Search

In the present study, I trace the shift of the modal adverb perhaps and its outdone rivals (i.e., peradventure, percase, and perchance) through the history of English. This analysis is based primarily on the OED and its quotation database, complemented by additional data from different datasets, namely, A Corpus of English Dialogues 1560–1760 in Early Modern English and the Brown family of corpora in contemporary English. After extracting instances of perhaps and related expressions from the datasets above, this study examines the following contextual factors: (i) the position occupied by the items in a clause, (ii) the combination of modal verbs in the same clause, and (iii) their parenthetical use. By analyzing the combined data, I provide a time chart for these adverbs and their competition for prevalence. The findings for the increasingly predominant perhaps demonstrate the grammaticalization of its epistemic function as a modal adverb and more recent (inter)subjectification to include pragmatic meanings concerning the relational basis between speaker and addresse

Japanese has (at least) five giving verbs: ‘yaru,’ ‘ageru,’ ‘sashiageru,’ ‘kureru,’ and ‘kudasaru.’ These are inherently deictic, and divided in a cross-cutting fashion in terms of (i) the viewpoint from which the giving event is described and (ii) the social/psychological factors relating to the speaker, the hearer, and the event participants. The giving verbs can also be used as auxiliary verbs with the same two-way categorisation. In this article, we reveal the legitimate and illegitimate combinations of giving verbs in a main-verb-auxiliary cluster, and propose a functional-pragmatic account: the doubling of giving verbs is possible only when a conventional implicature encoded in the main verb is consistent with a conventional implicature encoded in the auxiliary verb. The proposed account is based on theory-neutral assumptions/mechanisms, and it can thus be integrated with previous analyses, with the consequence of broadening empirical coverage.

Modal subordination is the phenomenon whereby a proposition in the scope of a modal expression is semantically subordinated to a proposition in the scope of another modal expression in a preceding sentence. This paper discusses to what extent reportative evidentials participate in modal subordination by describing and analyzing the modal subordination behavior of the German modal verb sollen in its reportative use. In doing so it contributes, indirectly, to the ongoing discussion in the literature on whether evidentials are a subcategory of epistemic modals, providing further evidence for the position that not all evidentials have an epistemic meaning component. It is shown that reportative sollen licenses modal subordination in first position, but can itself not be subordinated. This is explained by analyzing it as taking an informational rather than a realistic modal base (which is what epistemic modals combine with). While sollen cannot be subordinated, it nevertheless licenses inter-sentential anaphoric dependencies for anaphora in its scope when it occurs in second position, a property that is usually taken to be evidence for modal subordination. The paper puts forth the hypothesis that this is made possible by sollen being interpreted in coordination with the first modal, instead of subordination.

This paper investigates the use of possessive morphosyntax to express modal necessity, as in the English use of have to . We claim that possessive modality constructions arise because both possession and necessity express a relation of inclusion between two arguments of the same semantic type: possession involves a relation of inclusion between two 〈e〉-type arguments, while necessity involves inclusion between sets of worlds. Differences between the two arise from their different syntax: possessive have expresses possession via syntactic transitivity, while modals conceal one argument within the modal head. The similarities and differences are captured within a realizational approach to morphology, in which vocabulary items like have and must are inserted to spell out structures consisting of formal features. The proposal is then extended from have -possession languages such as English to be -possession languages, focusing on possessive modality in Hindi-Urdu and Bengali. We argue that the possessive/modal head can be “applicative-like,” licensing oblique case on an argument that raises to its specifier. This account explains why possessive morphosyntax is uniformly used to express modal necessity, and not other modal meanings: the universal force of elements like have (to) follows from the inclusion relation expressed by possession. Possessive modality thus sheds light not only on the semantics of possession but also on the compositional syntax of modal operators.

This study examines vowel epenthesis in Hebrew verbs with stem-medial glottals. The stem-internal epenthetic vowels of these verbs colloquially display variation in 3rd person feminine forms in some verb templates. While the normative epenthetic vowel in these cases is a, there are cases where it is colloquially e. We provide empirical evidence for the variation (or lack thereof), accounting for the differences 2 among verb templates and the variation by appealing to paradigmatic faithfulness constraints, and grammatical components deriving verbs. We argue that vowel selection is motivated by competing faithfulness to three different paradigms: other feminine 3rd person forms, base forms, and the general inflectional paradigm. The component of the grammar deriving verbs, lexicon vs. syntax, as well as the way verbs are stored in the lexicon, determine which paradigmatic relation is relevant in different cases. The results of this study show the importance of paradigm accessibility in morpho-phonological processes. In addition, they point to a high correlation between the degree of variation in such processes and the locus of application and storage.

Mood selection under doxastic, or belief, predicates in Spanish have been traditionally treated as cases of polarity subjunctive: affirmative predicates license the indicative and negated the subjunctive. In previous accounts of mood in Spanish, researchers (i.e. Haverkate, 2002 , Quer, 1998 , Mejias-Bikandi, 1998 , Mejias-Bikandi, 2016 ) argue that either subjunctive or indicative is felicitous under negation as mood in this environment is pragmatically motivated. Yucatec Spanish is a contact variety that allows the use of the subjunctive in both affirmative and negated doxastic predicates. I use Yucatec Spanish data to argue that mood is licensed by the speaker's epistemic uncertainty, not negation. As an alternative, I adopt the notion of (non)veridicality ( Giannakidou, 1998 , Giannakidou, 2015 , Giannakidou and Mari, 2020 ) and follow Mari's (2016) proposal of suppositional subjunctive selection to argue that mood is triggered by the evaluation of the proposition in Yucatec Spanish. When a speaker of Yucatec Spanish evaluates a proposition as nonveridical, she recognizes the possibility that the proposition could be either true or false. In such cases, the speaker selects the subjunctive in the embedded clause regardless of the presence or absence of negation. Other nonveridical environments (i.e. interrogatives and modals) are also briefly discussed as both license higher rates of subjunctive use in Yucatec Spanish.

The difference in the default prosodic realization of simple sentences with unergative vs. unaccusative/passive verbs (assigning early nuclear accent with unaccusative/passive verbs but late nuclear accent with unergative verbs) is often related to the syntactic distinction of their nominative arguments as starting off in different hierarchical positions. Alternative accounts try to trace this prosodic variation back to asymmetries in the semantic or pragmatic contribution of the verb to an utterance. The present article investigates the interaction of the assignment of default nuclear accent with the predictability of the verb. In an experimental study testing the acceptability of nuclear accent assignment, we confirmed that the predictability of the verb influences accentual preferences (such that highly predictable verbs are preferably not accented). However, the experiment also reveals that the unaccusativity distinction cannot be accounted for by means of pragmatic phenomena of this type: the two verb classes are associated with distinct accentual patterns in the baseline condition, that is, without the predictability manipulation. (C) 2014 Elsevier B.V. All rights reserved.

Spanish future and conditional morphemes may display inferential readings: Elena ganaria la carrera ayer, as in ‘Helen must have won the race yesterday’, a present deduction about a past event. They may also display readings known as concessive, dubbed ‘mirative’ here: Elena ganaria la carrera ayer, pero no esta contenta ‘Helen might have won the race yesterday, but she is not satisfied’. The proposal is that such futures and conditional affixes encode an evidential modal involving a body of indirect information, which the speaker may vouch for or not. This modal contributes to propositional content and can be syntactically and semantically embedded, so is not an illocutionary marker. It is a degree expression that does not reduce to necessity or possibility, so is reminiscent of gradable adjectives such as tall or probable. It displays the flexible anchoring characteristics of predicates of personal taste such as tasty, so may partially differ from canonical epistemic modals. Ordering sources and anchoring behavior combine in such a modal to trigger various levels of confidence in the information, resulting in variability in force, which may range from certainty/necessity to doubt/possibility in both inferentials and miratives.

Coherence relations differ in their tendency to be explicitly marked. How such relations are recognized and what determines their tendency to be marked is a matter of debate. The connective so represents a special case: it can be used to signal RESULT coherence relations and the more specific cause-effect relation of PURPOSE, but overt marking has been claimed to be required for PURPOSE and optional for RESULT. We present written corpus and experimental results on the use of so that show that RESULT and PURPOSE with this connective can be reliably distinguished from each other, and that the modal auxiliaries can/could and will/would are strongly associated with PURPOSE. In the corpus study, PURPOSE always occurs with explicit so, while RESULT is often left unmarked. These results are in line with recent claims based on annotated corpus data that implicit (unmarked) and explicit (marked) coherence relations can be qualitatively different (e.g. Sporleder and Lascarides, 2008; Webber, 2009). However, in our experiments using strongly purposive event pairs, 35-40% of examples were identified as PURPOSE without a connective or a modal verb cue. We argue that the difference between the corpus results and the experimental results can be explained as a difference between the tasks of speakers and hearers, and we outline an explanation for how marking can be obligatory for PURPOSE relations and yet optional for RESULT. We also propose that nonveridicality seems to play a key role in a marking requirement for PURPOSE, and explain why the unusual marking pattern found makes it difficult to give a pragmatic account similar to more well-known language asymmetries. (C) 2014 Elsevier B.V. All rights reserved.

In this paper I examine the syntax–semantics of the Romeyka infinitive, still to be found in an endangered Greek variety uninterruptedly spoken in the historical region of Pontus, Turkey. It is shown that the infinitive is found: (a) as a complement to negated past tense modals; (b) in before-clauses; (c) in counterfactuals. My proposal is that the Romeyka infinitive is licensed as a NPI. It is argued that antiveridicalidity (in the sense of Giannakidou, 1998 et seq.) licenses the infinitive and therefore explains the unavailability of the Romeyka infinitive in other nonveridical contexts such as: (i) questions, (ii) nonveridical conditionals, (iii) present and imperfect tense negated modals. The analysis set out here (i) proposes a new type of NPI, namely an infinitive; (ii) reinforces the disengagement between morphological negation and antiveridicality; (iii) highlights parallels with Romance polarity subjunctives, which, like the infinitive, also share a T-C dependency; the latter may have rendered the Romeyka infinitive diachronically more prone to developing a neg-dependency too (Sitaridou, 2014).

The purpose of this paper is three-fold: (i) I argue that topic A-movement exists in Chinese. Specifically, I show that argument displacement in Chinese raising modal constructions (RMC) containing raising modals like yinggai ‘should’ and keneng ‘likely’ is A-movement to spec-TP, yet semantically exhibits topicality; (ii) even though Miyagawa's (2010) extension of Chomsky, 2007 , Chomsky, 2008 feature inheritance hypothesis to the Topic feature provides a possible derivational system to capture topic A-movement in this construction, I show that object topic A-movement in this construction is underivable under Miyagawa's (2010) probe-driven system of movement based on feature inheritance. This is because of a timing problem regarding feature inheritance and Chomsky's (2001) Phase-Impenetrability Condition independently noted by Epstein et al. (2012) and Richards (2011) ; (iii) I propose that two major modifications are needed to maintain the basic assumptions of Miyagawa (2010) to derive object topic A-movement in RMC: (a) a more fine-grained characterization of topic A-movement on the basis of the dissociation of feature interpretability and feature valuation (see Pesetsky and Torrego, 2007 , Carstens, 2010 , Carstens, 2011 , Boskovic, in press ), and (b) the adoption of Boskovic’s (2007) moving-element-driven theory of movement.

Previous analyses of the double modal construction in Southern United States English (SUSE) have been unable to fully explain all of the observed data regarding negation, questions, and stranded quantifiers. In this article, I argue for a structure placing the first modal in an MP merged above TP and the second modal in T. This structure has the benefits of maintaining the distinction seen between the two modals and explaining the separation of the modals by negation and stranded quantifiers while being able to explain the raising of the second modal in main clause questions. To support the proposed merged MP structure, I argue that the first modal of a double modal pair is immune to sequence of tense effects, and I take this as evidence that first position modals lack syntactic Tense. This is further supported with crosslinguistic evidence for epistemic modality located above Tense. This article thus explains SUSE double modals as an instance of parametric variation regarding the availability of epistemic modality to be located above Tense. I also address issues associated with the proposed structure involving the Head Movement Constraint and the proposed EPP feature on M.

The present paper discusses the differences in distributional properties between schijnen and lijken : two Dutch raising-verbs that are often translated to English ‘seem’. According to the cartographic approach, for instance Haegeman, L. (2006a. Clitic Climbing and the Dual Status of Sembrare . Linguistic Inquiry 37, 484–501), the differences in distributional properties follow from a difference in their position in the syntactic structure. Lijken is proposed to be a lexical verb inserted in a low position. Schijnen on the other hand is a functional head encoding speaker-related evidential mood, inserted in a high position. This paper shows that postulating different syntactic positions is unnecessary. The distributional differences follow from restrictions on scope as a result of the subjectivity directly (or speaker-related mood). The subjectivity of schijnen results in positive polarity behavior and as such schijnen is ruled out in nonveridical contexts (cf. Ernst, T., 2009. Speaker-oriented adverbs. Natural Language and Linguistic Theory 27, 497–544). A benefit of the present approach is that it is able to fit in cases in which schijnen is used non-subjectively and can appear under modals and aspect. In addition, it is able to accommodate the impossibility to embed subjective lijkt-me under modals and aspect.

Certain modal auxiliary and control constructions in Thai differ only in the choice between a modal or control verb. Despite a possibility to analyze them as (subject) raising vs. control predicates, the present study argues that their counterparts with the marker thii may not be derived via the raising vs. control conventions. The reasons include the presence of thii caʔ which marks infinitival complements and a possible connection between modal auxiliaries and lexical control verbs revealed by a test on Thai clausal idioms. This study proposes an additional functional category, Mood (Irrealis) Phrase, M(I)P, to account for the co-occurrence restrictions of thii and non-assertive (irrealis) infinitival complements, and develops a unified analysis for modal and control constructions with thii , parallel to Kayne's (2000) approach to control infinitives with di in Italian. The analysis developed here is consequential to the syntax and semantics of modals and control infinitives, cross-linguistically. Firstly, it suggests that modals that behave identically to control predicates be analyzed by means of the control approach, rather than the raising predicate approach, which has been held customarily. Secondly, it establishes the role of the MIP, which enables the expression of a hypothetical tense that occurs in infinitival complements introduced by markers (i.e. thii in Thai, di in Italian, and de in French).

This paper develops a new perspective on the question of what type of verb movement the modern Celtic languages display, V to I movement or V to C movement. Under the standard assumption that the subject remains relatively low in these languages compared to Germanic languages, this category fails to be a diagnostic, since both verb movement operations cross the subject. The analysis presented here starts from the theory of verb movement developed in Koeneman (2000), which argues that the verb does not move to prefabricated functional head positions but moves in order to project one or more of its functional features (tense and agreement). It is shown how this theory offers a better understanding of verb movement (parameters) than mainstream approaches. After developing a concrete proposal for verb movement variation in Germanic, the paper tries to extend this analysis to the Celtic languages, so that a uniform analysis of verb movement (variation) ensues. The main conclusion is that it is very unlikely that Celtic languages have a verb movement operation involving an agreement projection. They are better analyzed as having a verb movement operation that projects tense features. The result is that the nature of verb movement in Germanic and Celtic is basically the same.

Evidentiality, the linguistic encoding of speaker's information source, is an understudied phenomenon in languages such as English and German, which do not encode evidential meaning in the grammar. However, there are several lexical means by which speakers of these languages can encode their source of information. It should thus come as no surprise that perception verbs – those verbs that denote sight, sound, touch, smell, and taste – are one of the most frequently used markers of evidentiality in English and German, for perception shapes most (if not all) of our epistemology. Although most studies of perception verbs have focused on vision, the primary sensory modality, this study will focus on hearing, the second most prominent sensory modality. That is, this will be a study of how verbs of auditory perception are used by speakers of English and German to indicate how hearing serves as evidence for the propositions they utter. A corpus-based analysis of English hear and sound , as well as German horen ‘hear’ and klingen ‘sound’ was conducted, and two aspects of the findings are examined: the specific construction types in which these verbs of auditory perception are used evidentially, and what specific evidential meanings are bound to these constructions.

Perfect modals are modals combining with (nonfinite) perfect have — may have , must have , might have , or should have . Under their non-root construal, perfect modals yield three temporal readings of the situation-time of the propositional modal complement under their scope: (i) a present perfect construal (past situation-time relative to a present reference-time); (ii) a future perfect construal (past situation-time relative to a future reference-time); (iii) a perfect future construal (future situation-time relative to a past reference-time). The goal of this paper is to provide a unified account of the temporal readings of perfect modals within the framework developed in our previous work, where tenses and aspects are dyadic predicates of spatiotemporal ordering projecting their time arguments in the syntax as Zeit -phrases. We argue that the temporal syntax of perfect modals supports the proposal that these Zeit-Phrases can enter into anaphoric dependencies and undergo phrasal movement (QR) to higher scope positions, just as ordinary DPs/QPs can.

This paper discusses the semantically parenthetical use of clause-embedding verbs such as see, hear, think, believe, discover and know. When embedding verbs are used in this way, the embedded clause carries the main point of the utterance, while the main clause serves some discourse function. Frequently, this function is evidential, with the parenthetical verb carrying information about the source and reliability of the embedded claim, or about the speaker's emotional orientation to it. Other functions of parenthetical uses of verbs are discussed. Particular attention is paid to the parenthetical uses of verbs which are standardly assumed to require their complements to be presupposed. It is demonstrated that when so used, these verbs are in no way presuppositional; that is, there is no presumption, or even pretense, that their complements have common ground status. It is further demonstrated that this loss of presuppositionality is not accompanied by a lack of commitment on the part of the speaker to the truth of the complement, as in the standard cases of non-presuppositional uses of these predicates. It is argued that this non-presuppositional use of factive verbs provides support for the (minority) view that presupposition is not a conventional property of lexical items.

This paper presents an analysis of sentences with the postverbal modal element dak in Cantonese and Hakka. We argue that dak is a verb, though in Cantonese and Hakka it is partly defective. It acquires different meanings depending on the position it occupies, which is either the position for modal verbs in the matrix clause or a modal position in the result denoting small clause. The analysis, if on the right track, has the consequence that there is no long distance Agree. Furthermore, it means that there is evidence for an IP-like functional projection in Chinese.

The paper discusses the structure of the Balkan Slavic Future tenses, with a particular reference to the Future tenses with modal clitics and finite verbs. It is shown that all the Balkan Slavic Future tenses have developed from restructuring configurations in which subjunctive constructions appear in complement positions of forms of the Old Slavic verb xotěti ‘will/want’. The development has gone through three stages, though not all Balkan Slavic languages have gone through all of them. In the first stage, represented in contemporary Serbo-Croatian, we have finite modal clitics plus infinitives or subjunctive constructions. In the second stage, which is to some extent represented in the South-Eastern Serbian dialects, non-finite modal clitics are followed by subjunctive constructions. In the third stage, represented in Macedonian and Bulgarian, non-finite modal clitics are followed by finite verbs. In contemporary Serbo-Croatian, the Future tense is constructed by finite modal clitics plus infinitives or subjunctive constructions. The Serbo-Croatian Future tense with subjunctive constructions is a mono-clausal raising configuration with a finite auxiliary and a finite lexical verb. Like the Macedonian and Bulgarian Future tenses, it has evolved from a restructuring configuration with two finite lexical verbs, though the evolution of the Macedonian and Bulgarian Future tenses has passed through three stages. In the first stage, the Macedonian and Bulgarian Future tenses had a structure such as the one in the Serbo-Croatian Future tense; in the second stage, they had another intermediate structure—the one witnessed in the contemporary South-Eastern Serbian dialects, in which a non-finite modal clitic is followed by a subjunctive construction; in the third stage, the non-finite modal clitics came to be followed by finite verbs. The change from finite to non-finite modal clitics relates to the loss of the subjunctive mood complementizer: When the ϕ-features of the modal clitics disappeared and T/AgrSP reappeared to the left of the lexical verb, the subjunctive mood complementizer became superfluous.

While there is a concensus that auxiliary verbs differ in one way or the other from main verbs, linguists do not always agree on what determines the difference. This is especially so in serialising languages where two or more independent verbs can occur in a clause. It is not always clear whether the ability of a verbal form to express a temporal, aspectual or modal notion in such languages is enough for them to be analysed as auxiliary verbs or whether the nature of the form is equally important. This paper argues that a reduction in material integrity of a verbal form is a very important criterion for distinguishing verbs in serialising languages.

In this paper it is shown how an account of the English auxiliary system that has been independently proposed to deal with problems in standard analyses also provides a natural treatment of microvariation among varieties of English. The phenomenon is the use of non-emphatic periphrastic/dummy do in positive declaratives (Mary did visit her brother), here called "spurious do," as found most famously in the English of the 1500s, but attested also in some modern dialects and registers and in child English, and closely related to the use of tun in colloquial German. The framework adopted dispenses with two standard but problematic claims about English INFL: the exceptional ability of be and have to raise to Tense, even across negation, and the existence of PF affix lowering. Instead it is claimed that English has overt verb raising and that finite be/have are base-generated in INFL, above negation; independent support for the latter is provided from VP ellipsis. The analysis of do is that it is an allomorph of the indicative value of the Mood head, whose other indicative allomorph is zero. Mood is above Tense and is where modals are base-generated. It is shown that this system cannot block the generation of spurious do, because this would require transderivational comparison. Thus, the narrow syntax makes spurious do freely available. Languages and dialects differ on the extent to which they make use of this option. All else equal, it should be dispreferred because it involves one more word than its counterpart without do, but numerous advantages, including processing and rhetorical benefits, can outweigh this. The conclusion is that do cannot be analyzed as a strictly last-resort device in the way proposed in Chomsky's classic analysis. (C) 2003 Elsevier B.V. All rights reserved.

This paper examines the object fronting construction of American sign language (Liddell, S., 1980. American Sign Language Syntax. Mouton, The Hague; Matsuoka, K., 1997. Verb missing in American sign language. Lingua 103, 127–149), and its relation to verb raising and aspectual inflection. In this construction, word order, canonically S–V–O, is modified to O–S–V. Matsuoka argues persuasively for a relationship between object fronting and overt verb raising, driven by the presence of an affixal verbal inflection [Lasnik, H. 1995b. Verbal morphology: syntactic structures meets the minimalist program. In: Campos, H. and Kempchinsky, P. (Eds.), Evolution and Revolution in Linguistic Theory. Georgetown University Press, Washington, DC, pp. 251–275]. Thus, her analysis reduces ASL object fronting to an instance of ‘object shift’ as occurs in the Scandinavian language family. However, evidence is provided here, from the distributions of adverbs and modals, and from the interaction of object fronting with question formation and raising contexts, which indicates that ASL object fronting targets a position high in the clause. A new model of ASL object fronting is developed against the backdrop of Matsuoka and the Lasnik theory of verbal morphology. This new framework has much in common with Matsuoka's account, while attaining greater empirical coverage. In this framework, as in Matsuoka, overt verb raising is triggered by the presence of an aspectual head (Asp). However, I argue that AspP is part of an articulated COMP layer [Rizzi, L. 1997. The fine structure of the left periphery. In: Haegeman, L. (Ed.), Elements of Grammar: Handbook in Generative Syntax. Kluwer, Dordrecht. pp. 281–337]. Object fronting targets spec-AspP, an A′-position. This model is shown to provide a better account of the distribution of adverbs and modals in object-fronted clauses, as well as explanations of the behavior of object fronted clauses in questions and in embedded contexts. Lasnik's model of verbal morphology is extended such that verbs in (some varieties of) ASL allow both derivational and lexical options for aspectual inflection. This mechanism provides for an explanation of the infelicity of object fronted clauses in embedded contexts. The analysis crucially relies on the categorial status of object-fronted clauses and the mechanism of c-selection.

This paper provides an explanation for aspects of the behaviour of the English modal auxiliaries based at the syntax–semantics interface. The core idea is a marriage of Kratzer's [Linguistics and Philosophy, 1 (1977), 337] classical analysis of modals as propositional operators with current minimalist ideas about the propositional nature of phases [Chomsky (1999) MITOPL, 18]. This gives rise to two scope positions for modals: one associated with the vP phase, and one with the CP phase. It is shown that the former scope position leads to root readings of modals and the latter to epistemic readings.

English, and a variety of Germanic languages, allow constructions such as The bottle floated into the cave , whereas languages such as Spanish, French, and Hindi are highly restricted in allowing manner of motion verbs to occur with path phrases. This typological observation has been accounted for in terms of the conflation of complex meaning in basic or derived verbs [Talmy, L., 1985. Lexicalization patterns: semantic structure in lexical forms. In: Shopen, T. (Ed.), Language Typology and Syntactic Description 3: Grammatical Categories and the Lexicon. Cambridge University Press, Cambridge, pp. 57–149; Levin, B., Rappaport-Hovav, M., 1995. Unaccusativity: At the Syntax–Lexical Semantics Interface. MIT Press, Cambridge, MA], or the presence of path “satellites” with special grammatical properties in the lexicon of languages such as English, which allow such phrasal combinations [cf. Talmy, L., 1985. Lexicalization patterns: semantic structure in lexical forms. In: Shopen, T. (Ed.), Language Typology and Syntactic Description 3: Grammatical Categories and the Lexicon. Cambridge University Press, Cambridge, pp. 57–149; Talmy, L., 1991. Path to realisation: via aspect and result. In: Proceedings of the Seventeenth Annual Meeting of the Berkeley Linguistics Society. Berkeley Linguistics Society, Berkeley, pp. 480–520]. I use data from Hindi to show that there is little empirical support for the claim that the constraint on the phrasal combination is correlated with differences in verb meaning or the presence of satellites in the lexicon of a language. However, proposals which eschew lexicalization accounts for more general aspectual constraints on the manner verb + path phrase combination in Spanish-type languages (Aske, J., 1989. Path Predicates in English and Spanish: A Closer look. In: Proceedings of the Fifteenth Annual Meeting of the Berkeley Linguistics Society. Berkeley Linguistics Society, Berkeley, pp. 1–14) cannot account for the full range of data in Hindi either. On the basis of these facts, I argue that an empirically adequate account can be formulated in terms of a general mapping constraint, formulated in terms of whether the lexical requirements of the verb strictly or weakly constrain its syntactic privileges of occurrence. In Hindi, path phrases can combine with manner of motion verbs only to the degree that they are compatible with the semantic profile of the verb. Path phrases in English, on the other hand, can extend the verb's “semantic profile” subject to certain constraints. I suggest that path phrases are licensed in English by the semantic requirements of the “construction” in which they appear rather than by the selectional requirements of the verb (Fillmore, C., Kay, P., O'Connor, M.C., 1988, Regularity and idiomaticity in grammatical constructions. Language 64, 501–538; Jackendoff, 1990, Semantic Structures. MIT Press, Cambridge, MA; Goldberg, 1995, Constructions: A Construction Grammar Approach to Argument Structure. University of Chicago Press, Chicago and London).

Prasada and Pinker's (1993) subjects provided past tense forms of nonce verbs. The subject's willingness to provide irregular past tense forms correlated with the verb's phonological similarity to existing irregular English verbs. However, there was no correlation between the number of nonce verbs assigned regular inflection, and the verb's similarity to existing regular verbs. According to the dual-route model, this is expected since irregular items are stored in associative memory, while regular items take an allomorph of -ed by rule. A singleroute connectionist simulation failed to duplicate the subject's behavior on regular verbs. Two instance-based models were applied to the data: Analogical Modeling of Language and the Tilburg Memory Based Learner. Each model employed a similarity algorithm to determine the behavior of all regular and irregular items. Both models successfully mirrored the subject's responses. Therefore, the data are consistent with an instance-based single-route model of morphology.

In the current literature, verb gapping is commonly said not to exist in Chinese. Given Johnson's (1994, 1996) analysis of verb gapping as ATB movement of the verb to a functional category, the apparent lack of verb gapping in Chinese seems to fit in nicely with another idea prevalent in Chinese linguistics viz., that verb movement is confined to the VP-shell (cf. Huang, 1992, 1994; Tsai, 1994). The present article argues against these claims and provides evidence for the existence of verb gapping — though restricted — in Chinese. Furthermore, it shows that Johnson's ATB movement analysis of gapping can be successfully applied to Chinese as well which leads to the typologically important result that the verb in Chinese can — under particular circumstances — leave the VP and raise to a functional category in overt syntax.

In this paper I will present and defend an analysis of (a sample of) the English modal auxiliary verbs using a relevance-theoretic semantic and pragmatic framework. I will start by discussing previous analyses of modality in English with an eye to explaining how a cluster of related meanings — epistemic, root, and other — is expressed by the same set of lexical items. I will then go on to develop a unitary semantic approach to the English modals, treating them as (mostly) incomplete propositional operators. After defending the details of my semantic account, I will show how the proposed semantics can give rise to the range of root interpretations modal verbs can receive in context. Epistemic interpretations require some further theoretical machinery, which will make crucial use of the notion of metarepresentation. Finally, I will sketch the differences between natural-language interpretations of modal operators and their alethic/logical uses.

Previous researchers have suggested that the base-generated word order in ASL is Subject- Verb-Object (Fischer, 1987; Liddell, 1980). However, there are some constructions which apparently contradict this assumption. Among these constructions are the phenomena of Verb Sandwich (Fischer and Janis, 1992), Verb Final (Romano, 1991), and Object Raising (Lid- dell, 1980). All three sentence types contain verbs with an aspectual marker. Unlike its unin- flected counterpart, an inflected verb in these constructions appears in the sentence-final posi- tion, which is not assumed to be the basic position for ASL verbs. I propose a unified account for these three ASL constructions and consider its theoretical implications for the theory of verbal morphology. I extend the idea of the Movement Analy- sis of Verb Sandwich, proposed in Fischer and Janis (1992). Based on the Minimalist approach proposed in Chomsky (1995), I argue that the derivations of all three sentence pat- terns involve overt verb raising to Infl. The verb movement to Infl occurs to save the other- wise stranded [asp] affix, even though the verb itself does not have motivation to raise overtly. Hence, this verb raising is a syntactic operation driven by Enlightened Self Interest (Lasnik, 1995a,b). In our discussion, it is shown that Chomsky's (1995) Checking Theory faces empirical problems in accounting for the ASL phenomena in question. I will present an alternative analysis, which is crucially based on a proposal of Lasnik (1995c) that regular verbs are selected from the lexicon uninflected. The Object Raising is analyzed as an instance of Holmberg's generalization (Branigan, 1992), which states that overt object shift is possible only when the verb is overtly raised out of VP.

Many researchers have argued that the verbs appearing in English inversion can be characterized as belonging to a particular syntactic or semantic class. For example, it has been claimed that such verbs are exclusively unaccusative (Levin, 1983; Coopmans, 1989; Hoekstra and Mulder, 1990; Watanabe, 1994, inter alia), exclusively presentational (e.g., Bresnan, 1994), or exclusively verbs of existence or appearance (e.g., Penhallurick, 1984). A study of a large corpus of natural language data, however, indicates that these restrictions do not account for the full range of verbs found in inversions. Instead, inversion is subject to a pragmatic constraint to the effect that the verb must not represent new information in the discourse. In the case of be, the verb is inherently informationally light; the data in this study indicate that other verbs appearing in felicitous inversions are likewise informationally light in that they represent evoked or inferrable information in context and therefore contribute no new information to the discourse. Thus, all verbs appearing in central position in inversion can be characterized by a single discourse-based principle.

The first aim of this article is to explore as exhaustively as possible the full range of modal usage found with the verbs need and dare in English, something which has never been done before. The second aim is to explain why these two verbs can exhibit modal-like behaviour. The reason is to be found in the fact that in certain uses they express a type of meaning which is analogous to that of the core medals, sharing with the latter the impression of ‘nonassertiveness’. This provides a valuable insight into the essence of the meaning of the modal auxiliaries in English and demonstrates convincingly that the morpho-syntactic behaviour not only of need and dare but also of the modals themselves is conditioned by meaning and not by syntactic rule.

We ask how children solve the mapping problem for verb acquisition: how they pair concepts with their phonological realizations in their language. There is evidence that nouns but not verbs can be acquired by pairing each sound (e.g., ‘elephant’) with a concept inferred from the world circumstances in which that sound occurs. Verb meanings pose problems for this word-world mapping procedure, motivating a model of verb mapping mediated by attention to the syntactic structures in which verbs occur (Landau and Gleitman 1985, Gleitman 1990). We present an experiment examining the interaction between a conceptual influence (the bias to interpret observed situations as involving a casual agent) and syntactic influences, as these jointly contribute to children's conjectures about new verb meanings. Children were shown scenes ambiguous as to two interpretations (e.g., giving and getting or chasing and fleeing ) and were asked to guess the meaning of novel verbs used to described the scenes, presented in varying syntactic contexts. Both conceptual and syntactic constraints influenced children's responses, but syntactic information largely overwhelmed the conceptual bias. This finding, with collatoral evidence, supports a syntax-mediated procedure for verb acquisition.

I examine Gleitman's (1990) arguments that children rely on a verb's syntactic subcategorization frames to learn its meaning (e.g., they learn that see means ‘perceive visually’ because it can appear with a direct object, a clausal complement, or a directional phrase). First, Gleitman argues that the verbs cannot be learned byb observing then situations in which they are used , because many verbs refer to overlapping situations, and because parents do not invariably use a verb when its perceptual correlates are present. I suggest that these arguments speak only against a narrow associationist view in which the child is sensitive to the temporal contiguity of sensory features and spoken verb, If the child can hyppthesize semantic representations corresponding to what parents are likely to be referring to, and can refine such representations across multiple situations, the objections are blunted;, indeed, Gleitman's theory requires such a learning process despite her objections to it. Second, Gleitmans suggests that there is enough information in a verb's subcategorization frames to predict its meaning ‘quite closely’ . Evaluating this argument requires distinguishing a verb's root plus its semantic content (what She boiled the water shares with The water boiled and does not share with She broke the glass ), and a verb frame plus its semantic perspective (what She boiled the water shares with She broke the glass and does not share with The water boiled ). I show that learning a verb in a single frame only gives a learner coarse information about its semanic perspective in that frame (e.g., number of arguments, type of arguments); it tells the learner nothing about the verb root's content across frames (e.g., hot bubbling liquid). Moreover, hearing a verb across all its frames also reveals little about the verb root's content. Finally, I show that Gleitman's empirical arguments all involve experiments where children are exposed to a single verb frame, and therefore all involve learning the frame's perpective meaning, not the root's content meaning, which in all the experiments was acquired by observing the accompanying scene. I conclude that attention to a verb's syntactic frame can help narrow down the child's interpretation of the perspective meaning of the verb in that frame, but disagree with the claim that there is some in-principle limitation in learning a verb's content.

This paper investigates the phenomena that come under the label ‘causative alternation’ in English, as illustrated in the transitive and intransitive sentence pair Antonia broke the vase / The vase broke . Central to our analysis is a distinction between verbs which are inherently monadic and verbs which are inherently dyadic. Given this distinction, much of the relevant data is explained by distinguishing two processes that give rise to causative alternation verbs. The first, and by far more pervasive process, forms lexical detransitive verbs from certain transitive verbs with a causative meaning. The second process, which is more restricted in its scope, results in the existence of causative transitive verbs related to some intransitive verbs. Finally, this study provides further insight into the semantic underpinnings of the Unaccusativity Hypothesis (Perlmutter 1978).

Lord (1976) explains the emergence of CP complements after verbs of saying, knowing, believing, etc., in some West African languages by hypothesizing that there was a verb say , which used to occur with such verbs in a serial verb construction (SVC), and which was subsequently reanalyzed into a that -complementizer. However, in , a Kwa language of Benin, the form which occurs after verbs of saying, knowing, believing, etc., exhibits (selectional) properties typical of a verb, rather than those of a C 0 like the English that -complementizer. Furthermore, the form occurs in contexts that must be analyzed as involving a SVC. This suggests that, contrary to Lord's hypothesis, the that appears after the verbs of saying, knowing, believing, etc., in is best analyzed as a verb which can occur in a SVC.

This paper examines the issue of the structural representation of the Serial Verb Construction (SVC). It is shown that SVCs consist of a matrix verb followed by an adjunct phrase or series of adjunct phrases as the case may be. Support for this comes from adjunct extraction facts in Yoruba and from the GB concepts of Barrierhood and L-marking.

Epistemic deontic Be (as in Keats was to die young and in This door is to remain closed all day ) only occurs in finite positions, even though the verb Be is not morphologically defective and has an infinitive, an -Ing form and a participle. This brings out the insufficiency of McCawley's (1971) argument that English modals occur only as finites because of their defective morphology. It is demonstrated that epistemic deontic Be cannot be accommdated in the rule system by means of which Akmajlan, Steele and Wasow (1979) stipulate the positions of English auxiliary verbs. Hence this system must be modified. It is argued in the light of the theory on tense in Rigter that, contrary to Akmajian et al's proposal, modals are auxiliary verbs and belong to the same position class as epistemic deontic Be . The syntactic stipulation in Akmajian (1977) and in Akmajian et al (1979) that complements of perception verbs contain neither a perfect nor a modal is shown to be semantically motivated. Evidence is adduced that the progressive does occur in structures underlying perception verb complements. It is suggested that the distribution of semantic material over stacked structures is subject to a semantico-syntactic ordering principle which places the more objective predications in the scope of the more subjective predications.

This study attempts to show that the model which I developed elsewhere to explain the semantic structure of the German modal system (Bouma 1973) is also generally applicable to the modals of English. The theory underlying my approach is based on the notion of (grammatical) meaning found in Joos (1964) and Jakobson (1971). The modal auxiliary system in both languages is seen as a grammatical category (relative assertion) which stands in specific opposition to the absence of a modal in a sentence (factual assertion). The general grammatical meaning conveyed by the modal auxiliary category is that the speaker sees the event as standing in a potential relation to reality. In both languages there is a type of scalarity of probability from precariousness (50/50 probability: KANN, DARF: CAN, MAY) to imminence (near certainty: MUSS, WILL; MUST, HAVE TO, WILL, WANT TO). Both modal systems can also be characterized in terms of the speaker's view of the involvement of the actor (grammatical subject). If the speaker views the actor as the initiator of the realization of the event he uses a subjective modal (e.g. German WILL: English WANT TO), and he uses an objective modal if he views the initiator as an agency outside of the actor (e.g. MUSS; MUST).

Positing two constraints on auxiliary co-occurrence, the Base Order Constraint and the Once-per-Clause Constraint, this study demonstrates that the auxiliary order reflects the historical sequence of grammaticalization, that the appearance of co-occurences developed in the same historical sequence, that the Once-per-Clause Constraint has changed its governance in the past two hundred years to a grammatical governance rather than a lexical one, and that the tense paraphrase, do , was prevented from spreading to clauses with other auxiliaries by the loss of tense and non-finite forms in the modals it has to precede.

Up to now it has been assumed that the epistemic use of modal auxiliaries such as must is entirely distinct from their deontic use, though the widespread association of the deontic with the epistemic in various languages suggests that they have a common semantic base. A consideration of practical examples of must in English shows that this base can be identified with the deontic notion of ‘demand’, which is closely related to ‘desire’ or ‘will’ (i.e. volition). may expresses the logically related notion of ‘permission’. The basically deontic function of must and may is easily detectable in all their uses including the epistemic. The epistemic lies at the extremity of a deontic gradient which also affects associated rules of syntax such as those affecting the scope of not and the use of have + en . The performative analysis is untenable. The notion of ‘demand’ or ‘will’ is present in all the deontic-epistemic modals and in will and shall , and the main source of their varied interpretability is the presence of unspecified arguments in the deep structure.

It is a frequently observed phenomenon that serial verbs are prone to reanalysis (i.e. changing of grammatical category). In African linguistics, for instance, there has been a lenghty debate on whether certain serial verbs have undergone reanalysis as prepositions (see for instance Givon (1975) and Lord (1973)). Reinterpretation of the verb ‘give’ as a valence increasing device which adds a recepient/benefactive to the verb frame is particularly frequent and occurs for instance widely in the New Guinea area, as for example shown for Abelam by Laycock (Laycock (1965: 55)) or for Awtuw by Feldman (1983). Similar phenomena are also attested from South-East Asia (e.g. Clark (1978)). A possibly unique type of reanalysis of serial verbs has occured in Imonda and some other closely related members of the same family. In these languages some serial verbs have undergone reanalysis as noun-classifying verbal prefixes. The purpose of this paper is to describe this unusual type of noun-classification system and to throw some light on how it may have arisen.

Future, predictions, desires, wishes, hopes, obligations, permission, and prohibition may all be considered categories of hypothetical information because the events they refer to are only potential, and not actually occuring. This paper explores some of the grammatical and semantic aspects of English verbs (in particular modals and catenatives) which express these types of hypothetical information. Experimental evidence for the semantic organization of hypothetical verbs is also presented. Subjects were asked to rate the semantic similarity of each pair of hypothetical verbs in the stimulus set. A hierarchical clustering analysis of similarity ratings can show some of the semantic components underlying a semantic field such as the one investigated here. Finally, a hierarchy of hypotheticals is proposed, whereby the hypothetical verbs are arranged from greatest to least probability of an event occuring. Evidence from double modal dialects and various sentence types suggests that within an utterance verbs higher in the hierarchy must be placed before less hypothetical elements.

Up to now, much work on the serial verb construction has been devoted either to the historical derivation of the construction or to its synchronic grammatical status. By focusing on a hitherto neglected aspect of the serial verb construction, that of the commutability constraints on verbs operating in concomitant serialization, this paper uses examples from Yoruba to highlight both the relatedness of verb serialization to other types of syntactic typology and the shortcomings of current research on the derivation of the serial verb construction. We set up a positional semantic classification of verbs which facilitates the prediction of permissible, doubtful, and non-permissible serial verb strings, showing (i) the commutability restrictions arising from such a functional-semantic classification and (ii) how various reductionist derivational explanations of the serial verb phenomenon can arise from undue concentration on one or the other of the various verb semantic functions. The utility of the approach dopted here is further illustrated by its capacity to handle an indefinitely long string of verbs, thus releasing verb serialization research from overconcentration on only two verbs in series.

This paper presents a hypothesis concerning the development of verb agreemnent in the Ugrian languages. It will be argued that with some modification Dik's reconstruction of the rise and fall of ergative systems (Dik 1978, 1980) can account for the agreement between verb and subject and between verb and object in these languages.

A re-analysis and classification of English modals are proposed. Most of the candidates for inclusion in the modal category are eliminated since they are independently generated as verbs, adjectives, and adverbs. Those included are further classified into the ‘pure’ (M) and the ‘quasi’ (Q) by their ability to combine internally in the auxiliary. On syntactic grounds, then, Chomsky's re-writing rule is revised as. Aux → Tense(M)(Q) n ( have + en )( Q ) n ( be + ing ) ( Q ) n . Based on this segmentation, both modals and quasi-modals are examined regarding their selectional characteristics as well as their relationships to certain transformation. The paper concludes that despite tenuous evidence surrounding their semantics, the syntax of English modal constructions can be described by a single rule.