Search Results

The distinction between coordination and subordination is claimed to be a structural universal. However, the structural criteria used to distinguish coordination from subordination do not match up across languages. This chapter proposes a functional analysis of complex sentence structure in terms of the Gestalt distinction between figure and ground. Coordination and complements constitute a complex figure (following Wierzbicka), while adverbial clauses and relative clauses constitute a figure-ground structure (following Talmy and Reinhart). Coordination and complementation are linked by grammaticalization, via serial verb constructions, and adverbial clauses and relative clauses are also linked by grammaticalization. Comparative and conditional relations are ambivalent, and expressed crosslinguistically by either complex figure or figure-ground constructions. A conceptual space is presented to account for Cristofaro’s implicational hierarchies for different types of semantic relations between situations and their encoding as balanced (asserted, and more coordinate-like) and deranked (nonasserted, less coordinate-like) complex sentence constructions.

The distinction between coordination and subordination is not clear-cut. The essence of coordination is the mental juxtaposition of structures construed as parallel and co-equal. Subordination has a number of dimensions, including form, the participation of one clause in the relationship profiled by another, and a clause's profile being overridden at the composite structure level. Constituency and profiling are often flexible, depending on factors like the size of clauses and their discourse function. The traditional division into adverbial, relative, and complement clauses is based primarily on how clauses are connected with one another. To the extent that these distinctions hold, they are based on semantic function rather than specific structural configurations. In the case of complements, the classic distinction between control and raising constructions is non-fundamental, the latter being just a special case of the former. Finite and nonfinite complements differ not just in form but in meaning and typical function. Predicates taking finite complements pertain to the epistemic status of propositions; those taking nonfinite complements pertain to the realization of occurrences. Complementation involves multiple conceptualizers and levels of conception. Different conceptualizers apprehend the same proposition each from their own perspective, assessing it with respect to their own conception of reality. Complement-taking predicates refer to different phases of this assessment. Impersonal constructions invoke a conceptualizer and the relevant scope of awareness in generalized fashion, suggesting that anyone would make the assessment under the circumstances.

In most languages of Europe, there is a polysemy pattern to the effect that one and the same marker (e.g., English who) is used to introduce both questions and subordinate clauses. This situation is cross-linguistically unusual, and this chapter attempts to account for why such a polysemy pattern exists and why it is, to a large extent, confined to the languages of Europe or to languages that have been in contact with Indo-European languages.

Supplements (i.e., parentheticals, appositives) provide one of the major sources of evidence for conventional implicatures discussed in this book. After some general discussion of supplements, the focus narrows to supplemental relatives and nominal appositives. The proposed syntax is conservative: supplements are argued to be fully integrated syntactically and morphologically, with their ‘peripheral’ feel attributed to their semantics and their comma intonation. Appositives are shown to manifest all of the properties Grice specified for conventional implicatures, and are analyzed in the terms of Chapter 3. The final major section extends the analysis to a range of sentence-level adverbs.

This chapter is concerned with relative clauses and wh-movement structures across languages, which is examined from the perspective of Minimize Domains, Minimize Forms, and Maximize On-line Processing. The chapter is organized as follows. Section 7.1 introduces background assumptions and details. Section 7.2 examines the Keenan–Comrie Accessibility Hierarchy from the perspective of MiD and MiF. Section 7.3 links the wh-fronting option to basic word order types, specifically to verb position, and accounts for the quantitative regularities in terms of MiD. In Section 7.4 numerous other increasing complexity hierarchies are proposed and tested, derived from MiD and MiF. Section 7.5 examines MaOP effects in wh-questions and relative clauses. Section 7.6 summarizes some apparent grammatical mysteries that make sense from a processing perspective, such as the ‘that-trace’ effect in English and subjacency violations in Japanese.

This chapter discusses the sequence of tense in complement clauses and object relative clauses. It shows how, given very simple semantic principles, the basic semantic phenomena associated with sequence of tense will follow. It then takes up the fundamental question of this chapter — the asymmetry between complement clauses and object relatives — and argues that an answer proposed since 1993, indebted to the proposal of Ogihara (1989), is incorrect. Finally, the chapter provides an alternative answer and remarks some questions that remain open.

The syntactic roles (grammatical relations) of subject and object are semantically irregular but their syntactic behavior is claimed to be syntactically unified, thereby justifying the existence of formal syntactic roles independent of meaning. Subject and object are certainly polysemous categories semantically, but syntactically they are no simpler. Ergativity shows that syntactic roles can vary across languages. Ergativity has been discounted in most syntactic theories by selectively ignoring certain constructions such as case marking and agreement (methodological opportunism). But the variation across and even within languages conforms to a universal implicational hierarchy, the Subject Construction Hierarchy: coordination < purpose clauses < relative clauses < agreement < case marking. If a construction patterns ergatively at some point on the hierarchy, then all constructions to the right also pattern ergatively. Language-specific syntactic roles can be mapped onto a conceptual space whose structure represents the semantic participant roles and the Subject Construction Hierarchy.

This chapter deals with the frequency development of the subjunctive and its competitors, namely indicatives and modal constructions, in adjectival relative clauses in the historical periods Old English (OE), Middle English (ME), and Early Modern English (EModE). Additionally, it discusses the linguistic and extralinguistic parameters influencing their distribution across these periods. The analysis of a corpus comprising nearly 3,000 relative clauses reveals that the subjunctive in adjectival relative clauses died out in the 16th century, that it was best preserved in text category STA containing legislative texts, and that it was favoured in combination with wh-relative markers and in constructions characterized by modal harmony, i.e. in combination with matrix clauses with verbal syntagms expressing root modality.

This chapter addresses the question of how children exploit different grammatical devices in comprehending relative clauses. It shows that comprehension of relative clauses follows different developmental patterns defined by the kind of device that cue a subject or an object relative clause interpretation.

This chapter analyses zhe in archaic Chinese as a determiner which selects a nominal or clausal complement and projects a DP. As a determiner, zhe can further serve as the external binder of a gap in its complement, thereby creating a relative clause. Consequently, this analysis offers s a unified account of the seemingly heterogeneous behaviour of zhe as nominalizer, relativizer, or topic marker.

This chapter examines synchronic cross-linguistic patterns in grammars and language use. It proposes that “variation-defining” universals delimit the scope of possible variation across languages. Examples of such universals include the Greenbergian implicational universals and the parameters in the Government-Binding tradition. It argues that variation-defining universals should be understood in terms of performance principles. It further suggests these same performance principles govern variation of structures within languages, dictating that following a verb, short prepositional phrases should precede long prepositional phrases.

This chapter treats finite and non-finite subordinate clauses, the structure, word order, and other properties of each type. Types of finite clauses are nominal clauses, relative clauses, and adverbial clauses. In connection with non-finite clauses, agreement, control, and raising are discussed.

This chapter examines various constructions types and their relationship to finiteness. Certain constructions, such as ‘indirect questions’ and relatives, are demoted finites: the properties that they display that are associated with finiteness do not save them from being necessarily subordinate. Indicatives, however, can typically be finite in subordinate clauses. Some other constructions are dedicated mood-markers: this is often the case with optative, hortative, and exclamative constructions, which are frequently highly idiomatic. Other constructions still, though often, even typically, occurring as subordinates, may be promoted to serve as less prototypical expressions of declarative or of more marked moods: this is characteristic of the basic non-finite construction that is usually called infinitive. Subjunctives, associated notionally with irrealis, and even counterfactuality, show properties often associated with the expression of declaratives, but are frequently demoted. Descriptions are offered for this range of construction types and their varying relationship with finiteness.

We start with a brief outline of kinds of clauses in Amazonian languages. We then turn to various techniques of putting clauses together into one sentence. Numerous Arawak, Carib and Tupí languages use nominalizations in the function of relative clauses, and in subordinate clauses. A number of languages north of the Amazon, and some in the south have what is known as ‘switch‐reference’: a clause‐combining technique which indicates whether the subject of the main clause is the same as that of a dependent clause, or different from it. This takes us to the issue of ‘pivot’ in clause combining. In a number of languages, including Aguaruna and the Quechua varieties, speech reports have many overtones to do with intention, internal thought and volition.

Chapter 6 introduces the actual relative clause constructions of Greek. The main constructions are: (i) postnominal relative clauses: a noun (traditionally the ‘antecedent’) is in some way modified by a following relative clause (the man whom you see; Socrates, whom you see); (ii) free relative clauses: in traditional terms, a relative clause that either has no expressed antecedent (what she saw) or has an antecedent ‘incorporated’ into the relative clause itself (what things she saw); (iii) semi-free relative clauses: a demonstrative pronoun is the traditional ‘antecedent’ to a following relative clause, and its only obvious function is to point forward to the relative clause (That which has happened may happen again); (iv) relative clauses in relative-correlative sentences: a sentence-initial relative clause is picked up by a demonstrative pronoun in a following main clause (What we don’t know, that we can’t teach). Structures (ii)–(iv) are inherently maximalizing relative clauses.

Chapter 9 is the first of four chapters considering Homer. It is devoted to the circumstances under which Homer chooses an inherently maximalizing relative clause and those under which he chooses a postnominal relative clause. Like the authors of the texts considered in chapter 8, Homer turns out to use postnominal relative clauses and inherently maximalizing relative clauses to say different kinds of things, with very little overlap. Taking advantage of the larger quantity of Homeric material available for study, the chapter offers a detailed and systematic demonstration of this point by considering all the inherently maximalizing relative clauses and all the postnominal relative clauses in Iliad 1–4. Exceptions to the general rule also help to explain it. Where either kind of construction is possible, Homer usually prefers an inherently maximalizing construction because maximalizing meaning (or definiteness) is inherent to the construction, and therefore unambiguous.

This chapter is concerned with the third major modifier category in the noun phrase (NP), localising operators and localising satellites, which relate to locative properties of the referent of the matrix NP. Examples of localising operators in NP are demonstrative pronouns and (in)definite articles, while examples of localising satellites are (localising/identifying) relative clause, possessive modifier, and locative modifier. Each of these constituents typically indicates that the referent of an NP has a place in the world of discourse (or that this place can be inferred). The grammatical expression of the notion location in the NP is discussed by focusing on demonstratives and articles, along with possessive modifiers, adpositional modifiers, and relative clauses.

Chapter 8 surveys the relative clauses of early non-hexameter texts. After summarizing an earlier study by the author on Mycenaean headings it turns to early alphabetic inscriptions; Anaximander; Heraclitus; iambic poetry; monodic and choral lyric; and elegy. Different relative clause constructions predominate in different kinds of text, yet the basic system remains remarkably similar. In this system inherently maximalizing relative clauses and postnominal relative clauses are used to say different kinds of things, with little overlap. Different inherently maximalizing constructions are also not simply interchangeable. The usual inherently maximalizing construction is the free relative clause, but semi-free relative clauses and relative-correlative sentences appear when the syntactic role of a free relative clause would be difficult to work out. The choice between relative-correlative sentences and semi-free relative clauses is determined almost entirely by text type: practical texts and lower literary genres favour relative-correlative sentences while higher genres favour semi-free relative clauses.

Chapter 14 returns to questions of linguistic history and prehistory. It begins with four syntactic phenomena sometimes considered relics from a stage before the development of Greek relative clauses as we know them: anaphoric hos, hē, ho; clause-final placement of postnominal relative clauses; verbless relative clauses; apodotic de. The chapter argues that none of these is actually such a relic, but that Greek relative clause syntax does include at least one genuine syntactic relic of an earlier stage: epic te. The earlier stage is, however, not a stage without relative clauses, but one in which relative clauses were necessarily already present. The chapter also makes the somewhat radical suggestion that one variety of early Greek relative clause may actually be a fairly recent development: the non-restrictive relative clause, which has often been taken to be the earliest Greek relative clause construction.

Outside of the subject, verb, and object in an independent clause, everything else in a sentence is modification, the specifying detail that often carries or constrains an argument. Patterns of modification depend on the types of modifier used, their placement, and the overall amount of modification. This chapter sorts through the arguer's modifying options first according to the grammatical profile of the unit involved. At the highest level are adverb and adjective clauses (subordinate or relative; noun clauses, though not actually modifiers, are also covered here). Next are phrases divisible into three types: those based on verbs (participial and infinitive phrases), those based on nouns (appositives and absolute, resumptive or summative structures), and prepositional phrases. Finally there are single-word modifiers, sometimes grouped in chains according to how dissociable they are from the word modified. Modifiers can be multiplied or embedded, and individual texts can carry heavy modification in proportion to the predication. The point of this chapter is not simply to review these sentence constituents but to see their potential argumentative consequences as in the epithetical style. For example, an appositive offers, sometimes tendentiously, an apparently equivalent term, and constructions like the absolute phrase allow the arguer to promote a feature of a mentioned noun into attention. Infinitive phrases often attribute purposes for actions and participial phrases subordinate one action to another. Depending on where modifiers are placed in relation to the main predication, they often predispose audiences to the arguer's interpretation of evidence or events.