The three lexical categories (or parts-of-speech) particles, markers, and adpositions are extremely vague to me. Some grammars call marker what others call particle or adposition. For example, the English marker "-'s" (as in "Mary's book"), the Maori particle "a" ("Te pukapuka a Mere"), the Japanese particle "no" ("Marī no shoseki") and the Spanish preposition "de" ("El libro de Maria"), all creates a genitive-like phrase, but each belongs to a different category.

The only patterns I can notice in those terminologies are the following (They are probably wrong):

  • Particles and markers are generally synonyms. In the English phrase "Mary 's", the bold word may be either a marker or a particle, depending on the grammar (others may also say it is a possessive/genitive inflection ending).
  • If the word is clitic or phonetically appended to another word, it is more likely it will not be called adposition. For example, in the Spanish phrase "para Maria" ("to Mary"), the bold word is phonetically independent, and it is a preposition; but in the English phrase "Maria 's", the bold word is phonetically dependent on the preceding word, so it is not an adposition.
  • If the word is used to mark predicates, instead of arguments or adjuncts, it is more likely it will not be called adposition either. For example, in the Maori phrase "ka mahi" ("to work"), the bold word, which turns something into a predicate instead of an argument or adjunct, is not a preposition; but in "a mahi" ("work's", genitive), the bold word may be called preposition.
  • If the word is used to mark adjuncts, mainly the locative ones (over, under, etc.), it is more likely it will be called adposition. For example, in the Spanish "sobre la mesa" ("over the table").
  • If the word is used to mark arguments of a predicate, it is more likely it will not be called adposition. For example, in the Maori phrase "i mahi", the bold word, which marks the object (accusative) or source (ablative) argument of a predicate, may not be a preposition.
  • If the word is used to mark grammatical functions in the same way as mainstream grammatical cases do, it is more likely it will not be called adposition. The words which mark a genitive-like or dative-like phrase are exceptions. If the word mark directional sense (as in allative or ablative), the terminology may be ambiguous.

So, what are the definitions of particle, preposition and marker. What are the differences between them (i.e., the words with belong to a category but not to another) and the intersections between them (i.e., the words with belongs to two or all tree categories)?

  • 4
    So you seem to be looking for a cross-linguistically valid definition of these POS terms, is that right? Nov 9, 2017 at 5:50
  • Yes. I want the definition of each one of these lexical categories (or POS, as you said). And also, analysing these three POS as sets of words, I want the intersections and exclusive disjunctions between them (i.e., the words considered to be exclusively in one of these POS, and the words considered to be in more than one of these POS).
    – Seninha
    Nov 9, 2017 at 14:44
  • I'm not certain you're going to find any universal agreement on this. Some Slavic languages, for example, have prepositions (under standard terminology) that are clitic/not phonetically independent (e.g. Polish w, z); Spanish has the "personal a" which marks certain direct objects but is regarded as a preposition; etc. Nov 16, 2017 at 15:57
  • When you are interested in part-of-speech, look at these two questions and their answers: linguistics.stackexchange.com/questions/12777/… and linguistics.stackexchange.com/questions/1138/… Sep 7, 2018 at 18:27

2 Answers 2


The first thing is that part-of-speech is something linguists assign to a certain word form and this assignment depends on the underlying analysis.

Second, the class particle is already a kind of container for "everything else, that was not previously sorted into other categories". Differentiating particle and marker as two separate classes for part-of-speech is rare.

Third, one word form can have different part-of-speech depending on its use in a sentence, e.g., the English word to is a particle in the infinitive construction to work, but a preposition in the phrase I go to work at 8 o'clock.

Fourth, it is very difficult to define even basic parts of speech cross-linguistically. One might come up with a set of tests that work well in one language, but are completely meaningless in another one. So there is always a language dependence in a set of parts of speech. There are lots of tries to unify sets of part-of-speech cross-linguistically, and a currently hot one is Universal Dependencies.

  • Your third point doesn't immediately explain why we should treat to as a particle in the infinitive case or as a preposition (proclitic) in the second case. See Zwicky (1985)'s "Clitics and Particles" in the journal Language, Vol. 61, No. 2 (Jun., 1985), pp. 283-305. His explanation is: "If to were a proclitic rather than an independent word, then we would expect no aspiration of the first /p/ in to perpetuate, since it would not begin a phonological word. The presence of aspiration there supports other evidence that to is not a clitic."
    – whatisit
    Oct 22, 2018 at 18:27

Some markers can have as origin a preposition, but it is a misuse of language to use interchangeably these concepts as in Spanish or French (preposition: I/Je come/viens from/de Paris ## genetive: The/le wine/vin of/de Paris). Particule is a subclass of the morpheme, it contrasts with morphemes that have a long connected speech, so you can use it to speak of prepositions or markers if they are morphemes that is composed of one or two sounds.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.