Laura A. Janda University of Tromsø Valery Solovyev University of Kazan’

1. What Constructional Profiles Reveal About Synonymy and Metaphor: A Case Study of Russian words for ‘sadness’ Laura A. Janda University of Tromsø Valery Solovyev University of Kazan’

2. “Happy families are all alike; every unhappy family is unhappy in its own way.”“Все счастливые семьи похожи друг на друга, каждая несчастивая семья несчастлива по-своему.” --opening line from Tolstoy’s Anna Karenina vs.

5. Overview • Constructional Profiles • What are they • Constructions relevant for ‘sadness’ & ‘happiness’ • Sadness synonyms • Dictionaries vs. Hierarchical Cluster Analysis • Antonyms: Happiness synonyms • Dictionaries vs. Hierarchical Cluster Analysis • Combined analysis of synonyms and antonyms

6. Where the data come from • Russian National Corpus (http://www.ruscorpora.ru) >120M words • Biblioteka Maksima Moškova (http://lib.ru/) >600M words • 500 sentences extracted and coded for use of preposition & case

7. Constructional Profile • A constructional profile is “the distribution of relative frequencies of constructions associated with a given word” • There are about 70 constructions of the form “[(preposition) [noun]case]” in Russian • The null hypothesis is that all nouns should have equal frequency in all constructions • But only about 6 (or fewer) constructions are needed to characterize a given noun • Example of constructional profile of vostorg ‘ecstasy’

8. ‘from’ ‘in’ ‘with’ ‘into’ ‘agent’

9. ‘Sadness’ in Russian • grust’, melanxolija, pečal’, toska, unynie, xandra • The constructions they appear in: • v + Acc ‘into’ • v + Loc ‘in’ • Inst: Agent • s + Inst ‘with’ • ot + Gen ‘from’ • (Direct Object) • (Other Constructions) Graphs will show only these five, as percentages Hierarchical Cluster Analysis is computed from all data

10. v + Acc ‘into’ Nesmotrja na saxarnuju bolezn’, on soxranjal ešče umstvennye sposobnosti, no smert’ blizkix ljudej povergla ego v krajnjuju grust’. [Not-looking on sugar disease-Acc, he retained still mental faculties-Acc, but death-Nom near people-Gen cast him-Acc in extreme sadness-Acc.] ‘In spite of his diabetes, he continued to retain all his mental faculties, but the death of people near to him cast him into extreme sadness.’

11. v + Loc ‘in’ Ej na um ne pridet, čto suprug iznyvaet v toske o poterjannyx minutax truda i vdoxnovenija. [Her-Dat on mind-Acc not comes, that spouse-Nom pines-away in sadness-Loc about lost minutes-Loc labor-Gen and inspiration-Gen.] ‘It doesn’t occur to her that her spouse is suffering in sadness over the minutes of work and inspiration that he has lost.’

12. Inst: Agent Kto iz živuščix, tomimyj unyniem, ne predavalsja takomu pereboru ne sostojavšixsja žiznennyx variantov? [Who-Nom from living-Gen, tormented sadness-Inst, not indulged such inventory-Dat not happened life’s variants-Gen?] ‘Who among mortals tormented by sadness, has not indulged in an inventory of all the things that didn’t happen in their life?’

13. s + Inst ‘with’ Ox, smertyn’ka moja prixodit, podumal Arefa s unyniem. [Oh, death mine-Nom approaches, thought Arefa-Nom with sadness-Inst.] ‘Oh, my death is approaching, thought Arefa with sadness.’

14. ot + Gen ‘from’: Healing from disease Samoe lučšee lekarstvo ot xandry èto čtenie. [The best medicine-Nom from sadness-Gen that reading-Nom.] ‘The best cure for sadness is reading.’

15. ot + Gen ‘from’:Cause Podumajte, ètot čelovek umer ot melanxolii! [Think, that person-Nom died from sadness-Gen!] ‘Just imagine, that person died of sadness!’

16. The ‘sadness’ nouns • They can’t all be the same: Uxodiš’, i ja gljažu vsled tebe s grust’ju, no bez toski. [Depart, and I-Nom look following you-Dat with sadness-Inst, but without sadness-Gen.] ‘You leave and I watch you go with sadnessgrust’, but without sadnesstoska.’ • Dictionaries differ on how synonyms are grouped: • Most usual grouping: grust’, pečal’, toska vs. melanxolija, unynie, xandra • Disagreements over unynie: • Apresjan et al. 1997: unynie goes with pečal’ • Aleksandrovna 1989: unynie goes with melanxolija and xandra • Evgen’evna 2001: unynie goes with grust’ and xandra • Švedova 2003: unynie goes with grust’, melanxolija and xandra

17. ‘Sadness’ Hierarchical Cluster xandra melanxolija grust’ unynie pečal’ toska

21. ‘Happiness’ in Russian • likovanie, naslaždenie, radost’, udovol’stvie, vostorg • Antonyms are words that are virtually identical, but differ in one value • ‘Happiness’ nouns focus on the same constructions in their constructional profiles as ‘sadness’ nouns • Dictionaries differ widely in grouping of ‘happiness’ synonyms • Aleksandrova 1998: naslaždenie & udovol’stvie vs. radost’ & likovanie vs. vostorg • Švedova 2003 and Abramov 1994: naslaždenie & udovol’stvie vs. likovanie, radost’ & vostorg

22. ‘Happiness’ Hierarchical Cluster naslaždenie radost’ udovol’stvie likovanie vostorg

25. ‘Sadness’ & ‘Happiness’ Hierarchical Cluster Group 1 Group2 Group 3 unynie pečal’ udovol’stvie grust’ radost’ melanxolija toska xandra likovanie vostorg naslaždenie

29. Conclusions • Constructional profiles give us data on the behavior of synonyms/antonyms • Constructional profile data can be used in Hierarchical Cluster Analysis • Hierarchical Cluster Analysis sheds light on the relations between synonyms and antonyms

30. Post-conclusion: Was Tolstoy right? • We can’t say for sure, but • ‘sadness’ and ‘happiness’ nouns have a lot in common • there are more “outliers” among ‘sadness’ nouns