For sure these things have been collected many times somewhere. It looks like a job that a professor could give to a seminar student. But how to put it in a simple way for a learner and  -  more than that  -  as a learner is another question. There are so many factors to take into account.

Is ´he may be able to run´ koşabilebilir?

Grammatically may not be incorrect but this form is not used at all.

Instead:

Muhtemelen koşabilir = Probably he is able to run

Morphologically, -(e)bil- can come after -meli- but not vice versa, is it?

No.

Let´s analyse:

-ebil- is actually v-a/-e bil- so -a/-e (hence -ebil-) needs a verb stem.

-meli also needs a verb stem

that means -ebil-meli is OK

Morphologically, -(e)bil- can come after -meli- but not vice versa, is it?

 

 

You know, I must have thought of the opposite but mixed it while writing...sorry.

______________________

Forms like et|me|yebil|ir ´may not do´ or koş|ama|yabil|ir ´may not be able to run´ really demonstrate the simple logics of Turkish morphology. Look at the place of the negation: it is attached to the part which is in fact denied, not to a grammatical hanger like in auxiliary languages. This is a very pedantic way to express the meaning. Actually you should translate them ´it may be that he doesn´t do´, ´it may be that he is not able to run´.

These epistemic examples also show that -(e)bil- still very much represents a full lexical verb even though due to agglutination it is usually seen as a grammatical morpheme.

Just a couple of modest notes, nothing important.

gele de bilirim, gelmeye de bilirim = I may as well come, I may not either come.

Yes, why not... It´s hard for the grammatical suffix to start an independent life while the original lexeme is alive and in frequent use in language.

I guess I understood why -memeli cannot mean ´doesn´t have to´. In Turkish modally modified verbs the negation marking is attached to the morpheme which is in fact denied. It is attached after it, not before it. Thus the negating morpheme -me- can only concern the verb root, not the necessitative -meli-. So, etmemeli has no other choice but to make ´it is necessary that he doesn´t do´. If we want to negate the necessitative only we have to use roundabout expressions like lazım as mentioned before in this thread.

What about değil? It is the usual way to negate an adjective. I even found occurances:

         Gitmeli değil miydik?

         Onunla birlikte melekler gelmeli değil miydi?

No. It gives different meaning:

Gitmeli = obliged-to-go (adj)

Gitmeli değil = not  obliged-to-go (not of that adj)

Gitmeli değil mi idik? = Were we not obliged-to-go? (didn´t we have to go? or wasn´t it the case that we had to go)

 

Even

Gitmemeli = obliged-to-not-go (adj)

Gitmemeli değil = not  obliged-to-not-go (not of that adj)

Gitmemeli değil mi idik? = Were we not obliged-to-not-go? (wasn´t it the case that we had not to go)

 

I leave the second example to you.

But that´s what I was looking for, si++. I was asking if değil also will do "if we want to negate the necessity only" which makes the meaning of ´not necessary, not obliged´. I mean -memeli cannot mean ´not necessary´ but maybe -meli değil can (just like gerekmiyor or lazım değil).

Edited (Nov 24) by Abla
Edited (9:09 am) by Abla