[OB] Glyph Translation Discussion (No Plot Spoilers)

[ccstat]

So I've taken another crack at the Taker of Secrets image. It's tentative, but I think I've got a lead. If I'm right, it contains the word "corrupted" twice.

First, I compared the new glyph phonemes from OB with the similar Thaylen alphabet we derived from the frostlands map in WoR:

As you can see, the majority of the letters match quite well. We got new phonemes not represented on the map, including vowels (which aren't written in Thaylen). EDIT: I've now confirmed our assignments from the frostlands map, concluding that the differences below reflect a divergence in the two writing systems, not errors on our part

• Blue: We assigned multiple English letters (e.g. J+G) to certain Thaylen letters, whereas the glyph phonemes have distinct symbols for each.
• Orange: The Thaylen letters for T and D are reflections of each other. In glyphs, which can reflect and rotate phonemes freely, that symbol is assigned to Th (not a unique letter in Thaylen), while T and D are assigned to related symbols.
• Green: The symbols for Sh and for Z are altered slightly between the two systems.

For making character assignments I used the OB phoneme chart, but supplemented it with our Thaylen deductions because the reflected T and D characters are both in evidence and vowels are entirely absent. Assuming the character orientation aligns with the reading order, it does appear to read from top to bottom on both sides, and left to right across the top, but right to left across the bottom. Multiple characters are flipped other directions though, so that assumption is suspect. Still, here is my transliteration using that convention:

(original image)

Spoiler

Left side:BSFZThKRRFThThThFSh

Top: KRRFThTh...VLFFThB

Right side: BNSFNThKFRFSK?ThSThN

Bottom: ThRKNSS

The lines assigned as B only occur at the corners, and could just be decorative to connect things visually. Where a Th occurs, I've formatted them by their orientation, with bold corresponding to the Thaylen letter T, and italics to Thaylen letter D.

(colored image, with color-coordinated text -- I'm sorry about the ugly lines. My image manipulation skills leave a lot to be desired)

Spoiler

Left side:BSFZThKRRFThThThFSh

Top: KRRFThTh...VLFFThB

Right side: BNSFNThKFRFSK?ThSThN

Bottom: ThRKNSS

The lines assigned as B only occur at the corners, and could just be decorative to connect things visually. Where a Th occurs, I've formatted them by their orientation, with orange corresponding to the Thaylen letter T, and yellow to Thaylen letter D.

I recognize that the color-coding ended up being overkill. Hopefully between that one and the original image you can see how I got to what I have.

I have concluded that the writing system in question is closely related both to the glyph phonemes and to the Thaylen writing of the frostlands map, but with some important differences. In particular, this text seems to freely reflect and rotate characters like the phonemes chart, but also utilizes the vowel omissions and character redundancies of Thaylen (e.g. P=F), with the additional redundancy of T=D=Th. 

If you are interested in helping to decipher these strings of consonants, these website may prove useful. OneLook is a dictionary search that allows wildcards, and Litscape provides soundex matching. (For the latter, note that soundex ignores repeated consonants, while this Thaylen text does not.)

This box shows my current guesswork in table form. Bolded words are those in which I have the most confidence. The top line is interpretation. The second line is the transliteration from the image, with subsequent lines providing alternative letters to account for the phonetic redundancies.

Spoiler

 

Co

R

Ru

P

Te

D

Vi

Le

oF

K

R

R

F

T

T

?

L

F

F

T

B

C

P

D

D

V

P

P

D

W

Th

Th

Th

V

(-)

S

Pie

S

iT

Co

R

Ru

P

Te

D

The

Pa

Ss

(ion)

B

S

F

Z

T

K

R

R

F

T

T

T

F

Sh

W

P

S

D

C

P

D

D

D

P

?

V

Th

Th

Th

Th

(-)

?

o

N

Ce

Ta

Ke

Pu

R

Po

Se

Co

N

Te

S

T

B

N

S

F

N

T

K

F

R

F

S

K

?

T

S

T

N

W

P

M

D

C

P

P

C

N

D

D

M

V

Th

M

Th

Th

(-)

Da

R

K

Ne

S

S

T

R

K

N

S

S

D

C

M

Th

 

This spoiler box contains previous guesswork and explanations, superceded by the above, though the formatting may be easier to read.

Spoiler

Here are some tentative guesses at interpretation. I think "Darkness" is correct for the bottom row, so I'm fairly well convinced that we are using Thaylen T and D in place of the glyph Th symbol. If so, some of the other Thaylen letter substitutions may be appropriate to apply, such as P for F, V/W for B, or M for N. Your suggestions are most welcome.

EDIT: While I still think T and D are valid assignments for the Th symbol, I am now less convinced that the orientation is an explicit clue, since it requires using them interchangably to get "corrupted" on the top and left, which I also think is likely to be correct. Note that this also uses a P for F substitution.

Left side:

B SFZ Th KRRFThTh Th FSh

Spies too corrupted to ____

Top:

KRRFThTh...VL F FThB

Corrupted...Vale of Fate

Right side:

B NS FN Th KFR F SK?ThS ThN

Knows pain the keeper of secret stone

Bottom:

ThRKNSS

Darkness

 

More Previous guesses

Spoiler

Left side:

BSFZ ThKR RF ThTh ThFY

_____ Taker __ Did Defy

Top:

KRR FThTh...VL F FThB

___Faded...Vale of Fate

Right side:

BNSF NThKFR F SK?ThSThN

____ Nature of __________

 

EDIT: I'm considering the possiblity that the points where symbols flip orientation (the Fs do this a lot) are indicative of word breaks, but I haven't really looked closely at that yet.

Edited December 19, 2017 by ccstat
new corrections and translation guesses

[Eri]

I posted it in previous thread but then I noticed this one is more alive, so… Some content back from WoK. Have fun with this! (I can't figure out those glyphs, but they look readable. Their origin involves a polar coordinates filter and the voidbinding chart. They may have flipped left half and right half, can't be sure.)

[Emily]

@Jofwu, I had a brief attempt at translating Kaladin’s chapter header.  I read it as O Ch I Ch O, but that assumes it is symmetrical and the I is highly distorted.

Edited January 3, 2018 by Emily
I added the actual chapter header

[Blazenella]

I've been staring at the sword stance glyph page for way too long and with the Calligraphic Glyphs it's almost impossible to decipher. In the "Title" glyphs, more specifically the second one, I could easily see a "Z", "V" and a "D"  something that could be an "M" "N" "H" "T" "L" or some other random thing, something that might be an "A" (Every glyph with a dot of some kind has had an A in it). I honestly can't tell if there are any other letters (Such as a B crammed in there or if it's just some rando F U lines. Looking at some more of the Glyphs you can recognize some of the phonemes, but obviously not all of them seeing as the Calligraphic can seriously screw with you.

The closest translation I got would be Vadzt though I really don't know with the placement of phonemes and the Calligraphic Phonemes

Also, does anyone know how glyphs are meant to be read? We know that you write the Glyphs out and the just mirror it, but is it read top to bottom? Right to left? And what about lines that run throughout the glyph? 

I'm gonna keep trying to translate some of these.

This is the glyph I'm talking about:

Edited January 3, 2018 by Blazenella

[Jofwu]

@Blazenella They seem to generally be read from top center to bottom outside, but there also seems to be some freedom involved.

The main problem here is that it represents an Alethi word. Your guess could be 100% right and we simply don't know what the word means.

[Blazenella]

@Jofwu Thanks for the clarification! Based on context it could mean stance seeing as it's a sword stance scroll and it's the "Title" of the scroll

[Stilgar12]

This isn't exactly helpful with translating anything, I just wanted to share what y'alls hard work has enable!

I put a post with more pictures over in the creators section. Thanks for all y'alls hard work! 

[Furry-And-Lovable-Grover]

On 11/15/2017 at 4:30 PM, Pagerunner said:

Airship translation:

Painrial translation:

Not quite sure on the first line of the airship one... I've triple-checked the letters, but unless there's  space missing in that second word, I don't quite get it.

Isn't there supposed to be a glyph or character that's a place holder to allow for symmetry? I thought it was supposed to be 'ha' but I don't really remember. I want to say that one of the characters talks about it in the text?

[ccstat]

Quick update: not much additional progress on the Taker of Secrets image.

I've spent a while puzzling over the border text translation (started in my post up above). While there are some good candidate words or phrases, I haven't been able to convince myself of many of them. The main difficulty is that the letter redundancy and lack of clear word breaks give a lot of freedom. Beyond simple guesswork, I have no good way to decide how close a given translation is to being correct. There isn't very much  to provide contextual clues.

For example, look at the left side. The transliteration that precedes KRRFTT (=corrupted) is BSFZT. Using the possible letter redundancies, both "- spies the" or "was fused" are equally valid, along with several other interpretations. I'll welcome any suggestions on candidate translations or brainstorms on how to approach decrypting it.

 

[Wonko the Sane]

I've been looking at the writing around this map: 

I've isolated the writing here, and adjusted the levels to improve readability. Because the regions outlined on the map are pretty close approximations of the Silver Kingdoms (albeit distorted ones), I've been working under the hypothesis that the writing translates as the names of those kingdoms, so that's how I've labeled them.

My first observations are that the language appears to use a lot of diacritics, and that either the first or last letter of each word (depending on reading direction) seems to be rotated or modified so that a trailing end of it strikes through the rest of the word.

The easiest to read, assuming I'm right about the translations, seems to be Natanatan:

There is a mark over the last (or first) 'N' which might be a diacritic of some kind, but which I'm unable to guess the meaning of. Regardless, I think this is looks to be about right.

I've done some work at identifying characters in the other words, but Natanatan is the only one I've got anything close to a complete idea of. I'll try and post everything I have later, but hopefully this will be useful for now.

Edited January 26, 2018 by Wonko the Sane

[Eri]

Wow, I didn't even notice there was a script on that image, great job!

 

I've tried to trace/clear the writings (white is for lines that are probably waves and may make other lines hard to read, dark blue for something that seems a main line / initial, colors are for letters that look similar, blue for other letters). Also, I've added the original location names, just in case. I hope this helps a bit.

 

Edit: I've tried to match the letters, but I'm too bad at this. 

Edited January 26, 2018 by Eri
new image

[Jofwu]

I've been wondering about these.

We know from in-world Words of Radiance that Urithiru was built/placed after/around the formation of the Knights Radiant. I say this because the script is very odd. I think we have evidence that symmetrical glyphs existed BEFORE the Silver Kingdoms and the Knights Radiants. My first guess was that this script might predate glyphs, but I don't think that can be the case.

The strikethrough line is very interesting and distinctive. It seems very reminiscent of women's script to me, which makes me wonder if this is an ancestor of that script? The curvy lines of course are definitely reminiscent of glyph graphemes of course.

I think the Silver Kingdom names are a very good guess. I wonder two things though... One is whether we have the actual names of those kingdoms. We have the VORIN names for them. Dalinar heard a few in his visions I think, but this could be a "translation" thing that we wouldn't get. And he didn't hear all of them. It's possible they went by something slightly different than what we know. The symmetry in particular is suspect I think. Also, we don't know how those names changed over time. Alethela eventually morphed to Alethkar. When did that happen in relation to this map's dating? Or perhaps I should say in relation to the writing on the map. It's possible that the notations we're trying to make sense of were added well after the drawing itself was created. They look like they were added afterwards, to me. (not to confuse them with Nazh's notations of course-- yeah, I know the "English" ones are him)

The leading N in Natanatan bears some similarity to the following two Ns, which makes me think that the letters change form depending on their position in the word. The script looks very Arabic, so that's another thing they would have in common. And that would mean we shouldn't put too much stock in comparing the leading characters to other instances of them.

I was wondering if the vowels might be full characters, and then the word is read backwards and forwards. So Natanatan would be written NATAN or Alethela would be written ALETH. Just another concept I thought I'd throw out. It seems possible in some instances, but in others it makes the diacritics more confusing and it doesn't work great for other names.

[Wonko the Sane]

On 1/26/2018 at 4:17 PM, Jofwu said:

We know from in-world Words of Radiance that Urithiru was built/placed after/around the formation of the Knights Radiant. I say this because the script is very odd. I think we have evidence that symmetrical glyphs existed BEFORE the Silver Kingdoms and the Knights Radiants. My first guess was that this script might predate glyphs, but I don't think that can be the case.

I originally suspected this was the Dawnchant; like you, I thought the image felt older than others. Now, I think this is one of Ellista's "in-between, weird languages" (OB I-1, hardcover pp. 335) where different cultures used a Dawnate script to transliterate their own tongues.

On 1/26/2018 at 4:17 PM, Jofwu said:

I was wondering if the vowels might be full characters, and then the word is read backwards and forwards. So Natanatan would be written NATAN or Alethela would be written ALETH. Just another concept I thought I'd throw out. It seems possible in some instances, but in others it makes the diacritics more confusing and it doesn't work great for other names.

A-HA! Thank you! I was having problems with asymmetry, but you've solved that for me. I'm now reasonably certain I've correctly translated the entire sample. Here's the final product:

Rishir:

Spoiler

Valhav:

Spoiler

Alethela:

Spoiler

Natanatan:

Spoiler

Thalath:

Spoiler

Makabakam:

Spoiler

Aimia:

Spoiler

Shin Kak Nish:

Spoiler

Iri:

Spoiler

Sela Tales:

Spoiler

 

Note that for all but one kingdom, there is a diacritic of two dots over the final character. I suspect that this indicates that the word should be mirrored for symmetry. If this is true, the different diacritic used in Valhav may be to indicate the additional 'H' before mirroring.

Also note that for Sela Tales, I'm mostly guessing, as large sections of the writing are covered up by the coastline. Similarly, I'm unsure about the first letter in Aimia. I've labeled it 'A', since that fits the word, but it doesn't resemble any other 'A' we have.

I'll try and put together a translation key with what I've got here, but I unfortunately don't have the time for that right now.

EDIT: Oh, one last thing. For the most part diacritics appear over vowels, probably indicating how the vowel is voiced. The exceptions are the two dots at the end of most words, the special diacritic at the end of Valhav, and the '>' over Valhav's 'V'. I'm not at all sure what that last one might mean.

Edited January 30, 2018 by Wonko the Sane
It was Aimia where I had trouble with the A, not Alethela.

[Jofwu]

@Wonko the Sane, brilliant work! @Harakeke, I think we're going to need a character chart for yet another Rosharan script. 

Assuming this is some kind of in-between script, bearing similarities to both glyphs and women's script, was curious if I could find any connections... Turns out there are 

Compare the glyph "I" phoneme  to the "I" in "Rishir", "Aimia", and "Shin Kak Nish". Practically identical, minus diacritics.

Compare the glyph "R" phoneme  to the "R" in "Rishir" and "Iri". Practically identical, just more script-y.

Compare the glyph "SH" phoneme  to the "SH" in "Rishir" and "Shin Kak Nish". Have the 'v' portion in common?

Compare the glyph "V" phoneme  to the "V" in "Valhav". Ignoring the weird diacritic, or merging them, gives you something similar. Edit: Whoops, that's the "U" phoneme. Don't see anything that links that one to the "V" in "Valhav".

Compare the glyph "N" phoneme  to the "N" in "Natanatan" and "Shin Kak Nish". Similar, if you turn top slant more script-y and add a twirl to the bottom one.

Compare the glyph "M" phoneme  tot he "M" in "Makabakam" and "Aimia". If you cut off the stem on the left side of this script, the resulting hump and long tail seem to be in common.

Not sure what to make of how similar "T" and "K" seem to be in this script. I suppose the glyph "K" phoneme  bears some similarity to a portion of the "K" in Makabakam.

The letter "E" in this script is essentially the same as "I" with a different diacritic. The glyph "E" phoneme  is sort of a warped version of the "I" OR a warped version of the (apparent) diacritic linked with "E". 

The glyph "A" phoneme is just a dot or small circle, compared to the full letter. But it's worth noting that one of the "A" varieties here uses a single dot as the diacritic.

I don't really see much to link "T", "TH", "L", or "S".

Also don't see any similarities to women's script beyond the whole center line running down the length of it thing.

[Wonko the Sane]

Hah. Beaten to the punch. I was going to point out many of the same parallels. It does make sense that these characters would appear in the glyphs; we knew already that the Dawnchant had a large influence on the early glyphs (and until proven otherwise, I'm going to operate under the assumption that this script is Dawnate). We may be able to use this to make guesses at the as-yet-unknown "Calligraphic Phoneme Set".

I think it's noteworthy that a Dawnate 'M' appears very plainly in Tsameth, along with a possible 'Th':

Nazh may be right that glyphs like these are older, and possibly adopted from another language, but I think he may be a little too quick to dismiss them as "unreadable".

Edited January 31, 2018 by Wonko the Sane

[Jofwu]

43 minutes ago, Wonko the Sane said:

I think it's noteworthy that a Dawnate 'M' appears very plainly in Tsameth, along with a possible 'Th':

Looks like a 'T' to me in there as well.

[ccstat]

Amazing, @Wonko the Sane. I spent a long time staring at those symbols yesterday and didn't get anywhere useful. Great work! 

[Subvisual Haze]

Brilliant work!  Talk about conservation of vowels, there appears to be only a single base symbol for all of the front vowel sounds (a e i), with only the diacretic marks to distinguish them from one another!

It reminds me of how Hebrew was originally written with only consonants and later scribes added in vowel marks to the words.  Is this a sign that the Dawnchant originally lacked vowels in its written form?  It would explain the difficulties in Thayley with vowels.

The diacretic over the "v" in valhav may indicate that the word originally started with a vowel sound?  There is some crossover between current languages having variable "v" or "w" sounds (like poor Chekov in Star Trek IV looking for Nuclear Wessels).  And "w" itself is a semivowel, so perhaps the city started with more of an oo sound?

The wierd tilda over the first vowel in Aimia is probably a diaresis.  It indicates that the "a" and "i" at the start of Aimia are two distinct vowel sounds and should not be combined into a single diphthong sound "ai" (like in the word "aisle"). Ah-ee-mee-ah instead of Ai-mee-ah.  It preserves the pronunciation symmetry, two vowel sounds at both the start and end of the word.

[Wonko the Sane]

4 hours ago, Jofwu said:

Looks like a 'T' to me in there as well.

Nice! That's nearly all the consonants, and I don't know how much luck we'll have with 'S', since we only have the distorted "capital" form.

11 minutes ago, Subvisual Haze said:

Brilliant work!  Talk about conservation of vowels, there appears to be only a single base symbol for all of the front vowel sounds (a e i), with only the diacretic marks to distinguish them from one another!

Not quite. 'A' and 'I' each have distinct symbols (Not counting whatever's going on with Aimia). 'E' appears to be divided between the two, depending on pronunciation.

13 minutes ago, Subvisual Haze said:

It reminds me of how Hebrew was originally written with only consonants and later scribes added in vowel marks to the words.  Is this a sign that the Dawnchant originally lacked vowels in its written form?  It would explain the difficulties in Thayley with vowels.

Possibly... but the symbols are fairly involved to be later additions, and remember, Thaylen didn't evolve from the Dawnchant. Both Dawnate symbols and a proto-Vorin ancestor were assimilated into the glyphs, but the languages themselves come from separate families.

16 minutes ago, Subvisual Haze said:

The wierd tilda over the first vowel in Aimia is probably a diaresis.  It indicates that the "a" and "i" at the start of Aimia are two distinct vowel sounds and should not be combined into a single diphthong sound "ai" (like in the word "aisle"). Ah-ee-mee-ah instead of Ai-mee-ah.  It preserves the pronunciation symmetry, two vowel sounds at both the start and end of the word.

Maybe so, but my main issue with that character isn't the diacritic, but the figure itself. It doesn't align with our other example of how 'A' should be capitalized, nor does it look like other 'A's -- or indeed, any other character. I could be persuaded to see a highly distorted, backwards version of the 'A' character, but the question remains: Why isn't it capitalized like in Alethela? Why does it flip backwards? I'm basically at a loss here.

[Subvisual Haze]

Oh, Sela Tales isn't starting with an "s" it's a "z".  Probably to mirror the later sound, English is terrible about pronouncing terminal "s" sounds as "z" sounds.

[Wonko the Sane]

22 minutes ago, Subvisual Haze said:

Oh, Sela Tales isn't starting with an "s" it's a "z".  Probably to mirror the later sound, English is terrible about pronouncing terminal "s" sounds as "z" sounds.

Huh? Why do you say that? It doesn't match 'z' in any other script, as far as I can tell.

Have I been pronouncing the name wrong? I say it as "SAY-la TALL-lace.

Edited January 30, 2018 by Wonko the Sane
Made pronunciation clearer.

[Jofwu]

So Isaac recently drew out the glyph for determination, KHAKH:

The 'KH' phonemes are well known and easy to spot here, leaving an 'A'. Guess what it sort of looks like...

[ccstat]

FYI, I've been in touch with Isaac via reddit and pointed him to this thread. He apparently dropped in to look at it sometime yesterday:

Quote

That's really fun to see you all deciphering the glyphs and writing systems over on the 17th Shard! Thanks for pointing that out to me!

He also said:

Quote

I double checked with Miranda, who created the Taker of Secrets piece. I can confirm that the word "corrupted" is in the border.

That would be Miranda Meeks, who also did the Shadows for Silence cover art and some fabulous Mistborn pieces. Sounds like we are on the right track there.

[Subvisual Haze]

22 hours ago, Jofwu said:

Compare the glyph "V" phoneme  to the "V" in "Valhav". Ignoring the weird diacritic, or merging them, gives you something similar.

That's a "u", not a "v" though, right?  The "v" phoneme is the third on the page, right next to "b" and has a very different shape that doesn't match the first letter of "Valhav"'s well at all. 

Which might explain the diacretic at the start of "Valhav"?

Because for the most part our diacretics are only over vowels.  The exception is the double dot at the end of the words indicating a point of symmetry.  "Valhav" is special here too though, as there is a bizarro otherwise unseen diacretic over the "L".  Perhaps it indicates a point of imperfect symmetry?  Valhav, you're the worst...  

Otherwise though the "~" tilda symbol above a vowel seems to indicate that the vowel is long.  A single dot seems to indicate the vowel is short.  And I think the marks that look like ">" or "<" indicate pauses or aspirates?  If Roshar languages are inspired by semitic languages, we are sorely lacking hard aspirate "h" sounds in our words

[Jofwu]

15 hours ago, Subvisual Haze said:

That's a "u", not a "v" though, right?

Thanks, good catch. Yeah, definitely no similarity there.

So "double-dot" is the symmetry indicator.

Looks like "single-dot", "tilde", and "less-than" appear on vowels only. I guess we really only have two vowel characters shown. Call them 'A' and 'E'. I see all three of these used on 'A', presumably giving a short "ah", a long "ay", and... something else? I see 'A<' used in "Valhav", "Makabakam", and "Tales". Maybe it's this? I suppose in "Kak" it doesn't have one of these. I expect one of them is assumed if you have the double-dots. I see all three used with 'E' as well, if we can assume the "Shin" one has a dot over it? I'm guessing that 'E`' is a short "ih", 'E~' is a long "ee", and 'E<' is a short "eh".

In summary, that's:

• A` = /æ/
• A~ = /eɪ/
• A< = /ɑː/ or /ɒ/
• E` = /ɪ/
• E~ = /iː/
• E< = /ɛ/

Then there's the weird marks of "greater-than" and "circumflex", both on "Valhav". I'm not going to try to understand the 'V>'... But I feel pretty comfortable assuming that the 'L^' is an aspirated 'L', and that the diacritic generally means an aspirated consonant.

Also wanted to point out, just in case it wasn't noticed, that while 'T' and 'K' appear to be the same symbol, the 'T' straddles the strike-through while the 'K' sits beneath it.