Hebrew has no multiple nominative construction (MNC) of the Japanese type, contrary to the claims of three recent studies. The alleged “broad subject” in the Hebrew constructions is nothing but a left-dislocated DP. A series of arguments is presented to the effect that this DP patterns with Ā-phrases and displays none of the subject properties characteristic of either standard or broad subjects.

The study of adjectival diathesis alternations lags behind the study of verbal diathesis and nominalization. This paper aims to diminish the gap by applying to the adjectival domain theoretical tools with proven success elsewhere. We focus on evaluative adjectives, which display a systematic alternation between a basic variant (John was rude) and a derived one (That was rude of John). The alternation brings about a cluster of syntactic and semantic changes – in the semantic type of the predicate, its valency and the mode of argument projection. We argue that the adjectival variants are related by the joint application of two operators: a lexical SATURATION operator (also seen in verbal passive) and a syntactic REIFICATION operator (also seen in nominalization). The analysis straightforwardly extends to similar alternations with Subject- and Object-Experiencer adjectives (proud, irritating). Among its important implications are (i) lexical saturation is not restricted to external arguments (internal ones may also be saturated), and (ii) ‘referential’ (R) roles are not restricted to nominal predicates (adjectives may assign them as well).

The unpronounced subject of infinitives, PRO, bears standard case, which is reflected on agreeing predicative elements in languages like Russian, Icelandic, Ancient Greek, etc. This case can be independent from the case of the controller DP, or identical to it (‘case transmission’). We report the findings of a novel study of case transmission in Russian, based on data collected from 30 speakers. The findings contradict some key generalizations that have gone unchallenged in the field for decades; specifically, case transmission is much more prevalent than previously assumed, often co-occurring with the option of independent case. The pattern of case transmission is determined by the interaction of a complex set of factors—the grammatical function of the controller, the shape of the complementizer, the type of control relation (exhaustive or partial), and more. The proposed analysis builds on “The Agreement Model of Obligatory Control (OC)” (Landau 2000, 2004, 2006) and strongly supports the claim that OC exploits two routes—either a direct Agree relation with PRO, or one mediated by the infinitival C. It is derivationally local and free of the “look-ahead” properties inherent to earlier accounts. Finally, we provide a description of the documented crosslinguistic variation in this domain, and situate it within a tight typological model.

This paper examines the solutions offered by Hornstein (2003) and Boeckx & Hornstein (2004) to the empirical challenges raised for the theory that Obligatory Control (OC) is a species of raising. The challenges include violations of the Minimal Distance Principle (MDP), partial control, OC into finite complements and into wh-infinitives, case-marking of PRO, PRO-gate effects, passive of subject-control verbs and sideward movement from OC complements. It is shown that none of the proposed solutions is adequate, and a superior account is generally available within an Agree-based approach to OC (Landau 2000, 2004). Furthermore, Hornstein's objections to that approach – in particular, to its treatment of partial control – are shown to be groundless, resting on misunderstandings of data or analysis.

Partial VP‐fronting, in which a verb is fronted with one argument, stranding the other one, is subject to a curious restriction in both Hebrew and English: The fronted VP‐portion must be a potential independent VP in the language. It is shown that both incremental merger and remnant VP‐fronting cannot explain the restriction, whereas an analysis incorporating late adjunction of the stranded argument can. Late adjunction, in turn, cannot apply too deeply, which explains why the same set of environments inaccessible to partial VP‐fronting force adjunct reconstruction. The analysis implies that not only Spell‐Out, but also interpretive constraints, like the θ‐Criterion, apply at the phase level. Furthermore, Condition A is shown to be another such constraint.

The fact that the specifier of T is subject both to the Extended Projection Principle (EPP) and to the Empty Category Principle (ECP) has remained an unexplained accident within Government-Binding Theory. I propose a principled account of this correlation. The EPP is a selectional requirement of functional heads (e.g., T, Top, C) that applies at PF—an instance of p-selection for an overt element. Like all selectional requirements, it applies to the head of the selected phrase, explaining why null heads cannot appear in EPP positions (thus deriving certain representational ECP effects). A wide range of empirical results follow, all unified by the exclusion of null-headed phrases from EPP positions: subject-object asymmetries in the distribution of bare nouns in Romance and sentential complements; failure of certain adjuncts to occur in clause-initial position; resistance of indirect objects to Ā-movement; and phonological doubling of heads of fronted categories. I argue against the agreement/checking view of the EPP and show that only the selectional construal allows a natural explanation of its puzzling properties.

The copy theory of movement receives the strongest form of support from instances of movement leaving phonetically visible copies. Such is the case in Hebrew V(P)‐fronting, where the fronted verb surfaces as an infinitive, and its “trace” is pronounced as an inflected verbal copy. This paper argues that V‐doubling is explained by the same algorithm that determines pronunciation of single copies in canonical chains. The phonetic resolution of chains is PF‐internal, strictly local, and need not appeal to cross‐interface recoverability constraints. Crosslinguistic variation in predicate clefts largely reflects different morpho‐phonological strategies of realizing the fronted predicate head.

Deriving the distribution of PRO in a principled manner is a central task for the theory of control. Traditionally, Case has been identified as the key to this problem: PRO was argued to bear no Case at all, or some special (‘‘null’’) Case. I argue that PRO bears standard case like normal lexical DPs; clear evidence comes from languages with case‐concord (Russian, Hungarian, Icelandic). Moreover, PRO (and obligatory control) may occur in finite clauses (Hebrew, Balkan languages). Conclusion: PRO's distribution must be completely divorced from Case, possibly because abstract Case does not exist. The alternative is to tie the distribution of PRO to the specific values of [T] and [Agr] on the I and C heads of the embedded clause (Landau 2004). A feature‐based algorithm predicts the distribution of PRO in a variety of clausal complements. It is shown that the system naturally explains some intriguing correlations between obligatory control and agreement in Basque and Welsh complementation structures.