This chapter argues that information structure plays a critical role in explaining the ditransitive construction’s distribution, scope properties, and interaction with long-distance dependency constructions. Specifically, since the recipient argument is a secondary topic and backgrounded in discourse, it tends to have wide scope over the theme argument and also tends to resist extraction when combined with long-distance dependency constructions. An alternative, the to-dative, is favored in these circumstances, since it has a more appropriate information structure.
The passive construction, one of the most scrutinized across varying theoretical and typological perspectives, sometimes gives rise to disagreements among linguists about the membership of particular cases. “Non-promotional” passives are a key example: they lack overt subjects but govern accusative objects and may be categorized as either passives or impersonal actives with null unspecified human subjects. Based on Irish, Polish, Ukrainian, Icelandic, and Pomo data, we argue that disagreements stem from two sources: (a) linguists’ differing theorizations of “passive,” including their tacit commitments to the importance of distinct constructional features for a theory of language; and (b) differing native speaker judgments, which reveal two distinct and persistent analyses of this configuration, arising from the syntactic ambiguity of verb forms with null surface subjects.
This paper examines the relationship among sentence focus, subject position (i.e., preverbal vs. postverbal) and genericity in Italian constructions involving two sub-classes of unaccusative verbs. It is shown that with unaccusatives denoting change of location (e.g., arrivare ‘arrive’), subject position depends on the number of arguments selected by the verb and the nature of the locative argument (i.e., overt vs. implicit); specifically, one-argument unaccusatives categorically disallow postverbal subjects, whereas two-argument unaccusatives require postverbal subjects only if the sentence refers to a specific situation and the locative is implicit. In contrast, two-argument unaccusatives selecting an experiencer/dative and a theme/subject argument (e.g., mancare ‘lack; be lacking’), always require postverbal subjects, independently of whether the sentence denotes a generic or a specific situation.
Despite the idiosyncratic properties of null-complement phenomena observed by Fillmore (1986), Lambrecht and Lemoine (2005), and others, many researchers have pursued large-scope, single-factor explanations of the distributional and interpretive restrictions on null complements – in particular, explanations based on Aktionsart (Rappaport Hovav and Levin 1998) and selectional restrictions (Resnik 1993, 1996). We argue instead for a limited implicational regularity tying the interpretation type of an omitted argument to the frame membership of its predicator. We show that our account is robust, that exceptions can be explained based on independently motivated principles, and that the proposed generalization can be motivated by reference to the discourse status of comparable overt arguments in both lexically and constructionally licensed omissions. Finally, we argue that successful generalizations in the realm of null complementation are likely to be narrow rather than broad in scope.
Construction grammarians analyze grammar in terms of conventional pairings of form and meaning (Fillmore et al. 1988; Goldberg 1995, 2006) that are largely limited to the sentence (Lambrecht 1994, 2004). Recently, construction grammarians have moved beyond sentential boundaries to consider grammatical structures in terms of interaction and discourse (Fried and Östman 2005; Fischer 2010; Fried 2010a, 2010b). Following Fleischman (1990), we argue that interactive frames are key to linking the concerns of grammarians with those of anthropological linguists, sociolinguists, and text linguists who call for richer analyses of the communicative context (Cook-Gumperz and Gumperz 2011; Park and Takanashi 2011). A frames-based approach to grammatical description situates language within a communicative context that includes factors such as the backgrounds of the speakers and hearers and the frames within which each participant places the interaction. Such contextualized descriptions demonstrate how interactive frames and grammatical constructions are implicated in the interpersonal function of language and the dynamics of meaning making.
This paper analyses a paradigm in Russian in which a preposed nominal stranding a numeral can show (paucal) number connectivity, with a gap following the numeral, or can appear in a non-agreeing (plural) form: (1) theater-paucal/plural, there were three. paucal__ Numerous syntactic diagnostics confirm that, when there is number connectivity, the nominal has been fronted via A′-movement, creating a syntactic A′-chain dependency. In the absence of connectivity, the construction involves a hanging topic related via discourse mechanisms to a base-generated null pronoun. These two constructions constitute a minimal pair and can be considered a counterpart to the better-known left edge topics observed in Romance or Germanic languages.
Final compression in French is examined from a theoretical perspective and illustrated with experimental data. In French, only final prosodic domains that are at least the size of a prosodic phrase (Φ-phrase) may be subject to final compression. Prosodic units that are smaller than Φ-phrases, such as prosodic words (ω-words) or syllables, cannot be compressed. Therefore, final compression is a phrasal phenomenon. This conclusion is confirmed in a comparison between sequences consisting of a noun + adjective and verb + argument on one hand, and verb + adjunct on the other. Syntactic expressions of the first kind are integrated into single Φ-phrases, and final compression does not take place. Syntactic sequences of the verb + adjunct variety, however, form separate Φ-phrases, and final compression may occur.
WH-Interrogatives in spoken French are exceptional due both to the diversity of syntactic forms available to express the same propositional content (Coveney 1996) as well as the parameters that influence their use. The causal interrogative adverb pourquoi ‘why’ has a notably deficient paradigm (Korzen 1985), as is seen in its incompatibility with stylistic inversion and in situ structures. In this chapter, all instantiations of pourquoi in a corpus of casual conversations, whether interrogative or not, are first categorized according to discourse function(s) and then mapped to structures. This usage-driven, discourse approach successfully narrows the number of observed forms per function and also further distinguishes them by integrating discourse markers and negation.
Causal conjunctions have been studied by linguists (e.g. Groupe Lambda-l 1975; Debaisieux 1994; Lambrecht, Bordeaux, and Reichle 2006) and by psycholinguists and cognitive psychologists interested in language processing (e.g. Millis and Just 1994; Millis, Golding, and Barker 1995). These approaches result in different descriptions of the French word parce que ‘because.’ It is predicted to be unmarked in its information structure, but marked in its processing due to the increased load it places on processing resources and working memory. This chapter compares these two possible moderating factors on the use of parce que by analyzing corpus data from native speakers and second language learners of French, and argues that the information structure of the conjunction plays a greater role than its processability in determining its use in spoken French.
Although the French c’est-cleft and the English it-cleft appear at first glance to share similar syntactic and pragmatic properties (they are both what Lambrecht 1994 calls “argument focus” constructions), their use in discourse is not always the same. One finds a number of situations in which the c’est-cleft is required but the it-cleft is pragmatically odd. The reason for this discrepancy has to do with French prosodic restrictions that do not exist in English, thus creating a motivation for the cleft in French that is not found in English. In addition, various c’est-cleft types and c’est-cleft “lookalikes” in French correspond to different types of constructions in English, demonstrating the importance of analyzing naturally occurring discourse to determine pragmatic well-formedness.
A number of studies of Left Dislocation (LD) in spoken French within the Interactional Linguistics (IL) framework (de Fornel 1988; Pekarek Doehler 2001; Chevalier 2011b) have been critical of the information-structure analyses of this construction as set forth in Lambrecht (1981, 1994) and Barnes (1985). This discussion attempts to clarify the original information-structure analysis, arguing that the pragmatic definition of LD should be limited to the explicit marking of the sentence-topic and its associated comment. This topic-comment configuration is compatible with a large variety of particular functions with respect to the larger discourse and to speakers’ interactional purposes. Explanatorily useful IL analyses are those that make clear the connection between the topic-comment configuration and the proposed interactional function.
This chapter argues that information structure plays a critical role in explaining the ditransitive construction’s distribution, scope properties, and interaction with long-distance dependency constructions. Specifically, since the recipient argument is a secondary topic and backgrounded in discourse, it tends to have wide scope over the theme argument and also tends to resist extraction when combined with long-distance dependency constructions. An alternative, the to-dative, is favored in these circumstances, since it has a more appropriate information structure.
The passive construction, one of the most scrutinized across varying theoretical and typological perspectives, sometimes gives rise to disagreements among linguists about the membership of particular cases. “Non-promotional” passives are a key example: they lack overt subjects but govern accusative objects and may be categorized as either passives or impersonal actives with null unspecified human subjects. Based on Irish, Polish, Ukrainian, Icelandic, and Pomo data, we argue that disagreements stem from two sources: (a) linguists’ differing theorizations of “passive,” including their tacit commitments to the importance of distinct constructional features for a theory of language; and (b) differing native speaker judgments, which reveal two distinct and persistent analyses of this configuration, arising from the syntactic ambiguity of verb forms with null surface subjects.
This paper examines the relationship among sentence focus, subject position (i.e., preverbal vs. postverbal) and genericity in Italian constructions involving two sub-classes of unaccusative verbs. It is shown that with unaccusatives denoting change of location (e.g., arrivare ‘arrive’), subject position depends on the number of arguments selected by the verb and the nature of the locative argument (i.e., overt vs. implicit); specifically, one-argument unaccusatives categorically disallow postverbal subjects, whereas two-argument unaccusatives require postverbal subjects only if the sentence refers to a specific situation and the locative is implicit. In contrast, two-argument unaccusatives selecting an experiencer/dative and a theme/subject argument (e.g., mancare ‘lack; be lacking’), always require postverbal subjects, independently of whether the sentence denotes a generic or a specific situation.
Despite the idiosyncratic properties of null-complement phenomena observed by Fillmore (1986), Lambrecht and Lemoine (2005), and others, many researchers have pursued large-scope, single-factor explanations of the distributional and interpretive restrictions on null complements – in particular, explanations based on Aktionsart (Rappaport Hovav and Levin 1998) and selectional restrictions (Resnik 1993, 1996). We argue instead for a limited implicational regularity tying the interpretation type of an omitted argument to the frame membership of its predicator. We show that our account is robust, that exceptions can be explained based on independently motivated principles, and that the proposed generalization can be motivated by reference to the discourse status of comparable overt arguments in both lexically and constructionally licensed omissions. Finally, we argue that successful generalizations in the realm of null complementation are likely to be narrow rather than broad in scope.
Construction grammarians analyze grammar in terms of conventional pairings of form and meaning (Fillmore et al. 1988; Goldberg 1995, 2006) that are largely limited to the sentence (Lambrecht 1994, 2004). Recently, construction grammarians have moved beyond sentential boundaries to consider grammatical structures in terms of interaction and discourse (Fried and Östman 2005; Fischer 2010; Fried 2010a, 2010b). Following Fleischman (1990), we argue that interactive frames are key to linking the concerns of grammarians with those of anthropological linguists, sociolinguists, and text linguists who call for richer analyses of the communicative context (Cook-Gumperz and Gumperz 2011; Park and Takanashi 2011). A frames-based approach to grammatical description situates language within a communicative context that includes factors such as the backgrounds of the speakers and hearers and the frames within which each participant places the interaction. Such contextualized descriptions demonstrate how interactive frames and grammatical constructions are implicated in the interpersonal function of language and the dynamics of meaning making.
This paper analyses a paradigm in Russian in which a preposed nominal stranding a numeral can show (paucal) number connectivity, with a gap following the numeral, or can appear in a non-agreeing (plural) form: (1) theater-paucal/plural, there were three. paucal__ Numerous syntactic diagnostics confirm that, when there is number connectivity, the nominal has been fronted via A′-movement, creating a syntactic A′-chain dependency. In the absence of connectivity, the construction involves a hanging topic related via discourse mechanisms to a base-generated null pronoun. These two constructions constitute a minimal pair and can be considered a counterpart to the better-known left edge topics observed in Romance or Germanic languages.
Final compression in French is examined from a theoretical perspective and illustrated with experimental data. In French, only final prosodic domains that are at least the size of a prosodic phrase (Φ-phrase) may be subject to final compression. Prosodic units that are smaller than Φ-phrases, such as prosodic words (ω-words) or syllables, cannot be compressed. Therefore, final compression is a phrasal phenomenon. This conclusion is confirmed in a comparison between sequences consisting of a noun + adjective and verb + argument on one hand, and verb + adjunct on the other. Syntactic expressions of the first kind are integrated into single Φ-phrases, and final compression does not take place. Syntactic sequences of the verb + adjunct variety, however, form separate Φ-phrases, and final compression may occur.
WH-Interrogatives in spoken French are exceptional due both to the diversity of syntactic forms available to express the same propositional content (Coveney 1996) as well as the parameters that influence their use. The causal interrogative adverb pourquoi ‘why’ has a notably deficient paradigm (Korzen 1985), as is seen in its incompatibility with stylistic inversion and in situ structures. In this chapter, all instantiations of pourquoi in a corpus of casual conversations, whether interrogative or not, are first categorized according to discourse function(s) and then mapped to structures. This usage-driven, discourse approach successfully narrows the number of observed forms per function and also further distinguishes them by integrating discourse markers and negation.
Causal conjunctions have been studied by linguists (e.g. Groupe Lambda-l 1975; Debaisieux 1994; Lambrecht, Bordeaux, and Reichle 2006) and by psycholinguists and cognitive psychologists interested in language processing (e.g. Millis and Just 1994; Millis, Golding, and Barker 1995). These approaches result in different descriptions of the French word parce que ‘because.’ It is predicted to be unmarked in its information structure, but marked in its processing due to the increased load it places on processing resources and working memory. This chapter compares these two possible moderating factors on the use of parce que by analyzing corpus data from native speakers and second language learners of French, and argues that the information structure of the conjunction plays a greater role than its processability in determining its use in spoken French.
Although the French c’est-cleft and the English it-cleft appear at first glance to share similar syntactic and pragmatic properties (they are both what Lambrecht 1994 calls “argument focus” constructions), their use in discourse is not always the same. One finds a number of situations in which the c’est-cleft is required but the it-cleft is pragmatically odd. The reason for this discrepancy has to do with French prosodic restrictions that do not exist in English, thus creating a motivation for the cleft in French that is not found in English. In addition, various c’est-cleft types and c’est-cleft “lookalikes” in French correspond to different types of constructions in English, demonstrating the importance of analyzing naturally occurring discourse to determine pragmatic well-formedness.
A number of studies of Left Dislocation (LD) in spoken French within the Interactional Linguistics (IL) framework (de Fornel 1988; Pekarek Doehler 2001; Chevalier 2011b) have been critical of the information-structure analyses of this construction as set forth in Lambrecht (1981, 1994) and Barnes (1985). This discussion attempts to clarify the original information-structure analysis, arguing that the pragmatic definition of LD should be limited to the explicit marking of the sentence-topic and its associated comment. This topic-comment configuration is compatible with a large variety of particular functions with respect to the larger discourse and to speakers’ interactional purposes. Explanatorily useful IL analyses are those that make clear the connection between the topic-comment configuration and the proposed interactional function.