Musings on Semantic Primes

Skip to first unread message

Joel Thomas

Apr 24, 2015, 6:37:24 AM4/24/15

Hi everyone,

While I am not actively involved with WAYK at the moment, it is never far from my mind. After a few months of mulling things over, I think I have a thought or two worth sharing.

One of the problems I experienced when using the method was the "obviousness" of language hunting (rather, the lack of it for some people). In my experience, young people in their late teens and twenties seem to cotton on quite quickly while monoglots above the age of 50 somehow seem to have a mental block. It seems to be a kind of mental agility that is lost with time.

Can it be regained?

Intuitively I believe it can. But how?

A couple of months ago I came across a concept that switched the light on for me: Semantic primes.

The road leading to this concept started with attempts to construct what we might call semantic languages, to break down the barrier between our natural vocabularies of disorganised and arbitrary words and the real world of meaning.

Often these tried to organize all known concepts into a hierarchy of categories and sub-categories. This seemed logical and straightforward enough, nevertheless all such attempts were doomed to failure. It turns out our universe of concepts is much more like a vast Venn diagram, and as such defies rigid, exclusivist classification in a top-down hierarchy.

The semantic primes idea is an answer to this conundrum and follows almost the opposite approach: Instead of slotting words into a larger hierarchy, what if we take the words themselves and "unpack" the concepts they contain until we find the basic building blocks, the atoms of language?

Regardless of the scientific debates surrounding this approach, it is a very useful mental exercise. In a way, you learn to approach language like a child again.

While growing up, our vocabularies are expanding at a fast rate. When we first grasp a concept we see it as a composite, a combination of simpler concepts. But as time passes we come to view the new concept as an simple concept in its own right. As we settle into adult life and reach the limits of the day-to-day universe of concepts, we come to see almost all concepts as simple and forget the web of semantic connections between them.

My theory is that this is why younger people find it much easier to "unpack" concepts than older people, because it was not so long ago that these concepts were new, composite concepts to them. This, in turn, makes language hunting much easier.

For example, I would expect a young person who does not know the word for "tea", say, to find it easier than an older person to come up with, "You know, you put hot water in a mug and then you put a little bag in it with a string and the water turns black." By extension I also think now that mastery of this skill is one of the biggest differences between a successful independent language learner and an unsuccessful one.

I think an effective way of getting adults back into the swing of language hunting could be to re-teach them the skill of "concept unpacking". Perhaps a useful exercise could be to get them to re-express concepts using simple words in their own language. For example, we could use the list of semantic primes in a game where the winner is the person who explains the concept of, say, "sandwich", with as few extra words as possible.

Semantic primes could also be a logical starting vocabulary, even more so than "the most common 100 words", because they go beyond simply giving learners a base to start communicating with, by providing the tools with which to expand that base. 

Language is a dynamic, day-to-day skill that was never meant to be taught in a classroom, with its intrinsic limitations of time, stimuli and real-world situations. Nobody is ever going finish a language course as a language speaker, but they just might finish as an effective language learner. Language hunting is a great project to that end, and I hope that my little brainstorm may contribute in some way.

Mishkin Berteig

May 5, 2015, 12:53:32 AM5/5/15
Hi Joel,

Great ideas here.  Reminds me a great deal of Douglas Hofstadter’s new book “Surfaces and Essences” where he uses language and analogy-making to explore the nature of intelligence.  For those interested in language learning, I strongly recommend reading it.


:: You are subscribed to the "Where are Your Keys?" Google Group.
:: To unsubscribe from this group, send email to
:: Also, see
You received this message because you are subscribed to the Google Groups "Where are Your Keys?" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
For more options, visit

Reply all
Reply to author
0 new messages