Earticle

The current paper has provided various example sentences that support Lee's (1997) discussion of the long-distance licensing of negative polarity items (NPIs) in Korean from non-finite clauses using the complementizers such as –ki/ko. This paper has also supported Lee's (2024a,b) claim that in Korean, NPIs derived from non-finite embedded clauses, which form a weak CP phase, are licensed by negation in the higher clause. Specifically, when licensed by a negative implicature, the NPI te isang 'any more' must be used in the complement clause of the negative implicature, which was extensively discussed. Theoretically, in non-finite clause boundaries, the CP functions as a weak phase, exempting it from the Phase-Impenetrability Condition (PIC), thereby allowing the NPIs in the subordinate clause to be licensed by the negation in the higher clause. This distinction is crucial because in Korean, long-distance NPI licensing is permissible only when the NPI appears in the infinitival complement, which contrasts with English where NPIs can be licensed across both finite and non-finite clauses. In future research, I plan to explore how multiple NPIs can be licensed from the embedded non-finite clause within the bi-clausal structure.

This paper investigates intervention effects in T(OUGH)- constructions in English. Hartman (2011) reports that the intervening attitude holder or judge PP prohibits the association of the matrix subject DP with the gap in the embedded infinitive clause of T-constructions. Hartman (ibid.) argues that intervention effects in these constructions make a strong case for the long-movement analysis for the matrix subject DP. In this paper we explore the different line of analysis, proposing à la Chomsky (1981) that the matrix TOUGH adjective is an individual-level (IL) predicate and converts the ensuing embedded infinitive form of verb into the same type of IL predicate. Consequently the two adjective and verb are combined into one complex IL adjective, mediated by raising of the functional category ⓘ-Pred that makes the embedded infinitive verb enter into IL predication. However, when the embedded infinitive clause is extraposed after the attitude holder or typical adjunct PP, ⓘ -Pred raising is disallowed, thereby failing proper predication in the matrix clause; this in turn accounts for intervention effects in T-constructions.

When the VP topic in English appears in the sentence whose predicate is individual-level, its trace within the individual-level predicate is bound to the event operator (i.e., preposed VP topic in the spec of CP). When the individual-level predicate appears in the lower clause of the periphrastic causative construction, the eykey-marked causee in the periphrastic causative construction obligatorily undergoes case-alternation to the lul-marked causee, which induces the lower individual-level predicate to get a caused event. The accusative- marked causee in the periphrastic causative construction undergoes movement from the matrix VP-adjoined position to the matrix spec of the vP to be assigned specificity against the matrix v. When an event operator (i.e., specific or topic argument with discourse effects here) in the matrix clause is coindexed with the pro or pronoun in the lower individual-level predicate of the periphrastic causative construction, the lower individual-level predicate comes to have a caused event.

The SCs selected by consider-type verbs and those selected by want-type verbs show different syntactic behaviors with respect to Binding Condition C, passivization, and topicalization. This fact naturally leads us to conclude that they should be treated as having different syntactic structures. This paper proposes proper syntactic structures for the two types of SCs on the basis of Basilico’s (1997) finding that consider-type verbs select SCs involving ILPs while want-type verbs select SCs involving SLPs. Especially, extending Diesing’s (1992) claim, which is that an SLP has an event argument, this paper argues that SCs with an SLP involve an event argument whereas SCs with an ILP do not. This point is crucially exploited in explaining the syntactic differences of the SCs selected by the two types of verbs in question. There is a potential counter-example to this paper’s proposal. It is the case that perception verbs, which select SCs containing an SLP, allow passivization of SC subjects. But it is not problematic if we adopt Higginbotham’s (1985) theta binding.

The main purpose of this paper is to provide an account of why in certain environments bare plural objects get a generic interpretation and what the semantic core of their genericity is. Generally, bare plurals have been regarded as having two different interpretations (i.e. generic and existential), and nor only in subject position but in direct object position bare plurals show this difference. There are two kinds of verbs that can have generic bare plural objects (i.e. individual-level predicates and habitual verbs), so proper theory must give an account of generic objects of both kinds of verbs. In this paper, I present two major types of previous approaches that deal with the problem of generic bare plural objects in terms of syntactic information and semantic roles respectively, and I point out each approach has its own problems. Adopting the notion of possible worlds and intensionality as relevant factors, I provide a fundamental explanation about the genericity of bare plural objects. Furthermore, it is also shown that the notion of possible worlds gives us a way to deal with generic objects of individual-level predicates and habitual verbs in a unified way.

This paper explores why the nun-marked nominal occurs as a contrastive topic in the embedded [Spec, CP] or in the sentence-medial position and why it occurs as a plain topic in the matrix [Spec, CP], centering around the tough and likely varieties. I argue that the contrastive topic and plain topic in the tough and likely varieties are more compatible with the stage-level predicate and individual-level predicate respectively. C. Lee (2002) distinguishes a contrastive topic of the stage-level predicate from a non-contrastive topic (i.e., plain topic) of the individual-level predicate in a simple clause. The nun-marked nominal in the sentence-initial position can be interpreted as a plain topic (Bak 1986), irrespective of whether the compatible predicate is an individual-level or stage-level one. If a nun-marked nominal is interpreted as a contrastive topic, it undergoes Chomsky's (2005) peripheral movement to the Spec of a phase under my system.

This paper proposes a mapping algorithm for generic constructions. By adopting, yet modifying Diesing’s (1992) proposal, I suggest a mapping structure for generic interpretation. In particular, primarily based on Korean data, I pin down TopP/FocP as the scope for the generic operator. This new mapping algorithm can be further applied to the so-called ECM construction where the semantic subject of the embedded clause carries the accusative case, which always begged the question about its precise structural position. If the proposed mapping system is true, it can shed new light on this matter. In addition, I also present evidence against inherently dividing verbs into stage-level and individual-level, arguing for a distinction purely dependent on syntactic contexts. Given this, the proposed mapping algorithm in this paper can provide an account for the generic representation in kind-referring indefinite nominals and in characterizing predicates together.

This article discusses the Individual-level Predicate Sentence consisting of a verb predicate. Based on the phenomena that the verb predicate does not have a tense or aspect meaning in some cases, Individual-level Predicate Sentences can be divided into three types. Each is established by different reasons, but from the perspective of Ecological Psychology, all are based on observer’s exploration activities. This article focuses on one type and finds that its establishment has undergone three processes: (1) implementation of exploration activity, (2) state findings, and (3) adding findings to things as its attribute. The various types of Individual-level Predicate Sentences form a continuum: at one pole, the sentence can only be used by specific people, at the other pole, it can be used by general people. Individual-level Predicate Sentences have a hierarchical structure like this. This article also revealed that the Stage-level Predicate Sentence is at the stage of exploration activities, while the Individual-level Predicate Sentence is at the stage of attribute granting. The two are at a different perceptional stage.