Creating dialects and language families

[GrinningManiac]

Hello all,

Most conlang language families I've seen have involved phonological change and orthographical variation and not much else i.e. protolang: ta: nekwi panki'i = 1) taa nipi pang'i 2) da neke: pagi'i 3) tha negwi fankiki

Occasionally I've seen some grammar changes - perhaps nipakogolo became ni kogolo ba or something.

I've never seen many examples of the kind of changes that take place in real languages - seemingly-random-but-explainable shifts in grammar, words being reused as new things etc.

I'm looking especially at the variation and yet slight similarities present across the Arabic dialects from Morocco to Yemen, Iraq to Chad.

I was wondering if anyone had
a) a guide or how-to on developing a dialect substratum in conlangs
b) an example of a conlang family in this vein that I could look to for tips
c) a document detailing the development of dialectal substratum IRL that could be appropriated as a how-to on conlang dialects.

Thanks

[Ser]

What is a "dialectal substratum"?

[GrinningManiac]

Curse my thinking that was a linguistics term and I was coming off all knowledgeable!

I seriously swear I've heard that in a linguistics capacity. I understood it to mean the distribution and variance of dialects within a language i.e. Maghreb, Iraqi, Egyptian etc. make up the Arabic substratum

just a combination of misremembering, tiredness and wanting to be fancy.

[Hallow XIII]

no a substratum is a language abandoned in favor of a new one that proceeds to influence the new language

[Cael]

no a substratum is a language abandoned in favor of a new one that proceeds to influence the new language

Pretty much. So a dialectal substratum would be a dialect being abandoned in favor of another dialect or language. I'm not even sure this is attested IRL.

EDIT: I'm dumb. Of course it's attested. Otherwise we wouldn't have people in the middle of a heavily dialectal community speaking Standard American English.

[Herra Ratatoskr]

You're not thinking of a dialect continuum, are you?

[WeepingElf]

You're not thinking of a dialect continuum, are you?

A language family, of course, is a dialect continuum that has broken up.

[GrinningManiac]

Educational and humbling though this has been, no one has addressed my question.

Are there any guides or resources on development of language families and dialects that could be followed as instructions or guidelines?

[vec]

What you do is apply sound changes to your language in stages. At each stage you can have things break off giving you a tree. At each stage you can also borrow words from neighboring languages that ignore previous sound changes for example, but then when you apply the next set of sound changes, are affected by those. Rinse and repeat as often as you can/want.

Useful tools for making language families are Zompist's Sound Change Applier and I would definitely read his Language Construction Kit (in book form esp.). And I would read his articles on:

Proto-Eastern
It's descendants Axunašin
Cuêzi
and Caďinor
and Caďinor's descendants
Verdurian
and so more: http://www.zompist.com/virtuver.htm

Other conlangers have done similar things.

Also, just reading about Historical linguistics, dialectal conitnua and so forth is useful. It is often good to start with exploring your native language's diachronics first and then branch out.

[Hallow XIII]

I strongly advise against most of vecfaranti's post, except the part where he recommends looking at historical linguistics in-depth. Probably the best thing to go and research is the history of the Romance languages, as they are extremely well-researched and provide great examples of all the things you were looking for.

Otherwise, well, Tolkien's languages are mostly all related, and he has some things akin to what you were seeking (there are several dialects and closely related languages to both Sindarin and Quenya), but the trouble with his stuff is that his conworld works in strange ways and so his languages do not change in the same way as real ones do. But, well, it's the closest we have.

[Ars Lande]

I strongly advise against most of vecfaranti's post, except the part where he recommends looking at historical linguistics in-depth. Probably the best thing to go and research is the history of the Romance languages, as they are extremely well-researched and provide great examples of all the things you were looking for.

Why is that? The method he suggests is a pretty useful model to start with.

[Salmoneus]

I strongly advise against most of vecfaranti's post, except the part where he recommends looking at historical linguistics in-depth. Probably the best thing to go and research is the history of the Romance languages, as they are extremely well-researched and provide great examples of all the things you were looking for.

Why is that? The method he suggests is a pretty useful model to start with.

Indeed. Although I hope nobody needs to be told by now that vec is a far, far better person to listen to in these matter than hollow.

[Hallow XIII]

You may think of my conlanging advice what you like, but if you're going to use my former name please do keep the vowels intact; it is not especially funny this way.

Also, here is the clarification that I accidentally sent to Ars in a PM rather than post on the board:

"You will notice that in GrinningManiac's OP he stated explicitly that phonological change was not what he was looking for, he rather wanted to go beyond that, so I ignored that part of the post. My post was about me not thinking it is necessarily a good idea to look at zompist's stuff, you understand."

[vec]

I will admit I wasn't entirely sure what the original question was. I don't know what language the examples are in and I'm not sure what GrinningManiac is talking about. But I assumed they were looking for examples of conlangs with more than just sound changes applied to them. Zompist explains syntactic and morphological changes and why they happen in each of his grammars. Besides that, his work is extensive, well-done, digestible and entertaining. In fact, he writes most of it as a how-to manual which is exactly something that was requested. You may not aesthetically appreciate Mark's work personally but I've found to to be an excellent reference. It's a far better source than Tolkien's Elvish in terms of realism.

In simple terms, you can say there are two types of spontaneous changes: sound changes and syntactic changes. Both affect languages globally. Analogy however, is a reactive change and can apply selectively to "patch up" holes in systems that are created by sound changes or syntactic changes. Morphological changes are generally the result of sound changes and/or syntactic changes occurring, potentially followed by analogical changes.

The reason why conlangers tend to focus on sound changes is because you can apply them with computer programs and easily track them with a spreadsheet or something like that. Also, conlangers tend to like phonology. Syntactic changes are more complicated to describe and therefore harder to track. Conlangers therefore tend to maybe focus on them less, and when we post about such things, few people generally comment because it's much more complicated to understand.

[Cedh]

The only small difficulty with Zompist's languages as a model for creating dialects and language families is that he created the protolanguage in retrospect, so some of the sound correspondences and grammatical developments in that family are not exactly what you'd see in natlangs. They're still a very good example of a conlang family though, and I agree with Vec that he describes morphosyntactic changes really well.

Some other examples (which were constructed going forward, and with realism in mind) can be found in the context of the collaborative conworld Akana:

• I would recommend looking at the Western languages descended from Dewrad's Proto-Western, especially the Coastal group consisting of Proto-Coastal-Western (PCW) and its three descendants Ìletlégbàku, Isho‘u ‘Ohu and Doayâu. PCW's sister languages Çetázó, Tmaśareʔ and Iŋomœ́ (and to a lesser extent Empotleʔá as well) also exhibit some interesting and realistic grammatical developments starting from Proto-Western, often accompanied by short notes of how new morphological and syntactic innovations came about, which sometimes explicitly includes analogy. There is also a comparative glossed translation for this family which allows side-by-side comparison of how the different languages tackle the same text.
• You might also want to have a look at the Fáralo branch of the Edastean language family, descended from Zompist's Fáralo. There are the two fairly closely related daughter languages Namɨdu and Woltu Falla, the somewhat more divergent Cəssın, and morphology sketches for two varieties intermediate between the first two of these, Farwo n-Abebbu and Fallo na Mendia. Both of the latter are genetically part of the Namɨdu subbranch (and thus classified as dialects of Namɨdu), but they show phonological and morphological developments similar to what Woltu Falla does. As in a real natlang dialect continuum, this is most obvious in the variety which is geographically closest to WF, namely Fallo na Mendia. Again, there's also a comparative translation available.

(Many of the other Akana languages are of course interesting as well, especially in terms of inspiration for changes in morphology and syntax, but not all of them are quality conlangs, and, more importantly with regard to the OP's question, most of the linguistic families other than the above suffer a bit from what I call the "conlanger's bias": making closely related languages too different from one another because of a desire to put in something unique.)

[GrinningManiac]

Thanks guys this was exactly what I was looking for. Cedh especially.

@Vec - yeah you're right that conlangers tend towards phonological change. I too tend to lean in that direction and the whole point of this thread was to educate myself enough to create more interesting divisions in the grammar and lexicon beyond the words sounding different.

With a lot of conlangs it seems like the ancient grammar is preserved almost wholly but the words become unrecognisable in pronounciation. Whilst this is fine and I am in no position to criticise anyone else's work I feel this is an incomplete interpretation of linguistic development.

Thanks again all

[Curlyjimsam]

My History of the Viksen Language contains a lot of information on morphosyntactic change (though I can't guarantee it's all necessary realistic). Obviously this is just one language rather than a whole family, but it might still be instructive.