Stormlight Archive Translation Guide

[Harakeke]

@ Marianmi: Yeah, see p. 2. Though most of them are still too tricky for me to make out.

~~~

Speculation time!

I believe the "Compass Rose" glyph is pronounced "Urithiru" -- or is at least somehow associated with the lost city. I don't have a good reading of the sub-glyphs yet, but the first syllable is very plausibly "uR", followed by what I'd tentatively say are "iTh" and "iR".

I speculate that Urithiru is/was located where the "Compass Rose" glyph is on the map of Roshar's southern hemisphere.

"Though many wished Urithiru to be built in Alethela, it was obvious that it could not be. And so it was that we asked for it to be placed westward, in the place nearest to Honor."

The "latitude" markings on the Frostlands map don't correspond to a bearing derived from the equator -- but rather from an arbitrary 0 mark at around 68°S. Roughly the same latitude as the Compass Rose on the Southern Hemisphere map! A non-glyph compass rose shows up at about the same location on the map in the full-color WoR endpaper (ceiling fresco?).

What other bearing would be important enough for maritime cartographers to use as the basis of their maps? The seat of the Heralds!

If Urithiru was out in the middle of the Southern Depths, this would also explain why the ship in Kharbaranth harbor has a Compass Rose glyph on its sail. I had assumed that Shallan had done the drawing, but it really isn't in her style. The presence of a ship from Urithiru implies that it is much, much older. It even looks like there are some storming glyphs hidden beneath the caption!

The Compass Rose glyphs on the various maps likely represent the locations of gateways into the "road" or "rapid transportation" method mentioned by King Nohadon. We see them all over the Frostlands map -- most notably in the Shattered Plains. The Map of Alethkar likewise has a Compass Rose at the location of the Shattered Plains.

Shallan notes that the gateways are very dangerous to activate. Perhaps a gateway mishap is what shattered the Shattered Plains.

Edited March 11, 2014 by Harakeke

[Argent]

Except we know where Urithiru is. More or less in the center of Roshar. Do you believe it was moved?

[Harakeke]

 

Except we know where Urithiru is. More or less in the center of Roshar. Do you believe it was moved?

 

Oh really? So much for that theory then. =P

I still think the compass rose glyph is suspicious...

[Argent]

I just went back to the (revised) original post, and have a couple of questions.

Regarding the Thaylen script - which letters / sounds are we missing? 'B' is one that comes to mind, but I'll need to go through the books and find all the Thaylen names I can, see if any others show up (we know 'b' must exist because of babsk)

Also, a few thoughts on the Alethy glyphs:

* The description explains the three ways of writing a glyphword well enough, but it doesn't really talk about how to read or write them. The key does that, but it's not very self-explanatory. I think I figured it out, but it definitely took me some time.

* Using shash as an example, it might be more intuitive to not include the "-S" glyphpair, but instead say that 1. each glyphpair must start with a vowel, and 2. if we are left with a standalone consonant, it gets a vertical bar to separate it from its mirror image. This rule, while more roundabout, also sounds more... reduced to me. It's like pulling a factor out of a mathematical expression in my head.

* Of course, Sadeas had to ruin my perfect rule like he ruins everything else. His name would be split into the three glyphpairs [sa][de][as] - which screws everything up. I can't see the Alethi Warcamps glyphs well enough to decipher how "Sadeas" would look like, but I'd be curious.

* The thing about glyphpairs starting with a vowel might be wrong.

Going to leave this post unfinished, got an interesting thought...

Edited March 11, 2014 by Argent

[Argent]

Alright, double-posting, but I am much excite. I am going to write some observations for now, and we'll see if I - or somebody else - can take them elsewhere later.

#1: The vertical line in the middle of the sas glyph is an odd one - it's not present in any of the Thaylen letters, and it's very rare in the illustrations (including the Orders / Surges chart). I now suspect it might be optional (fun fact, it looks like the Map of the Battle of the Tower and the Alethi Warcamps map have the highprinces' glyphs oriented differently; one of them has it upside down)

#2: I am going to guess that each glyphpair consists of two sounds (one vowel and one consonant, not necessarily in that order though), mirrored along the y-axis. Furthermore, it looks like we can take the "core glyph" of each consonant and rotate it along either axis to denote that this consonant is now part of a different glyphpair. For example, if we use 'L' to denote some glyphpair, then 'Г' might be this pair's consonant followed (or preceded) by a different vowel (e.g. 'L' could be "[a]" while 'Г' is "[e]")

#3: If I am correct in my previous assumptions, then there are 12 possible glyphpairs we can create using a single vertical line, a single "core" consonant glyph, and all the vowels. Each "core glyph" can assume one of four forms (default, rotated along x-axis, rotated along y-axis, and rotated along both axes). If we add the vertical line to the left, that's another 4. If we add it to the right, that's yet another four. This might be something we can work with, because there are 11 two-letter combinations we can form using a single consonant and all the vowels, as long as we include a "no vowel" as a unique option (e.g. [empty] and [empty] are the same).

#4: I am running out of time for now, so I'll have to go back and finish this, see if I can come up with a way to formalize everything. A set of rules that explain how to form any possible glyphpair.

Wow, brain is really fried right now. Wish I wasn't at work...

[1nikhil9]

There is also this:

 

 

Through the carriage window Shallan saw that they were finally approaching a warcamp flying Sebarial’s banner. It bore the glyphs sebes and laial stylized into a skyeel, deep gold on a black field.

 

Sebarial's banner has two glyphs... does that what a glyphpair means? But then how would it be symmetric?

[Argent]

It seems like glyphs are mirrored for symmetry. So instead of writing "ha" you would write "ahha."

It's possible that glyphs are also written outside in, but I doubt it. So the glyph "sas" is really a big "s" on the outside, surrounding an "a" in the middle.

[lilserf]

Is there a good resource with images of all the glyphs from the books collected together?

[1nikhil9]

From tWoK:

 

 

The slave’s black eyes glanced upward, toward Kaladin’s forehead, which bore three brands. The first two made a glyphpair, given to him eight months ago, on his last day in Amaram’s army. The third was fresh, given to him by his most recent master. Shash, the last glyph read. Dangerous.

 

So two glyphs make up a glyphpair..

 

 

Kaladin glanced to the side. This man’s forehead brand was older than Kaladin’s, the skin around the scar faded to white. What was that glyphpair? “Sas morom,” Kaladin said. It was the highlord’s district where the man had originally been branded.

 

And:

 

 

Deep blue with white glyphs—khokh and linil, stylized and painted as a sword standing before a crown. House Kholin. The king’s house.

[Harakeke]

@Argent: Great thoughts!

 

1. Yeah. My grasp on that is still somewhat in flux. Generally it seems like glyphs are written symmetrically outward and down, starting from the top middle.

 

2. The vertical bar might be purely decorative, or it might indicate that the word starts with a consonant. Either option seems plausible to me, pending further insights. I tried to structure my key template to accommodate either interpretation.

 

3. Yeah. Sadeas is a pain. =P I parsed his glyph phonetically: [-S] [aD] [aS], but this isn't necessarily correct.

 

4. Quite possibly! I arbitrarily picked the vowel-consonant ordering to narrow down the possible graphemes, but consonant-vowel is equally plausible, and perhaps more elegant.

 

1. Yeah - half of those glyphs are upside-down

 

2. Exactly

 

3. Yep. It looks like in the calligraphic form, the line can denoted with a dot instead. This may distinguish meaningful lines from decorative lines.

 

4. I look forward to seeing what you come up with!

Edited March 11, 2014 by Harakeke

[lilserf]

I think kholin is written in the negative (white) space, and tanat is written in the positive (black) space.  So they can sort of partially overlap. Alternately, I could just be wildly overreaching -- but anyways:

 

Not sure if you've noticed, but the Kholin and Tanat sides are not EXACTLY identical. The recent hi-res images at http://brandonsanderson.com/books/the-stormlight-archive/words-of-radiance/stormlight-2-maps-and-illustrations/ make it clear that there are some differences on the two sides.

[Harakeke]

Not sure if you've noticed, but the Kholin and Tanat sides are not EXACTLY identical. The recent hi-res images at http://brandonsanderson.com/books/the-stormlight-archive/words-of-radiance/stormlight-2-maps-and-illustrations/ make it clear that there are some differences on the two sides.

 

Yeah - that's why they only sort of overlap.  But I'm not particularly thrilled with my speculation about that glyphpair in general. =P

 

 

Is there a good resource with images of all the glyphs from the books collected together?

 

No, but that'd be a fantastic way for someone to help if they don't want to dive into the codebreaking side of things .

 

There are digital versions of the illustrations available at :

http://brandonsanderson.com/books/the-stormlight-archive/words-of-radiance/stormlight-2-maps-and-illustrations

and 

http://brandonsanderson.com/books/the-stormlight-archive/the-way-of-kings/stormlight-1-maps-and-illustrations/

[Argent]

I was actually thinking about Sadeas while I was forced to be away from the decoding project - and I am not sure his name is supposed to be translatable into glyphs. We know that Dalinar's last name can be turned into a glyphpair (Kholin = khokh + linil), but I didn't know that and was trying to write it out, I'd probably try to do it phonetically - and likely fail. Because his name is not [kho][li][n-], it's [khokh][linil] - however those two might look like. What I am pointing out here is that Dalinar's last name takes one (symmetrical) half of each of its "parent" glyphs, and merges those halves together into a single family name glyphpair.

khokh > kho|kh

linil > lin|il

Kholin > kho|lin[/code]

So maybe Sadeas' name is not supposed to be written as a series of glyphpairs, but a single glyphpair instead. If we use Dalinar's example and reverse-engineer it, we might be looking at something like

Sadeas > [sad|eas]

sad(+ das) > sadas (or possibly dasad)

eas(+ sae) > easae (which I'd wager is supposed to have an 'h' or two somewhere, because this does not sound like the name of an Alethi glyph!)

So for Torol Sadeas' family name glyphpair, we should try checking whether we can write the parent glyphs first (sadas = [ad][as], easae = ???), and only then try to combine them to form "Sadeas."

[lilserf]

So maybe Sadeas' name is not supposed to be written as a series of glyphpairs, but a single glyphpair instead. 

 

I think this is definitely true; and it wouldn't even be as literal as "sadas" and "easae" - Sebarial's glyphpair is "sebes" and "laial"!

 

Could be "sas" and something?

[lilserf]

Argh, even worse: Amaram's glyphpair as described in WoK is:

 

"a burgundy field blazoned with a dark green glyphpair shaped like a whitespine with tusks upraised. Merem and khakh, honor and determination."

 

That doesn't even phonetically add up to "Amaram", really...

[lilserf]

Ooh, one more from WoK: 

 

"The stark black glyphpair read shesh lerel - Sheler's company".

[Argent]

Stormfather, those are going to be a breech to figure out. I love it!

In the meantime, I took a quick look at The Way of Kings Endsheet (because the Surges have pretty clearly drawn glyphs there), and more specifically at the "tension" glyph one. From what I can tell, the outermost letter is definitely 'K' (rotated along the y-axis), but the other two lines stump me. If I try to translate "tension" to Thaylen, it would probably end up looking like "[t][n][sh][n]", but this one has 4 consonants, which suggests to me that the equivalent glyph combination will have four distinct lines... and "tension" has only three. The smallest one might be divisible, but I couldn't figure it out in a few minutes I could steal to work on this.

[lilserf]

It won't be "tension" -- it'll be the Alethi word that means tension.

[Harakeke]

Oh!  Clever!

Reminds me of how the number names (jes chach vev) concatenate to form dates (Jesachev).

[PeterAhlstrom]

The house names are usually written with a glyphpair (two glyphs) but it's also possible to write each of them as just one glyph.

[Argent]

Yea, not only did "tension" not work, none of the glyphs work in English. Makes sense, really. But there are so many glyphs, we should be able to do something with them...

Also, I am now obsessed.

[Harakeke]

The house names are usually written with a glyphpair (two glyphs) but it's also possible to write each of them as just one glyph.

Also, it seems like the shape of the glyphpairs can vary. Hasn't Kholin (kokh linl) been mentioned as being shaped both like a tower & crown, and like a sword & crown?

[lilserf]

I assume this just means that a highprince can ask his artists to design his glyph into whatever shape he wants - from generation to generation they could adapt the glyphs to whatever shapes this prince thinks represents him. So in the case of the Kholins they make a royal and non-royal version.

[Harakeke]

 

Stormfather, those are going to be a breech to figure out. I love it!

In the meantime, I took a quick look at The Way of Kings Endsheet (because the Surges have pretty clearly drawn glyphs there), and more specifically at the "tension" glyph one. From what I can tell, the outermost letter is definitely 'K' (rotated along the y-axis), but the other two lines stump me. If I try to translate "tension" to Thaylen, it would probably end up looking like "[t][n][sh][n]", but this one has 4 consonants, which suggests to me that the equivalent glyph combination will have four distinct lines... and "tension" has only three. The smallest one might be divisible, but I couldn't figure it out in a few minutes I could steal to work on this.

Yeah - of all the glyphs, those are the easiest to read. Unfortunately, we don't know what words/sounds they go with. Note that the "Voidbinding" chart on the opposite cover uses the same basic sub-glyphs, but rotates them rather than reflects them.

[Argent]

For the highprinces, I am going off the glyphs from the Alethi Warcamps, The Battle of the Tower, and The Battle of Narak. Those seem to be the standard glyphs for the highprinces' names and I've managed to identify with high confidence the glyphs for Kholin, Sadeas, Aldar, Roion, and Sebarial (even though the Alethi Warcamps map has Dalinar, Sadeas, and Roion upside-down; either that, or the two battle maps have everyone but Sadeas upside-down - either way, there is some discrepancy).

The Kholin glyphpair (khokh + linil) seems to mostly make sense at this point. I am having some issues with half of linil, but I am almost content with it. Here, let me show you. Oh, and a disclaimer - you may want to zoom in (Ctrl+ScrollUp, usually) on those images by about 500%:

This is the Kholin glyphpair, as taken from the Narak battle map.

This is the same glyphpair, but with what I think is the khokh half of it colored in red. With everything being so symmetric, it's a little difficult to decide whether the entire colored part is a single khokh, or whether only one side is the glyph, and the other is just a mirror image. I'd say that, for example, only the left (red) side says "khokh", and everything to the right of the vertical bar (which is added for decoration) is just a mirror image.

The reason I am so confident about this is because the Thaylen translation seems fairly straightforward, and their letter for 'K' is very distinctive ('Г'); which is exactly what we see here.

Finally, this has to be linil. Unlike khokh, this glyph has two consonants, so I figured it consist of two lines - one for each consonant. According to Harakeke's translation of the Thaylen alphabet, the red part closer to the central vertical line (the part I have colored blue here) is most likely the character for 'M' or 'N' - obviously 'N' in our case. So what's left, has to be the 'L' part... but it doesn't look like the Thaylen 'L' at all. Granted, the components of the Alethi glyphs are not going to be the same as the Thaylen letters, but we've seen enough similarities to expect them to be at least in the same neighborhood... And this 'L' is not.

Still, I am willing to call this a success and move on to other glyphs - until my hypothesis here proves to be so wrong, I can't do anything with it.