In order to make that conclusion [that this construction is grammatical], wouldn't you need to know the frequency with which [it] appears relative to other constructions that express the same idea? Or, perhaps, the frequency with which native listeners and readers have an adverse reaction to it? After all, if the phrasing struck *you* as wrong, and it struck *me* as wrong, I'll bet it strikes *lots* of people as wrong.What I found with my corpus search was that it was common and regular enough that I could not in good conscience simply dismiss it as a speech error. I found enough examples that, if this were a foreign language I were documenting, the prima facie conclusion would be that for at least some speakers, the way to turn a sentence like Someone attempted to kill others into a passive is to make both attempted and kill passive. Now I agree that there are probably a lot of speakers who, like Glen and me, find the sentence ungrammatical, or at least odd. From a linguist’s point of view, all this means is that people like Glen and me are speaking a different language from the people who have this construction. True, the languages are mostly alike, enough so that the speakers of each can easily understand each other, but they are different languages nonetheless. A more common way of referring to languages that are mutually intelligible (as these two are) is to call them dialects of the same language.
It might even strike some of the speakers as wrong. I can easily imagine saying trying to construct a passive voice sentence, realizing halfway through that it couldn't *quite* be done, and finishing incorrectly rather than starting over.Good point, if we’re talking only about spoken English. However, the other examples I found were in written English, which has had more time to be corrected by its authors. But even granting this point, suppose that all my examples are instances of people completing a sentence even though they know it’s not grammatical, just because they can’t think of a better way of doing it in the time they have. Now imagine you’re a kid learning the language. When you hear these sentences, as far as you know they ARE grammatical, and you infer a rule to generate them, without the mental baggage that the earlier speakers may have attached to them. And now in your grammar (and in the grammar of many of your peers who were exposed to similar input), these sentences are a completely legitimate part of your language.
As Mario Rizzo and Glen might put it, a “slippery slope event” has occurred: Letting the ungrammatical sentence pass has allowed speakers acquiring the language to accept it as grammatical. And now that their grammar will generate this kind of stuff, who knows what other surprises will be generated by whatever rule they’ve inferred? And from the rules that the next generation of speakers infer from what they hear this generation saying? This kind of slippery slope is a common way for language to change; John McWhorter, in The Power of Babel, describes it quite entertainingly. The process is so common that I even have an example from my own life, but since this posting is long enough, I’ll save it for my next one.