Skip to content
LangBytes.

LangBytes · Method

Situational fluency: why LangBytes starts from the situation you face, not a word list

· 4 min read

The short version

LangBytes does not start from a word list or a textbook unit. It starts from a situation you will really be in, such as ordering at a ramen counter, asking for directions or meeting a new team, and from what you need to get done there. The lesson is built around that outcome. This post explains the idea, why it is built this way, and what it has not shown yet.

What a situation is

A situation is a recurring real-life occasion with an intended outcome, for example "order a bowl of ramen and pay". LangBytes breaks it into four parts:

  • Your moves: what you say to reach the outcome.
  • Their moves: what the other person is likely to say back, so you can recognize it.
  • Critical details: numbers, names and conditions you must catch or say exactly, such as the price or a time.
  • Repair: what to say when the conversation breaks down, and when to use it: asking someone to repeat, to speak slowly, or to write it down.

Related situations are grouped into a family (dining, transit, work), so what you learn in one carries over to the next.

Why start there

  • The situation decides which words matter. A frequency list says which words are common in general. A situation says which words you need next week. The same person at the same ramen counter needs the same few moves again and again.
  • A word list never teaches repair. Real conversations break down. Knowing what to say when you did not catch the number is part of managing the situation, and it is rarely in a word list.
  • Practice matches use. Recognizing their move, saying yours and varying it when they phrase it differently is closer to the real exchange than matching a word to its translation.

This is design reasoning, not a measured result. See the section on evidence below.

Where the idea comes from

Situational Fluency is a method that LangBytes founder Brad Lishman developed over roughly ten years of teaching himself four languages while living in Spain, Germany and Japan, and through family life connected to Slovakia. He wrote most of the first manuscript in Málaga in 2015. LangBytes, conceived in 2025, is a software version of that practice: it keeps the loop of a real need, preparing the moves, using them, and revising after what happened.

What the research does and does not say

  • Established parts. Two pieces of the loop have strong laboratory support. Spaced practice beats massed practice for long-term recall (Cepeda et al., 2006), and testing yourself beats re-reading (Roediger and Karpicke, 2006). LangBytes uses both for review.
  • A familiar frame. Describing language ability as what a learner can do in tasks, rather than what they know about the language, is the action-oriented approach of the Council of Europe's CEFR.
  • The combination is not yet tested. Those studies mostly measure recall of words and texts. Whether learning from your own situations transfers to real conversations better than other methods is exactly what is still unproven. LangBytes has not published learner-outcome results, and its own usage numbers would not prove it. Treat it as a design with a research program behind it, not as a guarantee.

How to use the idea without LangBytes

  • Write down one situation you will face this week and the outcome you want from it.
  • Prepare five or six moves for you and the likely replies for them.
  • Add one repair line, such as asking someone to say it again more slowly.
  • Use it, then note what failed and fix only that.

Sources