Phrase Structure and Movement in Japanese

Japanese exhibits some unique features with respect to phrase structure and movement. It is well-known that its phrase structure is strictly head-final. It also provides ample evidence that a sentence may have more complex structure than its surface form suggests. Causative sentences are the best-known example of this. They appear to be simple sentences with verbs accompanying the causative suffix, -sase. But the causative suffix is an independent verb and takes a small clause vP complement in the syntactic representation. Japanese sentences can have a rich structure in the right periphery. For example, embedded clauses may contain up to three overt complementizers, corresponding to Finite (no), Interrogative (ka), and Report/Force (to). Matrix clauses may end in a sequence of discourse particles, such as wa, yo, and ne. Each of the complementizers and discourse particles has a selectional requirement of its own. More research is required to settle on the functional heads in the nominal structure. Among the controversial issues are whether D is present and whether Case markers should be analyzed as independent heads. Various kinds of movement operations are observed in the language. NP-movement to the subject position takes place in passive and unaccusative sentences, and clausal comparatives and clefts are derived by operator-movement. Scrambling is a unique movement operation that should be distinguished from both NP-movement and operator-movement. It does not establish operator-variable relations but is not subject to the locality requirements imposed on NP-movement. It cannot be PF-movement as it creates new binding possibilities. It is still debated whether head movement, for example, the movement of verb to tense, takes place in the language.

Attributive subcomparatives in Japanese

This paper aims to explain the difference in grammaticality between Japanese attributive subcomparatives with quantity adjectives (e.g., ooku/takusan ‘many’) and those with non-quantity adjectives (e.g., omosiroi ‘interesting’). My analysis assumes that Japanese clausal comparatives involve degree abstraction (Shimoyama 2012, see also Bhatt and Takahashi 2011). Degree abstraction is generally assumed to require movement of a null degree operator. In the case of attributive clausal comparatives, the movement takes place from the left branch position. If no operation that alleviates a left branch island violation is available, we would expect that the resulting sentences would be ungrammatical. I propose that in Japanese, quantifier float can play a role in ameliorating extraction out of the left branch islands. More specifically, I argue that (i) Japanese attributive subcomparatives with quantity adjectives involve quantifier float of the quantity adjectives, and that (ii) quantifier float carries the degree operators associated with the adjectives to a position where degree operator movement can originate.

Than clauses as embedded questions

Quantifiers in comparative than clauses often appear to take scope at the matrix level, a phenomenon that has spawned a large recent literature. Here I reopen an old line of investigation that seeks illumination in the strikingly similar behavior of quantifiers in embedded questions. A novel observation in this connection is that English clausal comparatives support quantificational variability effects. I explore the possibility of treating than clauses as embedded questions, sketching two implementations, and weigh this type of analysis against recent approaches that invoke degree pluralities. I also discuss multiple-wh configurations in clausal comparatives.

Smaller than small comparatives: the case of Basque

Adjectival comparatives like txiki baino txikiago ‘smaller than small’ in Basque exhibit some striking properties that present a challenge for previous analyses of inequality comparatives. The research on this unstudied type of adjectival comparatives in Basque – henceforth small comparatives – is especially interesting due to the rich morphology and freedom of word order that Basque displays. These two properties are vital for the testing of the hypotheses on the structure of these comparatives. First, the examination of the underlying structure of the standard of comparison reveals that these adjectival modifiers are inequality comparatives with a phrasal standard. Second, the study of the extraction constraints of the standard and the particular distribution of small comparatives evidence the fact that the standard marker in these comparatives behaves as a coordinating conjunction, and that these modifiers can appear inside a Determiner Phrase, in contrast with previously analysed adjectival comparatives. These properties are explained by assuming the functional analysis of adjectives and comparative markers (-ago ‘-er, more’), and proposing a coordination analysis of these phrasal comparatives. The study of small comparatives hence shows that the coordination analysis of comparative structures is necessary not only for clausal comparatives (Napoli & Nespor 1983 for Italian; Sáez 1992 for Spanish; and Lechner 2004 for English and German) but also for phrasal comparatives such as small comparatives.

Phrasal and clausal comparatives in Dutch

This paper presents a new analysis of Dutch comparatives and argues that they should be classified into phrasal and clausal comparatives (as e.g. Hankamer (1973) argues for English). Arguments for this classification come from differences in case marking and island effects. Merchant (2009) notes the same island effects in Greek comparatives, and by applying his analysis to Dutch we can explain the differences between the Dutch phrasal and clausal comparatives. Crucial in this analysis is the ellipsis of underlying structure, not only in the reduced clausal comparative, but also in the phrasal comparative, which reflects the similarity in their interpretation.