The following six Spanish sentences are different versions of the question/different questions corresponding to the unmarked declarative sentence Alguien más podría haber estado usando su ordenador (= Somebody else could have been using your computer) and all but (6) are acceptable, but how exactly are they derived under current minimalist assumptions, why is (6) excluded, and what consequences follow as to which of them is/are more ‘economical’ (and presumably ‘unmarked’) and which are less ‘economical’, ‘marked’ and reserved for special ‘informational’ uses?
(1) ¿Podría haber estado alguien más usando su ordenador?
(2) ¿Alguien más podría haber estado __ usando su ordenador?
(3) ¿Podría haber estado __ usando su ordenador alguien más?
(4) ¿Podría alguien más __ haber estado ___usando su ordenador?
(5) ¿Podría haber estado usando alguien más __ su ordenador?
(6) *¿Podría haber alguien más estado __ usando su ordenador?
[Word-by-word gloss: podría = could; haber = have; estado = been; usando = using; su = your; ordenador = computer; alguien = somebody; más = else]
As far as I know, under any version of late P&P or ‘minimalist’ syntactic theory that assumes the VP-Internal Hypothesis and a conventional hierarchy of projections – say, at least, CP > FocP > TP > PolarityP > ModalityP > AspP2 > AspP1 > VP (where ‘>’ = ‘dominates’) - and leaving aside and open the question of how verbal inflections, ‘phi features’, +Q, etc. are to be licensed/checked, since whatever technical mechanism is adopted in this respect will have to operate in all six sentences and, economy-wise, would not raise differences among them - I suppose it would be possible to derive each of them as follows:
Sentence (1) could be derived by leaving everything in situ (= zero displacements, maximal economy) and if so should be the optimal ‘solution’ and, presumably, the ‘unmarked’ version of the question. [Alternatively, of course, (1) would involve at least (Aux)V-Raising into Foc preceded by earlier (Aux)V-raising from Mod into T and lower (Aux)V-to-Infl raisings of usando, estado and haber into the respective Infl heads just above them, but if that is the mechanism by which verbal inflections and other features are ‘checked’, ‘licensed’ or whatever, then all other sentences (2-6) would also involve such displacements and we may as well disregard all such moves for the purpose of establishing which of (1-6) are more or less ‘economical’ and which unmarked/marked].
Sentence (2) could minimally be derived (leaving V-Raisings aside) by ‘long’ Subject-Raising (= one displacement, assuming that ‘long’ Subject Raising counts as a single operation, since raising into any intermediate specifier would presumably be gratuitous and impossible = a violation of Economy). However, if the subject may simply stay in situ, as assumed in the earlier derivation of sentence (1), its raising in (2) would seem to be gratuitous (a violation of Economy) and (2) should be bad. Since it is not, the derivation of (1) might be more complex than it seems and (1) would no longer qualify as the optimal solution and the ‘unmarked’ construction.
Sentence (3) could minimally be derived (again, leaving V-Raisings aside) by Subject-Postponement (= adjunction to VP) presumably triggered by the need to leave the subject in ‘End-Focus’ position (= one obligatory displacement, or two, if the subject must first raise to Spec TP, although that would be problematic, as just explained for (2)).
Sentence (4) could minimally be derived (again, leaving all V-Raisings aside) by ‘long’ Subject Raising, but with the same potential objection raised wrt sentences (2) and (3).
Sentence (5) would require just overt V-Raising of usando into the Infl head immediately above, but, if that is how it is derived, something must explain why usando must rise in (5) but not in (1), and the derivation of (1-4) might prove more complex than it seems.
Finally, the non-derivability of sentence (6) could be explained simply as a violation of Economy, as there seems no reason for the subject to raise into an intermediate specifier instead of directly into Spec TP, but since even the ‘long’ subject-raising into Spec TP looks like a possible violation of Economy to the extent that sentence (1), with the subject in situ, is grammatical, it is unclear to me whether that explanation stands.
I have deliberately used established terminology hopefully accessible to scholars familiar with some versions of Chomskian syntax, if not necessarily with the latest ones (which is about my own plight, I am afraid). The many ‘if’s’ and hedges above simply reflect the fact that it is unclear to me how the existence of sentences (1-5) and the non-existence of sentence (6) can be accounted for under current ‘minimalist’ assumptions. If somebody here can tell me what the current wisdom is in this respect, I will be very grateful.