Zompist Bboard Again

I know some of you have posted some critiques of David J. Peterson, but I can't find any of them and I'm talking about them to someone else.

There was once a thread in Ephemera where people posted their opinion on his book The Art of Language Invention. I only remember Jipí's review, which you can find on his website: https://ayeri.de/archives/5470

If I remember correctly, the general consensus was that while it's not a bad book, it's a bit too much over the place. Like, it should either have been about DJP's involvement as a conlang consultant in various movies and TV series, or about introducing linguistics. Leaving out the entertainy bits would've allowed to not rush through topics the book mentions as much, but I suppose that would've been too dry to DJP and/or the publisher, especially seeing as Penguin isn't exactly reknowned for textbooks as far as I'm aware.

(Also, hi. I rejoined, it looks like. Don't expect me to post much.)

His sense of humour is not really to my taste but the book itself is fine. I think his conlangs are quite respectable efforts and while I think the book could have been organised a bit better, the topics it covers make sense. I would have appreciated more references to real-world languages but these can be hard for some reasons (reference/source citation for example). I found the part about font creation the most useful.

I used to dislike Dothraki as being "too close to IE", but I have changed my mind on that. I think DJP was limited by the source material in terms of where the grammar could go and if you investigate it it's really not close at all to any specific IE language.

I think it is a good book. Indeed, after reading it, I considered giving up my own plans for a book about conlangs and conlanging and translating David's book instead. However, I decided not to, not because I feel that David's book wasn't worth translating into German, but because I have enough ideas how I could write a different book.

I haven't read the book. My impression is that he's a competent conlanger, but hardly of stellar talent. I guess it's easy to feel it rather unfair that he's profited so much from language creation when many others of equal or greater ability have not been able to.

I have seen better conlangs than his (though High Valyrian is a gem), but it is unfair to compare hired work with a deadline with something the author made for himself and working on it for many years. As far as hired work goes, I don't know anything better; David certainly set the bar higher in conlanging for hire.

Curlyjimsam wrote: ↑Tue Aug 14, 2018 6:56 am I haven't read the book. My impression is that he's a competent conlanger, but hardly of stellar talent. I guess it's easy to feel it rather unfair that he's profited so much from language creation when many others of equal or greater ability have not been able to.

I haven't read David's book or looked at Dothraki, but this seems like an odd charge. Assuming this is about Dothraki and Valyrian, people should bear in mind that conlangs for hire are subject to requirements or limitations that personal conlangs are not:

* there may be existing words or sentences the grammar must conform to
* the language must be pronounceable by English-speaking actors
* the client usually doesn't like diacritics or special characters
* ideally it shouldn't confuse non-linguist fans too much

All these things push the creator to somewhat easy, IE-ish languages. (The pronounceability thing can be a real killer. Sometimes, to use mostly-English sounds while not sounding like English, the best thing is to remove a few sounds.)

I've had quite a few clients myself by this time. Some are more adventurous, I'll note. But it's a very real thing for some clients to say, in effect, "This is great! Only... I don't get it."

Marc Okrand had some extra leeway because Klingon was supposed to "be alien". His main approach was to make it very alien to English-speaking fans... by using features that are absolutely commonplace to anyone who knows Native American languages.

I'm not criticizing the clients, here; accessibility is a valid design goal, and you can't demand that the client (or the eventual readers/viewers) take an introductory linguistics course.

And to add to what Zompist said, with the caveat that I have never read ASOFIAF nor watched Game of Thrones, GRRM seems to be absolutely abominable at naming things, which suggests that if there are snippets of language in his work they, too, are probably abominable.

Yeah, to be clear, there are plenty of Dothraki names from GRRM and quite a few sentences that had to be accommodated.

(E.g., khaleesi is from the books, so right away you're faced with either challenging his orthography, or giving it a really weird pronunciation.)

I agree about phonology and orthography; my criticism is of the morphosyntax. The thirty-odd phrases that appear in the books (source) create a few constraints in this regard, but not many.

zompist wrote: ↑Tue Aug 14, 2018 4:57 pm
Marc Okrand had some extra leeway because Klingon was supposed to "be alien". His main approach was to make it very alien to English-speaking fans... by using features that are absolutely commonplace to anyone who knows Native American languages.

And unfortunately, even apart from the actors' bad pronunciations, most of the "Klingon" that actually appears on Star Trek (well, DS9 and TNG) is total gibberish - even when the writers went to the cursory effort of consulting the Klingon Dictionary. It took until Star Trek: Discovery for them to take this more seriously.

To my mind probably the biggest problem in Klingon's construction - aside from the fact that it was evidently too difficult for most of the writers and production staff to bother with - was the way the grammar was developed, ad hoc for each line of dialogue that appeared in Star Trek III. Okrand himself has said he would do things differently if he was starting from scratch. I suspect the real reason Klingon doesn't have a word for "which" or a straightforward way of creating instrumental phrases is that (rather than a way of it being alien) these things did not come up in the first movie script.

Anyway, to go to Curlyjimsam's point, of course I'm jealous of the opportunity that DJP got, but it's always going to be hard for a conlanger who doesn't live in southern California to have their work appear in a major film or TV production. Certainly he had to demonstrate greater ability as a conlanger than most other people who get paid to design languages for major productions. Okrand got the job of creating Klingon because he happened to know the right people in Hollywood.

The point with art, set design, props and costumes for movies and TV will always be to look and sound convincing, evoking the right mood, etc, not technical excellence per se. You could definitely have designed Dothraki differently, but whether it would have worked "better" either as a language or as something that fitted the purposes of the production is hard to say. What is the "best" possible way of framing Dothraki? As identical to Mongol as possible? I honestly don't know.

And in any case, conlanging for the purposes of having someone on screen recite your lines I think might be less exciting than it sounds at first. Most likely you're going to be dealing with actors flubbing the lines constantly or (as happened with Star Trek - Into Darkness) a scene in a conlang being edited in a way that makes the dialogue nonsensical and having to go back and retrofit everything. I understand this has happened with Dothraki as well, a few bits of improvising by actors and writers leading to actually ungrammatical stuff appearing onscreen.

zompist wrote: ↑Tue Aug 14, 2018 10:32 pm Yeah, to be clear, there are plenty of Dothraki names from GRRM and quite a few sentences that had to be accommodated.

(E.g., khaleesi is from the books, so right away you're faced with either challenging his orthography, or giving it a really weird pronunciation.)

Well sure, but I have some issues with the decisions he took even based on that. My main issue is probably with the absence of labial stops in an otherwise very IE-like system, a gap which he tried to justify by saying that "they only occur in the name corpus", when by that logic half the consonants should be missing from his inventory. And imho opinion you could get away with quite a bit even within those confines (and given DJP seems to expect the actors to be able to pronounce a uvular stop with both High Valyrian and Dothraki it's safe to say he's not above those sort of things), hence why I proposed a Kabyle-style stop-fricative variation pattern when I did my own fan take on Dothraki phonology.

So Haleza Grise wrote: ↑Wed Aug 15, 2018 5:12 amAnyway, to go to Curlyjimsam's point, of course I'm jealous of the opportunity that DJP got, but it's always going to be hard for a conlanger who doesn't live in southern California to have their work appear in a major film or TV production. Certainly he had to demonstrate greater ability as a conlanger than most other people who get paid to design languages for major productions. Okrand got the job of creating Klingon because he happened to know the right people in Hollywood.

Really? I'd like to know more about how DJP actually got the job, do you have any articles on this?

The point with art, set design, props and costumes for movies and TV will always be to look and sound convincing, evoking the right mood, etc, not technical excellence per se. You could definitely have designed Dothraki differently, but whether it would have worked "better" either as a language or as something that fitted the purposes of the production is hard to say. What is the "best" possible way of framing Dothraki? As identical to Mongol as possible? I honestly don't know.

Well Mongol is pretty much out of the window, because it is clear to me at least that outside of the archetype of "steppe nomad" George R R Martin clearly knew nothing about actual Mongols; not language nor manner of dress nor much else about the culture really. If anything I'd say the Dothraki most make me think of Scythians more than anything else (which does make the IE-ness of the language much more reasonable).

TBH, it triggers my inner rantor when people complain about the "IE-ness" of languages. IE languages cover quite a typological range, some have split ergativity, some are VSO or strict SOV, it's quite likely that PIE or its preceding stage had active alignment...

@Frislander: Apparently there was a contest (I dimly remember a call for participants on the CONLANG list), and David won it. That said, David's conlangs are far better than Klingon - which is actually a good demonstration how not to create a conlang. It's one of the worst conlangs I have ever seen, worse even than Volapük. This shows that hiring a linguist to make a conlang for a film is like hiring a musicologist to write the score; as there are musicologists who can't compose, there are linguists who can't make a decent conlang. Hence, they don't have musicologists write the score, but composers. Of course, the snag is that there are no university degrees in conlanging, so the only way to find out that someone is a good conlanger is to look at his work. (David has a university degree in linguistics, though, but that not really was the relevant qualification. It was what he submitted in the contest.)

@hwhatting: Very true what you say about the typological diversity of Indo-European. Most people who say "IE" when it comes to typology of conlangs in fact mean "SAE" (Standard Average European), which in itself is poorly defined and controversial.

WeepingElf wrote: ↑Wed Aug 15, 2018 9:27 am@hwhatting: Very true what you say about the typological diversity of Indo-European. Most people who say "IE" when it comes to typology of conlangs in fact mean "SAE" (Standard Average European), which in itself is poorly defined and controversial.

True enough, in fact I'd probably like it best if Dothraki actually looked like Scythian, or even better a modern Iranian language, cause they're pretty out there even compared to other IE languages (or at least from what little I know of them).

Frislander wrote: ↑Wed Aug 15, 2018 2:06 pm

WeepingElf wrote: ↑Wed Aug 15, 2018 9:27 am@hwhatting: Very true what you say about the typological diversity of Indo-European. Most people who say "IE" when it comes to typology of conlangs in fact mean "SAE" (Standard Average European), which in itself is poorly defined and controversial.

True enough, in fact I'd probably like it best if Dothraki actually looked like Scythian, or even better a modern Iranian language, cause they're pretty out there even compared to other IE languages (or at least from what little I know of them).

I am like you in this respect. If I was to make a conlang for a given conculture, I'd look which historical culture comes closest and build a conlang resembling their language. So Dothraki would be either like Scythian, as you say, or like a Turkic or Mongolic language, perhaps a bit of both. Valyrian would resemble Greek to some degree - and David's Valyrian indeed does.

Of course one can say that similar culture does not necessarily imply similar language. That's true. But we are dealing with an artistic decision here, and what counts is what feels right. And to me, a language reminiscent of that of the closest Earthly equivalent does feel right. (By extension, I'd make something similar to Quenya or Sindarin, dependent on how lofty they are, for a culture of Tolkien-derived Elves. Indeed, my main conlang Old Albic owes quite something to those, even if these Elves are "merely" humans, though clearly inspired by Tolkien's Elves.)

Frislander wrote: ↑Wed Aug 15, 2018 6:42 amReally? I'd like to know more about how DJP actually got the job, do you have any articles on this?

The LCS was asked to create Dothraki, and had a contest to create it. I'm pretty sure it was advertised on the ZBB, because I submitted an entry.

Well Mongol is pretty much out of the window, because it is clear to me at least that outside of the archetype of "steppe nomad" George R R Martin clearly knew nothing about actual Mongols; not language nor manner of dress nor much else about the culture really.

That's for sure. What really bugs me is that Martin imagined that nomads would be super-ultra-macho and misogynist. Quite the opposite— they were better for women than the agricultural states around them.

WeepingElf wrote: ↑Wed Aug 15, 2018 9:27 amThat said, David's conlangs are far better than Klingon - which is actually a good demonstration how not to create a conlang. It's one of the worst conlangs I have ever seen, worse even than Volapük.

Wow, I disagree on both counts. Too bad Jay Shorten isn't here anymore to defend Volapük. It's kind of a Victorian monstrosity, yes, but that's what gives it its charm, unlike the streamlined but incoherent Esperanto.

Overall, I like Klingon. I can see Okrand's background and humor in it— plenty of ideas from Native American languages, a cheeky use of OVS, some nice Whorfian jokes. The agglutinative structure is the best way of getting complex ideas into a conlang without the difficulties of fusional inflection.

I don't like the phonology and orthography. For the first he was apparently trying for "miscellaneous", which is a pretty cheap way of making something alien. (Ideally, aliens should seem bizarre to us in some ways, but they should have an underlying logic of their own.) The orthography is a clever way of avoiding special characters (it was done before Unicode, so these would have been a big ask). But it's ugly and confusing, and it would be easy to redo it without the capitals.

The real problem is the scantiness of the grammar. It leaves out too much— indeed, the first edition didn't even tell you how to ask questions. It lacks glosses and a glossed sample text. Probably neither Okrand nor his publishers thought people would buy it except as a bit of merch.

zompist wrote: ↑Wed Aug 15, 2018 4:35 pm

WeepingElf wrote: ↑Wed Aug 15, 2018 9:27 amThat said, David's conlangs are far better than Klingon - which is actually a good demonstration how not to create a conlang. It's one of the worst conlangs I have ever seen, worse even than Volapük.

I don't like the phonology and orthography. For the first he was apparently trying for "miscellaneous", which is a pretty cheap way of making something alien. (Ideally, aliens should seem bizarre to us in some ways, but they should have an underlying logic of their own.) The orthography is a clever way of avoiding special characters (it was done before Unicode, so these would have been a big ask). But it's ugly and confusing, and it would be easy to redo it without the capitals.

agreed, although I don't mind the mixed case (except the stupid I (eye) vs. l (el) bullshit). i don't think it's that ugly or confusing; it's still pretty easy to learn and understand.