Semantic primes are basically the bare-bones building blocks of every language on Earth. You can think of them like the “starter kit” for human meaning, the same handful of core words that pop up whether you’re speaking English, Japanese, or yelling directions in traffic anywhere else. Linguists Anna Wierzbicka and Cliff Goddard have spent decades proving these 65 concepts show up in every tongue, which is pretty wild when you think about it. No matter how different we sound, we’re all working with the same linguistic toolbox, just assembling the pieces in our own way.
Below is the complete, official inventory of 65 semantic primes from Natural Semantic Metalanguage (NSM), as established through decades of cross-linguistic research by Anna Wierzbicka, Cliff Goddard, and the NSM community (current as of 2025).
Semantic primes are the fundamental building blocks of meaning that exist in every human language. These universal concepts form the foundation of Natural Semantic Metalanguage and provide a framework for understanding how meaning works across all languages and cultures.
The semantic primes are grouped into semantic categories for clarity. Each prime is presented in its canonical English exponent, but every semantic prime has exact equivalents in all tested languages.
Complete List of 65 Semantic Primes (NSM Framework)
| Category | Semantic Primes |
|---|---|
| 1. Substantives | I, YOU, SOMEONE, PEOPLE, SOMETHING/THING, BODY |
| 2. Relational substantives | KIND, PART |
| 3. Determiners | THIS, THE SAME, OTHER~ELSE |
| 4. Quantifiers | ONE, TWO, SOME, ALL, MUCH~MANY, LITTLE~FEW |
| 5. Evaluators | GOOD, BAD |
| 6. Descriptors | BIG, SMALL |
| 7. Mental predicates | THINK, KNOW, WANT, DON’T WANT, FEEL, SEE, HEAR |
| 8. Speech | SAY, WORDS, TRUE |
| 9. Actions, events, movement | DO, HAPPEN, MOVE |
| 10. Existence, possession | BE (THERE IS), HAVE |
| 11. Life and death | LIVE, DIE |
| 12. Time | WHEN~TIME, NOW, BEFORE, AFTER, A LONG TIME, A SHORT TIME, FOR SOME TIME, MOMENT |
| 13. Space | WHERE~PLACE, HERE, ABOVE, BELOW~UNDER, FAR, NEAR, SIDE, INSIDE, TOUCH (CONTACT) |
| 14. Logical concepts | NOT, MAYBE, CAN, BECAUSE, IF |
| 15. Intensifier, augmentor | VERY, MORE |
| 16. Similarity | LIKE~AS~WAY |
Notes on Notation
- CAPITALS indicate the prime (not the English word).
- ~ means “alternants” — e.g., MUCH~MANY = “much” (mass) or “many” (count).
- DON’T WANT is a single prime (not negation + WANT).
- BE (THERE IS) is the existential prime (not copula).
- TOUCH refers to physical contact, not emotional “touching.”
What Makes These Semantic Primes Universal?
These 65 semantic primes represent the irreducible core of human language and thought. They are claimed to be:
- Indefinable – Semantic primes cannot be paraphrased using simpler terms; they are conceptually fundamental
- Universal – Every semantic prime is present in every natural language studied
- Lexicalized – Each semantic prime has a word or fixed expression in each language
- Productive – Semantic primes can combine to express any meaning in any language
Official Source
Goddard, C., & Wierzbicka, A. (2018). “Minimal English and how it can add to Global English.” In Minimal English for a Global World, Palgrave.
Also available at: NSM Primes
Cross-Linguistic Research on Semantic Primes
The semantic primes inventory is the result of rigorous testing in over 30 languages, including Polish, Russian, English, Spanish, French, Italian, Swedish, Danish, Malay, Mandarin, Japanese, Korean, Lao, Mbula, Yankunytjatjara, Arrernte, Ewe, and others. This extensive cross-linguistic validation confirms that these semantic primes truly represent universal human concepts.
Applications of Semantic Primes
Understanding semantic primes is essential for:
- Semantic SEO – Understanding the fundamental meaning units that search engines use to interpret content
- Entity SEO – Building topical authority by connecting concepts through universal semantic relationships
- Cross-cultural communication – Creating messages that work across language barriers
- Natural Language Processing – Developing AI systems that understand meaning at a fundamental level
- Language teaching – Teaching vocabulary through universal concepts rather than translation