cws
Greetings Guest
home > library > journal > view_article
« Back to Articles » Journal
Proto-Languages Part 1
3▲ 3 ▼ 0
How to make a good proto-lang
This public article was written by [Deactivated User], and last updated on 17 Jan 2015, 13:51.

[comments]
?FYI...
This article is a work in progress! Check back later in case any changes have occurred.
<noindex>

There is a common misconception that proto-languages are somehow "lesser" forms of speech, more "primitive". After all, isn't that what 'proto-' means? Less developed? Not in a linguistic sense. Historical linguistics employs a set of prefices to denote reconstructed languages, which are just as real as recorded languages. In this capacity, 'proto-' indicates a language reconstructed through the comparative method, such as Proto-Indo-European, or Proto-Uralic, or Proto-Afro-Asiatic, to name but a few. To repeat, these are real languages (though anachronisms are virtually guaranteed), spoken by real people in a real place - they had living and changing grammars and lexicons, were subject to processes of language change, and were complex and dynamic. Not "primitive" in any sense.

So, when embarking on the construction of a proto-lang, you should think of it as another con-lang, with its own full and rich grammar. It should have its own quirks, such as PIE's deponent verbs, which are active verbs with only medio-passive forms. What's more, it should itself have a history - even the "First Language" grew and developed from a primordial brew of speech signals.

That said, if your proto-lang is for the purposes of creating a language family, you don't have to completely flesh out every aspect of derivation and morphology right away. Each language will innovate in their own ways, will lose category this and derivation that. Some features may even be lost by all daughters, in which case there is no need to add it to the proto-lang, except for completeness' sake. You don't even have to fully specify the realisations of the phonemes - you only need give enough for the purposes of sound changes.

This mirrors the state of reconstructed nat-langs, of which PIE is the best example. New research is constantly shedding light on aspects of morphology, theories are emerging of features indirectly attested in the daughters, proposals are formulated over the identities of long-known aspects of the language. Laryngeal theory, Jasanoff's h2e-conjugation theory, and even a recent idea I had on the nature of the labio-velars are excellent examples of this.

And even then, there will be disagreements over the language which may never be settled, such as the realisation of the three obstruent series (discounting /s/ and the laryngeals, which have problems of their own). The traditional model holds that they were voiceless stop ~ voiced stop ~ breathy-voiced stop, as in Sanskrit; glottalic theory insists they were fortis stop ~ ejective stop ~ lenis stop (with your own favourite fortis::lenis contrast); I believe that they were voiceless stop ~ voiced stop ~ voiced fricatives (with /s/ and the laryngeals completing the symmetry as a series of voiceless fricatives).

Ultimately, it's probably far simpler to just make a full and complete language, but if you really want to avoid all that work so you can focus on the daughters, build up the proto-lang at the same time as the daughters, adding to it as you need, but there's a trap here, that you may inadvertantly kitchen-sink the proto-lang. When starting each new daughter, look at what you already have, and what it could give you for this language, before you add more. The more basic a category is, the more languages will reflect it - very marked oppositions are likely to collapse in most languages, and even where they are maintained, they will likely be moved to a less-marked function.
Comments
privacy | FAQs | rules | statistics | graphs | donate | api (indev)
Viewing CWS in: English | Time now is 29-Mar-24 09:07 | Δt: 1572.0789ms