Orthography Evolution

I've posted about this before and made more progress, so this is an updated report about how my orthography has evolved since I made three posts about it at the end of '07, and things you might consider as you develop your own orthography.

I just couldn't leave well enough alone. I had my orthography, it worked, but each time I looked at it, there was still something that bothered me - ome little nagging itch in the back of my brain somewhere. So one day as I was sitting in a meeting, I starting listening to my itch to see what where it lead. Now these are little things, but maybe something I learned will benefit you, too. So here is the alphabet I had settled on previously:

This font, you might notice, is extremely light compared to the English letters. I realized quickly that I should make the letters thicker, but this change could wait. An item of interest: the Pitak characters have thicker horizontal strokes than vertical strokes. I thought it would be interesting to see how this looked, since our English alphabet characters are thicker on the vertical strokes (meaning, the sides of an O are thicker than the top and bottom, and the same goes for the other letters). It turns out that I like having thicker vertical strokes, and you'll be able to compare the two a little further down the page, but this kind of experimentation is good.

I realized one thing I wasn't entirely comfortable with was the letter order/arrangement. I designed the Pitak letters to manifest certain phonetic characteristics; for example, the voiced characters have middle strokes, or partial middle strokes if they are a combined sound (ch=t+sh, a combined sound). I had tried to arrange the letters so that the "related" letters flowed together, i.e. the unvoiced plosives were together, the voiced plosives were together, so that when you looked at the alphabet, you could see the relationships. The above letter order wasn't very conducive to that. I made a letter order I felt better about, and re-made the font, with vertical strokes now being thicker than the horizontal strokes, and I liked the way it looked much better:So now, the plosives are all on the left, or on the right, unvoiced or voiced, along with the unvoiced and voiced consonant combinations of ch and j, nestled in between the consonants that make up their sound. Then the second and fourth columns are fricatives, unvoiced and voiced. In the middle, nasals, a liquid, and h, because I wanted it to have a special heritage, if you will, of being a half letter, and that it is used in words for childhood, shortness, and etherealness. Also, semi-vowels are between the consonants and the vowels, showing their mixed heritage, and their letters are combinations of consonant and vowel shapes. Well the W and R are... the Y symbol is a bit of a stretch, to my mind. The ng sound is where it is because... well I was just experimenting, and thought I didn't want a voiced th sound, so substituted the ng sound in, so its close to the n sound, to which it is related.

This letter arrangement I liked much better, and I reworked a few characters to better reflect their new placement and their phonetic qualities. H is now a very minimal letter; only two strokes; this is to reflect the attributes I mentioned above, as well as a minimal effort being required to make this sound. Ch and J characters are better combinations of the letters for t+sh and d+zh.

Now originally, I had 30 characters in the alphabet, which, on a keyboard, still allowed me some keys to make up punctuation. With the new alphabet, now I had 33 characters and I'm getting short on keyboard space. I wanted to have one character for each sound (meaning no sounds that require two letters, such as th, sh, ch), but I don't want to have too many characters either. I started wondering if I should shorten my phonology. I also wanted to improve my font; in comparison with the English characters, my letters look so little - I wanted to beef them up more. So I decided to cut out the ch and j letters, and the ng. I moved the l into the place of the ng, because this column is all voiced consonants, and h isn't voiced (h was the only other consonant I was considering moving). I also moved the postion of w, y, and r to where they match up more closely with the vowels they are close to.
You can see now, in this third iteration, that the characters are bigger, "beefier," and don't seem as small when compared to the English letters, and I think they could be even thicker and look better. When I prepared the earlier alphabet graphics, I actually had to use a bigger font size each time. This time, I didn't. As you create your own font, you'll probably go through similar trial and error, until you know exactly how  thick, how tall, and how wide you need to make your letters. Again, I used High-Logic's FontCreator 5.6 to create this font.

So, this year I decided that my font STILL was not fit for public consumption and opened up FontCreator once again. But this time I was determined to create at least two good, solid fonts: an older, runic style and a newer, modern style. Here's the older, runic style. The graphic is a bit smaller than the others, but the letters are a better match size-wise to the English font, and I think the simplification of the strokes makes it nice, simple, and it still has a bit of a serif on the diagonal strokes to give it a little style.

Now here's the newer, modern style! I thought about what might make the above font look more modern or even a little futuristic, and the thing I kept coming back to was "fewer strokes" to make a character. So I got rid of the "neck" of the characters, making it look like an alphabet of the number seven, kind of. I also experimented with putting in a dot with the shorter vowel sounds, as a riff off of the "Electric Toaster" font which I'm fond of, and because in words, it almost looked like there was a space between some vowels and consonants, and the dot cleared that up, but it still looks weird.

This one I'm happy with, but I could be happier. Plus, I still want to make a cursive, elegant font as well. Every time I've tried one, I've been very lukewarm with the results and never finished it. I may just have to find a friend with a Wacom tablet and borrow it just for this.

Hope some of you got some ideas from this!

New ULD Word List! Link to GoogleDoc included

I posted awhile back about word lists you can use to guide you on creating a vocabulary for your conlang. I was recently looking at this again and I decided to look closer at the Universal Language Dictionary (click here for a good description of what this is). I found some great new stuff!

I downloaded the XML file from Rick Harrison's Universal Language Dictionary website and found that they added "levels" to the words! Now, in each category of words, a level is assigned to the words, from 0 to 3, in order of complexity. There are about 30 0 level words, and about 170 1 level words, so the first two levels is only about 200 words.

You could use this as a guide - creating all the level 0 words first, then the level 1 words, and so on. You might disagree with the level assignments, or you might be working on a conlang that emphasizes certain concepts or objects (for example, that old factoid about how eskimos have nine different words for snow) - so you can reassign those word levels.

Click here to access the Google Doc. You can download and modify this spreadsheet for your own purposes, but if you use this on the web, please give credit to Rick Harrison. He and his team have spent many, many hours on this project.

If you look, you can see that the first tab is the original spreadsheet as I downloaded it from the ULD website.

The second tab is all the categories with the words sorted by level.

The third tab is all the 0 level words, by category, and then all the 1 level words by category.

The fourth tab is just a pivot table showing how many words are at each level.

Enjoy!

How to Make a Conlang out of English

Ok.  My experience has been that some conlangers out there do not like it when your conlang is too... Englishey. This generally means your conlang has basically the same syntax and grammar as English, and the same sounds, too.  There might be a few twists in there - an extra case, some extra phonemes, a different alphabet, but overall, pretty close to English.

And really, who can blame them?  For those that take the time to learn and understand linguistics and all the concepts behind it, it looks and feels lazy and uninspired. For the record, I do not encourage conlanging snobbery, I'm just saying that I understand where it comes from.

But... if you DON'T know lots of linguistics, and don't care to study all the principles and so forth, what else can you do?  If you know a second language you can mash up the two languages you know.  But aside from that, how else can you build a language?

Being the conlang contrarian I am, I think you can transform English into an elegant, simple conlang if you understand at least a few linguistic principles, and because you're using English underneath it, you might be able to make it more sophisticated than you would otherwise be able to.

Here's my formula:


Step One: Restrict (and simultaneously simplify) the phonemes.  If you must, throw in some non-English sounds (like zh, a trill, a click, etc.)

Step Two: Figure out how you will substitute sounds as you translate words from English to your conlang.
Example: lets say the phonemes we picked in step one are P, T, K (P, T, & K are in almost every language), L, N, M, H, J, long and short vowels - I, U, O, and Y as a semi-vowel.
So, lets make a rule that any voiced plosive becomes non-voiced (b=p, d=t, g=k).
E's will become I's, A's will become U's.
Other semi-vowels (W, R) will beome Y.
Other fricatives will become H or be dropped.
Long vowel sounds will be shown by repeating the letter.  You know how to say "beet" but e's are now i's in this conlang, so you would have to write it as "biit," but it would sound the same.  There are a lot of issues we could get into here, but I'm just creating a framework to give an example of how this might work.
So, "cinnamon" would become hinnumon.  "Bulletin" would become pullitin.
"Keep this reference near you at all times" would become "Kiip hi yeheyen niu yu ut ul tiim" or something like that...

Step Three: Develop an altered grammar and syntax.
Example: Let's use a Yoda syntax and go for OSV (Object - Subject - Verb).
Lets use -im to make something plural.  Some "times" is now tiimim.
Let make ku- a prefix that shows command form of a verb.

So - according to the grammar rules, "Keep this reference near you at all times" would become "Near you, this reference keep at all times."
Then we apply the rest of the rules and "Near you, this reference keep at all times" would then become:

"Niu yu, hi yeheyen ku-kiip ut ul tiimim"

I don't know about you, but that does not look Englishey to me!  But hopefully you see my point, which is just that if you apply a few linguistic principles, you can make English, or whatever your native tongue is, into your own little conlang, and not have to worry as much about generating vocabulary and so on.  Try this out for yourself and see what you think.

How Do You Eat an Elephant?

Kam le elefant wape?
How you elephant(singular) will be eating?
How do you eat an elephant?

In wip siled in wip!
One bite after one bite
One bite at a time.

I was just thinking about the work I would still have to do on my conlang in order to get to a point where I felt like it was finished, and I was just feeling like I would NEVER get there.  Too many other more important things to take care of in the mean time, too many little details to wrestle with, too technical to understand, etc.

Anyone who's been reading my blog knows that I constantly cry SIMPLIFY and I realized I needed to take some of my own advice again.  When I first began getting into conlanging, so much of it was over my head (and there are a lot of things that still are!), but I just kept with it, taking in what I could, and utilizing what I did know as best I could.  Its good advice for beginners as well as higher-level conlangers.

I stopped myself and thought that instead of thinking of it all as a whole, I needed to just keep putting one foot in front of the other and taking little steps.  Get that list of words finished.  Later, test out the different tenses.  Later, translate that page of text.  On and on.  Small but consistent efforts are often more effective than larger but inconsistent efforts.

Bottom line is this: Don't try to make an entire language all at once.  Its a huge project, that will take a lot of effort.  Break it down and stay focused on the next step, not on how far you have to go. 

Have an awesome weekend everyone!

More on Ubese!

I recently got a request to expand on the Ubese language I deconstructed previously, and I had so much fun working on this the first time around, I just had to oblige (Thanks Libra!).

Okay, so I previously established a phonology and morphology; here they are again with a few changes:
Consonants- t, sh, ch, k, n, l, hh Semi-vowels- y Vowels- ee, ay, i, a, u, o
Morphology- CICV (I for Inflection)
Plus, some syntax and grammar rules:
Subject Verb Object, first vowel sound of word denotes inflection (i=I, a=you, o=he/she/it, ay=we, u=they), stress is commonly on second syllable of word - if first syllable is stressed this means something, based on the context (while holding a thermal detonator it means you're getting crazy up in there, so be careful with this!)

In the first post I did a lot with just the two-word-phrase that was pretty much all that was used in the clip from Return of the Jedi. Towards the end of the post I started trying to use more than two words in a phrase and raised some questions about how things should be ordered.

As I considered what rules to add for bigger phrases, I realized I might be going in the wrong direction.  Instead of trying to make the language more complex, I thought I should continue to keep it simple.  So I considered more phrases to translate (loosely related to tell a story):
They are my daughters.
I searched all over the town.
I could not find them anywhere.
But I did not give up hope.

They are my daughters. The possessive inflection in this sentence made me wonder if I had to create a rule of if this could be understood through context or some other way.  Lets make taynu the word for is/be, and naku the word for daughter. "They are being" would be Tunu, and daughters could be nakun, according to the rules I made previously. However, with out a possessive inflection, how can we tell if he's saying They are daughters/they are his daughters/they are my daughters?  Let's try this - we reinforce the CICV word syntax by making a new rule: you can use CICV on the second word in a word phrase to show possessive inflection, rather than a subjective inflection. So now we know a standard word phrase is structured thus: CICV(T) (CI/VCV) (X)

C=consonant V=vowel I=inflective vowel T=tense suffix X=helper word

So, They are my daughters would be translated thus: Tunu nikun

I searched all over the town. Alrighty, should be easy now. Lets make chayshay the word for 'search,' just because it reminds me of chercher - French word for search. Kayto will be 'town.' So, Chishayko kayto is easy enough; we substitute the ay in chayshay for i to make I the subject, and add -ko to make it past tense.  Literally, I searched town. Lastly, lets use reduplication - Chishay chishayko kayto - by doubling up on the searching, we show that the searching was intense. You'll notice I didn't add the past tense marker to both chishays... and honestly this is simply because it didn't feel right. I tried to imagine Boushh's voice saying chishayko chishayko and something about it didn't sound right. What do you think?

I could not find them anywhere.  This one is more difficult.  Yisha = able, tichay = find.  So, Yishako nay tuchay could mean "I was being able not to them-find."  This bizarre word order (bizarre to my English sentiments at least) raised a question: I had always thought of the helper words as being after the word they affect, just as Boushh said "Yatoh, cha." But, it sounded a little better to me to say Nay yishako tuchay.  I was about to create a new rule when I imagined a Boushh voice saying the phrase more like "Yishako NAAYYY tuchay!!" as he told the story, basically almost yelling the negating helper word to show frustration and anxiety, or perhaps just reduplicating the nay.  Either way it felt right so I continued.

A side-issue I thought of during the last translation was, if the first vowel is whatever it needs to be to show the appropriate inflection, how would they teach the words?  My thought was maybe they just show the words as y'sha, t'chay, t'nu, n'ku?  First vowel missing, because they know it becomes whatever is needed for the appropriate meaning. 

But I did not give up hope.  At first it looked tricky, but I realized it was pretty easy, once I looked past the metaphor.  I could literally boil the phrase down to "But, I hoped."  If yakee = hope, and yayt = but, or yet, then "Yayt, yikeeko."

SO! 
They are my daughters.                         Tunu nikun.
I searched all over the town.                 Chishay chishayko kayto.
I could not find them anywhere.            Yishako NAAYYY tuchay!!
But I did not give up hope.                   Yayt, yikeeko.

So theres a little more on Ubese.  I still think its pretty minimalistic, but we've gotten a little ways away from the couple of words used in ROTJ. Keep conlangin'.