This essay on toki pona's semantics is the result of a conversation at #sona-musi in ma pona pi toki pona. If you'd like to see that conversation and the progressive realization of the sheer absurdity of it, you may find that here. Otherwise, enjoy the show! :)

---

(click the image to use our (very real, mani-back guaranteed) highly-advanced technology to teleport to the post!)

---

It is a pretty decently common suggestion that in toki pona, words' semantic spaces be expanded to allow for not just the concepts presented in the dictionary, but of all ijo that using the word as a modifier represents. The argument is simple:

A pilin is an ijo. A tawa is an ijo.
Why can't tawa on its own refer to an ijo tawa?

Now, the reasoning is very understandable. ijo is often seen as a sort of umbrella term to refer to anything and everything in the least descript way possible. Furthermore, it is true that everything "boils down" to an ijo (or at least can be described as an ijo, even if its not the clearest thing to do). Seemingly even more inticing to this conclusion, we can (and do) use words like pona to refer to ijo pona occasionally! So, what's the big deal? tawa by itself contains all ijo tawa, simple as that.

But I suspect that those that say this aren't fully aware of the implication of their conclusion on the language as a whole. It is my belief that if X = ijo X is in fact true, we have just opened up a Pandora's box regarding the fundamentals that we understand toki pona to have.

With that in mind, let's have a closer look at what accepting X = ijo X actually means for the language. Who knows, maybe we could learn a thing or two about the role of ijo in the process!

The false sense of security: Head-only ijo X

Let's start with a pretty simple interpretation of this conclusion:

As the head of a phrase, X = ijo X.

tawa → (ijo) tawa
pilin pona → (ijo) pilin pona

This interpretation is generally not too bad to accept, actually! Additionally, it does allow for some unique usage that is actually pretty cool, such as referring to organisms with the word pilin or referring to a substitute for something as an esun.

However, it's also the simplest nasin to disregard. Why not just say ijo pilin or ijo esun? For situations in which constraints on available phrase space apply (such as songs or poems), this nasin might actually be worth considering! But for everything else, there doesn't seem to be a real incentive to being less clear like this.

Delving deeper into the chaos: Modifiers as ijo X

"But what about modifiers?", I hear you thinking to yourself. "If we are including ijo X into the semantic space of X, shouldn't that apply to all instances of X"? This is exactly the point at which this premise begins to take a decidedly... absurd turn. Consider the following interpretation of the conclusion that X = ijo X in all cases:

tawa → (ijo) tawa
pilin pona → (ijo) pilin (pi ijo) pona
pilin utala pona → (ijo) pilin (pi ijo) utala (pi ijo) pona
pilin pi utala pona → (ijo) pilin pi (ijo) utala (pi ijo) pona

Oh dear. This got quite complex to understand quite quickly, and we haven't even created a very long phrase! The amount of expressivity within a few words compared to a traditional nasin is astonishing, but it comes at a cost. Because all of these meanings are simultaneous, it requires a lot more effort on the part of the listener (or more contextual information than possible on digital platforms?) to decipher what meaning was intended by the speaker.

A curious note on transitivity

There is also an interesting side effect to how transitivity works if we accept that X = ijo X. Put simply, X = ijo X implements the monsutatesu¹ into every single word (as far as I can tell, at least) in toki pona.

mi moku e sina. → (mi (ijo) moku e sina)
➥ I eat you. / I make you (ijo) moku.
➥ I do something food-related to you. (⚠️)

This is hardly a problem given the contextual and relative nature of toki pona, but it is an interesting phenomenon nonetheless. There are further fun interpretations of transitives given this premise:

mi utala e sina → (mi (ijo) utala e sina)
➥ I fight you. / I make you battle (me).
➥ I do something conflict/battle-related to you.
➥ I make you an ijo utala. (⚠️ could mean "I make you a [combatant, weapon, person fighting with/against me, etc.]")

In this second example, the last interpretation contains a similar aspect to the previous example in that it technically ends up fitting the monsutatesu but also has elements that exist wildly outside typical range of meanings of the word on its own, resulting in another, unrelated interesting occurence with regards to transitives and X = ijo X.

Conclusion: With great power...

In the end, this all comes down to the function of ijo in toki pona. I think kili put it better than I could:

This idea greatly expands the semantic spaces of both types of words when considering X = ijo X, resulting in an even more ambiguous interpretation for the listener to attempt to make meaning from. However, this analysis also serves to elucidate the true power in (and our unconscious assumptions about) the role of the word ijo, which is an important idea to consciously consider!

X = ijo X is a decently comedic example of the Pareto principle as applied to toki pona. It doesn't add a new word, and the change that it makes is so simple that it can be described in a few words. However, its impact on the fundamentals of word interpretation of toki pona is so extreme that I'd strongly recommend avoiding this nasin (or at least using a ton of caution) if you thought at any point in this article "oh hey, this seems like a pretty fun nasin nasa. I'll try it out!"

o kepeken e wawa nasa sina tawa pona taso. o toki pona!

1. The "monsutatesu" is a hypothesized ambiguity present in some toki pona words where when applied transitively, the "apply X to" and "make into X" definitions collide into two opposing meanings. For monsuta, the ambiguity's namesake, it means being able to say "I make them afraid" and "I fear them" with the same sentence: mi monsuta e ona. One resolution for this ambiguity that doesn't rely purely on contextual information is to say that only sensory words actually utilize the "apply X to" definition (ex. mi luka e ijo for "I touch the thing (with my hand)").↩