I Spent 3 Months Learning English Vocabulary Without Opening a Single App
This isn't a story about discipline. It's a story about removing the need for it.
Three months ago, I set a rule: I would not open a language learning app intentionally. No scheduled study time. No evening sit-down sessions. No Duolingo streaks.
Instead, I let vocabulary come to me — in the margins of my actual workday.
The result: 547 technical English words learned and retained, confirmed by spaced repetition recall data. My reading speed on API documentation increased measurably. I stopped tabbing out to Google Translate while reading Stack Overflow.
Here's exactly how it worked.
The Problem With "I'll Study After Work"
I'd made that promise to myself before. Three previous attempts at consistent language learning, each abandoned within four weeks.
The failure wasn't motivation. In those four weeks, I was motivated. The failure was architecture — building a habit that depended on conditions that never aligned: enough energy after work, no urgent messages to reply to, no merge conflict to resolve before the day ended.
What I needed wasn't more discipline. I needed a system that didn't require any.
The Setup: 0 Intentional Study Time
The only thing I changed was installing Wordrop on my Mac and creating a word list of technical vocabulary I actually encountered at work.
The configuration I used:
That's the entire setup. No daily alarm. No study calendar entry. No commitment beyond "install the app and make a word list."
Month 1: The Awkward Phase
The first two weeks felt strange. Quiz pop-ups appeared at moments I hadn't chosen — after I sent a Slack message and was waiting for a reply, during the 4 minutes the CI pipeline took to complete, right after I closed my IDE to go get lunch.
I answered them. Sometimes wrong. Sometimes very wrong. "Idempotent" — I'd seen it hundreds of times in API docs and still couldn't define it precisely.
End of Month 1 stats:
Month 2: The Recognition Shift
Somewhere around week five, something changed. I started noticing words I'd been quizzed on appearing in documentation — and understanding them without stopping.
"Idempotent." I read it in a Stripe API doc and just… knew what it meant. Not because I'd studied it. Because I'd been quizzed on it seven times across three weeks, each time at the moment my brain was available rather than occupied.
The spaced repetition algorithm had done something I couldn't have done manually: found the exact moments to reinforce each word before I forgot it, and kept doing so until the memory was stable.
End of Month 2 stats:
Month 3: Actual Change in Reading Behavior
By month three, I noticed a behavioral shift I hadn't expected: I'd stopped opening new tabs to Google Translate while reading documentation.
Not because I knew every word. But because I knew enough words that the ones I didn't know were inferrable from context — the way native readers handle new vocabulary.
A colleague noticed I was moving through architecture review comments faster. I mentioned I'd been working on vocabulary. "How?" he asked. "When?"
"During CI builds, mostly."
End of Month 3 stats:
What Made It Work
Looking back, three things made this approach successful where previous attempts had failed.
1. No decision required at the point of practice. Previous methods required me to decide to study. This one delivered practice automatically. The decision was made once, at setup, and never again.
2. Timing was calibrated to my brain's availability. Every quiz appeared during a natural micro-gap in work — waiting for a build, between a sent message and a reply, during lunch. Never during deep work. Never competing with my actual job.
3. The vocabulary was directly relevant. Within the first week, I was quizzed on words I had already encountered in documentation but didn't know. The feedback loop was immediate: failure to recall a word on Monday, see it in prod docs on Tuesday, quiz about it again on Thursday. Three exposures with real context each time.
The Numbers, Honestly
This isn't a "10x your English in 30 days" post. Here's what 3 months of passive vocabulary learning actually produces:
| Month | New Words | Retention Rate | Daily Time |
|---|---|---|---|
| Month 1 | 142 | 61% | ~18 min |
| Month 2 | 170 | 74% | ~14 min |
| Month 3 | 235 | 81% | ~8 min |
547 total words over 3 months. Not fluency. But 547 words that appear constantly in developer documentation, code reviews, and engineering discussions — and that now require zero cognitive overhead to process.
Frequently Asked Questions
What word list did you use?
I pulled directly from documentation and GitHub issues I encountered at work over one week — I kept a running list of words I had to look up. Then I imported that list into Wordrop. High relevance, immediate motivation to learn them.
Did the pop-ups interrupt your flow?
The timing settings matter here. With IDE-detection enabled, quizzes never appeared while I was actively writing code. The 5-minute idle threshold meant they only appeared when I'd already stopped working.
How much English improvement is 547 words, realistically?
Linguist Paul Nation's research puts the top 1,000 technical words at ~85% coverage of developer documentation. 547 words, if targeted at the right vocabulary, is roughly half of transformative reading fluency — enough to notice a real difference, not enough to read everything without effort.
If You Try This
Keep the bar low. The only thing the system requires is a word list that's relevant to your work. Start with 50 words — ones you've had to look up in the last month. Add more as you go.
You don't need to study. You just need to be willing to answer a 30-second quiz when one appears.