Comments

Monday, June 20, 2016

Classical case theory

So what’s classical (viz. GB) Case Theory (CCT) a theory of? Hint: it’s not primarily about overt morphological case, though given some ancillary assumptions, it can be (and has been) extended to cover standard instances of morphological case in some languages. Nonetheless, as originally proposed by Jean Roger Vergnaud (JRV), it has nothing whatsoever to do with overt case. Rather, it is a theory of (some of) the filters proposed in “Filters and Control” by Chomsky and Lasnik (F&C).

What do the F&C filters do? They track the distribution of overt nominal expressions. (Overt) D/NPs are licit in some configurations and not in others. For example, they shun the subject positions of non-finite clauses (modulo ECM), they don’t like being complement to Ns or As, nor complements to passivized verbs. JRV’s proposal, outlined in his famous letter to Chomsky and Lasnik, is that it is possible to simplify the F&C theory if we reanalyze the key filters as case effects; specifically if we assume that nominals need case and that certain heads assign case to nominals in their immediate vicinity. Note, that JRV understood the kind of case he was proposing to be quite abstract. It was certainly not something evident from the surface morphology of a language. How do I know? Because F&C filters, and hence JRV’s CCT, was used to explain the distribution of all nominals in English and French and these two languages display very little overt morphology on most nominals. Thus, if CCT was to supplant filters (which was the intent) then the case at issue had to be abstract. The upshot: CCT always trucked in abstract case.

So what about morphologically overt case? Well, CCT can accommodate it if we add the assumption that abstract case, which applies universally to all nominals in all Gs to regulate their distribution, is morphologically expressed in some Gs (a standard GG maneuver). Do this and abstract case can serve as the basis of a theory of overt morphological case. But, and this is critical, the assumption that the mapping from abstract to concrete case can be phonetically pretty transparent is not a central feature of the CCT.[1]

I rehearse this history because it strikes me that lots of discussion of case nowadays thinks that CCT is a theory of the distribution of morphological case marking on nominals. Thus, it is generally assumed that a key component of CCT assigns nominative case to nominals in finite subject positions and accusative to those in object slots etc. From early on, many observed that this simple morphological mapping paradigm is hardly universal. This has led many to conclude that CCT must be wrong. However, this only follows if this is what CCT was a theory of, which, I noted above, it was not.

Moreover, and this is quite interesting actually, so far as I can tell the new case theorists (the ones that reject the CCT) have little to say about the topic CCT or C&F’s filters tried to address. Thus, for example, Marantz’s theory of dependent case (aimed at explaining the morphology) is weak on the distribution of overt nominals. This suggests that CCT and the newer Morphological Case Theory (MCT) are in complimentary distribution: what the former takes as its subject matter and what the latter takes as its subject matter fail to overlap. Thus, at least in principle, there is room for both accounts; both a theory of abstract case (CCT) and a theory of morphological case (MCT). The best theory, of course, would be one in which both types of case are accommodated in a single theory (this is what the extension of the CCT to morphology hoped to achieve). However, were these two different, though partially related systems this would be an acceptable result for many purposes.[2]

Let’s return to the F&C filters and the CCT for a moment. What theoretically motivated them? We know what domain of data they concerned themselves with (the distribution of overt nominal).[3] But why have any filters at all?

F&C was part of the larger theoretical project of simplifying transformations. In fact, it was part of the move from construction based G rules to rules like move alpha (MA). Pre MA, transformations were morpheme sensitive and construction specific. We had rules like relative clause formation and passive and question formation. These rules applied to factored strings which met the rules’ structural conditions (SD). The rules applied to these strings to execute structural changes (SC). The rules applied cyclically, could be optional or obligatory and could be ordered wrt one another (see here for some toy illustrations). The theoretical simplification of the transformational component was the main theoretical research project from the mid 1970s to the early-mid 1980s. The simplification amounted to factoring out the construction specificity of earlier rules, thereby isolating the fundamental displacement (aka, movement) property. MA is the result. It is the classical movement transformations shorn of their specificity. In technical terms, MA is a transformation without specified SDs or SCs. It is a very very simple operation and was a big step towards the merge based conception of structure and movement that many adopt today.

How were filters and CCT part of this theoretical program? Simplifying transformations by eliminating SDs and SCs makes it impossible to treat transformations as obligatory. What would it mean to say that a rule like MA is obligatory? Obliged to do what exactly?  So adopting MA means having optional movement transformations. But optional movement of anything anywhere (which is what MA allows) means wildly overgenerating. To regulate this overgeneration without SDs and SCs requires something like filters. Those in F&C regulated the distribution of nominals in the context of a theory in which MA could freely move them around (or not!). Filters make sure that these vacate the wrong places and end up in the right ones. You don’t move for case strictly speaking. Rather the G allows free movement (it’s not for anything as there are no SDs that can enforce movement) but penalizes structures that have nominals in the wrong places. In effect, we move the power of SDs and SCs from the movement rules themselves and put them into the filters. F&C (and CCT which rationalized them) outline one type of filter, Rizzi’s criterial conditions provide another variety. Theoretically, the cost of simplifying the rules is adding the filters.[4]  

So, we moved from complex to simple rules at the price of Gs with filters of various sorts. Why was this a step forward? Two reasons.

First, MA lies behind Chomsky’s unification of Ross’s Islands via Subjacency Theory (ST) (and, IMO, is a crucial step in the development of trace theory and the ECP). Let me elaborate. Once we reduce movement to its essentials, as MA does, then it is natural to investigate the properties of movement as such, properties like island sensitivity (a.o.). Thus, ‘On Wh Movement’ (OWM) demonstrates that MA as such respects islands. Or, to put this another way, ST is not construction specific. It applies to all movement dependencies regardless of the specific features being related. Or, MA serves to define what a movement dependency is and ST regulates this operation regardless of the interpretive ends the operation serves, be it focus or topic, or questions, or relativization or clefts or… If MA is involved, islands are respected. Or, ST is a property of MA per se, not the specific constructions MA can be “part” of.[5]

Second, by factoring out MA form movement transformations and replacing SDs/SCs with filters focuses on the question of where these filters come from? Are they universal (part of FL/UG) or language specific? One of the nice features of CCT was that it had the feel of a (potential) FL/UG principle. CCT Case was abstract. The relations were local (government). Gs as diverse as those found in English, French, Icelandic and Chinese looked like they respected these principles (more or less). Moreover, were CCT right, then it did not look like easily learnable given that it was empirically motivated by negative data. So, simplifying the rules of G led to the discovery of plausible universal features of FL/UG. Or, more cautiously, it led to an interesting research program: looking for plausible universal filters on simple rules of derivation.[6]

What should we make of all of this today in a more minimalist setting? Well, so far as I can tell, the data that motivated the F&C filters and the CCT, as well as the theoretical motivation of simplifying G operations, is still with us. If this is so, then some residue of the CCT reflects properties of FL/UG. And this generates a minimalist question: Is CCT linguistically proprietary? Why Case features at all? How, if at all, is abstract case related to (abstract?) agreement? What is anything relates CCT and MTC? How is case discharged in a model without the government relation? How is case related to other G operations? Etc. You know the drill.[7] IMO, we have made some progress on some of these questions (e.g. treating case as a by product of Merge/Agree) and no progress on others (e.g. why there is case at all).[8] However, I believe research has been hindered, in part, by forgetting what CCT was a theory of and why it was such a big step forward.

Before ending, let me mention one more property of abstract case. In minimalist settings abstract case freezes movement. Or, more correctly, in some theories case marking a nominal makes it ineligible for further movement. This “principle” is a reinvention of the old GB observation that well formed chains have one case (marked on the head of the chain) and one theta role (marked on the foot). If this is on the right track (which it might not be) the relevant case here is abstract. So, for example, a quirky subject in a finite subject position in a language like Icelandic can no more raise than can a nominative marked subject. If we take the quirky case marked subject to be abstractly case marked in the same way as the nominative is, then this follows smoothly. Wrt abstract case (i.e. ignoring the morphology) both structures are the same. To repeat, so far as I know, this application of abstract case was not a feature of CCT.

To end: I am regularly told that CCT is dead, and maybe it is. But the arguments generally brought forward in obituary seem to me to be at right angles to what CCT intended to explain. What might be true is that extensions of CCT to include morphological case need re-thinking. But the original motivation seems intact and, frow what I can tell, something like CCT is the only theory around to account for these classical data.[9] And this is important. For if this is right, then minimalists need to do some hard thinking in order to integrate the CCT into a more friendly setting.


[1] Nor, as I recall, did people think that it was likely to be true. It was understood pretty early on that inherent/quirky case (I actually still don’t understand the difference, btw) does not transparently reflect the abstract case assigned. Indeed, the recognized difference between structural case and inherent case signaled early on that whatever abstract case was morphologically, it was not something easily read off the surface.
[2] Indeed, Distributed Morphology might be the form that such a hybrid theory might take.
[3] Actually, there was a debate about whether only overt nominal were relevant. Lasnik had a great argument suggesting that A’-traces also need case marking. Here is the relevant data point: * The man1 (who/that) it was believed t1 to be smart. Why is this relative clause unacceptable even if we don’t pronounce the complementizer? Answer: the A’-t needs case. This, to my knowledge, is the only data against the idea that case exclusively regulates the distribution of overt nominal expressions. Let me know if there are others out there.
[4] Well, if you care about overgeneration. If you don’t, then you can do without filters or CCT.
[5] Whether this is an inherent property of movement rather than, say, overt movement, was widely investigated in the 1980s. As you all know, Huang argued that ST is better viewed as an SS filter rather than part of the definition of MA.
[6] I should add, that IMO, this project was tremendously successful and paved the way for the Minimalist Program.
[7] The two most recent posts (here and here) discuss some of these issues.
[8] Curiously, the idea that case and agreement are effectively the same thing was not part of CCT. This proposal is a minimalist one. It’s theoretical motivation is twofold: first to try to reduce case and agreement to a common “mystery,” one being better than two. Second, because if case is a feature of nominals then probes are not the sole locus of uninterpretable features. Case is the quintessential uninterpretable feature. CCT understood it to be a property of nominals. This sits uncomfortably with a probe/goal theory in which all uninterpretable features are located in probes (e.g. phase heads). One way to get around this problem is to treat case as by-products of the “real” agreement operation initiated by the probe.
            From what I gather, the idea that case reduces to agreement is currently considered untenable. This does not bother me in the least given my general unhappiness with probe/goal theories. But this is a topic for another discussion.
[9] Reducing nominal distribution to syntactic selection is not a theory as the relevant features are almost always diacritical.

40 comments:

  1. Hi! I think the CCT is dead.

    I don't think its desiderata have all been accounted for by other means; but some of them have, which is something that some CCT defenders seem to miss. And as for the remaining desiderata, I don't think the CCT does any better on them than its competitors. In fact, it arguably does more poorly.

    Let me elaborate. I agree with you that empirically adequate theories of overt case don't successfully regulate nominal distribution all on their own. But some of the data that was taken (by Vergnaud, Chomsky, and others) as support for the CCT can be successfully reanalyzed without it. For example, who's to say that John is fond *(of) Mary is really any different from I like him/*he (in other words, yeah, fond selects a particular form for its complements, but so does more or less everything; that is not something about distribution, that's something about the forms that things must take when they appear in a given environment).

    It seems to me that the two real pieces de resistance (sorry if my French is not up to snuff) for the CCT are: (1) the distribution of overt subjects in infinitives, and (2) why there is movement in passive/raising.

    Let's start with (1). To paraphrase Marantz (talking about something else), I think that in this domain, Postal's findings on wager-class verbs kill the CCT dead. Postal shows that there is a particular group of verbs that behave like ECM verbs but for the provision that the embedded subject cannot stay in situ. But the important thing is that this embedded subject can vacate through A or A-bar means. So Which horse did you wager to win? is fine even for speakers who find *I wagered this horse to win bad.

    Now if you endow A-bar movement with the ability to alter abstract case, the CCT's workings collapse anyway (e.g. you lose the CCT-based account for things like *Who was it arrested?). But that means that in order to account for the behavior of the wager class, we must allow for a non-CCT-related restriction that states "this is a position that you must vacate." (We now have more evidence for this type of thing; see for example Claire Halpert's 2015 OUP monograph.) Combine this with the restriction that arguments cannot bear more than one thematic role (sorry, Norbert), and you get much of the raising-vs.-control stuff without reference to the CCT. (You still need a theory where movement isn't free, since otherwise overt subjects of control clauses could vacate their positions to just anywhere, but I think we need that anyway. See the recent flurry of anti-filtering stuff on this very blog.)

    The cherry on top of this skepticism cake is that there are overt subjects in control clauses if you look hard enough (see Szabolcsi on Hungarian). They're morphologically nominative (in Hungarian anyway), though I concede that this particular detail probably doesn't matter in evaluating the CCT.

    CONTINUED BELOW...

    ReplyDelete
    Replies
    1. I was hoping for a lively reply, and I confess that I had you in my sites. You never disappoint. Let me take these points seriatim.

      You proposal for nominals and adjective complements is that they are effectively case marked DPs. They are not "really" PPs. That they look like them and strand like them (this strikes me as particularly significant, btw)is, on your view, somewhat of a coincidence (if I understand you correctly). Now this is, of course possible. But let's just say that the duck principle (walks like, quacks like etc) suggests that these are indeed PPs and why exactly PPs should be the selected element of choice seems like a question worth asking. Indeed, in English, it is typically a P that saves the day ("*(for) Bill to leave would..." This makes perfect sense in CCT. It is a stipulation in a selection account.

      Wager class verbs are, so far as I know, a problem without a current principled solution. Interestingly, the current CCT (minimized version) can actually provide the start of a solution. As you note this looks like a species of ECM (I think Kayne was the first to suggest as much for French analogues). What then is the standard structure of an ECM? It looks like a TP complement and the subject raises to case marking position of the ECM predicate (hence the Lasnik-Saito-Postal scope altering facts). Say this is so. Is wager an ECM verb? It looks like you are suggesting not. But we can test this. If we passivize the 'wager' then it should not longer allow for an ok WH sentence: 'Who was it wagered to have won'. To my ear this sounds pretty lousy, roughly equivalent to 'who was it believed to have won.' If os, then this is indeed an ECM verb. What's the complement? If it is a CP (strong phase etc) then the only way to get to the higher case position is via A'movement through the edge of v. There is gets case marked. Now, as you note, this option must be carefully constrained for we don't want this to be generally allowed. But then we don't want ECM to be generally allowed. It is clearly an "exception" as the name implies. However, this is all compatible with CCT, and offer the path towards an explanation that the brute statement that sometimes we just need to move fails to provide. Recall, I agreed that diacritics could be deployed to save the day. But diacritics do not an explanation make. They are rather tools suggesting that no explanation is needed. I return to the seance part of you comment anon. I have stuff to do now.

      Delete
  2. (CONTINUED FROM ABOVE)

    That leaves (2): why do Patients move in passive/raising. We know it's not for argument-structural reasons (cf. John was believed to be a liar, The chair appeared to be invisible). So yes, the CCT can explain this. But given that we have a "you can't stay here" feature (see above), can't we just say that the output of passive(V) endows its complement with this feature, and raising verbs come out of the lexicon bearing it? (This is more or less what we used to say about abstract ACC, anyway.) This would come close to working, with the residue being why A-bar movement is not enough to fulfill this requirement (that is, how can a non-CCT theory account for the badness of things like *Who was it arrested?).

    And now here is the choice we are faced with: adopt the CCT just because of this tiny residue, or complicate our "has to vacate" feature so that it comes in two variants – one in which only A-movement will do, and one in which any kind of movement will do. You know what I'm gonna say next. The ratio of postulates to yield is much better for the latter option. (Note: the marginal cost is only in complicating the must-vacate feature thusly; to repeat, both theories need a general must-vacate feature.)

    PS: Even if you don't agree with all of this, we should really stop calling the primitives of the CCT "c/C/Kase." That's hugely misleading given that it's now clear that the two have nothing to do with each other.

    ReplyDelete
    Replies
    1. This second section advances the "diacritic" (ahem) theory of nominal distribution. It effectively says that there is nothing principled going on. You either have a feature stay or go or whatever an nothing else counts. I am happy to concede that this "works." After all, how could it not. So, to repeat, the CCT is the only THEORY around that addresses these distributional concerns. From what I can tell, you agree. You are arguing that there is no there there and that a theory is what we don't need. Well, that is a position and I agree with the disjunction: either the CCT or it's just specialized features that ad holy do what needs doing. We likely diverge on what side of the disjunct we fall on.

      Last point and a wrap-up. One thing worth looking at are the "saving strategies" as regards CCT. You don't find nominals in subjects of non-finite clauses but you can get them there is you have a 'for' complementizer. Is it an accident that in English the C is prepositional? Nope. Is it on the diacritic theory? Sure. Same point I made above with PP complements in nominals and adjectives.

      Conclusion: I agree that they have nothing to do with one another. But from where I sit it's the morphologically inclined that have muddied the waters. CCT has been part of GG for well over 30 years. The issues were pretty well understood and we all knew that abstract case and morphological case were not the same thing. The terminological problem was hatched when some GGers assumed the opposite and then started talking about how CCT was shown to be inadequate, when what was meant is that CCT did not smoothly cover morphological case realization. But that was not news, at least to the CCTers. It's bad enough to have real empirical problems that bear on one's views, it's a gigantic pain to be saddled with problems that have nothing to do with one's views. So, sure let's distinguish them. I give you Kase and you give me Case or vice versa. I am even happy with JRV theory. But I don't agree that the confusion started with CCT. We were never confused. Which suggests...

      Delete
    2. A few points:

      1. I don't think your solution for wager predicates works. If these verbs select a CP, then how would This horse was wagered to win work? You would have to say something like the subcategorization changes from CP to TP under passivization – sounds like a diacritic if I've ever seen one :-)

      2. I'm not sure I agree with your negative assessment of must-vacate features. First, if you believe that there is "criterial freezing" (or whatever derives the ungrammaticality of *To whom do you wonder t Mary spoke t?), then this is just the other side of the same coin (the two sides being must-stay and must-vacate). Second, in my view the real assessment of this device versus the CCT depends on what independent support they each can marshall. When the CCT said something about the forms of nominals, that was independent support. Absent that, it is, from where I sit, every bit the "just so" story you accuse the must-vacate approach of being. Conversely, if we find that we need must-vacate features in contexts where "abstract case" could not possibly be the culprit, it would lend independent support for the need for such features. Moreover, from where I sit, the must-vacate theory looks simpler.

      3. About the unity of repair strategies: I agree with you that this is prima facie a good way to proceed. But I think that in this context, it turns out to be a red herring. There are plenty of languages where the complements of N and A cannot be analyzed as PPs. I think it is an independent point of variation whether a given, shall we say, "proto-case" (e.g. instrumental) will be realized in a given language as nominal case or as a PP. I think this was part of the central motivation for the KP analysis (which makes the structural similarity between the two explicit). And there might be interesting things to be said – e.g. instrumental will only be a case if dative is also a case, etc. But this is an independent point of variation among languages. To put this another way: it is a coincidence (of English and some other languages) that the complement of N and A, and the complementizer ('for') that licenses overt infinitival subjects, are of the same lexical category. And by the way, regarding prepositional complementizers, Tom McFadden has work showing that its distribution and "licensing" abilities are not about case either, but about selection. I suppose you might call this a diacritic theory, but like I said in (2) – from where I sit, saying that 'for' can "license a subject" and having no support for that except for the fact that it can license said subject, is just as much of a diacritic.

      Delete
  3. Just a brief point regarding 3. I don't think that the unity of repair strategies is a typological prediction of the CCT. As far as the CCT is concerned, it's entirely possible that some languages might use one repair strategy for overt subjects of infinitivals and another for DP complements of N. (And given that the CCT permits some variation in which heads can assign case, not all languages are necessarily going to use a repair strategy at all in either instance.) The question, I think, is whether we really want to say that the apparent parallel between 'for' insertion and 'of' insertion in English is just a coincidence. The CCT suggests that it is not, and yet leaves quite a bit of wiggle room for languages that don't pattern like English.

    ReplyDelete
    Replies
    1. @Alex: that's great, as far as I'm concerned. If the CCT makes no strong claims about the unity of repair mechanisms, then that's one less thing it does. So to summarize, we have a theory that makes no claims about the actual forms of nominals; no cross-linguistically robust claims about which base-positions will be in need of repair strategies and which won't (diacritics!); and no claims about the nature of the repair strategies.

      Haven't we arrived at my "must-vacate" features? Which, I'll remind you, are independently necessary to account for wager-class verbs (unless Pesetsky and Muller are right about derivational structure-removal/"exfoliation").

      Notice also that the CCT never quite explained why you can't use (the semantically bleached version of) for to "rescue" the complement of N, or use of to rescue the overt subjects of infinitives. The answer, it seems to me, was always "selection." And what McFadden shows is that once you avail yourself of selection in this fashion, you don't need the CCT at all to account for things like the distribution of prepositional complementizers.

      This is obviously a matter of scientific taste, but taking inventory of all this, I would say that this is when we should pronounce a theory dead.

      Delete
    2. There were a couple of alternatives to selection in the case of ‘of’ (a literal rule of ‘of’ insertion, and the genitive Case story in Knowledge of Language). If ‘of’ simply isn’t a complementizer, and (unlike ‘for’) is either inserted by a special rule or is a marker of genitive Case, then the distribution can be captured. So we have a principled account of why some kind of preposition-like thing gets inserted, a clunky set of rules/principles that do the dirty work of actually inserting it, and a boring story about why this preposition-like thing is sometimes pronounced ‘of’ and sometimes pronounced ‘for’. That seems more or less right. After all, English†, where ‘for’ serves both purposes, is presumably a possible language.

      I’m not really moved by the ‘wager’ verbs as they seem to be a problem for everyone. Your objection to Norbert’s story depends (if I understand correctly) on either (i) the assumption that A-movement out of an non-finite CP is impossible or (ii) the assumption that Spec,CP is not a possible intermediate landing site for successive-cyclic A-movement. I’m not sure how confident we can be that either of those principles holds.

      Overall, I don’t particularly disagree with your assessment of CCT. But a retreat to the least interesting story possible is a hard sell. If the best we can do after ~40 years of Case theory is point out that different things take different kinds of complement, and that some things have to move while other things don’t, shouldn’t we all just go home?

      Delete
    3. 1. I don't understand your first paragraph. If you view 'of' as related at all to genitive case, it seems you are conceding the point that Norbert objected to – namely, the fact that the complement of N in English is a PP is orthogonal to the issue at hand.

      2. My objection to Norbert's wager story depends on neither of the premises you stated. If nominals can A-move out of embedded CPs (be it through SpecCP or not), and that's how you derive This horse was wagered to win, then in the active version of the same sentence, the nominal should be able to move just far enough to be in the domain of ACC (recall, Which horse did you wager to win? is grammatical), and stay there. For concreteness, let's assume there=SpecCP. Why can't it stay there? Back to must-vacate.

      3. It's not a hard sell because I'm not trying to "sell" anything. My point is that the CCT, once properly confronted with its shortcoming, is itself the least-interesting-story of which you speak. It's a pile of diacritics that (a) doesn't quite do the job it is supposed to do; and (b) isn't supported by any independent evidence. My gesturing towards selection and must-vacate features is actually an attempt to sketch a slightly-less-uninteresting story. You can reject this alternative, but I would say that it is you who is then retreating to the even-less-interesting alternative, namely CCT.

      4. Re:giving up and going home – nah. First of all, I think there is much more to linguistics than figuring out the rules of nominal licensing. And even if there wasn't, and nominal licensing was the be all and end all, just because a problem is hard doesn't mean you go home. Maybe what we need is 40 more years! (I know you actually agree with this last part and were being facetious, but I thought I'd say this anyway...)

      Delete
    4. Sorry, been traveling. Some replies:
      First, the wager stuff requires something slightly more exotic than a vacate position feature given the * status of "who was it wagered to have won." This means that vacating is insufficient if the higher verb has been passivized. Why? Well if wager is a type of ECM verb we know why. For the same reason that "who was it believed won" is *. So, the vacate principle and needs to worry about the higher v/V, in other words, it seems that there is relationships between a nominal and a "case" assigner.

      As noted, this seems like a good first step. You noted cases like "John was wagered to have won" as problems. I ignored this as I always thought that these were not that good (at least as compared with "John was believed/expected to have won." They really degrade fast when the by-phrase is included: John was wagered by Bill to have kissed Mary is considerably worse than the analogue with 'believe' and 'expect.' So, I am not sure of the data (or more accurately analogous sentences seem to me quite poor (even small changes to the embedded verb stink up the example: John was wagered to kiss Mary is lousy).

      But say these are ok: then we have a case where v of ECM 'wager' has an EPP feature and forces movement in these cases. Interestingly, given current views both A and A' movement transit through this position (even in passives) so we can accommodate the data and explain why passivizing the wager and wh moving yields a bad result. Is this the right theory? I doubt it. But, the wager data is pretty unaccommodating and the story you offer is pure diacritic, so, not surprisingly it can be made to fit any data (though the simple one offered does not appear to fit it all).

      Re PPs in N and A. The observation is indeed language internal as Alex noted. In English N and A don't assign case. P does. So, interestingly when one gets complements to either they are wrapped in PPs. Why PPs? Because they do case mark in English and CCT requires that they be marked. Perfect account? Hardly. Partial, yes?

      So the main "shortcoming" is, in your mind is 'wager' verbs. I am not convinced that this is so. It looks like an ECM verb, albeit a quirky one given that A'-movement can feed it. This is not that surprising given what we have seen in French and the fact that Spec v is a landing site for both A and A' movement in current theory. The problem is why 'wager' verbs are not more ubiquitous. But then why is ECM not more ubiquitous? Dunno. I suspect that there is much being left out: for example, how would anyone learn wager class verbs? Surely they cannot that common! Here's a hunch: the good cases are parasitic on another structure: I wagered on Bill [PRO to win] and for some reason 'on' can delete marginally. But that's another story and it is just a hunch (BTW: who did you wager on to kiss Mary is fine. Drop the 'on' and it sucks).

      So how good is the CCT? Well if the only problem is wager, then things look pretty good to me.

      Delete
    5. This comment has been removed by the author.

      Delete
    6. @Norbert:

      1. We continue to speak across purposes re:wager verbs. The problem is not why the embedded subject can "transit through" Spec,v, but why it can't stay there (which, given Johnson 1991's observations on very-short-verb-movement in English, would yield the ill-formed John wagered the horse to win). I don't see a solution that's not in term of must-vacate features.

      2. Given (1), it is worthwhile to see just how much of CCT can be recapitulated in terms of must-vacate features. It seems to me that the answer is "almost all." I think we use the term 'diacritic' ratio of postulates to payoff is too high (correct me if I'm wrong); well, it seems to me that using the CCT to derive the residue of this "almost all" is the epitome of a high postulates-to-payoff ratio.

      Delete
    7. 1. Strong feature on v of ECM wager. This forces movement through its spec. DP must then vacate. Interestingly, if passive then cannot check/assign case which accounts for why vacating is NOT sufficient even if necessary. Not sure how simple vacate feature on lower Spec T gets the * on 'who was it wagered to win.'

      2. How much can one get? Not enough. But as I said, it looks like for you CCT fails primarily because of 'wager.' I am less impressed by the data and phenomena than you are. But to put things as you do: modulo 'wager' CCT works very nicely. The argument then boils down to the correct analysis of these and how good the 'wager' data is as an argument against CCT. IMO, not very. The data sucks and the proposal you gave given the data is inadequate. Hence...Viva CCT!!

      Delete
    8. Oh, btw: I am on vacation right now and some of my traveling partners think saving the CCT is a lousy way to spend my time. So, I will be signing off until sometime next week. I thus leave the rhetorical field to you, at least for a while.

      Delete
    9. 1. I still don't understand. If there is a strong feature on the matrix v, why can't the nominal stay there? That would give you John wagered the horse to win – or if Johnson 1991 is wrong, John the horse wagered to win – neither of which is grammatical. I still don't see how you rule this out without saying that Spec,v must be vacated, for non-c/C/Kase-theoretic reasons.

      1'. Re:Who was it wagered to win? – as I said earlier in this thread, I am quite convinced we need must-vacate features anyway. (The support for this is by now not restricted to wager verbs; see Claire Halpert's monograph.) So given this, the complication needed to get Who was it wagered to win? is to have general must-vacate features versus ones that can only be satisfied by A-movement. Ugly? Yes, I concede. But still a lot less ugly than the whole clockwork mechanism of abstract case sans the support of adequate morphological predictions, all in order to account for this tiny residue. (Remember: you need must-vacate features even if the CCT is true to the letter.)

      2. See above (must-vacate features are necessary beyond just in the explanation of wager verbs). And even if all of CCT is true, you still need to appeal to selection to explain why in English you can't use 'of' and 'for' interchangeably (see my conversation with Alex), as well as to solve many other similar issues. Combine must-vacate features with (independently-necessary) selection features and the empirical burden borne directly by the CCT is tiny (if it is extant at all). And the theory has no independent support beyond this, since it has been severed from its predictions about surface forms. That's a dead theory.

      Delete
    10. Sorry for the delay, but life is life. Ok Wager verbs. There are two problems: first is whether vacate features are enough and second whether just CCT is enough. You point out that being in a case position is not sufficient. I agree. The DP must move from its case position further up the tree. Why? Dunno. But then again neither does anyone else which is why 'wager' verbs are a problem. Your point seems to be that whatever solves this problem suffices to account for 'wager' nominal distributions. I don't see this. Let me explain.

      Recall that sentences like "*who was it wagered to win". These are bad yet they vacate whatever position the good ones vacate. So why *? Answer they have no abstract case. Why not? Because 'wager' is passivized and these have no ECM case to assign. So, even given a vacate feature something like CCT seems required. But, this means that 'wager' class verbs are NOT counterexamples to CC, which was my point.

      Truth be told, however, I have a larger point. And here we seem to agree. It was that MCT is not a real rival to CCT as they deal with non-intersecting phenomena. Your argument is not with this claim, but with the idea that JRV was right to try and replace unprincipled filters with a principled CCT. Your claim is that the principles do not work so we should return to F&C style diacritics. That's fine with me actually. I find the move unfortunate theoretically, but maybe the empirics demand it. This, of course, leaves many of the hard theoretical questions harder to address e.g. why these filters and not others, why filters at all, are there structural restrictions on filters etc. But, maybe this is what we need.

      Last point: I will now come even cleaner. I like CCT. Why? Because of Chomsky 1993. The appear showed how to reduce CCT to a species of movement. I like that idea. Lasnik and Saito (following Postal) and Brannigan and others showed that there is interesting empirical payoff wrt scope if we make this assumption. This was a NEW kind of fact and a very interesting one. One of the nicer consequences of adopting MP. So I would be sad to see it go away. This is my main interest in CCT. But given this interest I would hope that the empirical reasons for giving it up would be substantial. It really did simplify Gs and still does. It makes MP sense. Not bad for a theory and good reason to give it up very very slowly.

      Delete
    11. Oh, one more point: explaining why 'wager' verbs force movement to a higher position than the case one is a problem if we way that sentences like 'John was wagered to kiss Mary' are fine. If they are marginal (which is where I was going in an earlier comment) then we can account for this assuming that nominals get to the relevant higher case position via A' movement plus the assumption (required anyhow) that A' movement in English terminates in CP. So, how good are these passives? I find them odd. They get really bad when we add the by-phrase: John was wagered by Tom to have kissed Mary (compare: John was expected/believed by Tom to have kissed Mary). So, if these indicate that "john was wagered to kiss Mary" is * then why nominals don't remain the case position might be explicable. Maybe.

      Delete
  4. (Reply to Omer.)

    1. I don’t quite understand what you’re saying here either. It’s true that the Knowledge of Language theory loses the nice parallel between ‘of’ and ‘for’ (or at least risks doing so, depending on how exactly you finesse the notion of a case marker), but I only mentioned it as one alternative to a selection story. By the way, using selection to force the use of ‘of’ is not quite as bad as you make out. Yes, we could just say that the relevant nouns select ‘of’ PPs and leave it at that. However, CCT offers an explanation of why there are no nouns that select bare DPs. That is lost on a pure selection account.

    2. Yes, some kind of stipulation needs to be made to stop that happening. (No-one was claiming to have a nice theory of ‘wager’ verbs, as far as I can see.) The question was whether this stipulation must necessarily be contrary to CCT, or undermine it by introducing mechanisms that would make the principles of CCT largely redundant. I don’t see any reason to think this is the case. All that’s required is some stipulation that stops a DP moving from a theta position to Spec of non-finite CP, getting assigned Case by a higher head, and then staying put. This stipulation may (perhaps) have to be of the form “X must vacate Y”. However, if one such stipulation needs to be added to the theory, that in no way indicates that the principles of CCT ought to be thrown out and replaced by some more general theory of vacation.

    3. I guess I’m having difficulty seeing how the select plus must-vacate story is more than just a pile of diacritics.

    4. I wasn’t being entirely facetious. I think if it’s pretty bad news if we have to throw out CCT without any inkling of a better alternative.

    ReplyDelete
  5. This is for Dominique Sportiche. He is having some trouble commenting and so here is an assist:

    Assume we agree that there are positions in which (overt, or overt+wh-traces) DPs can occur and positions in which they can't.
    Two theories: characterize the set of positions S in which they can occur (fundamentally CCT) or its complement Cs (positions that must be vacated).
    The Filters and Control theory was a vacate theory.
    CCT was Vergnaud’s alternative. Its claim was that S is more naturally characterized than Cs: you have to be "governed" by the right kind K of element. And there is a coherent way to characterize K.
    So for example, of and for are of the right kind: P, while Ns are not - full disclosure: I do not quite quite believe this but it is coherent). The same would go for passive participles, they are Adjective-like (and incidentally do tolerate objects under certain circumstances – see Belleti on partitive case) or wager (there is substantial discussion of this in French – see Kayne’s “On certain differences…” or quite informatively in connection with participle agreement), all straightforwardly consistent with CCT, I think.

    So, I do not see that CCT is effectively challenged by anything yet.
    Rumors of CCT’s demise still strike me as exaggerated.

    This is not to say that CCT is right. It still is a set of stipulations (that seems “natural" given observables) but falls short of an explanation.

    ReplyDelete
  6. Well, it's starting to feel like the discussion is going around in circles. This is not a direct response to Dominique's post, but it is prompted by it. I don't understand what "observables" he is talking about – we have ruled out morphology as a relevant observable (the CCT doesn't work for that), so the only remaining "observable" that the CCT might work for is licensing. Saying that a head H can licensing a DP in position x with no independent evidence except that DPs can occur in position x is completely circular; but things are worse, since – as repeated ad nauseam in this thread – it doesn't even do a really good job at describing the possible positions of DPs. (As shown by wager verbs in English, but also a zillion other constructions in myriad other languages; see below for some examples.)

    Even if one has some magic potion to explain away wager verbs – and, despite promissory notes strewn throughout this thread, I have yet to see a working theory (except for "exfoliation") – you have to explain Zulu (Halpert's work), and Basque (my work, among many others), and ... In short, the valiant attempt to positively describe the set of licensing positions has failed. There will always be a residue of negatively defined positions (in Zulu if not in English). So you may as well try to make a uniform theory based on vacate rules, because the converse is clearly a dead end.

    In the spirit of some earlier comments in this thread, I think this will be my last posting on this topic. From my vantage point, it seems that the other camp is clinging to a shrinking corner that the CCT accounts for ("the distribution of some but not all DPs in English and a couple of other languages"). And to me, the price of this theory given its meager payoff borders on the absurd. I am quite confident that if the CCT hadn't proposed in the 70s-80s, and I came to you now proposing a theory that (i) has zero observables except licensing; (ii) is completely divorcible from the surface form of nominals, and (iii) doesn't even completely work for the very purpose it was put forth for (i.e., accounting for licensing) — I would be laughed out of the room.

    CCT defenders, I leave the floor to you.

    ReplyDelete
  7. I appreciate Norbert raising this important issue, and from a perspective that I standardly teach. I am not sure whether or not CCT is tenable, and I’m not sure whether we expect it to be given Darwin’s problem. A few thoughts to broaden the discussion a bit, though.

    The focus on wager-class verbs seems a bit odd. I’d thought that ever since we understood French object agreement we were comfortable with positions that an element can move through but can’t surface in.

    The fact that Case is non-identical to case has indeed been crucial from the beginning, and this does not deserve the scorn Omer expressed. I would expect here — as elsewhere — for the syntax-morphology mapping to be principled, even faithful, unless overridden by morphophonological properties of the language. This means it’s not simple, but not entirely arbitrary either = the most interesting situation!

    The version of the Case theory that only overt nominals were relevant also struck me as the most attractive version. The fact that OP in relative clauses need Case (Lasnik & Freidin 1981) was taken as disconfirming. This seems less obvious now, given theories of relative clauses that make no use of OP, e.g. involving ellipsis of the nominal inside the relative clause under identity (e.g. Sauerland 2000, Safir 1999). (I discussed this, and indeed many of the issues related to this post, in a 2011 talk in honour of Jean-Roger Vergnaud; unfortunately, it never became a paper rather than just slides.) On that approach, these data would indicate an ordering of the Case requirement before ellipsis.

    Another argument against this version of the Case theory may come from pro.

    No other version of CCT properly explains the distribution of PRO, and even this one needs the addition that PRO must be a grammatical subject. Actually, the much-maligned “null Case” fares better (particularly with the addition of Case being inherited from C, so we don’t have to worry about why only a subset of nonfinite Ts assign it); to adopt it, we’d need to reconcile ourselves to the fact that PRO often triggers standard case concord morphology. Maybe we should (Case vs case), but I understand null Case just feels like a hack to many people.

    ReplyDelete
    Replies
    1. As you might imagine, I find the PRO problem trifling. No PRO no case licensing of PRO. I also agree with your characterization that null case is a hack. Isn't it? A bespoke feature for a single item? If that is not stipulation what is? On this I agree with Landau and his contempt for dis"honest stipulations."

      Delete
    2. Yes, I know your take on PRO, Norbert. :)
      Anyone else want to engage on this issue from a non-movement perspective?

      Delete
  8. (continued)

    Those who argue against CCT tend to cite a parallel between the distribution of “for” and “that”.

    (i) optional in the complement of verbs
    1. I think (that) she read the paper.
    2. I want (for) her to read the paper.

    (ii) obligatory as grammatical subject
    3. *(That) she read the paper surprised me.
    4. *(For) her to read the paper surprised me.

    (iii) impossible in complementizer-trace contexts
    5. Who do you think (*that) read the paper?
    6. Who do you want (*for) to read the paper?

    The parallelism is only apparent, though.

    I always found most plausible the analysis of “optional” “for” as rather involving ECM with a null HAVE. (Kayne, den Dikken et al 1996)

    7. I want for her to read the paper.
    8. I want HAVE her to read the paper.

    For those who show the following pattern of grammaticality that I’ve seen reported, that means that HAVE under “prefer” is behaving like a wager-class verb

    9. I prefer *(for) her to win.
    10. Who do you prefer (*for) to win.

    Notice that supposed “for” deletion is not possible in the complement of an adjective, but “that” deletion is, breaking the supposed parallel:

    11. I’m happy (that) she won.
    12. I’m happy *(for) her to win.

    13. It’s likely (that) she won.
    14. It’s likely *(for) her to win.

    Notice that as we might expect on the HAVE analysis (if somebody is needed to HAVE something), (14) doesn’t allow “for” deletion under wh-movement (I can’t tell if (12) does.)

    15. *Who is it likely to win?

    It’s been suggested for (14) that the obligatory presence of “for” under adjectives is about the expletive. On that approach, we need to posit a direct relationship between the expletive and the clause, (rather unfortunately to my mind), and say that only clauses that can be subjects can associate with “it”.

    16. *It’s likely her to win.
    17. *Her to win is likely.

    Again, this claim breaks down with finite clauses:

    18. It’s likely she won.
    19. *She won is likely.

    This direct relationship would also need to have the odd property that “it” can’t associate with DPs, (which of course can be subjects):

    21. There/*It appeared a shadowy figure in the doorway.
    22. There/*It were several people arrested over the weekend.

    So the non-CCT approach doesn’t have the parallelism they rely on to make us turn away from Case as an explanation of the distribution of “for”, and doesn’t have an explanation for the obligatoriness of “for” under adjectives.

    Again, I’m not sure that CCT can be maintained, but the issues are more complex than they’re often made out to be.

    Norbert — you asked about quirky vs inherent — these terms get used inconsistently, but “quirky” indicates that the DP moves and behaves as though it needs licensing like any other DP. Some inherent cases are argued to be non-quirky — Pesetsky & Torrego have egs in their case overview chapter, Martha McGinnis has egs of what she calls “inert” case.

    ReplyDelete
  9. PS Examples like my (21), (22) above (adapted from a paper by Amy Rose Deal btw) remind us that a "vacate" feature for the theme of passives /raising is incorrect. So would:

    23. *Who did it appear in the doorway.
    24. *Who was it arrested over the weekend.

    So non-CCT approaches need a different account of this.

    Norbert's:

    25. *Who was it wagered to win.

    is similar, but since wager-class verbs actually do have a "vacate" feature, (25) further shows that "vacate" features co-exist with licensing, rather than being a replacement for licensing.

    ReplyDelete
    Replies
    1. @Norbert, @JAL: I already addressed this in my comments above. If vacate features are independently necessary, then introducing a second variant of "vacate!" which can only be satisfied by A-movement (and not A-bar) will handle all of this data, and is simpler than the entire clockwork of CCT.

      And if I already broke my promise and returned to this thread: people (like me) who claim that the CCT is completely divorced from case morphology are not making that claim simply because the syntax-morphology mapping is imperfect. We all agree with that. The point is that if you look at what you need to say to accurately capture case morphology – and you pick the simplest version of that (which, sorry @JAL, is still Marantz 1991 and its developments, e.g. Baker's work) – that version makes no reference to CCT primitives at all. So the fact that CCT is accidentally occasionally right about surface forms is not an argument for anything, since the right theory of case morphology is CCT-free.

      Delete
  10. Omer -- can you be clearer? A-movement "vacate" doesn't seem relevant to (21), (22), nor to the contrast between (25) and (26) (repeated from your comment above).

    (26) Which horse did you wager to win?

    ReplyDelete
    Replies
    1. I have to go now, so I'll be quick: (23) and (24) fail because "appear" (straight out of the lexicon) and "was arrested" (after passivization) endow their complement position with the A-mvmt variant of "vacate!". So A-bar movement in (23-24) doesn't fix that. (And there is no "free" movement - you can only move if there is a viable position that attracts you.)

      Ditto for (25).

      (26) is okay because there is an A-movement step prior to A-bar movement.

      Delete
    2. And as for (21-22), I'd adopt the (quite popular, I thought) theory that "there", but not "it", can start out with its so-called associate in a Big DP structure. All you have to add then is that subextraction of "there" is enough to satisfy vacate!.

      Delete
    3. Sorry, trying to move too fast for my own good – scratch what I said about (25) vs. (26), it (obviously) doesn't work. What I should have said is: (26) is a non-passivized "wager" verb, hence endows its complement domain with the generic vacate! feature. But in (25), the verb has undergone passivization. So, in keeping with what I wrote for (23-24), it would endow its complement domain with the A-movement only variant of vacate!. That rules out (25) but rules in (26). Sorry for the confusion.

      Delete
  11. Hmm, Amy Rose Deal's paper on selection and "there" I thought was nice evidence against the Big DP analysis.
    On this view, what's supposed to be regulating the appearance of "it"? (e.g. my discussion above surrounding (13)-(19))

    ReplyDelete
    Replies
    1. @JAL: fair point regarding Deal's findings.

      Perhaps, then, it's all about selection...? But let's suppose, just for the sake of argument, that a selection-based account of the distribution of 'it' vs. 'there' fails. Is the distribution of overt expletives, then, the piece de resistance for CCT?

      btw, Norbert seems content to ignore 'wager' verbs (if it turns out that he must) in order to salvage the CCT – though he may or may not be alone on that. I'd like to point out, in this respect, that overt expletives are crosslinguistically a far, far more marginal phenomenon than positions an XP can transit through but not stay in. If one of them is to be ignored in order to build the "right" theory, it's probably overt expletives. (Though, obviously, it would be better if we could handle both.)

      Delete
  12. I don't think you do have to ignore "wager" to save the CCT, because I agree on the existence of "vacate" features. If the facts are this:

    which horse did you wager to win?
    This horse was wagered to win.
    *I wagered this horse to win.
    Which horse was wagered to win?
    *Which horse was it wagered to win?

    Then CCT + (plain) vacate get the facts. So far we haven't seen, though, vacate allowing us to eliminate CCT, rather than being orthogonal to CCT.

    I want to reiterate that I'm not convinced CCT is right, but to me the most thought-provoking problems involve PRO, not wager-class verbs.

    ReplyDelete
    Replies
    1. Good presentation of the debate. I would resist assuming that wager passivizes that well, as i noted above. I think the data are less claer than Omer suggests. But JAL succinctly makes the point I was trying to make as well. The data as provided do not seem to indicate CCT is irrelevant. Only that it is not sufficient.

      Delete
    2. So, in the spirit of @Norbert's last comment, here's where I think we are

      1. CCT + vacate!-features seems to have good empirical coverage regarding the distribution of nominals in English. (Though it's not hard, at all, to find languages where the CCT's predictions break down.)

      2. An alternative is to jettison CCT and adopt vacate! features that come in two flavors (instead of one – which, to repeat, everybody needs): a general-purpose flavor and an A-movement only flavor. This seems to cover the data quite well, too, with the possible exception of the distribution of different kinds of overt expletives ('it', 'there') – though I'm not convinced handling this residue is hopeless.

      Now, where I think some of us differ is here:

      3. I think the CCT says nothing useful about surface forms. I think @JAL at least, and @Norbert possibly, disagree with this statement. One can formulate a complete theory of morphological case that makes no real reference to CCT's primitives, and is arguably simpler than what you need to say to "fix" CCT with respect to its morphological shortcomings.

      4. I think (2) is so much simpler (compared to (1)) that (1) bears an extremely heavy burden of proof, one that is hard to imagine it meeting, given (3).

      Like I said, it seems that (3) and (4) are the contentious parts here.

      Delete
  13. I think 2 is overstating the case. The distribution of “it/there” and “for/zero” are tools for investigating the licensing of nominals in English, and it’s exactly the type of data that they were worried about in 1977. Nothing said in a non-CCT vein here touches on that data.
    Regarding `A-movement only vacate’, while this is helpful for *Which horse was it wagered to win”, it doesn’t help with any of the other data. It also makes makes this particular fact unrelated to the general ungrammaticality of passive themes low with an “it” expletive, which seems rather unfortunate.

    ReplyDelete
  14. I'm a bit late to this party but I just want to mention that Milan Rezac's recent (2013) LI paper on DOC+ECM contains what I consider the most convincing and well-worked-out argument for something like CCT which has been proposed. Indeed I’d never been convinced at all that CCT was worthwhile until I read it. The argument therein comes from the ‘wager’-like difference between a,b and c-d, where ECM isn’t possible with a DOC frame for reasons not relating to selection.

    a. *We showed the reader these propositions to be false.
    b. *The reader was shown these propositions to be false.
    c. the propositions which we showed the reader to be false
    d. the propositions which the readers were shown to be false

    The suspicion that case is crucial is encouraged by contrasts between DOCs and related prepositional dative structures. Milan walks through the details extremely carefully and shows that the effect really can’t be dismissed as relating to case adjacency or other extraneous factors, and he develops an analysis where the good ones involve case assignment under A’-mvt (I am currently thinking of ways to support that component of the analysis). It’s also relevant that the problem element is the ECM subject, and that’s not selected by the matrix predicate, so it’s not easy to assimilate this to a ‘vacate!’ type analysis like is suggested above for ‘wager’. I am no expert on CCT or its brethren, but I feel like I know a good argument when I see it, so I would be v keen to hear if any anti-CCTers like Omer have a response in mind.

    Two other related comments:

    1. in an appendix Milan puts forth an analysis of 'wager'-class verbs and their quirks, one that doesn’t gut CCT of its content as far as I can tell. Incidentally it makes crucial reference to the oft-ignored fact that the object of ‘wager'-type predicates can often stay in situ if it’s a simple light pronoun, e.g. “We alleged {them/*the propositions} to be inconsistent”. This is surely a problem for a ‘vacate’ type analysis (right?).

    2. a lot of the apparent problem cases for one theory or another involve sticking quasi-expletive 'it' into sentences without any care for whether it's got an associated CP. just like arguments FOR CCT can sometimes lack the necessary controls to make their case (see e.g. McFadden on for-infinitives), the same can probably be said about the arguments against CCT as well, say from examples like "who was it arrested”. Generally I’m not convinced that ‘it’ has an expletive-like life when it’s not associating with a CP (see also Robyn Orfitelli’s diss on the argumental nature of 'weather it’).

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. I'm familiar with this Rezac paper, and was rather unmoved by it as far as the need for CCT is concerned. Here's why:

      Let's start by conceding that the data is a problem for "vacate!" type theories. (I'm not actually convinced that it is; but my argument here doesn't depend on there being a successful "vacate!"-style analysis of these data.) Milan's analysis involves Agree for phi features piggy-backing on Agree for A-bar features (e.g. [wh]) in configurations where the former would not be able to apply alone. This suffers two problems: (i) it would be completely ad hoc if it worked; (ii) it doesn't work.

      Let's start with (i): the proposed solution looks like little more than a redescription of the problem. Whence this piggybacking? Under what pairings of features <x,y> can Agree for x piggyback on Agree for y? We are not told. Absent a theory of this, CCT explodes anyway. And the kicker is that you can't even stipulate "okay, only when x=phi and y=wh"; even that explodes:

      (3) * John asked who Bill tried t to win.

      Both 'ask' and 'try' are accusative assigners (unlike, e.g., 'wonder'). Why can't accusative be assigned to 'who' in (3), piggybacking on Agree for [wh] features? And one can spend an evening churning out examples with slightly different structures than (3) that exemplify the same problem.

      On to (ii): the idea that case is assigned by Agree for phi features doesn't work anyway (see here). Nor does the idea that capital-C "Case" assignment is related at all to morphological form (see my discussions with @JAL, above), also a crucial piece in Rezac's argumentation. (Rezac tries to salvage this view, trying to explain away some of the central counterevidence from Icelandic in his fn. 17; but the Basque data I discuss in the linked-to post is not amenable these maneuvers, anyway.)

      So basically, Milan has identified what (to me) looks like data that is problematic for both views. Great, that's very valuable – but an argument for CCT this is not.

      Re:your (1), not really. His solution to this is based on carving out an exception: a kind of very short movement available only to (weak) pronouns. (It has to be based on such an exception, for CCT cannot afford to treat pronouns and other nominals differently: *Mary tried him to win is of course bad even with a weak pronoun.) Well, you see where this is going: this very short movement, available only to weak pronouns, will satisfy the vacating requirement.

      Re:your (2), this makes me quite happy. The upshot of my discussion with @JAL above was that there aren't really many problems that you need the CCT for that don't involve expletives; if "expletive it" is really not expletive per se, but originates on associate CPs, then it is by-definition outside the purview of CCT (which does not presume to regulate the distribution of CPs).

      Delete
    3. Ok thanks Omer, I figured the response would centre on a critique of the case assignment under A'-mvt analysis. I've not really got a dog in this fight, but I am inclined to give Milan's CCT-based analysis a fair crack of the whip as I feel like the problems with the "piggybacking" analysis are not insurmountable (and the empirical problem re "who did John try to win" and other such cases might be solveable with the right theory of infinitives) but for sure I can see why someone like yourself would feel rather unmoved. Certainly the problem with ECM+DOC looks a lot less like nominal licensing when we start to say loads of other apparent nominal licensing failures follow from independent factors.

      Delete