Search Results

This chapter argues that the occurrence of the to-infinitive after verbs expressing meanings like ‘intend’ or ‘want’, or the negative counterparts ‘fear’ or ‘avoid’ represents an innovation. Unlike the verbs discussed in the previous chapter, these verbs never occurred with a to-prepositional phrase, so that the presence of the to-infinitive requires a different explanation. The best predictor for the distribution of a to-infinitive after verbs of this group in Old English is that of the subjunctive þæt-clause. As subjunctive þæt-clauses and to-infinitives had occurred side by side as expressions of the adjunct of purpose and of the GOAL-argument, this may have led them to be analysed as each other’s equivalents, as if the to-infinitive had come to be regarded as a non-finite subjunctive clause. The chapter concludes with a discussion of the only environment in which bare infinitive and to-infinitive occur side by side: verbs with the meaning ‘begin’.

This chapter discusses the rise of to-infinitival Exceptional Case-Marking (ECM) constructions as in, He believes the results to be unscientific, in late Middle English. It argues that a distinction should be made between the construction after verbs like want (I want you to do it), which appears to arise out of a reanalysis of the to-infinitive as THEME with verbs of commanding and permitting, and the construction after verbs like believe (‘the verbs of thinking and declaring’). It is argued that emergence of ECMs with believe-verbs is connected with changes in information structure causes by the loss of verb-second. The register restriction on this type of ECM, as well as the restriction on embedded subjects (witness *They alleged the results to be unscientific), appears to suggest that the construction is still outside the core grammar of English and requires additional routines (viruses) that are acquired after the core grammar is in place.

This chapter identifies a second set of ditransitive verbs: the verbs of commanding and permitting. Unlike the verbs of persuading and urging discussed in the Chapter 3, which have a thematic structure of AGENT, THEME, and GOAL, the verbs of commanding and permitting have a thematic structure of AGENT, RECIPIENT, and THEME. The different roles are clear from the different cases that express them: THEMES are expressed by accusative NPs, and RECIPIENTS by dative NPs. Verbs of commanding and permitting have two ‘variants’, a three-place and a two-place one. This is not an accidental, idiosyncratic lexical property of these verbs, but a structural feature connected with the basic meanings of this group. It is precisely this structural feature of the verbs of commanding and permitting that is responsible for the appearance of the first to-infinitival ECMs in Middle English.

This chapter presents an extensive case study on diffusional change in the system of complementation, focusing on the most dramatic recent change of this kind in the history of English: the emergence of gerunds as verb complements. It shows that diffusion proceeds as a result of new regularities being inferred from existing usage. Generally, a given regularity is worked out in language use in a relatively short time, but no sooner is a regularity applied to a group of environments, than it can trigger some new regularity that again gives rise to new uses (analogical chain). It is also true that by the end of the process of diffusion, environments have to meet fewer conditions to qualify as hosts to the spreading construction (analogical snowball). This means that diffusion becomes easier as it proceeds, even though its progress does not necessarily become quicker. From the historical developments, it is clear that the use of a complement type is not determined by a single principle but is the outcome of a complex interaction between various factors, including different kinds of analogy and blocking.

This chapter argues that the to-infinitive cannot be analysed as a prepositional phrase in Old English. Its distribution is no longer that of a to-PP, its dative inflection has fossilized and does not behave like the dative inflection found on true N-heads, and it takes accusative objects rather than the genitive objects that would be expected if it was still nominal. Evidence from relative clauses, the position of the to-infinitival object, and the emergence of an overt complementizer — for — in early Middle English all argue for clausal rather than phrasal status. The most telling piece of evidence is the fact that the to-infinitive has entered into competition with the finite subjunctive clause and can be seen to gradually oust it in a number of environments. All the available evidence suggests that the to-infinitive was being analyzed as a non-finite subjunctive already in OE.

This chapter argues that some purpose adjuncts were reinterpreted as GOAL-arguments of conatives and verbs of persuading and urging. This involved a semantic change of the higher verb. Verbs meaning ‘exert oneself’ came to mean ‘exert oneself in order to’, hence ‘strive, try’, i.e., a conative, with the following to-infinitive, originally there as purpose adjunct, becoming a GOAL-argument of that verb. The GOAL-argument of verbs of persuading and urging underwent a similar change. That these to-infinitives are indeed arguments in Old English and no longer adjuncts is shown by syntactic tests. That the to-infinitive after these verbs started out as a purpose adjunct is clear from the fact that the other two expressions that are available to encode the GOAL-argument are the to-prepositional phrase and the subjunctive þæt-clause — both, like the to-infinitive, regularly used to encode purpose adjuncts.

This chapter addresses the LEX-headed inflectional pattern. In this construction, the lexical verb bears all the obligatory inflectional categories, and the auxiliary verb may appear in an uninflecting form, expressing only the category that it functions to encode. This construction is noteworthy insofar as the phrasal or syntactic head is generally the auxiliary verb, but the inflectional head is the lexical verb. In many descriptions, the auxiliary in the LEX-headed pattern is analysed as an uninflecting particle. However, when considering the functional semantics of the element and the fact that these historically originate from verbal elements, the semantics of which are in accord with the semantic developments typical of the process of auxiliation, it seems clear that these in certain instances should rather be considered as reflecting an AVC of the LEX-headed inflectional pattern. Lexical verbs, although bearing obligatory inflectional categories for the clause (other than those embodied or encoded by the auxiliary itself), may also bear an overt marker of dependency, further underscoring their presence in a grammaticalized AVC, albeit one in which the auxiliary itself bears no inflection. The LEX-headed pattern here also includes situations in which there is a dummy third singular ‘clausal’ subject marker found with the auxiliary, sometimes reflecting the construction’s origin in a reanalysed biclausal verb plus complement structure or in an ambient serialization construction.

This chapter focuses on a third type of ditransitive: commissives, with meanings like ‘promise’. The commissives are interesting in that finite clauses are still their preferred complement in Present-day English, unlike the other subject-controlled verbs where finite clauses have been ousted completely by the to-infinitive. Closer investigation suggests that this survival is due to the fact that coreference of the subject of the subclause and an argument of the main clause is less frequent than with the verbs discussed in the previous chapters, i.e., the embedded subject is not ‘controlled’. Secondly, the ratio of modals versus simple subjunctive form is far higher in Old English after the commissives than after other ‘control’ verbs, which suggests that there is a need to distinguish between these types by using different modals for different types, which explains the continued survival of finite clauses.

This chapter analyzes the Negative Movement (NM) pattern in Norwegian, where, as in other Scandinavian languages, objects containing a negative quantifier (e.g., no books) must appear to the left of the verb, albeit with a slightly archaic or marginal flavour. The chapter is organized as follows. Section 15.2 discusses Complement-Verb_main (XV) and Verb_main-Complement (VX) order, and gives a presentation of word order patterns in Early Norwegian that resemble scrambling, quantifier movement (QM), and NM in modern Germanic languages. These three types of XV word order are described in terms of semantically-driven movement regulated by information structure. QM and NM in nineteenth-century and contemporary Modern Norwegian is also discussed. Section 15.3 discusses NM as a remnant of these once-productive movement operations, framing the discussion in terms of Virus Theory. It examines four characteristics shared by grammatical viruses and point out the similarities between NM and other prestige constructions in Modern Standard English. The chapter concludes with a discussion of Virus Theory as a tool for the description and analysis of data on syntactic variation and change.

A variadic function approach to weather reports and similar constructions is designed to save the spirit, if not the letter, of the ‘standard view’: although the meteorological predicate does not project a locative argument, there is a systematic way of construing the predicate as locatively relational (as having an indefinite argument), and so weather reports pattern with complement deletion verbs. I shall raise some objections to Recanati’s suggestion by showing that putative variadic functions are severely constrained by syntactic conditions, and so the kind of indefinite interpretations variadic functions allow cannot be freely available as an extra-linguistic pragmatic mechanism to coerce the desired reading. If that is right, the ‘standard view’ construals are simply not available, even with the permissibility of variadic functions. The general conclusion will be that the variadic function approach is unduly concessive to the spirit of the ‘standard view’.