If natlangs were conlangs...

[Salmoneus]

I don't know whether initial prenasals/voiced stops are more common in nouns in POc - although I would caution that at early stages of AN, 'verbs' and 'nouns' were often not fully distinguished anyway.

Also, apparently equivalent processes have happened in indonesia and vanuatu, only here with verbs rather than nouns - the vowel in pronominal verbal clitics was dropped, leading to initial nasal + stop clusters, which turned into voiced stops, hence mutation between verbal forms.

Prenasalisation and its link with voiced stops on the one hand and simple clusters on the other seems to be a big thing in multiple branches of AN.

[Šọ̈́gala]

Are you talking specifically about the rise of prenasalised consonants in Proto-Oceanic? You might not be, because prenasalisation is a thing in a lot of Austronesian languages even outside of POc.

I don't actually know anything about the languages in question; I was just responding to a specific statement from the Original Poster ... or, at least, I was responding to what I thought he was saying.

[Seirios]

Are there any natlangs with >8 tones? Though there are three tone clusters that can occur on roots, and I don't know why they aren't counted as separate tones in Iau... Dananshan Hmong also has eight, but I can't think of any with nine or more.

Many Chinese languages do. (If we only count original tones but not those produced by grammatical, euphonic or other means, e.g. not by tone combination, diminutival tone-shift or sandhi)
E.g Standard Cantonese, though from a western perspective it only has 6~7 tones (Depending on whether the 1st tone is 55 or 53). Some Cantonese dialects have up to 10 tones in traditional Chinese perspective (checked tones are separated).
I don't know about them really well so I can't give much information, but I do know that many Wu dialects have a crazy number of tones, up to 15 or 16, though they usually don't contrast. In this way:
Early Middle Chinese (of 601 AD) has 3 tones in unchecked syllables and 1 tone in checked. Its initial consonants are usually analyzed into four groups: unvoiced obstruents, unvoiced aspirated obstruents (excluding fricatives), voiced obstruents, (voiced) sonorants.
Later, syllables with voiced initials developed into different tones (tone split), and in many Chinese langs voiced obstruents devoiced and made such difference phonemic (so any Chinese lang can have up to 8 "regular" tones, and many do have). However, in Wu, all voiced initials remain. In those Wu dialects I mentioned above, they split original four tones according to these four groups of initials. Some go to extremity and have like 15 tones.
And like Shanghainese, in sentences and words, many Wu dialects shift the tones of original syllables greatly by large-scale tone sandhi.

They must have been created by folks who wanted to be as musical as they could.

[CatDoom]

The Wobé language of Ivory Coast has been claimed to contrast 14 different phonemic tones, though many linguists are skeptical. Still, it probably has more than eight. A littler further east, the Cori language only has 6 tones, but it still breaks the IPA because all 6 are realized as level register tones.

Chatino, an Oto-Manguean language of Oaxaca state, has 10 phonemic tones, which may be the largest number in a language of the Americas.

For real kitchen-sinky goodness, though, you can't go wrong with the nearby Jalapa Mazatec language, which has oral and nasal modal, breathy-voiced, and creaky-voiced vowels (for a total of 6 phonation types), which freely combine with three register tones, which in turn may combine through morphophonetic processes to form 6 contour tones. As if that wasn't enough, Jalapa has been analyzed as distinguishing 3 degrees of vowel length as well, and has a three-way voiced-voiceless-glottalized contrast on every one of its sonorant consonants. It also has phonetic aspirated fricatives, but these are usually analyzed as clusters.

[----]

The other Mazatecan languages are not as well-described on Wikipedia, but I think they are all like that. The Chiquihuitlán variety has the same phonation distinctions (or similar effects), but even more tones.

[ol bofosh]

For real kitchen-sinky goodness, though, you can't go wrong with the nearby Jalapa Mazatec language, which has oral and nasal modal, breathy-voiced, and creaky-voiced vowels (for a total of 6 phonation types), which freely combine with three register tones, which in turn may combine through morphophonetic processes to form 6 contour tones. As if that wasn't enough, Jalapa has been analyzed as distinguishing 3 degrees of vowel length as well, and has a three-way voiced-voiceless-glottalized contrast on every one of its sonorant consonants. It also has phonetic aspirated fricatives, but these are usually analyzed as clusters.

Hmm, that's something remarkably close to what my amphimorphs could speak!

[CatDoom]

Hah! Actually, the Mazatecan languages, and I think some of the other Oto-Manguean languages as well, have whistled registers, wherein speakers can carry on simple conversations by whistling a sequence of tones. I've heard a recording of it; it's pretty much exactly like you'd imagine, and really cool, in my opinion.

Although I doubt the anphibimorphs have lips per-se, I imagine that they could produce probably tonal contours sufficient to do something similar themselves. You might even be able to make a whole language that way, using tones instead of the kinds of segments characteristic of human languages.

[ol bofosh]

Before their encounter with the monkees (see CCC culture) amphibimorph languages were "hummed" (because of their colourful throat sack), with few consonant or vowel-type inflections. That changed post-monkee (and monkees adopted more tones and phonation). At least, that's what I'm working on right now.

[TaylorS]

English: What happened to all the crazy Germanic noun morphology? What happened to grammatical gender? Where is the V2 world order? What is with this meaningless "do"? What is with all these French loanwords? THIS IS BS!!!

I'd give that to PIE, actually. 18 ways to form present tense what the fuck

IIRC PIE as commonly reconstructed had recently undergone a loss of unstressed vowels, which screwed up a lot of the morphology, triggered a shift from agglutinative to fusional morphology, and there had not been time for analogy to work, hence the mess.

[TaylorS]

I think he rather meant that Western Europe isn't a traditional zone for language isolates, unlike say the Laotian highland or Papua New Guinea. Even some erstwhile prestigeous romance languages have disappeared or nearly disappeared in that region, but somehow this ancient tartessic or vasconic whatever managed to survive to the present?

An interesting point I will make here is that highland, herding populations resist language assimilation very well. As a boon, whenever the lowlands get depopulated the highlanders can move down and take their place! E.g., compare the Vlachs (Romanians) and Slovaks to their lowland cousins the Romano-Thracians and Great Moravians. Both of whom got assimilated and later lost land to the highlanders. Some other populations like this in Europe are the Arnauts (Albanians) and aforementioned Basques.

This is one of the many reasons I think a nomadic society is the best society.

I have recently read a very interesting book on European prehistory called Ancestral Journeys by Jean Manco, and Manco argues that Basque is not an Iberian Paleolithic survival, but it is the descendant of the languages spoken by the neolithic Cardial Ware Culture that spread farming along the Mediterranean coast of Europe.

[R.Rusanov]

I'm not saying the Basques have been highland nomads forever, since time immemorial. Populations adapt to new environments...

You can look at the Ingush, their ancestors farmed the fertile crescent, drove it to an environmental crisis, and then moved to the Caucasus. They've been there ever since, herding and shit, whereas the populations that stayed - Sumerians, Akkadians, um... Babylonians... all got conquered and assimilated and genetically diluted.

[Birdlang]

I created Ebrié, it has 13 tones. And it uses punctuation marks for tone.

[Dē Graut Bʉr]

I have to say that using punctuation marks for tone doesn't strike me as a very good idea. You might want to use diacritics or numbers instead and use punctuation marks for what they're supposed to be used for.

[----]

Burmese does fine. Punctuation, like many linguistic features, is really not all that vital.

[Nortaneous]

http://en.wikipedia.org/wiki/Fraser_alphabet#Tones

[Bristel]

Lushootseed: "I'm real tired of your kitchen sink languages."

[Nannalu]

Kashaya: "Just a few more affixes, just a few more affixes.. There done. Perfect. Is the verb template big enough yet??

[StrangerCoug]

If I didn't know Japanese was a natlang, then I'd wonder if its creator had a hard time deciding on just what writing system to use and went "Screw this, let's use them all. But we're putting a method to my madness."

[Birdlang]

Taa, "KITCHEN SINK FUN TIME KITCHEN SINK FUN TIME!!!!!!!!" Sung to the tune of Peanut Butter Jelly Time.

[Birdlang]

Amuesha: "Tildes can be fun for anything, especially palatization!"

[Birdlang]

Zhuang: Ooh, number like letters. Let's use them for tone.
Gaeml: ooh adding letters for tone that don't occur in coda position.
By the way, the same guy invented Qiang and Japhhg Rgyalrong.