No. Word order, in this case (i.e., passive construct) doesn't alter the integrity of the semantic chain:
appoint someone something
someone is appointed [to be] something
Ex: Berkensauer is appointed [to be] captain.
Could the insertion of [to be] there tell us that
captain is an object complement?
I think, now we're on the same page. In my terminology, we'd have - in the passive sentence - an object complement complementing a subject. The subject is not merely a case of word order; the subject behaves in different ways concerning verb agreement (They appoint --> Beckenbauer is appointed) and case (see pronoun substituion "appoint him" --> "he is appointed).
If we were talking about German, where the case system is implemented not only on the pronoun level, we'd see that "captain" is in the dative case (following a preposition), a case object complements share with indirect objects. If we accept that English has a covert case system similar to German, than the translation test would also point towards an "object complement" (more accurately called "objective-case complement").
Thematic Roles
I think semantic roles play an important part but are not the primary criteria here. The key in determining whether a noun functions as an object or a complement is in whether or not its verb is non-stative and, most importantly, what kind of non-stative it is?
We do seem to be speaking primarily about participant roles, not so much about thematic roles. But "thematic roles" are again different from "subject" and "object", in that thematic roles change with inversion: "Him I can't abide!"
Jumping ahead, but the question is significant, could the noun
Berkensauer function as a complement, below, given its verb
appoint?
Ex: Germany appointed Berkensauer captian.
I don't think so. But it's a very interesting question.
First, complements and the nouns they complement are not always exchangeable, as the semantic relationship is often one-way.
Sometimes, their semantically exchangeable, and all that changes are the thematic roles:
Pixie is the name of my cat. --> The name of my cat is Pixie.
But at other times, the semantic relationship isn't equivalence; rather there a different levels of abstraction involved.
Pixie is a cat. --> *A cat is Pixie.
Similarly, "captain" is more abstract than "Beckenbauer", so that "captain" is a possible aspect of "Beckenbauer", while "Beckenbauer" is one element of the set of "captain". The verb appoint cannot - according to its meaning - "act" on an element of the objects set; rather it confers a new aspect on the object. I think that verb semantics would even override "unusual wordorder", given that you know about "Beckenbauer", "captains" and "appointing":
They appointed captain Beckenbauer.
Interestingly, though, if you remove the noun Beckenbauer, and talk in general only, you can have something like this:
They have the power to appoint captains.
And then even passivisation is possible:
Three captains were appointed by them this year, but none lived up to their expectations.
So apparantly it's possible to promote the object complement to object, by deleting the object. Interesting.
1. I lost face.
2. I lost heart.
Let's test them:
1. Face was lost.
2. Heart was lost.
If you can accept the grammaticallity of the passive forms, then both
face and
heart are direct objects. If not, and the meanings sit awkward with you, then they are complements. Here's another example,
Max hit the roof.
Literal: She used her hand to slap the roof.
Figurative: She got angry.
In which example does the noun
the roof function as an object?
The roof was hit. :tick:
=> She got angry :cross:
My hypothesis is that in idioms such as "She hit the roof!" the unavailability of the passive voice is not a function of grammar, but of pragmatics. Idioms work on pattern recognition, and breaking the pattern may keep us from recognising the idiom. Of course, this needs research.
So, basically, in expressions such as "She hits the roof," idiomacity prevents passivisation, but "the roof" still functions as an object. It's obviously still a patient: the roof is still affected by the hitting; that the entire action of "hitting the roof" is figurative doesn't change that, and neither does the fact that the collocation is idiomatic.
What I think we need is a methodology to assess the reliability of the passivisation method. Of course, if you
define complementation over the passivisation, then that's that. But I wouldn't consider that a very useful definition. We wouldn't have to go to semantics at all, for this, since passivisation can be described in purely syntactic/formal terms.
The way I see it, if a noun isn't an object, then it isn't acted upon by its verb. Which tells us we are dealing with a verb that's neither dynamic nor stative, but rather somewhere between the two.
I'm not sure how you're using "dynamic" and "stative" verbs here. To my mind, this is entirely unrelated to having an object or not; it's about progressive aspect or not, among other distinguishing features. Having objects is not one of them. Both stative and dynamic verbs can have objects.
***
After thinking about this a while, I think I'd just call "trains" in "changing trains" an adverbial, as it relates very closely to the verb, even before arguments are taken in consideration. To me, "trains" sounds more like part of the process than like a participant role.
As far as participant roles are concerned, I'd ask who's involved with "changing trains". If the context takes care of "trains", you can just leave it off.
"Take the train for
Brighton and change at
East Croydon Station for
East Grinstead then follow directions above." (
here)