A while back, an engineer who'd recently joined a team here said, in Japanese, that he had "pushed" his code. Except he didn't use the word "push." He'd looked it up, found the Japanese verb for physically pushing something, and used that instead. 押しました. Literally, "I pushed it," the way you'd push a door.
The room went quiet for a second. Not because it was wrong exactly, but because nobody says that. Everyone understood him eventually, but there was a visible half-second of "...pushed what?"
His Japanese was genuinely good. That was the strange part. He'd studied hard, he knew the grammar, and he'd done exactly what a textbook trains you to do: take the English word, find the Japanese equivalent, use the Japanese equivalent. For almost every situation, that's the right instinct. For technical vocabulary on a dev team, it's the one place it backfires.
Japanese dev teams speak in katakana English
In a Japanese engineering team, most of the technical verbs are just English words, pronounced in katakana, dropped into Japanese sentences.
You don't translate "push." You say プッシュ (pusshu). You conjugate it with a Japanese helper verb and move on:
プッシュしました (pusshu shimashita): "I pushed."
Same for almost the entire daily vocabulary of the job:
- マージする (māji suru): to merge
- デプロイする (depuroi suru): to deploy
- コミットする (komitto suru): to commit
- リリースする (rirīsu suru): to release
- レビューする (rebyū suru): to review
- リファクタリングする (refakutaringu suru): to refactor
The pattern is almost mechanical: take the English technical verb, say it in katakana, add する (suru, "to do") to make it a verb, then conjugate する normally. プッシュする becomes プッシュします becomes プッシュしました. You already know these words. You just have to trust that the katakana version is the correct one, not the "real" Japanese translation you'd find in a dictionary.
That trust is the hard part, because it runs against everything language study teaches you. The dictionary translation of "merge" exists. It's 統合する (tōgō suru). It's a perfectly real word. But if you say "この機能を統合しました" in a stand-up, you'll sound like you're presenting a corporate merger, not describing a git operation. The katakana word is the one that belongs in the room.
Why it works this way
Software as a field grew up in English. The tools, the docs, the error messages, Git itself, all in English. When Japanese teams adopted them, they didn't stop to invent native words for every command. They just absorbed the English terms into speech, the way English absorbed "café" or "tsunami." So the katakana isn't some casual shortcut. It's the register the work actually happens in.
The upside is almost unfair: if you know the English word, you already own the Japanese one. The gap is smaller than it looks. The catch is trusting that, which cuts against every instinct language study builds in you. Your brain keeps offering the dictionary translation, and you have to keep turning it down.
A few that trip people up
Not everything is a clean katakana swap, and a few common cases are worth knowing.
Some words have a native form that's genuinely used alongside the katakana one. "To fix" a bug is usually 修正する (shūsei suru), a real Japanese verb, not フィックスする. "A bug" itself, though, is バグ (bagu), straight from English. So you'll hear バグを修正しました (bagu o shūsei shimashita), meaning "I fixed the bug," mixing a katakana noun with a native verb in the same short sentence. There's no single rule; it's just what stuck.
And "push" specifically has one more wrinkle. In casual speech, people often shorten pull request to プルリク (puru-riku). Written in chat it's PR. Said out loud, it becomes puru-riku. So the full sentence in a stand-up might be PR出しました (PR dashimashita), where even 出す (dasu, "to put out / submit") is the native verb doing the work, and PR is read as puru-riku. Nobody spells any of that out for you. You just have to hear it a few times.
The takeaway
If you're learning Japanese to work on a dev team, your instinct to translate is correct almost everywhere except the technical core. There, do the opposite: keep the English word, say it in katakana, and staple する to the end. The engineer who said 押しました wasn't wrong about Japanese. He'd just missed one habit that nobody writes down anywhere, because to the people who already know it, there's nothing to write.
This gap, between the Japanese you study and the Japanese engineers actually speak at work, is the whole reason I built a course called Japanese for Software Engineers. It covers the stuff that only shows up once you're in the room: stand-ups, code reviews, incident calls, and yes, how to talk about your PR without confusing anyone. If that's useful to you, it's free through August 21, 2026: https://www.udemy.com/course/japanese-for-software-engineers/?couponCode=ABAD7DB87460AA4E2853
TL;DR
- Japanese dev teams say technical verbs in katakana English, not the dictionary translation: プッシュする (pusshu suru), not the native verb for "push."
- The formula is mechanical: English technical word in katakana + する (suru) + normal conjugation. プッシュする → プッシュしました.
- The dictionary translations exist (統合する for "merge") but sound out of place at work. The katakana word is the professional register.
- A few mix: バグ (katakana noun) but 修正する (native verb for "fix"). "PR" is written PR, said puru-riku.
- If you already know the English term, you already know the Japanese one. Trust the katakana.
Top comments (2)
I've always been curious — doesn't staring at all those katakana characters all day get exhausting?
"It's just a matter of getting used to it, so Japanese people don't really think twice about it. But hearing that katakana can be exhausting for others is a fresh perspective!"