Navigation – Plan du site

Basic Verbs of Possession

A contrastive and typological study
Åke Viberg

Résumés

Verbs of possession such as HAVE and GIVE have been extensively studied both typologically and from a cognitive linguistic perspective. The present study presents an analysis of possession verbs as a semantic field with a focus on the most basic verbs. It combines a corpus-based contrastive analysis with a sketch of a general lexical typology of possession verbs. The contrastive part consists of an analysis primarily of the Swedish verbs ge ‘give’, ‘get’ and ta ‘take’ and their correspondents in some genetically and/or areally related languages. Data are taken from two translation corpora, the large English Swedish Parallel Corpus (ESPC) and the Multilingual Pilot Corpus (MPC) consisting of extracts from Swedish novels and their published translations into English, German, French and Finnish. The study of ta is concerned in particular with the relation between the many concrete uses of the verb, which are based on the interpretation of taking as a goal-directed action sequence. The account of Swedish ge ‘give’ and ‘get’ are brief summaries of earlier studies concerned with patterns of polysemy and grammaticalization. In particular the verb ‘get’ has a complex and relatively language-specific such pattern including modal, aspectual and causative grammatical meanings. The meanings GIVE, TAKE, GET and HAVE are all realized as verbs with very high frequency in the Germanic languages. This appears to be a rather language-specific characteristic. The typological part presents a tentative typology and gives a brief overview of some of the ways in which the corresponding meanings are realized in languages that are not included in the corpus.

Les verbes de possession comme AVOIR et DONNER ont été intensivement étudiés à la fois sous un angle typologique et du point de vue de la linguistique cognitive. Cette étude présente une analyse des verbes de possession en tant que champ sémantique en mettant l’accent sur les verbes les plus basiques. Elle combine une analyse contrastive basée sur corpus avec l’ébauche d’une typologie lexicale générale des verbes de possession. La partie contrastive consiste en une analyse des verbes suédois ge ‘donner’, ‘recevoir’ et ta ‘prendre’ et leurs équivalents dans plusieurs langues apparentées génétiquement et/ou géographiquement. Les données sont tirées de deux corpus de traductions, le English Swedish Parallel Corpus (ESPC) et le Multilingual Pilot Corpus (MPC) composé d’extraits de romans suédois et de leurs traductions publiées en anglais, allemand, français et finnois. L’étude de ta concerne en particulier la relation entre les nombreuses utilisations concrètes du verbe, qui sont basées sur l’interprétation de ‘prendre’ comme une séquence d’actions orientée vers un objectif. Les analyses des verbes suédois ge ‘donner’ et ‘recevoir’ sont de brefs résumés d’études précédentes concernant les schémas de polysémie et la grammaticalisation. En particulier, le verbe ‘recevoir’ présente un schéma polysémique complexe et relativement spécifique à la langue comprenant les sens grammaticaux de modalité, d’aspect et de causation. Les sens DONNER, PRENDRE, RECEVOIR et AVOIR sont tous réalisés dans les langues germaniques par des verbes très fréquents, ce qui paraît être une caractéristique plutôt spécifique à la langue. La partie typologique présente une tentative de typologie et fournit un bref aperçu de quelques moyens par lesquels les sens correspondants sont réalisés dans des langues qui ne sont pas représentées dans le corpus.

Haut de page

Texte intégral

1 Introduction

1.1 Combining lexical typology and contrastive studies

  • 1  One of the anonymous reviewers of this article commented “The semantic analysis is corpus-illustra (...)

1The study of lexical semantics from various crosslinguistic perspectives is an area that is attracting more and more attention.1 Within general typology, the lexicalization of several semantic fields has been studied such as color word universals (Berlin & Kay 1969, Hardin & Maffi 1997), the lexicalization of the motion situation (Talmy 1985, Slobin 2004), postural verbs (Newman 2002) and the field of perception verbs (Viberg 1984, Evans & Wilkins 2000). Goddard (2001) and Koptjevskaja Tamm (2008) present reviews of lexical typology. There are also several related approaches such as the Natural Semantic Metalanguage (NSM, Goddard & Wierzbicka 2002), semantic typology (Levinson & Wilkins 2006) and the typology of semantic associations (Vanhove 2008).

2Within contrastive studies, corpus-based analysis has led to the revitalization of a field that was dormant for many years. The use of translation corpora and comparable corpora allows fine-grained semantic comparisons within corpus-based contrastive lexicology (Altenberg & Granger 2002, Johansson 2007, Viberg 2005, Gómez González et al 2008). Parallel corpora are also beginning to be used in general typology (Cysouw & Wälchli 2007). There are also extensive multilingual computer-based lexical semantic databases such as Global WordNet, Simple and Global FrameNet (see web addresses in the reference section).

3Basic verbs of possession were among the first to be studied semantically from a crosslinguistic perspective (Bendix 1966). From a typological perspective, the verb GIVE is studied by Newman (1996, 1997) and in the broad studies of HAVE and other types of predicative possession by Heine (1997) and Stassen (2009). Verbs meaning ‘get’ (or ‘acquire’) have been studied as an areal phenomenon in South East Asia by Enfield (2003) and in North European languages by Viberg (2002a, 2006a).

4The approach to crosslinguistic lexicology presented in this paper is a combination of lexical typology and corpus-based contrastive analysis. This combination of perspectives has been most developed in earlier studies of perception verbs (see Viberg 1983, 2001 for the typological approach and Viberg 2008 for a corpus-based contrastive study.) The present study represents a further development of studies carried out within the now completed project Crosslinguistic lexicology presented in Viberg (1996). This project included in particular a series of studies concerned with possession verbs by Burenhult (ms) on the verb ta ‘take’ in Swedish, a diachronic study of get by Gronemeyer (1999), a study of combining the meanings ‘give’ and ‘take’ in Sango by Thornell (2002) and a corpus-based contrastive study of Swedish Viberg (2002a), ge ‘give’ (Viberg 2002b) and ta (Viberg forthc.). Many semantically oriented studies of verbs have been concerned with individual verbs. The present paper will present a summary of earlier and ongoing studies of the most basic verbs of possession based on a corpus-based contrastive analysis of  the Swedish verbs ge ‘give’, ‘get’ and ta ‘take’ and their correspondents in English and a number of other European languages. Special attention is paid to patterns of differentiation within the field of possession verbs and to patterns of polysemy and grammaticalization. At the end of the paper, the result of this analysis is taken as a point of departure for a sketch of a typological approach towards the study of the verbs of possession.

1.2 The concept of possession

5The concept of Possession is complex and can only be discussed briefly in this paper. Even in the most straightforward case where the object is concrete, possession can be construed in various ways. Miller & Johnson-Laird (1976: 565) use the following example to illustrate this: He owns an umbrella but she's borrowed it, though she doesn't have it with her. (See also Heine, 1997, on possession). Using partly different terms than Miller & Johnson-Laird, we can say that He owns an umbrella refers to Ownership, whereas she borrowed it refers to Temporary possession. Ownership presupposes certain socially regulated rights to use an object which is regarded as the property of a certain individual. These rights can be transferred permanently (e.g. as a gift) or temporarily (e.g. as a loan). The exact conventions are complex and vary a great deal among different cultures. The last part of the example, she doesn't have it with her, refers to Physical possession. Availability for immediate use or, in more general terms, control seems to be the crucial notion behind this meaning. In the prototypical case, Possession involves both Ownership and Physical possession, which can be combined as in the traditional text-book example: Peter gave Mary an apple (in her hand and she could keep it). Temporary possession is a possible but marked interpretation with a verb such as give (Peter gave Mary a book as a loan).

6Physical possession should be regarded as different from uses where mere Location is involved as in Charles has a spider on his ear. In this case, even inanimate possessors are allowed: The table has a vase on it. Location on the other hand should be contrasted with the Part/whole relation which is constitutive and tends to be permanent whereas Location usually is temporary: Charles has pointed ears, The table has three legs.

7More loosely, I will refer to all of the aspects of possession mentioned so far as concrete possession provided that the possessor and possessed entity are concrete. When a verb of possession has an abstract subject or object, this will be referred to as abstract possession which is a cover term for a number of different uses.

1.3 From physical action to transfer of possession

8The most basic verbs of possession in Swedish ha ‘have’, ‘get’, ta ‘take’ and ge ‘give’ have all been derived from physical action verbs related to movements and manipulations with the hand. All except have the same historical origin as the corresponding English verbs and can be discussed together with them. OED says about take: “The earliest known use of this verb in the Germanic languages was app. to express the physical action ‘to put the hand on’, ‘to touch’ – the only known sense of Gothic têkan.” Similarly about give: “The verb seems, from the evidence of Goth., OHG., and OS., to have primarily denoted the placing of a material object in the hands of another person.” The verb have and several of its counterparts in other European languages have developed from verbs originally meaning ‘grasp’, ‘seize’, ‘hold’ (Buck-Darling 1949) and verbs with these and a number of related meanings serve as the source of ‘have’ also in many non-European languages (see Heine 1997, 47-50, on the Action Schema). The Swedish verb is etymologically related to present-day Swedish fånga ‘catch’.

9There are languages which still lack specialized verbs of possession and where the closest equivalents of the verbs of possession are physical action verbs describing handling and manipulation. A very illuminating description of such a language is found in Rice (1997) who deals with Chipewyan, which perhaps rather should be referred to with its indigenous ethnonym Dene Sųłiné (Rice 2009). A characteristic feature of Chipewyan and of Athapaskan languages in general is a classificatory verb stem system which categorizes objects participating in the event according to characteristics such as animacy, number, shape/consistency (e.g. round/compact or flat/flexible) or constituency (e.g. liquid or mushy matter) which is related to the type of container that must be used, such as closed container or open container. In Chipewyan, up to ten such categories are distinguished. GIVE and TAKE do not have any single equivalents in Chipewyan but a shared set of classificatory verb stems referring to controlled handling and manipulation are used to express these meanings together with other aspects of actions involving continuing manual contact with the object. A typical example is (1).

(1)

Keni

eritł’is-tili

Aniyes-tł’aghe-ye-į-ł-tą´

(Rice 1997:100)

Kenny

paper-pail

Agnes-palm-in-3SG:s-class-perf.handle a closed container

Kenny gave Agnes the box.

10The verb stem tą´ means ‘to handle a closed container’. The distinction between GIVE and TAKE is signaled by a postpositional phrase which is usually incorporated initially into the verb, which has a rather great number of slots in front of the stem. TAKE can be expressed by combining the same stem with a postpositional phrase meaning ‘out of the palm’. The type of possession is also signaled in the postpositional phrase; ‘in his/her palm’ refers to temporary possession whereas ‘to him/her’ refers to permanent possession. The set of stems referring to controlled handling and manipulation contrast with another set of classificatory verb stems which refer to forceful/uncontrolled action and are used to express meaning such as toss or throw things to someone or snatch or steal things from someone. The two sets of stems together with the rich possibilities afforded by elements of various types that can appear in the slots in front of the verb stem makes it possible to express a wide variety of meanings, but there are no basic verbs of possession, and with respect to the verb stems there is no clear distinction between various types of physical handling and possession. Rice (1997) sums up the situation in the following way: “The absence of either a schematic or an all-encompassing basic-level verb of giving or taking means that the coarsely-construed lexico-semantic category of transfer predications is populated exclusively by verbal hyponyms, or what George Miller has termed troponyms.” (op. cit. p. 98.) Another important characteristic is the restriction to literal use of the relevant verbs: “All things being equal, the potential for metaphorical, metonymic, or otherwise non-literal use of the classificatory verb system appears to be fairly limited in Chipewyan.”(op. cit. p. 129) This means that there is no equivalent to the use of the verbs of possession to express the various types of what was referred to above as abstract possession, which is very prominent in Swedish and similar European languages. Since the basic verbs of possession in many cases have developed out of physical action verbs describing actions with the hand and since this is a gradual process, particular attention will be devoted to traces of such meanings in the following description.

2 A corpus-based contrastive study of the Swedish verbs of possession

2.1 Introducing parallel corpora into cognitive linguistics

11Recently, there has been a strong (and healthy) concern for methodology within cognitive linguistics (see Gonzalez-Marques et al 2007 for an overview). The present study is primarily related to the interest in corpus-based approaches (Gries & Stefanovitch 2006, Stefanowitch & Gries 2006). One approach that is not (or only sparsely) represented in the cognitive linguistic tradition is corpus-based contrastive analysis based on multilingual corpora (see Johansson 2007 for an overview). The present paper is based primarily on data from a translation corpus that is being developed by the author: The Multilingual Pilot Corpus (MPC). The present study is based on extracts from 10 novels in Swedish and their published translations into English, German, French and Finnish (totally around 250,000 words in the Swedish originals). The limited size of this and several other parallel corpora is a temporary problem. At the time of writing, the MPC corpus is approximately twice as big and in the future much larger parallel corpora will probably be generally available. The analysis in my earlier studies referred to in this article to a great extent were based on the English Swedish Parallel Corpus (ESPC) compiled by Altenberg & Aijmer (2000), which contains original texts in English and Swedish together with their translations. The texts are divided into two broad genres: Fiction and Non-fiction with several subcategories. The original texts in each language contain around 500,000 words.

12Theoretically, the approach to crosslinguistic lexicology (Viberg 1996, 2006b) presented in this study1) is most closely related to (but not directly based on) Talmy (2000) among cognitive linguists. Corpus-based contrastive analysis has fairly direct applications in translation studies and studies of second language acquisition (SLA).

2.2 Basic verbs of possession

13The most basic semantic fields of verbs typically have several hundred members in languages with an open system of verbs such as English and Swedish. For that reason, it is meaningful to start a contrastive comparison by identifying a set of basic verbs. Frequency can be used as a simple criterion. The most frequent verbs within a semantic field will be regarded as the most basic ones. In European languages, which have an open class of verbs containing something in the range of 10 000 well-established verbs, the twenty most frequent verbs tend to cover close to 50% of all the occurrences of the verbs in running text. There is a striking similarity across languages with respect to the meaning of many of these verbs In particular, there are nine lexical meanings that tend to be realized as one of the twenty most frequent verbs in all European languages, which are referred to as nuclear verbs (Viberg 1993, 2006a). These meanings are distributed between the most basic verbal semantic fields (indicated within parentheses): ‘go’, ‘come’ (Motion); ‘give’, ‘take’ (Possession), ‘make’ (Production), ‘say’ (Verbal communication), ‘see’ (Perception), ‘know’ (Cognition), ‘want’ (Desire). The set of nuclear verbs have a universal tendency to belong to the basic verbs within their respective fields. In addition, there are language-specific basic verbs, which in individual languages may be very frequent and in some cases even more frequent than the nuclear verbs in the same language.

14In the Swedish SUC-corpus based on around 1 million written words, there are 188 verbs of possession according to a classification (done by me) of all the close to 4000 verb types in the corpus. This is not a complete list, since the largest Swedish dictionaries contain around 10 000 verb types. A large proportion of the verbs occur only once or a few times. The most frequent verbs within the field account for the majority of the tokens as is evident already in the list of the ten most frequent verbs of possession in Swedish and English shown in Table 1 (based on Kilgariff 1998 and SUC 1997). The intention is only to give a preliminary overview to inform the selection of verbs to include for further analysis, which will indicate more clearly to what extent the verb is to be regarded as a verb of possession. All uses of a verb are included in the frequency count and the classification is based on what is considered as the basic meaning. The aim of the corpus-based analysis presented below is to account for the different uses and their meanings.

Table 1. The the ten most frequent  verbs of possession in English and Swedish

English (based on BNC)

Swedish (based on SUC)

Rank

Verb

Frequency per

100 000 words

Verb

Gloss

Frequency per

100 000 words

1

have

1376

ha

'have'

1383

2

get

221

'get'

459

3

take

179

ta

'take'

229

4

give

131

ge

'give'

163

5

need

62

betala

'pay'

29

6

keep

50

köpa

'buy'

25

7

provide

47

sälja

'sell'

24

8

pay

36

sakna

'lack'

20

9

buy

25

äga

'own'

17

10

send

24

räcka

’reach’, 'hand'

16

Total number of

words in corpus

100 million

1 million

15We can observe that the four verbs meaning ‘have’, ‘get’, ‘take’ and ‘give’, which (for several reasons) will be regarded as the most basic verbs of possession in this paper also are the most frequent ones. There is a clear drop in frequency from ‘give’ with the rank of 4 to the next verb in both languages. (If less clear cases such as English need, keep and provide are not regarded as verbs of possession, the drop is more pronounced.) Clearly, the cut-off point is arbitrary in the sense that there is a continuum and that there is a great difference in frequency among the four basic verbs, in particular between ‘have’ and the following verb. We can observe that get and its closest correspondent are more frequent than the nuclear verbs ge/give and ta/take, which means that they represent important language-specific characteristics. We can also see that Swedish is approximately twice as frequent as get, which means that a comparison of these two verbs is particularly interesting in a contrastive study of English and Swedish (see Viberg 2002a).

16The three verbs ge, ta and together with ha stand out as more frequent and basic and have more complex patterns of polysemy than other verbs of possession in Swedish. That they are short and have irregular inflection is another feature that tends to characterize basic verbs. As has already been remarked, ge and ta belong to the set of nuclear verbs that have correspondents in many languages, whereas ha and are relatively language-specific. The analysis of ha and its various uses in the MPC has not been completed yet. Table 2 displays the basic verbs of possession in Swedish in a simple grid.

Table 2. The basic verbs of possession in Swedish in a simple semantic grid.

Swedish

Dynamic meaning:

Source-based

Goal-based

Causative

ge ‘give’

ta ‘take’

Inchoative

‘get’

State

ha ‘have’

17The distinction between giving and taking is described as Source-based and Goal-based with reference to the role of the subject and the direction of the transfer. The dynamic system containing the basic contrasts Causative, Inchoative (pure change) and State cut across all semantic fields (Viberg 1981). The empty space in the grid (opposite ) is not really empty but is occupied by förlora ‘lose’, ‘come not to possess (something valuable)’, tappa ‘drop, lose’ and bli av med ‘come not to possess (usually something you don’t want)’. The verb sakna ‘lack’ with rank 8 in Table 1 refers to the corresponding state. The space is left empty because those verbs do not reach a very high frequency and do not have characteristics which justify calling them basic to the same degree as the four verbs included in the table. Each of the basic verbs serve as the superordinates of a number of hyponyms, whereas a superordinate term is lacking for the empty space in the grid.

3 Swedish ta ‘take’ as a physical action verb

18English take is analyzed from a cognitive linguistic perspective in an early paper by Norvig & Lakoff (1987). The Swedish verb ta, which will be studied in this paper, has already been analyzed from a cognitive linguistic perspective by Ekberg, who focuses on metaphorical and grammaticalized uses (Ekberg 1993a, b). There is also a corpus-based unpublished study of Swedish ta by Burenhult (ms.). The following account is based on Viberg (forthc.) The analysis will focus on the more concrete uses which are more or less directly related to ta as a physical action verb leaving most of the abstract and metaphorical meanings out of consideration for the time being. The meaning of Swedish ta and its correspondents in other languages is hard to pin down even from this restricted perspective. As a point of departure we can use example (2), where the closest equivalents of ta in the languages included in the MPC are used as translations (Eng. take, Ger. nehmen, Fr. prendre and Fin. ottaa). In order to contextualize the example, the English translation is given first including the sentence preceeding the one to be analyzed.

(2)

[I don't think Aunt Ester and Aunt Beda are at home. Go into the dining room, open the little drawer on the left of the buffet, and inside there's a bag of chocolate creams they've hidden because Ma says I mustn't eat sweet things.]

Take four chocolate creams, but mind they don't catch you'.

Ta fyra praliner, men se dig för, så att du inte blir ertappad. IB

Nimm vier Pralinen, aber sieh dich vor, daß man dich nicht erwischt."

Prends quatre pralines, mais fais bien attention qu'on ne te surprenne pas.

Ota neljä karamellia, mutta pidä varasi, ettet jää kiinni.

19In this example, ta describes the transfer of possession of a concrete object to the agent who carries out the act of taking. In this particular example, the taking also involves physical action, i. e. actions carried out by moving various body parts. It will be argued that this should be regarded as a part of the meaning of ta and not exclusively as a pragmatically motivated interpretation filled in from the context. I will return to that later. The physical action consists of a sequence of bodily activities carried out with the arm and hand. The arm is extended (extension) until it touches the object (physical contact) and then the object is grasped with the hand (grasping). After that, the arm is moved while the object is still held in the hand (handling). The goal of the action is to control and have access to the object in some sense, i.e. to possess it.

Table 3. Taking as a goal-directed action sequence

Sequence of bodily actions

Goal/Function

Extend arm towards object

Touch object with hand

Grasp object

Move arm while holding object

a) Acquire object as a possession

b) Use object as an instrument

20Often the act of taking hold of an object is the first step in a sequence of actions where the object is used as an instrument. A typical example is shown in (3). ‘Take’ is used as a translation in all languages except French where a directional verb is used (sortir ‘move out of’).

(3)

Reine tog upp pennan ur väskan och började klottra på den påse han fått sin mat i. PCJ

Reine took out his pen and started jotting on the paper sack his food had come in.

Reine nahm den Stift aus der Tasche und begann auf die Tüte zu kritzeln, in der er sein Essen bekommen hatte

Gunnar sortit son stylo de son sac et commença à gribouiller sur la serviette en papier du Carrols.

Reine otti laukusta kynän ja alkoi raapustaa paperipussiin, jossa oli ostanut ruokansa.

21In (3), ta is constructed as an induced motion verb (Literally: Reine took up the pen out of (ur) the bag: V + Directional particle + Object + PPsource). This is the reason why French uses a directional verb in accordance with its being a satellite-framed language (Talmy 2000). Examples of this type are worth special attention, since ta is not an ordinary motion verb but the ultimate goal is to use the object as an instrument. This is even clearer in an example like He took the pen and started writing. The step from such uses and the development of ‘take’ into an instrumental marker is not far. Such developments represent a common type of grammaticalization in West African and Southeast Asian languages (Lord 1993).

22In Swedish, it is possible to find examples such as (4) and (5) where ta refers to the earlier steps in the action sequence, touching and grasping an object. Typically, ta in such uses is not combined with a direct object but with a preposition like ‘on’ as in (4), where ta refers to the examination of an object by touch. Verbs meaning ‘touch’ and ‘feel’ are used as translations in all the other languages.

(4)

De var lite klibbiga att ta på. POE

They [some young birds] were a little sticky to touch.

Sie fühlten sich ein bißchen klebrig an.

Ils étaient un peu poisseux à toucher.

Kun niihin koski ne tuntuivat hiukan tahmeilta.

23When touching by hand is profiled in the action sequence, ta can also appear in a construction that is characteristic of physical contact verbs such as slå ‘hit’, klappa ‘pat’ and stryka ‘stroke’, where the direct object is human and is followed by a PP designating a body part (Viberg 1999). The physical contact is combined with various interpersonal meanings (expression of emotion, non-verbal communication) as in (5), which literally means: they took her on the breasts.

(5)

De tog henne brösten och hon höll vindtygsjackan öppen, AP

They touched her breasts when she held her wind-cheater open,

24Greeting by shaking hands can be expressed ta i hand ’take in hand’ in Swedish and involves physical contact rather than physical possession and functions as a form of non-verbal communication.

(6)

Han lämnade bordet, tog doktor Pretorius i hand och bockade ARP

He left the table, took Dr Pretorius' hand and bowed,

25Example (7) provides an example, how the meaning can be further extended to physical contact that does not involve the hand (Literally: the knees almost took in the chin).

(7)

Sanna låg på sidan på en av sängarna med benen högt uppdragna så att knäna nästan tog i hakan. MPC2: ÅL

Sanna was lying on her side on one of the beds, her knees drawn up so that they were almost touching her chin.

26Table 4 gives a somewhat more detailed picture of taking as a goal-directed action sequence but is still sketchy in many respects and not all the related meanings of ta will be commented on in detail. A physical action sequence can be experienced both from within by the actor based on proprioception and from the outside by an observer based on visual and other sensory information (see Viberg 2004 for a similar characterization of Swedish slå ‘hit, strike’). At a more general level, the representation is inspired by Barsalou’s (1999, 2003) situated simulation theory according to which conceptual representations are multi-modal simulations that are distributed across modality-specific systems. A central assumption relevant for the view of the representation of ta in Table 4 is that a given simulation for a concept is situated and prepares an agent for situated action. Each step in the action sequence is linked to a number of characteristic functions which represent the goal of the situated action. A simple example is expressions like ta ett piller ‘take a pill’ or ta ett glas ‘take a glass’ which literally refer to the grasping of the object but tend to be extended to cover the goal, namely ingestion of the pill or (by metonymic extension) the contents of the glass.

27That the meaning of physical action verbs include sensorimotor information has been supported most clearly in neuroscientific studies. Damasio & Tranel (1993) have shown that verbs activate motor control regions, while nouns do not. More specifically, Pulvermüller (2001) has shown that words like kick, pull and lick activate appropriate areas in the motor strip. Obviously, our own actions are important for learning the appropriate sensorimotor associations of physical action verbs. In order to describe actions carried out by other people we must link these associations to our spatial perception. Thus both bodily activity and spatial perception form part of the meaning representation in Table 4.

Table 4. An extended version of taking as a goal-directed action sequence

Reaching & Touching

Grasping

Handling

Displacement

Sensorimotor:

Bodily activity of Agent

Extend arm towards object

Grasp object with hand

Move arm holding object

Walk while holding

Spatial perception

Approaching zero distance between Hand and Object

Object enclosed by Hand

Object moves   within the region of interaction of Agent

Joint motion of Object and Agent

Functional result

Physical contact with its associated meanings:

(a) examine by touch

(b) display of emotion

(c) non-verbal communication

Physical Possession,

Control

(a) Possession

(b) Manipulation

(c) Placing

(d) Removal

(e) Ingesting

(f) Change of  ownership

Change of place of object + Maintaining of physical possession

(followed by transfer to another person)

28Each step in the action sequence is related to functional interpretations that usually represent the profiled part of the meaning and that have a tendency to develop into meanings that are independent of the physical action sequence. Parallels to this appear in the semantic patterns of slå ‘hit’ and other Swedish physical contact verbs (Viberg 1999). Grasping has not been commented on so far, but this meaning can be profiled though it is usually expressed in combination with a cognate noun as a stressed particle ta tag i ‘take hold of’. As already noted in relation to example (3), ta often refers to induced motion of the object and appears in constructions characteristic of such verbs in combinations with spatial particles and/or PPs. In example (3), the Agent is stationary, which means that the manipulation of the object as an instrument is foregrounded. Ta is more similar to an ordinary motion verb, when the Agent moves together with the object as in (8). In this example, ta is translated with its closest correspondent only in German.

(8)

Kalle fick ta in sitt juiceglas bara. LM

Kalle was allowed to bring his juice in, nothing else.

Kalle durfte nur sein Glas mitnehmen.

Kalle a seulement eu le droit d'emporter son verre de jus d'orange.

Kalle sai vain viedä mehulasinsa TV-huoneeseen.

29In (8), an important aspect of the meaning is the fact that the Agent possesses and controls the object. (More typical induced motion verbs such as throw and put do not suggest continued control in this sense.) This is coded explicitly in the frequent combination of ta and the preposition med ‘with’, optionally combined as in (9) with the reflexive pronoun (sig). The preposition med is also used to express accompaniment and instrumentality, which are notions that tend to be associated with possession crosslinguistically and a frequent marker of predicative possession (in the Companion schema: X is with Y, Heine 1997). These meanings of med are closely related in the various uses of ta.

(9)

Hon tog med sig en kopp te upp på rummet, MF

She took a cup of tea up to her room.

Sie nahm sich eine Tasse Tee mit hinauf ins Zimmer,

Elle emporta une tasse de thé dans sa chambre,

Anna otti kupillisen teetä mokaan huoneeseensa,

30In (9), ta is both a motion verb and a possession verb. The object is held in the hand while the agent is walking in order to maintain control of the object and use it for a certain purpose on arrival. Actually, the purpose of the displacement of the object may be to transfer the possession to someone else on arrival, which can be expressed by adding a prepositional phrase with till ‘to’ indicating the receiver as in (10).

(10)

Ta med dig något gott från köket upp till ungarna du. ESPC: GT

Take something good with you from the kitchen up to the kids.

31The notion of accompaniment and control is present also when the object is human as in (11), even if the notion of physical control (holding etc) often is absent and replaced by some kind of verbal control ( is a dialectal spoken form of med ‘with’).

(11)

Ta pojken dej. MF

Take the boy with you.

Und den Buben nimmstmit.

Et emmène le gamin.

Ota poika mukaas.

32In (12), ta is constructed as an induced motion verb (V + Directional particle + Object + PPgoal). Also in this case the notion of control is present, even if it is exerted verbally, which is reflected in the French translation (appeler ‘call’).

(12)

Josefina hade tagit in honom och Eeva-Lisa i köket, POE

Josefina had taken him and Eeva-Lisa into the kitchen,

Josefina hatte ihn und Eeva-Lisa in die Küche geholt,

Josefina les avait appelésdans la cuisine, lui et Eeva-Lisa,

Josefina oli vienyt hänet ja Eeva-Lisan keittiöön,

33Table 5 shows the degree to which the closest correspondents of ta are used as translations in the MPC corpus when ta refers to concrete possession and to motion.

Table 5. Correspondents of ta in the MPC corpus

Swedish

English

German

French

Finnish

ta

take

nehmen

prendre

ottaa

All meanings

N

601

268

198

132

207

%

45%

33%

22%

34%

Concrete possession

N

79

49

55

31

62

%

62%

70%

39%

78%

Motion:

Induced motion

N

101

59

38

11

49

Self-motion

N

30

0

0

0

0

34As can be observed in Table 5, there are 601 occurrences of ta in total. Out of these, only 268 occurrences are translated with take in English in spite of the fact that the verbs are etymologically related. This represents 45% of the cases when ta is used in the Swedish originals. In the other languages, the correspondence is even lower. It is lowest in French, where prendre is used as a translation in only 22% of the cases. Still, prendre is the most frequent translation. Table 5 also shows the frequency of some of the major sub-senses of ta. In 79 out of the 601 occurrences of ta, the verb refers to concrete possession. In this case, the closest corresponding verb is used as a translation to a rather great extent. In 49 of the 79 cases, English uses take as a translation, which represent 62% of the uses of ta referring to concrete possession. In French, the correspondence reaches only 39% but this is higher than what is the case when all meanings are taken into consideration.

35As is shown in Table 5, the closest equivalents of ta as a concrete possession verb can be used as translations also when ta is used as an induced motion verb but to a much lesser extent. (Only absolute numbers are given for this and the following sub-sense.) In French, prendre is used as a translation only 11 times (out of 101, i.e. 11%). In many cases, French uses directional verbs (such as sortir) or one verb in a set of verbs contrasting with respect to motion to vs. from the focus of interaction and animacy: apporter/emporter (to/from focus) and amener/emmener (restricted to human Themes). It should be noted, that ta is counted as a motion verb whenever motion is explicitly coded which does not rule out that ta simultaneously indicates possession.

36In a reflexive form (ta sig ‘take’ + reflexive pronoun), ta can be used as an intransitive motion verb referring to self-motion as in (13), which literally means ‘The thieves had taken themselves in through a window’. This use is rather language-specific and is never translated with the closest correspondent of ta in the MPC languages in the present material.

(13)

Tjuvarna hade tagit sig in genom ett fönster på baksidan av huset. HM

The burglar came in through a back window.

Die Diebe waren durch ein Fenster auf der Rückseite des Hauses eingestiegen.

Les voleurs étaient entrés par une fenêtre à l'arrière du bâtiment.

Varkaat olivat tulleet sisään takapihan puolella olevasta ikkunasta.

37Unlike Swedish motion verbs in general, ta sig is not clearly associated with any specific manner of motion and is neutral in that respect but there is a tendency that ta sig indicates that some problem was involved as in (14) or that the manner of motion as in (13) was unconventional (to say the least).

(14)

Hon hade inte nån bil att ta sig ner till byn med. KE

She had no car to get down into the village.

Sie hatte kein Auto, um ins Dorf hinunterzufahren.

Elle n'avait pas de voiture pour redescendre au village.

Hänellä ei ollut autoa millä ajaa alas kylään.

38’Take’ and ‘give’ are often analyzed in a parallel way: John took the book from Mary vs. John gave the book to Mary. Actually, there are relatively few examples where Swedish ta is used in examples of this type. It is possible to use a stressed particle ifrån ‘from, off’ John tog ifrån Mary boken but it is also common to express the bereaved party as a possessive attribute as in John tog Marys bok ‘John took Mary’s book. An actual example of the latter construction is found in (15).

(15)

Om han blir fälld för misshandel så tar dom väl hans vapen? KE

If he's charged with assault, they'll takehis guns off him, won't they?

Wenn er wegen Körperverletzung verurteilt wird, dann nehmen sie ihm wohl seine Waffen weg?

-S'il est condamné pour coups et blessures, ils lui retireront son fusil, n'est-ce pas?

- Jos hänet tuomitaan pahoinpitelystä, niin ottaako ne hänen aseensa takavarikkoon?

39In German and French, taking is modeled on giving. In (15), the bereaved party is expressed with a pronoun in the dative in these two languages (but often with an alternative to prendre ‘take’ in French; retirer ‘remove’, ‘deprive’), whereas it appears as a possessive attribute in Finnish. It should be added that the parallelism is reflected in the choice of preposition (à) in French when the receiver or bereaved party is not pronominal (donner qc à qn/ prendre qc à qn). This represents part of an areal pattern in Europe. According to Janda (1997), 'take' is modelled on the verb 'give' in West and South Slavic languages, whereas in East Slavic 'take' is modelled on the Locative-Existential construction, which corresponds to the concept 'have'.

40It is not possible to go into general theoretical questions concerning the representation of polysemous words in any detail in this paper (see Croft & Cruse 2004 for an interesting proposal). The representation in Table 4 can be regarded as a sketch of the meaning potential of ta. When some form of ta appears in an utterance, this activates a number of typical interpretations which are selected by linguistic cues (such as the construction), information concerning the situation and prior discourse and general world knowledge. The functions associated with each step in the action sequence tend to be associated with a characteristic set of meanings exemplified above with Physical contact. The functional meanings usually represent the profiled part of the meaning and can also develop into independent meanings which no longer include the concrete action sequence. The analysis presented above is incomplete since many abstract and metaphorical meanings have not been taken into consideration, but the action sequence motivates a major part of the pattern of meanings expressed by ta.

4 The Swedish verb ge ‘give’

41As was shown in the previous section, ta still has many characteristics of a hand-action verb. What about Swedish ge? This verb appears to have a more abstract meaning. An attempt will nevertheless be done to provide a description of this verb similar to that of ta.

4.1 Can Swedish ge be regarded as a physical action verb?

42Newman (1996) gives a detailed and insightful account of the meaning of GIVE with reference to a number of different domains such as spatio-temporal, control, force-dynamics and human interest. The following account in accordance with the general theme of this paper concentrates on the question to what extent Swedish ge can be regarded as a physical action verb without denying the relevance of other dimensions of the meaning of ge. It proceeds from what Newman refers to as “the typical scenario of the act of giving”, where “there is a person who has some thing and this person passes over the thing with his/her hands to another person who receives it with his/her hands.” (Newman 1996, 37). In (16), ge appears in a ditransitive construction with the Receiver as an indirect object and the Gift as a direct object and is translated with its closest correspondents in the MPC languages (English give, German geben, French donner and Finnish antaa. In German and French, the Receiver is marked with dative case. In Finnish it is marked with the allative –lle: Reinelle ).

(16)

Morsan gäspade. Hon tog upp plånboken och gav honom en tia förutom det han redan fått. PCJ

Mom yawned, then took out her wallet and gave him a ten-kronor note to add to the one he had already.

Die Mutter gähnte. Dann nahm sie ihren Geldbeutel heraus und gab ihm noch einen Zehner zu dem, den er schon bekommen, hatte.

Maman bâilla. Elle sortit son portefeuille et lui donna un billet de dix couronnes en plus de ceux qu'il avait déjà.

Äiti haukotteli. Hän otti esiin lompakkonsa ja antoi Reinelle kympin, vaikka oli antanut jo yhden.

43This example depicts a scene where the act of giving can be interpreted as a goal-directed physical action sequence leading up to both physical possession and transfer of ownership as in Table 6. The interpretation of GIVE as a physical action verb is even more evident in the well-known type of alternation as in (17), where the Receiver is marked with a spatial directional marker. Only the word order alternates in Finnish which uses the allative in both (16) and (17). (This case is used also with concrete physical object such as sillalle ‘to the bridge’.) The English translation in this case is hand which clearly is a physical action verb.

(17)

Hon gav papperet till Ingvar Johansson. LM

She handed the paper to Ingvar Johansson.

Sie übergab das Blatt an Ingvar Johansson.

Elle donna le papier à Ingvar Johansson.

Hän antoi paperin Ingvar Johanssonille.

Table 6. Giving as a goal-directed action sequence

Sequence of bodily actions

Goal/Function

Hold object in hand

Extend arm towards Receiver while holding object

Loosen grasp and let go of Object when Receiver has grasped Object

a) Physical possession of object by Receiver

b) Transfer of ownership

44Actually, there is experimental evidence, provided by Glenberg and his associates, that the meaning of give has bodily grounding:

“As another example of grounding in bodily states, Glenberg and Kaschak (2002) asked each participant to judge the sensibility of sentences such as ‘You gave Andy the pizza’ or ‘Andy gave you the pizza’ by moving the hand from a start button to a Yes button. Location of the Yes button required a literal movement either toward the body or away from the body. Responding was faster when the hand movement was consistent with the action implied by the sentence (e.g. moving away from the body to indicate ‘Yes’ for the sentence, ‘You gave Andy the pizza’) than when the literal movement was inconsistent with that implied by the sentence. Apparently, understanding these action sentences called on the same neural and bodily states involved in real action.” (Glenberg et al. 2005, 117-118).

45Actually, Glenberg & Kaschak refer to the interpretation of the sentence but it is reasonable to assume that the meaning of the verb provides clues to this interpretation.

46Swedish ge does not in the same way as ta have uses that profile the earlier steps in the action sequence such as the extension of the arm. This is true only of räcka ‘reach’ which can be used to describe the transfer of physical possession in a similar way as hand (see example (18)).

(18)

räckte Ronja honom mjölkflaskan, AL (MPC)

Then Ronia handed him the jar of milk,

47The verb räcka can be used also with reference to the subevent of extending the arm as in (19).

(19)

Hon räckte ut handenefter  telefonluren. KE

She reached for the telephone.

48Räcka is an example of verb that can be used both to describe hand actions and as a verb of physical possession in a way that appears to be parallel to earlier historical stages of canonical possession verbs.

4.2 Extended meanings of ge

49Even though it is possible to find examples where reference to physical action is part of the meaning expressed by Swedish ge, such examples are less prominent than is the case with ta. An earlier study based on the complete set of occurrences of ge in the ESPC (Viberg 2002b) showed that only 17% of the examples referred to concrete possession in Fiction, and as little as 4% in Non-fiction. Table 7 complements these data with a summary of the degree to which ge in various meanings is translated with its closest correspondents in the MPC languages.

Table 7. The degree to which the closest correspondents of ge are used as translationsin the MPC

  • 2  sich begeben (not included in All meanings)
  • 3  abandonner (not included in All meanings)

Meaning

Swedish

English

German

French

Finnish

ge

give

geben

donner

antaa

All meanings

N

0

0

107

0

0

%

50.2

43.3

19.8

35.2

Concrete possession

N

52

43

36

26

43

%

82.7

69.2

50.0

82.7

Abstract possession

N

82

47

31

18

26

%

57.3

37.8

22.0

31.7

Self-motion

N

44

0

0/(2)2

0

0

Yield

N

39

30

31

0 /(13)3

16

Other meanings

N

30

4

9

5

2

50As can be observed, the degree of correspondence varies dramatically among the MPC languages when all meanings are taken into consideration, from 50% (round figure) to 20% in French. The correspondence is highest with reference to concrete possession, where it is as high as 83% in English and Finnish and reaches at least 50% in French.

51Abstract possession refers to the case where the object is an abstract noun such as in give an answer, give an idea, give a chance. In such uses, the degree of correspondence is much lower than in concrete possession, from 57% in English to only 22% in French. When the abstract nouns are classified semantically, it turns out that most of these nouns belong to a restricted number of semantic fields (see Viberg 2002b for a more detailed description). One such field is nouns referring to physical contact, which in many cases refer to an action closely related to the use of ge as a physical action verb as in (20). The arm is extended in order to give a punch or a slap with the hand. (Examples such as give a kick in the physical sense can be regarded as further extensions from examples of this kind.)

(20)

Birk gav henne en knuff. AL

Birk gave her a nudge.

Birk gab ihr einen Knuff.

Rik la poussa un peu

Birk tönäisi häntä.

52Similar to prototypical giving, examples of this type use the ditransitive construction and refer to interpersonal interaction. The same applies in principle to the most frequent type of abstract nouns which refer to verbal communication as in (21), even if the Receiver (or Addressee) can be understood as in this example. Rather often, a simple verbal communication verb is used as a translation as it is in this case in English, French and Finnish, in spite of the fact that all the languages have combinations of the type ‘give’ + Verbal communication noun.

(21)

- Kan du ge en kommentar kring Christina Furhages kvarskrivning? frågade hon snabbt. LM

"Can you comment on Christina Furhage being off record?" she swiftly continued.

"Können Sie einen Kommentar zu Christina Furhages Sperrvermerk abgeben?" fragte sie schnell.

- Pouvez-vous commenter l'absence d'inscription de Christina Furhage ? s'empressa-t-elle de demander.

- Voitko kommentoida Christina Furhagen tapausta? hän kysyi nopeasti.

53Across languages, the abstract nouns which often are combined with ge as an object tend to belong to the same semantic fields as the verbs that have a tendency to appear in the ditransitive construction. Such nouns also evoke the same semantic roles (or frame elements) as the ones that tend to be expressed by grammatical markers (case, adpositions) of the Dative (Receiver, Addressee, Experiencer; but not to the same extent Beneficiary).

54Similar to ta, ge in the reflexive form is used as a verb of self motion in combination with various particles and other spatial markers. Most of these combinations refer to departure (for example ge sig iväg lit. ‘give onself away’ and ge sig av ‘give oneself off’). A frequent English translation is leave as in (22). The rest of the MPC languages also use verbs of departure as translations (German weggehen, French s’en aller and Finnish lähteä ‘leave’). As can be observed in Table 7, this represents a language-specific use of ge. The correspondents of ge, never appear as translations except for two occurrences of sich begeben in Geman, which is a reflexive form of geben combined with the prefix be-. (This form has a closer correspondent in Swedish bege sig which is not associated with departure).

(22)

Hon fick inte ge sig av. KE

She was not to leave,

Sie durfte nicht weggehen.

Elle ne devait pas s'en aller.

Tämä ei saanut lähteä tiehensä.

55A further use which reaches a certain frequency but not will be discussed here is called Yield and refers to the correspondents of give up and give in in English. There is, however, one use of Swedish ge which reached a high frequency in the Non-fiction part of the ESPC (25% vs. 4% in Fiction) and that is the use of ge as a verb of production (see Viberg 2002b). Typically, the product or result appears as a direct object of ge without any overt indirect object as in (23) and (24). In most cases, there is an understood Receiver or Interested party (‘us humans in general’). This use was much more frequent than in English where it appears to be more restricted (e.g. Cows give milk).

(23)

Förbränning av fossila bränslen ger också stora utsläpp av koldioxid som bidrar till växthuseffekten. ESPC:EVIR

The burning of fossil fuels also produces large amounts of carbon dioxide which contributes to the greenhouse effect

(24)

Däremot förefaller det sannolikt att mineralullsfibrer ger ögonirritation. ESPC:BJ

On the other hand, it does appear probable that mineral wool fibres cause eye irritation.

5 The Swedish verb

56The verb is one of the most frequent verbs in Swedish. It has the rank of 4, if verbs are ordered according to descending frequency. It has a complex pattern of polysemy including frequent grammaticalized uses as a modal and as a periphrastic causative (see Viberg 2002a, 2006a; Ramnäs 2006). Etymologically is derived from a verb meaning ‘catch’ (present-day Swedish fånga ‘catch’). Not many traces are left of the original meaning, the closest being the notion of Success in certain constructions (see Viberg 2002a). The following material is taken from work in progress based on the complete set of occurrences of in the MPC. (Due to the complexity of this verb, the presentation here represents a brief summary of work in progress.) Already as a lexical verb of possession with the meaning ‘come to possess’, it is relatively language-specific as can be seen from examples (25)-(27) from the MPC.

(25)

du får tio öre. IB

You'll get ten öre.

du kriegst zehn Öre

Je te donnerai dix öre.

saat kymmenen äyriä.

(26)

Två gånger om dagen eller natten fickhan vatten och mat. HM

Twice each day or night hewas given water and food.

Zweimal am Tag oder in der Nacht bekamer Wasser und Essen.

Deux fois par jour ou par nuit, onlui donnait de l'eau et de la soupe.

Kaksi kertaa päivässä tai yössä hänsai vettä ja ruokaa.

(27)

Hon fick aldrig fler barn. MA

She never had any more children.

Sie hat nie weitere Kinder gekriegt.

Elle n'a jamais eu d'autres enfants.

Hän ei saanut muita lapsia.

57The closest equivalent of in English is get but, as can be observed in Table 8, the frequency of give as a translation, especially in passive form, approaches that of get, when has a concrete noun as object (Concrete possession). In French, it turns out that the most frequent translations are avoir ‘have’ and ‘donner ‘give’, whereas recevoir which superficially may appear to be the closest equivalent of is used as a translation only 18 times (11% of all cases). Often, as in (26), donner has an impersonal form with the ‘generic’ subject on ‘one’. In (27) Frenchavoir ‘have’ used in the past perfective form which indicates a transition when used with a basically stative verb. This justifies the conclusion that French does not have any direct equivalent of but rather uses some other basic possession verb or one of the number of more specific verbs at a level below the one that is basic in Swedish. Other verbs of possession that are used as translations of are more specific in meaning (number of occurrences within parentheses): obtenir (8), toucher (5; with object referring to ‘money’), offrir (4), (re)trouver (6), attribuer (2), distribuer (2), laisser (2), payer (2), prendre (2). This means that French has no verb with the general meaning ‘come to possess’ but a set of verbs that semantically correspond to a set of hyponyms of . (See Ramnäs 2006 for a more detailed analysis of the French equivalents.)

58German has two rather direct equivalent of as a verb of possession, bekommen which has highest frequency in written varieties and kriegen which is more frequent in spoken varieties. Finnish has a rather close equivalent saada, which also shares many of the extended uses of .

Table 8. The most frequent translations of as a verb of possession in the MPC.

N

English

German

French

Finnish

Concrete

possession

166

get

45

be given/give

35

bekommen

104

kriegen

25

avoir

25

donner

25

recevoir

18

saada

135

Abstract

possession

228

get

43

be given/give

13

bekommen

56

kriegen

9

avoir

36

donner

2

recevoir

16

saada

104

59The most frequent translations in the MPC languages when refers to Concrete possession are used also when is combined with an abstract noun in object position as in (28) but to a much lower extent. The closest equivalent is used as a translation except in French which does not have a clear equivalent and in this case uses recevoir.

(28)

hon ger en kyss och får många tillbaka.

She gives him one kiss and gets many in return.

sie gibt einen Kuß und bekommt viele zurück.

Christina reçoit plus qu'elle ne donne.

antaa suudelman ja saa monta takaisin.

60In many cases, a simple verb is used as a translation which incorporates the meaning of the abstract noun. Often this verb appears in passive form as in the English translation in (29), which corresponds to få besked om ‘get information about’ in the Swedish original. German and French use an active verb with a generic subject (‘one’) as translation, whereas Finnish uses saada kuula ‘get (to) hear’.

(29)

Han fick besked om att Åkeson var i Malmö hela dagen. HM

but [he] was told that Åkeson would be in Malmö all day.

Man teilte ihm mit, daß Åkesson den Tag über in Malmö sei.

On l'informa qu'Åkeson était à Malmö.

Hän sai kuulla, että Åkeson olisi koko päivän Malmössä.

61As already mentioned, has a complex pattern of polysemy. Only a few examples can be given here of the two major grammatical uses as a modal and as a periphrastic causative. As a modal, is combined with a bare infinitive. It has primarily a deontic meaning, which, depending on the situation, can be interpreted either as Permission (30) or as Obligation (31). (Example 8 above contains another example of in the Permission sense).

(30)

Dom som ville fick vara uppe hela natten... PCJ

Anyone who wanted to was allowed to stay up all night...

Wer wollte, durfte die ganze Nacht aufbleiben...

Ceux qui voulaient ont eu le droit de rester debout toute la nuit...

Jos halusi sai valvoa koko yön...

(31)

Nån av er får ta tag i saken. HM

One of you will have to look into it.

Einer von euch muß die Sache in die Hand nehmen.

Quelqu'un doit s'en occuper.

Joku teistä saa ottaa asian hoitaakseen.

62The two meanings are represented by two contrasting sets of translations in English, German and French, whereas the Finnish verb saada ’get’ can be used as a translation in both cases (but to a more limited extent in the Obligation sense). The major translations are shown in Table 9. (Other alternatives exist and will be accounted for in an article in progress together with statistical information. There are also complicating factors such as negation of permission.)

Table 9. The major translations of modal in the MPC.

Meaning

Translations into

Modal:

English

German

French

Finnish

Permission

be allowed

can

dürfen

können

pouvoir

saada

Obligation

have to

müssen

falloir

devoir

saada

joutua

63The second grammatical use is the use as a periphrastic causative (see Rawoens 2008 for a detailed description), where is combined with a direct object and an infinitive with the infinitive marker att ‘to’ as in (32).

(32)

TV-kamerornas blåaktiga ljus lyste upp scenen och fick snöflingorna att gnistra. LM

The TV cameras' bluish lights were illuminating the whole scene, making the snowflakes sparkle.

Das bläuliche Licht der Fernsehkameras erleuchtete die Szenerie und ließ die Schneeflocken funkeln.

La lumière bleuâtre des caméras éclairait la scène et faisait crépiter les flocons de neige.

TV-kameroiden sinertävä valo valaisi tapahtuma paikan ja sai lumihiutaleet säteilemään.

64The major translation equivalents of as a periphrastic causative are shown in Table 10 together with information about the major equivalents of the other frequent Swedish periphrastic causative låta ‘let’ (see Viberg 2009 for a more detailed description). The closest correspondent of is used as a translation only in Finnish, where saada ‘get’ signals coercive causation in contrast to permissive causation which is signaled by another basic possession verb antaa ‘give’. In the other MPC languages, permissive causation is expressed with verbs that are etymologically related to låta ‘let’. In German, lassen, is used both to signal permissive and — in alternation with bringen — coercive causation.

Table 10. The major translations of causative låta ‘let’ and in the MPC.

Meaning

Swedish

English

German

French

Finnish

Permissive Causation

låta

let

lassen

laisser

antaa ‘give’

Coercive Causation

make

lassen

bringen

faire

saada ‘get’

65The pattern of polysemy which distinguishes has an interesting areal distribution. The Norwegian cognate shares most of the meaning patterns with the Swedish verb, whereas the Danish cognate has similar uses primarily as a verb of possession and as a periphrastic causative. Interestingly, Finnish has the etymologically unrelated verb saada, which has a pattern of polysemy that closely resembles the Swedish one (except that modal obligation is not as prominent as in Swedish). Semantically, English get shows many parallels to at a general level in the sense that the verb basically refers to possession and also has modal, causative and inchoative extended meanings, but as shown in Viberg (2002a) the meanings tend to be different at a more precise level to such a large extent that the two verbs serve as the translations of one another only in a limited number of cases. The polysemy of get is treated in Gronemeyer (1999), which also accounts for the diachronic development of its meanings from Middle English to present-day English. The extension of into various non-physical domains (referred to as abstract possession in this study) has parallels in the development of other acquistional verbs in English as demonstrated by Nordlund (2008) in her corpus-based study of the extension from physical to mental acquisition of the verbs gather, grasp, seize, acquire, buy and receive.

66In particular with respect to the modal meanings, there are many interesting parallels to the semantic extensions of Swedish in Mainland Southeast Asian languages described in Enfield (2003), who refers to the spread of these characteristic patterns of polysemy as linguistic epidemiology. Van der Auwera, Kehayov & Vittrant (2009) refer to Northern Europe and South(east) Asia as two hotbeds of ‘acquisitive modality’ in a revised version of modality’s semantic map (van der Auwera & Plungian 1998). For another parallel, see the notes on Swahili in the next section.

6 Putting the result in a typological perspective

67There is no description of the semantic field of possession verbs from a general typological perspective. In what follows, a brief summary will be given of relevant studies and examples from some languages that represent various alternative ways of structuring the field. I hope to be able to return to this in a more systematic study.

6.1 Basic verbs of possession in the MPC languages

68Before summing up the result of the corpus-based study, something must be said about HAVE which has not been treated in this paper. From a typological perspective, this verb and its correspondents have been extensively studied first by Heine (1997) and more recently in even greater detail by Stassen (2009). As a background to the typological sketch below, a brief account will be given of Stassen’s work according to which there are four basic types of predicative possession based on the encoding of the grammatical functions of the Possessor (PR) and the Possessee (PE). The Have-Possessive is syntactically transitive and the possessor takes the encoding characteristic of the agent NP in the language and the possessee is encoded like the patient: Peter (PR) has a book (PE). Heine (1997, 47-50) refers to this as an instantiation of the Action Schema. The other basic types identified by Stassen are all intransitive. In the Locational Possessive, the Possessor is encoded as some oblique, adverbial case form, while the Possessee is encoded as the grammatical subject of the predicate which is existential/locative. Schematically, Stassen represents this type: At/to PR, (there) is/exists a PE. In the With-Possessive, the predicate is also of the existential/locative type, but the Possessor is encoded as grammatical subject, whereas the Possessee is oblique or adverbial: PR is/exists with a PE. Even the last type, the Topic Possessive, has a locative/existential predicate. Similar to the Locative type, the possessee is constructed as the grammatical subject but in this type the possessor is constructed as the sentence topic: (As for) PR, PE is/exists.

69Table 11 shows in schematic form the systems of basic possession verbs in the languages described in this paper using the semantic grid introduced in Table 2.

Table 11. The realization of the basic grid of verbs of possession in English, German, French and Finnish

Agrandir

70The grid presented in Table 11 accounts primarily for the appearance of simple verb roots. Obviously, the concepts GIVE, TAKE, GET and HAVE can be expressed even in many of the languages that do not have a simple verb root expressing one or more of these concepts. It is interesting to look at all the possible ways to express the meanings in the semantic grid on which Table 11 is based, both with lexical and non-lexical coding strategies and the description should be extended in this way. For HAVE, we already have such descriptions in Heine (1997) and Stassen (2009). For GIVE and TAKE, the account given above can be complemented with the typological study of the encoding of three-participant events presented in Margetts and Austin (2007). Several of the semantic classes of three-participant events in that study (presented in their Table 1 p. 398) involve possession as a central concept. Margetts and Austin are concerned with the way the participants of such events are syntactically coded as arguments and describe this in terms of a number of encoding strategies of three-participant events. When all three participants are expressed as direct arguments of the verb, there is a direct argument strategy (of the type: A gives B C ). However, three-participant events that are syntactically expressed with three-place predicates in English and other well-known languages are expressed with a variety of constructions that are syntactically two-place in many languages. In such cases, three-participant events are expressed with a number of encoding strategies such as the causative strategy, the applicative strategy (see example 36), the oblique or adjunct strategy, where a third participant is marked with an oblique case or an adposition (A give C to B), or a serial verb strategy. From the perspective of lexical typology, such strategies can be regarded as a means to express the concepts that are not coded as simple verbs in a certain language. They can also be used to describe the contrasting ways in which argument structure is realized in languages that have verbs that correspond to one another, for example the various argument structures appearing with TAKE in the European language treated above. Encoding strategies of a similar type can be described for the expression of possessive two-participant events such as the examples below that show how GET tends to be expressed with a passive or impersonal third person form of GIVE in several European languages.

71The motivation for trying to a establish a lexical typology and for paying special attention to the extent to which various semantic components are encoded as simple verb roots rather than more analytically is that this is thought to have major consequences for what meanings tend to be expressed most frequently within a certain semantic field, for the syntactic organization in particular with respect to argument structure, for what is easy to process in production and comprehension and for the way the semantic lexicon in general is organized. The semantic aspects are most important for lexical typology. To begin with, an economic way of expressing a meaning such as a simple verb root rather than a more analytic construction can influence what aspects of a situation tend to be expressed explicitly. For motion verbs, Slobin (2003) has shown that manner of motion is encoded more frequently in satellite-framed languages with many manner of motion verbs than it is in verb-framed languages. With respect to the overall organization of the lexicon, the meanings of the basic verbs of possession serve as superordinates of more specific verbs (see 6.2). Each of the basic verbs tends to be associated with recurrent patterns of polysemy across languages (see also Heine & Kuteva 2002 for grammaticalized meanings). This means that the semantic grid can serve as a starting point for predictions (things to look for) as soon as the type is known and it can also serve as a way to organize these patterns of polysemy.

72In Swedish and English, the most basic verbs of possession meaning ‘have’, ‘get’ and  ‘give’ in combination with abstract nouns  form  a very productive system generating complex predicates which represent states, inchoatives or causatives. The same dynamic contrasts are basic when complex predicates are formed with adjectives: ‘be’, ‘become’ and ‘make’. This represents the core of the dynamic system. Sometimes it is possible to form a complete set of parallel predicates as  shown schematically in Table 12 (taken from Viberg, 1981), in which various ways to form emotive predicates related to the concept happiness are shown.

Table 12. The role of the basic possession verbs in the dynamic system in Swedish.

Word class

Dynamic meaning

Noun

Adjective

Verb

STATE

X hade glädje av Y

X var glad (åt Y)

X gladdes/ gladde sig

INCHOATIVE

X fick glädje av Y

X blev glad (åt Y)

(åt Y)

CAUSATIVE

Y gav X glädje

Y gjorde X glad

Y gladde X

have/get/give happiness

be/become/make happy

Emotion verb

73To express the inchoative, for example, it is possible to say either X fick glädje ‘X got happiness’, X blev glad ‘X became (got) happy’ or to use a passive (gladdes) or reflexive (gladde sig) form of the verb glädja ‘make happy’, which in its basic form has a causative meaning. (The example is somewhat idealized, since all possibilities are seldom realized with a single concept such as ‘happiness’).

74Among European languages, the system with four basic verbs of possession GIVE, TAKE, GET and HAVE appears to be characteristic primarily of Germanic languages. (As we have seen, this holds a little less clearly in English where get alternates with a passive form of give to express concrete possession.) The Romance languages appear to be similar to French with respect to the lack of a general equivalent of GET. The Baltic and Slavic languages follow several patterns. With respect to predicative possession according to Isačenko (1974), Serbo-Croat (sic.), Czeck, Polish and Lithuanian are Have-languages, while Russian and Latvian are not. Some other Slavic languages seem (according to Isačenko) to be in a stage of transition to Have-languages, namely Ukrainian and Belorussian. With respect to GIVE and TAKE, TAKE is modelled on the verb GIVE in West and South Slavic languages, whereas in East Slavic TAKE is modelled on the Locative-Existential construction, which corresponds to the concept 'have' as already mentioned with reference to Janda (1997). GET seems in general to have a correspondent in Slavic, but the degree to which this verb is used is difficult to assess without a corpus-based study. The data in (33a-f) provided by Ljuba Veselinova for Bulgarian and by Maria Koptjevskaja-Tamm for Russian as an answer to a simple questionnaire shows some of the characteristics of Bulgarian and Russian basic possession verbs.

(33)

a.

Petar

ima

mnogo

knig-i

Peter

have.3.sg.pres

many

book-pl

a.RU

U

Petr-a

mnogo

knig

at

Petr-gen.sg

many

book:gen.pl

Peter has many books.

b.

Petar

poluchi

kniga

Petar

got

book

b.RU

Petr

poluchi-l

knig-u

Petr

get.pfv-past.m.sg

book-acc

Peter got a book.

75(33b) indicates that Bulgarian and Russian have a verb corresponding to GET but the use of GIVE in an impersonal third person plural form is a frequent alternative. The Bulgarian (33b) is said to be somewhat odd unless the giver is specified. Especially in a more colloquial style, one of the verbs davam ’give’ or podarjavam ‘give as a present’ is used as in (33c).

(33)

c.

Na

Petar

mu

podar-ixa

kniga

to

Petar

3.sg.m.dat

give.as.present-3.pl.aorist

book

c.RU

Petr-u

podari-l-i

knig-u

Petr-dat.sg

give.pfv-past-pl

book-acc

Peter got a book.

76(33d) shows GIVE in its basic syntactic frame.

(33)

d.

Maria

podar-i

edna

kniga

na Petar

Maria

give.as.present-3.sg.aor

a

book

to Petar

d.RU

Marij-a

podari-l-a

knig-u

Petr-u.

Maria-nom.sg

give.pfv-past-f.sg

book-acc.sg

Petr-dat.sg

Maria gave Peter a book.

77In Bulgarian, a construction with a pronominal receiver is preferred with davam ’give’ as in (33e). (The pronominal clitical construction is typical of the Balkan languages and not applicable to Russian.)

(33)

e.

Maria

mu

da-de

kniga-ta

(na

Petar)

Maria

3.sg.dat

ge-3.sg.aor

book-3.sg.f.def

(to

Petar)

Maria gave the book to Petar.

78The constructional parallelisms between TAKE and GIVE in Bulgarian and between TAKE and predicative possession in Russian are shown in (33f). In Bulgarian, both the Receiver and the Bereaved party are marked with the preposition na ‘to’ and/or a dative pronoun as in (33f).

(33)

f.

Harry

mu

vze

kniga-ta

(na

Petar)

Harry

3.sg.dat

take.3.sg.aor

book-3.sg.f.def

(to

Petar)

Harry took Peter’s book (from him).

79In Russian, on the other hand, the Bereaved party is marked with the locative preposition u in a way that parallels predicative possession in (34aRU).

(33)

f.RU

Harry

vzja-l

knig-u

u

Petr-a

Harry

take.pfv-past.m.sg

book-acc.sg

at

Petr-gen.sg

Harry took Peter’s book (from him).

80It will not be possible to give more than a few examples of basic possession verbs from non-European languages. Four more examples are given in Table 13, which will be commented on in the following sections.

Table 13. Basic verbs of possession in Chipewyan, Sango, Turkish and Swahili

Agrandir

81Chipewyan has already been commented on in section 1.3. Thornell (2002) describes how Sango (spoken in the Central African Republic) has developed a general verb that covers  both of the meanings TAKE and GIVE, which are differentiated by using the verb in different constructions as shown in (34a-b). The verb is adopted from the neighbouring language Ngbandi, where it originally meant TAKE. Partly, this is a contact phenomenon since Sango is regarded as a kind of lingua franca. (Opinions vary as to the exact status of the language in this respect.) It does not represent an intermediary step in an evolutionary sequence between languages with no basic verbs and verbs with GIVE and TAKE. HAVE as described in Thornell (1997, 135) is realized as a With-possessive.

(34)

a.

Kêtê

mamâ

lo

kâsa

lo

small

mother

rel

3sg

take

vegetable

of

3sg

The lady who takes her vegetables… (Thornell 2002, 273)

b.

Ë

na

lo

yê tî dutï

1pl

give

to

3sg

thing to sit on

We gave him something to sit on. (Thornell 2002, 273)

82The closest equivalent of GET is expressed with the verb wara which has the basic meaning ‘find’ as shown in (34c). This verb requires that the giver is indicated in some way as it is in (34c) with the phrase na tï tî (‘to/from hand of’) in order to express GET without ambiguity. Otherwise it could also mean ‘Peter found a knife’. The verb mû ‘give/take’ may also be used as in (34d), which rather means ‘one gave the book to Peter’ (Christina Thornell, personal communication).

(34)

c.

Pierre

a-wara

fadë

zembe

na

Henry

DEM

3sg.find

recent past

knife

to/from

hand

of

Peter got a knife from Henry.

d.

A

na

Pierre

mbêtî

3sg

give

to

book

Peter got a book.

83Turkish appears to represent a frequent type of language, where GIVE and TAKE are realized as simple and frequent verbs. Both vermek ‘give’ and almak ‘take’ appear among the 20 most frequent verbs (Göz 2003). No other verb of possession comes close to that frequency. HAVE is expressed by var ‘existing’ and yok ‘not=existing’ together with the possessor in the genitive (the Genitive schema in Heine, 1997; a non-standard form of the Locative Possessive in Stassen 2009). Almak ‘take’ can be used also to express GET, but in a short Swedish text translated into Turkish, both almak ‘take’ and the passive of vermek ‘give’ appeared as translations of . Turkish examples are shown in (35a-f).

(35)

a.

Ayșe’-nin

çok

kitab-ı  

var.

Ayșe-gen

many

book-poss.3sg

existing

Ayșe has many books.

b.

Ayșe’nin

kitab-ı

yok.

neg.existing

Ayșe doesn’t have any book.

c.

Ali’-den

bir

hediye

al-dı.

Ali-ablative

a

present

take/get-past

She got a present from Ali.

d.

Bir-i

cüzdan-ım-ı

al-dı.

one-3sg.poss

wallet-1sg.poss-acc

take/get-past

Someone has taken my wallet.

e.

Hasan

Ayşe'-nin

kitab-ı-nı

(Ayşe'-den)

al-dı,

Ayşe-gen

book-3sg.poss-acc

Ayşe'-ablative

take/get-past

Hasan has taken Ayşe’s book (from her).

f.

Hasan

Ayșe’-ye

bir

kitap

ver-di.

Ayșe -dative

a

book

give-past

Hasan gave Ayșe a book.

84GIVE appears to be the most widely attested basic possession verb. Haspelmath (2005) in a study of the coding of ditransitive clauses could base the description on the verb ‘give’ in 378 languages. Obviously, there are languages where this meaning is not expressed with a simple verb root, but such languages appear to be relatively few. Already in Newman’s (1997) book, Roberts (1997) provides a detailed description of Amele of Papua New Guinea. In this language as in the other languages of the Gum family, there is no verb stem ‘give’ but a set of affixes marking tense/aspect and subject and object agreement on verbs appear in the verb slot without any verb stem. In this way, the concept GIVE is signalled without overt expression. Another interesting case has been documented in Margetts (2008), who describes Saliba, a Western Oceanic language spoken on an island at the eastern tip of Papua New Guinea. In this language, there are two ‘give’ verbs in complementary distribution. One is used when the recipient is first or second person and the other when it is third person. The two verbs are also constructed in different ways, one being transitive and the other transitive or ditransitive. Obviously, there are languages where GIVE is derived from GET or HAVE. According to Margetts & Austin (2007), ‘give’ is derived by applicativizing ot ‘get’ (or as an alternative atada ‘do for the benefit of’) as in (36a-b):

(36)

a.

Banda

n=ot

yan

Bakan

Banda

3sg=get

fish

be.big

Banda caught a big fish.

b.

Banda

n=ot-ik

yak

yan

Banda

3sg=get-appl

1sg

Fish

Banda gave me some fish.

85In this paper, there has been as certain focus on GET and I will give a few more examples of languages where this concept is lexicalized in an interesting way. Swahili in principle realizes GIVE, TAKE and GET as simple verbs. Predicative possession is expressed as a With-Possessive (see 37a-b). Historically, GET -pata is derived from -pa ‘give’ with the ‘contactive’ suffix –ta which is not fully productive any more. Actually, as demonstrated in (37c-e) GET can be expressed both with -pata and with the passive form of -pa ‘give’ (pa + wa > pewa-) (Abdulaziz Lodhi, personal communication).

(37)

a.

Hamisi

a-li-kuwa

na

ki-tabu

cl1-past-Copula

with

cl7-book

Hamisi had a book.

b.

Hamisi

a-na

ki-tabu

cl1-with

cl7-book

Hamisi has a book.

c.

Hamisi

a-li-pata

ki-tabu

cl1-past-get

cl7-book

Hamisi got a book

d.

Hamisi

a-li-ki-pata

ki-tabu

cl1-past-cl7.obj-get

cl7-book

Hamisi got the book.

e.

Hamisi

a-li-pe-wa

ki-tabu

cl1-past-give-passive

cl7-book

Hamisi got/was given a/the book.

f.

Ali

a-li-ki-chukua

ki-tabu

kutoka kwa

Juma

cl1-PAST-cl7.obj-take

cl7-book

from

Ali took the book from Juma.

86Judging from Johnson (1937/1971), the verb -pata ‘get’ has developed a wide range of meanings including grammaticalizations and is one of the most frequent verbs in the language, even more frequent than -pa ‘give’ (Bertoncini 1973). It can be used to refer to Abstract possession in a number of expressions such as pata nguvu ‘gain strength, become strong’; pata baridi ‘catch a cold’ and a more general sense of change (glossed ‘happen’, ‘become’, ‘occur’) as in pata kujua ‘get to know, find out’. In combination with an infinitive, pata is used to express a range of meanings related to possibility, permission and success as in examples (37g-h).

(37)

g.

A-li-pata

ku-fanya

kazi

Cl1-PAST-get

INF-do

work

He was able (or permitted) to work.

h.

Ni-me-pata

ku-fanya

1sg- PERF-get

inf-do

I have succeeded in doing it.

87Thus, Swahili provides one more example of the tendency of GET to develop a rich polysemy including modal possibility in an area outside Northern Europe and South(east) Asia.

88As is well known, in certain varieties of English, GET can be extended to cover HAVE (Peter (has) got a car).This has been established in Tok Pisin, where a separate verb kisim (<’catch them’) realizes the meaning GET as can be observed in (38a-c) taken from the questionnaire briefly described in Viberg (1986)

(38)

a.

Pita gat gan.

Peter has a gun.

b.

Pita bin kisim naip long Heri.

Peter got a knife from Harry.

c.

Heri bin givim Pita wanpela naip.

Harry gave Peter a knife.

(bin is a PAST marker and long a general LOCative marker. –im appears as marker on transitive verbs)

89An interesting source of GET is found in Setswana (the major Bantu language spoken in Bostswana), where bona ‘see’ is established as a realization of GET. (39) is taken from the same questionnaire as (38).

(39)

Pita

o

ne a  

bona

thipa

mo go

Heri.

cl1

T/A

see

knife

loc

Peter got a knife from Harry

(T/A = Tense/Aspect)

90The languages looked at so far have lacked one or more of the four verbs that are basic in Germanic languages. There is, however, one language that requires the addition of a further dimension, social deixis, to account for the basic verbs of possession. In Japanese, GIVE is expressed with a set of verbs that are differentiated with respect to the social status of the giver and the receiver. In particular, there are several verbs whose use depend on the speaker’s status relative to other people concerned. The system can be said to be based on social deixis. The following account is based on Kuno (1973), in which a special chapter is devoted to giving and receiving verbs. (40a) shows the typical argument structure of clauses with a verb of giving: Giver ga Receiver ni Gift o Verb.

(40)

a.

John

ga

Mary

ni

kono

hon

o

yatta

NOM

DAT

this

book

ACC

gave

John gave this book to Mary.

(Kuno 1973, 129)

91Yaru is used when the receiver is lower in status than the giver or a close friend, whereas another verb ageru must be used when the giver is lower or equal in status to the receiver. Another requirement that applies to both of these verbs is that the receiver is not identical to the speaker. There is another verb sasiageru that basically requires that the giver is identical to the speaker and simultaneously is inferior to the receiver as in (40b).

(40)

b.

Boku

ga

sensei

ni

kono

hon

o

sasiageta

I

NOM

teacher

DAT

this

book

ACC

gave

I gave the book to the teacher.

(Kuno 1973, 129)

92There are also two verbs that basically require that the speaker is identical to the receiver. Kureru is used when the giver is lower or equal in status to the speaker/receiver, whereas kudasaru is used when the giver is superior in status. The system is more complicated than indicated here. For example, the role of the speaker could be extended to refer to someone who is related to the speaker. (40b) would be acceptable also if the subject referred to the speaker’s brother. There are also two verbs of receiving that are differentiated with respect to social deixis: itadaku, which basically indicates that the speaker is the receiver and is inferior in status to the giver and morau which does not require that the receiver is identical to the speaker but indicates that the receiver is superior or equal to the giver. The verbs of giving and receiving can also be used as auxiliaries with various benefactive meanings as shown in (40c)

(40)

c.

sensei

ga

boku

ni

hon

o

yonde

kudasatta.

teacher

NOM

I

DAT

book

ACC

reading

gave (the favor of)

The teacher read the book to me (for me).

(Kuno 1973, 131)

93There is a study by Östlund (2006) who compares Swedish and the Japanese verb of receiving morau based on a Swedish novel and its translation into Japanese. The translations of all the occurrences of in the Swedish original were analyzed. The correspondences between and morau turned out to be extremely scarce and even in other respects the comparison showed that the translations of were “various and seemingly haphazard” (quoted from the Abstract). This indicates that the systems differ profoundly, at least at a more fine-grained level with respect to the use of individual verbs. At a more general level, there seems to be some interesting semantic parallels. When denoted permission, benefactive auxiliary constructions with verbs of giving appeared as translations in some cases.

94Even if there is to date no general typological study of the verbs of possession, it appears that the Germanic system of four basic possession verbs corresponding to GIVE, TAKE, GET and HAVE probably is very language-specific from a typological perspective, not only within Europe but even more so in a world-wide perspective. This forms a parallel to the motion verb typology. Levinson & Wilkins (2006, 527) say with respect to (Talmy 1985, 2000) that their sample suggests “that the Germanic satellite-framed pattern may be very restricted typologically.” (For similar typologically specific patterns at the grammatical level of Germanic and (to various degrees) other European languages, see Hammarberg & Viberg 1978, Haspelmath 1998, van der Auwera 1998).

6.2 Elaborations of the basic system

95The system of basic verbs of possession can be conceptually elaborated by adding more specific conceptual information of various types. Most of these involve Ownership rather than Physical possession. Elaborations of the basic concepts GIVE and TAKE include temporary possession (lend/borrow), economic obligations (buy/sell; pay) and combinations of these concepts (rent/let/hire/lease). In some languages, the semantic relatedness with the basic verbs is overt as in the basic correspondents of borrow and lend in Turkish: ödünç almak (loan take) ‘borrow’ and ödünç vermek (loan give) ‘lend’ or Polish sprzedawać ‘sell’ a derived form of dawać ‘give’. Many of these concepts are organized as opposites on the model of GIVE and TAKE (in the sense ‘dispossess’). A radical contrast to the system in European languages is described in Dixon (1973), see Table 14. Whereas European (and many other languages), have verbs of possession such as buy and sell, which crucially involve economic obligations, the Australian language Dyirbal (spoken in North Queensland) has verbs which incorporate concepts of various types of kinship obligations. (Kinship relations play a certain role also in Western languages in verbs such as inherit and bequeath.) Table 14, which is based on the analysis presented in Dixon (1973), shows a sketch of the conceptual elaborations of GIVE in English and Dyirbal. The set of subordinate verb meanings might be regarded as a third dimension in the semantic grid, which only shows the top nodes of the hierarchies.

Table 14. Conceptual elaboration of GIVE in English and Dyirbal (based on Dixon 1973).

Agrandir

7 Conclusion

96A typologically oriented characterization must be formulated at a rather general level to be meaningful, such as the systems of basic possession verbs (Table 11). One of the major advantages of corpus-based contrastive analysis is that it makes it possible to refine the analysis by showing that languages that are similar at a general level and belong to the same type can contrast in a striking way with respect to usage patterns. Some examples of that are given in the sections on Swedish ta, ge and . The focus on basic verbs makes it possible to draw attention to some of the major systematic contrasts, in particular with respect to patterns of polysemy.

97The lexical organization of the semantic field of possession verbs shares one characteristic with other major verbal semantic fields such as the verbs of motion or the verbs of perception. Even if the total number of verbs within the field may amount to several hundreds in languages with open verb classes, there are a small number of basic verbs that dominate in terms of frequency and centrality. Some of these, the nuclear verbs, GIVE and perhaps TAKE in the case of the verbs of possession, tend to be basic in most languages, whereas others such as GET and HAVE represent more or less language-specific features.

98Possession is a rather abstract concept. Across languages, the basic verbs of possession have a strong tendency to be derived from physical action verbs describing acts carried out with the hands and as we have seen traces of this are left even in some of the verbs like Swedish ta that overall take on many abstract meanings. The ability to carry out complex hand actions is a general human characteristic biologically constrained by both the brain and the anatomy of the hand. At a more detailed level, the latter is reflected, for example, in the existence of verbs such as Swedish plocka ‘pick’ (e.g. plocka blommor ‘pick flowers’), a troponym of ta ‘take’. The meaning of this verb refers to one of the characteristic grips that can be achieved with the human hand, the precision grip, where the thumb is placed in contact with the finger tips as when holding a pen (Trew & Everett 2005). Acts like handing things also appear to be a form of social cooperation that is acquired early both in enculturation of children and in the evolution of human culture. Superimposed on this, we find culturally variable concepts of ownership, economic obligations and obligations based on kinship relations. Such concepts are further elaborated in special registers such as legal language. Many metaphors related to the concept of possession also appear to be culture-specific, such as the classical example of concepts Americans live by, like TIME IS MONEY (Lakoff & Johnson 1980). The role of complex and culture-specific elaborations of the most basic concepts of possession has only been briefly indicated in this paper, but even in modern Western societies where there exists an extensive legal and economic terminology referring to possession, a few basic verbs of possession predominate in ordinary discourse in terms of frequency and availability and these verbs tend to be historically derived from verbs referring to actions with the hand.

Haut de page

Bibliographie

Altenberg, B. & K. Aijmer. 2000. The English-Swedish Parallel Corpus: A resource for contrastive research and translation studies. In : Mair, C. & M. Hundt (eds.) Corpus Linguistics and Linguistic Theory. Amsterdam – Atlanta/GA: Rodopi, 15-33.

Altenberg, B. & S. Granger (eds.). 2002. Lexis in contrast. Amsterdam: Benjamins.

Barsalou, L. 1999. Perceptual symbol systems. Behavioral and Brain Sciences 22: 577-660.

--- Situated simulation in the human conceptual system. Language and Cognitive Processes 18(5/6): 513-562.

Bendix, E.H. 1966. Componential analysis of general vocabulary: the semantic structure of a set of verbs in English, Hindi, and Japanese. The Hague: Mouton.

Berlin, B. & P. Kay. 1969.Basic color terms. Berkeley: University of California Press.

Bertoncini, E. 1973. A tentative frequency list of Swahili words. In : Annali Vol. 33, Fascicolo 3. Instituto Orientale di Napoli, 297-363.

Buck-Darling, C. 1949. A dictionary of selected synonyms in the principal Indo-European languages. Chicago & London: The University of Chicago Press.

Burenhult, N. (ms.) On the polysemy of the Swedish verb for take.

Croft, W. & D.A. Cruse. 2004. Cognitive linguistics. Cambridge: Cambridge University Press.

Cysouw, M. & B. Wälchli (eds.). 2007. Parallel texts. Sprachtypologie und Universalienforschung  60(2).

Damasio, A R. & D. Tranel. 1993. Nouns and verbs are retrieved with differently distributed neural systems. Proceedings of The National Academy of Sciences 90: 4757-4760.

Dixon, R.M.W. 1973. The semantics of giving. In : Gross, M., M. Halle & M-P. Schützenberger (eds.) The formal analysis of natural languages. The Hague: Mouton.

Ekberg, L. 1993a. Verbet ta i metaforisk och grammatikaliserad användning (“The verb ta ’take’ in metaphorical and grammaticalized use”. In Swedish.) Språk & Stil 3: 105-139.

--- 1993b. The cognitive basis of the meaning and function of cross-linguistic take and V. Belgian Journal of Linguistics 8: 21-42.

Enfield, N.J. 2003. Linguistic epidemiology. Semantics and grammar of language contact in mainland Southeast Asia. London & New York: Routledge Curzon.

Evans, N. & D. Wilkins. 2000. In the mind’s ear: The semantic extension of perception verbs in Australian languages. Language 76:546-592.

Glenberg, A.M., D. Havas, R. Becker, & M. Rinck. 2005. Grounding language in bodily states. In : Pecher, D. & R.A. Zwaan (eds.) Grounding cognition. The role of perception and action in memory, language and thinking. Cambridge: Cambridge University Press, 115-128.

Glenberg, A.M. & M.P. Kaschak. 2002. Grounding language in action. Psychonomic Bulletin & Review 9: 558-565.

Goddard, C. 2001. Lexico-semantic universals: A critical over-view. Linguistic Typology 5: 1-65.

Goddard, C. & A. Wierzbicka (eds.) 2002. Meaning and universal grammar.  Theory and empirical findings. Amsterdam: Benjamins.

Gómez González, M., J. L. Mackenzie & E. González Álvarez (eds.). 2008. Current trends in contrastive linguistics. Amsterdam: Benjamins.

Gonzalez-Marquez, M., S. Coulson & Mittelberg, I. (eds.). 2007. Methods in Cognitive Linguistics. Amsterdam: Benjamins.

Göz, I. 2003. Yazιlι türkçenin kelime sιklιğι sözlüğü. Ankara.

Gries, S. Th. & A. Stefanowitsch (eds.). 2006. Corpora in cognitive linguistics. Berlin & New York: Mouton de Gruyter.

Gronemeyer, C. 1999. On deriving complex polysemy: the grammaticalization of get. English Language and Linguistics3: 1-39.

Hammarberg, B. & Å. Viberg. 1977. The place-holder constraint, language typology, and the teaching of Swedish to immigrants. Studia Linguistica 31(2): 106–63.

Hardin, C.L. & L. Maffi (eds.). 1997. Color categories in thought and language. Cambridge: Cambridge University Press.

Haspelmath, M. 1998. How young is Standard Average European? Language Sciences 20(3): 271-87.

--- 2005. Ditransitive constructions: the verb ‘give’. In : Haspelmath et al (eds.).

Haspelmath, M., M. Dryer, D. Gil & B. Comrie (eds.). 2005. The World Atlas of Language Structures(WALS). Oxford: Oxford University Press.

Heine, B. 1997. Possession. Cognitive sources, forces, and grammaticalization. Cambridge: Cambridge Univeristy Press.

Heine, B. & T. Kuteva. 2002. World lexicon of grammaticalization. Cambridge: Cambridge University Press.

Isačenko, A.V. 1974.  On ‘have’ and ‘be’ languages. In : Flier, M. (ed.) Slavic forum: Essays in linguistics and literature. The Hague: Mouton, 43–77.

Janda, L.A. 1997. GIVE, HAVE, and TAKE in Slavic. In : Newman (ed.).

Johansson, S. 2007. Seeing through multilingual corpora. Amsterdam: Benjamins.

Johnson, F. 1971 (first published: 1937). A Standard Swahili-English Dictionary. Oxford: Oxford University Press.

Kilgariff, A. 1998. BNC lists. http://www.itri.brighton.ac.uk/peopleindex.html. Accessed March 2002.

Koptjevskaja-Tamm, M. 2008. Approaching lexical typology. In : Vanhove (ed.), 3-52.

Kuno, S. 1973. The structure of the Japanese language. Cambridge/Mass.: MIT Press.

Lakoff, G. & M. Johnson. 1980. Metaphors We Live By. Chicago: University of Chicago Press.

Levinson, S. & D. Wilkins (eds.). 2006. Grammars of space. Explorations in cognitive diversity. Cambridge: Cambridge University Press.

Lord, C. 1993. Historical change in serial verb constructions. Amsterdam: Benjamins.

Margetts, A. 2008. Learning verbs without boots and straps? The problem of ‘give’ in Saliba. In : Bowerman, M. & P. Brown (eds.) Crosslinguistic perspectives on argument structure. New York & London: Lawrence Erlbaum Associates, 111-137.

Margetts, A. & P.K. Austin. 2007. Three-participant events in the languages of the world: towards a crosslinguistic typology. Linguistics 45(3): 393-451.

Miller, G.A & P. Johnson-Laird 1976. Language and perception. Cambridge: Cambridge University Press.

Newman, J. 1996. Give. A cognitive linguistic study. Berlin: Mouton de Gruyter.

Newman, J.(ed.) 1997. The linguistics of giving. Amsterdam: Benjamins.

--- (ed.) 2002. The linguistics of sitting, standing, and lying. Amsterdam: Benjamins.

--- (ed.) 2009. The linguistics of eating and drinking. Amsterdam: Benjamins.

Nordlund, M. 2009. From physical to mental acquisition : a corpus-based study of verbs. Saarbrücken: VDM Verlag Dr. Müller.

Norvig, P. & G. Lakoff. 1987. Taking: a study in lexical network theory. Berkeley Linguistic Society 13: 195-206.

Östlund, J. 2006. Verbs of receiving. A comparative study of Swedish and Japanese morau in the written language. Bachelor Thesis. Department of Linguistics, Stockholm University.

Pulvermüller, F. 2001. Brain reflections of words and their meanings. Trends in Cognitive Sciences 5: 517-24.

Ramnäs, M. 2006. Étude contrastive du verbe suédois dans un corpus parallèle suédois-français. PhD dissertation. Gothenburg University.

Rawoens, G. 2008. Kausativa verbkonstruktioner i svenskan och nederländskan. En korpusbaserad syntaktisk-semantisk undersökning. (”Causative verbal constructions in Swedish and Dutch. A corpus-based syntactic-semantic study.” In Swedish.) Göteborgsstudier i Nordisk Språkvetenskap 11.

Rice, S. 1997. Giving and taking in Chipewyan: the semantics of THING-marking classificatory verbs. In : Newman (ed.), 97-134.

--- 2009. Athapaskan eating and drinking verbs and constructions. In Newman (ed.), 109-152.

Roberts, J.R. 1997. GIVE in Amele. In Newman (ed.), 35-65.

Slobin, D.I. 2003. Language and thought online: cognitive consequences of linguistic relativity. In : Gentner, D. & S. Goldin-Meadow (eds.) Language in mind: advances in the study of language and thought. Cambridge, Mass.: MIT Press.

--- 2004. The many ways to search for a frog: linguistic typology and the expression of motion events. In : Strömqvist, S. & L. Verhoeven (eds.) Relating events in narrative: typological and contextual perspectives. Mahwah, NJ: Lawrence Erlbaum Associates, 219-257.

Stassen, L. 2009. Predicative possession. Oxford: Oxford University Press.

Stefanowitsch, A. & S. Th. Gries (eds.). 2006. Corpus-based approaches to metaphor and metonymy. Berlin & New York: Mouton de Gruyter.

SUC. 1997. SUC 1.0. Stockholm Umeå Corpus. Produced by Dept. of Linguistics, Umeå University and Dept. of Linguistics, Stockholm University. CD-Rom.

Talmy, L. 1985. Lexicalization patterns: semantic structure in lexical forms. In : Shopen, T. (ed.) Language typology and syntactic description. III. Grammatical categories and the lexicon. Cambridge: Cambridge University Press, 57-149.

--- 2000. Toward a cognitive semantics. Vol. 1-2. Cambridge/Mass.: The MIT Press.

Thornell, C. 1997. The Sango language and its lexicon. Travaux de l’insitut de linguistique de Lund 32.

--- 2002. The uses of the Sango verb ‘transfer’. Afrika und Übersee 85: 267-299.

Trew, M. & T. Everett (eds.). 2005. Human movement. An introductory text. London: Elsevier.

Tummers, J., C. Heylen & D. Geeraerts. 2005. Usage-based approaches in Cognitive Linguistics: A technical state of the art. Corpus Linguistics and Linguistic Theory 1(2): 225-261.

Van der Auwera, J. 1998. Conclusion. In : Van der Auwera, J. (ed.) Adverbial constructions in the languages of Europe. Berlin: Mouton de Gruyter.

Van der Auwera, J. & V.A. Plungian. 1998. Modality's semantic map. Linguistic Typology 2: 79-124.

Van der Auwera, J., P. Kehayov & A. Vittrant. 2009. Acquisitive modals. In : Hogeweg, L., H. de Hoop & A. Malchukov (eds.) Cross-linguistic semantics of tense, aspect, and modality. Amsterdam, Benjamins, 271-302.

Vanhove, M. (ed.) 2008. From Polysemy to Semantic Change. Towards a typology of lexical semantic associations. Amsterdam: Benjamins.

Viberg, Å. 1981. Studier i kontrastiv lexikologi (Studies in contrastive lexicology. In Swedish). Ph.D. dissertation. SSM Report 7-8. Dept. of linguistics, Stockholm University.

--- 1984. The verbs of perception: a typological study. Linguistics 21: 123-162.

--- 1986. The study of crosslinguistic lexicology. In : Ö. Dahl (ed.) Papers from the Ninth Scandinavian Conference of Linguistics. Dept. of Linguistics, Stockholm University, 313-327

--- 1993. Crosslinguistic perspectives on lexical organization and lexical progression. In : Hyltenstam, K. & Å. Viberg (eds.) Progression and regression in language. Cambridge: Cambridge University Press, 340-385.

--- 1996. Crosslinguistic lexicology. The case of English go and Swedish . In : Aijmer K., B. Altenberg & M. Johansson (eds.) Languages in contrast. Papers from a Symposium onText-based Cross-linguistic Studies [Lund Studies in English 88]. Lund: Lund University Press, 151-182.

--- 1999. Polysemy and differentiation in the lexicon. Verbs of physical contact in Swedish. In : Allwood, J. & P. Gärdenfors (eds.) Cognitive semantics. Meaning and cognition. Amsterdam: Benjamins, 87-129.

--- 2001. The verbs of perception. In : Haspelmath, M., E. König, W. Oesterreicher & W. Raible (eds.) Language Typology and Language Universals. An International Handbook. Berlin: De Gruyter, 1294-1309.

--- 2002a. Polysemy and disambiguation cues across languages. The case of Swedish and English get. In : Altenberg, B. & S. Granger (eds.), 119-150.

--- 2002b. The polysemy of Swedish ge ‘give’ from a crosslinguistic perspective. Proceedings of Euralex 2002, Copenhagen University, 669-682.

--- 2004. Physical contact verbs in English and Swedish from the perspective of crosslinguistic lexicology. In : Aijmer, K. & B. Altenberg (eds.) Advances in corpus linguistics. Amsterdam/New York: Rodopi, 327-352.

---  2005. The lexical typological profile of Swedish mental verbs. Languages in Contrast 5(1): 121-157.

--- 2006a. Towards a lexical profile of the Swedish verb lexicon. In : Viberg, Å. (guest ed.) The typological profile of Swedish. Thematic  issue of Sparachtypologie und Universalienforschung 59(1): 103-129.

--- 2006b. Crosslinguistic lexicology and the lexical profile of Swedish. In : Bardel, C. & J. Nystedt (eds.) Progetto Dizionario Italiano Svedese. Atti del Primo colloquio. Acta Universitatis Stockholmiensis. Romanica Stockholmiensia 22: 79-118.

--- 2008. Swedish verbs of perception from a typological and contrastive perspective. In : de los Ángeles Gómez González, M., L. Mackenzie & E. González Álvarez (eds.) Languages and Cultures in Contrast: New Directions in Contrastive Linguistics. Amsterdam: Benjamins.

--- 2009. The meaning patterns of Swedish låta ‘sound’; ‘let’. A corpus-based contrastive study. In : Bernardini, P., V. Egerland & J. Granfeldt (eds.) Mélanges plurilingues offerts à Suzanne Schlyter à l'occasion de son 65ème anniversaire. [Études romanes de Lund 85], 455-480

--- (forthc.) Cognitive linguistics and corpus-based contrastive analysis: The case of the Swedish verb ta ‘take’ (15 pages). To appear in the proceedings from: New Directions in Cognitive Linguistics, 2nd Conference of the UK-Cognitive Linguistics Assoc. Cardiff University August 27-30, 2007

Haut de page

Annexe

Electronic sources:

The English Swedish Parallel Corpus (ESPC). For a description, see: http://www.englund.lu.se/content/view/66/127/

FrameNet: http://framenet.icsi.berkeley.edu/.

Global WordNet and EuroWordNet: http://www.globalwordnet.org/

OED. Oxford English Dictionary. Electronic version:http://www.oed.com/

SIMPLE: http://www.ub.es/gilcub/SIMPLE/simple.html

Swedish WordNet: http://www.lingfil.uu.se/ling/swn.html

WordNet: http://wordnet.princeton.edu

Haut de page

Notes

1  One of the anonymous reviewers of this article commented “The semantic analysis is corpus-illustrated rather than corpus-based (to use a distinction introduced by Tummers et al 2005). The author uses his/her intuition about various aspects of each verb's meaning and then illustrates these aspects using corpus examples.” Since I think that this way of reasoning is representative of people who favour the introduction of multivariate statistics and similar rigorous methods, I take the opportunity to answer. The statistical analysis is primarily descriptive and does not live up to the standards set by a work such as Tummers et al (2005), but it nevertheless accounts for all occurrences of the verbs in the corpus and in that sense is systematic and corpus-based. Besides, more detailed information is provided in the earlier studies which are referred to. In several cases, the differences in usage patterns between languages are so great that a simple frequency count is enough to prove the point. A more sophisticated distributional analysis based on larger corpora would in many cases make it possible to go deeper, but the crosslinguistic comparison has its own merits. (It’s up to the readers of this article to decide whether they agree.) The semantic analysis is based on a close reading of all individual examples in the corpus. Although I regard greater statistical sophistication as something to strive for, we should remember that the major limitation of corpora is that they only represent linguistic forms. (This is an even greater limitation than the fact that they are off-line, as Tummers et al 2005 correctly point out.) When small children start acquiring language, many conceptual primitives are already well-established and adult language users rely on a vast conceptual system, which is only partly linguistically based, in order to make sense of utterances. For that reason we must rely on our intuition to a rather large extent when we prepare semantic analyses as long as background knowledge and situations of use are not explicitly represented. (Greater rigour in this area is, of course, also something to strive for.) As indicated in sect. 3, the semantic analysis is inspired in a general way by Barsalou’s (1999, 2003) situated simulation theory according to which conceptual representations are multi-modal simulations that are distributed across modality-specific systems and prepare an agent for situated action.

2  sich begeben (not included in All meanings)

3  abandonner (not included in All meanings)

Haut de page

Pour citer cet article

Référence électronique

Åke Viberg, « Basic Verbs of Possession », CogniTextes [En ligne], Volume 4 | 2010, mis en ligne le 17 mars 2010, Consulté le 27 mars 2017. URL : http://cognitextes.revues.org/308

Haut de page

Auteur

Åke Viberg

Department of Linguistics and Philology, Uppsala University, Box 635, SE-75126 Uppsala, Sweden

Haut de page