PAX East 2018: Final Fantasy XIV’s main scenario writing and localization

Oh, these bugs.
It’s easy to have little to no idea about how Final Fantasy XIV is localized. Obviously the localization team has advanced beyond the days of Final Fantasy Tactics (which apparently was translated by someone with Babelfish and a rampaging hangover), but it’s still pretty easy to picture the localization as a matter of the Japanese staff dropping a stack of untranslated text on someone’s desk with a laugh and a note to have fun figuring it out.

The note, presumably, would also be in Japanese.

This is not just wrong, but it bears absolutely no resemblance whatsoever to the actual localization process. I had the chance to talk with main scenario writer Natusko Ishikawa and localization lead John Crow, who helpfully went into some details on both the localization process and their personal feelings about the story and characters therein. You can also check out the embedded footage of the panel below, which goes into more detail on the writing process.

The localization, as explained in the panel and as explained by both Ishikawa and Crow, is a collaborative process rather than a direct one. While Ishikawa might be the one to write the original line, any ambiguity is discussed with Ishikawa before deciding on a translation, and the goal of the teams is entirely on preserving the emotional response on both sides. That means that any parts of the script you do or don’t like in English have the same intended meaning in Japanese; it’s just a difference of implication and phrasing.

Of course, that doesn’t mean that there are no differences, per se; it just means that those changes come down to something other than “how accurate is this.” For example, while Ishikawa stressed that she found no characters particularly difficult to localize, Crow said that he found Urianger particularly difficult to localize simply because the easiest way to render his particularly formal dialect in Japanese is via archaic English, which he has a hard time writing. The characters from 1.0 in general often require a gentle touch, just to keep their localized voice consistent.

We also talked about the split between the game’s lore team and the scenario team; as Ishikawa framed it, her job was to write the story, while the lore team provided the larger context. That being said, neither one is “more important” than the other; the story needs to go where the story needs to go, and both teams work with one another to help ensure that they work together instead of apart. You may be noticing a theme of interconnection here.

Of course, for Crow’s job of localizing, it’s all a big pile of stuff that needs to be made accessible in a different language, so it seems a bit more uniform.

Both of them were quite happy with the reception that Stormblood has received from fans, with the new characters being received very positively on a whole and players excited to explore lands beyond the boundary of Eorzea proper with the Eastern portion of the story. Every content release gets lots of feedback, and as always, the team is well aware of memes (such as frequent “oh no, he’s hot” cries about Zenos yae Galvus).

This is probably fine.

Of course not all of that feedback is positive, but Ishikawa stressed that the team doesn’t look at individual points of feedback to alter the overall arc. If players don’t necessarily like a character, that isn’t taken as a sign that the direction of the story needs to change or that said character should be abandoned; it means that the portrayal of that character needs to improve.

Crow has, however, been somewhat surprised at the amount of backlash Lyse received over her role in the story from some fans. (It’s hardly universal, but it’s present.) Ishikawa, for her part, has found that player reactions usually match what she expects, in part because she loves playing the game herself. She does often need help writing parental dialogue herself, so she worries that some such interactions which she pens might seem a bit off.

Neither of them, again, was surprised about how many people found Zenos pretty.

Of course, Stormblood also had a much more actively rotating cast of characters than Heavensward, which was both unintentional and intentional. In a broader sense, it was about serving the needs of the story, but in a more specific sense it spoke to a different sort of Final Fantasy game. Where Heavensward had a much smaller cast traveling with you, Stormblood hearkens to games with a larger rotating lineup. (Think Final Fantasy X compared to Final Fantasy IV.)

Ishikawa also talked a little bit about character death. Sometimes, the story requires characters to die; sometimes that affects sidequests and needs NPCs to be rewritten based on where you are in the overall story. It’s inevitable. That having been said, a character’s survival or death is never contingent upon sidequests, but upon the needs of the story.

That comes into play looking at Haurchefant and Gosetsu’s respective death scenes, which happen to both occur at around the same level and at similar points in the overall story. The parallels there, though, are coincidental. Haurchefant’s death was clearly a shock and a tragedy, the logical extension of his life lived as a knight but not something he sought. Gosetsu, by contrast, had spent most of the expansion stating that he expected to die; his death was something he wasn’t surprised by, but something he stepped up to embrace.

When asked about a wish to introduce certain characters earlier in the hopes of players growing more accustomed to a character, Ishikawa made it clear that there are always going to be elements of that, but there’s also always more to explore. It’s the sort of thing best considered after the story is done, not before.

I said that he uses one, not that it does him any good.

Ishikawa also clarified the flow between story content and dungeons. As she puts it, much like many of the parts of the game, it’s a collaborative effort; the scenario team figures out the overall arc of the story, and that includes discussions about which dungeons or primals will be required to fill out the content. There needs to be a backbone to tie all of these concepts together, after all. This can also be important if, say, a hard mode for a given dungeon is already being planned or developed. It’s a process of working together, not in isolation, and the story team does not dictate dungeon flow unilaterally; nor is the inverse true.

When asked about influences on their writing, Ishikawa stressed that there were many influences on her work, but also that she was heavily influenced by other Square titles including several of the company’s classic RPGs like Crono Trigger and the Seiken Densetsu series. (Better known as the Mana series in the US, with Secret of Mana serving as its highest-profile release.) Crow, meanwhile, pointed to lots of outside influences like the Discworld novels and – obviously – the musical Hamilton.

He stressed that when you’re writing a story about liberation, you sort of have to reference Hamilton. He wants to avoid making the game’s localization feel too insular, giving a wide range of touchstones for players to enjoy.

So, given Ishikawa’s avowed love of playing the game as well as writing for it, what job does she play as her main? “Bard,” she replied with enthusiasm. That’s why she’s never done any of the writing for the Bard storylines; she gets to see them fresh. Crow admitted that most of his play time is on the internal test server, but he personally favors Dark Knight while also being the one who tends to localize those quests.

Last but not least, I asked Ishikawa about future story content. She just laughed and said that there was more to come, but if she gave anything away, Yoshida would be angry.

“Everyone dies,” Crow added. “Except Gosetsu. The kami are still not done with him.”

We’d like to thank both Ishikawa and Crow for their time and answers to our questions.

Massively Overpowered was on the ground in Boston for PAX East 2018, bringing you expert MMO coverage on everything (and everyone!) on display at the latest Penny Arcade Expo!

No posts to display

newest oldest most liked
Subscribe to:
Alexander Dragonfang

And this might very well be one of the worse readings the game has to offer.

But is not fault of the writing alone, the main story altogether is bad, as it tries to convey so many things at the same time, gets none of them across. Doma was a mistake, should have been, its own expansion, or mid expansion release maybe.


I think they need to find an established author and pay them some money. The writing is objectively bad.


Full of grammar mistakes? Misspellings? Wrong words chosen?

If it’s not any of the mechanics of the language that are bad, it’s not objective, it’s subjective.


The writing, not the grammar or spelling. There’s a difference. Sentences are added for no reason other than to extend the length of the paragraph, whole blocks of text are written for no conceivable reason. It’s like they are being paid by the word and are determined to maximize their pay.

Writing has many objective purposes; to convey information & to amuse your audience are the two most important ones here. Unless someone is amused by a deluge of words conveying little to no meaning, it’s an utter fail.


Sentences and dialog are chosen to convey a certain emotion, a certain pacing in a scene, a certain idea, a certain character element/trait…and often to convey a mix of some or all of these at once. Choose any sentence or phrase in the game – literally any of them – and I bet I can describe why it was put in the game and what context it adds to the scene, and that it wasn’t just to up the word count.

I’m not musing idly about it either. I literally want you to choose a random phrase or sentence, and I guarantee you’ll have the context for it for any sentence you choose.

This ain’t high school English where the lazy kid in the back is rephrasing the same sentence to meet a three page requirement, this is the professional world. Honestly, at the pace SE has them working, the opposite would be more likely to be true, that they’d be trying to reduce how much they have to write.

You are right about one solitary thing though – writing will have an objective purpose: to convey information. Whether it does that in an entertaining manner, or in the way intended by the writers, or a way the audience enjoys, etc.,? All that is subjective, and that’s what you’re currently trying to pass as an objective matter.


If it is subjectively appealing to only a small % of the target audience, I would call that objectively bad. And yes, you would be right to say I’m not familiar with the target audience of SE. However, that audience would have to both be poorly read, AND tolerant of overly wordy prose in order to have the tripe passed off as writing appeal to them, so I’m making an educated guess.


And…you’re still doing it.

You’re still trying to pass a subjective judgment as an objective one, based seemingly solely on how you don’t like it (alongside a spurious claim with no evidence to speak of).

Until you get past that particular hurdle, I think this discussion is done.


Yeah, beating a dead horse here. If you think all writing is equal, with the only objective criticism possible being that of spelling or grammar, well, I think we have gone back and forth enough.