it crawls about”) + -í (“nominalizer”); and beeʼeldǫǫhtsoh (“cannon”), from the verb beeʼeldǫǫh (“explosion is made with it”) + the adjectival enclitic tsoh (“big”); and the particle bikááʼ (“on it”) plus the verb dah naaznil (“they sit up”) plus -ígíí (“nominalizer”)
So "a big one that explosions are made with that is sitting up on a car that crawls about" I guess

Re: Odd natlang features thread

Posted: Tue Jul 31, 2012 6:20 am
by ol bofosh
"Hey, we've got to leg it because I just saw a chidí naaʼnaʼí beeʼeldǫǫ..." BANG!!! :o

Re: Odd natlang features thread

Posted: Tue Jul 31, 2012 6:52 am
by Yng
Treegod wrote:"Hey, we've got to leg it because I just saw a chidí naaʼnaʼí beeʼeldǫǫ..." BANG!!! :o
It's a little known fact that tanks were first developed during the Indian Wars to take advantage of this very weakness

Re: Odd natlang features thread

Posted: Wed Aug 01, 2012 5:28 am
by jmcd
Xephyr wrote:
jmcd wrote:
Travis B. wrote:But by context you can generally tell what people mean by "British accent", if they use that term at all; e.g. most Americans, e.g. patiku, in saying that would really mean either one of Received Pronunciation or Estuary English, especially since these are probably the only English English varieties most Americans have any awareness of in the first place (except for maybe Cockney, but that is always thought of as some historical thing from back in the days of Dickens and like, not as something anyone actually speaks today).
So basically because they don't know much about it.
And hey, look, apparently everybody in Scotland speaks exactly the same, too!

Yeah I know I'm digging up old shit, I don't care.
Saying that there is such a thing as "English with a Scottish accent" is different from saying everybody in Scotland speaks the same (that is what you're getting at, is it? Or is it something else?). If I'd said everybody in Scotland says x word [Eks wV4d] I would be wrong. But I didn't say that. In fact, the very point of that post was to show how people in Scotland do speak differently from one another. On the other hand, my choice of words in that post you linked to was poor; "more common" would have been preferable to "standard".

However, there are generally some commonalities between different Scottish accents, even if these commonalities are not present in all accents. Phonological features common to most Scottish Anglic varieties and the lack of which makes an accent less markedly Scottish include:
Diphthongs being limited to /ai oi au/ (the corollary of this being that /e i o u~}~y/ are usually monophthongs)
Long vowels occurring before voiced fricatives and word boundaries
A lack of a distinction between /U/ and /u/, dropping the /U/
A lack of labialistion of /r/
Pronunciation of coda /r/

Admittedly, the following features, commonly ascribed to Scottish English, are actually specifically found in the Lowlands:
consistent dark l
@i~ai variation
They are also found outside the Lowlands, particularly in Canada.

By contrast, there are no phonological differences which distinguish British English from other non-British varieties of English without specifying further than that, even if there are vocabulary and orthography differences. When people do compare British and American English pronunciation, it's generally specifically RP and GA they're comparing. (See page 129 here)
Also, in models of differentiation within English accents, even Northern England and Sourthern England are often divided. (http://books.google.com/books?id=wawGFW ... sh&f=false).

Re: Odd natlang features thread

Posted: Wed Aug 01, 2012 8:35 pm
by finlay
"Even"? That's like the basic subnational divide in England.

As for this one:
Long vowels occurring before voiced fricatives and word boundaries
it's more specific than that, because long vowels occur in these positions in the majority of accents. In Scottish English, they occur in these positions, plus before morpheme boundaries (hence brood/brewed is a minimal pair, and I know I have a few compound words where I have a long vowel before the boundary although I can't think of them off the top of my head), and not before voiced plosives (in most other accents they occur before all voiced obstruents, so brood/brewed both have a long vowel). They also occur before R, in the standard formulation of that rule.

(Additionally, in other accents the rule is usually formulated as a "clipping" of phonemically long vowels to phonetically short vowels before voiceless obstruents, so essentially backwards compared to the way you've formulated it. There might be evidence that this happens in Scottish English too, just because you'd never have, for instance, a long /ɪ/. It doesn't happen with all the vowels, essentially.)

Re: Odd natlang features thread

Posted: Wed Aug 01, 2012 9:56 pm
by Travis B.
finlay wrote:As for this one:
Long vowels occurring before voiced fricatives and word boundaries
it's more specific than that, because long vowels occur in these positions in the majority of accents. In Scottish English, they occur in these positions, plus before morpheme boundaries (hence brood/brewed is a minimal pair, and I know I have a few compound words where I have a long vowel before the boundary although I can't think of them off the top of my head), and not before voiced plosives (in most other accents they occur before all voiced obstruents, so brood/brewed both have a long vowel). They also occur before R, in the standard formulation of that rule.

(Additionally, in other accents the rule is usually formulated as a "clipping" of phonemically long vowels to phonetically short vowels before voiceless obstruents, so essentially backwards compared to the way you've formulated it. There might be evidence that this happens in Scottish English too, just because you'd never have, for instance, a long /ɪ/. It doesn't happen with all the vowels, essentially.)
The rule I am used to is basically that all monophthongs and primary diphthongs are short when there once was* a fortis obstruent phoneme somewhere between it and where there once was* the next lenis obstruent phoneme, the next vowel, or end of the utterance, whichever is closer, including across word boundaries, and all other monophthongs and primary diphthongs are long, including historical short vowels.

Secondary diphthongs**, triphthongs**, and overlong vowels*** behave like monophthongs and primary diphthongs with adjacent glide(s) that do not participate in underlying vowel length.

* This gets tricky because consonants that are never pronounced can still condition vowel length; for instance, the stressed vowels in it'd and gonna are short due to an unpronounced /t/ in each, while couldn't reduced to a single syllable ending in /t/ has a long vowel due to the elided or assimilated (to /n/) /d/ still blocking the vowel from "seeing" the /t/; this could be interpreted as actually being phonemic vowel length, but that analysis has major problems of its own.
** These are from a monophthong or primary diphthong getting extra glides attached to it, e.g. from following /l/, preceding unstressed /i/ and /u/, and vowels after elided intervocalic consonants (which in the process lose their own assigned length).
*** These are from elision of a lenis intervocalic consonant combined with the preceding vowel stealing extra length from a schwa or identical vowel in the following syllable.

(Yes, this may seem special-snowflake-y, but it seems pretty similar to what I have heard other Americans from the northern (eastern) US having...)

Re: Odd natlang features thread

Posted: Wed Aug 01, 2012 11:04 pm
by Chagen
Nortaneous wrote:Saanich:
2.2.1. ∥-í∥ ‘persistent’.
1. This is an aspectual morpheme opposed to unmarked ‘non-persistent’. It indicates that the activity expressed in the stem to which it is attached continues past inception as a state. The absence or presence of this suffix allows differences in meaning comparable to the differences in such English pairs as ‘figure out/know’, ‘look at/watch’, and ‘take/hold’.
This affix is morphophonemically unusual in that it has qualities of the radical morphological processes (see §2.3). It is classified here as a suffix since it always follows and never directly affects the root phonologically. But it does directly affect other suffixes. ∥-í∥ ‘persistent’ is what might be called a "parasitic" morpheme. Its placement requires the presence of another suffix having an underlying /ə/ that acts as "host". This /í/ assumes the position of the rightmost /ə/ of a host suffix that is not preceded by a suffix with a non-schwa. This morpheme has never been recorded occurring without another suffix having an underlying ∥ə∥.
Every time I try to think of a cool idea for my conlangs, a natlang comes along and goes "Hah hah, how WEAK".

Dammit. ANADEW is so so true.

Re: Odd natlang features thread

Posted: Thu Aug 02, 2012 4:17 am
by jmcd
finlay wrote:"Even"? That's like the basic subnational divide in England.
Indeed. I said "even" because they didn't just separate the four constituent countries of the UK but also split England in two parts.

Re: Odd natlang features thread

Posted: Thu Aug 02, 2012 1:39 pm
by Whimemsz
How about some Amazonian languages with unusual phonologies! (My info comes from Dixon & Aikhenvald's The Amazonian Languages rather than from Wikipedia, though in some cases the inventories Wikipedia gives may well be more accurate and based on better/more complete data...)

Maxakalí's consonants are: /p t tʃ k ʔ/, /b d dʒ ɡ/, /h/.

Karajá has already been mentioned.

Xavánte lacks velars: /p t tʃ ʔ/, /b d dʒ/, /w ɾ h/.

Some dialects of Madí (Jamamadi and Banawá) have a pretty unbalanced consonant inventory, especially among the stops: /t k/, /b d ɟ/, /ɸ s h/, /r w/.

Jebero has only two ejectives: /kʼ/ and /ɾʼ/ -- the full consonant inventory is: /p t c k ʔ/, /kʼ/, /s ʃ/, /m n ɲ ŋ/, /ɾ ɹ l ʎ/, /ɾʼ/, /w j/.

Another language with an unbalanced consonant inventory, especially in the area of stops, is Trumai: the stops are /p t̪ t k ʔ/, /t̪ʼ tʼ kʼ/, /d/.


Beyond phoneme inventories, there's interesting phonological processes too. In a number of Amazonian languages, "nasal" is a property of syllables rather than individual segments, and there are various types of nasal harmony and nasal spreading, e.g. Tuyuca: bia-ja (close-IMPER, "close (it)") = [biaja], but põõ-ja (open-IMPER, "open (it)") = [põõɲã].

Kaingáng is especially cool in this respect. Basically, "nasal" stops have different realizations depending on whether they border a nasal vowel, oral vowel, or word boundary. If an oral vowel precedes the stop, the stop's onset is pronounced as an oral stop, which then transitions to a nasal stop; if an oral vowel follows, the nasal stop has an oral stop release. A preceding or following nasal vowel means the stop's onset or release, respectively, is fully nasal. Word boundaries work like nasal vowels in this respect. Some examples:
kanɛ̃ = [kadnɛ̃] "eye"
ɛ̃mɛ = [ʔɛ̃mbɛ] "blue sky"
pãnĩ = [pãnĩ] "back"
nɛn = [ndɛdn] "thing"
no = [ndo] "arrow"
nĩm = [nĩm] "to give a long object"
ti no = [tidndo] "his arrow"
no nĩm = [ndodnĩm] "to give arrows"
kanɛɾ = [kadndɛɾɛ] "smooth"

Re: Odd natlang features thread

Posted: Thu Aug 02, 2012 8:31 pm
by Vuvuzela
Ju|'hoan (a grammar of which I received recently) has adjectives as a closed class. I'm not sure how unusual it is, but I find it interesting.
There are exactly seventeen words in the language which function as adjectives, as descriptive functions are usually performed by verbs. These are the words for "female"; "male"; "other" as in "remaining"; "other" as in "strange"; "true"; "old"; "new"; "a certain"; "each"; "all"; "some"; and the numbers one through four.