Knigamŭ bo jest’ neiščetnaja glubina: Essays in Honour of Irina Lysén

Essays in Honour of Irina Lysén

Edited by

Книгамъ бо есть

неищетная глубина

Книгамъ бо есть неищетная глубина

!

Книгамъ бо есть неищетная глубина

Essays in Honour of Irina Lysén

!!

!

!

!

Книгамъ бо есть неищетная глубина

Essays in Honour of Irina Lysén

Edited by

Hanne Martine Eckhoff and

Thomas Rosén

!

!

!

!

!

!

!

!

!

!

!

Uppsala, 2015

! kolofon?

!

!

Table of contents

Tabula gratulatoria

Hanne Martine Eckhoff and Thomas Rosén

Irina Lysén as Teacher, Scholar and Colleague ... 11 Per Ambrosiani

Recently Identified Cyrillic Incunabula in Sweden ... 14 Hanne Martine Eckhoff

Death of a Construction: Old Church Slavonic Touch Verbs ... 24 Mats Eskhult

Swedish Scholarly Interest in the Karaites around 1700 ... 46 Atle Grønn

Параллельный корпус в системе университетских

лингвистических курсов ... 55 Elisabeth Löfstrand

Några vittnesmål från skilda tider om svenskarnas syn

på ryssarnas ikoner ... 64 Lennart Lönngren

Сложные прилагательные с нулевым суффиксом в русском языке ... 81 Tore Nesset

Old Languages, New Ways? ... 91 Олег В. Никитин

«Подлая» речь в деловых документах XVI–XVII веков ... 102 Trond Gunnar Nordenstam

Роль музыки в богослужении Православной Церкви ... 113 Jussi Nuorluoto

On Possible Diagraphemes in the Early Glagolitic and Cyrillic ... 124

!

Małgorzata Anna Packalén Parkman

”Att ha sanningen i spetsen…”

Om Katarzyna Suchcickas lyriska värld ... 136 Raimo Raag

A Criterion for Old Age ... 153 Virve Raag

Ploščad´ Proletarskoj Diktatury: On Ideologized Names and the Re-

establishment of Pre-Soviet Names in Estonia ... 167 Thomas Rosén

Вымирание надстрочных знаков (на материале российской

дипломатической корреспонденции XVIII века) ... 184 С. Ю. Темчин

Eвангельские чтения по гласам Октоиха ... 198 Cynthia M. Vakareliyska

The Synaxarion to the Dobrejšo Gospel ... 209

!

Irinae Lysén,

conlegae, magistrae, familiari nostrae.

Nos, quorum nomina infra scripta sunt, hunc librum tibi donamus sperantes eum tibi gaudio fore votisque piissimis optantes, ut omnes anni consequentes feliciter praetereant.

Postridie Id. Oct. MMXV.

Per Ambrosiani Umeå & Stockholm

Rogier Blokland Uppsala Bernt Brendemoen

Oslo

Hanne Martine Eckhoff Tromsø & Oslo

Erik Egeberg Tromsø Josefine Ekroth

Stockholm Lennart Elmevik

Uppsala Mats Eskhult

Uppsala

!

Gunilla & Ivan Fredriksson Uppsala

Antoaneta Granberg Göteborg Atle Grønn

Oslo

Marianne Gustavsson Uppsala Roger Gyllin

Uppsala Lennart Hagåsen

Uppsala Julie Hansen

Uppsala Hans Helander

Uppsala Lisa Kraft Stockholm Hans Kronning

Uppsala Gösta Lavén,

Uppsala Ivett G. Larsson

Stockholm Jonas Lembke

Stockholm Elisabeth Löfstrand

Stockholm

!

Tamara & Lennart Lönngren Stockholm

Witold Maciejewski Warszawa Gernot Müller

Uppsala Johan Muskala

Uppsala Tore Nesset

Tromsø Oleg V. Nikitin

Moskva

Trond Gunnar Nordenstam Oslo

Jussi Nuorluoto Uppsala Alf Nyholm

Uppsala Sture Packalén

Uppsala

M. Anna Packalén Parkman & Stefan Parkman Uppsala

Olof Paulsson Fjärås Ludmilla Pöppel

Tullinge Virve & Raimo Raag

Uppsala

!

Gunilla Ransbo Uppsala Thomas & Lina Rosén

Uppsala Jelena Spasenić

Uppsala Lars Steensland

Lund

Dessislava Stoeva-Holm Uppsala

Lars-Göran Sundell Uppsala Anna Sundmark

Stockholm Sergejus Temčinas

Vilnius Tone Tingsgård

Uppsala

Cynthia M. Vakareliyska Eugene, Oregon Jan Erik & Birgitta Walter

Åkersberga Christine Watson

Uppsala Jardar N Østbø

Oslo

Stockholms universitetsbibliotek Uppsala universitetsbibliotek

Åbo Akademi

Irina Lysén: Scholar, Colleague, Teacher and Friend

Hanne Martine Eckhoff and Thomas Rosén

For most of her career, Irina Lysén has been affiliated with Uppsala Univer- sity, Sweden. She began there as an undergraduate, continued as a graduate student and a teacher, and, in 2008, was made Professor of Russian. How- ever, her life and work in Uppsala was interposed for much of the 1990s and early 2000s by work as an Associate Professor of Russian at Oslo Univer- sity, Norway. Over the years, in Uppsala, Oslo and elsewhere, Irina has made many friends, a number of whom have agreed to contribute articles to the present volume.

Some of the contributors – notably the authors of this short introduc- tion – have the privilege of knowing Irina in more than one capacity; in our case, as a PhD supervisor, a colleague and a friend. For readers who are not intimately familiar with Irina’s scholarly work, we have attempted to produce this short introduction in which we, among other things, present a selection of her publications.

As a scholar, Irina Lysén is best known for her work in four fields:

textual criticism, the study of mediaeval theories of translation and their ap- plication, comparative linguistic studies of source texts and their translations and, finally, textual editions. Her production in these fields includes work on specific textual traditions as well as investigations of a more general, metho- dological nature.

In her dissertation (Ogren, 1989), Irina investigated the Slavonic trans- lation from Greek of Paraenesis, a work of Ephraim the Syrian, a fourth- century Theologian and Hymnographer. Inspired by experiences made dur- ing her dissertation work, she published a second monograph (Ogren, 1991) in which she dealt with a difficult problem that arises in the study of medi- aeval Slavonic translations: namely the discrepancy between the mediaeval Slavonic manuscripts and more recent, printed editions of Greek texts.

In 1995, Irina published two major works: an article on translation

technique in the earliest Slavonic translations (Ogren, 1995) and a concord-

ance to the oldest preserved copies of the Old Church Slavonic translation of

the Gospels (Lysén, 1995). The latter was characterised as “an extremely

useful handbook which to a large extent facilitates textological and lexico-

logical investigations of this literary monument” (Temčin 2010, 122. Our translation – HME and TR).

During her years at Olso University, Irina took on a very challenging research problem. The problem in question was the Slavonic translation of the Book of Esther, “one of the most problematic and controversial texts in the Slavonic corpus of translations”, as she herself characterised it (Lysén 2001, 379). The main difficulty lay in determining from which language the original Slavonic translation had been made. Had it been from Greek, or di- rectly from Hebrew? A few years prior to the publication of Irina’s mono- graph on Esther, two well-known scholars, Horace G. Lunt and Moshe Taube, had published an investigation in which they concluded that the orig- inal from which the Slavonic translation had been made was Greek (Lunt and Taube 1998). Following a painstaking analysis, Irina arrived at a rather different result, concluding instead most convincingly that the translation had been made directly from Hebrew.

Since the beginning of the 2000s, Irina Lysén has continued her re- search on mediaeval Slavonic translations, e.g. on the Book of Psalms (Lysén, 2009).

Irina’s expertise has come to good use also outside her usual fields of activity. From 2010 to 2013, she participated in the project Digitalised descriptions of Slavic Cyrillic manuscripts and early printed books in Swedish libraries and archives in co-operation with Per Ambrosiani (Umeå), Antoaneta Granberg (Gothenburg) and Alexander Pereswetoff- Morath (Stockholm). The project was funded by Riksbankens jubileumsfond, the tercentenary fund of Sweden’s National Bank.

Since 2010, Irina has been a valued member of the editorial board of the Uppsala-based journal Slovo (see References below).

Irina is not only a careful and dedicated scholar, but also an excellent teacher. For many years, she has taught Russian to countless Scandinavian students, who will remember her as a structured, insightful and exacting, but also warm and humorous teacher, with a precise understanding of the partic- ular difficulties the Russian language poses to native speakers of Swedish or Norwegian. Besides having taught grammar and syntax at all levels, she also took part, in the 1980s, in the establishment of a special Russian curriculum for future business administrators.

In academic writing, Irina has always valued brevity and lucidity. We will therefore try to keep this introduction short. However, an introduction such as this one would not be complete if it did not also mention the indivi- dual’s life outside the workplace.

Besides her university career, Irina has, for three decades, taken active

part in the authorisation of translators and interpreters, organised by Sweden’s

Legal, Financial and Administrative Services Agency (Kammarkollegiet).

In Irina’s case, private life has meant bringing up two children, Marie- Louise and Alexander, and she is now also a grandmother. There has also been time for some gardening, mushroom-picking and other such activities.

As mentioned at the beginning, the essays contained in this volume are written by colleagues – former as well as present – and friends of Profes- sor Irina Lysén in celebration of her sixty-fifth birthday. With the essays, we jointly offer Irina our heartfelt thanks for her unwavering friendship over the years, for clear-headed advice, honesty and for many good times.

References

Lunt, Horace G. and Moshe Taube. 1998. The Slavonic Book of Esther: text, linguistic analysis, problems of translation. Cambridge, Mass: Distri- buted by Harvard University Press for the Harvard Ukrainian Re- search Institute.

Люсен, Ирина. 1995. Греческо-старославянский конкорданс к древней- шим спискам славянского перевода евангелий (codices Marianus, Zographensis, Assemanianus, Ostromiri). Acta Universitatis Upsali- ensis, Studia Slavica Upsaliensia, 36). Uppsala.

Люсен, Ирина. 2001. Книга Есфирь: к истории первого славянского перевода. Acta Universitatis Upsaliensis, Studia Slavica Upsaliensia, 41). Uppsala.

Люсен, Ирина. 2009. “Проблематика древнейших двухступенчатых переводов: Старославянский перевод Псалтыри”. Studies on Lan- guage and Culture in Central and Eastern Europe 3:47–60.

Огрен, Ирина. 1989. Парененсис Ефрема Сирина: к истории славянского перевода. Acta Universitatis Upsaliensis, Studia Slavica Upsaliensia 26. Uppsala.

Огрен, Ирина. 1991. К проблеме использования печатных изданий греческих текстов при исследовании древних славянских пере- водов: на примере славянксого перевода Паренесиса Ефрема Сирина. Acta Universitatis Upsaliensis, Studia Slavica Upsaliensia 31. Uppsala.

Огрен, Ирина. 1995. “К вопросу о теоретическом и практическом бази- се древнейших славянских переводов”. Подобаѥтъ памѧть съ- творити: Essays to the Memory of Anders Sjöberg, edited by Per Ambrosiani, Barbro Nilsson and Lars Steensland). Acta Universitatis Stockholmiensis, Stockholm Slavic Studies 24. Stockholm: Almqvist

& Wiksell International: 157–172.

Slovo: Journal of Slavic languages, literatures and cultures.

http://www2.moderna.uu.se/slovo/ index.html (accessed 27.08.2015).

Temčin 2010. Темчин, Сергей Ю. Исследования по кирилло-мефодие-

вистике и палеославистике. Krakowsko-Wileńskie studia slawis-

tyczne, Tom 5. Kraków: Scriptum.

Recently Identified Cyrillic Incunabula in Sweden

Per Ambrosiani, Umeå University

The article offers an identification and short description of thirty-four earlier unidentified printed leaves with Cyrillic text currently preserved at the Swe- dish National Archives in Stockholm. The leaves belong to three Cyrillic incunabula editions printed in Cracow in the 1490s: a Horologion, a Lenten Triodion, and a Festal Triodion. They constitute, to the best of our know- ledge, the first Swedish Cyrillic incunabula fragments described. The leaves arrived in Uppsala as filling in the binding of a copy of Apuleius Asinus Au- reus. The copy was taken as war booty in Poland in the 1620s, and the lea- ves were later transferred to the National Archives.

1. Introduction

During the work within the project Digitalised Descriptions of Slavic Cyril- lic Manuscripts and Early Printed Books in Swedish Libraries and Archives (2010–2013), in which I had the privilege to cooperate with Irina Lysén,

the main focus of my part of the work was on printed books. From the begin- ning, this work was based on the inventory with short descriptions published by Kjellberg (1951), which includes no incunabula – the oldest item dates from 1519.

Similarly, in the recent catalogue by Wolfgang Undorf (2012), no Cyrillic incunabula are included.

1

The two other participants in the project were Antoaneta Granberg (project coordinator), and Alexander Pereswetoff-Morath. Earlier publications, based on the results of the project, in- clude Ambrosiani and Granberg 2010, Ambrosiani 2012, 2014a, 2014b, 2014c, Pereswetoff- Morath 2013, 2014.

2

See Kjellberg 1951, no. 1: “Liturgie. Venise 7.VIII 7028=1519”, with a handwritten shelf- mark “Ksl. 130”. Nemirovskij 2009a, 333–338, no. 26, mentions 96 copies of this edition, including the one in Uppsala (with a reference only to the University Library, without any shelfmark).

3

Overall, at least 30,000 incunabula editions/titles are known to have existed. Of these, more

than fifty are printed in Slavic languages (cf. Bošnjak 1968, 148–170, who lists five Glagolit-

ic titles, ten Cyrillic, and altogether thirty-nine titles printed with Roman letters, of which

thirty-five are identified as Czech). Until recently, the known Slavic incunabula in Sweden

include six copies and fragments of four Czech editions: 1) Bible, printed in Prague in 1488

(Bošnjak 1968, no. 8, with one copy at a “Swedish library”, see Undorf 2012, no. 696

[http://libris.kb.se/bib/9464979]); 2) Chronicon Martymiany dictum, printed in Prague in

1488 (Bošnjak 1968, no. 9, with one copy at the National Library of Sweden in Stockholm,

However, the project also opened possibilities for reviewing informat- ion on other early Cyrillic prints in Sweden. One such “additional” col- lection included thirty-four earlier unidentified Cyrillic leaves located at the Swedish National Archives (Riksarkivet) in Stockholm.

As it turned out, the thirty-four leaves could be shown to belong to altogether four copies of three different Cyrillic editions printed in the 1490s. The purpose of the pre- sent article is to give a short description of these leaves and discuss their provenance.

2. Three Cyrillic incunabula

Among the eight attested Cyrillic titles printed before 1500,

fragments of the following three, all printed in Cracow by Schweipolt Fiol in the 1490s, have been identified at the National Archives in Stockholm: two double lea- ves from a Horologion, sixteen single leaves from a Lenten Triodion, and altogether sixteen single leaves from at least two different copies of a Festal Triodion.

2.1 Horologion, 1491

Two double leaves (“H/A1+A2, H/B1+B2”) from a Horologion, printed in Cracow in 1491 (cf. Karataev 1861, no. 2, 1883, no. 2; Nemirovskij 1996, 87–108, no. 2; Nemirovskij 2009a, 198–202, no. 2 “Časoslov”; Bošnjak 1968, 165f, no. 46 “Časoslovec´”; ISTC ih00484300; GW 13447; PI, no.

B1), see Table 1. The two preserved double leaves constitute the two outer-

see Undorf 2012, no. 1121 [http://libris.kb.se/bib/8930511]); 3) Bible, printed in Kutná Hora in 1489 (Bošnjak 1968, no. 12, with copies and fragments at Uppsala University Library, the National Library of Sweden in Stockholm, and a “Swedish library”, see Undorf 2012, no. 697 [http://libris.kb.se/bib/8930176]); 4) Jacobus de Voragine: Pasionál čili Knihy o životech svatých, printed in Prague in 1495 (Bošnjak 1968, no. 19, with a copy at a “Swedish library”, see Undorf 2012, no. 2098 [http://libris.kb.se/bib/9465147]) – for an explanation of the repos- itory designation “Swedish library” see Undorf 2012, VIII.

4

I want to thank Lars Steensland and Jan Brunius for bringing my attention to this collection.

5

Parts of the information provided below were presented at the workshop “Krigsbyten och metadata” (‘War booty and meta-data’), held at Uppsala University Library on September 25, 2014.

6

Cf. Nemirovskij 2009a, nos. 1–4, printed in Cracow between 1491 and 1493, and nos. 5–8, printed in Cetinje in 1494–1495; Bošnjak 1968, 165–170, nos. 45–48 and 50–53. Bošnjak also mentions a Psalter (p. 167, no. 50) and a Four Gospels (p. 170, no. 54), of which no cop- ies are preserved.

7

The Horologion edition includes on the last folio a colofon, which indicates that the printing was finished in 1491 (cf. Karataev 1883, 5, Nemirovskij 1996, 93, etc.), but there is still no established scholarly consensus on in which year(s) the Lenten and Festal Triodions were printed. Karataev (1861, 1, 1883, 9f, 12–14), Bošnjak (1968, 166), Lukjanenko (1993, 15f, 19–21), and, recently, Wronkowska-Dimitrowa (2010, 20–22), all date both Triodions to

“1491” or “c. 1491”, whereas Nemirovskij for both editions offers a careful argumentation for

a later dating, “c. 1493” (cf. Nemirovskij 2009a, 51–56, 203, 208). Not being able to provide

any additional argument for either dating, I will indicate the time of printing of both Triodion

editions as “before 1494”.

most leaves of the fourteenth gathering, which originally consisted of four double leaves.

Fragm.

leaf no.

Leaf no. in Russian State Lib- rary copy

(digital surrogate)

Signature in lo- wer right corner of recto page

First words on line 1 of recto page

H/A1 105 к҃з [27] съвъкоупльше сѧ

H/B1 106 – влⷣкоу и г҃а ,

H/B2 111 – троⷫ҇ глаⷭ҇ . д҃ .

H/A2 112 – нѫѧ нарекыи себѣ

Table 1. Horologion, two double leaves 2.2 Lenten Triodion, before 1494

Sixteen single leaves (“LT/A–P”) from a Lenten Triodion, printed in Cracow before 1494 (cf. Karataev 1861, no. 4, 1883, no. 4; Bošnjak 1968, 166, no.

47 “Triod´ postnaja”; Nemirovskij 1996, 109–138, no. 3; Nemirovskij 2009a, 203–207, no. 3 “Triod´ postnaja”; ISTC it00427980; GW M47501;

PI, no. B2), see Table 2. The preserved leaves originally belonged to the third (ff. LT/A–D) and seventh (ff. LT/E–F) gatherings, and they also in- clude all ten leaves of the thirty-first gathering (ff. LT/G–P).

8

The original setup of the gathering was thus: leaf A–ff. 105 and 112, leaf B–ff. 106 and 111, leaf C (not preserved)–ff. 107 and 110, leaf D (not preserved)–ff. 108 and 109. For an over- view of all gatherings see Nemirovskij 1996, 272f.

9

The leaf nos. refer to a non-original foliation of a copy at the Russian State Library, Mos- cow (cf. Nemirovskij 1996, 95, copy no. 2.5), a digital surrogate of which is available at http://dlib.rsl.ru/viewer/01003441794.

10

! Cf. Nemirovskij 1996, 273f for an overview of all gatherings. !

Fragm.

leaf no.

Leaf no. in Russian State Library copy (digital surro- gate)

Leaf no. in Bib- lioteka Naro- dowa copy (digital surro- gate)

Signature in lower right cor- ner of recto page

First words on line 1 of recto page

LT/A 21 20 а҃і [11] на ⷮлѵⷢ҇ри блж҃ена

LT/B 22 21 в҃і [12] даждь ми

просѧщꙋ .

LT/C 29 28 оу҆мръ́щбенїа

ризѫ прїеⷨ .

LT/D 30 29 своимь

бестⷭ҇рїе҆мь

LT/E 64 63 л҃д [34] блю҆ди д҃ше

моѧ .

LT/F 67 66 въ тъжⷣе срⷣѣ

LT/G 301 300 рн҃а [151] быва

ѫть г҃лѧ

LT/H 302 301 рн҃в [152] и҅деть . въсѫ

землѧ

LT/I 303 302 рн҃г [153] къ дх҃овныⷨ

подвигомь

LT/J 304 303 рн҃д [154] въ раѝ пребыⷭ .

LT/K 305 304 рн҃е [155] чимь . даже и҆

до шествїа

LT/L 306 305 силаеть свое҅го

LT/M 307 306 гшꙋ . иDже и҆з

ѡ҆ного

LT/N 308 307 б҃ъ к тебѣ

посла мѧ .

LT/O 309 309 єⷭс и҆стинно

зримое .

LT/P 310 – дѫ . бжⷭтвныи

моиси Table 2. Lenten Triodion, sixteen single leaves

11

The leaf nos. refer to a non-original foliation of a copy at the Russian State Library, Mos- cow (cf. Nemirovskij 1996, 126, copy 3.8; Nemirovskij 2009b, 85f, copy 3.6), a digital surro- gate of which is available at http://dlib.rsl.ru/viewer/01003441762.

12

The leaf nos. refer to a non-original foliation of a copy at the National Library of Poland,

Warsaw (shelf mark SD Inc.F.1349, cf. Nemirovskij 1996, 135f, copy 3.31), a digital surro-

gate of which is available at http://www.polona.pl/item/412064/3/.

Fragm.

leaf no.

Leaf no. in Biblio- teka Narodowa copy (digital surrogate)

Signature in lo- wer right corner of recto page

First words on line 1 of recto page

FT1/A 71r л҃ѕ [36] нїи прои҆зволениє҅мь :~

FT1/B 72r л҃з [37] сFъ же роуцѣ

ѡDмывааше .

FT1/C 73r л҃и [38] тъжде аDнтифѡнь г҃

глась, в҃ .

FT1/D 74r л҃ѳ [39] хлѣбь насытившаго

мнѡ҆жьство .

FT1/E 75r м҃ [40] вь єⷭ҇ и҆ спастн [sic]

могыи .

FT1/F 76r м҃а [41] днⷭ҇с црковнаа завѣса .

FT1/G 77r сти намь плода

животнаго .

FT1/H 78r слава троⷰ҇ оц҃а и҆ сн҃а и҆

дх҃а

FT1/I 79r вань бысть .

FT1/J 80r є҆моу крⷭ҇ть .

FT1/K 81r ѡ҆на сьдръжить

таинество .

FT1/L 82r чныи цр҃ь новы иD

адамь .

FT1/M 303r р҃Ѯ [160] сныхь намь и҆збавити

сѧ

FT1/N 306r є҆си оуDтѣшителѣ

нбⷭ҇ныѫ Table 3. Festal Triodion, fourteen leaves from copy “FT1”

2.3 Festal Triodion, before 1494

Sixteen single leaves (“FT1/A–N, FT2/A–B”) from two different copies of a Festal Triodion, printed in Cracow in c. 1493 (cf. Karataev 1861, no. 5, 1883, no. 5; Bošnjak 1968, 166f, no. 48 “Triod´ cvětnaja”; Nemirovskij 1996, 139–166, no. 4; Nemirovskij 2009a, 208–214, no. 4 “Triod´ cvetnaja”;

13

The leaf nos. refer to a non-original foliation of a copy at the National Library of Poland,

Warsaw (shelf mark SD Inc.F.1350, cf. Nemirovskij 1996, 163, copy 4.23), a digital surro-

gate of which is available at http://www.polona.pl/item/416187/8/.

ISTC it00428000; GW M47500; PI, no. B4),

see Tables 3 and 4. The pre- served leaves include all leaves from the original seventh gathering (ff.

FT1/A–L) and two leaves from the twenty-sixth gathering (ff. FT1/M–N).

The preserved leaves also include the first and last leaf of the seventh gathe- ring of a second copy of the same edition (FT2/A–B).

Fragm.

leaf no.

Leaf no. in Biblio- teka Narodowa copy (digital surro- gate)

Signature in lower right corner of recto page

First words on line 1 of recto page

FT2/A 71r л҃ѕ [36] нїи прои҆зволениє҅мь :~

FT2/B 82r чныи цр҃ь новы иD адамь .

Table 4. Festal Triodion, two leaves from copy “FT2”

3. Provenance

Currently, all thirty-four leaves are preserved in a box with the label “Depo- sition från KB, A 103, 3 (“Ur 32:92”)”. The text indicates that the leaves were earlier preserved at the National Library of Sweden.

However, as it has now been possible to ascertain with a high degree of probability, the leaves were earlier used as filling in the binding of a copy in the Uppsala University Library of Apuleius’ Asinus Aureus, printed in Bologna in 1500.

On the inside of the front cover of the Uppsala copy there is a

14

According to Lukjanenko (1993, 19) and Nemirovskij (2009b, 87), the edition includes two states, with differences between the two states reported for ff. 240–241, 249, and 267–270.

Błońska (1968, 51f) gives examples of textual differences on ff. 240r, 267r, and 268r, and provides facsimiles of f. 240r in a Warsaw copy (“wariant A”) and a Cracow copy (“wariant B”) (see insert between pp. 56 and 57). However, none of the indicated readings are repre- sented among the Stockholm leaves, and it has thus been impossible to identify them as be- longing to either of the two states.

15

Cf. Nemirovskij 1996, 274 for an overview of all gatherings.

16

The fact that the leaves include two copies each of the first and last leaf of the seventh gathering, indicates that the sixteen leaves originate from at least two copies of the Festal Triodion edition, but the exact distribution of the preserved leaves between the two copies has not been possible to define, and all leaves except the two “doublet” leaves have been consid- ered to belong to the “FT1” copy.

17

The abbreviation “KB” here clearly refers to “Kungliga biblioteket” ‘The Royal Library’, the Swedish name for the National Library of Sweden.

18

I wish to thank Jan Brunius at the Swedish National Archives, Anna Wolodarski at the National Library of Sweden, and Håkan Hallberg at Uppsala University Library for help with tracing the path of the leaves from Uppsala University Library to the National Library, and from there to their present location at the National Archives.

19

Current shelfmark “Copernicana 30”, see http://libris.kb.se/bib/8929772, Czartoryski 1978, 374, no. 23. Undorf 2012, 75, no. 316, lists the Uppsala copy as “incunable 143” (cf. Collijn 1907), providing only general information on the binding: “Gilt-stamped tooled brown calf”.

For the edition, see GW no. 2305

handwritten note in Swedish: “Ur detta bands pärmar har vid obekant tid- punkt (ca. 1905?) ett 40-tal blad äldre kyrkslaviska tryck uttagits (inlagda i fragmentsamlingen)”.

As is clear from other inscriptions in the Uppsala copy, the volume was earlier owned by the chapter library of Frauenburg (Frombork) in Poland, and it was included in the book collection taken in Frauenburg by Swedish troops as a war trophy in 1626 (cf. Collijn 1907, no.

143). According to Kolberg (1913, 110f, with a reference to Collijn 1907), the Apuleius copy had been donated to the Frauenburg chapter library by Johannes Langhannius (died 1567). Whether the incunabula leaves were put in the binding by Langhannius or by an earlier owner of the book is, unfortu- nately, not possible to verify. However, it seems reasonable to assume that the insertion of the Cyrillic leaves printed in Cracow into the binding of the Bologna copy must have taken place some time between 1500 and 1567, presumably in Poland.

4. Conclusions

The identification of the Stockholm/Uppsala leaves makes it possible for the first time to add three Cyrillic incunabula editions from the 1490s to the in- ventory of known incunabula in Sweden (cf. Undorf 2012, 2013/14). The description of the Stockholm/Uppsala leaves also draws attention to a speci- fic “path of acquisition” of early printed books, namely, fillings in bindings of other books (sometimes also serving as complete bindings). This acqui- sition method is well known for fragments of parchment manuscripts, in- cluding Cyrillic manuscripts,

but has been more seldom described when it comes to early prints.

The identification process also highlights the im- portance of recent digitalisation projects undertaken by institutions in Rus- sia, Poland, and other countries. These efforts now make it possible to access a considerable number of digital surrogates of early Cyrillic printed books, some of which are comparably rare, through the Internet.

Finally, the identification of the Stockholm/Uppsala Cyrillic in- cunabula leaves offers an example of the positive effects of the collaboration between researchers at different universities, libraries, and archives that has been activated through the Digitalised Descriptions of Slavic Cyrillic Manu- scripts and Early Printed Books in Swedish Libraries and Archives project.

(http://www.gesamtkatalogderwiegendrucke.de/docs/GW02305.htm), ISTC no. ia00938000 (http://istc.bl.uk/search/search.html?operation=record&rsid=127044&q=0).

20

‘At an unknown time (c. 1905?), some 40 leaves from older Church Slavonic printed books were extracted from the binding of this copy (now placed in the fragments collection).’

21

For an overview of the Swedish acquisition of the Frauenberg chapter library, including the Langhannius collection, see Walde 1916, 72–82.

22

Cf., for example, Steensland 2005, Brunius 2013, 68–70, Pereswetoff-Morath 2013, 2014, 298–300.

23

Cf., however, Wolodarski 2013, 210 and passim (Wolodarski does, however, not refer to

any Cyrillic fragments).

It has tried to expand the investigation of Cyrillic early prints and manu- scripts beyond the separate archival institutions, striving to create a union catalogue of the early Cyrillic manuscripts and prints presently located in Sweden.

References

Ambrosiani, Per. 2012. “Slaviska språk i en hyllningsbok till Gustav III.”

Slovo. Journal of Slavic Languages and Literatures 53:7–28.

[Available at http://www2.moderna.uu.se/slovo/Archives/2012- 53/PerAmbrosiani.pdf, accessed January 23, 2015]

Ambrosiani, Per. 2014a. “ЯГЛАБА, МРАВЕНЕКЧ, ГРОМГАДТДОС.

Cyrillic Words in Leonhard Thurneysser’s ‘Melitsah’ (Berlin, 1583)”.

In Med blicken österut. Hyllningsskrift till Per-Arne Bodin (Stock- holm Slavic Papers 23), edited by Per Ambrosiani, Elisabeth Löf- strand and Ewa Teodorowicz-Hellman. Skellefteå: Artos & Norma, 2014, 33–42.

Ambrosiani, Per. 2014b. “Types of Books and Types of Records: A Short Presentation of the CGS Database of Cyrillic and Glagolitic Books and Manuscripts in Sweden”. Scripta & e-Scripta 13:9–24.

Ambrosiani, Per. 2014c. “Johan Gabriel Sparwenfelds kyrilliska och glago- litiska 1600-talstryck utgivna av Sacra Congregatio de Propaganda Fide i Rom”. Slovo. Journal of Slavic Languages and Literatures 55:9–17. [Available at

http://www2.moderna.uu.se/slovo/Archives/2014- 55/2_Ambrosiani.pdf, accessed January 23, 2015]

Ambrosiani, Per and Antoaneta Granberg. 2010. “Slavonica Glagolitica and Cyrillica in Swedish Repositories. The Project Digitalised Descrip- tions of Slavic Cyrillic Manuscripts and Early Printed Books in Swe- dish Libraries and Archives”. Slovo. Journal of Slavic Languages and Literatures 51:107–113.

[Available at http://www2.moderna.uu.se/slovo/Archives/2010- 51/Per%20Ambrosian%20Antoaneta%20Granberg-slutversion.pdf, accessed January 23, 2015]

Bošnjak, Mladen. 1968. A Study of Slavic Incunabula. Zagreb: Mladost.

Brunius, Jan. 2013. From Manuscripts to Wrappers. Medieval Book Frag- ments in the Swedish National Archives (Skrifter utgivna av Riksarki- vet 35). Stockholm: The Swedish National Archives.

Collijn, Isak. 1907. Kataloge der Inkunabeln der schwedischen öffentlichen Bibliotheken. II. Katalog der Inkunabeln der Kgl. Universitäts-Biblio- thek zu Uppsala. Uppsala: Uppsala universitetsbibliotek.

Czartoryski, Pawel. 1978. “The library of Copernicus”. In Science and His- tory: Studies in Honor of Edward Rosen (Studia Copernicana 16).

Warszawa: Ossolineum, 355–396.

GW = Datenbank Gesamtkatalog der Wiegendrucke, http://www.gesamtkatalogderwiegendrucke.de.

Heintsch, Karol. 1957. Ze studiów nad Szwajpoltem Fiolem. Cz. 1: Mate- riały do życiorysu i działalności Fiola. Wrocław: Zakład Narodowy im. Ossolińskich.

ISTC – British Library Incunabula Short Title Catalogue, http://istc.bl.uk.

Karataev 1861: Каратаевъ, И. Хронологическая роспись славянскихъ книгъ, напечатанныхъ кирилловскими буквами. 1491–1730.

Санктпетербургъ: Императорская Академія Наукъ.

Karataev 1883 = Каратаевъ, И. Описанiе славяно­русскихъ книгъ, напе- чатанныхъ кирилловскими буквами . Т. 1. Съ 1491 по 1652 г (Сборникъ Отдѣленiя русскaго языка и словесности Импера- торской Академiи Наукъ, 34:2). Санктпетербургъ: Императорская Академія Наукъ.

Kjellberg, Lennart. 1951. Catalogue des imprimés slavons des XVIe, XVIIe et XVIIIe siècles conservés à la Bibliothèque de l’Université royale d’Uppsala. Uppsala: Uppsala universitetsbibliotek.

Kolberg, Josef. 1913. “Die Inkunabeln aus ermländischem Besitze auf schwedischen Bibliotheken”. Zeitschrift für die Geschichte und Alter- tumskunde Ermlands 18:94–137.

Lukjanenko 1993 = Лукьяненко, В.И., Издания кириллической печати XV–XVI вв. (1491–1600 г.). Каталог книг из собрания ГПБ. Санкт- Петербург: Российская национальная библиотека.

Nemirovskij, Evgenij L. 1996. Gesamtkatalog der Frühdrucke in kyril- lischer Schrift. Band I. Inkunabeln. Baden-Baden: Koerner.

Nemirovskij 2009a = Немировский, Евгений Л., Славянские издания кирилловского (церковнославянского) шрифта 1491–2000. Т. 1:

1491–1550. Москва: Знак.

Nemirovskij 2009b = Немировский, Евгений Л., Книги кирилловской пе- чати 1491–1550. Каталог (Книжные памятники Российской государственной библиотеки). Москва: Пашков дом.

Pereswetoff-Morath, Alexander. 2013. “Klinckowströms samling: Stafsunds slotts och Kungliga bibliotekets kyrkoslaviska pergamentsamlingar och deras tillkomst”. Slovo. Journal of Slavic Languages and Litera- tures 54:114–132.

[Available at http://www2.moderna.uu.se/slovo/Archives/2013- 54/8_Pereswetoff.pdf, accessed December 7, 2014]

Pereswetoff-Morath, Alexander. 2014. “Ord och blad. Martin Aschaneus och allt det ryska”. In Med blicken österut. Hyllningsskrift till Per- Arne Bodin, edited by Per Ambrosiani, Elisabeth Löfstrand and Ewa Teodorowicz-Hellman. Skellefteå: Artos & Norma, 291–307.

PI = Granberg, Antoaneta, Per Ambrosiani, Irina Lysén and Alexander Pere-

swetoff-Morath. 2010. Preliminary Inventory of Slavic Cyrillic and

Glagolitic Manuscripts and Early Printed Books in Sweden. Göte- borg: Göteborgs universitet.

Steensland, Lars. 2005. “Trash and Treasure. Russian Parchment Fragments in Swedish Archives”. In Medieval Book Fragments in Sweden. An In- ternational Seminar in Stockholm 13–16 November 2003 (KVHAA Konferenser 58), edited by Jan Brunius. Stockholm: Kungl. Vitterhets Historie och Antikvitets Akademien, 210–225.

Undorf, Wolfgang. 2012. Catalogue of books printed in the 15th century in Swedish collections. Wiesbaden: Harrassowitz.

Undorf, Wolfgang. 2013/14. “Svensk inkunabelcensus. En inventering av 1400-talstryck i Sverige”. Biblis 64:73–78.

Walde, Otto. 1916. Storhetstidens litterära krigsbyten. En kulturhistorisk- bibliografisk studie, 1. Uppsala: Almqvist & Wiksell.

Wolodarski, Anna. 2013. “Klemming och jakten på de försvunna inkunab- lerna”. In Fragment ur arkiven. Festskrift till Jan Brunius, edited by Marie Lennersand, Åsa Karlsson and Henrik Klackenberg. Stock- holm: Riksarkivet, 191–210.

Wronkowska-Dimitrowa, Mirosława. 2010. Triod kwietny z krakowskiej ofi-

cyny Szwajpolta Fiola (1491 r.). Bydgoszcz: Uniwersytet Kazimierza

Wielkiego.

Death of a Construction: Old Church Slavonic Touch Verbs

Hanne Martine Eckhoff, UiT The Arctic University of Norway

In this article I will use an usage-based constructionist approach to examine an argument structure construction on its deathbed, namely, the Old Church Slavonic (OCS) bare locative argument construction. Why does the con- struction even exist when the pattern it instantiates is otherwise dead? What happens to the verbs involved, and which argument structure constructions take over?

! I take a usage-based approach to syntactic productivity to examine what happens in such cases. In many ways, this approach can account for the observed development. The final stronghold of the locative construction was a small group of highly semantically coherent verbs. When the con- struction was lost, the verbs joined other argument constructions, preferably constructions that were also highly semantically coherent, but not necessa- rily very type frequent. However, syntactic productivity alone cannot explain the patterns of attraction: we must also look at the semantic and formal fit.

1. The locative construction

At the time of the earliest attestation of Slavic, OCS, we find the bare loca- tive on its way out of the Slavic languages altogether. It has mostly lost its spatial meanings and most of its (other) adverbial usages (Bauer 1963) – and is generally being replaced by preposition + locative.

! Nonetheless, a bare-locative argument structure construction does still

exist. A small set of verbs denoting ‘touching’ are the prime collocates with

this construction: prikosnǫti sę, prikasati sę and kosnǫti sę. The locative ar-

gument structure construction shows clear signs of decay in OCS. However,

unlike in its adverbial usages, the locative is not being replaced with preposi-

tion + locative. Instead, we see that the locative is partially homonymous

with both the genitive and the dative, leaving us with a number of uncertain

attestations. In addition, there are a number of unambiguous attestations of

genitive and dative constructions. The bare locative construction also com-

petes with various PP constructions, but these do not contain a locative nom-

inal.

What has saved the locative construction from disappearing altogether? I will explore the following four factors:

(1) The verbs in this class are semantically very similar. They all have semantics revolving around contact, approximation (movement up to) or proximity (location near)

(2) The verbs also have very similar form (frequently the pri prefix, to some extent the na prefix)

(3) The prefixation is semantically motivated and corresponds to locative prepositions

(4) Many of the verbs are reflexive, which precludes the regular transitive construction.

2. Data

This article is based on data from the Codex Marianus as found in the PROIEL/TOROT treebanks,

encompassing 58,357 word tokens. The tree- banks are lemmatised and have comprehensive annotation for morphology and syntax, which enables us to extract argument structure frames for every verb occurrence and to identify reflexive verbs. The text is also tagged for various semantic and derivational features. The Marianus is aligned

at token level with the Greek New Testament (Tischendorf 1869–1872). The Greek text has verb class tags: all Greek verb lemmas have been classified into 58 verb classes. This classification may be transferred to the OCS text via the token alignments insofar as the OCS verbs are aligned with Greek verbs (this holds for 13,505 out of 14,779 verb tokens in the Marianus data set).

The locative construction has very low frequency in this material:

There are only 40 occurrences of verbs with a locative oblique argument (alone or with some other argument). 24 of these are touch verbs, out of a total of 34 touch verb occurrences, all of which render Greek haptomai

‘touch’, which takes a genitive oblique argument. Thus, the locative occur- rences themselves do not lend themselves to statistical approaches. What we can do, however, is to situate the locative construction relative to its compet- itors, and see how it is similar and how it differs from them.

1

TOROT (The Tromsø OCS and Old Russian Treebank, https://nestor.uit.no) is an expansion of the Slavic part of the PROIEL corpus (Pragmatic Resources in Old Indo-European Lan- guages, foni.uio.no:3000). The Codex Marianus (edition: Jagić 1883) was fully annotated in PROIEL, but is also included in TOROT along with the Greek Gospels. The data were all drawn from TOROT in 2014. All data and an R script necessary to replicate this study are publicly archived at the Tromsø Repository for Language and Linguistics (opendata.uit.no) with the permanent URL http://hdl.handle.net/10037.1/10180

2

The texts were automatically aligned, and the alignments were then hand-corrected.

3. Tools from the usage-based constructionist toolbox

In this article I will use tools and ideas from usage-based construction gram- mar. The usage-based model, as developed by Bybee (1985, 1995), is funda- mental to Langacker’s Cognitive Grammar (Langacker 1987, 1991) and the construction grammars of e.g. Croft (2001) and Goldberg (1995, 2006). It assumes linguistic knowledge to be organised in schematic networks, which emerge through actual language use and experience, and generalisations over this input. The networks are bottom-up; much linguistic knowledge is assumed to be completely specific, or to consist of lower-level generalisa- tions in the shape of partially specific schemas. Generalisations on higher levels – fully schematic constructions – are also assumed to exist, but only when motivated by high type frequency. The level of generalisation and the strength of each node is deemed to be an empirical question – it depends on the construction’s actual usage frequency (Langacker 1987, 46; Croft 2001, 28). The approach is well suited to a study of the relative strength and pro- ductivity of argument structure constructions, where we frequently see a highly type frequent and schematic construction (the transitive construction is the example par excellence) coexisting and competing with several con- siderably less type frequent and schematic argument structure constructions.

It is especially well suited to look at such situations diachronically: how constuctions lose and gain schematicity and type frequency as verbs aban- don one construction and join another (see e.g. Eckhoff 2009).

3.1 Syntactic productivity

The subject matter of this article – the death of an argument structure and the

subsequent fate of the verbs that once belonged to it – is clearly a story about

syntactic productivity (or the lack thereof). Barðdal (2012) takes the usage-

based constructionist approach as her point of departure, and sees syntactic

productivity as a function of type frequency and semantic coherence: A

highly type frequent construction must necessarily be semantically very

open, and thus is easily extensible to many items. If a construction is less

type frequent, on the other hand, it needs to be semantically coherent in or-

der to be able to attract new (and semantically compatible) items.

Figure 1. The relationship between type frequency and semantic coherence in syntactic productivity, from Barðdal 2012.

By extension, this notion of syntactic productivity should also be able to tell us something about construction death. In our case, the locative construction has very low type frequency, but strong semantic (and phonological) cohe- rence, i.e. it is found in the lower right corner of Figure 1, where Barðdal places a question mark. If Barðdal’s notion of syntactic productivity holds, we can predict that the locative construction should be only marginally pro- ductive, if at all, and very vulnerable to more productive patterns

3.2 Semantic representations

In order to make judgements about semantic coherence and semantic fit, we need an apparatus for analysing verb semantics. I will informally model events in terms of causal chains, following Croft’s (2012) formulation of a force dynamics approach to verb semantics. His approach rests on the Cau- sal Order Hypothesis: “a simple verb in an argument structure construction construes the relationships among participants in the event it denotes as forming a directed, acyclic and nonbranching causal chain” (Croft 2012, 221–222). Hence, even though not all verbs denote causal, acyclic and non- branching events, they are nonetheless conceptualised as such if they are realised as simple verbs. Thus, argument structure constructions conform to the Causal Order Hypothesis regardless of the event type. Croft proposes the following linking rules:

Regularity-Generality-Open Schema

Different Degrees of Productivity

?

(1) The verbal profile is delimited by Subject and Object (if any) (2) Subject is antecedent to Object in the causal chain: SBJ → OBJ (3) Antecedent Oblique is antecedent to the Object in the causal chain;

Subsequent Oblique is subsequent to the Object in the causal chain:

A.OBL → OBJ → S.OBL

(4) Incorporated arguments are between Subject and Object in the causal chain: SBJ → INCORP → OBJ

The causal chain is simple: the subject will act on the antecedent oblique, the antecedent oblique will act on the object, and the object will act on the sub- sequent oblique. In Croft’s example “Sue broke the coconut for Greg with a hammer”, the causal chain looks like this:

Sue → hammer → coconut → Greg

SBJ A.OBL OBJ S.OBL

Cross-linguistically, oblique markers will generally encode either antecedent obliques or subsequent obliques – but one marker will rarely cover both types (Croft 2012, 223). Table 1 lists roles that are typically marked as ante- cedent and subsequent obliques respectively cross-linguistically.

Causal roles Spatial metaphor Noncausal roles Antecedent

oblique

passive agent, cause, comitative, instrument, man- ner, means

ablative figure, posses- sum

Subsequent oblique

result, benefici- ary, “maleficiary”

allative ground, posses- sor, recipient Table 1. Summary of major antecedent oblique and subsequent oblique roles, (after Croft 2012, 275)

Naturally, as table 1 indicates, many events are non-causal, and hence not naturally directed, for instance spatial or possessive relations, or mental events. Such events are construed in terms of a causal chain too, but since there is no natural directionality, they are more open to multiple construals.

For spatial and possessive events, there are robust cross-linguistic patterns of construal (figure before ground, possessum before possessor). In mental events, on the other hand, the event is essentially bidirectional: the experi- encer directs its attention to the stimulus, and the stimulus changes the men- tal state of the experiencer. Thus, either the experiencer or the stimulus may be construed as the origin of the force transmission (Croft 2012, 233–234).

It is also important to note that many events are cyclic, crucially this is

the case for reflexives and reciprocals (Croft 2012, section 6.2.3.2). In a true

reflexive event (“He nominated himself”), the same participant is the initia-

tor and the endpoint simultaneously. In reciprocal events (“They congratu- lated each other”), there are two events of the same type, where the force transmission goes in two directions. Participant 1 is the initiator of event 1 and the endpoint of event 2, and vice versa. However, such events are also construed as acyclic. There are several ways of doing this. Many languages have grammaticalised a middle marker for this purpose (Kemmer 1993), and this is the case for OCS as well: the reflexive clitic sę is transparently an ac- cusative form of the reflexive pronoun sebe, but it has a distribution that goes well beyond true reflexives (passives, reflexiva tantum).

Reflexive verbs with middle markers may be conceptualised as “in- transitive” (i.e. no force transmission), but also as two-participant events: the initiator acts on an endpoint but is her/himself affected by the outcome. In OCS, this endpoint cannot be encoded as an object, but must be expressed by an antecedent or subsequent oblique. The middle marker may also be used to signal the reciprocality of an event: even though one participant is selected as the initiator and the other as the endpoint, the middle marker sig- nals that the force transmission is really bidirectional.

4. The core examples: OCS touch verbs

4.1 The examples

24 of 38 locative construction occurrences in the data set occur with touch verbs. If we count reflexive and non-reflexive verbs as separate lemmas, there are four such verbs: kosnǫti , kosnǫti sę, prikosnǫ̨ti sę and prikasati sę.

All four mean ‘touch’, and they almost exclusively translate Greek haptomai (+ genitive) ‘touch’.

kosnǫti kosnǫti sę prikosnǫ̨ti sę prikasati sę

locative construction 0 4 14 6

genitive construction 0 0 1 0

vъ+acc. construction 2 0 0 0

transitive construction 6 0 0 0

object + vь+acc con- struction

1 0 0 0

Table 2. Touch verb distribution across argument structure constructions As we see in table 2, the three reflexive verbs primarily occur in the locative construction (1–3).

(1) kosnǫ sę vьskrilii rizy ego

touched REFL edge. LOC cloak. GEN his

hēpsato tou kraspedou tou himatiou autou touched the edge. GEN the cloak. GEN his

‘she touched the edge of his cloak’ (Codex Marianus, Luke 8:44) (2) kto prikosnǫ sę rizaxъ moixъ

who touched REFL robes. LOC my. LOC tis mou hēpsato tōn himatiōn who my touched the robes. GEN

‘Who touched my robes?’ (Codex Marianus, Mark 5:30) (3) vьsь narodь iskaaše prikasati sę emь

whole people sought touch REFL him. LOC

pas ho okhlos ezētoun haptesthai autou all the crowd sought touch him. GEN

‘all the crowd sought to touch him’ (Codex Marianus, Luke 6:19) Note that the locative is frequently homonymous with either the dative or the genitive. The single genitive construction occurrence in table 2 is unambigu- ous (4), but there are two examples counted as locatives which may as well be genitives (5) and nine that could be datives (6).

(4) ašte tokmo prikosnǫ sę rizy ego spsna bǫdǫ if only touch REFL robe. GEN his saved will-be ean monon hapsōmai tou himatiou autou sōthēsomai if only touch the robe. GEN his will-be-saved

“If I only touch his cloak, I will be healed.” (Codex Marianus, Matt.

9:21)

(5) i pristǫpь is prikosnǫ sę ixъ and having-approached Jesus touched REFL them. LOC / GEN

i reče and said

kai prosēlthen ho Iēsous kai hapsamenos autōn and approached the Jesus and having-touched them. GEN

eipen said

“And Jesus came to them and touched them and said” (Codex Maria- nus, Matt. 17:7)

3

The locative-dative ambiguities are probably best interpreted as locatives – while there is

evidence of unambiguous genitive arguments with touch verbs both in the Marianus and

elsewhere in the OCS canon (there are several occurrences in the Suprasliensis), no unambig-

uous dative arguments were found neither in the Marianus nor in the Zographensis or Su-

prasliensis (data from TOROT). There are, however, attestations with kъ+dative goal

arguments in the Suprasliensis.

(6) ne prikasai sę mьně not touch REFL me. LOC / DAT

mē mou haptou not me. GEN touch

“Do not touch me” (Codex Marianus, John 20:17)

Non-reflexive kosnǫti, on the other hand, never occurs in the locative con- struction in the Marianus material.

It is mostly found in the transitive construction (7), but also with a vъ+accusative goal argument, with (8) or without (9) a direct object.

(7) isъ kosnǫ i Jesus touched him. ACC

hēpsato autou touched him. GEN

‘Jesus touched him’ (Codex Marianus, Matt. 8:3) (8) tъgda kosnǫ ě vъ oči

then touched them. ACC in eyes. ACC tote hēpsato tōn ophthalmōn autōn then touched the eyes. GEN they. GEN

‘Then he touched their eyes’ (Codex Marianus, Matt. 9:29) (9) kosnǫ vъ odrъ

touched in stretcher. ACC

hēpsato tēs sorou touched the stretcher. GEN

‘He touched the stretcher’ (Codex Marianus, Luke 7:14)

In every cited example, we see that the Greek has haptomai ‘touch’ with the touched entity occurring as a genitive oblique, and this is the case in virtu- ally all of the other occurrences as well. In other words, where the Greek makes do with a single verb and a single argument structure construction, OCS differentiates considerably: there is a choice of four different (albeit related) verbs and four to six different argument structure patterns. What semantic differences can we find between the various OCS constructions?

4

The same holds for the Codex Zographensis and Codex Suprasliensis, which are also found in TOROT.

5

The choice between prikasati sę and prikosnǫti sę is in all likelihood aspectual (see Eckhoff

and Haug 2015 and references within for a comprehensive debate on the expression of aspect

in OCS). The aspectual distinction does not seem to be relevant to the argument structure

choice here.

4.2 Touch verb semantics

In terms of the causal chain approach, it is somewhat surprising that the touch verbs are not uniformly transitive in OCS (and never transitive in Greek). Intuitively, this seems like a straightforward case of two-participant force transmission, which should be conceptualised with an initiator subject and an endpoint object (as in example 7):

toucher → touchee SBJ OBJ

In OCS, most manipulation verbs are transitive. What is it about the touch verbs that make them different? Why are they not consistently transitive like most closely related verbs? Why is there an alternation between a transi- tive construction and a (family of) reflexive oblique construction(s)? Does this alternation carry meaning?

There are several reasons why touch verbs are not typical impact verbs. Note that in FrameNet (https://framenet.icsi.berkeley.edu/fndrupal/), the English verb touch is not primarily assigned to the Impact frame, but to the Manipulation frame, probably mostly due to the range and properties of possible participants. The frame states that the manipulation is performed with a body part, which is thus a natural antecedent oblique for the verb. It also states that the patient is affected, and may, but needs not undergo a change, suggesting that the patient is a less typical object. The frame also lists Locus as a non-core element: “The Agent's manipulation of an Entity may be further specified as being localized to some part of the Entity, a Lo- cus.” With touch and other manipulation verbs, the Locus is often expressed alongside the Entity, as in She touched him on the nose (so-called possessor raising). The verbs assigned to this frame naturally all entail touching, such as nudge, kiss, pinch, rub. They usually affect their Patients in the sense that the Patient senses the manipulation (with pain, pleasure, irritation etc.), but they rarely change them. Note that touch itself is clearly one of the less inva- sive verbs in this group: the patient may be minimally affected and may not even notice being touched.

Touch is also assigned to the Locative_relation frame. Only occur- rences of static touching belong to this frame, but the fact that static touch- ing situations exist, points to another important property of touch verbs: the force transmission may be minimal or even non-existing. This is also a fac- tor that disfavours the straight initiator-endpoint conceptualisation.

Finally, we should take into consideration the fact that touching is not

necessarily an acyclic event, in fact it is inherently reciprocal: what you

touch will necessarily touch you as well. Although the event is typically

asymmetrical (the initiator is in control), it is actually more likely that the

toucher has a sensory impression of the touch event than that the touchee has one. It seems likely that this is the motivation for the fact that both OCS and Greek have touch verbs with middle markers.

Touch situations, then, typically have weak, but bidirectional force transmission and little-affected patients. Greek solves this by sticking to a single reflexive verb that takes a genitive oblique in all contexts. In OCS, however, several reflexive oblique constructions alternate with the transitive construction. Does this alternation carry meaning?

As it turns out, as far as we can tell from our sparse data, it does. All save one of the examples of non-reflexive kosnǫti are used in contexts where Jesus is touching people to heal or bless them. In example (7), he heals a man from leprosy, and in example (8), he makes two blind men see. The only exception is an example where Jesus touches a stretcher, but even then he very soon raises the man lying on the stretcher from the dead (9), which suggests that this example may possibly belong to the touch-to-change class as well. Note also that (9) has no direct object.

touch to change touch, no change

non-reflexive 8 1

reflexive 3 21

Table 3. The touch alternation. P < 0.0002, Fisher’s exact test, two-tailed.

As table 3 shows, reflexive touch verbs are very rarely used in such contexts.

There are only two examples of reflexive prikosnǫti sę (10) and one of re- flexive kosnǫti sę with this kind of meaning in the data set.

(10) i prikosnǫ sę rǫcě eję i ostavi jǫ and touched REFL hand. LOC her and left her ognь

fire

kai hēpsato tēs kheiros autēs, kai aphēken autēn and touched the hand. GEN her and left her ho puretos

the fever

“He touched her hand, and the fever left her” (Codex Marianus, Matt.

8:15)

However, the vast majority of reflexive touch verbs are used in cases where the touching does not change the state of the patient and where the agent has no intention that it should. Quite the opposite, frequently, as in examples (1–

4), the agents are actually the ones hoping to be changed by the touch (and

some of them are even hoping that Jesus will not notice) – contexts where

the bidirectionality of touching is important.

In Croft's terms, then, touching is always construed as a directed, acy- clic causal chain in OCS, and the toucher is always construed as the initiator and subject. However, there is an alternation between a straight transitive construal where the verb has no marker and the patient is construed as an OBJ (11a: Jesus touched the leper), and a construal with a verb with a mid- dle marker and the patient construed as a subsequent oblique (11b): The woman touched Jesus.).

As far as we can see from the sparse data, (11a) is chosen when the toucher expects a real force transmission and a real change in the patient. 11(b), on the other hand, is chosen when force transmission is unimportant or the bidirectionality of force transmission is important, and when the toucher does not expect to affect the patient much.

(11) a. Jesus → leper (change state) SBJ OBJ

b. woman ⇄ Jesus SBJ S.OBL

Interestingly, (11b) covers the majority of situations, and is consistently real- ised as a locative subsequent oblique – a marginal pattern on the brink of extinction, shared by very few other verbs.

It is worth noting that not even non-reflexive kosnǫti patterns entirely with other transitive verbs with affected objects. The locus, if any, is consist- ently realised as a vъ+accusative PP (8). We never see the locus occurring as a direct object and the patient as an external possessor, as with prěbiti

‘break’ in (12).

(12) prъvoumu prěbišę goleni first. DAT broke shins. ACC

tou men prōtou kateaxan ta skelē the PTC first. GEN broke the shins. ACC

‘They broke the shins of the first one’ (Codex Marianus, John 19:32) The OCS external possessor construction seems to require that the locus of impact is a body part and that the locus itself goes through a physical change of state. The vъ+acc construction does apparently not require that the locus changes, and may be less restrictive.

5. The peripheral examples

Apart from the touch verbs, there are only 14 more examples of the locative construction in the Marianus data set. It is found with nine different verb

6

For the possibility of the genitive construction involving an antecedent oblique, see section

6.