aquamotion verbs in slavic and germanic

28
Uncorrected proofs - John Benjamins Publishing Company chapter 13 Aquamotion verbs in Slavic and Germanic A case study in lexical typology Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina Stockholm University / University of Sheffield, UK / Institute for the Russian Language, Moscow, Russian Federation e paper contrasts the verbs plyt’/plavat’ in Russian and płynąć/pływać in Pol- ish with their correspondences in Dutch, English and Swedish against a broader typological background. e three Germanic languages use several verbs for what is covered by a pair of derivationally related verbs in each of the two Slavic languages. e Germanic languages lexicalize the activity/passivity of motion, but vary considerably as to how they carve up the conceptual space. Russian and Polish, on the other hand, use plavat’/plyt’ independently of the activity/pas- sivity of motion and focus on the uni- or non-unidirectionality of the motion. Nonetheless, it appears that the different lexicalizations in the Swedish-English- Dutch systems of aquamotion verbs are reflected in constructional differences in the Russian-Polish systems. 1. Introduction e present paper views the verbs plyt’/plavat’ in Russian and płynąć/pływać in Polish against a broader typological background, stemming from a systematic lexical-typological study of more than 50 genetically, areally and structurally diverse languages (Maisak & Rakhilina 2007b). In order to highlight the main 1. e international project on aquamotion (2002–2006) supported by the Russian Foun- dation for the Humanities (RFBR#05-06-80400a) involved a systematic lexical-typological research based on a methodology (questionnaires, checklists and corpora) elaborated, tested and improved by a group of language experts, who collected and analyzed the data in close col- laboration with native speakers. e directors of the project were Timur A. Maisak (Institute of Linguistics, Russian Academy of Sciences) and Ekaterina V. Rakhilina (Institute for the Rus- sian Language, Russian Academy of Sciences), who also both carried out research on several languages. Maria Koptjevskaja-Tamm and Dagmar Divjak contributed to the project with the

Upload: birmingham

Post on 19-Nov-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

nychapter 13

Aquamotion verbs in Slavic and GermanicA case study in lexical typology

Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina RakhilinaStockholm University / University of Sheffield, UK / Institute for the Russian Language, Moscow, Russian Federation

The paper contrasts the verbs plyt’/plavat’ in Russian and płynąć/pływać in Pol-ish with their correspondences in Dutch, English and Swedish against a broader typological background. The three Germanic languages use several verbs for what is covered by a pair of derivationally related verbs in each of the two Slavic languages. The Germanic languages lexicalize the activity/passivity of motion, but vary considerably as to how they carve up the conceptual space. Russian and Polish, on the other hand, use plavat’/plyt’ independently of the activity/pas-sivity of motion and focus on the uni- or non-unidirectionality of the motion. Nonetheless, it appears that the different lexicalizations in the Swedish-English-Dutch systems of aquamotion verbs are reflected in constructional differences in the Russian-Polish systems.

1. Introduction

The present paper views the verbs plyt’/plavat’ in Russian and płynąć/pływać in Polish against a broader typological background, stemming from a systematic lexical-typological study of more than 50 genetically, areally and structurally diverse languages (Maisak & Rakhilina 2007b).� In order to highlight the main

1. The international project on aquamotion (2002–2006) supported by the Russian Foun-dation for the Humanities (RFBR#05-06-80400a) involved a systematic lexical-typological research based on a methodology (questionnaires, checklists and corpora) elaborated, tested and improved by a group of language experts, who collected and analyzed the data in close col-laboration with native speakers. The directors of the project were Timur A. Maisak (Institute of Linguistics, Russian Academy of Sciences) and Ekaterina V. Rakhilina (Institute for the Rus-sian Language, Russian Academy of Sciences), who also both carried out research on several languages. Maria Koptjevskaja-Tamm and Dagmar Divjak contributed to the project with the

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

316 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

typological properties of these Russian and Polish verbs we will contrast them with the corresponding verbs in three Germanic languages.2 For this purpose we will look briefly at Swedish simma, segla, driva, flyta etc., at swim, sail, drift and float in English, and at zwemmen, varen, drijven in Dutch (see Divjak & Lemmens 2007; Koptjevskaja-Tamm 2007; Golubkova & Rakhilina 2007 for detailed case-studies on these languages, and Divjak 2006 for a comparison of Dutch and Russian). All these verbs are primarily used for horizontal transla-tional (and sometimes stationary) motion of a non-liquid Figure in or on the sur-face of a liquid Ground (prototypically in water); this domain of Aquamotion is complementary to Terra-motion (motion on land) and to Aero-motion (mo-tion in the air). The verbs listed are opposed to mersion verbs (such as dive and sink), to self-contained motion verbs (bob) and to verbs expressing motion of the liquid itself (stream, flow).

As is clear from the above enumeration, all three Germanic languages use sev-eral different verbs for what is covered by a pair of derivationally related verbs in each of the two Slavic languages. Such cross-linguistic variation in how languages categorize or carve up the conceptual aquamotion domain lay at the core of the Aquamotion project (http://aquamotion.narod.ru) and cross-linguistic variation in lexicalization patterns will likewise constitute the main focus of the present paper. This difference is particularly interesting given that all five languages under scrutiny belong to the same type of satellite-framed languages in Talmy’s (2000) classification. This is in fact one of the reasons for why we have chosen to compare Slavic and Germanic instead of comparing Slavic with more typologically diverse languages: we aim to draw attention to the considerable diversity that co-exists with cross-linguistic similarities stemming from genetic and typological related-ness among languages.

descriptions of Swedish and Dutch aquamotion systems. We are also grateful for the INTAS support (05-�000008-79�7).

�. The data on the Slavic and Germanic languages discussed in the present paper come from various sources including several corpora (Russian National Corpus, http://www.ruscorpora.ru/; Korpus Języka Polskiego Wydawnictwa Naukowego PWN, http://korpus.pwn.pl/; the cor-pora at the Instituut voor Nederlandse Lexicologie in Leiden, www.inl.nl; the Parole corpus at the Swedish Language Bank, http://spraakbanken.gu.se). In working on the present paper we have benefited from the SALT Dutch-Swedish parallel corpus compiled by Gudrun Rawoens (Ghent University) who has kindly provided us with numerous examples. The examples for which the source is not explicitly given are constructed by the authors but have been checked with several native speakers. In some cases we have opted for this strategy to illustrate certain basic facts and simplify the presentation, although we are very much in favor of using authentic examples in our research.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 317

In this paper, we will show that Dutch, English and Swedish lexicalize the degree of activity/passivity of aquamotion (i.e., primarily Manner of motion) by means of different verbs, i.e., they encode the source of propulsion on the verb, leaving the interpretation of directionality to optional satellites or to contextual inference. However, the three Germanic languages vary considerably as to how exactly they carve up the conceptual space and what precisely they encode on the verb. Russian and Polish, on the other hand, use plavat’/plyt’ independently of the activity/passivity of motion involved, and focus on the uni-/non-directionality of the motion that has grammaticalized. Nonetheless, it turns out that the different lexicalizations in the Swedish-English-Dutch systems of aquamotion verbs are reflected in constructional differences in the Russian-Polish systems.

We will start from a general description of the aquamotion system in Swed-ish, English and Dutch (Section 2); this will introduce the general typological background relevant for the domain (Section 3) that will be used in the analysis of Russian and Polish. In Sections 5 and 6 we will look more specifically into the differences among the closely related languages. Section 6 will sum up the main findings illustrated throughout the paper.

�. A Germanic triad: Swedish, English and Dutch

Swedish has an elaborated lexical system for talking about the horizontal motion of a non-liquid Figure in or on the surface of a liquid Ground (typically water) that includes six designated (or proper) aquamotion verbs (simma, segla, driva, flyta, ro, paddla) and a number of general motion verbs.

Simma denotes active self-propelled aquamotion with an animate Figure accomplished by the Figure’s (controlled) movements of body and/or limbs. The typical Figures here are human beings, but many animals likewise swim (mam-mals, fish, water birds and snakes). Simma cannot be used whenever aquamo-tion involves a vessel and/or whenever the Figure’s own contribution to motion is minimal or non-existent (passive aquamotion); a limited contribution might be caused either by the fact that the Figure’s bodily movements are not (fully) re-sponsible for motion (for an animate Figure), or because the Figure is inanimate. Let’s consider these different cases in turn.

For aquamotion involving vessels the situation is quite complex. The most prominent verb here is segla (the cognate of sail), which is obligatorily used for motion of sailing boats and/or of people aboard them. Consider ex. (�a) in which the word båt ‘boat, ship’ is used as the subject to seglade ‘sailed’. Although båt can denote any (or most) kinds of vessel, in this example it unambiguously refers to a sailing boat because of the predicate – neither rowing boats, nor canoes or

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

318 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

motor-driven boats and ships “sail” in Swedish. Segla can be said both about moving sailing boats and about the people aboard, i.e., segla combines with dif-ferent kinds of Figure/subject (an inanimate Figure referring to a vessel vs. a human Figure). In addition, for human Figures there are different constructions available depending on whether the human being is actively navigating (with segla used as a transitive verb and the vessel encoded as its direct object, ex. (�b)) or is a passenger (with segla used intransitively and the vessel encoded as its co-mitative adjunct, ex. (�c)).

(�) a. Båt-en segla-de från G. till M. på e-n dag. boat-def.c.sg sail-pret from G. to M. in one-c day3 ‘The sailing boat sailed from G. to M. in one day.’ b. Jan segla-de vår båt från G. till M. på e-n dag. Jan sail-pret our.c.sg boat from G. to M. in one-c day ‘Jan sailed our sailing boat from G. to M. in one day.’ c. Jan segla-de med vår båt från G. till M. på en dag. Jan sail-pret with.our.c.sg boat from G. to M. in one-c day ‘Jan sailed with / on our sailing boat from G. to M. in one day.’

Different from sailing boats, neither rowing boats nor canoes can be construed in Swedish as actively moving Figures. Their motion can only be described indi-rectly, via reference to the corresponding activities (rowing – ro and paddling – paddla) of the human Agents aboard (2a). Strangely enough, motor-driven boats and ships do not have any designated aquamotion verbs at all; reference to their motion and to the motion of people aboard has to recruit generalized verbs of motion (åka/fara ‘to go by a vehicle’, gå ‘to go on foot’, komma ‘to come’) or even verbs that do not express motion at all (e.g., ta ‘to take’, etc.) (2b)).

(2) a. Vi/*Båt-en ro-dde / paddla-de över vik-en. we/*boat-def.c.sg row-pret / paddle-pret over bay-def.c.sg ‘We / *The boat rowed / paddled over the bay.’ b. Titta, vilk-en stor ångare som komm-er hit! look which-c.sg big.c.sg steamship rel come-pres here ‘Look what a big steamship is coming!’

Finally, passive aquamotion is covered by two verbs – driva and flyta. In both cases the Figure’s motion (or the absence of any significant motion) is to a large extent determined by the dynamics of its liquid Ground (prototypically water), sometimes in combination with another external force (typically the wind). Driva

3. Glossing conventions follow the Leipzig Glossing Rules available at http://www.eva.mpg.de/lingua/resources/glossing-rules.php.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 319

refers to passive motion that is completely out of control and often (but not nec-essarily) takes on an undesirable course and/or is chaotic in an unpredictable and disturbing way. The latter connotation is primarily present when the Figure is a vessel or a human being, for which controlled aquamotion is the default (or at least a desideratum). Flyta lacks negative connotations and is more neutral as to whether the motion is uncontrolled or controlled. What matters is that the Figure is situated in water or another liquid (typically on its surface) and moves together with it without sinking. Note that if motion is determined by the dynamics of the water, it is not necessarily translational, i.e., the Figure does not necessarily change its location. In the absence of a current, when the water itself is not mov-ing, objects on its surface will remain more or less located in one and the same place – such situations of location on the surface of water are also covered by flyta, cf. ex. (3). For inanimate Figures, except for vessels, passive aquamotion (includ-ing location on the surface of water) is the only possible kind of aquamotion; for vessels and animate Figures it competes with others, more active kinds of aqua-motion, cf. ex. (4)–(6).

(3) Location in/on the surface of water/liquid Död-a fisk-ar flyt-er vid strand-en. dead-pl fish-pl float-pres by shore-def.c.sg ‘Dead fish is floating by the shore.’

(4) Aquamotion of a human Figure Jan simma-r (a.) / seglar (b.) / flyt-er (c.) / driv-er (d.) Jan swim-pres (a.) / sail-pres (b.) / float-pres (c.) / drift-pres (d.) mot klipp-or-na. towards rock-pl-def.pl a. ‘Jan is swimming towards the rocks.’ b. ‘Jan is sailing (on) a sailing boat towards the rocks.’ c. ‘Jan is floating towards the rocks.’ d. ‘Jan is drifting towards the rocks.’

(5) Aquamotion of sailing boats as a Figure Båten *simma-r (a.) / seglar (b.) / flyt-er (c.) boat-def.c.sg swim-pres (a.) / sail-pres (b.) / float-pres (c.) / driv-er (d.) mot klipporna. / drift-pres (d.) towards rock-pl-def.pl a. *‘The sailing boat is swimming towards the rocks.’ b. ‘The sailing boat is sailing towards the rocks.’ c. ‘The sailing boat is floating towards the rocks.’ d. ‘The sailing boat is drifting towards the rocks.’

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�0 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

Tabl

e 1.

The

Swed

ish sy

stem

of a

quam

otio

n ve

rbs

Act

ive

mot

ion

of a

n an

imat

e Fi

gure

Mot

ion

of v

esse

ls a

nd p

eopl

e ab

oard

Pass

ive

mot

ion

Saili

ng b

oats

Mot

or-d

rive

n ve

ssel

sR

owin

g bo

ats

Can

oes

Mot

ion

out o

f co

ntro

lN

eutr

al m

otio

n /

Loca

tion

simm

ase

gla

(no

spec

ific a

qua-

mot

ion

verb

s)ro

(ind

irect

ly)

padd

la (i

ndire

ctly

)dr

iva

flyta

Tabl

e 2.

The

Engl

ish a

nd D

utch

syst

ems o

f aqu

amot

ion

verb

s

Act

ive

mot

ion

of a

n an

imat

e Fi

gure

Mot

ion

of v

esse

ls a

nd p

eopl

e ab

oard

Pass

ive

mot

ion

Saili

ng b

oats

Mot

or-d

rive

n ve

ssel

sR

owin

g bo

ats

Can

oes

Mot

ion

out o

f co

ntro

lN

eutr

al m

otio

n /

Loca

tion

Engl

ishsw

imsa

ilro

w (i

ndire

ctly

)pa

ddle

(indi

rect

ly)

drift

float

Dut

chzw

emm

enze

ilen

roei

en (i

ndire

ctly

)pa

ddele

n (in

dire

ctly

)dr

ijven

vare

n

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3�1

(6) Aquamotion of other inanimate Figures Isflak-en driv-er / flyt-er mot klipp-or-na. ice-floe-pl.def drift-pres / float-pres towards rock-pl-def.pl ‘The ice-floes are floating / drifting towards the rocks.’

The Swedish facts considered above are summarized in Table �.The aquamotion verb systems in the two related Germanic languages Eng-

lish and Dutch show distinctions akin to those in Swedish, but with somewhat reduced degrees of elaboration. Thus, the distinctions among swim, drift and float in English (Golubkova & Rakhilina 2007) are fairly similar to those among their cognates simma, driva and flyta in Swedish. English, however, has generalized the use of sail to cover most kinds of vessels (both sailing and motor-driven boats and ships), whereas the motion of rowing boats and canoes is treated indirectly, in the same way as in Swedish (row, paddle).

The Dutch system (Divjak & Lemmens 2007) is reduced compared to both Swedish and English in two respects. First, it has only one verb for passive aqua-motion (including location in water), drijven (cognate to drift and driva). And, second, it has one generalized aquamotion verb, varen, which covers all motion of vessels and people on vessels, even though there are also other optional verbs with more specific applications (zeilen, roeien, paddelen). The English and Dutch facts are summarized in Table 2.

The differences among Swedish, English and Dutch notwithstanding, all three languages clearly differentiate among several unrelated verbs, each re-sponsible for its own kind of aquamotion. The main distinctions are between active self-propelled motion of an animate Figure, active motion of vessels and people aboard (with further specifications of vessels) and passive motion of both animate and inanimate Figures (with location on the surface of water as the end-point of aquamotion).

Now, in what way are these Germanic facts relevant for a volume on motion verbs in Slavic? In order to see how, we first need to consider a typology of aqua-motion verbs.

3. From Germanic to typology: Sub-domains of aquamotion and aquamotion language systems

The Germanic facts considered in the previous section are highly relevant in a more general typological perspective as they find parallels in many languages. In fact, as shown in Maisak & Rakhilina (2007a), languages tend to carve up the aquamotion domain according to the distinctions illustrated by the three

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�� Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

Germanic languages, e.g., by using different lexemes and/or constructions for reference to different “kinds” of aquamotion. The main parameter underlying semantic categorization within the aquamotion domain across languages is the degree of activity/passivity of motion, which can be viewed as a scale with three main zones, or sub-domains, each centred on a particular prototype.

�. One sub-domain corresponds to active, self-propelled motion of animate Figures accomplished by the Figure’s controlled movements of body and/or limbs in water (swimming).�

2. Its opposite is passive uncontrolled and non-agentive aquamotion, where-by the Figure simply follows the movement of the water in which or on the surface of which it happens to be. Since passive Figures lacking control and agentivity are prototypically inanimate, passive aquamotion commonly in-volves inanimate Figures. And, since motion in these cases depends on the dynamics of the water, passive aquamotion can occasionally be stationary, whereby the Figure remains more or less located in one and the same place. Passive aquamotion is therefore particularly sensitive to the presence/absence of direction, with the concomitant distinction between the maximally passive location in or on the surface of water (floating) and passive, but directed motion in water (drifting).

3. Finally, there is motion of vessels and/or people aboard (sailing), which com-bines properties typical of both active and passive aquamotion and can there-fore be viewed as in-between the two. A sailing situation can be construed as motion of different kinds of Figure – the navigator, the vessel and passengers on board. Sailing involves, of course, agentivity of an animate (and prototypi-cally human) entity who navigates the vessel, but whose own motion is com-pletely dependent on the latter’s motion. The vessel itself, although moving, lacks agentivity and control over the motion, whereas “passengers” on board are typically passive, when it comes to their contribution to the motion.

Figure � bellow summarizes possible participation of different kinds of Figures in different kinds of aquamotion and the relative degree of activity of motion typical for each of the aquamotion subdomains.

As Maisak & Rakhilina (2007a) demonstrated, although languages on the whole tend to distinguish among these sub-domains by means of lexical opposi-tions and/or constraints on interpretation in the corresponding contexts, there is a great deal of cross-linguistic variation. One major difference is the sheer number

�. The four main English aquamotion verbs will be used as meta-linguistic labels for the aqua-motion sub-domains, even though the distinctions among the verbs and the sub-domains do not completely coincide.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3�3

of aquamotion verbs in a language, or the degree of elaboration within its lexical aquamotion system. There are, roughly speaking, three groups of systems.5

– Swedish, English and possibly Dutch belong to languages with rich aquamo-tion systems, i.e., they have more than three verbs. Indonesian, with its four-teen aquamotion verbs (Lander & Kramarova 2007), provides an example of a particularly richly elaborated aquamotion system.

– Languages with middle systems have two main aquamotion verbs normally distinguishing between active and passive motion, and use a general motion verb for motion on vessels. They are, on the whole, fairly infrequent – Tamil, Persian and Maninka are the only languages with middle aquamotion sys-tems par excellence among the 50 languages in the Aquamotion project.

– Finally, quite a few languages have poor systems that seem to neutralize the distinctions among the aquamotion sub-domains or at least make these dis-tinctions peripheral. Turkish, for instance, has only one aqua-motion verb, yüzmek, which alone covers swimming, sailing and drifting/floating. Poor lexical aquamotion systems are also found in the Slavic languages Rus-sian and Polish, to which we turn in the next section.

An interesting question is how the distinctions among the aquamotion subdo-mains relate to the components in a Motion event and to the different lexicaliza-tion patterns as suggested by Talmy (2000). Thus, for instance, in the face of such contrasts as Peter was swimming/floating/sailing, where one and the same human Figure is participating in different kinds of aquamotion, it might be reasonable

�. Note that even languages with comparable systems may show significant variation in de-tails, e.g., with respect to the precise ways in which the different aquamotion verbs divide the semantic space. For the sake of space we will not elaborate on this point in the paper.

Kind of Figure Animates Vessels Inanimates

primarily humans

Kind of aquamotion Active self-propelled aquamotion (swimming)

Motion of vessels and/or people aboard (sailing)

Passive motion (directed motion, drifting, and lo-cation in water, floating)

Activity of motion

more active more passive

Figure 1. The activity-of-motion scale underlying the distinctions among the main aquamotion sub-domains and possible kinds of Figure participating in each of them

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�� Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

to suggest that the three English verbs swim, float and sail have lexicalized the Manner of aquamotion. On the other hand, all the three conflate Manner with Ground; in addition, swim conflates both Ground and Manner with Figure (since only animate Figure can swim).

�. From typology back to Russian and Polish

For reference to aquamotion, both Russian and Polish have only two derivational-ly related imperfective verbs, plyt’/plavat’ in Russian and płynąć/pływać in Polish, where the basic distinction between the two verbs has to do with unidirectionality of motion. The first verb in each pair is unidirectional (7a), whereas the other one is non-(uni)directional (7b–c). (Although the presentation and examples in this section will be based on Russian, it applies also to Polish, mutatis mutandis.)

(7) Russian a. Petja ply-l k skal-am. Petja “plyt’”-pret.m.sg towards rock-dat.pl ‘Petja was swimming to the rocks.’ b. Petja plava-l k skal-am (každyj den’). Petja “plavat’”-pret.m.sg towards rock-dat.pl (every day) ‘Petja swam to the rocks (every day).’ c. Petja plava-l u skal. Petja “plavat’”-pret.m.sg by rock-gen.pl ‘Petja was swimming by the rocks.’

Aquamotion verbs in Russian and Polish, like motion verbs in general, have thus lexicalized a semantic distinction that appears to be absent from the correspond-ing lexical field in Swedish, English, Dutch, as well as in many other languages.

At the same time, the parameter of activity/passivity of motion, underlying the distinctions among the aquamotion sub-domains within the Swedish, English and Dutch systems and, further, in many other languages of the world, does not appear to be relevant for Russian and Polish. For instance, the Russian examples in (8)–(�0) show ambiguities that are absent from their Swedish, English and Dutch translations.

(8) Human Figure: Active self-propelled motion vs. motion aboard Ply-vi k skalam! “plyt’”-imp towards rock-dat.pl ‘Swim/sail/row/paddle to the rocks!’

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3��

(9) Vessel: Motion of vessels vs. passive motion Lodk-a plyl-a vniz po rek-e. boat-nom.sg “plyt’”-pret.f.sg down on river-dat.sg ‘The boat was sailing/drifting/floating down the river’ or ‘Someone was rowing the boat down the river.’

(�0) Animate non-human Figure: Active self-propelled motion vs. passive motion (i.e., location in/on the surface of water)

V bassejne plava-l krokodil. in swimming-pool-loc.sg “plavat’”-pret.m.sg crocodile.nom.sg ‘A crocodile was swimming in the swimming pool / There was a crocodile in

the swimming pool.’

Let us examine to what extent the Swedish-Dutch-English and the Russian-Pol-ish systems are different. As we hope to show, the systems in the two groups of languages are sensitive to similar cross-linguistically relevant semantic distinc-tions within the aquamotion domain, even though their manifestations can be striking.

First of all, the difference between plyt’ and plavat’ in reference to passive aquamotion easily translates into the distinction between drifting and float-ing. In the prototypical cases drifting constitutes a unidirectional motion that involves a current or a stream, flowing in a certain direction. Therefore it does not come as a surprise that drifting is expressed by the unidirectional verb plyt’, while the non-directional plavat’ refers to non-directional (or even static) float-ing. Thus, the normal interpretation of ex. (��a) requires that both the boat and the black logs are moving in a certain direction, while the black logs in ex. (��b) are either static or are moving chaotically, without any specific direction.

(��) a. Rjadom s nash-ej lodk-oj ply-l-i alongside with our-instr.f.sg boat-instr.sg “plyt’”-pret-pl čern-ye tjažel-ye brevn-a. black-nom.pl heavy-nom.pl log-nom.pl ‘Heavy black logs floated downstream alongside our boat.’ (= were moving) b. Rjadom s nash-ej lodk-oj plava-l-i alongside with our-instr.f.sg boat-instr.sg “plavat’”-pret-pl čern-ye tjažel-ye brevn-a. black-nom.pl heavy-nom.pl log-nom.pl ‘Heavy black logs floated around/here and there near our boat.’ (= were located)

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�6 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

Second, plyt’ and plavat’ show some asymmetries in their propensity to combine with expressions referring to the ground, source or goal of motion that can be related to the semantic distinctions among the three main aquamotional sub-domains.

To unveil these asymmetries we have compared the relative frequencies of plyt’ in the Russian National Corpus (RNC) used in different constructions and referring to different kinds of aquamotion. The relevant constructions (or con-texts) are as follows:

– plyt’ not followed by ground, source or goal– plyt’ followed by ground (po ‘on’ + NP in the Dative case)– plyt’ followed by goal (k ‘towards’ + NP in the Dative case, v ‘into’ + NP in

the Accusative case, or do ‘to’ + NP in the Genitive case)– plyt’ followed by source (iz ‘from’ + NP in the Genitive case, or ot ‘from’ + NP

in the Genitive case)– plyt’ followed by source and goal (iz ‘from’ + NP in the Genitive case and v

‘into’ + NP in the Accusative case, or ot ‘from’ + NP in the Genitive case and do ‘to’ + NP in the Genitive case)

The occurrences of plyt’ have been categorized into the by now familiar cate-gories:

– motion of vessels and people aboard, – aquamotion of other inanimate (including dead) Figures (i.e., passive mo-

tion), and – active aquamotion, where we have distinguished between two kinds of ani-

mate Figures – human Figures and animate non-human Figures (both not aboard), for reasons which will become evident below.

We have restricted ourselves to a one-sentence context, which has sometimes not been sufficient for understanding the details of the motion (i.e., who/what is moving and in what way). However, the absolute majority of the examples could be categorized unambiguously. Note also that we have excluded the numerous metaphorical and extended uses of plyt’. We have applied different strategies for selecting and counting the occurrences of plyt’ without reference to the ground, source and goal, on the one hand, and in all the other contexts. In the first case, we have randomly selected �000 examples among the 6362 occurrences of plyt’ in the Russian National Corpus (RNC). For the other contexts we have analyzed all the occurrences in the corpus. That means that the comparison of plyt’ across the different contexts should be based on relative rather than absolute frequencies.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3�7

Plyt’ not followed by ground, source or goal. As is clear from Table 3, in this context (which is the by far most frequent one) plyt’ tends to refer more or less equally often to motion of vessels/people aboard and to active aquamotion, while its passive-aquamotion readings have a significantly lower frequency.

However, explicit reference to the ground, source or goal of aquation may influence the distribution of the preferred vs. dispreferred readings of plyt’. These constructions are on the whole relatively infrequent, with ground being men-tioned much more often than both source and goal, and source being men-tioned in very few cases (for the figures cf. Table 4).

Plyt’ accompanied by ground (po ‘on’ + NP in the Dative case). In this con-text, motion of vessels and people aboard constitutes the most frequent reading of plyt’ (cf. ex. (�2a)). In addition, these contexts also favour passive-motion read-ings of plyt’, as in ex. (�2b), even though although reference to passive motion is in general fairly infrequent. On the other hand, this context seems to disfavour active self-propelled motion of human Figures, i.e., prototypical active aquamo-tion. In other words, when referring to active self-propelled motion of animate entities, plyt’ prefers not to combine with an overt indication of the ground (with (�2c) being a rare example of the opposite). Perhaps somewhat unexpectedly, ac-tive motion of other animates – fish, birds, snakes, dogs, horses, fairy figures etc. freely allows specification of the ground (ex. (�2d)).

(�2) Plyt’ accompanied by ground (po ‘on’ + NP in the Dative case). a. Po sin-emu morj-u ply-l znakom-yj on blue-n.dat.sg sea-dat.sg “plyt’”-pret.m.sg familiar-m.nom.sg belyj korabl’ s dvumja vysoki-mi white.m.nom.sg ship.nom.sg with two-instr tall-instr.pl naklonny-mi truba-mi. inclined-instr.pl funnel-instr.pl ‘A familiar white ship with two tall inclined funnels was sailing on the

blue sea.’ (G. Alekseev. Zelenye berega (�983–�984), quoted after RNC) b. Plyli po vod-e stog-i sen-a, “plyt’”-pret.pl on water-dat.sg stack-nom.pl hay-gen.sg brevn-a, plot-y, oblomki izb i, log-nom.pl raft-nom.pl fragment-nom.pl hut.gen.pl and (…) dostignuv plotiny, stalkivalis’ drug s drugom, nyrjali, opjat’ vyplyvali i

sbivalis’ v kuču v odnom meste. ‘Haystacks, logs, rafts and fragments of huts were floating/drifting

downstream on the water and, having reached the dam, dove into the water, emerged again and bunched together in one and the same place.’ (M. E. Saltykov-Ščedrin. Istorija odnogo goroda (�869–�870), quoted after RNC)

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�8 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina RakhilinaTa

ble

3. P

lyt’

refe

rrin

g to

the

diffe

rent

sub-

dom

ains

with

in a

quam

otio

n w

hen

not f

ollo

wed

by

expr

essio

ns o

f gro

und,

sour

ce o

r goa

l in

the

Russ

ian

Nat

iona

l Cor

pus (

amon

g th

e �0

00 ra

ndom

ly ch

osen

occ

urre

nces

of p

lyt’)

Act

ive

aqua

mot

ion

Mot

ion

of v

esse

ls a

nd

peop

le a

boar

dPa

ssiv

e aq

uam

otio

n of

an

inan

imat

e (in

cl. d

ead)

Fi

gure

(not

a v

esse

l)

Unc

lear

con

text

sTo

tal

Aqu

amot

ion

of a

hum

an

Figu

re

Aqu

amot

ion

of a

n an

i-m

ate

non-

hum

an F

igur

e

23.3

% (�

06)

�3.6

% (6

2)33

.4%

(�52

)6.

6% (3

0)23

.�%

(�05

)45

536

.9%

(�68

)

Tabl

e 4.

Ply

t’ re

ferr

ing

to th

e di

ffere

nt su

b-do

mai

ns w

ithin

aqu

amot

ion

whe

n fo

llow

ed b

y ex

pres

sions

of g

roun

d, so

urce

or g

oal

in th

e Ru

ssia

n N

atio

nal C

orpu

s

Act

ive

aqua

mot

ion

Mot

ion

of v

esse

ls

and

peop

le a

boar

dPa

ssiv

e aq

uam

o-tio

n of

an

inan

im./

dead

Fig

ure

(n

ot a

ves

sel)

Unc

lear

co

ntex

tTo

tal

Aqu

amot

ion

of a

hu

man

Fig

ure

Aqu

amot

ion

of a

n an

im. n

on-h

uman

Fi

gure

Expr

essio

ns o

f gr

ound

Plyt

’+ p

o DA

T�.

6% (6

)��

.2%

(4�)

47.5

% (�

74)

�3.7

% (5

0)26

% (9

5)36

6

Expr

essio

ns o

f go

alPl

yt’ +

k D

AT47

.5%

(47)

�6.2

% (�

6)25

.3%

(25)

7.�%

(7)

4% (4

)99

Plyt

’ + v

ACC

0%

0%94

.7%

(36)

5.2%

(2)*

0%

38Pl

yt’ +

do

GEN

00

92.3

% (�

2)

07.

7% (�

)�3

Expr

essio

ns o

f so

urce

Plyt

’ + iz

GEN

0%0%

87.5

% (�

4)�2

.5%

(2)

0%�6

Plyt

’ + o

t GEN

�2.5

% (�

)0

87.5

% (7

)0

08

Expr

essio

ns o

f so

urce

and

goa

lPl

yt’ +

iz G

EN v

ACC

0%

0%�0

0% (7

)0

0%7

Plyt

’ + o

t GEN

do

ACC

00

�00%

(2)

00

2

* Inc

ludi

ng o

ne o

ccur

renc

e in

the

cont

ext p

lyt’

+ po

DAT

+ v

ACC

(i.e.

, ply

t’ +

grou

nd +

goa

l)

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3�9

c. Mne snilos’, čto ja ply-vu I:dat dream:pret.n.sg that I:nom “plyt’”-pres.�sg po bezbrežn-omu i spokojn-omu morj-u, on boundless-n.dat.sg and calm-n.dat.sg sea-dat.sg (…) tolkaja vperedi sebja jarko-krasočnyj mjač. ‘I dreamed that I was swimming/floating on a boundless and calm sea,

pushing a bright-coloured ball in front of me.’ (R. Naxapetov. Vljublennyj (�998), quoted after RNC) d. Černyj zmej ply-l po vod-e, (…) black-n.nom.sg. serpent-nom.sg “plyt’”-pret.m.sg on water-dat.sg (…) a sled za soboj ostavljal krasnyj, počti krovavyj, i Miše stalo strašno. ‘The black serpent was swimming [forward] in the water, leaving a red,

almost bloody trace after itself, and Misha got scared.’ (A. Jašin. Sladkij ostrov (�960), quoted after RNC)

Plyt’ accompanied by goal (k ‘towards’ + NP in the Dative case, v ‘into’ + NP in the Accusative case, or do ‘to’ + NP in the Genitive case). Explicit reference to the goal of aquamotion is, on the whole, much less frequent than reference to the ground and can be expressed in different ways. In our corpus data, the most fre-quent goal expressions consist of “k ‘towards’ + NP in the Dative case”. These are compatible with the different readings of plyt’ (cf. (�3c) for passive motion), but are particularly frequent with reference to active self-propelled motion of animate entities (ex. (�3a–b). The other two goal expressions – “v ‘into’ + NP in the Accu-sative case”, or “do ‘to, as far as’ + NP in the Genitive case” – are almost exclusively restricted to motion of vessels and people aboard (ex. (�3. Interestingly, one of the two examples with plyt’ followed by a “v+ACC”-phrase and referring to passive motion (ex. (�3d)) contains also a reference to the ground (a “po+DAT”-phrase) which, as shown above, constitutes a favourable context for such readings.

(�3) Plyt’ accompanied by goal a. Marija ply-l-a k bereg-u, (…) M.:nom.sg “plyt’”-pret-f.sg to shore-dat.sg bojazlivo pogljadyvaja vverx. ‘Maria was swimming towards the shore, casting timid glances upwards.’

(A. Dmitriev. Doroga obratno//“Znamja”, 200�, quoted after RNC) b. S živ-oj pojmannoj ryb-oj with living-f.instr.sg captured-f.instr.sg fish-f.instr.sg vo rt-u vydra ply-l-a k gusto-mu in mouth-loc.sg otter.n.sg “plyt’”-pret-f.sg to thick-m.dat.sg

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

330 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

kustarnik-u, kotorym byl pokryt bereg. shrubs-dat.sg (…) ‘Holding a living captured fish in its mouth, the otter was swimming

towards the trick shrubs covering the shore.’ (I. Sokolov-Mikitin. Vydry (�923–�928), quoted after RNC) c. Želty-e list’ja, plyvu-šč-ie yellow-nom.pl leaf.nom.pl “plyt’”-part.pres.-nom.pl k vodostok-u to gutter-dat.sg ‘Yellow leaves floating downstream towards the gutter’. (E. Šklovskij. Sutra pjatogo patriarxa (�990–�996), quoted after RNC) d. Tak poka on tam naladčikom byl, ply-l-i brevn-a (…) “plyt’”-pret-f.sg log-nom.pl po Isterv-e v Vytekl-u, on Isterva-dat.sg into Vytekla-acc.sg (…) ni razu ni za čto sučkom ne zadevši. ‘So while he worked as an adjuster there, logs floated downstream the

Isterva into the Vytekla without any bough ever touching anything.’ (E. Lukin. Katali my vaše solnce (�997), quoted after RNC) e. My seli na kater, plyv-em do Novorossijsk-a (…) “plyt’”-pres.�pl as.far.as Novorossijsk-gen.sg (…) i popadaem v sil’nejšij štorm. ‘We took a launch, are sailing as far as Novorossijsk and get caught in a

very heavy storm.’ (È. Gerštejn. Perečen’ obid (�997), quoted after RNC)

Plyt’ accompanied by source (iz ‘from’ + NP in the Genitive case, or ot ‘from’ + NP in the Genitive case), and plyt’ accompanied by source and goal (iz ‘from’ + NP in the Genitive case and v ‘into’ + NP in the Accusative case, or ot ‘from’ + NP in the Genitive case and do ‘to’ + NP in the Genitive case). Although these con-texts are, on the whole, fairly infrequently, in our data they are almost exclusively restricted to motion of vessels and people aboard. This is particularly true when both source and goal are mentioned, as in ex. (�4a). Example (�4b) illustrates one of the rare cases in which plyt’ designates passive aquamotion when followed by an explicit reference to the source of motion.

(�4) Plyt’ accompanied by source or by source and goal a. Jaxta s kapitan-om Tolst-ym i šest’j-u yacht.nom.sg with captain-instr.sg T.-instr.sg and six-instr preobraženc-ami s potuše-nn-ymi preobrazhenec-instr.pl with put.out-pass.part-instr.pl ognj-ami ply-l-a iz Peterburg-a light-instr.pl “plyt’”-pret-f.sg from Petersburg-gen.sg

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 331

v Kronštadt. in Kronstadt.acc.sg ‘A sailing boat with captain Tolstoy and six soldiers of the Preobrazhenski

Regiment was sailing from Petersburg towards Kronstadt with extin- guished lights.’

(È. Radzinskij. Knjažna Tarakanova (�999), quoted after RNC) b. Reka sovsem očistilas’ oto l’da, i tol’ko izredka na nej pokazyvalis’ belymi

pjatnami zapozdavšie l’diny; verojatno, oni plyli (…), they.nom “plyt’”-pret-pl iz kakogo-nibud’ bojk-ogo gornogo from some-m.gen.sg vivid-m.gen.sg mountain.adj-m.gen.sg pritok-a. tributary-gen.sg ‘The river had become completely clear of ice, only some late ice-floes

appeared as rare white spots; they were probably floating downward from some vivid mountain tributary.’

(D. N. Mamin-Sibirjak. Na reke Čusovoj (�9�2), quoted after RNC)

The constructional differences related to the distinctions in the interpretation of plyt’ can also be found, mutatis mutandi, in the behaviour of plavat’, even though the situation there is slightly more complicated.

Summarizing this section, we can conclude that the cross-linguistically re-current distinctions among the three basic aquamotion sub-domains are in fact present in Russian (and Polish) as well, yet in other ways than in languages with lexically rich aquamotion systems in general and in Swedish, English or Dutch in particular. The three Germanic languages have several individual and deri-vationally unrelated lexemes, each specializing in a particular sub-domain. The main parameter underlying these distinctions is the degree of activity/passivity of motion, which, to a certain degree, can be interpreted as the differences in the manner of motion of a non-liquid Figure on a liquid ground (dependent, in turn, on the nature of the Figure). Russian and Polish, on the other hand, have two derivationally related lexemes that in the first place lexicalize unidirection-ality/non-directionality of motion of a non-liquid figure on a liquid ground, rather than the degree of its activity/passivity (or Manner). The Swed-ish-English-Dutch and the Russian-Polish systems are, however, not as distinct as they appear. First, directionality (i.e., presence/absence of aquamotion) is relevant also for Swedish and English, even though this distinction is found in only one sub-domain of aquamotion (drifting/floating within passive aquamotion) and is expressed by two unrelated lexemes. Conversely, the semantic distinctions among the main aquamotion sub-domains underlie some interesting syntactic asymme-tries in the behaviour of plyt’ and plavat’, i.e., their combinability with expressions

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

33� Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

referring to the ground, source or goal. In other words, different lexicaliza-tions in the Swedish-English-Dutch systems of aquamotion verbs are reflected in constructional differences in the Russian-Polish systems.

�. Genetic vs. typological similarities: Aquamotion of a non-liquid vs. liquid Figure

Since the main impetus for cross-linguistic and, in particular, typological research is finding out what aspects of a particular phenomenon are more specific or more universal, one of its tasks consists in distinguishing between cross-linguistic simi-larities that stem from genetic relatedness, on the one hand, and commonly attest-ed phenomena that do not stem from genetic relatedness. These commonalities, in turn, call for an explanation. The different ways of lexicalizing the degree of passivity/activity of aquamotion of a non-liquid Figure, demonstrated by Swed-ish, English, Dutch, as opposed to Russian and Polish, seem to have a strong ge-netic component; among other things, they involve cognates in Germanic and the inherited derivational opposition in Slavic.

We will now turn to a phenomenon where similarities among languages cross-cut these two families, namely, aquamotion verbs that can apply to both non-liq-uid and liquid Figures. These are found in two of the languages considered in the present paper – in Polish and Swedish; the description of Polish and the relevant examples are quoted from Prokof ’eva (2007).

In Polish, the verb płynąć (i.e., the unidirectional member of the pair płynąć/pływać considered in Section 3) can, in addition, describe the motion of a river, i.e., of a liquid Figure, as in ex. (�5)

(�5) O jest rzeką nizinną, płynie doliną o Ob is river lowland płynąć.pres.3sg valley.instr about szerokości do 60 km i uchodzi do Morza Karskiego. width.instr up.to 60 km and go:pres.3sg to Sea.gen Kare.gen ‘Ob is a (lowland) river that flows through a valley that is up to 60 km wide

and goes into the Kara Sea.’ (Zat. Obska, http://transsib.com.pl/rzeki.html, visited on November ��, 2008)

Płynąć competes with several other verbs for motion of liquids, each of which has its own specific meaning and range of uses: sąszyć się ‘ooze out, trickle’ (about a little stream of liquid), ciec ‘run, flow’ (about bodily liquids and thick liquids such as honey or lava that move on a surface without normally following a par-ticular path), and lać się ‘flow, pour, stream’ (about a considerable amount of

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 333

liquid moving in a stream without any contact with a surface, e.g., water pouring from a tap or from a ceiling).

The core uses of płynąć among these verbs have to do with motion of (wa-ter in) rivers. Crucially, rivers flow in riverbeds, their movement (or rather, the movement of the water in them) is, thus, severely restricted in its path and is ide-ally unidirectional, which agrees very well with the unidirectional meanings of płynąć when applied to motion of non-liquid Figures. On the contrary, the water in a sea or an ocean cannot be described by płynąć. The two other main contexts compatible with płynąć are liquids running in pipes (e.g., oil or water, ex. (�6a)) and flow in blood vessels (ex. (�6b)), which both show obvious similarities with water flowing in rivers e.g.,

(�6) a. Woda ogrzana w kotle płynie water.nom.sg heated.up.f.sg.nom in tank.loc “płynąć”.pres.3sg rurami wgórę do kaloryfer-ów, pipes.instr.pl up to radiator-gen.pl (...) a woda zimna innymi rurami spływa w dół do kotła. ‘Water heated up in the tank flows up trough pipes to the radiators, while

cold water flows down through other pipes to the tank.’ (“Horyzonty Techniki”. 20.�959, quoted after KJPWN) b. W rozgrzan-ych i elastyczn-ych żyłach in warmed.up-loc.pl and elastic-loc.pl veins.loc.pl płynie bez przeszkód krew. “płynąć”.pres.3sg without obstacle:gen.pl blood:nom.sg. ‘In warmed up and elastic veins blood can flow without obstacle.’ (J. Wojda. Za oceanu, quoted after KJPWN)

Swedish too uses one and the same verb, flyta, both with non-liquid and liquid Figures. As shown in Section �, for a non-liquid Figure flyta describes its passive aquamotion including location in/on the surface of water. In addition, it can apply to (the water in) rivers, as in ex. (�7):

(�7) Swedish a. Ungern-s huvudstad Budapest bruka-r kalla-s för Hungary-gen capital Budapest use-pres call.inf-refl for “Donau-s pärla” eftersom flod-en flyt-er mitt Donau-gen pearl because river-def.c.sg “flyta”-pres amidst genom stad-en och dela-r de-n i Buda och Pest. across city-def.c.sg and divide-pres it-c in Buda and Pest

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

33� Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

‘The capital of Hungary, Budapest, is often called “the pearl of the Donau” because the river runs across the middle of the city and divides it into Buda and Pest.’ (www.lfv.se/upload/Flygplatser/Umea/Program.doc, visited on November �� 2008)

b. I älv-en flöt ren-t vatten. in river-def.c.sg “flyta”.pret clear-n.sg water ‘Clear water was running in the river.’ (Parole)

Swedish has several verbs for motion of liquid, primarily water (ströma ‘to stream, to flow, to pour’, forsa ‘to rush, to come in torrents’, droppa ‘to drop’, etc.), but the main “rival” and quasi-synonym of flyta is rinna. Rinna normally denotes rela-tively rapid unidirectional motion (often downwards) of a relatively small amount of water that is conceived of as a relatively narrow stream and is therefore often applied to brooks, small currents, and the like. Flyta, on the contrary, presupposes a broad flow, a large amount of water flowing slowly, smoothly and is therefore often used for depiction of big and broad rivers. However, rinna is preferred in geographic contexts for generic descriptions of the location of rivers, of where they have their sources, where they flow out etc. This can be due to the differences in the perspectivization typical of rinna and flyta: on a map rivers look like thin lines running in one particular direction. In many contexts with water as Figure, flyta lacks (almost) any elements of motion, emphasizing the big quantity, abun-dance of water, often undesirable, as in (�8):

(�8) Det flyt-er omkring e-n massa vatten på köksgolv-et. subj “flyta”-pres around a-c lot water on kitchen.floor-def.n.sg ‘There is a lot of water (flowing) on the kitchen floor.’ (Parole)

Summing up, we see that both Polish and Swedish converge in having a verb that can describe aquamotion of both non-liquid and liquid Figures, with mo-tion of (water in) rivers constituting the core of the latter uses (flowing). It is also clear that the liquid-Figure uses share significant semantic similarities with the non-liquid Figure uses of the same verb. In Polish, the distinctive property of płynąć is the pronounced unidirectionality of the motion it refers to, which explains why rivers, pipes and blood vessels provide the best conditions for such motion. In Swedish, the distinctive property of flyta in non-liquid-Figure uses is the passive, slow character of motion or even the absence of any motion, all depending on the dynamics of the water involved (which, in turn, presupposes a considerable amount of water). This slowness or even absence of motion and the presence of a big quantity of water is also characteristic of flyta in its liquid-Figure uses. It would in fact be rather unexpected if liquid-Figure uses would be found in the swimming verb to the exclusion of the floating verb in a system

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 33�

distinguishing the two. An interesting contrast between Polish and Swedish is provided by descriptions of blood, sweat and tears. Since streams of blood, sweat and tears can evoke an image of a channel, they can occasionally be depicted as Figure to płynąć; however, the relative small quantities of liquid involved in such situations are not compatible with the basic requirements on the use of flyta.

Now, in contrast to Polish, Russian distinguishes between aquamotion verbs combining with liquid vs. non-liquid Figure: plyt’/plavat’ apply to non-liquid Fig-ures, while motion of liquid Figures is described by such verbs as lit’sja and teč.6 The same is true for such Slavic languages as Czech, Bulgarian and a few others. On the other hand, examples like (�9) attested in Old Russian (at least in the �6th–�8th centuries), Ukrainian, Belorussian and in some Russian dialects testify to the existence of a situation akin to the one found in Polish:

(�9) a. Old Russian (�688) polacъ by-lъ sozda-nъ rostrum.nom.sg be-pret.m.sg make-part.pret.pass.nom.m.sg nadъ mor-emъ, takъ čto vod-a morsk-aja over sea-instr.sg so that water-nom.sg sea.adj-nom.f.sg pod nego ply-l-a. under it.acc “plyt’”-pret-f.sg ‘A rostrum was built over the sea and the sea water was flowing under it.’

(Rimskie dejanija, �83) b. Ukrainian Plyv-l-a rik-a, zabut-a cily-m plysti-pret-f.sg river-nom.sg forgotten-f.sg.nom whole-instr.m.sg svit-om. world-instr.m.sg ‘The river was flowing, forgotten by the whole world.’ (L. Malkovič, “Tanec’ samotnosti” (2005). Literaturno-mistec’kij al’manax’,

2–3: 69, www.pu.if.ua/data/ukr/lib/e-book/alkos-2-3.pdf, last visited on November �� 2008)

Similarly, within Germanic, languages such as German, Dutch and English maintain the distinction between floating and flowing, in contrast to Swed-ish; Danish behaves like Swedish, while Norwegian and Icelandic have certain restricted contexts where the cognates of flyta can apply to liquid Figures. Some

6. There is a handful of examples in the Russian National Corpus where plyt’ describes the motion of rivers and of liquids, e.g., Kuda plyveš, Volga? ‘Where are you flowing, Volga?’ (Valentina Oseeva. Dinka (�959), quoted after RNC). We have no explanation for these ex-amples which strike us as very marked.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

336 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

other Indo-European languages outside of the Germanic and Slavic families show similar phenomena, e.g., fluitare in Latin (Gruntova 2007) and plaukti in Lithu-anian (Arkad’ev 2007) which both can describe aquamotion of both non-liquid and liquid Figures. We cannot judge to what degree these syncretic phenomena are inherited from their proto-languages or are results of later independent de-velopments in particular languages or in particular subgroups of languages. The typologically interesting fact is that there are abundant cross-linguistic parallels to the Swedish and Polish verbs conflating aquamotion of liquid and non-liquid Figures – e.g., in Persian (šenāvar budan), Hindi (bah-), Finnish (solua/soljua), Japanese (nagareru), Hakass (ağarğa), etc. (Maisak & Rakhilina 2007a: 5�–52). The direction of the semantic extension is not clear, i.e., whether it starts with the motion of liquid Figures and later spreads to non-liquid ones, or the other way round. Both scenarios involve simple metonymy.

Thus, the (non-)existence of an aquamotion verb that can apply to both liquid and non-liquid Figures is an additional parameter of cross-linguistic variation, where even closely related languages differ. The next section will continue on the issue of closely related languages in lexical typology.

6. Lexical typology, closely related languages and diachronic change

Closely related languages often offer instructive examples of significant typologi-cal differences that cast light on possible ways in which language systems arise and develop, e.g., expand or shrink. While this is an acknowledged and theoreti-cally significant fact for grammatical and phonetic typology, this is probably even more evident in lexical-typological comparison, since historical changes in the lexicon take on the whole less time than grammatical changes and sound changes. Here we will briefly mention a couple of relevant phenomena in the Slavic and Germanic aquamotion systems.

All the Slavic languages have poor aquamotion systems, but this poverty shows somewhat different degrees across Slavic. On the one hand, the Old Rus-sian system was richer than the systems of modern Russian and of most other Slavic languages in having a special verb bresti for active self-propelled motion of animate Figures (swimming) while limiting plouti/plavati to motion of ves-sels/people on vessels and to passive motion/location (Rakhilina 2007). Bulgar-ian, on the other hand, has an extremely poor aquamotion system, both within Slavic and within the whole sample of languages examined for the aquamotion project (Ganenkov 2007). More specifically, while some speakers do maintain the directionality-related distinction between the two verbs pluvam/plavam, others have generalized pluvam to (almost) all aquamotion. The only gaps in its uses are

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 337

passive motion of inanimate Figures, including location in or on the surface of water that are covered by nosja se.

The different Germanic languages, while operating with similar sets of cog-nates, distribute their uses in markedly different ways. Thus, the aquamotion sys-tem of Modern German has been subject to a historical change that has brought it closer to the poor Slavic systems (cf. Šemanaeva 2007). Although German has two special verbs for passive aquamotion, driften and treiben, these are used option-ally. They have to compete with the verb schwimmen that has been generalized to mean both active self-propelled motion (swimming) and passive motion (drift-ing/ floating). Its only “gap” is the motion of vessels and of people aboard, for which German uses general motion verbs as well as segeln ‘sail’, rudern ‘row’, etc.

Active motion of vessels and people aboard is, on the whole, an interesting domain within Germanic, where the languages differ quite a lot, as was made clear in Section � (cf. Tables � and 2). Swedish and German split this zone into four sub-zones in accordance with the kind of vessel involved: the motion of row-ing boats and canoes, and sailing boats. This splitting is easy to understand given that motion in each of these cases is accomplished in its own particular way and is driven by its own particular force. Motor-driven boats stand out here: since they are relatively late additions to the traditional vessels, the need to have a label for their motion arose quite recently. Now, although motor-driven boats move in water, they also share certain similarities to other motor-driven vehicles, such as trains, cars etc., which might be the reason for why German and Swedish use general motion verbs for this kind of aquamotion. English, on the other hand, has extended sail to these vessels, thereby reducing the number of sub-zones within motion of vessels and people aboard to three. This step is quite reasonable since motion in motor-driven boats has partly taken over the role of sailing as the fast-est and most efficient type of aquamotion.

Dutch has undertaken a more radical step in having one verb covering aqua-motion of all vessels and people aboard – varen, which however co-exists with optional specialized verbs for motion involving particular kinds of vessels. Inter-estingly, its German cognate, fahren, is the general verb for going/riding/travel-ling-not-on-foot. It seems that Dutch first used the general motion verb varen both for “not-self-propelled” aqua- and terra- motion, but later limited it to the former, while terra-motion is covered by rijden. The Dutch system finds there-fore certain parallels in the languages with middle aquamotion systems (e.g., Per-sian, Tamil, Maninka (Maisak & Rakhilina 2007a)), which have two designated aquamotion verbs – one for active self-propelled motion and one for passive mo-tion/location – and use general motion verbs for motion of vessels and of people aboard. The present Dutch system is, however, still richer than these systems in

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

338 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

restricting varen to aquamotion and in having optional specialized verbs for mo-tion involving particular kinds of vessels.

7. Conclusions and implications

The main points of the paper can be summarized as follows:

�. Languages vary considerably as to how they carve up, or categorize the aqua-motion domain by means of words and lexicalized expressions; they differ primarily with respect to the number of different aquamotion verbs they have, ranging from one in the “poor system” of Turkish to eight in the “rich system” of Indonesian, and secondarily with respect to the division of labour or the parameters underlying the choice among them. The main parameters underlying semantic categorization within the aquamotion domain across languages are the degree of activity/passivity of motion, which can be viewed as a scale with three main sub-domains – here labelled swimming, sailing, drifting/floating, – each centred around its own prototype and repre-senting a particular Manner of motion often limited to a particular kind of Figure), and, the presence/absence of direction, for distinguishing between drifting and floating. An additional typological parameter is the existence of a verb that can apply to aquamotion of both liquid and non-liquid Figure. The cross-linguistic variation according to these two parameters seems to be independent of each other.

2. The aquamotion systems in Slavic are poor, typically limited to two deriva-tionally related verbs like plyt’/plavat’ in Russian that cover the whole do-main. Although the main distinction between the two verbs is apparently of a completely different nature than the distinctions among the aquamotion sub-domains, the latter are still lexicalized as the different senses within the two polysemous lexemes. This is manifested in the asymmetries in the uses of plavat’ and plyt’ and in the different constructions (in)compatible with each of the senses. Thus, the Slavic opposition, in all its uniqueness, is an interest-ing variation on the generally attested cross-linguistic theme.

3. Closely related languages show significant typological differences in their aquamotion systems and provide invaluable insights for fine-grained se-mantic analysis and for our understanding of historical processes that create, modify and obliterate such systems. In general, data from closely related lan-guages often deserve more attention in typological studies than they normally receive, and this is particularly true for lexical typology.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 339

A large portion of studies on motion verbs carried out within the latest two de-cades starts from the binary distinction between verb-framed and satellite-framed languages, building on Talmy’s (2000) three-fold classification of languages with respect to the encoding of spatial events. This is also true for a number of articles in the present volume. Although we have no doubts about the importance of this work in the domain of motion, we would like to emphasize that there are many more aspects to motion that deserve to be studied and many more parameters to motion verbs along which languages demonstrate systematic cross-linguistic variation.

As we hope to have shown in the present paper, aquamotion is an excellent example of a domain where languages show systematic cross-linguistic variation that has previously remained unnoticed. It is worth mentioning that Talmy’s types do not necessarily predict other important aspects of cross-linguistic variation within motion verbs. Thus, the division of the European languages according to the presence/absence of deictic verbs, studied by Ricca (�993), crosscuts Talmy’s types. The same goes for aquamotion, where the two groups of satellite-framed languages, Germanic and Slavic, turn up to be very different in their lexicalization patterns. Cf. the following quote from a paper by Slobin (2003):

S[atellite-framed]-languages allow for an economical expression of manner of motion in the main verb of a clause. Apparently as a consequence, these languag-es make habitual use of manner verbs when encoding motion events, and have developed large lexicons with many fine-grained distinctions of manner, in com-parison with smaller and less differentiated manner lexicons in V[erb-framed]-languages. One can say that the semantic space of manner of motion is “highly saturated” in S-languages, in comparison with V-languages. For example, French bondir doesn’t distinguish between the manners of motion encoded in English by jump, leap, bound, spring, skip, gambol; Spanish escabullirse can be translated as creep, glide, slide, slip, slither. A detailed study of ��5 English manner-of-mo-tion verbs found only 79 French counterparts, many of them of low frequency in comparison with English manner verbs (Jovanović & Kentfield �998). By con-trast, a similar study of Russian and English showed these two S-languages to be comparably saturated on this dimension (Dukhovny & Kaushanskaya �998). (Slobin 2003: �6�)

Slobin uses these cross-linguistic differences in the number of manner-of-motion verbs as evidence for a straightforward connection between a language type ac-cording to Talmy and the elaboration of its verbal lexicon with respect to manner of motion; this connection provides, in turn, support for Slobin’s “thinking-for-speaking” research. Although the connection pointed out in Slobin’s quote may very well hold for the lexicon of motion verbs on the whole, its different subparts can obviously show their own peculiarities. Thus, counter to the expectations, the

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

3�0 Maria Koptjevskaja-Tamm, Dagmar Divjak and Ekaterina Rakhilina

satellite-framed Slavic languages turn up to have significantly fewer manner-of-motion verbs than the verb-framed Romance languages Portuguese, Spanish and French. The latter, on the other hand, side with most of the Germanic languages, which are satellite-framed, but show a much higher degree of elaboration within their aquamotion verbal systems than the Slavic ones. Facts like these question the validity of Slobin’s straightforward connection or at least call for its modifications.

We hope that the future will see more detailed lexical-typological studies of particular domains, or of particular lexical groups within verbs of motion as a complement to the more general and often imprecise holistic classifications of languages on the basis of their overall motion verb systems.

References

Arkad’ev, P. 2007. Glagoly peremeščenija v vode v litovskom jazyke. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 3�5–333.

Divjak, D. 2006. Bewegen in water. Een zaak van lexicon of grammatica? In De taal van Pe-ter. Russisch-Nederlandse contacten en contrasten, E. Waegemans (ed.), 55–67. Leuven- Voorburg: Acco.

Divjak, D. & Lemmens, M. 2007. Lexical conflation patterns in Dutch aquamotion verbs. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), �52–�74.

Dukhovny, E. & Kaushanskaya, M. �998. Russian Verbs of Motion. Ms, Department of Psychol-ogy, University of California-Berkeley.

Ganenkov, D. 2007. Glagoly peremeščenija v vode: južnoslavjanskie jazyki. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 286–305.

Golubkova, E. & Rakhilina, E. 2007. Glagoly plavanija v sovremennom anglijskom jazyke. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), �06–�27.

Gruntova, E. 2007. Latinskaja sistema glagolov plavanija i ee razvitie v romanskix jazykax (fran-cuzskom, ital’janskom, ispanskom). In Glagoly dviženija v vode: Leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 23�–266.

Jovanović, J. & Kentfield, M. �998. Manifold manner: An exploratory analysis of French and English verbs of motion. Ms, Department of Psychology, University of California- Berkeley.

Koptjevskaja-Tamm, M. 2007. Švedskie glagoly dviženija v vode. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), �28–�5�.

Lander, Yu. & Kramarova, S. 2007. Indonezijskie glagoly plavanija i ix sistema. In Glagoly dviženija v vode: Leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 664–693.

Maisak, T. & Rakhilina, E. 2007a. Glagoly dviženija i naxoždenija v vode: leksičeskie sistemy i semantičeskie parametry. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 27–75.

Maisak, T. & Rakhilina, E. (eds). 2007b. Glagoly dviženija v vode: Leksičeskaja tipologija. Mosk-va: Indrik.

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny

Aquamotion verbs in Slavic and Germanic 3�1

Prokof ’eva, I. 2007. Peremeščenie v vode i peremeščenie vody: Glagoly płynąć/pływać ‘plyt’/pla-vat’’ v pol’skom jazyke. In Glagoly dviženija v vode: Leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 305–3�4.

Rakhilina, E. 2007. Glagoly plavanija v russkom jazyke. In Glagoly dviženija v vode: Leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), 267–285.

Ricca, D. �993. I verbi deittici di movimento in Europa: Una ricerca interlinguistica. Firenze: La Nuova Italia Editrice.

Slobin, D. I. 2003. Language and thought online: Cognitive consequences of linguistic relativ-ity. In Language in Mind: Advances in the Study of Language and Thought, D. Gentner & S. Goldin-Meadow, S. (eds.), �57–�92. Cambridge MA: The MIT Press.

Šemanaeva, O. 2007. Vyraženie peremeščenija v vode v nemeckom jazyke. In Glagoly dviženija v vode: leksičeskaja tipologija, T. Maisak & E. Rakhilina (eds.), �75–�97.

Talmy, L. 2000. Toward a Cognitive Semantics, Vol. I & II. Cambridge MA: The MIT Press.

Data sources for the examples quoted in the paper

KJPWN = Korpus Języka Polskiego Wydawnictwa Naukowego PWN, http://korpus.pwn.pl/Parole = the Parole corpus at the Swedish Language Bank, http://spraakbanken.gu.seRimskie dejanija (Gesta Romanorum). �878. Sankt Petersburg, v. �–2.RNC = Russian National Corpus, http://www.ruscorpora.ru/

Unc

orre

cted

pro

ofs

- J

ohn

Ben

jam

ins

Publ

ishi

ng C

ompa

ny