Aesthetics of a Proto-Language

[cerealbox]

Having finally after many years of "not quite perfect enough" stuggling, I've finally gotten down some basic grammar and vocabulary for my Conlang, I immediately realized I had a new problem awaiting me: the proto-language. My conlang will not be complete until I've derived it from a Proto-language and created sister languages The problem I have is getting the aesthetics down for the proto-language. My first reaction was to create as primitive a language as I could and evolve it through regular sound laws, however, this has proven quite difficult and I've come to realize that the proto-language will never be basic enough or simple enough or logical enough. You'd think creating a pigin would be easy enough, but alas, it is quite difficult when you have some idea of how you want things to end up and I'm stuck on the difficulty of getting complex grammatical rules derived from fairly simple changes of a fairly simple language.

So my question is, what's your approach to creating a proto-language? What qualities are you trying to achieve? What's your process?

[Aurora Rossa]

Protolanguages are no simpler than any other kind of language. The term simply refers to languages which have given rise to language families. The languages which spawned those families had all the functionality and complexity of modern languages. You need only look at Proto-Indo-European to see that simplicity has nothing to do with it.

[cerealbox]

Protolanguages are no simpler than any other kind of language. The term simply refers to languages which have given rise to language families. The languages which spawned those families had all the functionality and complexity of modern languages. You need only look at Proto-Indo-European to see that simplicity has nothing to do with it.

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic. But for my purposes, I thought my conlang needed a "starting point" and that starting point would be a pigin since I can't very well go back to the dawn of language itself. My instinct (and maybe it's an instinct I should abandon) is to derive complexity from simplicity as simply as I can.

[GreenBowTie]

Protolanguages are no simpler than any other kind of language. The term simply refers to languages which have given rise to language families. The languages which spawned those families had all the functionality and complexity of modern languages. You need only look at Proto-Indo-European to see that simplicity has nothing to do with it.

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic. But for my purposes, I thought my conlang needed a "starting point" and that starting point would be a pigin since I can't very well go back to the dawn of language itself. My instinct (and maybe it's an instinct I should abandon) is to derive complexity from simplicity as simply as I can.

i don't think it's an instinct you should necessarily abandon; just realize that it's kind of a two-way street. complexity derives from simplicity, but simplicity also derives from complexity. as some aspects of a language grow more complex, develop exceptions, whatever, different aspects (or even those same aspects!) simplify as well, so to be realistic you need to have both currents running through your languages' histories

[GreenBowTie]

Protolanguages are no simpler than any other kind of language. The term simply refers to languages which have given rise to language families. The languages which spawned those families had all the functionality and complexity of modern languages. You need only look at Proto-Indo-European to see that simplicity has nothing to do with it.

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic. But for my purposes, I thought my conlang needed a "starting point" and that starting point would be a pigin since I can't very well go back to the dawn of language itself. My instinct (and maybe it's an instinct I should abandon) is to derive complexity from simplicity as simply as I can.

i don't think it's an instinct you should necessarily abandon; just realize that it's kind of a two-way street. complexity derives from simplicity, but simplicity also derives from complexity. as some aspects of a language grow more complex, develop exceptions, whatever, different aspects (or even those same aspects!) simplify as well, so to be realistic you need to have both currents running through your languages' histories

[Pole, the]

I want to see a language family based on Volapük!

[Vuvuzela]

But for my purposes, I thought my conlang needed a "starting point" and that starting point would be a pigin since I can't very well go back to the dawn of language itself.

Pisin gatim samting no isi tasol, olsem long tok no pisin. Empela nogatim traipela namba long tok, empela nidim tok bilong tingting toktok long samting.
Loose translation: Pidgins have complicated aspects, too. They don't have many words, so they have to use those words cleverly to express things.

This results in creoles steeped in idiom, for instance "gat bel" or "has belly" in Tok Pisin for "is pregnant". Also, pidgins sometimes end up keeping weird bits of grammar from the L1s of the people who use them, for instance, Tok Pisin has the "-im" suffix on every transitive verb, and pronouns like "yumitripela", which is in the first-person-inclusive-trial. It's simpler in some ways than English, but also more complicated than others. Even if you're making a pidgin, there's no way to make it perfectly simple and regular and still be naturalistic.

[Ollock]

There is a very good reason, when doing historical conlanging, to start with the proto-lang and work forwards, rather than the other way around. It makes things easier.

[sucaeyl]

[A]s a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic.

Do you believe this tendency to be more than coincidence?

[Ars Lande]

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic. But for my purposes, I thought my conlang needed a "starting point" and that starting point would be a pigin since I can't very well go back to the dawn of language itself. My instinct (and maybe it's an instinct I should abandon) is to derive complexity from simplicity as simply as I can.

Proto-languages are not any less - or more - complex than attested languages; the key to simplifying them is, I believe, is that a great deal of their complexity may have been lost.
(For instance, Old Chinese probably had a complex derivational and inflection morphology, comparable to written Tibetan. There isn't enough data to reconstruct it, though).

The proto-language I've used is intended to be realistic, but I've mostly worked on a word list. The syntax and morphology I've developped as it allowed to develop the descendant languages, so it is rather minimalistic.
I'll complete it as I go, or when I work on another language from the family. I assume that otherwise, whatever I haven't described has been lost.

[Anguipes]

My method is, if your proto-lang is a tool, not a project in and of itself where detailed knowledge of grammar and word use is important: be fuzzy. That way you can pretend there are greater details under your simple outline. Apply a series of related concepts to a root as meanings, and give a grand shrug as to what it meant precisely to the original speakers. Sometimes don't even bother defining whether some root was a noun, a verb, or whatever. Have a basic idea of word order and inflections (if any), but again don't detail it. Use terms like "broadly", "usually", "probably" and "tenuous". Then add or change details, or whole chunks your original sketch, as you work on daughter langs - just as long as you end up being able to derive all the features of each daughter somehow, it'll be fine.

[cerealbox]

[A]s a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic.

Do you believe this tendency to be more than coincidence?

I do. I can't repeat the original argument I heard too well, but I do know that the number of words the language contains skyrockets (makes sense). Small, pre-literate societies tend to lose words at a great pace, which results in the creative destruction of morphology. I would conjecture myself that as the number of words a language has increases, the less need there is for complex morphology. Conversely, the fewer words a language has, there is a greater need for complexity elsewhere.

If these views are heterodox, I apologize. it's merely my own (lay) understanding of how languages work.

[Herr Dunkel]

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic.

So, Latin was spoken by apemen and Georgian still is, while Chinese is spoken by Übermenschen?
That's like reverse Whorfianism on a cultural and technological scale!

Old French was far more analytic than Latin, but their culture was simpler and rarely who was literate anymore. Pre-literate societies don't have less words than modern-day ones, they just have different.

For example, Northern Sami has sixty plus terms for raindeer, while English gets by with only one (raindeer); once, English had a multitude of terms for young of farm animals (calf and kid are a good example), and now they aren't as used anymore.
So yes, that's heresy what you're preaching.

[Ollock]

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic.

So, Latin was spoken by apemen and Georgian still is, while Chinese is spoken by Übermenschen?
That's like reverse Whorfianism on a cultural and technological scale!

Old French was far more analytic than Latin, but their culture was simpler and rarely who was literate anymore.

He actually is talking about a real effect. Languages that are more isolated (less trade) or are spoken by smaller groups (which often goes along with isolation), tend to be more morphological complexities. Also, I don't know anything about Georgian, but Latin is certainly not anywhere near any ideal of complexity -- try comparing it to Navajo. Also, yes, Chinese was spoken not be übermenchen, but by the people of a vast empire that managed to hold together (in a way) for a couple thousand years over many different changes in leadership, largely because they invented bureaucracy early.

Pre-literate societies don't have less words than modern-day ones, they just have different.
For example, Northern Sami has sixty plus terms for raindeer, while English gets by with only one (raindeer); once, English had a multitude of terms for young of farm animals (calf and kid are a good example), and now they aren't as used anymore.
So yes, that's heresy what you're preaching.

Who said anything about number of words? It's rather difficult to count the words in a natural language, anyway -- a living language has new words coined every day, and old ones loosing favor almost as often.

[Pole, the]

I didn't mean to imply that Proto-Langages tended to be simpler; as a matter of fact, more technologically primitive cultures have a tendency to have greater morphological complexity than modern, literate, industrialized societies who tend toward the analytic.

So, Latin was spoken by apemen and Georgian still is, while Chinese is spoken by Übermenschen?
That's like reverse Whorfianism on a cultural and technological scale!

It's more like reverse Marrism.

[Curlyjimsam]

If "aesthetics" are your main problem, then I would say that the aesthetics of a proto-language have fairly minimal bearing on the aesthetics of its daughters, provided the time period involved is long enough. English does not look much like Proto-Indo-European in any of its reconstructed variants, nor much like its more distant IE relatives. Or, to take some conlang examples, look at the following from my own work. (1) is from Proto-White-Viksorian; (2) from its descendant, modern Viksen (note that both sentences mean the same thing, and many of the words in (2) are directly and regularly derived from those in (1)). (3) is from Proto-South Atlian, and (4) from its own descendant, Greater Atlian (again the same things apply).

(1) gazba mhéma lia ethè rhena ei arak:èdém lia len:ì byé húlaí zudi
(2) djig yin yi mimi wugaz yitsjin ye yin edu wag-æz yæ

(3) tí potesum wél, hai kwó kúti is tí lekel
(4) pu saupu kuwissu yokiuno pudessi u pu

These are just random examples, not chosen because they exhibit a particularly great deal of change or anything like that. So I would suggest that it is the changes, not what you start with, that are the real determinants of the aesthetics of a language, and as such the aesthetics of the proto-language aren't particularly importance.

[LinguistCat]

Also, which Chinese? Mandarin? Cantonese? One of the other dozens of "dialects" that are basically distinct languages if they weren't all spoken in the same country?

[Ars Lande]

Ollock, cerealbox: I take it that what you mean by 'complexity' refers to degree of synthesis.

A few counterexamples:

- An appreciable number of African languages tend towards the isolating. IIRC, the Khoisan languages, spoken until recently by hunter-gatherers (and in any case, by a very isolated people) are isolating.
- Latin and Greek, though spoken by agriculturalists, as opposed to their more pastoralist ancestors, have morphologies comparable to that of PIE.
- proto-Uralic is reconstructed has having less cases and more postpositional constructions than Finnish, even more so compared to Hungarian.
- Chinese languages likewise show no sign of morphological simplification through history. Old Chinese had some synthetic elements, but was still more analytic than English. If anything, Mandarin is acquiring synthetic features, not losing them.
- Classical Nahuatl is comparable to other Uto-Aztecan languages.

So far I'd say there is little correlation other than geographical.

[Click]

I think I should add most of Balto-Slavic languages as another counterexample.

[Vuvuzela]

Austronesian languages tend to be morphologically and phonologically rather simple, and , despite the advances that many groups of speakers had (see: Hawai'ian empire, and those head things on Mata Nui), most Austronesian groups tended to be rather small and isolated, and none of them AFAIK developed literature before Western contact.
Arabic and Hebrew, on the other hand, are more like seriouslywhatthefuckisthisthere'stoomuchablautGAAAAAAAAHHHHHHHH.

[cerealbox]

Wow. I think I accidentally touched on some sort of taboo around here. I didn't make up the morphological complexity thing.

Also, pointing out that there exists exceptions to a general statement does not invalidate the mathematical correlation found. Not every langauge is going to fit on the line, but a correlation can nonetheless be measured.

[Salmoneus]

Wow. I think I accidentally touched on some sort of taboo around here. I didn't make up the morphological complexity thing.

Also, pointing out that there exists exceptions to a general statement does not invalidate the mathematical correlation found. Not every langauge is going to fit on the line, but a correlation can nonetheless be measured.

So go on, measure it.

Repeating an urban myth is not more valid than citing evidence, which is what Ars Lande was doing.

Your theory has, so far as I can tell, zero evidence to support it, and does not make sense. [Ollock's theory makes some sense, but is similarly unsupported by evidence]

[cerealbox]

My original source was "Through the Language Glass: Why the World Looks Different in Other Languages" (took me a hell of a time to look through my amazon history to figure out where I first heard it). Alas, I don't have the book anymore and cannot site the sources he uses. Google search does not seem to be turning up very accurate results.

Still don't know why everyone is being so hostile about it.

[cerealbox]

Anyway, didn't mean this to turn into a sociolinguistics debate or whatever this is. I'm just wondering what sorts of things you guys look for in a proto-language. is it even that important to you? As someone who has never cared that much about history or things like that i find myself oddly compelled towards etymology and historical linguistics. Don't know much about it yet, but I think to create my 'perfect' language it needs a semi perfect history complete with perfect imperfections and flaws. A pidgin seemed the easiest way to go, but I'm learning it's hard for me to derive complexity from simplicity (and then, frankly, back to a kind of simplicity). What approaches do you all use? What are you trying to get out of your proto-languages?

[sucaeyl]

We weren't being hostile; we were being skeptical, as is to be expected when such a claim as yours is made. I have to agree, however, that from a cursory glance at WALS, there seems to be slightly more synthesis in languages spoken by traditionally simple-technology societies. I don't believe that the trend is too great to be more than coincidence.