When the pieces fit together

[DorotheaBrooke]

In my conlang Hayakan there is a verbalizer -amati which means to "to lose X." For example waka "voice, speech" → wakamati "lose one's voice." Hayakan is heavily head initial, so I justified this as one of several derivational suffixes borrowed from Sewaqli, the traditional liturgical/prestige language, which is head-final and suffixing. I made this choice many years ago.

Today I was reviewing some old notes on Sewaqli, which I am finally getting around to making, and remembered this suffix; I realized I would have to justify it. Then I realized that Proto-Kuzitic has a verb *áma which means "drop, give away," (and in many descendants, though not Sewaqli, becomes the verb for "to give.") Proto-Kuzitic is the ancestor of Sewaqli, and the -ti is very easy to derive, perhaps from an intensive or applicative. So I ended up coming up with a great historical justification for this affix, without any work. (I don't think it's possible I invented this form and forgot about it, as I derived everything in Awkwords).

Something I think about a lot is how art, including conlanging, is often about creating a lot of loose ends and then realizing with a start that they link to each other. What are some moments for you all where you feel the pieces come together in your languages?

[äreo]

Great question and cool example.

I'll have to think about this a bit to find some good examples, but Msérsca has definitely given me some moments like that. It contains some features from defunct conlangs I worked on as a teenager, but of course those features fit within a different context now, and that can play out in serendipitous ways.

[bradrn]

What are some moments for you all where you feel the pieces come together in your languages?

I have such moments every so often. Probably the most recent was when I worked out that I can derive coverbal roots from ideophones, giving me just the right set of properties ‘for free’.

[WeepingElf]

This happens to me every now and then when I am working on Hesperic/Albic project, especially with etymologies. A recent example is Proto-Hesperic *hak'wah 'river', which I fancy to be the source (by borrowing) of Latin aqua 'water' and Old High German aha 'river' - I found out that it could be a thematic noun *h₁oḱwos from the PIE root *h₁eḱu- 'swift' (also underlying PIE *h₁eḱwos 'horse'). Another example: Old Albic sighlam 'wing, sail', borrowed into Germanic as *sigla- 'sail', giving Engl. sail etc., from PIE *selh₂- 'jump' > 'fly' in Hesperic (the development *-lh₂- > -gh- is regular in Old Albic).

[xxx]

It's typically these eureka moments,
that border on linguistic orgasm,
that are the driving force behind conlanging...

I'm still having HPPD from the overwhelming one,
that lasted several days,
generated by the nesting of the semantic primitives,
at the source of Contamoḻinilam...

Since then, it's rather the semantic use of Contamoḻinilam,
which sometimes allows for a bold metaphor,
that changes my perspective on the wor(l)d,
that shakes me up in this way...

[WeepingElf]

Working on the Hesperic/Albic project, I sometimes have the feeling as if I was discovering a lost linguistic reality rather than inventing a conlang family. When I uttered this sentiment on the CONLANG mailing list a few years ago, someone replied that I perhaps remembered a language I spoke in a previous incarnation. Well, maybe I am in morphic resonance with an actual extinct language of Bronze Age Britain But no, I can't reconstruct that language (well, I doubt anyone can), only re-create it.

[rotting bones]

But no, I can't reconstruct that language (well, I doubt anyone can), only re-create it.

Britain is probably too far off from surviving communities for an educated guess. Without proof, I somehow suspect that languages of pre-Indo-European nomads in Southeast Europe shared areal features with the polysynthetic grammars of Northwest Caucasian languages, but that they belonged to completely different language families. Pre-Indo-European farmers were more likely to have spoken agglutinative languages like Etruscan, Basque or Georgian. The phonologies of these families can only be guessed at from toponyms and substrates. The lexicons are mostly lost.

[WeepingElf]

But no, I can't reconstruct that language (well, I doubt anyone can), only re-create it.

Britain is probably too far off from surviving communities for an educated guess. Without proof, I somehow suspect that languages of pre-Indo-European nomads in Southeast Europe shared areal features with the polysynthetic grammars of Northwest Caucasian languages, but that they belonged to completely different language families. Pre-Indo-European farmers were more likely to have spoken agglutinative languages like Etruscan, Basque or Georgian. The phonologies of these families can only be guessed at from toponyms and substrates. The lexicons are mostly lost.

Fair. We can only speculate on these languages. My idea behind Hesperic is that the Bell Beaker people, who originated in the steppe and spread across Western Europe about 2500 BC, spoke a lost branch of Indo-European, which may have been related to Anatolian (Hittite, Luwian etc.). Celtic came later, not before the Urnfield expansion about 1300-1200 BC, more likely even as late as the Hallstatt culture about 800 BC. In my personal scenario, Urnfield brought Proto-Celtic to southern Germany (and Proto-Italic to Italy), and Hallstatt spread it across Western Europe. The extant Celtic languages are IMHO too similar to each other to have Proto-Celtic as early as 2500 BC. Around 1 AD, British and Gaulish were essentially the same language, with only some minor dialectal differences, and Irish differed from that mainly in lacking the change /kw/ > /p/ that had happened in British and Gaulish. The divergent typology of the Insular Celtic languages (VSO word order, initial mutations and all that) is the result of a small number of far-reaching changes that occurred in the "Dark Ages" between 400 and 700 AD. Whether there is a substratum involved is anyone's guess (the late date speaks against that, but some preparatory changes, such as subphonemic lenition of postvocalic stops, that do not show up in the sparse written records of earlier Insular Celtic may have been earlier); but the frequently adduced Semitic substratum is IMHO bullfrogs.

[Man in Space]

What are some moments for you all where you feel the pieces come together in your languages?

• Uniting Proto-O and CT to make the Beheic family by working backwards.
• AW’s most recent revision of the verbal (and thence nonconcatenative) morphology.
• When I worked out the Kgáweq’ noun class system, verb tenses, general address practices, and phrasal concord.
• When I standardized Common Caber’s glyph composition system.
• My recent breakthroughs in CT cuneiform and hieroglyphics.

[äreo]

Just had a great example of this. At this stage of Msérsca's development, I have some grasp of what its ancestor c. 1000-2000 years prior would have sounded like, and I've gotten pretty good at "justifying" whatever new coinage or discovery I have by thinking through the diachronics at the same time. In other words, usually when I find a new word in Msérsca, I at least consider how it would have developed from its ancestral equivalent. But I still haven't fully and systematically fleshed out the proto-language. So I'm setting out to do this.

With that in mind, I just combed through my lexicon to ensure that there were no words with intervocalic -s-, as earlier -s- should have become voiced and then rhotacized. Msérsca should allow -r- (former -z-) and -ss-, but no -s-. And as it turns out, I had either already accounted for this or unintentionally respected this rule; only one word violated it, and this was a word I had coined years ago that also violated another rule.

Thinking about it now I'm pretty sure I did account for this last time I did a major chunk of work on Msérsca, and had left that word alone intentionally because I wasn't sure what I wanted to do with it at the time. But still it feels good to know there wasn't much adjustment to do. It also helps that my lexicon is still only some ~500 words.