WE ARE MOVING - see Ephemera
http://www.incatena.org/
They actually seem to be. I searched for "Taoripi" as it's listed in your list and found the same consonant inventory that has [m ~ v] as /m/ and lacks /n/ altogether. Both the UPSID entry and the SIL phonology sheet use the same original source so there seems to have been some error at some point.Nortaneous wrote:Are Taoripi and Toaripi different languages?
Presumably, this must be taken with a great deal of salt. Hurrian samples are a) relatively few and b) only written, and written in a syllabary not designed for the language. There may well be distinctions not recorded in the script, or recorded in an unclear way (eg by geminated letters). Indeed, we only guess there was /f/ because different scribes couldn't agree whether to spell some words with <p> or with <w>.cromulandt wrote:Hurrian: /p t k ts) f s x m n l j w/.
There are some scholars who assume that this is just an artifact of a massively underspecifying writing system, and the language was really more like a Northeast Caucasian language.cromulandt wrote:Hurrian: /p t k ts) f s x m n l j w/.
Dialects can also lack /d/ and /ŋ/ though, getting the count as low as 11.Nortaneous wrote:By 'small' I mean 12 or fewer consonants, which isn't completely arbitrary since the smallest consonant inventory in Europe (Finnish) has 13, if you ignore the glottal stop and all the loan phonemes.
Raw counts are not gonna be particularly useful though, since the data is not independent. E.g. everything Polynesian having /p/ but no /b/, or everything Iroquioan not having /p/ don't quite count as 10 independent data points each. Trimming 'em down to a single average, though, also doesn't seem fair…Yaali Annar wrote:In case anyone wondered about the frequency of each phonemes listed
As I said, some scholars assume such a massive underspecification; others disagree, pointing out how unlikely it is that a language with 40+ consonant phonemes appears to have only 12 just due to limitations of the script.cromulandt wrote:You can't underspecify much more massively than reducing a NEC inventory to Polynesian size. Consider that one recanted.
I disagree about the data independence - obviously they aren't independent, but I think it doesn't matter too much, for two reasons. First is that data independence is a larger concern with smaller samples and a smaller concern with larger samples, whereas here we are attempting a sample size of 100% to the extent data is available. And second, given the first, it is also important that the fact that having a minimalist phonology and thus appearing on the list is also not independent. You would generally expect the relatedness of the data to be roughly comparable to the relatedness of languages' having a minimalist phonology. So your objection is substantially mooted (but not completely, I agree).Tropylium wrote:Raw counts are not gonna be particularly useful though, since the data is not independent. E.g. everything Polynesian having /p/ but no /b/, or everything Iroquioan not having /p/ don't quite count as 10 independent data points each. Trimming 'em down to a single average, though, also doesn't seem fair…Yaali Annar wrote:In case anyone wondered about the frequency of each phonemes listed
/g/ as the only voiced stop? How does that work?Whimemsz wrote:Ikpeng which has 12: /p t ʧ k g m n ŋ l r j w/
Possibly...I used to speak a bit of Inuinnaqtun, and it has a cluster /bl/ that is cognate to some Inuktitut /ll/. You could probably analyze as an allophone of /p/ before /l/, except that in Inuit languages clusters must be of one sort of consonant (stop plus stop, nasal plus nasal; /pl/ doesn't work). I don't know if Inuinnaqtun distinguishes /bl/ and /vl/, though.Theta wrote:Maybe a situation like Inuktitut, which only has /g ɢ/ for voiced stops? *b and *d apparently lenited since it has /v/ and /l/. Speaking of Inuktitut, a couple dialects of it are sooo close to fitting the phoneme number requirement. Inuinnaqtun apparently has /p t k q h v l j g ɢ m n ŋ/, which is 13.
Just looked it up. Wow. Doesn't look at all Asian (or even human)... Chomsky would have a field dayTheta wrote:Oh, found another. Central Miyako has /m n p t k ɾ f s ʋ/, which is 9 consonants. Also notable for allowing fricatives, nasals, and /ʋ/ to be syllable nuclei, as well as allowing initial geminates, and overlong consonants can also appear.
Absolutely, you just need to think how to best draw the borders of the contrastive slots for consonant phonemes between the differnt languages. After that you just calculate Pr(Y|X) or in other words the frequency of the languages that have X also having Y for what ever X and Y you want to look at.KathAveara wrote:Is it possible to do an analysis of these along the lines of 'If a language has X, it also/normally/sometimes has Y'?
Code: Select all
::POLYNESIAN::
Hawai'ian p m n k ʔ v h l 8
Tahitian p m t n ʔ f v h r 9
Māori p m t n k ŋ f w h r 10
Niuean p m[t] n k ŋ f v [s] h l 10
Samoan p m k ŋ ʔ f v s h l 10
Rarotongan p m t n k ŋ ʔ f v s h r 12
Tongan p m t n k ŋ ʔ f v s h l 12
::OTHER OCEANIC::
Roro p b m t n k ʔ h r 9
Mekeo p m t n k ŋ ʔ f s l 10
Lote p m t n k ŋ s x h l r 11
MussauEmira p m t n k ɣ ŋ β s l r 11
Wuvulu p b m t n ʔ f w h l r j 12
::IROQUIOAN::
Onondaga [t d]n ʤ [k g] ʔ w s h j 9
Cayuga [t d]n ʦ k kʷ ʔ w s h r j 11
Cherokee m t n ʦ k ʔ ɰ s h l j 11
Mohawk [t d]n ʤ [k g] kʷ ʔ w s h[l r] j 11
Oneida t n ʦ k kʷ ʔ w s h l j 11
Seneca [t d]n ʣ ʤ [k g] ʔ w s ʃ h j 11
Tuscarora t n ʧ k ʔ w θ s h r j 11
::CADDOAN::
Pawnee p t ʦ k ʔ w s h r 9
Wichita t [n]ʦ k kʷ ʔ w s h [ɾ] j 10
::ALGONQUIAN::
Cheyenne p m[t] n[ʦ] k ʔ v s[ʃ x]h 10
Menominee p m t n ʧ k ʔ w [s ʃ] h j 11
Arapaho [p b] t n ʧ k ʔ w θ s x h j 12
::NASAL-ALLOPHONY AMAZONIAN::
Pirahã p[b m]t [n] k[g] ʔ s h 8
Cubeo p[b m]t[d n] ʧ k w x r j 10
Xavante p[b m]t[d n|ʦ~ʧ|ʣ~ʤ ɲ] ʔ w [s] h r [j] 10
Barasana p[b m]t[d n] [ʧ] [ʤ] k[g ŋ] w [s] h r [j] 10-11
Tuyuca p[b m]t[d n] [ʤ ɲ]k[g ŋ] w s h r [j] 11
Karajá [b m]ɗ[d n] ʧ ʤ k w θ ʃ h l ɾ 12
::OTHER AMAZONIAN::
Puinave p m t n k (w) s h (j) 7-9
Barí b m t d k s h ɾ r j 10
Iquito p m t n k w s h r j 10
Tiriyó p m t n k β [s~ʃ] h [ɾ~ɽ] j 10
Arabela p m t n k w s ʃ h r j 11
Jamamadí b m t n [ɟ] k (ʔ)ɸ w s h̃[l~r] [j] 11-12
Huaorani p b m t[d]n [ɟ]ɲ k g ŋ w [ɾ] [j] 12
Ikpeng p m t n ʧ k g ŋ w l r j 12
::PAPUAN MISC::
Iau [b m]t[d n] k f s 6
Rotokas p b t d k g 6
I'saka [p|b m]t[d n] k [ɸ]w s j 8
Nasioi p b m t d n k ʔ 8
Taoripi p m t k f s h l 8
Abau p m[t d]n k w s h[l r] j 9
AitaRotokas p b m t d n k g ŋ 9
Gadsup p m t d n k ʔ β j 9
Namia p m t n k w l ɾ j 9
Ekari p b m t d n k w gʟ j 10
Sentani p m t n k f w h l j 10
Koiari b m t d n k g f ð h r 11
Kwomtari p b m t[ɖ]n k g [ɸ β] s [ɭ]r 11
Tigak p b m t n k g ŋ s l r 11
Awa p b m t n k g ʔ w s r j 12
Ese p m t n ʧ ʤ k ʔ β s h r 12
Fas p m t n k ʔ f w s r j 12
Tifal [p b]m t[d]n [k ɣ]ŋ f w s l[ɾ] j 12
Yimas p m t n [c] ɲ k ŋ w [s] ʎ r j 12
::OTHER::
Miyako [p b]m[t d|n] [k g|ŋ] f v [s ɕ|x] ɾ 9
Keuw p b t d k g w s l j 10
Maxakali p[b m]t[d n] k[g ŋ] ʔ ʃ h j 10
Palauan [p b]m t d[n] [k g]ŋ ʔ [w|θ ð]s l r [j] 10
Ainu p m t ʦ k w s h r j 11
Irantxe p m t n k ʔ w [s ʃ] h[l~r] j 11
Maranao p m t k ŋ ʔ w l r j 11
Warao p m t n k kʷ w s h r j 11
Bandjalang p m t c ɲ k ŋ w l r j 12
Comanche p m t ʦ k kʷ ʔ w s h j 12
Ket [p b]m t d n [k g]ŋ [q ɢ] s[ç] h ɮ [ʝ] 12
Meänkieli p m t n k ŋ ʋ s h l r j 12The file doesn't open for me.Nortaneous wrote:^ You've never heard of Lau because it's not actually a langu--dammit yes it is. But it has over twice as many consonants as Iau -- /b t d k g kʷ gʷ f s m mʷ n ŋ l r/. Sans-serif fonts make certaln ianguages dlfficuit.
Iau source: http://sealang.net/archives/nusa/pdf/nu ... p29-42.pdf