The Actual News:
Just a small little notice that I’ve updated the Galarian Word List with a whole bunch of different text I’ve picked up from here and there.
That said, I’ve worked out a little more about the Galar alphabet. Long story short, it’s definitely looking like the Galar language is “consistent gibberish”. Here’s the deets:
- On the train station’s platform sign, the words
and 
appear; since the first word appears after “1” and the second appears after “2/4/5”, it’s clear that those words are “car” and “cars”, respectively. As in, the train heading to [somewhere town] on Platform 1/2 will have 1 car or 2 cars. - The names used on the uniforms for the Gym Leaders Nessa and Kabu … well, it might just be their names like with Milo/Yarrow, or it might be relating to their Numbers (which are used as “goroawase” or number wordplay in Japanese). Namely, since
fits neither “Nessa” or “Rurina/ルリナ“, I THINK it might therefore be related to Nessa’s League Number, 049, which is a goroawase 泳ぐ oyogu, meaning “to swim”. Since sometimes the final U in Japanese words is silent-ish (kinda), Nessa’s Galar name might therefore be “OYOG”. Not only does it seem to fit the pattern, but the final Galar character might in turn reference the ぐ gu character. - Kabu’s name
, on the other hand, probably is just meant to be “KABU”. My only guess towards that is that is used as an “A” in Galarian “Game Freak”. - A poster in Hop’s room shows a bunch of PokéBalls on it, or as they’re known in Japan, Monster Balls, which in turn was written as
- Wedgehurst, in Japanese, is “Brassie Town“, which probably explains the double in its Galar name:
… That said, it’s probably supposed to be “BRASSY”.
So as I mentioned, these all seem to add more evidence to the “consistent gibberish” hypothesis, as these new words continue to support the idea that they make SOME kind of sense within themselves, such that they aren’t purely random characters… but that sense only applies to themselves and not to the entire Galar language as a whole.
OK, that’s all I have to share for now. Keep an eye out for more info! And if you have anything to share, please do! Let’s build a list of Galar words together!
6 Comments
From the point of view of a Japanese player, your argument for ‘OYOG’ on Nessa’s uniform doesn’t seem convincing to me, although your knowledge of the language is quite right on devoicing of terminal u’s. It seems more possible, if anything, that the 4-letter word stands for ‘RURI’, which itself means lapis lazuli and its colour, whence Nessa’s Japanese name Rurina obviously comes. However, the truth will be, unfortunately, that neither of our hypotheses is correct.
While I’m among those who had an interest in Galarian language even before the release, now I’m rather pessimistic to the decodability, since there are too many incoherent texts.
Evidently, there is no ‘global consistency’ because three types of e’s in ‘DANDE’ (Leon in jpn), ‘ROSE’, and ‘GAME FREAK’ do not coincide, even though these three names have the most definitely certain reading among all the Galarian texts.
Thus, in my opinion, we inevitably have to pursue some ‘local consistencies’, that is, a spelling could be consistent in not a few cases if limited in each single image and we cannot compare texts over two or more images.
You might have already noticed the same idea of consistency as you collected Galar words whose spellings very often contradict each other. Probably my idea can be identical to the ‘consistent gibberish’ in your terms.
If so, there is scarce use in just blindly collecting Galar words. What is required would be a change in approach. I’m suffering a setback in this very point. Do you have any good ideas?
Frankly I was just taking a stab in the dark with “OYOG”. Like, I didn’t consider it initially because neither the three character Katakana text (RU-RI-NA) nor either of the Latin text (Rurina or Nessa) fit. I mean, I guess neither of the goroawase forms fit either, but it was the only idea that made the most sense to me… relatively speaking. Ultimately I guess “RURI” could work just the same.
For the record, anytime a Japanese character with a terminal-U is transliterated into Korean/Hangul, we use two different vowels for it depending on what it originally was. Specifically, we have a “hard U” vowel (ㅜ) as well as a “soft U (-eu)” vowel (ㅡ)… the latter vowel is used for words like 泳ぐoyogu, thus 오요그 versus 오요구. It’s not REALLY a purely “silent-U”, but it’s definitely not stressed; the Korean vowel -eu is also used for English words which end in a consonant in a similar manner to Japanese… you probably get what I mean.
>> now I’m rather pessimistic to the decodability, since there are too many incoherent texts.
Yeah, I’m with you on this. However, this hasn’t diminished my interest in it, but rather increased it. Like, to me there is clearly some kind of system they used when generating words, because even if they was no UNIVERSAL guide, there was still some kind of methodology behind it. But what? Did the top brass just give the graphic artists free reign to use whatever characters seem to fit? Is there maybe a way to detect “Artist A’s” style and preference from “Artist B’s”, in the same way the New Testament Gospel Books of Matthew and Luke pulled text from the Book of Mark as well as a mysterious “Q Source“?
OK, maybe I’m over-hyping the significance of all of this… but honestly, I feel like there might be something that can be pulled from understanding the Galarian text, even if it won’t be any ACTUAL reliable one-to-one translations.
>> If so, there is scarce use in just blindly collecting Galar words. What is required would be a change in approach. I’m suffering a setback in this very point. Do you have any good ideas?
TBH, I can’t think of a change in approach which would work… because there is nothing that can be gleaned from just looking at the text at face value. Like, it seems as if everyone has done the “find a Galar word, see if it is related to another Galar word” thing to death and it’s gone nowhere.
This is therefore why I still want to collect Galar words in general. Partially it’s for character frequency purposes, as that would benefit from making sure I don’t accidentally repeat words which would skew the data. But most importantly, I feel like collecting them will definitely lead to something bigger later on, because at least having them all together in one page might make spotting patterns a bit easier. It’s definitely a step up from just comparing screenshots together, especially since I’m using text-based emoticons which makes using a text editor to search for pattern searching easier.
That said, I’m open to whatever ideas anyone else has about this. While I don’t mind typing out Galar words for now, I’m also not keen on doing needless work if someone else worked out something more efficient.
>> to me there is clearly some kind of system they used when generating words, because even if they was no UNIVERSAL guide, there was still some kind of methodology behind it.
I agree with you, or, to be more exact, I feel like believing that there is some kind of laws behind it. One thing is for sure, it cannot be a simple substitution cipher.
While it’s plausible that the graphic artists didn’t have ‘free reign’ for the aesthetic sake, it must be pointed out at the same time that they did make some errors and commit negligence in constructing Galarian words.
In fact, the text on the signboard above the information desk (next to the food court) in central stations (Motostoke and Hammerlocke) reads ‘GALAR LEGAUE’, instead of ‘LE*AG*UE’, compared to Champion Cup posters and so on. This is an example both of careless error and of lax negligence.
Another example of negligence is found on the menu boards in the cafés. The headline of the left one is the same as the signboard of hair salons and there is of course hardly any relevance between hair salon and café menu.
Thus the biggest question before us is that we cannot tell which Galarian words are authentic and which are erroneous or irrelevant to their contexts. As to the last example, does that 10-letter word have connection with hair or coffee or neither?
>> Partially it’s for character frequency purposes, as that would benefit from making sure I don’t accidentally repeat words which would skew the data.
This is exactly what I was worrying when I read the frequency analysis on your earlier article. Since due to the artists’ negligence there are quite a few identical texts, it’s not appropriate for an accurate analysis that we count up all the ‘raw’ Galarian letters as they are. In case a certain word with Z were repeated many times, perhaps we would be liable to identify the letter as E or T mistakenly.
Anyway, it’s true that the effort to collect Galarian words will work as a prerequisite for detecting those errors and negligence. Your endeavour will not prove futile. Likewise, I’ll try to tackle this issue once again from the beginning.
>> I feel like believing that there is some kind of laws behind it. One thing is for sure, it cannot be a simple substitution cipher.
Well I’m glad I’m not the only one to realize that SOMETHING is going on here. Frankly it would’ve been far less fun and challenging than if it WAS just a simple substitution cipher!
…
>> Thus the biggest question before us is that we cannot tell which Galarian words are authentic and which are erroneous or irrelevant to their contexts.
This is one of the reasons why I want to catalog Galar words; if I/we can at least see where a certain words appears more often, it might be easier to determine which context is the error case.
For example, maybe right now that 10-letter word might not fit in either a hair salon or café situation… but if after—say—eleven more appearances of that 10-letter word and it showed up at the hair salon in all those other cases, then its appearance at the café was an error. OR it might just be that that 10-letter word actually appears at Pokémon Stadiums and thus BOTH of its salon and café appearances were the error.
More data helps clarify a situation. Speaking of which…
…
>> it’s not appropriate for an accurate analysis that we count up all the ‘raw’ Galarian letters as they are
I’m legit not too worried about that… but don’t get me wrong, that has crossed my mind. The thing is, unless it’s FULL of errors, one or two bad examples can be rendered less and less statistically relevant with more and more data.
Alternatively, if the errors actually DO make up most of the examples, then it just means the accurate/consistent use of Galar by the game’s artists wasn’t something that they strove much for.
In any case, my quest for ANY answers means that I’ll be satisfied with whatever answer I get. And whatever that answer may be, more and more data can only help bring that into sharper focus.
MOAR DATA
This is something that struck me while decoding. But first, I agree that there is a local consistency where each Galarian letter seem to look similar to a certain Latin letter. However I do see that many local consistencies do appear similarly, so even if we can’t create a global consistency we will be able to end at something inbetween local and global. If that makes sense. So, the thing that struck me is:
• that the building next to Motostoke’s eastern Pokecenter is the fire station (deducted from PokéJobs logos and names), and the text on it is readable somewhat as FAJESTAESHUN. The STAESHUN-part is somewhat trickier to make out and I am not happy with it. The backwards galarian R equals “ST”, the second last galarian letter is “SH” and he last galarian letter is “U” (but with a silent N, which appears in this case).
• Motostoke’s stadium text on the outside is readable as “Cosmos Stadium”. The backwards R make a reappearance as “ST” and the second last galarian letter returns as “U” (this time without the silent N).
• Hammerlock’s gym reception text above the two “receptionists” is readable as “RISUPSH(U)N”. This time the backwarss R is “R”, then “I”, “S”, “U” (it’s back again), “P”, “SH” (it’s back again too) and “N” (backwards R being an N this time). It’s not pretty, and maybe this one is forced, but it aligns somewhat with how the galarian letters could be used if the only requirement is that they somewhat look similar to a letter and that the sum of all letters sounds like the phonetics or the “japanization” of an English word.
Interesting observations! I didn’t think about some characters representing actual phonemes! But now that I think about it, it might help explain why there’s 39 characters; consider something like Aurebesh from Star Wars, and although it’s a cipher script of English for the most part, it did include some extra characters which aren’t normally in English, like “SH”, KH”, and “NG”.
In that case, while it might look weird, it could also be that the language—or at least the artist who made up those particular examples—could have wrote it to be more phonetic, in de saim wai this sentenss iz ritten.
Either way, it’s definitely starting to look like the meanings of the various examples of Galarian depend on whichever artist decided to design that particular instance of it. Now I wonder if they told the CG modelers “this picture means XXXX, that picture means YYYY” when they delivered their pictures… or it was the other way around, where the bosses said “give me some text for the reception desk” or “give me some text for the fire station”.
But yeah, thanks for your post! This really is something to add to the pile. Maybe when everything is said and done, I’ll make a video about it!