Is Toki Pona a Standard Average European language?

First posted on ma pona on 2025-01-25.

Standard Average European is a sprachbund of languages centered around Central and Western Europe. Though most of them are genetically related (having PIE as their common ancestor), they have converged on some common linguistic features that were not present in their past. Some of these features are markedly more present in Europe than elsewhere in the world.

Toki Pona is a constructed language that was designed by jan Sonja with influence from around a dozen languages, spoken both in and outside Europe. The community generally prides itself on rejecting eurocentrism. As such, the natural expectation is for Toki Pona to not be a Standard Average European language. However, Toki Pona is also largely spoken online by a community dominated by American and European users, which may have influenced the language to conform to the expectations of the speakers’ native tongues.

Standard Average European is not a strictly delineated set of languages; instead, it is defined by a set of features, concentrated in the geographical centre of the sprachbund and dropping off towards the periphery. Let us look at the list of features from [Haspelmath 2001] and consider whether Toki Pona has those.

Features

2.1. SAE has definite and (to a lesser extent) indefinite articles.
✘ Toki Pona does not have articles.

2.2. SAE builds relative clauses with a pronoun that inflects for case within the relative clause.
✘? Toki Pona does not have (embedded) relative clauses. Instead it uses a sequence of two syntactically independent clauses. In place of an SAE relative pronoun we find ni in the main clause and usually, though not always, ona in the (implicitly) relative clause. If counted together, the ni-ona structure could be described as a relative pronoun; however, the existence of sentences without ona (e.g. reported speech) and the fact clauses are not embedded is, in my opinion, enough to consider this a strategy sufficiently different from SAE.

2.3. SAE has a perfect construction constructed using an auxilliary verb meaning ‘to have’.
✘ Toki Pona has no such auxilliary verb.

2.4. Verbs related to sensation, emotion, cognition, and pcrception in SAE predominantly mark the experiencer as a nominative (subject), rather than a dative (indirect object).
? Toki Pona exhibits both strategies: dative (… li pona tawa Exp) and nominative (Exp li olin e …, Exp li kute e …). To properly compare it to SAE, we’d need a full list of verbs compared in past studies.

2.5. SAE has a passive constructed using the verb meaning ‘to be’ or ‘to become’ and a passive participle.
✘ Toki Pona has a preverb meaning ‘to become’ which may be used in constructions that resemble a passive: ‘mi jelo e sinpin ~ sinpin li kama jelo’. However, this interpretation only works with predicates that are labile, and does not actually require the preverb to achieve a ‘passive’ meaning: ‘sinpin li jelo’. For a non-labile predicate, e.g. ‘jan li esun e poki. *poki li kama esun’ the meaning does not (or cannot) appear passive.

2.6. In SAE, it is more common to derive intransitive (inchoative) verbs from transitive ones, than to derive transitive (causative) verbs from intrantive ones.
✘ All Toki Pona predicates can be used with or without a direct object, making them all (actively or passively) labile. Some people interpret the accusative marker ‘e’ as deriving a causative verb from a previously intransitive one, which is a framework I personally don’t use, but if accepted it only strengthens the distinction from SAE.

2.7. In SAE, a possessor can be expressed outside the noun phrase using a dative.
✘ This does not occur in Toki Pona, to my knowledge all possession either occurs within one noun phrase or is periphrastic with unrelated strategies.

2.8. In SAE, when a negative indefinite pronoun is the subject, the verb is not negated: ‘nobody comes. niemand kommt.’
✔ Although Toki Pona does not have indefinite pronouns as a specialised word class (the same can be said about most classes of pronouns), an analogous structure with equivalently broad semantics exists, and does not negate the verb: ‘jan ala li kama’.

2.9. SAE uses a specialised particle, typically related to pronouns, to construct a comparative.
✘ There is notoriously an unusual amount of variation in Toki Pona when it comes to comparative constructions, from polar constructions intended by jan Sonja found in early texts, to case-marked constructions (with tawa, tan, or lon), to exceed-type constructions (with ete), and others (with la) - comprising, in total, almost every strategy found crosslinguistically. Amusingly, the SAE particle is one of the only strategies not found in Toki Pona.

2.10. In SAE, equative constructions (of the form ‘as … as …’) are based on adverbial relative clauses.
✘ Equative constructions in Toki Pona are marked with sama.

2.11. In SAE, verbs have bound morphemes that indicate the subject, but the subject still has to be explicitly stated even as a pronoun.
✘ Toki Pona does not have bound morphemes that indicate the subject. The subject is only marked once.

2.12. In SAE, there is a reflexive that is also used as an intensifier (indicating the head as central, not peripheral to the action taken).
✘ For some speakers of Toki Pona ‘sama’ may be considered a reflexive, but most seem to instead repeat the same personal pronoun used earlier. Intensives definitely exist in Toki Pona, but are unrelated to ‘sama’: most commonly, ‘a’, but in certain circumstances perhaps also ‘mute’ or ‘taso’.

3.1. In SAE, verbs are sometimes fronted in questions.
✘ There is no verb fronting in Toki Pona; word order in interrogative sentences is identical to positive sentences.

3.2. In SAE, there is a special adjective form, used in comparative constructions.
✘ Although as previously mentioned comparatives exhibit large variation in Toki Pona, in most constructions the “adjective” (usually a predicate) is unmarked compared to a positive adjective. Some constructions add “mute”, but are relatively less common.

3.3. In SAE, the “and” conjunction is symmetric (as opposed to the asymmetric “with” strategy), and the word order / syntax is of type “A and-B”.
✔? As Toki Pona only has analytic case marking, it seems impossible to determine how symmetric the marking is. Moreover, an explicit marker is only found in the subject position. It does, however, unambiguously belong to the “A and-B” type.

3.4. SAE languages use the same marker for instrument (= instrumental) and accompaniments (= comitative).
✘ In Toki Pona, instrumental and comitative are differentiated: instruments are marked with ‘kepeken’, accompaniments with ‘lon (poka)’, or somewhat more rarely ‘en’.

3.5. In SAE, the cardinal number ‘one’ has a suppletive ordinal (‘first’).
✘ In Toki Pona, ‘wan’ and ‘nanpa wan’ form part of the regular paradigm.

Conclusion

Out of 17 features examined, only one (2.8) is unambiguously shared between SAE and Toki Pona, with three more (2.2, 2.4, 3.3) being difficult to interpret due to great typological difference between Toki Pona and languages in Europe. In (Haspelmath, 2001), which presents a combined map of 9 features (with greatest data availability), this gives it a score equivalent to Welsh, Georgian, and Armenian, i.e. highly peripheral to the Standard European core.

An interesting observation is that even in cases where European speakers could have influenced Toki Pona (2.9), this has not resulted in convergence with SAE.