Sxak:

Tempodivalse:
Can someone provide other cases where adjectival roots turn into transitive verbs when they take -i? I am unable to think of one.

simil
sam
and "interes" is an adjective root by any dictionaries

I was a little surprised and confused the first time I saw the construction "simili ion"; I guess somehow I didn't expect it to have an object; I would have said "simili al io". But after that I realized that this use actually makes perfect sense, and I'm used to it now. But I can't recall seeing "sami ion"...

Sxak:

Tempodivalse:
Can someone provide other cases where adjectival roots turn into transitive verbs when they take -i? I am unable to think of one.

simil
sam
and "interes" is an adjective root by any dictionaries

Hmm. A good test of transitivity is to try to form a passive participle.

Estas similata/ita and Estas samata/ita don't make obvious sense, in contrast to estas interesata or estas egalata.

Are you sure this isn't just a case of the accusative replacing a preposition like je, according to rule 14 of the Fundamento?

Tempodivalse:Estas similata/ita and Estas samata/ita don't make obvious sense, in contrast to estas interesata or estas egalata.

But are you sure that's not just because in English we have "equalled", whereas we don't have "samed" or "similared"? A non English speaker might not necessarily find "egalata" any more intuitive than "similita" or "samata".

tommjames:

Tempodivalse:Estas similata/ita and Estas samata/ita don't make obvious sense, in contrast to estas interesata or estas egalata.

But are you sure that's not just because in English we have "equalled", whereas we don't have "samed" or "similared"? A non English speaker might not necessarily find "egalata" any more intuitive than "similita" or "samata".

I'm not seeing any attestation for the passive of simili, neither in the Tekstaro, nor elsewhere (aside from one dubious entry at the Greek Wiktionary). I have certainly never seen it.

The closest equivalent I can think of now is the Russian походить, however, this is also intransitive.

Oн походит на Алексея. (Он похож на Алексея.) = Li similas je Aleksejo.

Perhaps this is not much different from some other cases where we'd use je, for example, Li ridas je tio -> Li ridas tion (accusative substitution via Rule 14 of Fundamento). Trying to rewrite that as Tio estas ridata de li somehow sounds off - it's because tio isn't the object.

This is just splitting hairs, I know. Perhaps one should not care too much about transitivity at all, and just hope context will clear things up for you.

The problem then becomes that you can get away with almost anything:

Vi, edzino, estas mia vivo / vi la ĉiutaga voj' irata

This is poetry, so more leeway is allowed - but in normal texts I would find this construction patently wrong. It's a category mistake, just as *Mi brulas la lignon fails to see that bruli just isn't the sort of action that plausibly permits of an object.

Tempodivalse:The problem then becomes that you can get away with almost anything:

Vi, edzino, estas mia vivo / vi la ĉiutaga voj' irata

This is poetry, so more leeway is allowed - but in normal texts I would find this construction patently wrong. It's a category mistake, just as *Mi brulas la lignon fails to see that bruli just isn't the sort of action that plausibly permits of an object.

I'm afraif this so called "inner accusative" is common (mi iras - mi iras longan vojon; mi vivas - mi vivas komfortan vivon) and Zamenhofian. I'm too lazy now to check whether also Fundamental.

Kirilo81:

Tempodivalse:The problem then becomes that you can get away with almost anything:

Vi, edzino, estas mia vivo / vi la ĉiutaga voj' irata

This is poetry, so more leeway is allowed - but in normal texts I would find this construction patently wrong. It's a category mistake, just as *Mi brulas la lignon fails to see that bruli just isn't the sort of action that plausibly permits of an object.

I'm afraif this so called "inner accusative" is common (mi iras - mi iras longan vojon; mi vivas - mi vivas komfortan vivon) and Zamenhofian. I'm too lazy now to check whether also Fundamental.

I guess I want to know: it's an accusative, but is it also an object? Or simply a substitution for a preposition? If it's the latter, then the passive participle should not work.

Lia vivo estis komforte vivata actually doesn't cause immediate discomfort to my ear - though I think I'm influenced by English.

On the other hand, Russian does not permit of such passive constructions. You can add -ся (-iĝi) sometimes to indicate a passive of sorts (the informal живётся хорошо - something like "it [life] is lived well" ), but definitely not the participle. From a more Slavic perspective, these almost look likе anglaĵoj.

Kirilo81:

Tempodivalse:The problem then becomes that you can get away with almost anything:

Vi, edzino, estas mia vivo / vi la ĉiutaga voj' irata

This is poetry, so more leeway is allowed - but in normal texts I would find this construction patently wrong. It's a category mistake, just as *Mi brulas la lignon fails to see that bruli just isn't the sort of action that plausibly permits of an object.

I'm afraif this so called "inner accusative" is common (mi iras - mi iras longan vojon; mi vivas - mi vivas komfortan vivon) and Zamenhofian. I'm too lazy now to check whether also Fundamental.

Also known as cognate objects which exist in various languages. Live a good life, dance the first waltz, sleep the sleep of the righteous...

Tempodivalse:I guess I want to know: it's an accusative, but is it also an object? Or simply a substitution for a preposition? If it's the latter, then the passive participle should not work.

Ad 1: Yes, it is an accusative.
Ad 2: Yes, it is an object (complement of the verb and subcategorized by it).

However, this doesn't turn intransitive verbs miraculously into transitive ones. Those verbs will keep on behaving intransitively in all other environments.

Peter slept a troubled sleep. --> OK.
?A troubled sleep was slept by Peter. --> Critical.

Vikipedio's commentary on the subject, foro what it's worth. I need to see what PAG and PMEG have to say.

The cognate object per se does not bother me. Russian, my go-to language for parallels:

Жить счастливой жизнью. = Vivi feliĉan vivon, or literally Vivi per feliĉa vivo.

However, you can't put this in a passive construction. You need to change the verb.

Жизнь счастливо прожита. = Vivo feliĉe travivita.

This leads me to think the cognate object indicates nothing more than "faux" transitivity. But Tekstaro brings up:

la sumo de ŝiaj vivitaj jaroj estis tranĉita laŭ du partoj

ekzistis nur foresto de vivitaj impresoj

Interestingly, in this case the passive sounds OK, but active voice equivalents don't: *Vivi impresojn, and Vivi jarojn looks more like the accusative of time, not object accusative.

In any case, this doesn't seem to answer my question over the status of sami and simili.

This seems related (see best answer).

Also, I fail see a direct relation between transitivity and the possibility to form a passive.

I'm still thinking about simili. The first example that came to mind of a sentence in which it takes an object is this quote from the song "Procesio Multkolora":

Kajto:
La petolaĵoj, la malicaĵoj
De la marioneta rond'
Similas vere la kutimojn
De nia propra homa mond'.

This is a rather long sentence for using as a grammatical example. So, I will say "Marionetoj similas homojn." I think we've established that this is a grammatically correct sentence. Now the question is, is it correct to say "Homoj estas similataj de marionetoj."? For what it's worth, the Firefox spell-checker for Esperanto recognises "similata" as a word, but not "samata".

Interestingly, while we wouldn't usually use an accusative object with "loĝi", but almost always a prepositional phrase; one finds examples of the word "loĝata". We wouldn't usually say, for example, "Ana Pana loĝas Novjorkon", but "Novjorko estas dense loĝata" sounds normal.