anonymous asked:

I'm a Linguistics student at your almamater and my non-Linguistics student boyfriend, as lovely as he is, seems to be a bit misguided in the fact that he is absolutely convinced that it is possible to one day invent a language which completely eliminates all ambiguity. I pointed him in the direction of pragmatics, arguing that no invented language could eliminate something like pragmatic ambiguity, but he seems absolutely convinced that it could and *should* be eliminated. What do you think?

If we were really better off without ambiguity in our languages, we wouldn’t have it. But you should ask him to define ambiguity first. I gather he probably wouldn’t distinguish between the different types of ambiguity—e.g. syntactic vs. lexical. But if you go down that rabbit hole, you’ll quickly find there’s no bottom. By that I mean we couldn’t really have lexical items if we wanted to be certain that we only ever referred to unique objects, people, or processes. Take a word like “man”. Terribly vague. It could be anybody that fills a certain number of characteristics (and these characteristics, incidentally, are not anything that can be absolutely determined. Not all people agree what those characteristics are). So a sentence like “I talked to a man yesterday” is specific for me, in that I’m referring to some specific man I talked to, in my estimation, but the word is not. The word is ambiguous. If you go that route, you may as well get rid of all human nouns and replace them all with unique identifiers (note that names aren’t unique enough for this).

And then moving on to objects, consider that all objects are situated in space and time. Even if you’re talking about a ball bearing manufactured in the best ball bearing factory in the world which ensures that every single ball bearing is identical as possible, each one will only ever occupy one specific place in time and space. So they’re all unique, and each will require its own unique designation—with a time stamp (recall how different we humans are depending on the time you find us. Like I don’t care how tough you are, me at 20 can beat you up at 1 year old. Boxer? Weightlifter? Soldier? Don’t care. MY 20 YEAR OLD SELF WILL TAKE ON YOUR ONE YEAR OLD SELF AND PUT YOU IN THE GROUND).

And think about actions—like the verb “kick”. We know what it means. But has there ever been any two kicks that were identical?! How imprecise that we just have one word for all these different types of motions! And yet we tolerate it!

And don’t even get me started on modifiers like “hot” or “red” or “acceptable”. Like “acceptable”, really?! Just…no.

So yeah, talking about eliminating ambiguity is misunderstanding the nature of…everything, I guess. You can only eliminate ambiguity within a specific domain of experience. The result probably wouldn’t be too useful, though.

For an attempt at this, see John Quijada’s Ithkuil, which is a great language, but which picks its battles. Even so, though, note that John isn’t pushing this as a language that everyone should learn and use in their daily life. That’s not the point. It’s a philosophical experiment—and a good one—but it’s not supposed to be a chatting language you use as a primary means of communication.

I was reading up on Ithkuil, a language constructed to be brief, precise, and very complicated. The famous Monty Python line “My hovercraft is full of eels” is translated on the webpage as Tî akt’asalb abjatļud. I still don’t know much about the grammar, but after going through the site for a while, I think that literally translates to something like

Part of my hovering transport vehicle, which I legally own, which exists as a discrete whole, whose existence I am not unsure of, and with a natural relationship between its parts, is overwhelmed with a single group made up entirely of eels, whose existence I am sure of, which came together naturally, unidentified listener or listeners.

It’s precise, at least!

How Good Are The Matsuno Brothers At English + Languages They Want To Learn


  • ??? doesn’t know shit. 3/10
  • but maaaaybe wants to learn Korean or something similar to his native language.


  • actualy good omg??
  • can hold a simple conversation at least. 8/10
  • wants to learn something ~cool~ and ~romantic~ so probably French or Spanish.


  • ehh knows words and phrases here and there. 7/10
  • French sounds nice???? or Russian???
  • he probably wants to go for something difficult…. LIKE ARABIC OR ICELANDIC
  •  ur rising choro pls stop


  • doesn’t even bother. 2/10
  • if he had to choose a language it’d be German.
  • or Swedish. something like that.


  • nah. 1/10
  • he knows how to say ‘baseball’ in 12 different languages tho
  • either this or is secretly fluent in 40 languages.


  • fucking fluent and nobody can tell me otherwise. 9.5/10
  • already knows Korean, French, Norwegian, and is starting Swahili soon.
  • either this or knows languages nobody heard of, like Sellkup or Ithkuil.

anonymous asked:

As a deaf person with limited/no context for how words sound, I'm finding the conlang community's fixation on starting with sound to be a huge stumbling block when trying to learn how to make conlangs that aren't signed. Do you have any advice on how to work around that?

Two three actually four things:

(1) “Phonology” doesn’t simply refer to sound. (I mean, etymologically it does, but not in linguistics.) Phonology refers to how a language uses unanalyzable, meaningless units (phonemes) to create larger chunks that do have meaning (words or affixes). In a spoken language, this refers to spoken phonemes (e.g. /p/, /b/, /d/, etc.). In a sign language, this refers to places on the body, motion, and handshape. If you know ASL, compare the sign for APPLE to the sign for ONION. These are almost exactly the same sign. You put your hand in the 1 handshape with the first two segments completed curled under, you put it up to your face and twist a couple times. The difference is whether you put your hand up next to your eye, or up at your cheek. That’s the only difference between those two words. Thus, the difference between APPLE and ONION in ASL is the same difference (quantitatively) as the difference between English “meet” and “mitt”. I share this example to show you how the principle behind the arrangement of both systems—signed and spoken—is the same. They differ in their expressions (i.e. through speech sounds and through movements done with the hands and body in particular places), but the notion is that there are certain things that have no meaning (for example the place next to the eye in ASL), but you can use those things in combination with other bits that have no meaning to form meaningful units.

There aren’t a lot of signed conlangs because many conlangers aren’t as familiar with sign languages as they are spoken, and also because they are very difficult to record on paper. Any hearing person who’s studied ASL will be familiar with this: You go to class and learn, but what do you write down? I imagine every learner kind of tries to invent their own notation system to help them remember, but ultimately you just have to memorize it. That doesn’t really work for someone creating the language, though. Video is the best way to capture a sign language, but it’s not super practical (though it’s getting easier). I tried to create a phonetic transcription system for sign languages called SLIPA. I’m not sure if anyone has used it, but I think the principle is sound (or sound enough). Plus, as with the relationship between narrow transcription and romanization, I think it makes sense for the creator of the language to create a more streamlined system for use with their language that can then be explicated in a page with SLIPA.

(2) There are still other conlang types that make no reference to sound but aren’t signed. I made one called X. It’s a purely visual language (think hieroglyphs but with no phonological component whatsoever). There’s a lot to be done in this area of conlanging. You can go the picture/glyph route I did, or you could just do something totally different, as with Sai and Alex’s UNLWS. You could also do something like this:

  • #$% = cat
  • #$%* = cats
  • ##$% = big cat
  • ##$%* = big cats
  • #$$% = small cat
  • #$$%* = small cats

After all, even letters are just symbols. They can stand for whatever you want, or nothing at all! As long as you can describe what’s going on, that’s all that’s necessary.

(3) To your main concern, saying “the conlang community’s fixation on starting with sound” is, to put it mildly, unfair. I start with the sound system in the spoken conlangs I do, and I mostly do spoken conlangs. If I’m writing a book on how to create a language, that’s where I start, because I’m writing it. But just because I do it that way doesn’t mean most do. Even if you go to a forum or mailing list and you see most people falling into that pattern, that doesn’t mean that’s representative of the community either, because there are any number of people who simply aren’t replying or aren’t volunteering their methods. We’ve had the discussion within the community time and time again about where one starts a conlang, and there’s a significant chunk that start with the syntax. They’ll use English words or just nonce forms to realize the grammatical idea they’re interested in, and only grudgingly turn to the phonology after they’re done. Some don’t even get that far, because their interest wanes when it comes to phonology. This split even exists in linguistics, where we refer to P people (phonetics/phonology) and S people (syntax/semantics). Ask any linguist: these camps don’t always understand one another. The same is true in conlanging. A P person sees an S person’s awesome subordinate clause marking system with a makeshift phonology and says, “Is that your phonology? It’s a little unrealistic.” An S person looks at a P person’s incredible naturalistic vowel harmony system and says, “Why waste your time on that if you’re not even going to speak it? It obscures the morphology. I can’t make heads or tails of it. Just show me the interlinear.” And these are all hearing conlangers! If you’re only finding conlangers who are talking about phonology, then you need to look for other conlangers—like the Jeff Jones and Gary Shannon type of conlangers. This was, admittedly, an easier task when the community was smaller—before social media, ironically. But I swear to you: There are TONS of conlangers who share your interests.

(4) There are also lots of spoken conlangs that don’t bother too much about phonology. There are minimalist conlangs which, by definition, don’t really have a lot of material to work with, so there’s not much to design/learn in the way of phonology (e.g. three vowels, seven or eight consonants, no consonant clusters). There are also a priori auxlangs or otherwise non-natural spoken languages where you don’t find assimilation or dissimilation, or anything like that. If there are five vowels and ten consonants with ©V© syllables, then there are 555 possible syllables (if I counted right), and every syllable is valid and pronounced exactly as it’s supposed to be, and can occur next to any other syllable. Then there are other conlangs with complex but non-natural phonologies, where there are many distinctions to be made (many of which wouldn’t exist in a natural language), and the speaker must make them. I’m thinking of Ithkuil. There, is admittedly, some small amount of variation is allowed, but otherwise the way that sounds are arranged is almost mathematical. There is no concern for how the sounds fit together, or whether two words sound too similar: The grammar says what sounds go where, and that is that. Any type of project like I’ve described above incorporates aural phonology but in a way that I think makes a little more sense to an S person.


If you have a particular project in mind but the approaches you’ve seen don’t match it, do a little digging and find a similar project, and see how they got going. If you have the start of something and want to know about a similar project, just send another ask, and I’ll see what I can find (most of the early conlangers have websites that are still up. Btw, newer conlangers: Even though no one does websites anymore, we need a way to see your work! Hunting through Tumblr posts/tweets/FB group posts doesn’t work!). But if I can add a tl;dr to this: THERE IS NO RIGHT OR WRONG WAY TO CONLANG! You’re good, I swear! And hey, if no one can approach the way you want, why not invent it yourself and detail it? You may be creating a method/approach that will be a great help to others down the line!

List of Languages in The Art of Language Invention

Below is a list of all the languages, created and natural, either used or mentioned in The Art of Language Invention. Some (like Arabic) are used a lot; some are just mentioned. Either way, this is all of them (conlangs marked with an asterisk)! The Art of Language Invention comes out September 29th. (Preorder here!)

  • !Xóõ
  • Afrihili*
  • American Sign Language
  • ámman îar*
  • Arabic
  • Astapori Valyrian*
  • Ayeri*
  • Balaibalan*
  • Brithenig*
  • Castithan*
  • Chamorro
  • Cherokee
  • Da Mätz se Basa*
  • D’ni*
  • Dothraki*
  • Dyirbal
  • Egyptian (Middle Egyptian)
  • Esperanto*
  • Finnish
  • Fith*
  • French
  • Georgian
  • German
  • Greek
  • Gweydr*
  • Hausa
  • Hebrew
  • High Valyrian*
  • Hindi
  • Hungarian
  • Icelandic
  • Idrani*
  • Indojisnen*
  • Inuktitut
  • Irathient*
  • Ithkuil*
  • Japanese
  • Kamakawi*
  • Kēlen*
  • Khangaþyagon*
  • Klingon*
  • Kobaïan*
  • Latin
  • Linear B
  • Lishepus*
  • Mandarin
  • Minza*
  • Mongolian
  • Moro
  • Moten*
  • Na’vi*
  • Njaama*
  • Novial*
  • Okuna*
  • Paku/Pakuni*
  • Phoenician
  • Portuguese
  • Quenya*
  • Rikchik*
  • Ro*
  • Rotokas
  • Russian
  • Sanskrit
  • Shiväisith*
  • Sindarin*
  • Skerre*
  • Slovianski*
  • Sondiv*
  • Southern Paiute
  • Spanish
  • Swahili
  • Tagalog
  • Tamil
  • Teonaht*
  • Thai
  • Thhtmaa*
  • Tsez
  • Tukang Besi
  • Tundra Nenets
  • Turkish
  • Ubese*
  • Urdu
  • Væyne Zaanics*
  • Vai
  • Valdyan*
  • Volapük*
  • Yi

Man, some of my favorites didn’t make it in! (Tok Pisin, Sodemadu, mërèchi, ea-luna, Akkadian—HAWAIIAN?!?! That can’t be right… Let me look up my draft. Oh, duh, no, of course there’s Hawaiian. I guess it just didn’t make it into the index… Oops! I guess Zhyler also got a brief mention…) Maybe if a second edition comes out some day. As it is, that’s at least 87 languages (89 if you count the two I just found here—probably more than that), so that’s pretty good.

Oh, no wait, it’s 90. Forgot one: English. lol English counts! Some weird ass stuff in English…

Note: Also, Valdyan is now called Ilaini! My bad!

anonymous asked:

Who are the most famous colangers after you and Tolkien?

Conlangers. And I guess it depends if you mean “someone who’s created a language” or “someone who creates languages regularly”. Not many of the latter are famous. But there are plenty of well-known people who are famous either for creating a language, or are famous for some other reason and also create languages. For example, John Wilkins is famous as a philosopher, and also created a language. L. L. Zamenhof has created the best known language, Esperanto, though I gather the language is more famous than he is. Marc Okrand has created languages for Star Trek (Klingon and some others) as well as Disney’s Atlantis (Atlantean). David Salo created the language used in the Hobbit trilogy; Christine Schreyer created the Kryptonian language for Man of Steel; Paul Frommer created the Na’vi language for Avatar. Hildegard von Bingen is a famous saint, but also is the first recorded language creator we know of (her conlanging is so little discussed, though, that in the movie about her life there was absolutely no mention of it). C. S. Lewis was a friend of Tolkien’s, and tried to create languages a little bit due to his influence, but didn’t really do a proper job of it. John Quijada’s Ithkuil is one of the best created languages ever, but was unknown until it was written up in an article in The New Yorker. So it really depends how you mean the question. Those are some, though!