Skip to main content
SpringerLink
Log in

Animacy and differential object marking in Old Church Slavonic

Одушевленность и дифференцированное маркирование объекта в старославянском языке

  • Published:
Russian Linguistics Aims and scope Submit manuscript

Abstract

This article explores the synchronic variation between the nominative-accusative (NA) and genitive-accusative (GA) in the oldest layer of canonical Old Church Slavonic (OCS), using parallel Greek and OCS data with principled information status annotation. Firstly, the data are used to clarify the claims made about the pragmatic properties of the alternation in the previous literature. There is a good case for claiming that OCS GA marking functions as a limited type of definiteness marking, i.e. that GA objects will nearly always be previously mentioned or contextually accessible. Secondly, the data are used to examine whether the GA–NA variation correlates with any other discourse properties known to be important in differential object marking systems. The NA is found to be a marker of referential persistence: a new referent will typically be NA-marked if it is an important participant in the further narrative. Third, the focus is shifted to the relationship between subject and object properties. There are indications that the GA is preferred even with new object referents if the subject has low prominence. Thus, the variation is best understood as a situation of differential object marking conditioned by several discourse properties: definiteness, referential persistence and perhaps subject-object asymmetry.

Аннотация

В статье исследуется синхронная вариативность в употреблении форм винительно-родительного (ВИН-РОД) и винительно-именительного (ВИН-ИМ) падежей в древнейших текстах старославянского канона. Анализ опирается на параллельный корпус старославянских и греческих текстов, размеченный в том числе на уровне информационной структуры предложения. Вначале рассматриваются существующие гипотезы о прагматических свойствах вариативности ВИН-РОД/ВИН-ИМ в свете корпусных данных. Есть основания утверждать, что ВИН-РОД в старославянском функционирует как подобие маркера определенности, т.е. если прямой объект стоит в форме ВИН-РОД, то почти всегда он либо употреблялся раньше, либо доступен из контекста. Затем проверяется, коррелирует ли вариативность ВИН-РОД/ВИН-ИМ с какими-либо другими дискурсивными параметрами, релевантными для дифференцированного маркирования объекта. Выясняется, что ВИН-ИМ служит маркером значимости для дальнейшего повествования. Наконец, рассматривается взаимосвязь свойств субъекта и объекта. Есть некоторые указания на то, что если выделенность субъекта в дискурсе невысока, то для объекта будет предпочтительна форма ВИН-РОД, даже если референт является новым. Таким образом, вариативность ВИН-РОД/ВИН-ИМ следует описывать как случай дифференцированного маркирования объекта, в котором выбор формы обуславливается рядом дискурсивных параметров: определенностью, значимостью референта для дальнейшего повествования, и, возможно, асимметрией субъекта и объекта.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. Note that Krys’ko (1994, pp. 63–65) rejects Meillet’s proposal on the basis of the behaviour of nouns denoting animals, but he does not himself examine the nouns denoting human beings.

  2. His ‘strong tendency’, as it turns out in the latter case, boils down to a distribution of 17 NAs with indefinite reference and 10 with definite reference.

  3. All data sets and statistical code necessary to replicate this study are publicly archived at the Tromsø Repository for Language and Linguistics (opendata.uit.no), hdl:10037.1/10190.

  4. foni.uio.no:3000.

  5. For a fuller description of the corpus, annotation and alignment, see Haug et al. (2009).

  6. Tromsø Old Russian and OCS Treebank (TOROT), https://nestor.uit.no. Note that the OCS data from PROIEL are also found in TOROT, which uses the OCS data from PROIEL as the basis for an expanded OCS and Old / Middle Russian corpus. The Suprasliensis data are still under development, at the time of data extraction (January 2014) chapters 1–28 were fully annotated.

  7. p=0.0009 for common nouns, Fisher’s exact test, two-tailed.

  8. Chapters 6 (Encomion on the 40 martyrs of Sebasteia), 20 (John Chrysostom, Homily on the Annunciation part 1), 25 (Vita of John the Hesychast) and 27 (John Chrysostom, Homily on the resurrection of Lazarus after four days part 2).

  9. http://titus.fkidg1.uni-frankfurt.de/texte/etcs/slav/aksl/zograph/zogra.htm (29 January 2015).

  10. The tag was only used if the antecedent occurred within a 13-sentence window. If the antecedent was outside that window, the tag old-inact (old inactive) was used. The two tags are collapsed in all statistics in this article.

  11. In the sense that the discourse referent in question only exists inside certain embeddings, such as negation, modality etc., short-term referents in Karttunen’s (1969) terms.

  12. For a fuller description and problematisation of the annotation scheme and its theoretical background, see Haug et al. (2014).

  13. For interannotator agreement statistics, see Haug et al. (2014).

  14. The algorithm was written by Dag Haug and myself. The scores were adjusted during various experimental runs on a different dataset.

  15. old: 15, acc_sit: 13, acc_inf: 10, acc_gen: 5, acc_gen: 2, new: 0. In addition old_inact is scored 4 and anchored referents (new, but with an old dependent) are scored 8.

  16. Subject: 10, object: 5, oblique argument: 2, complement clause: 1, adverbial: 1, other syntactic relations are scored 0.

  17. The OCS nouns in the PROIEL corpus are annotated for animacy at lemma level, following Zaenen et al. (2004), for more details on the tagging, see Haug et al. (2009). The scores are human: 10; org (human collective): 5; animal: 3; concrete: 3; time, place, nonconc, veh(icle): 0.

  18. The linearly first of the topic candidates is scored 15. Note that the word order is mostly taken from the Greek original, but the OCS word order is highly consistent with the original.

  19. +5 for personal pronouns, +5 for human proper nouns, +30 for prodrops unless they are non-spec, kind, or new.

  20. +10 for the candidate with the highest relative saliency (number of mentions in the 30 preceding sentences).

  21. +2 each if the antecedent outranks the intervening tokens with information status on these hierarchies.

  22. This is not to deny that the GA also occurred with nouns denoting animals in the OCS, cf. Krys’ko (1994). They are, however, very rare in the Marianus / Zographensis data set.

  23. No clear differences were found between direct objects and complements of accusative-governing prepositions. The two groups were therefore collapsed in the analysis, and both are referred to as ‘objects’ throughout, unless the distinction is important.

  24. Note in particular that all occurrences headed by viděti ‘see’ were excluded, since the verb can marginally take the genitive regardless of animacy.

  25. Of course, we do not have the exact originals of the Gospel translations either. The PROIEL corpus uses the Tischendorf New Testament edition, which has multiple discrepancies from what we would expect from the real original. However, the discrepancies are small and contribute only a small amount of noise to this study.

  26. Examples are given in transliteration, with supralinear letters taken down, but unexpanded. Abbreviation marks are not indicated in the transliteration.

  27. Note that the counts do not match the counts in Table 6 since bogъ ‘god’ has been excluded from the variation set.

  28. The three types of accessibility coded in the PROIEL corpus are collapsed in both tables: world knowledge accessibility (3 NA, 47 GA in Table 8; 3 NA, 7 GA in Table 9), accessibility by inference (12 GA in Table 8) and accessibility by deixis (6 GA in Table 8; 4 GA in Table 9). Anchored nominals are coded as new in the corpus, but anchoring has been computed by looking for old dependents with new, non-specific and kind nominals.

  29. Fisher’s exact test, two-tailed, in both cases.

  30. There is one final example with no information status, a PP occurring as a predicative complement (Mar. Lk 7:11).

  31. The same passage has a NA in the Zographensis.

  32. rabъ ‘servant’, gospodinъ ‘master’, drugъ ‘other’, pastyrь ‘shepherd’, cěsarь ‘king’. The two occurrences of the latter may well be nominatives (Mar. Zogr. Jh 19:21).

  33. p-value < 0.00001 for both sets, Wilcoxon test. The Wilcoxon test is commonly used as a substitute for the t-test when the population cannot be assumed to be normally distributed, and should be a good fit for this data set.

  34. p-value = 0.1979 for the full set and 0.2053 for the variation set, Wilcoxon test.

  35. p-value = 0.00240, Wilcoxon test.

  36. p-value = 0.1212, Wilcoxon test.

  37. p-value < 0.00001 in the full set (66 NAs and 65 GAs), p-value = 0.00025 in the variation set (35 NAs, 23 GAs), Wilcoxon test.

  38. Only four of these belonged to lemmata which were attested in variation in the data set, and all of those belonged to the same lemma: drugъ ‘other’.

  39. Subject ranks were calculated both for finite verbs and for control infinitives and conjunct participles, both of which have external subjects, which are indicated in the PROIEL syntax.

  40. p-value = 0.00651, Wilcoxon test.

  41. p-value = 0.1093, Wilcoxon set.

  42. The exception is i privedošę kъ njemu gluxъ gǫgnivъ ‘Some people brought him a deaf man who also had a speech impediment’ (Zogr. Mk 7:32; translation from the International Standard Version, quoted from http://biblehub.com/mark/7-32.htm). This example also has a null subject, but the referent is new (‘some people’). Mar. has a GA object in this passage.

  43. The example is a complicated one. This is a quotation from Zechariah 13:7: “ ‘Awake, sword, against my shepherd, against the man who is close to me!’ declares the Lord Almighty. ‘Strike the shepherd, and the sheep will be scattered, and I will turn my hand against the little ones’ ” (translation from the New International Version, quoted from http://biblehub.com/zechariah/13-7.htm). Only the last part is quoted, but the shepherd has an antecedent in the source text, and should be understood as at least accessible in the passage from Matthew. The same passage has a GA object in Mar. Mk 14:27.

References

  • Aissen, J. (2003). Differential object marking: iconicity vs. economy. Natural Language & Linguistic Theory, 21(3), 435–483.

    Article  Google Scholar 

  • Chiriacescu, S., & von Heusinger, K. (2009). Pe-marking and referential persistence in Romanian. In A. Riester & E. Onea (Eds.), Focus at the Syntax-semantics interface (Working Papers of the SFB 732 ‘Incremental Specification in Context’, 3, pp. 1–17). Stuttgart.

    Google Scholar 

  • Comrie, B. (1978). Genitive-accusatives in Slavic: the rules and their motivation. In B. Comrie (Ed.), Classification of grammatical categories (Current Inquiry into Language and Linguistics, 21. Slavic Series, 6, pp. 27–42). Edmonton.

    Google Scholar 

  • Corbett, G. G. (1991). Gender. Cambridge.

    Book  Google Scholar 

  • Dalrymple, M., & Nikolaeva, I. (2011). Objects and information structure (Cambridge Studies in Linguistics, 131). Cambridge.

    Book  Google Scholar 

  • Haug, D. T. T., Jøhndal, M. L., Eckhoff, H. M., Welo, E., Hertzenberg, M. J. B., & Müth, A. (2009). Computational and linguistic issues in designing a syntactically annotated parallel corpus of Indo-European languages. Traitement Automatique des Langues, 50, 17–45.

    Google Scholar 

  • Haug, D. T. T., Eckhoff, H. M., & Welo, E. (2014). The theoretical foundations of givenness annotation. In K. Bech & K. Eide (Eds.), Information structure and syntax in Germanic and Romance languages. Amsterdam.

    Google Scholar 

  • Huntley, D. (1993). Old Church Slavonic. In B. Comrie & G. G. Corbett (Eds.), The Slavonic Languages (pp. 125–187). London, New York.

    Google Scholar 

  • Karttunen, L. (1969). Discourse referents. In Proceedings of the International Conference on Computational Linguistics COLING 1969 (pp. 1–38). Stroudsburg.

    Google Scholar 

  • Klenin, E. (1983). Animacy in Russian. A new interpretation (UCSL Slavic Studies, 6). Columbus.

    Google Scholar 

  • Krys’ko, V. B. (1994). Razvitie kategorii oduševlennosti v istorii russkogo jazyka. Moskva.

    Google Scholar 

  • Lambrecht, K. (1996). Information structure and sentence form. Topic, focus, and the mental representations of discourse referents (Cambridge Studies in Linguistics, 71). Cambridge.

    Google Scholar 

  • Lunt, H. G. (2001). Old Church Slavonic grammar (7th rev. ed.). Berlin, New York.

    Book  Google Scholar 

  • Malchukov, A., & de Swart, P. (2009). Differential case marking and actancy variations. In A. Malchukov & A. Spencer (Eds.), The Oxford handbook of case (pp. 339–355). Oxford.

    Google Scholar 

  • Meillet, A. (1897). Recherches sur l’emploi du génitif-accusatif en vieux-slave. Paris.

    Google Scholar 

  • Nørgård-Sørensen, J. (2006). Aspect and animacy in the history of Russian. Developing the idea of parallel grammaticalization. In O. Nedergaard Thomsen (Ed.), Competing models of linguistic change. Evolution and beyond (Amsterdam Studies in the Theory and History of Linguistic Science. Series IV – Current Issues in Linguistic Theory, 279, pp. 289–305). Amsterdam.

    Chapter  Google Scholar 

  • Prince, E. (1981). Toward a taxonomy of given-new information. In P. Cole (Ed.), Radical pragmatics (pp. 223–255). New York.

    Google Scholar 

  • Riester, A., Lorenz, D., & Seemann, N. (2010). A recursive annotation scheme for referential information status. In N. Calzolari et al. (Eds.), Proceedings of the Seventh International Conference on Language Resources and Evaluation: LREC 2010 (pp. 717–722). Valletta.

    Google Scholar 

  • Timberlake, A. (1997). Čemu jesi slěpil brat svoi: Templates and the development of animacy. Russian Linguistics, 21(1), 49–62.

    Article  Google Scholar 

  • Zaenen, A., Bresnan, J., O’Connor, M. C., Carletta, J., Garretson, G., Koontz-Garboden, A., Nikitina, T., & Wasow, T. (2004). Animacy encoding in English: why and how. In Proceedings of the 2004 ACL Workshop on Discourse Annotation. 25–26 July 2004 (pp. 118–125). Barcelona.

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. University of Tromsø – The Arctic University of Norway, Tromsø, Norway

    Hanne Martine Eckhoff

Authors
  1. Hanne Martine Eckhoff
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hanne Martine Eckhoff.

Additional information

I would like to thank all the members of the PROIEL project (University of Oslo), especially Dag Haug. This article could not have been written witout our communal effort at tackling information structure issues in ancient texts. I would also like to thank the members of the CLEAR group (UiT The Arctic University of Norway) for valuable comments to several drafts of the article, as well as conference audiences in Lawrence (Kansas), Frankfurt, Osaka, Reykjavik and Leipzig for input and discussion.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Eckhoff, H.M. Animacy and differential object marking in Old Church Slavonic. Russ Linguist 39, 233–254 (2015). https://doi.org/10.1007/s11185-015-9148-3

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11185-015-9148-3

Keywords

Search

Navigation