Preposition stranding

[dotdv]

I made some updates to "preposition stranding" section:

1) Originally it said to add a remote C and I changed it to "add the preposition's object as a Remote" 2) I added an example with a Remote P: ``[The wedding]$_P$ [I went [to$_R$ \rem{wedding}$_P$]$_A$]$_A$ was$_F$ beautiful$_D$

• do you agree with it? we haven't really discussed yet what to do in cases such as this where a relative clause elaborates a P/S 3) I disabled the example with 'work' because it's not clear whether it should be a P or a C, no?

[nschneid]

"the wedding I went to": I don't think "I went to" is an A-scene. I think the relative clause requires an elaborator:

[the wedding_P]_C [I went [to_R (wedding)_P]_A]_E

[omriabnd]

Nathan, I think your solution may lead to problems is there are more elements to the Scene.

[the_F service_P]_C [I_A witnessed_P]_E wasF efficient?

In fact, it seems that the relative clause here either takes scope over the entire scene, in which case it should be:

[the_F service_P]_C- [I_A witnessed_P (service)_A ]_E [was_F efficientD ]-C?

OR:

[the_F service_P]_A- I_A witnessed_P [was_F efficientD ]-A?

On Tue, Nov 13, 2018 at 5:18 PM Nathan Schneider notifications@github.com wrote:

"the wedding I went to": I don't think "I went to" is an A-scene. I think the relative clause requires an elaborator:

[the wedding_P]_C [I went [to_R (wedding)_P]_A]_E

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-438303117, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG861BkrTQwOd1O5P7E-faMOTPMt8GMks5uuuK5gaJpZM4YbdtT .

[nschneid]

I would have done:

[[the_F service_P]_C [I_A witnessed_P]_E]_A was_F efficient_S

[omriabnd]

And what about "The service was efficient". Don't we annotate it as?

the_F service_P was_F efficient_D

On Sun, Nov 18, 2018 at 5:46 PM Nathan Schneider notifications@github.com wrote:

I would have done:

[[the_F service_P]_C [I_A witnessed_P]_E]_A was_F efficient_S

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-439702453, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG866b8SyvmNloxBFhu30Tj6s-bq8cRks5uwYDVgaJpZM4YbdtT .

[dotdv]

And what about "The service was efficient". Don't we annotate it as? the_F service_P was_F efficient_D

Yes, that's the solution we've been following

[omriabnd]

OK, so what do you think about the solution of making it two Hs, with a remote:

[The_F service_P ]_H- [I witnessed (service)_P]_H [was_F efficientD]-H

?

On Mon, Nov 19, 2018 at 10:36 AM dotdv notifications@github.com wrote:

And what about "The service was efficient". Don't we annotate it as? the_F service_P was_F efficientD … <#m-489297567830099087_> On Sun, Nov 18, 2018 at 5:46 PM Nathan Schneider @.***> wrote: I would have done: [[the_F service_P]_C [I_A witnessed_P]_E]_A was_F efficient_S — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#47 (comment) https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-439702453>, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG866b8SyvmNloxBFhu30Tj6s-bq8cRks5uwYDVgaJpZM4YbdtT .

Yes, that's the solution we've been following

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-439811536, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG86zYftAi-wZcPqKiKt57hNXk1jHI_ks5uwm2MgaJpZM4YbdtT .

[omriabnd]

I meant: [The_F service_P ]_H- [I witnessed (service)_A ]_H [was_F efficientD]-H

[nschneid]

If "the service was efficient" is

the_F service_P was_F efficient_D

then UCCA is removing the information structure of where the main syntactic predication is, making it equivalent (modulo was_F) to "the efficient service". In which case more information structure is preserved when adjectives modify non-scene-evokers (attributive "the [tall_S (boy)_A]_E boy" vs. predicative "[the boy]_A was tall_S").

If that is the policy, then I think "the service I witnessed was efficient" should be the same as "I witnessed the efficient service" / "I witnessed the service (that) was efficient", i.e.

[the_F service_P]_A- I_A witnessed_P [was_F efficientD]\-A

[omriabnd]

That's acceptable in my view. Dotan?

[dotdv]

That's acceptable in my view. Dotan?

I see that you continued this discussion in other issue ( #48 ) and that your conclusion was that you prefer the two H solution. If I understand correctly, you prefer that we generally use the two H solution when dealing with RCs over P/S so that's what I'll follow unless you raise it up again.

[omriabnd]

Yes. Let's do the two Hs

On Thu, 22 Nov 2018, 06:46 dotdv <notifications@github.com wrote:

That's acceptable in my view. Dotan?

I see that you continued this discussion in other issue ( #48 https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/48 ) and that your conclusion was that you prefer the two H solution, so that's what I'll follow unless you raise it up again.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-440911589, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG868Z0aM394HJ7Dn47_xcUdopxb8Pxks5uxiwIgaJpZM4YbdtT .

[dotdv]

The guidelines re stranding prepositions refer to cases where the A is missing: "In some cases, an A is missing but its preposition is in place. We mark the preposition as an A, with an R inside of it, and add the preposition's object as a Remote" ``The_F book_C [I_A 'm_F looking_P [for_R \rem{book}_C ]_A]_E''

What do we if the A is not missing from the scene (see examples below)?, if we follow the guidelines in such cases as well, it seems like we are adding an unnecessary Participant.

[This path]_A has already been walked [on_R (path?)_C]_A What_A are_F you_A talking_P [about_R (what?)_C]_A?

We were wondering if we need to consider an alternative solution for these cases?

[nschneid]

What do we if the A is not missing from the scene

Indeed, it doesn't make sense to have something as a participant AND a remote participant of the same scene. Shouldn't it be a discontiguous unit in those cases?

[[This path]_C]_A– has already been walked [onR]\–A [[What]_C]_A– are you talking [aboutR]\–A

[omriabnd]

makes sense

On Wed, Jan 23, 2019 at 4:12 PM Nathan Schneider notifications@github.com wrote:

What do we if the A is not missing from the scene

Indeed, it doesn't make sense to have something as a participant AND a remote participant of the same scene. Shouldn't it be a discontiguous unit in those cases?

[[This path]_C]_A– has already been walked [onR]–A [[What]_C]_A– are you talking [aboutR]–A

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/47#issuecomment-456815362, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG863Ets8MrZDzAMJpFUL4lNZ7xv9izks5vGG3HgaJpZM4YbdtT .

[dotdv]

Yes, makes sense, thanks.