Archive for December, 2007

Syntax 101

Monday, December 17th, 2007
So I wanted to pick up where I left off on the last post before Christmas, and talk about syntax and grammer a bit.

"Fa meshsak sosha o kulntht to tisiks afshra pefsi."
This pancake is going from zero to sixty within five seconds.
Or, more literally, This pancake is going from stop until sixty within five.

We're going to break this sentence down and figure out how you can construct grammer and syntax. Syntax could be defined simply as how words are ordered in a sentence. In English, we describe nouns with adjectives, or tell what the noun is, like so: "the bright room," "the room was bright," "the smooth, round ball," "the ball was smooth and round." In the case of "the room was bright," the room is the subject, was is the verb, and the brightness is the object. English is a SVO language, meaning that the ordering of the subject, verb and object are respectively, first, second, and third. In an SOV language, the sentence would be, "The room bright was." And its not hard to rearrange the syntax into all six possible combinations and see how the sentence changes. A fine article is here, by Rick Murneau, goes into many of the finer points here, and I highly recommend it.

This sentence is SVO, just like English. If it was SOV, it would be more like:
"Fa meshsak o kulntht to tisiks afshra pefsi sosha."
This pancake from zero to sixty within five seconds is going.

Also, this has a modifier-head format, where the modifiers come before what they are modifying, also, just like English. Lets change the SOV example into a head-modifer format:
"Meshsakfa kulntht to tisiks o pefsi afshra sosha."
Pancake(the) zero to sixty from five seconds within is going.

Now, I don't know about you, but the "is going" part of that last sentence doesn't sound right to me. It sounds like it should say "will go," or "will be going." And here is another issue to think about: tenses and cases. Tenses and cases DROVE ME CRAZY the first 100 times I thought about them, but again, the solution was just to simplify. For this mock-up lang we are playing with, lets create three tenses and three cases: future tense, present tense, past tense, and the tense case (which we just divided into three tenses), the descriptive case (adjectives, adverbs, etc.) and the plural case. We'll show what case or tense a word by adding a vowel sound to the end of a word: future tense= e (ey), present tense=a(au), past tense=o, plural case=i(ee), descriptive case=u(oo). So if we wanted to change sosha from "is going" (present tense) to "will go" (future tense), it would become soshe. What if we pluralized it? Soshi: what would that mean? Goings-on? Walks? Journeys? You decide in the end; whatever makes the most sense for you, and fits into the pattern of your conlang best.

So, our sentence has changed quite a bit from the beginning of the post.
From: "Fa meshsak sosha o kulntht to tisiks afshra pefsi."
To: "Meshsakfa kulntht to tisiks o pefsi afshra soshe."

Take a moment and just play. Look at my previous post called "Phonotactics," create your own SIMPLE phonotactics system, and create a few words and a simple sentence. Then play with your syntax and grammer.

More Phonotactics

Friday, December 14th, 2007
After reading Rick Morneau's wonderful summary of morphology for the umpteenth time, I thought I should write a post, in my words, about the relationship between phonology and morphology, or phonotactics. I think once this relationship is understood better, it makes your conlanging more enjoyable and quicker.

A quick and dirty definition of phonology is that it is the sounds permitted in your conlang. Anything not in your phonology, speakers of that language would have a hard time saying (kind of like how Japanese are famous for speaking Ls like Rs). Lets break down the phonemes of your language into a few categories: consonants, clusters, vowels and semi-vowels. Just these four categories, for now. In fact, lets make up a phonology for the purposes of this post. P, t, k, f, th, s, sh, m, n, r and l for consonants. Ee, ei, au, oo, and o for vowels. 11 consonants, 5 vowels.

Now, phonotactics. Lets keep explanations, and these phonotactic rules, simple. The phonology should include how consonants, vowels, semi-vowels, diphthongs and clusters can or cannot be ordered within a word. C= p, t, k, f, th, s, sh, m, n, r, l. V= i, e, a, u, o (but pronounced the way I spelled them above). S= ... hmm, we didn't specify any semi-vowels in our phonology did we? Let's say that r is a consonant but ALSO a semi-vowel. S= r. As for diphthongs, in some morphologies, you might be limited as to which vowels can be put next to which others, but to keep things simple and neat, we'll just say any of our vowels can be paired to form a diphthong; D= V1V2 (subscript added to show that a diphthong is not two of the same vowel). Now, what types of clusters do we want? I'm going to say that we are having only ending clusters in this morphology, but we'll make them moderately complex for fun: K=[L][N][F]. The brackets mean there may or may not be one of the indicated phoneme, and L means liquid, N means nasal, F means fricative, and P means plosive. So an ending cluster will have either a liquid, a nasal, a fricative, or combinations of these, but not a plosive.

So how can these phonemes be combined? Again, let's keep it simple: a basic word will be [C][S]V[K][C]. So you can have a word be simply a vowel, like "o" (let's say that o means "from"), or basic like "sosh" (lets say sosh means "go"), all the way up to "kulntht" (and lets say that kulntht means "stop"), where we have an ending cluster with a plosive at the end! Ok, I just think thats outrageous and hard to pronounce, but fun. A few more examples: frith (remember, its pronounced "freeth" and lets says that is means "bird") is a word this morphology could make, but wriths is not. A) because w is not part of the phonology, and B) because an ending cluster cannot be just a fricative (th) and a fricative (s). If we had spelled wriths like the English word, wreaths (you know, those things everyone puts on their doors at Christmas), it would also be unacceptable because, although we technically allowed any vowel to be next to any other vowel to make a diphthong, we didn't include any diphthongs in the morphology we defined above. In order to allow a word like wriths, or wreaths, we could redefine the morphology to include FF clusters, and perhaps redefine the phonology to include w, although we might just forego that and spell it riths instead, OR we could say that some words follow another, separate morphology from the one we already created, and it looks like this: [C]V[K2], and define the second cluster type as being FF. With this second morphology defined, we can work out words like "afs" (means "in") or "meshth" (means "flat"), which we couldn't with only the first morphology.

Lets throw in one more twist before this post is done: prefixes and suffixes. In your morphology you can also make special definitions for how these are constructed, or adapted out of existing words. So lets define that, in this limited conlang, we can have ONLY prefixes (SF= 00), and that there are two morphologies for them: CV-, or you can take a [C]V[K2] word and shave off the last F in the cluster to make it a prefix. Not sure how I would notate that, like I've been trying to make short notation on everything else, but maybe something like this: [C]V[K2]-/[K2]=F1F2/=F1. I dunno. But lets say we want to make the word "pancake" and decide to translate it as "flatcake;" the word for "cake" is sak, so "flat-cake" would be meshsak, because we shave off the th at the end of meshth.

A CV prefix could be something like "po-" (means "more") or "she-" (means "without") so that when the prefix is added to a word, it changes the meaning. Pososh could change the meaning of "go" into a command form, like "Go!" Or it could mean "go quickly." But if pososh meant "go quickly," what would pokulntht mean? Stop quickly? Maybe the prefix could mean both things, and its just defined by the context. This is starting to overlap the arena of grammer at this point, so I'm going to back off for now. You ultimately decide if you like how it flows, how it sounds. If you don't like it, try tweaking the structure some more. Remember, if you're having trouble, keep it simple, at least at first, to get a good handle on how all these have an effect on each other. Oh, and just for kicks, here's a sentence using most of the words we defined, even though we haven't talked about syntax or grammer:
"Fa meshsak sosha o kulntht to tisiks afshra pefsi."

"This pancake is going from stop to sixty within five seconds." And I'll end on this note, because I don't think this post can get much better than this today!

Simplicity In My Conlang

Tuesday, December 11th, 2007
I stated in a previous post that there were two main things I did after the second LCC that helped me really firm up my first conlang. The first was digging into the conlang card game. The second was I decided to try and make things REALLY simple. I thought, "I don't really want it to be this simple, but I'm just going to experiment and see what happens." Here's a few other things that lead me in the direction I took:

I had previously found this page on Huttese, the language of Jabba the Hutt and Tatooine from Star Wars. I liked the sound of it. I wondered how I might make my conlang sound more like it, but with my phonology (the one that had TONS of phonemes, remember?). I realized after studying it for a week or two that what I really liked about it was the open syllable structure. "Tolpa da ponki nu puti cha naga." It just sounded right; it sounded good to my ears.

So I started pulling out phonemes and making sure that I had mostly open syllables, meaning CVCV (consonant-vowel-consonant-vowel) and not CVCCVC. Or, an even better example might be to contrast the Huttese phrase above with something, say, in English: "I walked down the road to the Seven Eleven." A lot of closed syllables there. Let's change them to open syllables and see how it sounds; I'll just change the morphology around a bit: "I walko downu roada to SevenElevena." A little better. Then I tried it with a much smaller phonology, shifting the now-extinct phonemes into nearby still-existing phonemes: "Ee waulko taunoo rota to Seifein Eileifeinau." Ooo... Now THAT sounded cool to me.

And that's basically how I developed the basic rules for Pitak, the proto-language to Fauleethik. A very SMALL AND SIMPLE phonology and morphology. There are (so far) only five cases: future, present, and past tense, plural, and "descriptive" case, good for adjectives and adverbs. Each means a different vowel sound tacked onto the end of the syllable. So, the truth is, the syllables by themselves are generally CVC, or CVCVC, or sometimes CVCCVC, in the case of some compounded words, which means they are closed, BUT with the addition of the case markers, it becomes a very open syllable language. Only singular nouns have no vowel at the end.

I felt like I had finally, truly wrapped my head around a lot of the linguistic principles at this point. And I did it by getting REALLY simple; by stripping out a lot of stuff I kind of wanted in there to get something that was simple, but worked. I think a lot of conlangers should try this and make something functional, then start building on it. Instead of 30 phonemes, use half that, to start with at least, then a basic but functional morphology, then add in simple syntax, grammer, develop a basic lexicon, and then start adding in more stuff. Work on the different layers of the language, seeing how each one influences the next, and just keep building.

Simplicity In My Conlang

Tuesday, December 11th, 2007
I stated in a previous post that there were two main things I did after the second LCC that helped me really firm up my first conlang. The first was digging into the conlang card game. The second was I decided to try and make things REALLY simple. I thought, "I don't really want it to be this simple, but I'm just going to experiment and see what happens." Here's a few other things that lead me in the direction I took:

I had previously found this page on Huttese, the language of Jabba the Hutt and Tatooine from Star Wars. I liked the sound of it. I wondered how I might make my conlang sound more like it, but with my phonology (the one that had TONS of phonemes, remember?). I realized after studying it for a week or two that what I really liked about it was the open syllable structure. "Tolpa da ponki nu puti cha naga." It just sounded right; it sounded good to my ears.

So I started pulling out phonemes and making sure that I had mostly open syllables, meaning CVCV (consonant-vowel-consonant-vowel) and not CVCCVC. Or, an even better example might be to contrast the Huttese phrase above with something, say, in English: "I walked down the road to the Seven Eleven." A lot of closed syllables there. Let's change them to open syllables and see how it sounds; I'll just change the morphology around a bit: "I walko downu roada to SevenElevena." A little better. Then I tried it with a much smaller phonology, shifting the now-extinct phonemes into nearby still-existing phonemes: "Ee waulko taunoo rota to Seifein Eileifeinau." Ooo... Now THAT sounded cool to me.

And that's basically how I developed the basic rules for Pitak, the proto-language to Fauleethik. A very SMALL AND SIMPLE phonology and morphology. There are (so far) only five cases: future, present, and past tense, plural, and "descriptive" case, good for adjectives and adverbs. Each means a different vowel sound tacked onto the end of the syllable. So, the truth is, the syllables by themselves are generally CVC, or CVCVC, or sometimes CVCCVC, in the case of some compounded words, which means they are closed, BUT with the addition of the case markers, it becomes a very open syllable language. Only singular nouns have no vowel at the end.

I felt like I had finally, truly wrapped my head around a lot of the linguistic principles at this point. And I did it by getting REALLY simple; by stripping out a lot of stuff I kind of wanted in there to get something that was simple, but worked. I think a lot of conlangers should try this and make something functional, then start building on it. Instead of 30 phonemes, use half that, to start with at least, then a basic but functional morphology, then add in simple syntax, grammer, develop a basic lexicon, and then start adding in more stuff. Work on the different layers of the language, seeing how each one influences the next, and just keep building.

Simplicity In My Conlang

Tuesday, December 11th, 2007
I stated in a previous post that there were two main things I did after the second LCC that helped me really firm up my first conlang. The first was digging into the conlang card game. The second was I decided to try and make things REALLY simple. I thought, "I don't really want it to be this simple, but I'm just going to experiment and see what happens." Here's a few other things that lead me in the direction I took:

I had previously found this page on Huttese, the language of Jabba the Hutt and Tatooine from Star Wars. I liked the sound of it. I wondered how I might make my conlang sound more like it, but with my phonology (the one that had TONS of phonemes, remember?). I realized after studying it for a week or two that what I really liked about it was the open syllable structure. "Tolpa da ponki nu puti cha naga." It just sounded right; it sounded good to my ears.

So I started pulling out phonemes and making sure that I had mostly open syllables, meaning CVCV (consonant-vowel-consonant-vowel) and not CVCCVC. Or, an even better example might be to contrast the Huttese phrase above with something, say, in English: "I walked down the road to the Seven Eleven." A lot of closed syllables there. Let's change them to open syllables and see how it sounds; I'll just change the morphology around a bit: "I walko downu roada to SevenElevena." A little better. Then I tried it with a much smaller phonology, shifting the now-extinct phonemes into nearby still-existing phonemes: "Ee waulko taunoo rota to Seifein Eileifeinau." Ooo... Now THAT sounded cool to me.

And that's basically how I developed the basic rules for Pitak, the proto-language to Fauleethik. A very SMALL AND SIMPLE phonology and morphology. There are (so far) only five cases: future, present, and past tense, plural, and "descriptive" case, good for adjectives and adverbs. Each means a different vowel sound tacked onto the end of the syllable. So, the truth is, the syllables by themselves are generally CVC, or CVCVC, or sometimes CVCCVC, in the case of some compounded words, which means they are closed, BUT with the addition of the case markers, it becomes a very open syllable language. Only singular nouns have no vowel at the end.

I felt like I had finally, truly wrapped my head around a lot of the linguistic principles at this point. And I did it by getting REALLY simple; by stripping out a lot of stuff I kind of wanted in there to get something that was simple, but worked. I think a lot of conlangers should try this and make something functional, then start building on it. Instead of 30 phonemes, use half that, to start with at least, then a basic but functional morphology, then add in simple syntax, grammer, develop a basic lexicon, and then start adding in more stuff. Work on the different layers of the language, seeing how each one influences the next, and just keep building.

Make A Lang Card Game – Part Two!

Friday, December 7th, 2007
With it being the holiday season, I've been pretty busy, so I'm sorry about the lack of posts recently. But wow, I got a lot of response to the card game post, so I wanted to post more about it! I thought that today I would write some more about the additional functionality I am building into the game.

I am designing the game to be playable with a group but also for solo play. With solo play, I figured you could use it to generate a random language, or you could put up the cards that would mostly describe your own conlang, and then be able to play around with the language by substituting, adding, or taking away some of the cards. Also, you could more easily understand other conlangs by putting up the cards that create that conlang. After doing this a few times, you would sense patterns between the cards for how certain languages sound and behave. If you an amateur linguist (like me), developing a recognition for these patterns could be really helpful to wrapping your head around and learning linguistic principles.

In addition to all this, after I'm done making all the cards, I want to make a flash game which you can play, and make it even easier to play with language, being able to click and drag cards around and such. Also, I want to make it so that you can develop a conlang in the game, and have it spit out a long code which you can copy and paste. Then you could post about your language online, include the code, and others would be able to paste the code into their flash game and see your conlang pop up on the screen! I think could work much more quickly than having to write out your entire phonology, morphology, syntax, grammer, etc. Especially if the code is in a file that can include a lexicon and notes on the language. It could make Conlang Relays a smoother process and more fun! But thats probably a couple years away still.

Please comment if you have any helpful ideas or if you think this can't be done. I'm interested to hear your opinion either way!

Make A Lang Card Game – Part Two!

Friday, December 7th, 2007
With it being the holiday season, I've been pretty busy, so I'm sorry about the lack of posts recently. But wow, I got a lot of response to the card game post, so I wanted to post more about it! I thought that today I would write some more about the additional functionality I am building into the game.

I am designing the game to be playable with a group but also for solo play. With solo play, I figured you could use it to generate a random language, or you could put up the cards that would mostly describe your own conlang, and then be able to play around with the language by substituting, adding, or taking away some of the cards. Also, you could more easily understand other conlangs by putting up the cards that create that conlang. After doing this a few times, you would sense patterns between the cards for how certain languages sound and behave. If you an amateur linguist (like me), developing a recognition for these patterns could be really helpful to wrapping your head around and learning linguistic principles.

In addition to all this, after I'm done making all the cards, I want to make a flash game which you can play, and make it even easier to play with language, being able to click and drag cards around and such. Also, I want to make it so that you can develop a conlang in the game, and have it spit out a long code which you can copy and paste. Then you could post about your language online, include the code, and others would be able to paste the code into their flash game and see your conlang pop up on the screen! I think could work much more quickly than having to write out your entire phonology, morphology, syntax, grammer, etc. Especially if the code is in a file that can include a lexicon and notes on the language. It could make Conlang Relays a smoother process and more fun! But thats probably a couple years away still.

Please comment if you have any helpful ideas or if you think this can't be done. I'm interested to hear your opinion either way!

Phonoaesthetic Considerations

Tuesday, December 4th, 2007






























Here is the handout made by John Quijada for his talk at the 2nd LCC:

If you click on the image, it will take you to my Flickr account and you can see the full-size version; even print it out if you like.

It seemed like such a basic idea when I first heard it, but it really is at the heart of making a language: how do you want your language to sound? And even deeper than that, what character does your language have?

There was a word John showed us: sprachgefuhl. It means "the character of a language," among other things. We all know that French is a pretty soft language, the "language of love," and so forth. We know that Japanese, German, and Norwegian sound completely different from each other. These languages have a very different "feel" from each other. How will your language feel different? Or do you want it to feel different? Maybe you want your conlang to feel similar to some other language. What characteristics will your language have that will make it unique? Or what are the unique characteristics of the language you want to emulate? On this handout, John Quijada remarks on how Quenya (Tolkien's elvish language) sound nothing like Polynesian or African Bantu languages, despite having similar phonemic inventories. Interesting, huh?

There are a lot of options on this sheet that could go over your head if you look at them all at once. I suggest a strategy or exercise (whichever you prefer): consider each item individually, write out your selection on a page somewhere, then look at your answers and try to think out how this language might look or sound. Generate some words and put together a few sentences. Then go through it again with a new page and change a few things. See how it changes and compares to the last page. Try this a few times and get a feel for how the pieces of language fit together and change the overall picture of a language.

Phonoaesthetic Considerations

Tuesday, December 4th, 2007






























Here is the handout made by John Quijada for his talk at the 2nd LCC:

If you click on the image, it will take you to my Flickr account and you can see the full-size version; even print it out if you like.

It seemed like such a basic idea when I first heard it, but it really is at the heart of making a language: how do you want your language to sound? And even deeper than that, what character does your language have?

There was a word John showed us: sprachgefuhl. It means "the character of a language," among other things. We all know that French is a pretty soft language, the "language of love," and so forth. We know that Japanese, German, and Norwegian sound completely different from each other. These languages have a very different "feel" from each other. How will your language feel different? Or do you want it to feel different? Maybe you want your conlang to feel similar to some other language. What characteristics will your language have that will make it unique? Or what are the unique characteristics of the language you want to emulate? On this handout, John Quijada remarks on how Quenya (Tolkien's elvish language) sound nothing like Polynesian or African Bantu languages, despite having similar phonemic inventories. Interesting, huh?

There are a lot of options on this sheet that could go over your head if you look at them all at once. I suggest a strategy or exercise (whichever you prefer): consider each item individually, write out your selection on a page somewhere, then look at your answers and try to think out how this language might look or sound. Generate some words and put together a few sentences. Then go through it again with a new page and change a few things. See how it changes and compares to the last page. Try this a few times and get a feel for how the pieces of language fit together and change the overall picture of a language.