The method begins with language that already means something
A learner rarely needs another abstract rule floating by itself. They need to recognise structure while meaning is still present. Metkagram therefore begins with a complete sentence. A Mark points to a useful feature inside it; the learner does not have to jump between a diagram and the sentence to discover what the cue refers to.
The next step is abstraction, but only as far as it helps reuse. A sentence such as “It’s not that X; it’s that Y” can become a Frame with replaceable slots. The original example remains available, so the learner can move back and forth between a concrete utterance and the more general structure.
Frame describes the form; Move describes the job
A Frame is language-specific. An English Frame and a German Frame can look quite different even when they solve the same conversational problem. A Move is the more general job: correcting a framing, limiting a claim, setting a condition, drawing a conclusion, conceding a point, or asking for clarification.
This separation matters because useful language is not only a collection of shapes. Speakers choose a shape because they are trying to do something. Metkagram keeps both layers visible: how the expression is built and why a speaker might choose it.
Learning continues after recognition
Recognition is comfortable, which is exactly why it can be misleading. The method adds Contrasts and Choices so the learner has to discriminate between nearby Frames, then retrieve one before seeing the answer. Routes group several objects around a communication job, and Bridges connect reviewed Frames across languages without pretending that translation is always one-to-one.
The end of the loop is production. Keep the relationship or reasoning move, change the people, topic, tense or context, and build a sentence that you could actually use. The aim is not to admire the pattern library. It is to make the pattern available when the situation changes.
A design informed by research is not the same as a proven product
Several components of this workflow have independent research traditions: formulaic language, attention to form, retrieval practice and spaced practice. Metkagram combines them in one concrete system, but evidence for those mechanisms does not automatically establish the learning effect of the complete Metkagram method. That larger claim needs direct testing.
This boundary is deliberate. The useful question is not “can we attach the word science to the homepage?” It is “which part of the learning loop is supported by existing evidence, which part is our design interpretation, and what should we measure next?”