Where are the African conlangs?

[WeepingElf]

This matter came up in this thread where it does not really belong. So the question is: Where are the African conlangs? How many conlangers modelled their conlangs on African languages?

I haven't done any African-modelled conlang yet, though the idea of a non-Bantu pygmy lostlang somewhere in the Democratic Republic of the Congo occurred to me. However, the project never got past phoneme inventory considerations. I did not even decide on a phoneme inventory, except that it would have a set of voiceless obstruents like /p t s k/, a matching set of prenasalized voiced obstruents, and two or three plain clicks with nasalized counterparts. And of course a few vowels.

Then I put the thing aside, together with several other minor projects, as I took the decision to focus on a small number of more important projects.

[Vardelm]

On my current work, I've looked at a few African languages, including Fula (IIRC). This was mostly for consonant mutation and vowel alternation only.

I've had in mind for a while to do an African inspired language. I'd like it to be a pidgin or creole used by nations or an empire engaged in a lot of trade. I think Kikuya might be a prime source of inspiration. I haven't done any work on it other than jot a few notes down and look at some sound inventories.

The African languages do have some pretty cool stuff, so yeah, it's surprising it's not used more in conlanging.

[Miekko]

I have the intention of making an african-inspired lang, but as always, ... my plans for when to complete it keep being adjusted.

[----]

If I ever did an Africanlang, it'd have to be a Dahalo-esque beast with clicks and epiglottals and all that jazz, and probably with other influence from North African languages in the grammar. Probably [+/-ATR] too.

[Nortaneous]

Gehui phonology was modeled on Khoisan, though that's the only thing I've done more recently than Kett that's not modeled on Pacific (well, SE Asian to Australian, that area) languages. (Amqoli was supposed to be a Georgian clone, like Kannow was, but a week after I started I dug up a Kayardild grammar for it, so, that didn't last long.)

What general areas do y'all base your conlangs on? Seems like North America is the most popular one. There aren't any Mesoamerican conlangs either AFAIK, and that would be interesting. Only a few SE Asian, and that's all like Khmer and shit -- no one rips off Yi.

(My usual method is to combine two different languages -- V'eng is Mwesen + Tibetan, Kett is a Hittite triggerlang, and so on. If I do another conlang it'll probably be Mayan + Bantu. Or just Yi. /a æ ɔ i ɯ u v̩ v̩̠ ɿ ɿ̠/ is a hell of a vowel system.)

I tried to make a Hadza clone a while back -- remember Tnerakhii? It ended up looking more like Haida than Hadza, and I'm not really sure what happened there.

[Vardelm]

What general areas do y'all base your conlangs on? Seems like North America is the most popular one. There aren't any Mesoamerican conlangs either AFAIK, and that would be interesting. Only a few SE Asian, and that's all like Khmer and shit -- no one rips off Yi.

My Tibetan Dwarvish went from being the bastard child of Kēlen and Tibetan to having much more influence from Qiangic and Samoyedic families. There's a bunch of other inspiration from various Himalayan langs: Nepali, Pashto, Punjabi, etc. and also a little bit from Scandinavian langs.

I think I've hit a wall, though. I may end up backing up to make a proto-lang so that all the crap I'm trying to do actually makes sense. It's a daunting thought to restart again.

(My usual method is to combine two different languages

Yep, I think that's a great way to kick things off. Pick most of the features from either language, and then fill in with inspiration from a few other langs.

[Buran]

What general areas do y'all base your conlangs on? Seems like North America is the most popular one. There aren't any Mesoamerican conlangs either AFAIK, and that would be interesting. Only a few SE Asian, and that's all like Khmer and shit -- no one rips off Yi.

(My usual method is to combine two different languages...)

I use a methodology much like that. It can lead to some interesting bash-langs.

I've currently got four proto-languages in mind, and for all of them I've bashed two or more langs (usually representing families) together (at least for the aesthetics), e.g. moving from west urheimat to east, one is Semitic + Polynesian, another IE (minus Celtic and Italic) + Turkic + Uralic (mostly Finno-Ugric), another Altaic + Uralo-Siberian + Eskimo-Aleut + Sino-Tibetan (less Sino, more Tibetan), the last a two-pronged one, with a Bantu + Algic + Austroasiatic branch and a Bantu + Sino-Tibetan (Yue) + Indo-Aryan branch.

I've had a thing recently for Tibeto-Burmese and Siberian langs. I feel like I should be ripping off more languages of the Americas (Haida + Welsh, anybody?).

[Salmoneus]

[My Rawàng Ata is primarily Austronesian-inspired, although not exclusively (eg it has cases, which Austronesian languages don't tend to have).]

I suppose one issue with African languages, beyond general ignorance of African languages, is: what makes something distinctively 'african' as opposed to just 'a conlang'? One reason north american languages seem common is probably that any time you get a language with a lot of synthesis and big vowel clusters, it'll look pretty north american (or caucasian - so I guess kinaesthesis and verbyness probably help make it NA too). But what makes something look african?

I guess if you have clicks, that might suggest khoisan. And if you have prenasalised voiced stops, coarticulated bilabial-velars, and prefixial noun classes, it'll look Bantu. But beyond those specific flourishes, what is africanness in a language?

Tones, I guess?

[----]

Taahu was originally devised to be like an Iroquoian language, but I think now it's closer to Siouan. Dbe is blatantly PNG~Oceania.

[Salmoneus]

Well, that made me look at WALS. So, some things:

- An Afrolang probably shouldn't have a small vowel inventory. It'll probably have a large vowel inventory, though it may be only average
- Specifically, vowel harmony of some sort is quite likely
- Voicing is likely to be distinctive, probably even on fricatives
- There may or may not be velar nasals - if there are, they can probably be initial
- There aren't front rounded vowels
- There might be nasal vowels
- There might be implosives, but probably not ejectives
- There may be labial-velars
- There will be tone

...probably enough to be going on with

[roninbodhisattva]

I suppose one issue with African languages, beyond general ignorance of African languages, is: what makes something distinctively 'african' as opposed to just 'a conlang'?

This is a great point, and it's kind of interesting that this has come up. I was just having a discussion with some other graduates at my university who are working on African languages and this came up, but without the conlang bit. There isn't an `African' language type, really. Africa as a continent is not a linguistic area. There are definitely linguistics within Africa, but you can't talk generally of an `African' language type. I think a lot of people think of when they think of 'African language' is Niger-Congo, and that basically means Bantu cuz it's the most well studied branch.

So really, making an `African' conlang really doesn't exist as such. Maybe something relativized to a specific type of language zone/type/family that exists within Africa. This point can also probably be extended to North America to maybe a lesser degree, but that's a different discussion.

On the topic of my own conlanging, I've been kicking around a lot of ideas based on things I'm working with currently. But nothing has come to fruition because I'm busy.

[Torco]

Most likely the languages most representative of africa by population are like Arabic, French and English. Also africa is huge and what ronin said is true.

EDIT: according to the wiki, the most people in Africa speak Arabic, then Swahili, then some form of Berber, and then Hausa. Of those, only two are what one would prototypically think of as "african": swahili and hausa. Arabic and Berber have a much different sound to them.

I would also like to add that, incidentally, Berber's consonant clusters are fucking amazing.

[Risla]

What general areas do y'all base your conlangs on? Seems like North America is the most popular one. There aren't any Mesoamerican conlangs either AFAIK, and that would be interesting. Only a few SE Asian, and that's all like Khmer and shit -- no one rips off Yi.

Andean languages (ahem, Aymara) have been pretty pervasive influences on my conlangs for a long while, and there's going to be a lot more Aymara influence in Arrakum as I develop it.. Also, South Eresian is phonologically more Mesoamerican than anything else.

[Matrix]

What general areas do y'all base your conlangs on? Seems like North America is the most popular one. There aren't any Mesoamerican conlangs either AFAIK, and that would be interesting. Only a few SE Asian, and that's all like Khmer and shit -- no one rips off Yi.

I don't tend to necessarily base my languages on real-world linguistic areas. My original idea for Nahakhontl was Russian + Mayan, but knowing very little about either at the time (Nahakhontl was like my second conlang), it looks completely different. Otherwise, I usually just use whatever features I feel like.

[Soap]

In the late 90s I was working on a conlang called Echo that started out influenced by Middle Eastern languages, especially Hebrew, but "moved along" to central African. This is really only about phonology though; I've always used 100% a priori vocabulary words and the grammars I was using at that time were unlike anything on Earth.

[Tiamat]

Wycoval had a west african conlang

[Pazmivaniye]

I was working on a language a long time ago that was pretty much just Luganda with different words and a slightly different phonology. Other than that, I probably don't have much stereotypical "African-ness" in my conlangs.

[Hallow XIII]

Yeah the one thing that says "African" to me really is "Bantu", probably because the other big branch that people know (i.e. Semitic and closely related offshoots of Afro-Asiatic) is == Middle East rather than "African" as it exists in the general consciousness (which seems to mean "Sub-Saharan" really). Nilo-Saharan is too much of a mess, and too deep, to present a coherent impression and nobody does Khoi-San both because it's almost a bit too distinctive and generally insane.

That said, it's not like there's a lot of Australianesque conlangs around either -- if you ask me, really, T:conlang RESEMBLE > X(Language Family) is not too great an approach anyway. That said, a distinctly African feature (widespread among basically everything that isn't Afro-Asiatic), and one that really ought to be more used, is morphological tone. With all the love that non-concatenative morphology gets you'd think more attention would be paid to this, because there are some really batshit crazy tone things going on in the mother continent.

Addendum: tying in with this, the Bantu impression is probably generated by anything that has multiple stop series preferably including voiced, aspirated and prenasalized and a CV syllable structure; tone is peripheral (even though things like depressor consonants are like *the* hallmark of Bantu).

[roninbodhisattva]

That said, a distinctly African feature (widespread among basically everything that isn't Afro-Asiatic), and one that really ought to be more used, is morphological tone.

Agreed.

[WanderlustKoko]

Karuslnko incorporates quite a bit of Swahili within its structure but it still has quite a ways to go. Saharaina was supposed to be a heavily influenced African language but it ended up being too much like a cipher for Swahili and then I just gave up on it entirely due to the lack of resources. My resources are still limited but I do make use of my Lonely Planet Swahili (which includes basic grammar and "Phrase-builders") and a Swahili-English/English-Swahili dictionary I bought a few years ago. But I guess this is why we have the internet or Wiki-Cheatia err.. Wikipedia too. I definitely want to incorporate more African languages in a future project (Yoruba, Igbo, Kongo, Wolof, Somali, etc) though.

[Yng]

Yeah on the one hand there is no distinctively 'African' language type and on the other people tend to know very little about African languages - I get the feeling, possibly unfairly, that they are understudied or underaccessible wrt grammars and so on. I only have a couple of grammars myself and most of them are old and rubbish or else are of languages that are not, in fact, that stereotypically 'African' (i.e. they aren't Bantu).

I have long had an idea of doing an a-priori real world African conlang. I guess Tsi is inspired vaguely by Africa in that it has clicks, but not much else (the idea in fact was to try and make a very NON-African language with clicks).

[WeepingElf]

Yeah on the one hand there is no distinctively 'African' language type and on the other people tend to know very little about African languages - I get the feeling, possibly unfairly, that they are understudied or underaccessible wrt grammars and so on. I only have a couple of grammars myself and most of them are old and rubbish or else are of languages that are not, in fact, that stereotypically 'African' (i.e. they aren't Bantu).

True. Africa is a big continent with many different languages (even if one leaves the northern part of the continent that is inhabited by Afrasian-speaking Europoids out of the game), and many of them are indeed underdescribed. Many features often thought "typically African" - prenasalized stops, labial-velar coarticulation, implosives, a particular type of tone system that is completely different from East Asian tones, noun class prefixes, whatnot - only occur in some African languages and in others not, or also occur in many non-African languages. It is perhaps the Bantu languages that are often considered "most typically African", but that is just one subbranch of the Benue-Congo branch of Niger-Congo.

An entertaining question is: What was spoken in the large swath of land that is now Bantu-speaking before the spread of Bantu? There must have been a bewildering diversity of languages down there. Perhaps many languages with clicks? Languages of a type that has been completely lost due to the spread of Bantu? Plenty of stuff to be explored by lostlangers.

I have long had an idea of doing an a-priori real world African conlang. I guess Tsi is inspired vaguely by Africa in that it has clicks, but not much else (the idea in fact was to try and make a very NON-African language with clicks).

Ah, clicklangs. I have had three. The first, X-2, was an attempt at a stack-based loglang, which had one lone click phoneme. It was abandoned soon. The second was an early draft of Caledonian, an Albic language, but the clicks were revised away soon because I felt that they did't really make sense in that language. The third is the one which I mentioned in the opening post of this thread. It is neither abandoned, nor have the clicks been revised away, but the project is mothballed.

[Ketumak]

Vortex mentioned Wycoval's African conlang. It's called Tseeyo and is a Niger-Congo style language. It's still viewable here: https://sites.google.com/site/wycoval/

[Nortaneous]

relevant: http://annierialland.free.fr/Clements_Rialland.pdf
also: http://email.eva.mpg.de/~gueldema/pdf/05MacroSudan.pdf

Sudanic conlang would be interesting -- haven't seen anyone use labial flaps

[roninbodhisattva]

That Sudanic paper looks like a nice one. Some neat features as well.