Natural Semantic Metalanguage

Natural Semantic Metalanguage(NMS) is a framework theorizing that every language has a set of base words, called semantic primes, which can be used to describe all other words in that language. It was first formalized by Anna Wierzbicka in 1972¹. It is theorized that all languages have semantic primes. I first learned of NMS through Aragües y Oroz’s paper Simplex sigillum veri: a minimal language approach to Latin Teaching²

Semantic Primes

The core of the metalanguage are indefinable words called semantic primes. A prime word is

• Universal. All languages have a word for it.
• Undefinable. Can not be defined using simpler words.

Constructing a sentence using semantic primes means that one could translate to and from another language without adding or taking away meaning of the sentence. Doing so often means turning one sentence in to multiple, where each bit of meaning is captured in its own sentence (also called a frame). Goddard³ gives an example of breaking down the phrase ‘X felt sad’ in to semantic primes.

x felt sad

Someone X felt something bad
Someone can feel something like this when this someone thinks like this:

• “I know that something bad happened”
• “I don’t want things like this to happen”
• “I can’t think like this: I will do something because of it now”
• “I know that I can’t do anything”

These frames convey just a single piece of information that can’t be broken down further. NSM explications often boil down to “Something happened” or “someone did/said something”.

There are believed to be 65 semantic primes. I provide them below at the end of this post.

Semantic Molecules

Molecules are words That could be broken down further in to primes, but may be universal enough to be understood without doing so. As an example: explicating on the words ‘cat’ and ‘dog’ could use the molecule ‘animal’, without needing to further break down the concept of an animal. Goddard says that Molecules allows for “conceptual chunking” to manage semantic complexity. He estimates the English language uses around 150-250 molecules.

Practical Applications

Using a minimal vocabularly such as the one provided by NSM could prove effective in language learning. David Bullock has put together a brilliant website for English Learners based on NSM: Learn These Words first. It uses Around 300 words,the primes and the 263 molecules, to describe 2000 words. From there, 70000 words in the Longman English Dictionary can be described from the core 2000.

Appendix

The 65 semantic primes In English