Most of those English examples sound terrible. Some of them are clearly mis-translations from Japanese, a language in which even intransitive verbs can be passive, and passive verbs can be chained to your heart's content.

It's come up on Language Log tangentially, in a post by Mark Liberman last February (linking to Neal's post).

I'm also surprised that it hasn't been attempted to be discussed in Language hat.

This strikes me as one of those very common borderline cases where a stylistic principle is sometimes interpreted as (or attempted to be turned into) a grammatical rule. Rewriting eliminates the problem.

The building is scheduled to be demolished -- The building is scheduled for demolition next week

The piece was originally intended to be played on the harpsichord --> The piece was originally intended for the harpsichord.

Paul D: I think most of the examples I collected are from native English speakers; I left out those I identified as mistranslations or things written by nonnative speakers.

John E: The building is scheduled to be demolished isn't a problem at all; in fact, it's not even what I classify as a double passive (though usage manuals lump it under the same heading). If you have the active sentence They scheduled the building to be demolished, then your sentence involves just one passivization, of schedule. OTOH, there is no similar active form of a true double passive such as the building was attempted to be demolished. If you say, *They attempted to be demolished the building, it's ungrammatical even (I'd say) for the people who accept was attempted to be demolished.

The double passive has long stalked legal writing, at least in the US. Bryan Garner wrote about it in his Modern Legal Usage (1987) and, to much the same effect, in Modern American Usage (1998). His analysis parallels that in your posting. Many lawyers still seem to feel that they aren't earning their fees if they don't write something slightly off-kilter every few sentences, and the double passive is among their favorite tools. I have colleagues who simply cannot differentiate "the meeting was scheduled to be held" from "the meeting was attempted to be held," and I am powerless to help them. Perhaps the local chapter of Double Passives Anonymous will take them in.

What I was saying is that if double passives don't seem right, they should be eliminated by rewriting. For reasons of style, not grammar.

Passives are used when you don't want to specify a subject. So we're starting from "Someone we don't want to name has scheduled the demolition of this building by someone else we don't want to name". There are various ways of making that into a readable sentence, but the double passive we're talking about wasn't one of them.

"Don't want to name" sound over-dramatic, like "She Who Must Not Be Named" or J*H*V*H. Usually specifics are just left out in order to avoid raising complicating issues irrelevant to the topic you want to discuss.

Ten points for anyone who can write a triple passive.

Martin:

“The meeting was attempted to be rescheduled to be held on Tuesday, however schedules were conflicted and the motion was requested to be tabled.”

I would never say that, to me it’s essentially ungrammatical, but I've heard others use all the components before.

There are ten points to be had for any triple passive someone has attempted to be written.

I tried to write a maximum verb. I came up with "would have been being done". A modal past perfect progressive passive, I think.

Once I stick agents in there I get something that approximates grammaticality: “The meeting was attempted by Ford to be rescheduled by Arthur to be held on Tuesday by Zaphod.” I don’t get a WTF reaction from this sentence, but this may be because the complexity of the structure prevents me from processing the entire sentence as a whole.

Wait, I can do better:

"There are ten points to be had for any triple passive that has been attempted to be written."

John: "It would have been being done now had it not have been delayed being started then."

Probably modal present perfect progressive passive.

I suppose you can also string more passives onto those original sentences cited as "perfectly acceptable," or versions thereof, almost indefinitely, for example:

The building is believed to be scheduled to be demolished next week.

The building is rumored to be designated to be scheduled to be demolished next week.

The man is said to be planning to be designated to be scheduled to be operated on next week. (Quintuple, and counting.)

All the attempts at triple passives in the comments here have involved ordinary passives that just work out to being two passives in a row (as in was scheduled to be demolished), or well-accepted complex passives that nonetheless do not have an active counterpart (is rumored to be...). But here is a genuine triple passive I constructed in my draft paper: She was forgotten to be attempted to be contacted (i.e. someone forgot to attempt to contact her).

These contorted constructions are had been tried to be accounted for very well by Whitman indeed. At a pedant's cursory glance all that could be dared to be faulted was inconsistency in italicising punctuation, by which the mustard is hardly cut as useful criticism. O, and perhaps excessive length, now that of cutting is begun to be spoken.

It is seemed to by me this (though I don't know whether the point has been made by someone else). We make passives of forms like spoken of because we already make passives of forms like beaten up, even though the of is not dispensable in the way the up is:

Ali was beaten up.
Ali was beaten.

Let the war not be spoken of.
*Let the war not be spoken.

On a general softening (up) is brought by this sort of move, leading to, if not the worst excesses of the French Revolution, a revolting turn in our own tongue.

Here are the French examples that I found (with some typos fixed):

- Ils (=les IP loggers) permettent de déterminer les ports qui ont été tentés d'être ouverts, ou les machines envoyant du traffic ICMP.
- (...) ... l'esprit communautaire qui a été tenté d'être mis en place.
- Pour une fois ce n'est pas un personnage de film qui a été tenté d'être ennéatypé, mais l'histoire d'un bateau mythique (le Titanic) au travers d'un film à son image.
- Dans la plupart des disciplines, le problème de recrutement a été tenté d’ être résolu de la même manière: (...)
- Le topic a été oublié d'être fermé.
- Le passeport de Maurice a été oublié d’être tamponné ...
- Voila ce qui a été oublié d’être commenté dans le SCIENCE ET VIE (...)
- Une clause permet à ceux qui ont été oubliés d'être rajoutés et s'ils répondent aux critères, ils seront gradés comme les autres.
- (...) mais je dirais plutôt qu'ils ont été oubliés d'être arrosés !

I saw this construction "was attempted to be [something]" for the first time somewhere in the works of Patrick O'Brian. He uses it in dialogue somewhere. It was a surprise to me but I got the impression that it might be an Irish version of English, and I found it very infectious.

I'll have to hunt for the reference but I think someone says "it cannot be attempted to be done".

You won't find many examples in German because German has very tight restrictions on the AcI. You can literally translate "I see him work", but that basically is it. It's forbidden with "forget". It's also impossible with "attempt/try", because the outcome would be interpreted as "has been tempted". (Funny, actually. Normally German has lots more verb prefixes than English and uses them to disambiguate meanings that can only be distinguished from context in English.) This eliminates most possibilities -- and all of the French examples above.

That said, "the piece was intended to be played on the harpsichord" can be translated almost literally: Das Stück war dazu/dafür gedacht, auf dem whatever a harpsichord is, I'm too tired to look that up now gespielt zu werden. Note how far the two passives are apart (war ... gedacht, gespielt zu werden): so far that nobody gets the idea of giving the phenomenon of "double passive" a name.

Works in French, too, with the passives much closer, erm, more closely together: La chanson (disons) à été écrit pour être jouée sur le...

a language in which even intransitive verbs can be passive

:-o

What would "he has been come" mean, then? Or have I picked a bad example? Does it work like "he has been disappeared"?

There are ten points to be had for any triple passive someone has attempted to be written.

These are just two passives: "someone has attempted" is active. So, there are ten points to be had for any triple passive that is attempted to be written (by anyone).

You won't find many examples in German because German has very tight restrictions on the AcI. You can literally translate "I see him work", but that basically is it. It's forbidden with "forget". It's also impossible with "attempt/try", because the outcome would be interpreted as "has been tempted". (Funny, actually. Normally German has lots more verb prefixes than English and uses them to disambiguate meanings that can only be distinguished from context in English.) This eliminates most possibilities -- and all of the French examples above.

That said, "the piece was intended to be played on the harpsichord" can be translated almost literally: Das Stück war dazu/dafür gedacht, auf dem whatever a harpsichord is, I'm too tired to look that up now gespielt zu werden. Note how far the two passives are apart (war ... gedacht, gespielt zu werden): so far that nobody gets the idea of giving the phenomenon of "double passive" a name.

Works in French, too, with the passives much closer, erm, more closely together: La chanson (disons) à été écrit pour être jouée sur le...

a language in which even intransitive verbs can be passive

:-o

What would "he has been come" mean, then? Or have I picked a bad example? Does it work like "he has been disappeared"?

There are ten points to be had for any triple passive someone has attempted to be written.

These are just two passives: "someone has attempted" is active. So, there are ten points to be had for any triple passive that is attempted to be written (by anyone).

I know I've seen this exact subject mentioned in Danish - in one the newspaper advice columns about language (one of the reasonable, researched, non-prescriptive ones, i.e, not written by a "mendacious old windbag").

Unfortunately I can't locate any discussions of it now.

The only example I recall is "bedes afleveres" ("is asked to be delivered" (I think) as in "Bottles for recycling are asked to be delivered at the back door"). Google does turn up a few hits for this string.

It should "bedes afleveret" with a normal ppP ("is asked delivered"?).

I'm not good at coming up with examples, but "is asked to be answered" and
"is asked to be informed" give similar results.

As a French expatriate I find the French examples above to be very strange, in fact pretty awful. As a French teacher I would have suggested rewritings if English- speaking students had written such sentences. However, as a linguist I should be only mildly surprised at this proliferation of passives, which is typical of the translationese from English which has been overwhelming French written syntax, because of the massive and hurried translation of material originally in English (e.g. American news reports and commentaries, as well as other written works of a non-literary nature, including ads for American products).

It is getting to be so that I hate to read such works in French, even if written by French speakers: I don't recognize my own language in them and the awkwardness of the new style makes me cringe several times per page. But French-to-English translators must find their work becoming much easier as French syntax is losing its own specificity by aligning itself on English syntax.

- Une clause permet à ceux qui ont été oubliés d'être rajoutés et s'ils répondent aux critères, ils seront gradés comme les autres.
- (...) mais je dirais plutôt qu'ils ont été oubliés d'être arrosés !

This sentence contains a pun on two types of passives:

1) une clause permet [à ceux qui ont été oubliés] d'être rajoutés:
one clause allows those who were forgotten to be added (= those who were forgotten can now be added)

Here the second passive is a complement of the preposition de after the verb permet, not after the passive être oubliés) - this is traditional syntax and the justaposition of the two passives is fortuitous.

as against

2) ils ont été oubliés d'être arrosés = they were forgotten to be "watered" (a word I hesitate to translate literally without knowing the context) (= someone forgot to "water" them)

Here the preposition comes right after the passive ont été oubliés, and the sequence of passives qualifies as one of the new double passives. This sentence structure is one that I find totally foreign.

I would have written: On a oublié de les arroser, but the very useful and general on seems to be on its way out in written French as it rarely has a direct equivalent in English.

I wonder whether languages with a middle voice wouldn't gravitate toward it for these chains, because the agency gets so fuzzy.

Languages with a middle voice? Do you include English, MMcM? Certainly agency and patiency get fuzzy in English because of active, middle, and passive uses of our "active" forms (alternative analyses in terms of ergativity aside). But what did you have in mind? Can you give an example that might support your interesting speculation, preferably an English one?

I meant a morphologically apparent middle voice, like Sanskrit, Greek or Icelandic. Middle voice is common enough after a verb of causation. So the question is, does it get even more prevalent in a chain. I tried looking for a few variations of að something-ast að something-else-ast in the wild, but I may just not have thought of the right verbs. Or there may be some famous supporting or counter example from some ancient philosopher that I'm unable to call up directly right now.

Marie-Lucie,
The French double passives also surprised me because I could think of a rewording with the indefinite pronoun "on" for each of them.
The Dutch pronoun "men" (the counterpart of "on" in French and "man" in German) is still going strong. I think that's why I have found only two double passives in Dutch so far:
- “Alle oude culturen, alle oude geloven, zijn geprobeerd omvergeworpen te worden met als grondslag haat.” (”All old civilizations, all old religions, were tried to be overthrown based on hatred.”)
- “Dus je ziet hoe de A-G zo argumenteert dat de feitelijke en harde informatie wordt versluierd en geprobeerd wordt ontkracht te worden.” (Roughly: ”So you see how the A-G reason(s) such that the factual and hard information is being obscured and being attempted to be disproved.”)

It may be fairly far afield, but Turkish, which loves the passive voice in bureaucratic, academic, and "elevated" writing, often uses double passives for impersonal expressions with modal verbs. I ran across such a sentence today on the internet:

PKK K. Irak’ta kurulacak Kürt devletinin Türkiye’ye kabul ettirilmesinin bir aracı olarak kullanılmak isteniyor olabilir. (Literally "The PKK may be being desired to be used as a means for Turkey to be made to accept a Kurdish state that would be established in N[orthern] Iraq.")

"Isteniyor" is the present passive of "to want", i.e., "it is being wanted/desired", with the "olabilir" coming afterwards adding a note of possibility.

A more idiomatic translation for the sentence as a whole would be something like "There may be a desire for the PKK to be used as a means to get Turkey to accept a Kurdish state that would be established in Northern Iraq."

In my last post on double passives, I give some examples from Turkish that Jaklin Kornfilt wrote about in a few papers. According to her, Turkish even allows recursive double passives (i.e. "triple passives"), but double passives can occur only with three verbs: 'want,' 'try,' and 'begin.'

but the very useful and general on seems to be on its way out in written French as it rarely has a direct equivalent in English.

Are you sure? Isn't it rather avoided because it has become ambiguous, now that its most common use is "we"?

The Dutch examples (three comments above) remind me that in German not just the more standard and less regional word for "try/attempt", versuchen, lacks a passive (that's the one that also means "tempt"), but so does the more Austrian probieren, which would be unambiguous in the passive. A both literal and idiomatic translation of the Dutch examples is impossible.

Hm... actually versuchen can be put into the passive, but only when the context makes it extremely obvious that "tempt" is not meant: das ist bereits versucht worden "this has already been tried". And even this reeks of "politician's passive" ("mistakes have been made"). Strange, actually, because people don't talk about tempting except in church, and even then rarely!

Stupid me. I got it right yesterday night. What is missing in German is not the passive of "try" and "forget", but the possibility of making an AcI with the passive of these verbs. "They have been forgotten" is idiomatic, "they have been forgotten to be [verb]" is impossible.

(French on)

The indefinite subject pronoun on has also had the meaning 'we' for a very long time, although this was not obvious in written French where nous was preferred: when quoting direct speech, for instance in citing interviews, journalists automatically translated the colloquial on as the more literary nous (as the subject of a verb). This was very obvious in quoting children's speech, where the quoter would transform the children's normal on into the stilted nous (which has to be followed by a more specific verb form). Ambiguity in speech between the two meanings of on is extremely rare in practice: if there is a possibility of doubt, the meaning 'we' is normally indicated by using nous, on .... for 'we', instead of just on.

The huge increase in passives seems directly related to literal translations from English, causing people to become accustomed to reading a lot of passive forms, therefore to start using them in spontaneous writing. With indefinite on disappearing from written sources as the result of the overuse of passives, and with greater tolerance for some features of the spoken language, on meaning 'we' has become more acceptable and prominent in writing. This does not mean that indefinite on is disappearing from spoken speech.

However, there is yet another pressure from English on indefinite on, that is the use of vous or tu as indefinite, equivalent to the indefinite use of you in English - again a result of literal translation.