Early human-based conlang

[JuRi]

This conlang is based on the anatomical capabilities of early Homo species. As such it looks very different from usual languages. Here's the phonology:

/p b ɸ t d s r l j w ʔ h/
/ɘ ɜ/

/j/ is phonetically [ð̞] word-initially, [z̞] intervocalically and [e̯] in coda, while /w/ is [β̞] word-initially and intervocalically and [o̯] in coda. /t d s/ are [t̪ d̪ s̪] word-initially, while /ʔ/ is [ʔ̞] intervocalically.

Some dialects have [ɫ̪] (more accurately [l̪͡ɤ̯]) as coda allophone of /l/, while others vocalize it to /w/. Some dialects monophthongize /ɘj ɘw ɜj ɜw/ to [e o ɛ ɔ̈~ɞ~ʌ̈].

This kind of inventory — while highly simplistic — does have natlang parallels (see NWC for vowel number and Rotokas and Pirahã for consonant number).

[lurker]

Looks neat. I've done a bit of research into early human language. There seems to be a lot of emphasis put into anatomical features needed to produce the full range of modern human speech sounds. It seems you've accounted for that in the vowel inventory.

I personally don't think a fully developed vocal tract is needed to have a fleshed out language (which again is I think what you're going for), though I personally like the theory that early language was at least partially signed. Makes sense to me given how much humans rely on vision, and how much more capable signing is of being iconic.

I'll have to keep an eye on this.

[JuRi]

The orthography is as follows:

' /ʔ/ b /b/ d /d/ e /ɘ/ é /ɜ/ f /ɸ/ h /h/ l /l/ p /p/ r /r/ s /s/ t /t/ w /w/ y /j/

Here are some sample words:

Wéw /wɜw/ [β̞ɜo̯~β̞ɔ̈~β̞ɞ~β̞ʌ̈] 'one'
Yey /jɘj/ [ð̞ɘe̯~ð̞e] 'two'
Yé /jɜ/ [ð̞ɜ] 'three'
Péyfé /pɜjɸɜ/ [pɜe̯ɸɜ~pɛɸɜ] 'water'
Téfeywew /tɜfɘjwɘw/ [tɜɸɘe̯β̞ɘo̯~tɜɸeβ̞o] 'fire'
Réyyé /rɜjjɜ/ [rɜe̯z̞ɜ~rɛz̞ɜ] 'to roast'
Eypé /ɘjpɜ/ [ɘe̯pɜ~epɜ] 'stone'
Téye /tɜjɘ/ [tɜz̞ɘ] 'clothing'
'ewyé /ʔɘwjɜ/ [ʔɘo̯z̞ɜ~ʔoz̞ɜ] 'meat'
Fépé /ɸɜpɜ/ 'root'
Lesel /lɘsɘl/ [lɘsɘl~lɘsɘɫ~lɘsɘo̯~lɘso] 'to sleep'

[TBPO]

I think that early humans' langs should have extremally simple grammar, simpler that in any natural languages. It means that 50% words are onomatopoeias and 50% are gestures; no tense, no aspect, no modality, cases are unreachable dreams, number only by lenghtening or reduplication. Lexicon should be small and extremally basic.

[DV82LECM]

This is an *actual* proto-lang. How far back are we saying?

[JuRi]

Homo antecessor level. I think they were advanced enough to have a decently large vocabulary, yet I agree on the grammatical aspects. Maybe even word order wasn't a thing.

[JuRi]

One could reach further back and come up with a consonant inventory of /p t j w ʔ/, where /j/ and /w/ are actually [e̯] and [o̯] at all positions. Vowel inventory could either be the same or be just /ə/.

[Khemehekis]

This looks like an original idea for a conlang! The early-hominin setting is not overused. And it seems you did research into the vocal capacity of early humans.

[lsd]

languages based on semantic primes are serious candidates...
reduced grammar, reduced basic lexicon,
can therefore be signed easily,
few users capable of producing incorrectness up to naturalism...
for instance, think to toki pona...

[Znex]

I'm not sure how helpful I can be, but I am reminded of JBR's Pleistocenese conlang where in some grammatical areas he actually made it more complicated, since as he argued, early human language may have been less streamlined and harder to learn than we'd expect. As an early human language, it may have been more clunky and unwieldy, not necessarily more simple.

[JuRi]

Here are a few more words:

Wéwse /wɜwsɘ/ [β̞ɜo̯sɘ~β̞ɔ̈sɘ~β̞ɞsɘ~β̞ʌ̈sɘ] 'sun'
Repé /rɘpɜ/ 'moon'
Yéwdéy /jɜwdɜj/ [ð̞ɜo̯dɜe̯~ð̞ɔ̈dɛ~ð̞ɞdɛ~ð̞ʌ̈dɛ] 'star'
Hése /hɜsɘ/ 'ground'
Bey /bɘj/ [bɘe̯~be] 'mother'
Péy /pɜj/ [pɜe̯~pɛ] 'father'
'éy /ʔɜj/ [ʔɜe̯~ʔɛ] 'fat'
Péyse /pɜjsɘ/ [pɜe̯sɘ~pɛsɘ] 'seed, grain'
Félé /ɸɜlɜ/ 'hot'
Béwéy /bɜwɜj/ [bɜβ̞ɜe̯~bɜβ̞ɛ] 'cold'

[lurker]

I’m excited to see where you go for the grammar. Some theories of early language lean on pidgins and creoles, how they usually have SVO word order, few grammatical categories and sparse inflection, etc.

[Keenir]

This looks intriging, and I agree with the others here: please continue this.

languages based on semantic primes are serious candidates...

reduced grammar, reduced basic lexicon,
can therefore be signed easily,

And you're basing that upon which sign language?

[lurker]

This looks like an original idea for a conlang! The early-hominin setting is not overused. And it seems you did research into the vocal capacity of early humans.

Indeed. I like the themes of cavemen and the Stone Age, be they more whimsical or more grounded. I’ve represented this interest in my own work with the tree dwellers.

[lurker]

My only nitpick is that you have a vowel with a diacritic in your romanization even though there are only two vowels in the language.

But yes please continue this. Ooga ooga me want reject modernity and embrace inner caveman. Me want eat mammoth steak, me want make cave painting, me want sell car insurance.

[DV82LECM]

You should call this Croodish.