Tempodivalse:This is a real mess, the more I look into it. The differences between PAG and PMEG are very illuminating, but one walks away not sure which, if either, to believe.

@NoordZee, PMEG actually directly discusses the use of passive okupi. The gist is that the actual difference between estas okupita and estas okupata is minimal - similar to what I said in my first post in this thread. So far, my intuitions have not led me astray.

But according to PMEG, there is an asymmetry between -int and -it that I failed to account for.

Translation: "-INT forms always show an action that happened before another action. -IT forms show an action, or the fulfillment of an action, which gave a result. -IT forms can indeed show a time earlier than some other time, but very often it is not so."

Later: "-ANT forms always show, that an action is continuous, unfinished, or repetitive. -AT forms in normal Esperanto show precisely the same nuance."

Remarkably, the example sentence is: Hieraŭ la perditaj aferoj estis trovataj. In other words, the lost things were re-found yesterday, over some period of time.

But this contradicts PAG, which indeed explicitly lists trovi as a result-only verb that should use only the past participle.

PMEG is closer to my intuitions that esti + participle should be parsed as two simple tenses, one in relation to the other. There is no talk in PMEG of imperfect or perfect.

I hope I haven't provided too much information or made things even more confusing. Maybe the upshot is that one should use -iĝ to indicate passive instead, where convenient. Much simpler.

EDIT: I just saw your reply, NoordZee. I was so fascinated by the point brought up by Sudanglo that I forgot that your query was initially about something different! Looks like all this was one large digression, my apologies.

There is absolutely no problem with digressing from original questions. In the spate of the 2 pages of responses so far, I have found out a lot thanks to the input. I am going to spend much more time reading in both the PAG and PMEG. I got both of them on my computer. You are quite right. Things are not so straight forward as many Esperantist would have you believe.

There is no need to get bogged down with deciding whether this or that verb falls into a particular category. All you need is a knowledge of the world and an awareness of what you want to say.

In the nature of the world the breaking of a vase is not something that you would normally see as having temporal duration.

However if you are a materials scientist and have filmed the breaking of a vase with a high speed electron microscope camera, then you may say Dum la vazo estis rompata, ni observis ke la kristala strukturo estas unue deformita .... But for most of us a vase is intact and then it is rompita.

If you are looking for somewhere to sit in a theatre during the interval, when some patrons may have gone to the bar or be taking a toilet break, you might enquire of a vacant chair Ĉu ĉi tiu seĝo estas okupata?, as you conceive of the occupation of the chair as having started before the interval and going to be resumed after the interval (duration).

But equally you might view the critical issue to be the result of the initial occupation (with its implication of resumption rights) and enquire Ĉu ĉi tiu seĝo estas okupita.

As regards Hieraŭ la perditaj aferoj estis trovataj, I would think it more natural to be concerned with the state of being found (no longer lost) rather than the process, so would say 'trovitaj'

In the nature of the world the breaking of a vase is not something that you would normally see as having temporal duration.

Surely one can break an object over some extended period of time. With the vase, perhaps this isn't obvious - our imagination might immediately lead us, for instance, to it falling off a table, rather than to it being more gradually hacked apart with, say, a small trowel.

But there are plausible everyday scenarios where something is broken bit-by-bit. "Mia aŭto estas rompata de malkompetentuloj!" - shrieks the unhappy owner of the Mercedes in Bob's Car Repair.

This is why I disagree with PAG - it has a very unimaginative, rigid idea of how certain verbs should be used, that does not correspond to usage, nor to intuitions.

But you're right, there is no particular need to get pedantic here. I apologise for the excess verbiage - but I learned a lot from it, and this facet of verbs highly interested me.

tempodivalse:La vazo estis rompita. = The vase had been broken already, relative to some past event. (Normally similar to the pluperfect in English.)

'Estis X-ita' passives can have two meanings in Esperanto. What you have given is one of them, but in my own experience it's the less common one. It's more common for "La vaso estis rompita" to mean the vase was broken right at that time.

It's the same situation in English, which also forms the passive voice periphrastically with an auxiliary verb. "The vase was broken" might mean either the vase got broken right then, or the vase was in a broken state (because someone broke it earlier).

tommjames:It's more common for "La vaso estis rompita" to mean the vase was broken right at that time.

Yes. But conceptually, I wonder how great the difference is. Consider the possible interpretations of La objekto estis rompita.

At the time of estis, that is, relative to some past event,

a) the object had been broken (and finished being broken) right then.

b) the object had already been broken for some time. (Closer to pluperfect)

In neither case is it assumed that the event of being broken was still ongoing at the time of estis. The difference is one of immediacy only - technically both are past events and hence fair game for -ita - at least if one accepts my premise that Esperanto ultimately contains only simple tenses that, by themsleves, can carry no further aspectual, temporal, etc. information.

If we were witnessing an object being broken piece by piece, we would need to use rompata. Yes, this isn't the way things are characteristically broken, but we don't need to imagine a particularly bizarre situation where that might happen.

Consider the following line out of La Koro de la Serpento:

-> La veninta silento estis rompata nur de plioftiĝinta spirado de la emociiĝintaj teranoj.

Or, here's one from Doktoro Jekyll kaj Sinjoro Hyde:

-> La senbrueco estis rompata nur de la sono de piedo.

Surely the shade of meaning would be quite different with rompita.

This is why I disagree with PAG - it has a very unimaginative, rigid idea of how certain verbs should be used

It is undeniably true that with certain verbs -ata is more commonly used, with others -ita is the norm, and with others one may encounter either -ata or -ita.

But it is not the case that there is a grammatical rule that certain verbs impose a certain passive form. It's all to do with what those verbs represent. When I said that rompi is a verb of result without duration, I was pointing out the meaning of rompi, not putting it into a special grammatical class.

there are plausible everyday scenarios where something is broken bit-by-bit. ... If we were witnessing an object being broken piece by piece, we would need to use rompata.

Really? Would rompi be the right verb here anyway.

The essence of rompi is discontinuity. At first the relevant object is intact and then it is broken, or no longer functional, or in the case of something more abstract disrupted (eg break a silence).

A breaker's yard may break up a car (over a period of time). But the idea of Bob's Car Repair breaking the customer's car seems odd, though of course they could break a headlight, then rip a seat, then fail to reconnect a hose. At what point is the car broken?

In the case of -ata, quotations from published works may not be convincing, as certainly some books were published in the period of the ata/ita debate with atismoj.

At one point I had some in my personal library. The authors/translators were of a Dutch or German or Danish mother tongue background.

From Marta (Trad Zamenhof)

ekregis granda silento, interrompata nur de la kraketado de la fajro.

This creates an image of silence interrupted from time to time by noise of the fire (-ata here by virtue of repetition).

La veninta silento estis rompata nur de plioftiĝinta spirado de la emociiĝintaj teranoj.
La senbrueco estis rompata nur de la sono de piedo.

The thing about silence is that unless the noise is absolutely continuous the the disruption is very likely silence noise, silence noise, silence noise and so on.

Tempodivalse:Consider the following line out of La Koro de la Serpento:

-> La veninta silento estis rompata nur de plioftiĝinta spirado de la emociiĝintaj teranoj.

Also consider Zamenhof himself:

LLZ:Post la eliro de la knabino en la ĉambro ekregis granda silento, interrompata nur de la kraketado de la fajro, kiu brulis sur la kameno, kaj de la bruo de la strato, kiu obtuze kaj neklare atingadis la mansardon.

In the Fundamento and hence canonical:

LLZ:Sed, prezentante parton de la fundamento, tiuj ĉi arĥaismoj neniam estos elĵetitaj, sed ĉiam estos presataj en ĉiuj lernolibroj kaj vortaroj samtempe kun la formoj novaj, kaj tiamaniere ni havos la certecon, ke eĉ ĉe la plej granda perfektiĝado la unueco de Esperanto neniam estos rompata kaj neniu verko Esperanta eĉ el la plej frua tempo iam perdos sian valoron kaj kompreneblecon por la estontaj generacioj.

I don't think that this rompata conveys any aspect or aktionsart. It is a plain passive.

The fact that many atists were or are German is obvious due to the Fundamento:

LLZ:ŝi est'as am'at'a de ĉiu'j, sie wird von Allen geliebt.

sudanglo:Really? Would rompi be the right verb here anyway.

In English, in some contexts it will sound perfectly normal to say "is being broken", "was being broken" (presumably to indicate duration, as opposed to "is broken", "was broken" ). "My doghouse is being broken by the neighbourhood hooligans," complains the house owner to the police over the phone, while looking out the window concernedly. This seems generally equivalent to Esperanto's rompata - though I'm loath to draw connections between the two languages this way.

I am not at your level of Esperanto knowledge but have followed the discussions with great interest. Just a few more observations.
Whether one translates 'Ŝi estas amata de ĉiuj' into German as 'Sie wird von allen (von jedem) geliebt'(Alle lieben sie) or into Dutch 'Zij wordt van iedereen gehouden'(Iedereen houdt van haar)it is clear to me that this state of being loved is continuing from the past when it started into some undetermined future. I would indeed be inclined to use the -ata form.

As regards the conversation about 'rompi', what do you think of this: you walk into a room and discover a broken vase. You could call out: "Oh, this vase is broken." "Ho, tiu vaso estas rompata." You describe a present state or would you still prefer the -ita form?

I continue to follow this discussion with interest.