17 Principles Behind Phrasing

In this detailed episode of Phrasing FM, the host breaks down the scientific foundations of the Phrasing app. Drawing from linguistics, memory science, and second language acquisition research, he explains how Phrasing combines proven principles into one cohesive, highly effective system for learning languages faster and more enjoyably than traditional methods.

00:09 – Introduction The host shares how he spent over three years building Phrasing by integrating research-backed techniques to maximize language acquisition.

01:31 – Core Principles

Spaced Repetition (01:31)

• Reviews material at increasing intervals for long-term retention.

• Uses a battle-tested algorithm similar to Anki but optimized for enjoyment and habit formation.

• Can dramatically boost recall; Phrasing makes it addictive rather than a “daily grind.”

Effortful Retrieval (05:31)

• Actively recalling information (vs. passive review) strengthens memory.

• Implemented via typed answers, cloze deletions that expand from single words to full phrases, forcing real cognitive effort.

Contextualized Learning (09:40)

• Learning words/phrases in meaningful context instead of isolated translations.

• Highlights phrases in the target and translation; users must piece together meaning from surrounding context.

Affective Filter Hypothesis + Zone of Proximal Development (12:08)

• Reduces stress/anxiety while keeping challenge at the optimal level.

• Configurable difficulty target dynamically adjusts card selection to keep accuracy in the “just right” zone (~70% for many users).

Tolerance of Ambiguity (18:23)

• Training the brain to handle uncertainty, essential for real-world use.

• Long native-level expressions (15–20 words) build comfort with ambiguity from day one.

Category Induction (22:26)

• The brain naturally extracts grammar patterns from repeated examples.

• Grammar is presented retrospectively through explanations rather than upfront rules.

Contextual Variability & Language Graph (26:20)

• Seeing the same word in multiple expressions and connections (roots, compounds, characters).

• Creates a rich web of understanding; adding 2–3 examples often produces a “magical” unlock.

Encoding Variability (28:54)

• Multiple ways of engaging with material (reading, listening, typing, explanations, shapes) create stronger, more flexible memories.

Interleaving (39:30)

• Mixing multiple languages in reviews improves discrimination and long-term retention.

• Users report better separation between similar languages (e.g., Croatian/Macedonian).

Batching & Language Momentum (42:00+)

• Balanced approach to grouping same-language reviews for flow while maintaining engagement.

Spacing (43:53)

• Multiple short sessions throughout the day outperform one long session.

Input Hypothesis (47:45)

• Benefits of comprehensible input via long expressions that create asynchronous processing (you can’t hold the whole sentence in memory).

Extensive Reading & Phrases / Multi-word Units (53:25+)

• Phrases and linguistic chunks make learning more natural and effective than single words or full sentences.

Linguistic Chunks & Shapes Feature (55:00+)

• Breaks expressions into meaningful processing units (slightly different from phrases).

• Shapes view helps users understand complex sentences by exploring how chunks combine.

26:04 – Conclusion These principles work synergistically in Phrasing to create engaging, effective, lifelong language learning. Try it at phrasing.app.

SEO keywords: science of language learning, spaced repetition principles, effortful retrieval, zone of proximal development, tolerance of ambiguity, category induction, linguistic chunks, Phrasing app, input hypothesis, interleaving languages.

AI Transcript

00:00 All right. Welcome back to another episode of Phrasing FM, a podcast where I talk about languages and my experience building an app to learn them.

00:09 So today I want to talk about the science behind Phrasing. I've been building Phrasing now for over three years, and one of the things I wanted to do with Phrasing was draw from linguistics and memory science and second language acquisition, and take all the research I could find about how can we learn more language and put them all together in one cohesive experience.

00:42 And so today, I want to walk through some of the different principles that I've applied to Phrasing that have had significant effects in my experience using the application, and that backed up by the literature, can have significant impacts on the amount of language that you can learn.

01:31 Top of the list, of course, is spaced repetition. If you haven't heard of spaced repetition before, it's just the idea of showing you words at increasingly long intervals to make sure you never forget them... This is something that is pretty core to the experience of Phrasing. And we lean on it quite heavily.

03:00 The whole experience is built around this because again, this can tenex your recall in any field or in any language... I wanted to build spaced repetition that you'd actually enjoy — flashcards that you can't put down.

05:31 The next scientific principle that I want to talk about is the idea of effortful retrieval... There's this idea of effortful retrieval, where there is a desirable amount of difficulty that you should have when remembering something to increase your memories of different things.

06:36 We use close completions... We start by hiding one word in a sentence, and then we will hide two words, and then three words, and then four words and up to entire phrases.

09:40 The third principle I want to talk about is contextualized learning... Before we used to hide a single word, and then I would give you the translation of the word that you need to fill in. And so this is non-contextualized learning... Now I don't show you the missing word. And instead what I do is I highlight the phrase that the word belongs to and the corresponding phrase in the translation.

12:08 The fifth and the sixth items... The affective filter hypothesis... And the zone of proximal development... We don't use due dates in the application. We try to keep you at the zone of proximal development. So in Phrasing, you can configure the difficulty target that you want... And we will choose words that will put you at that difficulty.

18:23 The next principle I want to talk about is number seven, the tolerance of ambiguity... How well can you tolerate ambiguity when you're learning a foreign language?... With longer expressions... you're starting from this ambiguous state and you're working your way through it.

22:26 Number eight, which is category induction... The brain's innate ability to unconsciously extract grammatical patterns from repeated examples... We don't start with the grammar... We just show you patterns... After seeing these patterns for a couple different times, you'll just pick it up.

26:20 Number nine is contextual variability... With the idea of the language graph... We create this very rich web, this graph of language... Adding just a second expression with that word, the clarity that you get... is remarkable.

28:54 And then there is number ten the encoding variability hypothesis... Having seen the words on screen, hearing the words playback... recalling the word, typing in the word and reading the explanations, then also opening up the phrases and seeing these expressions as sets of phrases...

39:30 Number eleven is interleaving... Mixing multiple languages in reviews... Users report better separation between similar languages.

42:00 Number twelve is batching... I implemented language momentum... We will show you words in a very secretly batched manner... balancing that dopamine principle, but also giving you this batching principle.

43:53 Number thirteen is spacing... Doing several smaller review sessions is better than one big review session... The application is designed and optimized to be used as many times as possible throughout the day.

47:45 Number fourteen is the input hypothesis... My hypothesis for what makes input is if you cannot remember the beginning, by the time you get to the end, then it is input... With expressions around fifteen to twenty words long... it puts you into kind of this asynchronous flow.

53:25 Number fifteen is extensive reading... Number sixteen is just phrases in general and multi-word units... This is where Phrasing gets its name.

55:00 And then number seventeen... linguistic chunks... When you open up the shapes feature... it opens up the entire expression in a color coded view... split again into phrases... This is the idea of linguistic chunks... It helps you understand sentence structure, word order, and just word grouping.

01:06:16 And with that, that is seventeen scientific principles that I have been leaning on when building Phrasing... Hopefully these are interesting to you... If you want to try any of these out, you can try them out on phrasing.app... Thank you very much for listening this long and have a great day. Ciao!

Listen to the episode on YouTube: