Particles, postpositions and suffixes

[Tengado]

Recently I have got myself a little confused.

What is th differecnes between adpositions (prepositions and postpositions) and affixes (prefixes and suffixes)?

Obviously I know the obvious answer (that's the key thing about it being obvious) "adpositions are words and affixes are parts of words," but how do you define that more rigorously and precisely? For example, English prepositions have no stress and behave pretty much like prefixes, except that they can be separated from their object sometimes (where'd he go to?).

And then what about particles? Is there really even a contrast here between adpositions and particles? Is it that adpositions must follow a certain word, whereas particles don't?

For example, in a conlang I have little unstressed words that follow nouns. Apart from the "can they ever be separated from their nouns" test as in English, how else can I tell if they are postpositions or suffixes?

I guess stress is one - if they affect the stress pattern of the word they must be part of it. But if they don't affect the stress pattern they are not necessarily not part of it. How else can I test to see if they are postpositions or suffixes?

[Salmoneus]

There is no 'rigorous' way, cross-linguistically, although there may be a clear distinction in your language.

I think questions you might ask would be:
- do they effect the realisation of the word?

- can they be divided from the word? If so, only by other debateable 'words', or by things that definitely are words?

- are they identical in form to words that appear in other circumstances?

- how likely is a speaker to speak that morpheme individually? That is, asked the equivalent of "did you say 'went to' or 'went from'?", how likely are they to say "from" or "to", and how likely to say "went from" or "went to"? If they feel they must always include the verb, that's a fairly good sign the morpheme is not independent. Contrariwise, if you ask "next to the dog or next to the cat?" and they say "the dog", that suggests that 'next to' is free, whereas if they always say 'next to the dog', that might indicate that 'next to' is an inflection - although I'm less sure about this one than about the questions comparing the affixes/adpositions themselves

- do there exist unambiguous affixes and/or adpositions in the language? If so, do the debateable morphemes pattern like either or neither of these?

[eodrakken]

The only thing I can think of to add to Sal's answer is that you can look at examples of how far speakers back up when they make a mistake. "It was next to the dog-- I mean, the cat" suggests again that "next to" is not bound (and that "the" might be). Speakers don't usually start a repair from the middle of a word.

[Jetboy]

I'm not sure about this, but I think one big difference is that adpositions can be applied to more than one word at once, i.e., he threw it towards the house and farm, as opposed to, he threw it homeward and farm-ward.

[Miekko]

I'm not sure about this, but I think one big difference is that adpositions can be applied to more than one word at once, i.e., he threw it towards the house and farm, as opposed to, he threw it homeward and farm-ward.

and oppositely, in some langs, several adpositions can apply to one noun.

(however, in some langs, a noun can be in a position where it fills syntactical positions requiring two different cases, but only marks one of them; tricky stuff)

[Radius Solis]

Recently I have got myself a little confused.

What is th differecnes between adpositions (prepositions and postpositions) and affixes (prefixes and suffixes)?

Obviously I know the obvious answer (that's the key thing about it being obvious) "adpositions are words and affixes are parts of words," but how do you define that more rigorously and precisely? For example, English prepositions have no stress and behave pretty much like prefixes, except that they can be separated from their object sometimes (where'd he go to?).

And then what about particles? Is there really even a contrast here between adpositions and particles? Is it that adpositions must follow a certain word, whereas particles don't?

Yes, I think you've gotten a bit confused. First let's separate affixes from particles: affixes are bound grammatical morphemes and particles are free grammatical morphemes. Decent tests for "bound" vs. "free" are listed by others above.

Next, clitics: their most usual definition is "phonologically bound but syntactically free". The specifics of this are messy and language-dependent, and some will argue that no "particles" are truly phonologically free so it's a moot category, but the three-way distinction is still the canonical approach.

Now, adpositions: they are not properly defined by their free/bound status (in linguistics, as opposed to popular definition). They are defined by what they accomplish in the morphosyntax, and are sometimes hard to tell apart from cases. Frequently adpositions are free and cases are bound, as in most European languages, but this is not always true. Both cases and adpositions can be any of free, bound, or cliticized. For free case markers, look no further than English to (free) and 's (cliticized), and some of the Japanese postnominal particles like ga and o. For bound adpositions, look at the non-core "cases" of Finnish (everything but the first four on this list).

There are also languages where the case/adposition distinction is not necessarily very meaningful.