Evolution of language ex nihilo

[Otto Kretschmer]

Ok

Assuming that a group of people conducts an experiment and raises their children on an isolated island with a grammatically complete language but very limited vocabulary (approx 100 words) how does the language develop?

The children of the first inhabitants will know the concept of language and will be able to communicate at a very basic level (pronouns, greetings, body parts, basic verbs like eat, sleep, walk) but will lack words for vast majority of things and phenomena and will have very limited vocabulary to create it from

How do they create the necessary vocabulary? If you see an aircraft for the first time you may call it a metal bird or a flying car but you cannot if you do not know the word for flying

[Rounin Ryuuji]

Well, for "fly", you can construct the word "air-walk" or "air-go"; if you don't have "air", then "sky".

[Linguoboy]

Why do you even need a word for "fly"? Why not just say "go in the air"?

[zompist]

How do they create the necessary vocabulary? If you see an aircraft for the first time you may call it a metal bird or a flying car but you cannot if you do not know the word for flying

This isn't exactly ex nihilo. But anyway, the community will do as all linguistic communities do: 1) create new words out of their existing stock of roots, and 2) create new words from scratch.

It's a bit challenging, sure, to have such a small stock of roots. Several conlangs, notably Toki Pona, have set themselves that challenge; you can get surprisingly far with a very small starting lexicon. Worth a try if you've never done it.

People sometimes forget about method 2. Think about onomatopoeia to start with-- you can directly represent sounds (knock, tap, drum, splash, smash), some bodily actions, animals, etc. Completely invented words are not unknown, and might be favored in this setup.

I'd also note that, besides conlangs, languages do get created ex nihilo in the real world-- sign languages. Arguably they're easier to create because so many words can be iconic.

[MacAnDàil]

This resembles an idea that Derek Bickerton had and he mentioned in his autobiography / summary of his work as creolist, Bastard Tongues.

My own hypothesis is that some elements are likely to occur e.g. starting out with an analytic language due to the nature of the experiment whereas others e.g. describing a plane will vary.

[Raphael]

Hm, what about the idea that the ancestors of pretty much all word components, if you go far enough back into the distant past, originally started out as onomatopoeia? I mean, it's basically either that, or derivation from earlier roots, or borrowing, and in the latter two cases, that leads to the question of where these earlier roots or words in other languages originally came from.

Wait - while I wrote the previous paragraph, I could think of one other possible original source: baby talk, such as in the case of "mama" and "papa". So make that two basic original sources for word roots.

[xxx]

it depends on the nomothete,
if it is an a priori language, in the philosophical sense of the term,
a limited number of semantic primes could well be enough to name/define everything...
(that's how I conlang...)

[alice]

Hm, what about the idea that the ancestors of pretty much all word components, if you go far enough back into the distant past, originally started out as onomatopoeia? I mean, it's basically either that, or derivation from earlier roots, or borrowing, and in the latter two cases, that leads to the question of where these earlier roots or words in other languages originally came from.

Wait - while I wrote the previous paragraph, I could think of one other possible original source: baby talk, such as in the case of "mama" and "papa". So make that two basic original sources for word roots.

I thought of asking how one would onomatopoeia-ize a plant, then read this post talking about roots, and decided not to bother in case it confused people.

[Raphael]

it depends on the nomothete,
if it is an a priori language, in the philosophical sense of the term,
a limited number of semantic primes could well be enough to name/define everything...
(that's how I conlang...)

Oh, sorry, I didn't mention that I was basically asking about words in natlangs.

[Raphael]

I thought of asking how one would onomatopoeia-ize a plant,

Well, where did the first words for specific plants in the first human proto-languages or pre-languages or almost-languages come from?

[Rounin Ryuuji]

For a plant, I might guess rustling for leaves, a sort of admirative sound for flowers, the sound of chewing or indicating something is delicious for a fruit or certain vegetables, a disapproving sound for plants that are poisonous, possibly also a knocking sound for dead wood, or for trees...

[Raphael]

Thank you, those are all very interesting options!

[Rounin Ryuuji]

I can see a large number of them used together with gestures to indicate the rough shape of a plant, and then the onomatopoeia made with them shifting to be words understood as the names of the plant, and so on and so on. It's a rather amusing mental exercise.

[zompist]

Hm, what about the idea that the ancestors of pretty much all word components, if you go far enough back into the distant past, originally started out as onomatopoeia? I mean, it's basically either that, or derivation from earlier roots, or borrowing, and in the latter two cases, that leads to the question of where these earlier roots or words in other languages originally came from.

Wait - while I wrote the previous paragraph, I could think of one other possible original source: baby talk, such as in the case of "mama" and "papa". So make that two basic original sources for word roots.

Max Muller considered both ideas about 150 years ago, and gave them silly names:

* The mama theory. The easiest syllables attached to the most significant objects.
* The bow-wow theory. Imitations of natural sounds
* The pooh-pooh theory. Interjections and instinctive criies.
* The ding-dong theory. Words created by synaesthesia.
* The yo-he-ho theory. Rhythmic chants accompanying work.

Linguists don't take these very seriously, partly because no one really believes that all words originated this way, partly because we just don't know.

I'd add that I don't see why people wouldn't have just made conlangs. I mean, why not just create a bunch of nonsense words and try them out? This is still an available method, especially in slang. People seem to prefer to use existing words, but they may expand their options if few words exist.

[Rounin Ryuuji]

I suspect the evolution of language was at least at first not a conscious process, but that all of the above things probably contributed to it in some measure.

[Raphael]

Thank you, zompist, very informative!

[xxx]

Oh, sorry, I didn't mention that I was basically asking about words in natlangs.

not believing in Adamic languages, gifts of God(s), I think that every language goes through a conlinguistic phase...