Using Proto-European languages to make your own fantasy languages?

Has anyone ever used proto versions of languages to make your own language? I've been thinking about this for months but my settings have yet to have a made up language for any reason

For example, for an elvishish language, using Proto-Celtic to have a language that sounds similar to Irish or Welsh while still being completely made up, along with using it for names(Finn/Gwyn are common elements of names from the same Proto Celtic word meaning fair/white, so making up a new declension such as Venn, and then using another one for war(Cath/Cad), Cad, combining the two makes the name Cadvenn, meaning fair warrior, which can also be done with lots of other word and name roots in a similar way

Thoughts?

Other urls found in this thread:

omniglot.com/writing/venetian.htm
unboundworlds.com/2017/05/names-arent-neutral-david-j-peterson-on-creating-a-fantasy-language/
twitter.com/AnonBabble

>venetian

>not an italian dialect

Silly comic is silly.

This is a neat idea. It'll sound realistic while being completely created. Check out the various constructed language communities for help.

Not exactly, but I do like to think that Chondathan (the major language of the Sword Coast of Faerûn) sounds like English if English had never undergone a great vowel shift in Middle English, so that the vowels retain a more Continental sound (ah, ay, ee, oe, uu, like in Italian or Spanish; rather than ah, eh, ih, oh, uh).

Gotta tell ya, English with Continental vowel pronunciation sounds weird as Hell, especially if you add in other Romance tendencies, like silent H, or J as Y, or no "th" sound. For example:

The quick brown fox jumps over the lazy dog

Tay queek broen foeks yuumps ovaer tay lahzee doeg.

I saw that image today in a slideshow. You wouldn't happen to be in my Spanish class?

It's the same as German or Chinese; the line between a language and a dialect depends on how politically beneficial each is.

Not taking any language courses, just really like the design of that pic

omniglot.com/writing/venetian.htm

>The language is more closely related to French and Spanish than it is to Italian.

>serbian
>croatian
>bosnian

REEEEEEEEEEEEEEEEEEE

If I read your post correctly, you want to steal the evolution, but not the lexicon.
See, second to hardest thing in conlanging is generating a lexicon. It's a bitch, just believe me. You have to go balls deep in etymology, to model borrowing patterns (which are dependent on the history of speakers), even as far as trying to understand just which parts of the lexicon are more likely to be based on onomatopoeia etc. It's a rabbit hole really, and you wouldn't even have an actual linguistic theory to back you up, it's still blank spots on the map.

Modelling sound change is rather easy, if you know how to translate the aesthetic you are aiming at to the language of linguistics. Like, what makes Germanic words sound Germanic? It's actually syllable structure and patterns of onset-coda consonant pairing. It's subtle, but it's easily modelled anyway.

Now, the hardest thing is linking your conlang to actual usage, as in semantics. It's a soul-draining exercise in futility. Ever tried making up idioms without them being arbitrary? Well, never even try, it is, for all intensive purposes, impossible to do well.

So, if your design goal is to make a language that sounds similar to Irish OR Welsh (they are very different phonetically), you will first have to decipher the aesthetic you are after. Then think, how to build a model that will generate you that aesthetic. Try to be lazy when you can afford to.

Reminder Low German is genetically closer to English than it is to High German

Not OP, but what do you suggest doing "to be lazy" so to speak.

That's not incorrect

>Has anyone ever
Yes, Tolkien and everyone who has ever read about Tolkien

Don't dabble into semantics for example. There's invisible work that IS going to have effect on how the end result is perceived, even subconsciously, and there's invisible work that's not gonna amount to anything at all.

I could make a conlang that sounds like Irish, written like Welsh, has a V2 syntax, uses morphological patterns from Modern Greek, and is identical to Russian in actual word usage and meanings. I also could start from zero with all that, and spend one hundred times more effort to get a result that is not even guaranteed to be better, rather the opposite.
And the fact that I've stolen morphology from Greek is going to be visible either to a very experienced linguist or a Greek native.

So, ask yourself a question: what I'm trying to accomplish with my conlang? If it's for a fantasy setting, I could make you a good one in two hours. No need to spend months agonising over a hard task with little returns.

That pic of from a lovely webcomic called Stand Still, Stay Silent, drawing off of scandinavian mythology, written by a Finn. Great stuff, would heartily recommend.

I'm working on one right now. I want it to be similar to english in terms of syntax and use of particles and compound words, but still sounding pretty different. The rules are looser, so it's more like EME, where shakespeare was free to do whatever the fuck he wanted.
Phonologically and phonotactically, it's more like a very slavic language, heavily influenced by mediterranean languages such as Greek/Latin.

For an alt history setting, I want to design a language contemporary to proto-celtic, within the celtic family. But I'm practicing with something closer to home.

Honestly I've been really lazy about it lately. I should get back on top of it. I want to be able to speak it fluently and write an epic in it.

I'm trying to create a naming language that can be morphed and changed so it has descendants and stuff and a really basic grammar. I'm studying linguistics right now, and that's taught me enough to know that I should set my sights low. Admittedly, what do you mean by "don't dabble in semantics?" I understand the folly of the example you provided, but a language with no meaning is, well...meaningless.

Or are you suggesting isolating the features of a language and creating nonsense words that sound sufficiently, say, "german" rather than making up some nonsense words with meanings that sound german.

My project's not really supposed to mimic anything, it's just supposed to sound reasonable.

Steal morphemes. Simple as that. Find a language, learn it's morphology and fucking steal away.

You'd have to come up with an algorithm to convert those morphemes to your phonetic/phonotactical framework. Sure it will result in something that would look unnatural (as in, such changes just don't happen in natural languages) under a close scrutiny by an experienced linguist, but it also gives you ways to add flavourful details on a whim, stuff like fancy alternations, semi-regularity, inject stuff that superficially looks like unproductive derivation, whatever.

To provide a quick example. Say, you have an isolating SOV language.
Inventory is m p b f n t d s r l k g h j w a e i o u
Phonotactics is C1-V-C2, where C2 is restricted to nasals and liquids.
Now, we decide to steal from Russian. The root "to sit" has at least 9 different forms (caused by palatalisations, alternations and other crazy IE stuff): sad, ses, sazh, sid, se, sizh, sed, sedj, sjad.
To properly convert these, we would have to do something with the impermissible codas. I suggest that we incorporate the following thematic vowel, if it's a verb, and one is indeed present. Otherwise, we use u as a velarised epenthetic vowel, and i as a palatalised.
Converting: sadi, sesi (the full word is sestj, the "tj" part triggers regressive assimilation), sagi, sidu, se, sigi, sedu, sedi, siadu.
There you go, a wonderful array of alternating roots that are going to make anyone question your sanity.

Also, I probably have a slightly unfair advantage of speaking four rather dissimilar languages.

>four rather dissimilar languages
Yeah, I've just got english, french, and italian under my belt.
So if I'm understanding you correctly, you pick another language whose sound you like and figure out what words you're going to need from it. Then you convert the relevant morphemes into the conlang and fiddle until it fits the parameters. If that's right, I think I'm set.
Though, one more question, in your example, what would you do with those nine converted russian forms of "to sit?"

Bear in mind that I'm not an expert, and certainly not somebody training to be a linguist.
unboundworlds.com/2017/05/names-arent-neutral-david-j-peterson-on-creating-a-fantasy-language/
This article might help you. It specifically talks about how to create a naming language. However, it is written for people who have no idea what the fuck they're doing (like me!) so it may drag on a bit for you.

>whose sound you like
A very important distinction: you have to watch out for incompatible phoneme classes in certain positions. Say, Russian has a lower than expected frequency of CVC roots that have dental stops both in onset and coda. English is much more permissible with that, and if you were to convert English phonemes into a Russian-sounding end product, you would get this unnerving feeling of not quite right. Read more about "Same Place Avoidance theory".

>If that's right, I think I'm set.
Yeah, you're on a right track. It's gonna take a few attempts, it always does.

>what would you do with those nine converted russian forms of "to sit?"
Use them. Since Russian is my native language, I'd know when to use which variant. In an isolating language, it'd actually be fucking hilarious, because it would make people think of shit like Early Middle Chinese, which was rather isolating, but had weird vestiges of the earlier Old Chinese stage where it had an actual functioning morphology. So, it'd boil down to redundancy/double signalling for flavour.

Note that this feature has cost me around 5 minutes of work. That's how you do lazy.

You know your stuff, but intents and purposes, not *intensive purposes. Ykeep throwing the dogs, we'll keep trying with lex and sem, hokay?

Who are you? I bet you're from the ZBB.

Can I fuck your brain?

I always write "intensive purposes" to see if somebody corrects me.

The fuck is ZBB?

You're not the first on this board. The previous user liked to imagine I'm a hot linguist chick. Oh well.

It stands for the Zompist BBoard, a forum where people discuss worldbuilding. Started by a guy named Mark Rosenfelder who wrote a guide about how to create languages.

I think it's listed as Similar place avoidance (same place just brings up Veeky Forums and russian Veeky Forums), but it's all scholarly articles that are a liiiiitle over my pay grade, you know of a good summary or a link to an article that explains it rather than articles that are about how one weird minutae about it is being applied.

Hey, that previous user was me! Aw, you are wrought of manflesh. But yeah the ZBB is a conlang community. 'Sone I was once part of. Glad to see there's another conlang nerd around! I always dealt in semantics, which is why I grew needlesome on that whole shebang. After all, meaning is fun to play with!

&also all three posts yresponded to were me.

You. I like you.

>Similar place avoidance
That must be it. It's been quite some time since I read that.

So, okay, I'll try to get the gist of it across to you.
You know how there's always this nice hyperbola popping out when you try to graph phoneme frequency in a language?
It means that while for a single morpheme constituent phoneme choices are effectively random, there are constraints on the macro-level.
Now, imagine a language that only has C1VC2 morphemes, and three places of articulation for consonants: labial, coronal and dorsal. Let's denote them L, C and D.
If we say that L has frequency of 1/8, C 1/2, and D 1/4, we could draw a table of frequency values for all the possible C1-C2 pairs.
LVL = 1/64
LVC = 1/16
LVD = 1/32
etc.

The resulting table can be visualised as picrelated.
Everything looks perfectly symmetrical and fine, but it just doesn't happen in reality. For some combinations of C1 and C2 the real frequency is much lower than the expected frequency. Most often, this happens with LVL, CVC and DVD.
Why? Well, we switch to acoustics for a second. A typical vowel is articulated roughly in the center of the mouth, a jaw and tongue position that is different from any of L, C, and D. But nevertheless we can say that any of LVL, CVC, DVD requires movement from one position to another, and back. That is ineffective. Compare with something like CVD, which doesn't require movement direction inversions. Languages are very good at self-optimisation, so these inverse-movement cases tend to get pruned with time.

So, tl;dr: simple phoneme frequency information is not enough to probabilistically describe a language's phonetics. Other constraints are also in place.

Gotcha, gotcha. Yeah, I've learned a little bit about that, but under a different set of terminology (probably babby's intro terminology). While I was reading it, I thought, "I think I'd be the only one who'd notice that," but when I put it in practice and tried to say stuff, you're right, it'd be obvious even to a non-linguist. I hope this prolonged discussion has helped the OP and anyone else who is interested.

Nope. I'm lazy so I just use my tribe's native language and Romanize it a little so it seems exotic without being too alien for my players. It's real fun, since I get to use obscure legends and stories passed down from my elders in a somewhat meaningful way.

What language?

Meskwaki, or Fox as we're sometimes called

You won't find a lot of language resources for it out there. I'm lucky that I have access to people who grew up speaking it as well as obscure books that you won't find at your local library. It's interesting to see how the language has changed, as the oldest books document words and pronunciations that are no longer used. I actually spoke with our tribe's historian on how this sort of linguistic "bottlenecking" would affect the language in the future, what with the only native speakers being really old folks, and the young speakers learning, at best, alongside English, which will of course affect how they learn and their view of our language.