Here's the frequencies for some other 6 consonant languages:
Obokuitai (b t d k s h)
d – 39%
k – 22%
s – 14%
b – 12%
t – 9.1%
h – 5.2%
Iau (b t d k f s)
d – 40%
b – 21%
s – 16%
f – 8.4%
t – 8%
k – 5.6%
Rotokas (p t k β ɾ g)
t – 34%
ɾ – 33%
β – 20%
k – 8.2%
p – 4.7%
g – 0.2%
North Mekeo (b k β m ŋ l)
ŋ – 28%
m – 22%
k – 22%
b – 18%
l – 7%
β – 1.9%
And to recap, Orokolo (p t k m l h)
/l/ – 39%
/m/ – 23%
/k/ – 15%
/h/ – 15%
/p/ – 7%
/t/ – 0.7%
Notes:
- Take these data with a grain of salt; they're from varying sample sizes and qualities, although in all cases continuous text not wordlists.
- /d/ is common in Obokuitai because it's also the only coda; it's even more common in Iau despite Iau having only coda /f/. 40% is nuts.
- In three of the languages, one phoneme acounts for around
40% of all consonants which is crazy.
- In two languages there's a phoneme with a frequency under 1%, which is very unexpected. Rotokas /g/ seems to be even rarer than Orokolo /t/.
- It seems frequency distribution is fairly random; Obokuitai drops off smoothly, Iau drops rapidly then suddenly levels off, Rotokas has two very common consonants then drops off steeply, NMek has a plateau in the middle, and Orokolo drops of smoothly save for a mini plateau halfway through.
- The nature of the consonants preferred or disprefered is completely random. Stops are favoured in Obokuitai and Iau, but dispreferred in Rotokas and Orokolo; nasals are very popular in NMek and Orokolo, the two languages they actually appear in. Coronals are the most common consonant in Rot, Oro, Ob and Iau, but /l/ is fairly uncommon in NMek.
What 7-consonant languages are there? East and West Mekeo, Namia and Sikaritai I think? (Weird that there's more 6C langs than 7C langs)
East Mekeo (p k ʔ f m ŋ l)
ŋ – 28%
k – 22%
p – 16%
ʔ – 12%
m – 9.7%
f – 6.0%
l – 5.7%
West Mekeo (p k b g m ŋ l)
ŋ – 26%
g – 21%
m – 13%
p – 12%
k – 10%
b – 8.8%
l – 8.3%
Namia (p t k m n l r)
k – 20%
l – 19%
r – 16%
m – 14%
n – 12%
p – 11%
t – 7.2%
Sikaritai (p b t d k s w)
k – 27%
b – 19%
d – 18%
t – 13%
p – 9.1%
w – 8.1%
s – 6.1%
At this point I can't help quoting a brilliant gloss in the Sikaritai discourse paper I got my sample out of:
David L Martin wrote:They just looked up. THE SCROTUM was hanging down. Hanging down very big.
On a more serious note, there's much of the same again. Namia is unusually non-steep, without any super-rare or super-common consonants. The top frequency tends to be closer to 30 than 40, and there's no very low-frequency consonants. Now it's velars that have their time in the sun, and coronals which are dispreferred. Basically I think there's no relation between how common consonants are cross-linguistically and how common they are in a given language. And I feel like maybe Mitsiefa could do with dropping the frequency of one consonant a bit so its not so shallow. I'll try and not use so much /k/ in the future.
