Prosodic weight, word order, and the ...
Document type :
Autre communication scientifique (congrès sans actes - poster - séminaire...): Communication dans un congrès avec actes
Title :
Prosodic weight, word order, and the structure of the Latin clause
Author(s) :
Conference title :
55th Annual Meeting of the Societas Linguistica Europaea
Conference organizers(s) :
University of Bucharest
City :
Bucharest
Country :
Roumanie
Start date of the conference :
2022-08-24
English keyword(s) :
Syntax
Word order
Latin
Word order
Latin
English abstract : [en]
My aim is to investigate the influence of prosodic weight on constituent order in Latin. In the wake of Behaghel (1932), in the philological literature it is often claimed that heavy objects are more likely to surface ...
Show more >My aim is to investigate the influence of prosodic weight on constituent order in Latin. In the wake of Behaghel (1932), in the philological literature it is often claimed that heavy objects are more likely to surface postverbally than lighter ones, as an effect of a broader tendency for heavy material to gravitate towards the right edge of the clause. I report on two corpus-based experiments, which reveal that in Latin the empirical picture is more complicated.In a first experiment, I looked at a sample of ca. 8.000 clauses featuring a finite verb and a nominal direct object, drawn from a corpus covering the period from ca. 200 BCE until 600 CE. In this sample, the factor ‘prosodic weight’, operationalized as ‘number of orthographic words’, cannot be shown to have any statistically significant effect on the OV/VO-alternation, neither synchronically nor diachronically. An alternative way of operationalizing the weight variable, which simultaneously takes into account both the number of words and the average word length, also fails to yield a significant result. I propose that this (non-)result can be understood in terms of the typological observation that the direction in which heavy material is shifted co-varies with the basic directionality of complementation (Hawkins 1994). Concretely, rightward extraposition of heavy XPs is restricted to head-initial contexts, whereas in head-final contexts, heavy constituents are more likely to undergo leftward movement. To test whether this line of reasoning is on the right track, drawing on Faghiri et al. 2018 I carried out a follow- up experiment, looking at a sample of 2.500 clauses featuring a finite form of the ditransitive verb dare ‘give’ and two NP internal arguments. The results confirm (i) that when both the direct and the indirect object precede the verb, the heavier one is more likely to precede the lighter one, and (ii) that when both appear postverbally, the heavier object is more likely to follow the lighter one. As predicted, the factor ‘weight’ does not come out as significant if the verb sits between both objects.I conclude that prosodic weight is indeed correlated with constituent order in Latin, but the direction of Heavy NP shift systematically co-varies with the headedness of the VP. In the closing sections of the talk I sketch how the experiments reported on here can be further enriched to arrive at a multifactorial characterization of the data under discussion. Additional predictors include (i) priming effects (e.g., Labov 1987), (ii) internal (syntactic) complexity of NP objects (Wasow 2002), and (iii) information structure (Taylor and Pintzuk 2012). Preliminary results suggest that all three of these are also correlated with variable constituent order, in conjunction with the factor ‘prosodic weight’.Show less >
Show more >My aim is to investigate the influence of prosodic weight on constituent order in Latin. In the wake of Behaghel (1932), in the philological literature it is often claimed that heavy objects are more likely to surface postverbally than lighter ones, as an effect of a broader tendency for heavy material to gravitate towards the right edge of the clause. I report on two corpus-based experiments, which reveal that in Latin the empirical picture is more complicated.In a first experiment, I looked at a sample of ca. 8.000 clauses featuring a finite verb and a nominal direct object, drawn from a corpus covering the period from ca. 200 BCE until 600 CE. In this sample, the factor ‘prosodic weight’, operationalized as ‘number of orthographic words’, cannot be shown to have any statistically significant effect on the OV/VO-alternation, neither synchronically nor diachronically. An alternative way of operationalizing the weight variable, which simultaneously takes into account both the number of words and the average word length, also fails to yield a significant result. I propose that this (non-)result can be understood in terms of the typological observation that the direction in which heavy material is shifted co-varies with the basic directionality of complementation (Hawkins 1994). Concretely, rightward extraposition of heavy XPs is restricted to head-initial contexts, whereas in head-final contexts, heavy constituents are more likely to undergo leftward movement. To test whether this line of reasoning is on the right track, drawing on Faghiri et al. 2018 I carried out a follow- up experiment, looking at a sample of 2.500 clauses featuring a finite form of the ditransitive verb dare ‘give’ and two NP internal arguments. The results confirm (i) that when both the direct and the indirect object precede the verb, the heavier one is more likely to precede the lighter one, and (ii) that when both appear postverbally, the heavier object is more likely to follow the lighter one. As predicted, the factor ‘weight’ does not come out as significant if the verb sits between both objects.I conclude that prosodic weight is indeed correlated with constituent order in Latin, but the direction of Heavy NP shift systematically co-varies with the headedness of the VP. In the closing sections of the talk I sketch how the experiments reported on here can be further enriched to arrive at a multifactorial characterization of the data under discussion. Additional predictors include (i) priming effects (e.g., Labov 1987), (ii) internal (syntactic) complexity of NP objects (Wasow 2002), and (iii) information structure (Taylor and Pintzuk 2012). Preliminary results suggest that all three of these are also correlated with variable constituent order, in conjunction with the factor ‘prosodic weight’.Show less >
Language :
Anglais
Peer reviewed article :
Oui
Audience :
Internationale
Popular science :
Non
Collections :
Source :