How to define colors in the Natural Semantic Metalanguage?

Question

The Natural Semantic Metalanguage (NSM) attempts to reduce the semantics of all lexicons down to a restricted set of semantic primitives. But in NSM, colors are not semantic primitives. How then to define words like "blue"? More generally, how to define word representing qualia (i.e. individual instances of subjective, conscious experience such as the redness of an evening sky, pain of a headache, the taste of wine, etc.)?

Answer 1

Anna Wierzbicka wrote a chapter in her 1996 text Semantics: Primes and Universals on the semantics of colour terms. In this chapter she presents a theory where colours are understood according to their similarity to exemplars (a type of prototype theory). For example, here are two explications for the English colour terms red and yellow:

X is red. =
when one sees things like X one can think of fire
when one sees things like X one can think of blood

X is yellow. =
when one sees things like X one can think of the sun

These explications are not fully reduced to the NSM primes, but use semantic molecules (fire, blood, sun, etc).

She first considers individual languages' colour terms, showing that the exemplars of each colour are language specific. For example, although all of these colour terms overlap, they are not equivalent, and have different exemplars:

X is niebiski (Polish) =
at some times people can see the sun above them in the sky
when one sees things like X one can think of the sky at these times

X is blue (English) =
(a) at some times people can see the sun above them in the sky
  when one sees things like X one can think of the sky at these times
(b) in some places there is a lot of (very much) water
  when people are far from these places
  they can see this water
  when one sees things like X one can think of this

X is aoi (Japanese) =
(a) at some times people can see the sun above them in the sky
  when one sees things like X one can think of the sky at these times
(b) in some places there is a lot of water
  when people are far from these places
  they can see this water
  when one sees things like X one can think of this
(c) in some places things grow out of the ground
  at some times there is water in those places
  when one sees things like X one can think of this

Later in the chapter she considers Berlin and Kay's research on the proposed hierarchy of basic colour terms, and whether there could be a universal basis to colour terms. For the first level binary division she hypothesises that light, the sun, and fire could be the basis of the "macro-white" and "macro-black" concepts. For the secondary level base colours of "macro-red" and "grue" (cool/blue/green) she suggests that reddish colours are focused on fire and to a lesser extent the sun (but not blood, even though it is an important part of the definition of the English colour red), and, contrary to expectation, "grue" is not actually focused on either the sky or vegetation, but is instead negatively defined as "not-red". This is because at that time "grue" had never been identified as being focused in the intermediate blue-green region, but was only found to have blue and green focal regions. This cross-linguistic exploration of possible colour universals is quite tentative though, and Wierzbicka and other NSM proponents may have changed their models substantially since the chapter was written.

Answer 2

You simply have to build up phrases from prior phrases, beginning with semantic primitives, to define new terms.

Colors are in principle defined through the objects that bear the color, sky blue, blood red, leave green etc. So first you need to define those, which is out of scope for the question. There are the primitives see and like, so it may very well be possible.

@amegnunsen's answer implies a semantic molecule for colors was proposed, however I have no access to the books to see and confirm an example sentence defining colors.

Answer 3

This theory considers that colours are not an universal concept, so they are not included as semantic prime. But, another conceptual tool, which takes into account cultural categories as the colour, is used, it is called semantic molecule. Yet, they don't go further in the description.

In an article "Why there are no ‘colour universals’ in language and thought (Wierzbicka, 2008)", this author proposes to sweep the colour concept aside and instead to retain as concept the visual semantic. By doing this, the semantic prime SEE can be used to describe these meanings.

But if you are not satisfied by their proposal, you can offer your own description. Thus, you can choose a language or languages which have a wealthy vocabulary about colours and consider them as your enters for other languages.

Riffian: azeggwagh azeggwagh azegza azegza

English: red red green blue

French: rouge roux vert blue

From this example, you can see that French is much more accurate than English and Riffian. So French should serve to elaborate your primitives.

You can also choose to use codes which describe the colour spectral such as RGB, HEX, ...

Yet, If you mean by "primitive" the colour concepts shared by all the languages of the world, so there are only two which are found everywhere and which are: light/white & dark/black. Others argue that there are six universal colours: https://wals.info/chapter/132

The comparative method is the only one which can help you to determine your primitives.