This blog has moved:

In addition to my current writing, all the old posts are collected on the new page.
(You can use your browser's "find" function to find what you're interested in there.)
Your browser does not support Javascript.
This site requires Javascript.
You can see where this becomes a problem.
Without Javascript,
Many posts will look wrong
Comments are inaccessible
Interactive dialogues won't function
Hidden text will never be revealed
The sidebars will not open

If you choose to continue, be warned
That you are missing crucial elements
Of I Am Not's design.

Sunday, January 10, 2010

~My Alphabet

My Alphabet

When I Grow Up, I Want To Do Everything (26/7/2009)
Step 2: Arellian
I had almost finished my logical alphabet, where the connections between letters are perfectly intuitive and logical. I needed to start working on the vocabulary. It would use a modified version of Hebrew's "root" logic, because it just makes everything more sensible. In addition, you could get the opposite of any word by spelling it backwards. And the longer a word, the more specialized its meaning. There would be short words talking about general concepts, so that you could have a basic conversation without having an advanced vocabulary. But adding on more letters to the start and end would add on subtleties and contexts and connotations and iterations. Starting from these rules, I'd eventually deduce the single most logical language in the world. I believed I'd find that in the end, there would only be one solution to this problem.

This post is very long and thorough, and includes both images and an audio file. If you are short on time, or are not interested in the minutiae of useless things, you may want to skip most or all of this post.

The first few years of my life were spent in America, where the name "Mordechai" might sound weird, so my parents gave me a nickname. At the time I first figured out how to read, that name was spelled "Morry". An overthinker even then, I went to my mother and protested. "If it's spelled 'Morry', it has to rhyme with 'sorry'. But I'm not 'Mahr-ee', I'm 'More-ee'!" So we took an R out of my name, and I was satisfied. But of course I was making a naïve mistake: I was expecting language to make sense. Language hasn't made sense in thousands of years.

Ancient Hebrew- now there was a sensible language. In Hebrew every word is made up of a shoresh (meaning root) of three letters that says what the essence of the word is. You can detach those three letters from the "structure" that contains them, so that you can understand what the word means just by recognizing those two elements. The logic of the system is clear: just by knowing one word (any word) from a particular root, you can infer all the rest. There are rules covering every aspect of spelling and grammar. And in ancient times, the Hebrew letters were all distinct from each other. None of this nonsense like in English where C and S can make the same sound, and sometimes an S can sound like a Z, and C, K and Q can all be the same. English derives from languages that derive from other languages that derive from other languages that eventually go back to ancient Hebrew, and it makes sense that with such a logical language as ancient Hebrew the other languages would want to build on that, but over the millennia all sorts of inane compromises of sanity have crept in to the point where none of the elegance remains.

(I'm sure there's a long history explaining every misstep that led to where we are now, but I'm not too curious because all the people I could assign blame to are long dead.)

In Israel I learned Hebrew, and those rules were like a breath of fresh air after a lifetime of pollution. But modern Hebrew is not ancient Hebrew. Many mistakes and contrivances which got into other languages somehow found their way into Hebrew. What keeps it reasonably sane to this day is that we've got many books written in ancient Hebrew, from which we can figure out how the general grammar works. (Modern grammar is different in many little ways, but it's mostly the same.) But we don't have an audio recording telling us how it's meant to be pronounced. So in learning the rules, there was a bit of a frustrating disconnect where I understood the logic of the rules, but didn't see them being applied consistently.

For instance: in Hebrew there are six letters which sometimes have dots in them. Without a dot the sound is continuous, with a dot it's a single sound. (I actually think that's exactly backwards from the way it used to be, but that's a different topic.) Take the letter pei: with a dot it's pronounced like a P, without it's pronounced like an F (or "PH"). That makes sense. It's roughly the same sound, except that one is continuous and one isn't, so they're the same letter and there are rules saying when you use one or the other. But in the rules of spelling, there are six letters like that: the equivalents of B, G, D, K, P, and T. Of those, only the B, K and P equivalents make a different sound with a dot. That's clearly wrong. In ancient Hebrew each of those had a continuous form, but we've forgotten or discarded those.

Things like that are just inconsistent, and that drove me crazy. If it's going to be a sensible language with sensible rules, it's got to go all the way. You can't just take out some of those rules and expect the result to still make sense. There are letters whose sound we've forgotten, so we pronounce them the same as other letters. The result is redundancies and confusion, which is not how Hebrew is supposed to be. Hebrew is supposed to be a holy language, not a random one. When we read out of our holy Torah every Shabbat, we're reading it wrong. It's not supposed to sound the way we make it sound, and I imagine the real thing is probably much more beautiful.

I admired the ambition of Hebrew, didn't care for how it had ended up. So in fifth grade, I started working on a replacement. By the standards of this confused world Hebrew is still pretty logical. But not nearly enough for my liking, and that's what this new language I imagined would have to be. I started with the alphabet, trying to work in all the (American) English sounds but in a structured way which was detached from the fuzzy logic of modern alphabets.

Here it is:

Back then I called it the Arellian alphabet after my middle name; now I call it the Hee alphabet, for reasons that I will explain later. There's a lot going on here, so it might take a while to explain fully. (The scanner chopped off the sides a bit. Just pretend it didn't.)

But first, please note that this is not exactly what I came up with in grade school. What I came up with then had flaws, which I've corrected in order to post them here. This is, in my view, the definitive version of the alphabet, and if ten-year-old me saw what I've done (and heard my reasoning for the changes), I have little doubt that he'd come around to my way of thinking. 35 of the 52 symbols here are exactly as scribbled on all my school notebooks, and the other 17 I created yesterday to fill in the gaps and correct the mistakes while staying fairly true to the original scribbles through all the revisions.

Anyway. As you see, both the vowels and the consonants are categorized into families. Similarly to Hebrew words but unlike Hebrew letters, just by looking at any letter in this alphabet you can understand exactly how it relates to all the other letters. In both English and Hebrew there are some sounds which are combinations of other sounds, and that goes unacknowledged. That's unacceptable to me. If a common sound is made up of other sounds, you know it because you can see that the symbol for the letter is made up of two other letters. So that's what all these vowel and consonant "combinations" are.

Here's an MP3 file of all fifty-two sounds, some of which you've most likely never heard. I'm going to get to an explanation of each letter in a minute, but first I'd like to clarify where I'm coming from with this.

It probably looks like a tremendous number of letters, and I guess it is. There are twice as many sounds here as in the modern Hebrew language, or probably even the (more diverse) biblical Hebrew language. It even has more sounds than English, a language where each letter can sound twenty different ways depending on the time of day. On the other hand, it's not even close to containing all the sounds out there, so it's not useful for a thorough examination of language either. The question that must be asked is what exactly it's for. My answer is that it was a personal way for me to make sense of the sounds that I already knew and used on a daily basis. English and Hebrew didn't make sense, so I needed to satisfy myself.

Every single sound which I use in casual English speech is contained in this alphabet, even those which I use only when saying Hebrew words in English. The reason there are more letters than both languages combined, even though authentically Hebrew sounds are not accounted for, is twofold. First, the letters are consistent. Each letter will always make the same sound, with no exceptions. Second, the rules need to be thorough. If there are four sounds to be made of a certain category, I can't just include three. I need the full set, or else I can't justify including that category to begin with. This second rule is a lot less satisfying to me now than the first, but back then I thought the alphabet was a lot more thorough than it actually is so I didn't have a problem with it.

Okay, let's go through bit by bit. Keep in mind the whole time that I speak with an American accent. If you have a different accent, you're almost guaranteed to be confused at times unless you remember that.


First up are "ah" and "uh". Similar sounds, so much so that when Israelis speak English they think there's only "ah" and no "uh" and no one thinks to correct them. "Ah" is the A in "blah", the O in "pot" and the first E in "en garde". "Uh" is the U in "run", the O in "cover", the OO in "blood", the A in "mesa" and the E in "the".

Next are the short A, E, and I, which have always sounded to me more like three parts of a specific range of sounds than three distinct entities. You can't go from "a" to "i" without passing through "e" on the way. At least, that's how I see it. They are drawn accordingly, and note that all three are only drawn above the middle of the line. (I drew on two lines to better emphasize this.) Anyway, "a" is the A in "cat" and the EA in "yeah". "E" is the E in "pet". "I" is the I in "think", the O in "women" and for that matter the E in "women" as well, the E in "glasses", the second O in "horizon", the U in "rhesus", the Y in "cyst" and the dot in "Mrs.". :)

"O" is the O in "or", but not the O of "most". It is the sound made by the Hebrew letter vav, and I believe it's also like the O you'd hear in Spanish though I'm not certain of that. It's a simpler sound than the English O; if you don't know what I'm talking about you'll have to listen to the MP3. "Oo" is the OO in "caboose", the O in "move", the U in "ruse" and the W in "ewe". What I call "u" is probably not what you'd expect; I'm referring specifically to the U of "put" and the OO of "foot". The old version of "Arellian" had a fourth leter in this category, and the whole family looked more congruous for it, but I just realized now that it's actually made of two sounds. It's hard to shake what I've learned. Anyway, this is a family of three because the mouth is a similar shape in all three, but it's not a range of sounds like the "e" family.

Finally there's "ee": the EE of "sweet", the EA of "leaf", the EI of "either", the E of "mete" and the I of "pizza". This letter is in a category of its own. It looks like the Hebrew letter yod, which makes the same sound.

Vowel Combinations

The first line is what you get when you add "ee" to the end of all eight of the other vowels. There are two "I"s: the first is the Y in "fly" and the second is the I in "flight". They are not the same sound; the first is "ah-ee" and the second is "uh-ee". The third I put in parentheses because it's hard to describe, but it's a short "a" with "ee" added. For some reason it makes me think of pirates. "Ay" is the A of "date" and the EI of "neigh" and the É of "café". Like "I", we English-speakers think of this as one sound but it's actually made of two: "eh-ee". After that is "i-ee", which is again not English. The MP3 file will tell you how it sounds. Then to the "oo" family: "oy" is the OY of "boy" and the oi of "noise". "Ooy" isn't English except under certain pronounciations of "buoy", but it's straightforward: just try to say "Shmooi!" as one syllable, and you've got it. And finally "u-ee", which is the U of "put" plus "ee" though that might be hard to imagine without hearing it.

The second line is what you get by adding "oo" to the other vowels. I'm just going to skip to the parts that are English sounds, because you understand the routine by now. "ow" is the OW in "now", which is actually "a"+"oo". "ew" is not in any words except "Ew!", but you know it from there: "i" plus "oo". Then there's "oh", which is the English O in "boat" and "comb" and snow". And at the end of the line I should have put in "ee-oo", as in the word "Eeeeeew!", but I didn't think of it until just this moment. My bad. The old school notebooks probably had that, though, because it just makes sense. It would look like a backwards "ooy". (So I guess there are actually 53 letters, not 52.) There's no "u-oo", because the two sounds are so similar that I personally can just barely perceive the combination as being different from a simple "oo", though that's probably just me.

On the third line is one of the changes from the original that I mentioned earlier. "Aw" used to be in the "oo" family as its own basic sound, and that always gave me lots of trouble but I couldn't figure out why until today. It's actually a combination of "o" and "uh", isn't it? If I had realized this back in the day, I might have filled out all the rest of the vowel+"uh"s, because they're all pretty distinct sounds, but today I don't see the need. Only one of those nine sounds is in my life. "Aw" is the AW of "saw" and the AU of "pause".

And that's all for the vowels! [phew] This didn't seem so complicated when I made it up...


What I found about consonants, when I thought about them, was that they could generally be broken into categories by two criteria. First, some sounds were short sounds after which you move on, and others were continuous. Secondly, some sounds sounded dry and some sounded soft. I don't know how else to describe it.

Let me give you an example of what I mean, using the first family of consonants I wrote out. (There isn't really a proper order to any of this, it's just how I decided to write it this time.) P is a noncontinuous sound. If you make it continuous (down), you have an F. If you make the P's sound softer (in the sense of texture, not loudness), you have a B. That's written to the right. And if you make the P both soft and continuous, you have a V. A family of four. Now take notice of how these letters are written; their appearance reflects their places in the family. The P and F are made of sharp angles, while the B and V are made of soft curves. Also, the F and V's lines end up back where they started. That's a principle I was particularly happy about: if a consonant is continuous, you know it because it closes a loop. By the way, the resemblance to a lowercase B and capital F was not unintentional. It made the letters easier for me to remember.

The T family follows the same principles. The T and D equivalents look much like lowercase T's and D's, but more importantly the T is made of straight lines and the D is curved. The continuous form of T is Th, which is the Th of "math". The Th of "there" I'm calling Dh, because it's the continuous form of D.

The third family is less English-sounding. The soft version of K is G. The continuous version is Kh, like the "ch" in "Mordechai" or "Bach" or "Blecch!". That's the Hebrew khaf, not the hebrew khet, because I don't pronounce khets correctly even when speaking in Hebrew. The final part of the family I can't really describe in words, because it is neither a part of English nor a part of modern Hebrew. (I like to think it was in ancient Hebrew, though.) I call it Gh, and if you want to know what it sounds like you'll just have to hear it in the MP3. Alternatively you could figure it out for yourself, because there's only one sound that could possibly complete the sequence. It's the continuous form of G and the soft form of Kh. Now, what I find funny is that everyone thinks the continuous form of G is J, but the two sounds aren't connected to each other in any way! I'll explain a little bit later what the J sound is, but it has nothing to do with this family.

The fourth family is an odd one. S and Sh are clearly connected, as is Z and Zh (the S of "decision"), and the Z's are certainly the soft version of the S's, but all four sounds are continuous. So they all start with a line in the middle and then close a loop, but they close that loop in different ways.

In the bottom left corner is H. That's not connected to anything, because it's kind of the prototypical sound. You don't close your mouth at all to make it, you just exhale. So the design is a simple two lines- first there's silence, and then sound. This does not conform to the closing-a-loop rule, but the line goes along with the rule I wrote to the right of the big word "Consonants": drawing a line in the middle coming out from any letter continues its sound. (This rule replaces any cases where one letter repeats itself, which would be redundant.) So in a sense you're just continuing the most basic sound. Because of that, H is to my mind almost a vowel, which is why I didn't make it loop.

Speaking of almost-vowels, the two letters to its immediate right are just vowels masquerading as consonants. Y is what you get when you continue the end of an "ee", W is what you get when you continue from an "oo". Each is drawn as its vowel with a line continuing it. (With both H and these, to lengthen the sound you just draw out the line further. You don't add a second line.) Y and W should be considered a family of two.

All the way on the right are the four leftover sounds. These cannot fit into any possible categories, but they are common sounds in both English and Hebrew and therefore need to be represented. I would have loved to make whole families out of each of them, but that would involve filling those families almost entirely with sounds I don't know. The fact that I didn't do that makes me feel better about not completing the "aw" line. But even if I had, whatever came out would most likely not resemble the four regular consonant families. The R is pronounced like an American R. If the N comes before a G or a K, it is not combined with those sounds as in English but the two sounds are kept separate. You'll notice the lines sticking out from the R and L, with parentheticals pointing out that they aren't always there. That's because I find that R and L are most peculiar letters: they have a bit of a vowel built in. Every time your mouth moves to form an R or L and air is still coming out of it, you have to make an "u" sound of sorts in the process. That's how it seems to me. So the line completes the V-shape that's the "u" vowel. The line does not appear when an R or an L is the first letter of a word. I'm okay with adding a weird rule like that, because it's just codifying how I already talk.

Consonant Combinations

This bit could have been longer, including sounds like X (=K+S) and maybe some new ones, but I don't remember having any combined sounds other than these four back in elementary school so these are the only ones I'm including. The first line is straightforward: I'm adding the S family to T and D. "Ts" is the Hebrew tzadi, so I use it regularly to say things like names. In English it's the double-Z of "mezzo". "Dz" is just D and Z together; I can't think of any words which use this as a single sound, but it's the softened version of "ts" so it has to be here. The other two you'll recognize: T+Sh="ch", and D+Zh="j". "Ch" is the CH of "which", "J" is the J of "Jew" and the G of "gem". How these letters got connected to C's and G's, I have no idea.

And finally, something which isn't so much one letter after the other as it is the synthesis of two sounds. "Ng" looks like an N inside a G. I mentioned earlier that just an N with a G after it makes two disconnected sound, so here's the way to write the English NG of "ring". Unlike in English, this sound is not restricted to the ends of words; it's a letter like any other.

So that's the Hee (formerly Arellian) alphabet. Here's how my full name is written:

Now, the Arellian alphabet was supposed to be a precursor to the Arellian language. I never did that. I started it, and gave up almost immediately afterward because the task I'd set for myself was just too big. There are a few principles that all words in this language would need to follow. First, any word written backwards is its opposite. Second, the shorter the word the more general its concept. The more specialized the field up for discussion, the longer the words get. Third, changing any single letter in a word (consonant or vowel) to another letter in its family (or a combination letter made from it) gets you a word which means almost the same thing but not quite. Fourth, by looking at a word you've never seen and comparing it to other words you know you should sometimes be able to figure out what it means for yourself.

You can see why I gave up. I'm not sure if it's theoretically possible to make an entire language with such precise logic to it.

I might as well tell you where the name "Hee" comes from. In twenty years or so, I hope to have gotten up to the point where I can make my fantasy-RPG idea. In that game there will be many different races, one of which I call Hee. They're fundamentalist atheists in white burkas who strive to achieve perfection by destroying anything which doesn't fit their very narrow view of what's logical. (Like all the races I have planned for the game, they're actually me in disguise.) This imaginary language I wanted is perfectly suited for them, being a naïve attempt at perfection of thought. "Hee" is the only name they could possibly have. Like I said, any letter can be changed for another letter of its family. That means that any word is not the be-all and end-all of its concept, except for a word which has only one syllable where all the letters come from one-letter families. That's H and "ee". The word written out looks like a backwards four, which is a striking enough symbol to put on flags.

(By the way, while trying to find a suitable name for the Hee I first went through the two-letter families, and accidentally stumbled across the Jewish name for God! That gave me a lot more respect for the unappreciated logic of correctly-pronounced Hebrew.)

I do not intend to create the Hee language, Tolkien-style (or Avatar-style, more topically). Back when I was in school it seemed like a good idea, but when I was in school I was a lot more bored than I am now. It just seems like too much work for something with too little purpose. If anyone reading this would like to give it a shot, be my guest.



I like your idea of the alphabet. I also thought about making one, but the combination of all the sounds used in correct Hebrew and their families was just too much for me. I do have some corrections for you though.

You forgot to talk about M and N. Coincidentally, I have a correction to your classification. I think M should be related to the B family. Similarly, N should be related to the D family and ng should be related to the G family. It's not very hard to make the connection. Each one is just making that lip position and breathing out of your nose instead.

[If you have a cold and can't breath out of your nose, you will probably use the one you blow out of your mouth for (B instead of M). Try it! Just replace the letters M N and ng for B D and G, ad you idstadtly soud codgested!]

Anyways, ng should not be classified as a consonant combination, rather a consonant on its own (related to G, as stated above).
Your system isn't completely consistent, there are still some things it doesn't cover:

How do you know when to change to the next syllable? Words can be spelled the same way, yet still divided differently. In my name, for example, B-e-ts-ah-L-e-L how do you know that the division is ts-ah-L e-L and not ts-ah L-e-L?

Similarly, you don't talk about accenting. Which syllable is accented is undetermined by your system.
Just a random comment, your gh is like many people pronounce the reish in Hebrew (although its correct pronunciation is not in your alphabet). There is also a letter for it in Arabic, the ghin. In arabic it's related to the 'Ain.

M and N are grouped in the "left over" category together with R and L. It's not a particularly elegant way to include them. And yes, NG isn't really a combination but a sound of its own. However, it would not fit into any of the families because there's such a strict logic of four members to a family. Just because a person who can't speak clearly pronounces one letter as another doesn't mean they're related, and I see no reason to put M and B together.

If there is no space between one word and the next, there is no break in syllables and they flow one into the next. Your name would only be pronounced Betsah Lel if it were two separate names. Accents are more of an issue, but I think putting the accent on any syllable at all would be acceptable for this theoretical language.

They're not related because people can't pronounce them well, that's just a way of demonstrating it. The fact is that you're making the same mouth position in the whole family and in that letter. The difference is that you're breathing from your nose.

Even if you do decide to keep them separate, there's no reason not to include ng in the same family as N and M. [The reason there's no equivalent for the forth family (s/z) is that it's actually part of the second (t/d) family. See]

The problem is in names such as mine is that the normal syllable division is different than just consonant-vowel. In my name (how I pronounce it in English), the syllables should be Be-tsal-el. The alternate (more logical) way of dividing them would be Be-tsa-lel. There's nothing here to differentiate between the two ways of division.

I guess I can see what you're saying about M being related to the P family. If that's the case, it seems like NG ought to be grouped with the K family somehow. And that means that there are other rules that need to be taken into account for grouping consonants together other than the two I identified, because you can't have just two families of five. If those families can lead to those sounds, whatever principle that is ought to be applicable to the other families to get other sounds which aren't from English or Hebrew.

The more I think about this alphabet, the less elegant it seems.

What I'm not convinced is a problem is what you're saying about breaking up words. Frankly, I don't see what it is you're objecting to. When I say "Betzalel" I'm not breaking it up "Be-tza-lel" or "Be-tzal-el", because I'm not breaking it up at all. I'm saying the whole thing together in a continuous legato sound. Each letter leads right into the next, without any breaks, so your insistence that it needs to be broken up into distinct chunks seems to be at odds with how we actually talk. We don't pause between syllables while talking, we only pause between words.

The alphabet could still be simplified at the price of phonetic accuracy. I mean you could still put things in these sort of groups without actually taking into consideration all of the related sounds.

Our forefathers also thought like you when they put together the alphabet. They didn't think of NG. The special rules in Hebrew for the letters B G D K P T concerning when they're voiced and when they aren't show that they had some kind of grouping with them. The S and Sh appear in the same letter. In the ancient script the M and N are very similar. I think W and Y are too. Tet and Kuf are clearly Taf+'Ayin and Kaf+'Ayin. The 3-letter roots have a common 2-letter root allowing you to figure out pretty much what things mean. [P-R-x for example all mean things like open up, unravel, etc.] So I think language was originally very logical, it's only been made less logical throughout the millenniums of changes it had. I don't know about backwards spelling though.

There's no complete stop between syllables, but there is a difference in the way we speak. It's the difference it the nikud is on the lamed in my name or on the aleph. If you use a glottal stop or not.

What you're talking about is going way over my head, I think. So it's safe to say that fifth grade me would not have thought of any of it. This alphabet has not been a serious proposal for more than a decade now.


Post a Comment