This introductory chapter provides background on the phenomena of Pseudo-Coordination (PseCo) and Multiple Agreement Constructions (MACs) with the aim of familiarizing readers with major trends in previous research on these varied phenomena. Common structural and functional properties used to identify PseCo and MACs are described, along with a detailed discussion of the features that make crucial differences within each phenomenon in individual languages and cross-linguistically. We also observe interesting similarities between the two phenomena and across related and unrelated languages. We maintain a pre-theoretical view here that is compatible with the different approaches represented in the volume.
Italo-Romance varieties present at least three types of constructions that cluster together two verbs displaying double tense and double subject agreement and are taken as Pseudo-Coordinations (PseCos) or Multiple Agreement Constructions (MACs). In this paper, we follow Cardinaletti and Giusti’s (1998, 2001, 2003, 2020) hypotheses and claim that unification between the PseCos with a and the MACs with mu/mi/ma or ku in Southern Italian dialects is not viable. We adopt a diagnostic tool, which we call a protocol, that clusters the predictions of theory-driven analyses and apply it to the ‘take and’ construction, which is widespread across dialects and productive in Italian. In doing so, we discuss unobserved facts arising in the well-studied dialectal structures and make fine-grained observations about the less studied ‘take and’ PseCo in Italian.
In South Italian varieties of Apulia, Calabria and Sicily a restricted number of control/raising verbs, including stay/be, go, come and want embed finite complements, either bare or introduced by a. These are not necessarily languages with so-called subjunctive particles; in any event, the latter have a different form. Under monoclausal analyses, verbs like stay/be, go etc. are functional heads embedding an inflected predicate. Here we adopt a biclausal analysis under which embedding under stay/be, go etc. is a normal clausal embedding. We argue that the biclausal analysis is not only feasible, but also advantageous, from a morphosyntactic point of view. Focusing on the progressive, we also consider whether the bisentential analysis is compatible with semantic interpretation and how it fares in a typological perspective.
This paper discusses the paradigmatic configuration (or ‘morphome’; Aronoff 1994) that Pseudo-Coordination (V1[TAM.Agr] a V2[TAM.Agr], as in Jivu a ffici la spisa. ‘I went and did the shopping.’) displays in the preterite indicative in Deliano: i.e. the ‘W-Pattern’ (Di Caro 2019a; Di Caro and Giusti 2018). In the first part, the suppletive nature of the preterite paradigm of the V2s licensing this construction is discussed. These V2s all feature perfective roots (i.e. PYTA roots; Maiden 2018b), such as fici ‘I made/did’ and dissi ‘I said’, which are the ones allowed in the construction, and imperfective roots, such as facisti ‘you (sg.) made/did’ and dicisti ‘you (sg.) said’. In the second part, new data from a grammaticality judgment-based study on Pseudo-Coordination in Deliano are discussed, with reference to the emergence of the W-Pattern in a specific paradigm. The results clearly show that this morphome is consistently present throughout the sample (11–80 y.o. participants, N = 140) and has a “psychological reality” (cf. Maiden 2018b: 1–10), in the sense that it does not seem to be affected by variables such as age or gender, or to be subject to ordering effects.
In this paper, I discuss a periphrastic construction involving the verb go in Sicilian that is used to express surprise and unexpectedness with respect to a past event. I show that the special meaning and function of this structure is best accounted for by postulating that in this construction the verb go is now a functional verb associated with a mirative conventional implicature. In this use, the construction is grammatically in the present tense, but is used within a narrative context to foreground an unexpected or surprising event that happened in the past. To account for the present-tense morphology, I propose that the conversational backgrounds – and in particular the ordering source defining the set of expectations of the conversation participants – can be indexed to the present time. I finally explore the hypothesis that the mirative use of this construction can shed light on the development of the Catalan go-past.
This paper presents a preliminary classification of the verbal structure (a) lua și X (‘(to) take and X’) in Romanian, showing that it represents a special case of pseudocoordination. The structure behaves differently from both coordination structures and other pseudocoordination structures with respect to the tests proposed by de Vos (2005) and Ross (2013) (e.g. the Coordinate Structure Constraint, coordinator substitution, semantic bleaching, VP-deletion, etc.), as shown by an exploratory acceptability judgment task with 52 native speakers of Romanian testing for 16 structural properties. The results suggest that the existing classification of pseudocoordination structures should be revisited in order to accommodate Romanian ‘take’ as an additional type.
In this paper, we offer some comments about the syntax of pseudo-coordination in colloquial registers of Brazilian Portuguese and Polish. Focusing on V1-take (and) pseudo-coordination, we suggest that, in both of them, V1-take (and) belongs to the expressive realm of language and we analyze V1-take (and) as an appositive element adjoined to vP in the extended projection of V2. In addition to the meaning of the structure, evidence for the expressive nature of V1-take (and) comes from the fact that it can be ignored for ellipsis purposes in contexts such as verb-echo answers, polarity contrast, verb-doubling and VP-topicalization. Evidence for the positioning of V1-take (and) at the vP edge is provided by distributional patterns, including the placement of adverbs and sentential negation with respect to V1-take (and) and V2. We propose that two minimally different structures are available for pseudo-coordination, depending on whether a coordinator accompanies V1.
This chapter investigates the use of the verb jít (‘go’) in two construction types in the Czech language; they have in common a binary coordinative structure where the verb jít is coordinated with any other verb with the coordinator a (‘and’). These constructions are prototypical coordination (ProCo) and pseudo-coordination (PseCo). The main claim is that even though these two types share the same surface structure jít-a-V2, they represent distinct phenomena.
The resolution criteria are based on a two-part analysis. First, PseCo is analysed as a complex predicate. This analysis immediately accounts for a number of properties of PseCo compared to ProCo. Second, the formal features of the construction are linked to its semantic structure: PseCo expresses aktionsart via coordination over sub-stages of events.
I argue that ProCo is a biclausal structure coordinating two separate events, while PseCo coordinates two verbs into one complex predicate and the coordinator a (‘and’) serves for a coordination of sub-stages of this combined event. It appears that the first verb expresses the preparatory phase for the activity denoted by the second verb. The pseudo-coordinative verb in the first conjunct lexicalises a manner component in the internal event structure. The verb ‘go’ is desemanticized and instead of the meaning of physical motion expresses dynamic aspects of the second event.
This research is based on 1611 examples from the Czech National Corpus, subcorpus SYN2005, from which 923 examples are analysed as ProCo and 668 as PseCo.
This study provides a discussion of the development of subjective meaning associated with the motion-verb pseudocoordination gå och V ‘go/walk and V’ and the posture-verb pseudocoordination sitta och V ‘sit and V’, using historical and present-day linguistic data. It is claimed that an interpretation in terms of item-based analogy and entrenchment of frequent meaning clusters is the most plausible analysis for the development of subjective (and pejorative) meaning associated with gå och V. The study of sitta och V is preliminary, but the results indicate that the subjective meaning of this construction is less entrenched that that of the gå och V construction and that the subjective overtone of subjectivity may be a result of the combination of the social/cultural meaning of the posture and certain intrinsically pejorative verbs, together with certain locatives.
The phenomena of pseudo-coordination and, to a lesser degree, pseudo-subordination have been recognized to play an important role in Semitics linguistics, notably in the realm of converb (gerund) and serial verb constructions, albeit under different scholarly labeling. As the distinction between coordinated (paratactic) and subordinated (hypotactic) structures is often blurred in this context, this paper additionally refers to the concept of “para-hypotaxis”. As will be shown, this choice of terminology, further elaborating on a basic model proposed by Yuasa and Sadock (2002: 91), is useful to describe an analyze a number of phenomena in the domain of complex predicates, in both ancient and modern Semitic languages.
This paper addresses Japanese pseudo-coordination containing an existential verb by focusing on the two types of existential verbs, one taking an animate subject and the other taking an inanimate subject. Though both can form a pseudo-coordination expression, one is two-way ambiguous in aspectual interpretation, and the other is not. Moreover, dialectal variation is observed. This paper attempts to explain the difference between the two interpretations of Japanese pseudo-coordination and assigns a different structure to each interpretation. More specifically, in one interpretation, existential verbs are truly existential verbs of a lexical category, and in the other interpretation, they are functional categories. Based on this analysis, dialectal variation is also explained based on the notion of externalization.
In Turkish nominal phrases and clauses where a coordinated possessor or subject is related to agreement morphology on the possessee or the nominalized predicate, respectively, agreement is realized either in full or in partial expression. The choice between the two is determined in certain cases by syntactic phenomena, suggesting that agreement must figure in syntax. However, partial possessor agreement appears to result from a relationship between the possessee or nominalized predicate, and the last conjunct only, hinting that it is also subject to linear locality conditions. We conclude that the agreement phenomenon in languages results from conditions that apply in syntax proper and from conditions that apply in a post-syntactic component separately, which can alter the output of syntax proper where applicable.
There have been very few attempts to date to provide an explicit semantics/pragmatics for Pseudo-Coordination (PseCo) expressions. This chapter is an attempt to fill that gap, zooming in on the ‘go-(and-)get’-type. To do so, I first provide a syntactic account of PseCo, which derives from a standard coordination structure (which I label Junction), onto and from which a compositional semantic account is derived. The signature pragmatic properties of PseCo of negative-emotive factivity are also derived. Aside from providing the first systematic and cross-modular analysis of PseCo, the chapter also provides a number of new diagnostics for identifying and classifying PseCo expressions which may be useful in future work on the topic.
Verbal pseudocoordination (as in English go and get) is often seen as an idiosyncratic phenomenon described in exceptional terms. This paper establishes the typological context to explain key properties of pseudocoordination, integrated into a more general typology of multi-verb constructions. At the same time, principled motivations are given for the arbitrary list of traditional properties attributed to Serial Verb Constructions (SVCs). The broader category of Multi-Verb Predicates (MVPs) is proposed as any monoclausal multi-verb construction with two verbs forming a complex predicate. Subtypes of MVPs are distinguished by their form: pseudocoordination with a linker ‘and’, while SVCs have no linker. Structural properties of MVPs, such as shared inflectional features on each verb, are readily explained as due to monoclausality.
This introductory chapter provides background on the phenomena of Pseudo-Coordination (PseCo) and Multiple Agreement Constructions (MACs) with the aim of familiarizing readers with major trends in previous research on these varied phenomena. Common structural and functional properties used to identify PseCo and MACs are described, along with a detailed discussion of the features that make crucial differences within each phenomenon in individual languages and cross-linguistically. We also observe interesting similarities between the two phenomena and across related and unrelated languages. We maintain a pre-theoretical view here that is compatible with the different approaches represented in the volume.
Italo-Romance varieties present at least three types of constructions that cluster together two verbs displaying double tense and double subject agreement and are taken as Pseudo-Coordinations (PseCos) or Multiple Agreement Constructions (MACs). In this paper, we follow Cardinaletti and Giusti’s (1998, 2001, 2003, 2020) hypotheses and claim that unification between the PseCos with a and the MACs with mu/mi/ma or ku in Southern Italian dialects is not viable. We adopt a diagnostic tool, which we call a protocol, that clusters the predictions of theory-driven analyses and apply it to the ‘take and’ construction, which is widespread across dialects and productive in Italian. In doing so, we discuss unobserved facts arising in the well-studied dialectal structures and make fine-grained observations about the less studied ‘take and’ PseCo in Italian.
In South Italian varieties of Apulia, Calabria and Sicily a restricted number of control/raising verbs, including stay/be, go, come and want embed finite complements, either bare or introduced by a. These are not necessarily languages with so-called subjunctive particles; in any event, the latter have a different form. Under monoclausal analyses, verbs like stay/be, go etc. are functional heads embedding an inflected predicate. Here we adopt a biclausal analysis under which embedding under stay/be, go etc. is a normal clausal embedding. We argue that the biclausal analysis is not only feasible, but also advantageous, from a morphosyntactic point of view. Focusing on the progressive, we also consider whether the bisentential analysis is compatible with semantic interpretation and how it fares in a typological perspective.
This paper discusses the paradigmatic configuration (or ‘morphome’; Aronoff 1994) that Pseudo-Coordination (V1[TAM.Agr] a V2[TAM.Agr], as in Jivu a ffici la spisa. ‘I went and did the shopping.’) displays in the preterite indicative in Deliano: i.e. the ‘W-Pattern’ (Di Caro 2019a; Di Caro and Giusti 2018). In the first part, the suppletive nature of the preterite paradigm of the V2s licensing this construction is discussed. These V2s all feature perfective roots (i.e. PYTA roots; Maiden 2018b), such as fici ‘I made/did’ and dissi ‘I said’, which are the ones allowed in the construction, and imperfective roots, such as facisti ‘you (sg.) made/did’ and dicisti ‘you (sg.) said’. In the second part, new data from a grammaticality judgment-based study on Pseudo-Coordination in Deliano are discussed, with reference to the emergence of the W-Pattern in a specific paradigm. The results clearly show that this morphome is consistently present throughout the sample (11–80 y.o. participants, N = 140) and has a “psychological reality” (cf. Maiden 2018b: 1–10), in the sense that it does not seem to be affected by variables such as age or gender, or to be subject to ordering effects.
In this paper, I discuss a periphrastic construction involving the verb go in Sicilian that is used to express surprise and unexpectedness with respect to a past event. I show that the special meaning and function of this structure is best accounted for by postulating that in this construction the verb go is now a functional verb associated with a mirative conventional implicature. In this use, the construction is grammatically in the present tense, but is used within a narrative context to foreground an unexpected or surprising event that happened in the past. To account for the present-tense morphology, I propose that the conversational backgrounds – and in particular the ordering source defining the set of expectations of the conversation participants – can be indexed to the present time. I finally explore the hypothesis that the mirative use of this construction can shed light on the development of the Catalan go-past.
This paper presents a preliminary classification of the verbal structure (a) lua și X (‘(to) take and X’) in Romanian, showing that it represents a special case of pseudocoordination. The structure behaves differently from both coordination structures and other pseudocoordination structures with respect to the tests proposed by de Vos (2005) and Ross (2013) (e.g. the Coordinate Structure Constraint, coordinator substitution, semantic bleaching, VP-deletion, etc.), as shown by an exploratory acceptability judgment task with 52 native speakers of Romanian testing for 16 structural properties. The results suggest that the existing classification of pseudocoordination structures should be revisited in order to accommodate Romanian ‘take’ as an additional type.
In this paper, we offer some comments about the syntax of pseudo-coordination in colloquial registers of Brazilian Portuguese and Polish. Focusing on V1-take (and) pseudo-coordination, we suggest that, in both of them, V1-take (and) belongs to the expressive realm of language and we analyze V1-take (and) as an appositive element adjoined to vP in the extended projection of V2. In addition to the meaning of the structure, evidence for the expressive nature of V1-take (and) comes from the fact that it can be ignored for ellipsis purposes in contexts such as verb-echo answers, polarity contrast, verb-doubling and VP-topicalization. Evidence for the positioning of V1-take (and) at the vP edge is provided by distributional patterns, including the placement of adverbs and sentential negation with respect to V1-take (and) and V2. We propose that two minimally different structures are available for pseudo-coordination, depending on whether a coordinator accompanies V1.
This chapter investigates the use of the verb jít (‘go’) in two construction types in the Czech language; they have in common a binary coordinative structure where the verb jít is coordinated with any other verb with the coordinator a (‘and’). These constructions are prototypical coordination (ProCo) and pseudo-coordination (PseCo). The main claim is that even though these two types share the same surface structure jít-a-V2, they represent distinct phenomena.
The resolution criteria are based on a two-part analysis. First, PseCo is analysed as a complex predicate. This analysis immediately accounts for a number of properties of PseCo compared to ProCo. Second, the formal features of the construction are linked to its semantic structure: PseCo expresses aktionsart via coordination over sub-stages of events.
I argue that ProCo is a biclausal structure coordinating two separate events, while PseCo coordinates two verbs into one complex predicate and the coordinator a (‘and’) serves for a coordination of sub-stages of this combined event. It appears that the first verb expresses the preparatory phase for the activity denoted by the second verb. The pseudo-coordinative verb in the first conjunct lexicalises a manner component in the internal event structure. The verb ‘go’ is desemanticized and instead of the meaning of physical motion expresses dynamic aspects of the second event.
This research is based on 1611 examples from the Czech National Corpus, subcorpus SYN2005, from which 923 examples are analysed as ProCo and 668 as PseCo.
This study provides a discussion of the development of subjective meaning associated with the motion-verb pseudocoordination gå och V ‘go/walk and V’ and the posture-verb pseudocoordination sitta och V ‘sit and V’, using historical and present-day linguistic data. It is claimed that an interpretation in terms of item-based analogy and entrenchment of frequent meaning clusters is the most plausible analysis for the development of subjective (and pejorative) meaning associated with gå och V. The study of sitta och V is preliminary, but the results indicate that the subjective meaning of this construction is less entrenched that that of the gå och V construction and that the subjective overtone of subjectivity may be a result of the combination of the social/cultural meaning of the posture and certain intrinsically pejorative verbs, together with certain locatives.
The phenomena of pseudo-coordination and, to a lesser degree, pseudo-subordination have been recognized to play an important role in Semitics linguistics, notably in the realm of converb (gerund) and serial verb constructions, albeit under different scholarly labeling. As the distinction between coordinated (paratactic) and subordinated (hypotactic) structures is often blurred in this context, this paper additionally refers to the concept of “para-hypotaxis”. As will be shown, this choice of terminology, further elaborating on a basic model proposed by Yuasa and Sadock (2002: 91), is useful to describe an analyze a number of phenomena in the domain of complex predicates, in both ancient and modern Semitic languages.
This paper addresses Japanese pseudo-coordination containing an existential verb by focusing on the two types of existential verbs, one taking an animate subject and the other taking an inanimate subject. Though both can form a pseudo-coordination expression, one is two-way ambiguous in aspectual interpretation, and the other is not. Moreover, dialectal variation is observed. This paper attempts to explain the difference between the two interpretations of Japanese pseudo-coordination and assigns a different structure to each interpretation. More specifically, in one interpretation, existential verbs are truly existential verbs of a lexical category, and in the other interpretation, they are functional categories. Based on this analysis, dialectal variation is also explained based on the notion of externalization.
In Turkish nominal phrases and clauses where a coordinated possessor or subject is related to agreement morphology on the possessee or the nominalized predicate, respectively, agreement is realized either in full or in partial expression. The choice between the two is determined in certain cases by syntactic phenomena, suggesting that agreement must figure in syntax. However, partial possessor agreement appears to result from a relationship between the possessee or nominalized predicate, and the last conjunct only, hinting that it is also subject to linear locality conditions. We conclude that the agreement phenomenon in languages results from conditions that apply in syntax proper and from conditions that apply in a post-syntactic component separately, which can alter the output of syntax proper where applicable.
There have been very few attempts to date to provide an explicit semantics/pragmatics for Pseudo-Coordination (PseCo) expressions. This chapter is an attempt to fill that gap, zooming in on the ‘go-(and-)get’-type. To do so, I first provide a syntactic account of PseCo, which derives from a standard coordination structure (which I label Junction), onto and from which a compositional semantic account is derived. The signature pragmatic properties of PseCo of negative-emotive factivity are also derived. Aside from providing the first systematic and cross-modular analysis of PseCo, the chapter also provides a number of new diagnostics for identifying and classifying PseCo expressions which may be useful in future work on the topic.
Verbal pseudocoordination (as in English go and get) is often seen as an idiosyncratic phenomenon described in exceptional terms. This paper establishes the typological context to explain key properties of pseudocoordination, integrated into a more general typology of multi-verb constructions. At the same time, principled motivations are given for the arbitrary list of traditional properties attributed to Serial Verb Constructions (SVCs). The broader category of Multi-Verb Predicates (MVPs) is proposed as any monoclausal multi-verb construction with two verbs forming a complex predicate. Subtypes of MVPs are distinguished by their form: pseudocoordination with a linker ‘and’, while SVCs have no linker. Structural properties of MVPs, such as shared inflectional features on each verb, are readily explained as due to monoclausality.