This chapter offers a descriptive and theoretical account of ditransitives and reassesses the contribution of diachronic research to their analysis and understanding. It opens with some introductory remarks about the syntactic and semantic status of ditransitives from a functional-typological perspective. Then, it provides an updated state of the art on the relevant literature on the topic, showing that scholarship has thus far predominantly dealt with ditransitives from a synchronic viewpoint. However, given that one of the characteristic traits of ditransitive verbs and constructions is precisely their high degree of synchronic variation in terms of structural alternation and alignment split, the diachronic approach can shed light on distinct routes of evolution followed by these verbs across languages. The present chapter focuses on the main developmental pathways along which ditransitives change; it examines which factors play a role in determining the emergence or decay of competing ditransitive constructions, as well as the rise of new meanings and functions; finally, it discusses the general principles that seem to be involved in the functional reorganization of coexisting ditransitive constructions.
Complex predicates built with dar ‘give’ and a noun carrying the primary eventive information are common in Spanish. Our focus in the present chapter rests on a specific subclass of such predicates, used to refer to physiological experiences and characterized by the presence of a sentence-initial human participant coded in the dative case (cf. Me dio hambre ‘I felt hungry’, lit. ‘To me gave hunger’). The literature on oblique subject experiencers contains many examples of parallel structures featuring a variety of light verbs. The objective of our corpus-based study is to shed light on the diachronic process by which Spanish literal ‘give’ developed its supporting function in the physiological expressions under analysis. We locate the source of the historical development in one of the polysemous senses of dar (‘hit, strike’), product of the entrenchment of an old collocational pattern (give + blow), and we trace the evolutionary path leading from the source meaning to the experiential domain, drawing special attention to a sequence of analogical extensions guided by similarities with extant forms and templates. The outlined history accounts for the fact that the experiential predicates with dar do not exhibit the reflexive marker which had to be expected if a passive or anticausative type of derivation, with an original ‘giver’ suppressed, had been involved.
This chapter aims at discussing whether the indirect object is a pure grammatical relation, ready to host a heterogeneous set of roles, or a motivated coding form of a coherent family of roles. After justifying the choice of the sample of languages (Italian, French, Spanish, Latin and English) and explaining the theoretical assumptions about the autonomy of syntax and its limits, the chapter provides some arguments for the hypothesis that the indirect object is a grammatical relation within the examined languages - namely, the lack of inner coherence of the set of roles associated with the indirect object, the lack of a biunivocal correlation between grammatical relations and roles, and the behaviour of the ditransitive construction when transferred onto two-place verbs. A diachronic section illustrates the shift in coding regime that leads an iconic coding form of expression of a consistent family of allative roles in Latin to become the form of expression of a grammatical relation in Romance languages. The conclusion is that the indirect object is a grammatical relation, which implies that the ditransitive construction is in the first place a formal network of grammatical relations, including a subject, a direct object and an indirect object, filled up with a three-place verb ready to provide any grammatical relation with an argument.
This chapter explores the competition between the use of the dative case and the prepositional construction featuring ad ‘to’ to encode the third argument of ditransitive verbs in Merovingian Latin. Unlike the dative, the prepositional strategy is semantically transparent, since ad ‘to’ has a clear allative meaning; accordingly, its gradual expansion in the functional domain of ditransitivity is expected to follow a revealing path of decreasing semantic motivation, where metaphorical transfers played a major role in the increasing spread of the construction from classical Latin onwards. The Merovingian data discussed in this study add further details on the historical drift towards the grammaticalization of a/à in Romance. In such diatopic variant of late Latin, the gradual extension of the prepositional construction to third arguments of ditransitive verbs continues along the same semantically constrained path identified in earlier stages, but occasionally spreads to additional types of verbs which do not imply any kind of movement. The data discussed in this chapter thus witness a transitory phase where we begin to get glimpses of a substantial shift towards the grammaticalization of the indirect object as a void grammatical relation in Romance languages.
This chapter deals with the variety of alignment types, as identified by typological research, to which ditransitive verbs gave rise in Old Italian, where the Recipient may be expressed not only as an indirect object, as is proper of ditransitive constructions in Modern Italian (showing indirective alignment), but also as a direct object: this mirrors the Latin state of affairs, where neutral alignment and secundative alignment were allowed beside indirective alignment. In this contribution it will be shown how the set of changes undergone by ditransitive constructions from Latin to Modern Italian is in principle larger and more complex than previously recognized, and how Old Italian is crucial when attempting to reconstruct the diachronic paths taken by this class of verbs. In particular, the factors will be investigated which can explain evolution, partial continuity and loss in the argument structure of Italian ditransitives.
Present-day Dutch displays a productive argument structure alternation between a double object construction with bare NP Theme and Recipient objects and a prepositional-dative in which the Recipient is marked by the preposition aan. The latter construction is the younger of the two: it is a post-Middle Dutch innovation, which became a well-established part of the grammar of Dutch in 17th century language only. The present chapter investigates the emergence of an embryonic dative alternation in the immediately preceding period, through a detailed investigation of the presence of aan-datives in three nonbelletristic texts from the latter half of the 16th century. It will be argued that the degree of semantic overlap between the DOC and the “new” aan-pattern increased through the combined effect of chains of local extensions in various sub-regions of “dative” semantic space.
In this study, we revisit the history of the English dative and benefactive alternations in the light of De Smet et al.’s (2018) model of language change. Countering traditional competition models, this model admits that the degree of functional overlap between two distinct linguistic units may increase over time, a process the authors label attraction, which ultimately rests on analogy. Adopting this perspective, we propose that the restructuring of these alternations took place over at least the following four stages: 1) in Old English, the double object construction (DOC) is found with verbs of transfer such as give and send as well as verbs of creation and (possibly) pure benefaction, coding both dative and benefactive scenarios; 2) starting from the Middle English period, attraction takes place between DOC and a periphrastic construction featuring the preposition to (to-POC) with transfer verbs such as give as well as verbs of creation such as make; 3) later on, possibly from the end of the Middle English period, to-POC with verbs of creation such as make disappears in favour of DOC and a new periphrastic construction featuring the preposition for (for-POC), used also with verbs of pure benefaction; 4) the last stage involves the shedding of pure benefaction from DOC, which instead is still found with for-POC.
This chapter discusses the development of the ditransitive alternation with the verb verkaufen (‘sell’) in the history of New High German (1650-present) from a probabilistic point of view. The alternation concerns the constructional variation between the “Indirect Object Construction” (IOC) with dative case (e.g. der Frau ‘the woman’) and the “Prepositional Object Construction” (POC) (e.g. an die Frau ‘to the woman’). A random sample of 700 IOC and 741 POC attestations extracted from the HIST-Archiv, the Deutsches Textarchiv (1650-1899) and the Kernkorpus (1900-1999) was annotated for sixteen variables. The study focuses on (i) the relative proportions of both constructions over time, (ii) constituent order preferences of THEME and RECIPIENT and (iii) potentially motivating factors for the alternation. The corpus analysis reveals that POC imposes itself in the course of the 18th century, followed by a further gradual increase of the construction in the 19th century. Conversely, IOC with THEME-RECIPIENT order gradually falls into disuse. A logistic regression analysis provides evidence for the effect of ten variables on the alternation. A comparison with findings from previous studies of the corresponding alternation in English and Dutch reveals many similarities in the history of the three languages, but also a number of differences.
This chapter explores the diachronic behavior of ditransitive predicates in Vedic Sanskrit, the language of the so-called Vedas and their ancillary texts, the Brāhmaṇas. Vedic Sanskrit has a relatively consistent nominative-accusative alignment system, allowing for a broad range of object (P) alternation patterns but being more restrictive as regards subject (A/S) realization (cf. e.g. Dahl 2009, 2014a, 2014b, 2019). Ditransitive predicates show three morphosyntactically distinct argument realization patterns. Certain verbs have an indirective pattern, where P selects accusative case and R dative case, a secundative pattern, where P has locative case marking and R accusative, and a neutral pattern, where both P and R receive accusative case marking. Among these, the indirective pattern is predominant, because it attracts a larger number of verbs than the other two patterns. Diachronically, ditransitive predicates remain stable in their argument realization options throughout the history of Vedic Sanskrit. This fact may suggest that lexical semantic factors play a more important role than general linking rules in argument realization in Vedic Sanskrit, at least as regards ditransitive constructions.
This chapter investigates the properties and the diachronic development of a poorly described Italian construction, vedersi ‘to see oneself’ followed by the infinitive, with the aim of defending its Recipient passive interpretation. According to the recent typological literature (Malchukov, Haspelmath, and Comrie 2010), Italian, as a language that follows an indirective coding pattern, should not allow the passivization of the Recipient argument in ditransitives. The case under scrutiny represents a special strategy whose motivations are to be searched for, on the one hand, in the tendency for Recipients to take the subject role in passivization (Haspelmath 2015), and on the other at the pragmatic level in a response to the need for expressing the (mostly negative) subject affectedness. The diachronic analysis illustrates the grammaticalization path of vedersi + infinitive leading from the reflexive construction, in which vedersi has a perceptive meaning, to a Recipient passive construction through texts from Old to Modern Italian. The semantic expansion of the Recipient construction beginning in the late 19th century, through its ability to attract a variety of ditransitive predicates, is illustrated with corpus-based examples. The earlier but parallel development of a Theme passive in the construction vedersi + past participle (Giacalone Ramat 2017) confirms the main lines of change and suggests a new assessment of passivization patterns in Modern Italian.