Definite referential null objects in Vedic Sanskrit and Ancient Greek

Abstract

Ancient Greek and Early Vedic are pro-drop languages which allow for referential null objects. In this paper we give an overview of the various conditions under which null objects are licensed and compare their use in both languages. In Greek and Vedic null objects occur frequently in conjuncts, be it clauses or sentences. They are also attested with participles embedded into finite sentences. A third type is the syntactically unrestricted discourse conditioned null object, which is typically an anaphora. In Vedic, however, it can also be used cataphorically and with extratextual reference.

Keywords::

• null objects
• conjuncts
• participles
• Vedic Sanskrit
• Ancient Greek
• pro-drop
• anaphors

Notes

¹ Early Vedic, the language of the Rigveda, is the oldest attested Indo-Aryan language. The Rigveda is a collection of 1028 metrical hymns composed for usage in rituals. The oldest part of the corpus goes back to at least 1500 BCE, the most recent hymns to ca. 1200 BCE (Witzel 1995). The collection as we know it today was compiled ca. 700 BCE. Since then the text was transmitted orally. Although the hymns were composed in a time span of roughly 300 years it is reasonable to treat the whole corpus as one synchronic stratum for syntactic investigations (Hettrich 2007; Dahl 2010).

² Contrary to Sanskrit, for which we limit our corpus as specified in fn.1, research on Greek spans over the whole literary sources in our possession. Previous research (see especially Luraghi 2003) has revealed that there is virtually no difference in the conditions under which null objects appear at different diachronic stages; in particular, conditions on null objects in Homeric Greek (ca. 1200 BCE) seem to be pretty much the same as in Classical Attic-Ionic prose (fifth − fourth century BCE).

³ Oblique forms of autón when used as anaphoric pronoun are glossed as 3rd person pronominal forms; elsewhere they are glossed as forms of the demonstrative.

⁴ Occurrence vs. non-occurrence of min (and presumably other special clitics) in Herodotus under similar reference conditions is dependent on the wider discourse context, see Luraghi (2013 and forthcoming); such conditions possibly override conjunction reduction in example (6).

⁵ The third occurrence is Th. 7.52.2, in which the object pronoun is contrastive with the DO of the following clause.

⁶ Example (14) provides the preceding context to example (3). It is also discussed in Luraghi 2003.

⁷ This verb can passivize in Classical Sanskrit, so it can be considered transitive at that language stage. Unfortunately, there is no evidence for passivization in Vedic Sanskrit.

Additional information

Notes on contributors

Götz Keydana

Götz Keydana teaches courses in General Linguistics, Theoretical Linguistics and Indo-European Studies at the Linguistics Institute of Göttingen University. His research focuses on the syntax of ancient Indo-European languages and phonological reconstruction.

Silvia Luraghi

Silvia Luraghi is Associate Professor of Linguistics at the University of Pavia. She is the co-editor of Journal of Historical Linguistics. Her interests include language change, Indo-European linguistics, and language typology.