The Position of Perché

What about the position INT in main clauses? As main yes/no questions are not introduced by any special morphological marker in Italian, we do not have direct evidence on the presence and position of INT in main questions. Nevertheless, there are good reasons for concluding that a small class of Wh elements fills a higher position than ordinary wh elements, and it is tempting to conjecture that this position is INT. Let us consider the relevant evidence.

Wh elements corresponding to arguments or lower adverbials require inversion in Italian main interrogatives, as is shown by the fact that the subject cannot intervene between the Wh operator and the inflected verb:

(16)a * Che cosa Gianni ha fatto?

             ‘What Gianni did?’

      a’    Che cosa ha fatto Gianni?

             ‘What did Gianni?’

       b * Dove Gianni è andato?

             ‘Where Gianni went?

       b’ Dove è andato Gianni?

            ‘Where went Gianni?’

       c  * Come Gianni è partito?

              ‘How Gianni left?

       c’   Come è partito Gianni?

               ‘How left Gianni?’

This may be due to the Wh Criterion (Rizzi 1996): the inflectional node carrying the feature Wh must move to the C system in interrogatives to create the required Spec-Head configuration with the Wh operator, and as the word order shows, this operation has not taken place in (16)abc. I to C movement has applied in (16)a’b’c’, which then satisfy the Wh Criterion (the subject appears in final position in these examples, and not in between the Aux and the past participle possibly for case theoretic reasons, along the lines of Rizzi & Roberts (1989)).

The necessity of I to C movement in these cases is also illustrated by the distribution of certain adverbs: short adverbs like già, ancora, solo, typically occur in between the auxiliary and the past participle, but may also occur, at least in some varieties, immediately to the left of the auxiliary (or other inflected verbs):

(17)a    I tuoi amici hanno già fatto il lavoro

            ‘Your friends have already done the work’

       b   I tuoi amici già hanno fatto il lavoro

            ‘Your friends already have done the work’

(18)a   Che cosa hanno già fatto?

            ‘What have (they) already done?

       b * Che cosa già hanno fatto?

             ‘What already (they) have done?’

(19)a   Voi siete già andati a Milano

            ‘You have already gone to Milan’

      b   Voi già siete andati a Milano

            ‘You already have gone to Milan’

(20)a   Dove siete già andati?

           ‘Where have (you) already gone?

       b  * Dove già siete andati?

               ‘Where already (you) have gone?’

Evidently, the richly articulated structure of the IP (Cinque 1998) specifies a position for such adverbs in between the subject and the inflected verb in cases like (17)b, (19)b. But in questions like (18), (20) the inflected verb must move to C to satisfy the Wh Criterion, thus bypassing this high adverbial position; whence the ill-formedness of (18)b, (20)b.

A number of alternatives to the analysis involving the Wh Criterion and I to C movement have been explored in the literature. The evidence based on adverbial distribution (18)-(20) is important because it suggests  that the facts of (16) should not be  uniquely attributed to some peculiarity of the preverbal subject position, as some alternative analyses propose.  I will not discuss these issues here;  what is relevant in the context of the present discussion is that some Wh operators behave differently from ordinary operators like those in (16), (18), (20). Perché (why) and other higher adverbials, like the near synonym come mai (how come), do not require inversion:

(21)a   Perché Gianni è venuto?

            ‘Why Gianni has left?’

      b  Come mai Gianni è partito?

           ‘How come Gianni has left?’

They are also consistent with short adverbials preceding the inflected verb:

(22)a   Perché (i tuoi amici) già hanno finito il lavoro?

            ‘Why (your friends) already have finished the work?’

       b  Come mai (voi) già siete tornati a Milano?

            ‘How come (you) already have come back to Milan?’

Clearly, these elements do not require I to C movement. Why is it so? Suppose that perché, come mai, may occupy the specifier position of INT in Italian. This is not implausible: the Spec of INT is presumably filled by a null operator in main and embedded yes/no questions, so it may be specialized for other operator-like elements which can be base generated there. Perhaps, INT selects in its Spec clausal operators, which are first merged there, in the sense of Chomsky (1998). If INT is intrinsically endowed with the feature Wh, no inversion is needed in cases like (17). Wh arguments and lower adverbials such as those in (16), contrary to higher sentence adverbials cannot be first merged in the Spec of  INT because of their selectional and interpretive requirements, demanding first merge in some lower, IP-internal position;  so , they can only meet the Wh Criterion by being moved to a suitable landing site in the left periphery (the specifier of the FOC position in main clauses, or the lower position involved in embedded clauses like  (14)c, whatever its label) and by triggering inversion.

Clear independent evidence that perché, come mai  fill a position distinct from and higher than the position of ordinary Wh elements in main clauses is that they can cooccur with focus:

(23)a Perché QUESTO avremmo dovuto  dirgli, non qualcos’altro?

         ‘Why THIS we should have said to him, not something else?’

       b  Come mai IL MIO LIBRO gli ha dato, non il tuo?

            ‘How come MY BOOK you gave to him, not yours?’

The order is fixed: the focussed element cannot precede perché and come mai, on a par with se (see (7)):

(24)a * QUESTO perché avremmo dovuto dirgli, non qualcos’altro?

             ‘THIS why we should have said to him, not something else?’

       b * IL MIO LIBRO come mai gli hai dato, non il tuo?

             ‘MY BOOK how come you gave to him, not yours?

Remember that ordinary  wh elements are incompatible with a focussed element in main questions (see (13)), a property that we have interpreted as showing that ordinary wh elements move to the specifier of FOC in main questions, thus competing with focussed elements. So, (23)-(24) and the contrast with (13) are directly accounted for if perché is first merged in the Spec of INT, while other Wh elements are moved from their first merge position to the Spec of FOC, lower than INT.

We have seen that in embedded clauses Wh elements do not have to move to the Spec of FOC, and they are consistent with a preceding focussed element (see the contrast (13)-(14)). No such main/embedded asymmetry is found concerning perché and similar elements: they are consistent with a following focus in both main and embedded clauses:

(25)a Mi domando perché QUESTO avremmo dovuto  dirgli, non qualcos’altro

         ‘I wonder why THIS we should have said to him, not something else’

       b  Non so come mai IL MIO LIBRO gli ha dato, non il tuo

           ‘I don’t know how come MY BOOK you gave to him, not yours’

These special distributional properties follow from the assumption that in both main and embedded questions special adverbial wh operators like perché can be base generated (first merged) in the Spec of INT in (10), a head endowed with the wh feature (hence no inversion is needed) and inherently selecting a clausal interrogative operator (hence unable to function as the landing site of movement, and uniquely  consistent with elements that are base-generated there).

As is expected, both topic positions higher and lower than INT can be filled, thus surrounding perché (the topics are expressed here by the Clitic Left Dislocation construction, as in previous cases):

(26)   Il mio libro, perché, a Gianni, non glielo avete ancora dato?

          ‘My book, why, to Gianni, you still haven’t given it to him?’

It is well-known that perché can also be construed with an embedded clause in cases like the following:

(27)          Perché ha detto che si dimetterà?

‘Why did he say that he will resign?’

This sentence is ambiguous: it may be asking the reason of his saying something, or of his resigning. Clearly, in case of  long distance construal, perché cannot be base-generated in the Spec of INT: movement from the embedded clause must be involved. We therefore predict that in case of long distance construal perché should behave as any other wh element: incompatible with FOC in main clauses, requiring I to C movement, etc. Judgments are not easy, but the prediction seems to be correct. For instance, if a focus is inserted, the sentence seems to cease to be ambiguous, only the local construal with the main clause remaining available:

(28)          Perché A GIANNI ha detto che si dimetterà (non a Piero)?

‘Why TO GIANNI he said that he will resign (not to Piero)?’

Along similar lines, Contreras (1989) pointed out that the Spanish equivalent does not require inversion when construed locally, but it does  when construed long distance. That locally construed pourquoi  (“why”) in French may allow and in fact require base generation in the left periphery is suggested by the fact that it doesn’t naturally license Stylistic Inversion (a construction which is parasitic on a genuine operator-variable configuration: see Kayne (1983)), nor can it be left in situ in an IP-internal position.