Размещено на http://www.allbest.ru/

• Table of Contents
• 1. Introduction
• 1.1 Theory block
• 1.2 Multifunctionality of the instrumental grams
• 1.3 Semantic maps
• 1.4 The aim of the paper
• 2. The complexity of instrumental meanings
• 3. Methodology
• 4. The competition of ways to express the instrumental meaning
• 4.1 Adpositions competition
• 4.1.1 Hebrew discussion
• 4.1.2 Arabic discussion
• 4.1.3 Berber discussion
• 4.1.4 Thai discussion
• 4.2 Cases competition
• 4.2.1 Permyak languages discussion
• 4.2.2 The other Uralic languages
• 4.3 The competition of cases and adpositions
• 4.3.1 Russian discussion
• 4.3.2 Indic language family discussion
• 4.3.3 Buryat discussion
• 4.4 The competition of applicative and adpositions
• 4.4.1 Ainu discussion
• 4.4.2 Abaza discussion
• 4.5 The competition of applicatives and cases
• 4.5.1 Adyghe discussion
• 4.5.2 Dyirbal discussion
• 5. Grammaticalization patterns
• 5.1 Instrumental and comitative
• 5.2 Instrumental and locative
• 6. Conclusion
• 7. References



1. Introduction

1.1 Theory block

The notion of an instrument as a semantic role (SR) was introduced by Charles J. Fillmore in late 60s and further improved in (Fillmore 1981a), but up to this day there are some problems of how exactly to specify the instrumental meaning.

The prototypical instrumental meaning covers a participant of the situation used by the agent and affecting the patient, which is the case in (1.1). But even this instrumental meaning can be further divided into 2 categories by most authors and into 4 categories according to others (Muravenko 1994).

Rwanda

(1.1)Umugabo a-ra-andik-iish-aibaruwa ikaramu

Manhe-pres-write-instr-aspletterpen

`The man is writing a letter with the pen'

The 2-categories classification includes an instrument that cannot end during the usage (e.g. to write with a pen) and an instrument that is somehow limited (e.g. to draw with paint - paint will run out eventually).

The 4-categories classification includes the following 4 types:

· Instrument (e.g. to write with a pen, to cut an apple with a knife);

· Bound instrument (e.g. to fix the hair with a hairpin, to mend the roof with wood);

· Free means (e.g. to clean teeth with a toothpaste);

· Ingredient means (e.g. to draw with gouache).

The mentioned classifications are considered to be sufficient enough to serve the purposes of the description of argument structure and semantic role of instrument itself. The further we refine and develop these classifications the subtler would be the differences between the meaning types. However, some languages mark the instrumental object differently depending on its features - I go more into details on the instrumental meanings in chapters 2 and 4.

So, the first sub-aim of this paper is to examine a large sample of languages and find out what means are used in the language to express the instrumental meaning.

Different languages have different inventories of grammatical features, and, therefore, use different means to express the instrumental meaning. The most common ways to express the instrumental meaning are via adpositions, cases, applicatives and lexical-grammatical affixes (1.2-1.6).

Hebrew

(1.2)hu xatax tapu'ax [im sakin | be-sakin] he cut.pst.3sg.m apple [com knife | loc knife]

`He cut the apple with a knife' (Postposition usage)

Dyirbal

(1.3)balandugumbilbangulyarangubangu

there-nom-IIwoman-nomthere-erg-Iman-ergthere-inst-IV

yugungubalgan

stick-insthit-pres/past

`The man hits the woman with a stick' [R.M.W. Dixon 1972](Instrumental case usage)

Abaza

(1.4)a- и^jk'^w?na-mhaи'^wa y-ax?^jad-a-la-y?-c-и^ja-t'

the-boythe-spoonsg.m-sistera3sg.h-p3sg.n-instr-p3sg.m-com-eat-dyn

`His sister ate with boy with the spoon'(Applicative usage)

Adyghe

(1.5)se wate-m-??e ???ит??ne?-r?-xe-s-??aлтa-?

Ihammer-obl-insnailsg.io-instr-loc-1sg.erg-to hammer-pst

`I hammered the nail'(Applicative and case usage)

Comanche

(1.6)tihiya-seu-siH-kwii-i

horse-cntrhim-sih(with feet)-shoot-cmpl:asp

`A horse kicked him' [Jean Ormsbee Charney 1997](lexical-grammatical affix usage)

Out of these ways to express the instrumental meanings the most complicated one is the applicative, so, I believe that it requires additional explanation in order to understand the competition among the grammatical features. instrumental discussion grammatical pattern

Applicative is a raising actant derivation, introducing a new participant during the situation which is described by the verb. Morphologically, applicative is usually represented by a special marker on the verb and introduces a direct or indirect object. Semantically, applicative constructions introduce participants, which semantic roles usually are benefactor, malefactor, locative, comitative, instrument and some others, but in this work I focus on the instrumental applicatives.

The prototypical applicative construction should have the following features:

· The applicative construction increases the verb valency by one point;

· The applicative construction presents the alternative way to express the syntactic and semantic peripheral object, so there must be another way to express the same meaning;

· The applicative construction cannot introduce an obligatory participant (e.g. agent or patient/theme).

For purposes of this work, the most important feature of an applicative construction is that if the language can express the instrumental meaning via an applicative construction, then it must also be able to express the instrumental meaning via some other grammatical features.

This type of rivalry creates the competition among several grammatical features to express one meaning. So, for example, in the Russian language the instrumental meaning can be expressed by both a prepositional phrase or by the instrumental case (cf. `Иван резал хлеб ножом/ Иван резал хлеб при помощи ножа' - Ivan cut bread with a knife).

So, the second sub-aim of this paper is to investigate different types of competitive forms expressing the instrumental meaning and find out what factors usually imply the usage of one grammatical feature over another (e.g. why native speakers in one context would use the case marker and not the prepositional phrase).

Coming back to the instrumental meaning, the prototypical instrumental meaning is not the only one that is relevant to this work. In most cases, the one grammatical feature that expresses the prototypical instrumental meaning also might express a number of peripheral instrumental meanings, which might include the locative or temporal functions. For example, in the Adyghe language there are around 20 meanings that are expressed by the instrumental case, and 14 of them might be also expressed by the instrumental applicative (more details in 4.5.1). In Hebrew, these 20 are divided between the comitative and locative postpositions. In Permyak languages, there are 19 instrumental functions. And on Fig. 1 we can see the instrumental meanings in language of the world according to (Narrog, Ito 2007). On Fig.2, we can see the grammaticalization patterns of the instrumental domain.

Fig.1: Polysemy of the instrumental markers



Fig.2: Revised diachronic map of the instrumental/comitative domain (Narrog 2010)

So, the third sub-aim of this paper is to investigate the diversity of the meanings that might be expressed by the grammatical features that also represent the prototypical instrumental meaning. By looking at the prototypical and peripheral instrumental meanings and the grammatical features that are used to express them, I hope to make generalized statements on what factors influence the choice of one grammatical representation over others.

1.2 Multifunctionality of the instrumental grams

In this section, by the instrumental grams I mean both function words expressing the instrumental meaning, such as adpositions and affixal categories, such as cases and applicatives that also express the instrumental meanings. I am using the terms grammatical morphemes and grams for short, following (Haspelmath 2000), and later in the paper I will use the term instrumental grams to refer to all the non-lexical ways to express the instrumental meaning.

So, in this section, I talk about different approaches to multifunctionality of grammatical morphemes, such as the monosemist, polysemist and homonymist positions, and also talk about semantic maps and why these terms must be discussed for a cross-linguistic research. So, the example of the multifunctional grammatical morpheme can be seen in (1.7, 1.8).

Hebrew

(1.7)hu xatax tapu'ax be-sakin he cut.pst.3sg.m apple loc-knife

`He cut the apple with a knife'

(1.8)hu nasa be-otobus he travel.pst.3sg.m loc-bus

`He went by bus'

In the examples above, the locative preposition be- in (1.7) implies a meaning different from the meaning in (1.8) and goes from the prototypical instrumental meaning (described in (1.1)) to the means of transport meaning. So, in Hebrew the locative preposition can be used for both of these meanings, but in English, for instance, these two meanings would require different prepositions. From that we can conclude that the locative preposition in Hebrew is a multifunctional grammatical morpheme with its functions being language-specific.

Different approaches analyze this type of multifunctionality in different ways. The monosemist approach states that grammatical morphemes bear only a very abstract meaning, and we receive the elaborated meanings because of the gram's interaction with context. So, according to this position, we should perceive the locative preposition in (1.7) and (1.8) as something like `abstract instrument' and the difference in the derived meanings would appear because of the different properties of the instrumental object, knife and bus. However, this type of analysis does not quite fit the Hebrew language, as the main usage of the locative preposition is the locative meaning, so we would need even a vaguer abstract meaning for this analysis to work.

The other approach is the polysemist position, which states that there are different senses and meanings to each grammatical morpheme, but all those meanings are related in some way and have restrictions. So, I believe that according to the polysemist position, we should analyze the locative preposition in (1.7) and (1.8) differently from the locative preposition in its most prototypical sense.

The third approach is the homonymist position, and it basically states that each grammatical morpheme should have its unique meaning.

However, the topic of multifunctionality and these three positions is quite controversial, and it is very difficult to take into account the vast number of different grammatical morphemes and the meanings that they represent, so none of these approaches seem to be able to explain all the complexity. Therefore, Martin Haspelmath suggests using semantic maps as a method of explaining the different meanings of grammatical morphemes, especially in the cross-linguistic setting.

I would also like to say, that even though that I mostly use Martin Haspelmath's works on the semantic maps, the notion of the semantic maps itself origins to (Anderson 1982) and (Anderson 1986) and supplements heavily by (Haspelmath 1997), (van der Auwera & Plungian 1998), (Croft 2001), (Haspelmath 2003) and (Malchukov 2004).

1.3 Semantic maps

`A semantic map is a geometrical representation of functions in “conceptual/semantic space” which are linked by connecting lines and thus constitute a network' (Haspelmath 2000).



Fig. 3: A semantic map of typical dative functions/ The boundaries of English to

The semantic map above shows the range of functions of a typical dative case, and we can see how the English to covers four functions: direction, recipient, experiencer and purpose. The box with the curved lines indicates the boundaries of the grammatical morpheme's multifunctionality. As examples for this semantic map, Martin Haspelmath provides the following four examples from the English language:

a. Goethe went to Leipzig as a student (direction)

b. Eve gave the apple to Adam(recipient)

c. This seems outrageous to me(experiencer)

d. I left the party early to get home in time (purpose)

So, according to the semantic map we can see that the English to lacks a few functions that are more typical to the dative case cross-linguistically. For example, in English we cannot use to in a beneficiary function (cf. `I can buy this car for you/ *I can buy this car to you') while at the same time, the dative case in Russian can be used to express the beneficiary function (e.g. `Я могу купить тебе эту машину' (I can buy this car for you)).

Semantic maps have quite a few advantages over other classical approaches in typology, such as the list method and the general-meaning method. The list method basically introduces all the possible functions of a particular grammatical morpheme in a form of a list and labels them. It is a good and classic approach to get a quick review of functions of a gram, but it implies no theoretical claims, it lacks illustrativity and it is difficult to use in terms of cross-linguistic comparison. The general-meaning method identifies grammatical morphemes as abstract meanings opposed to other grammatical morphemes that the former are in contrast with. Basically, with such a method it is quite difficult to figure out what exactly the functions of the grammatical morpheme are, because they are not listed, but rather united in some abstract meaning.

On the other hand, semantic maps are much better for cross-linguistic comparability; they are more illustrative, and, therefore, it is much easier for the researcher to keep track when working with a large sample of languages. They also might help to discover some semantic universals considering particular grammatical morphemes, and they might create a number of implicational universals at the same time. Also, while this is not particularly significant for this work, semantic maps might bring out the expectations about diachronic change.

The main reason for using semantic maps in this work is their compatibility for the cross-linguistic research, as I am going to analyze a decent number of languages in this paper. With the help of semantic maps, I also hope to make some generalizations considering the instrumental meanings in a typological perspective.

I think that it is also important to notice, that the idea of semantic universals is also controversial in some sense. Discussion in (Bybee 1988) suggests that grammatical meanings are derived from system-internal contrasts, which implies that the languages would develop very differently from each other, and the universals should be almost impossible. However, the typological work proves that there are similar distinctions of grammatical morphemes and their meanings, independent of the genetic and geographical states of the languages. So, the fact that languages make the same distinctions for grammatical morphemes proves that there is some sort of a limited range of functions, and that is why the `semantic maps' method is a very good tool for cross-linguistic comparison.

However, semantic maps are a tool for illustration and representation of the grammatical meanings in languages - they do not show the cognitive relationship between several meanings. Basically, we cannot state why the same grammatical morpheme is used for two different grammatical meanings according only to the semantic maps; we need an extra cognitive-semantic analysis to be able to state any theories.

Coming back to the instrumental meanings, Haspelmath also provided a few instrumental semantic maps (fig.4-fig.7) in (Haspelmath 2000).

Fig.4: Instrumental and related functions

Fig.5: The boundaries of English, German, and French prepositions



Fig.6: The boundaries of Seychelles Creole ek

Fig.7: The boundaries of the Russian instrumental and s `with'

On these semantic maps we can see the functions that are covered by the certain prepositions and cases. For example, on fig.4 we can notice that the English preposition with covers only three functions, that are co-agent, comitative and instrumental, while the Seychelles Creole preposition ek has a much wider range of functions and covers all the functions in the semantic map, except for beneficiary.

For purposes of this work, I have also decided to use semantic maps to illustrate the range of instrumental meanings covered by different instrumental grams in different languages. I took this type of maps (cf. fig.4) and expanded it by adding the peripheral instrumental meanings. I also created another semantic map considering only the type of the instrumental object; so, basically, such a map takes into account only the prototypical instrumental meaning but considers different types of the instrumental objects, such as weapons or body parts. I introduce both of these semantic maps in the chapter 2, and use them actively in the chapter 4.

One last remark I posit in this chapter is that I hope to make the semantic maps as illustrative and full as possible, but for most languages I collected the data through various grammars, and this source of information often lacks subtle semantic facts, so the semantic maps might miss something that actually exists in the languages but is not marked in the grammars.

1.4 The aim of the paper

Concluding the chapter, I would like to review the sub-aims I posited in 1.1 and the usage of semantic maps in 1.3 and posit the main aim of the paper.

So, the aim of the paper is to find out what instrumental functions are expressed by the instrumental grams in different languages; to create illustrative semantic maps of instrumental functions and instrumental objects for a cross-linguistic comparison; to find out what factors influence the choice of instrumental grams in a typological perspective.



2. The complexity of instrumental meanings

In the first chapter, I have already posited some insights onto the instrumental meanings and the complexity surrounding them. In this chapter, I will go into more details considering both the prototypical and the peripheral instrumental meanings.

Firstly, I introduce the two semantic maps, mentioned in the chapter 1.3, namely, one general instrumental semantic map with various functions of the instrumental grams in various languages, and one instrumental object semantic map, that includes only the prototypical instrumental meaning with different instrumental objects. The second map is also necessary for this work, as in many languages the choice of instrumental grams only differs by the features of the instrumental object.

Fig.8: The general semantic map of instrumental functions



Fig.9: The instrumental object semantic map

The instrumental main and peripheral functions on fig.8 are taken from the following sources (Hasplemath 2000), (Narrog & Ito 2007), (Serdobolskaya 2010) and (Arsentyev 2016). The various types of instrumental objects on fig.9 are taken from (Beloshankova & Muravenko 1985) and (Saint-Dizier et al. 2006). Overall, on the semantic maps there are more than twenty instrumental functions and eight types of the instrumental object in the prototypical instrumental meaning, which shows the complexity of the instrument notion in general. Below, I go through all the meanings listed in the semantic maps.

The instrumental functions are following:

Instrumental - denotes a participant used by the agent and affecting the patient, prototypically taking form of a physical inanimate object.

Manipuri

(2.1) m?hak-n? thaN-n? u k?ki

he-nom knife-ins tree cut

`He cut the tree with a knife' [Narrog 2009]

Comitative - denotes a participant in the accompaniment relationship with another participant, both are typically animate.

English

(2.2)The pantomime gave a show with a clown [Schlesinger 1979]

Locative - indicates a location.

Hebrew

(2.3)Ha'xayal met ba'emek the-soldierdie.3pst.sgloc-valley `The soldier died in the valley'

Passive agent - denotes a theme or a patient of the main verb in the clause with passive voice.

German

(2.4)Ich wurde von Hunden gebissen.

`I have been bitten by dogs.' [Haspelmath 2000]

Source - denotes a starting point of an action or motion.

Seychelles Creole

(2.5)Mon ganny pansyon ek gouvernman.

`I get a pension from the government.' [Michaelis & Rosalie 2000]

Conjunctive - denotes a connection of two clauses.

Seychelles Creole

(2.6)dan zil Kosmoledo ek Asonpsyon

`on the islands of Cosmoledo and Assomption' [Michaelis & Rosalie 2000]

Co-agent - denotes a second agent in the clause, taking part in the same action, as the first agent.

Russian

(2.7)Petya podralsya s Vasei `Petya fought with Vasya'

Recipient - denotes a participant receiving some object from the agent.

Swahili

(2.8)Juma a-li-m-nunul-i-a m-toto ki-tabu Juma 1sa-pst-10a-bring-app-fv I-child -book

`Juma brought the child a book' [Deo Ngonyani 1995]

Beneficiary - denotes a participant, who the action is completed for.

Tukang besi

(2.9)No-wila-ako te ina-no i daoa.

3r-go-applcoremother-3possobl market

'She went to the market for her mother.' [Donohue 1999]

Cause - denotes a participant, which is also the reason for the action described by the verb.

Seychelles Creole

(2.10)Pa kapab reste laba ek moustik.

`It was impossible to stay there because of the mosquitoes.' [Michaelis & Rosalie 2000]

Peripheral functions:

Means - the type of an instrument that is limited and might run out (Apresyan 1974).

Hebrew

(2.11)ha'yeledkotev im dio | be'dio the-boywrite-pres.3sgcom-ink/loc-ink `The boy writes with ink'

Transport - the instrument acts as a means of transport.

Adyghe

(2.12)awtwebw?s-?m-??e?-r?-kт?a-? bus-obl-inssg.io-instr-go-pst `He went by bus'

Prolative - denotes the locative meanings, expressing the motion through or across the landmark.

Adyghe

(2.13) и'?le-m ??eg??-m-??e q?-?-r?-kт?a-?

village-oblroad-obl-insdir-3sg.io-instr-go-pst

`I came to the village by the road'

Language - denotes a language, acting as an instrument in the process of speech.

Adyghe

(2.14)se ad?gabze??es-e-g??љ??e

Iadyghe-inssg.abs-dyn-speak

`I speak Adyghe'

Temporal - denotes a temporal mark.

Hebrew

(2.15)ani ozev be'od yoma'im Ileave-1sg.presloc-twoday-pl `I leave in two days'

Causal - denotes an instrument, which assists a particular purpose. I differentiate it from the cause instrumental meaning, because of the positive implication the causal instrument has.

Bengali

(2.16)dillI paryatan-er khAtire samriddha Delhiberich-erinstrtourism `Delhi is rich because of tourism' [Saint-Dizier et al., 2006]

Price - denotes the price of an object. Price is acting as instrumental object in such situations.

Hebrew

(2.17)kaniti et ha'ca'acu'a haze be'љnei dolar buy.pst.1.sgacc the-toy this loc-two dollar `I bought this toy for two dollars'

Exchange - denotes an object, that is the matter of exchange. Usually, the object that is being changed for is marked.

Hebrew

(2.18)hu hexlif et ha'sefer be'iparon he change.pst.3sg.m acc the-book loc-pencil `He changed the book for a pencil'

Comparison - denotes an instrument that acts as measurement item.

Hebrew

(2.19)Peter gavo'ah mi-mary be-xamesh santimerim Peterto.be.3sgcomp-maryloc-fivecentimeter `Peter is taller than Mary by five centimeters'

Possessed object close to the subject (physical proximity) - denotes an instrumental object, located close to the subject or the agent of the situation. I think this meaning is close to what is referred to physical proximity in (Narrog 2010).

Hebrew

(2.20)bati elexa [im| be- magafayim] meluxlaxot come.pst.1g to-you com OR loc-boots dirty `I came to you in dirty boots'

Instrumental objects:

Proto-instrument - see Instrumental above;

Contact instrument - usually a movable instrument having direct contact with another stationary instrument (Beloshankova & Muravenko 1985).

Hebrew

(2.21)Hu nigen ba'gitara im/be mafret he play.3MSG.PST in-guitar with/in pick

`He played the guitar with a pick'

Body part - the similar meaning to the prototypical instrumental meaning, but the instrumental object is represented with a body part.

Comanche

(2.22)[nahniiihtumagwainiiti]

naHniii-H/tu=ma-wai-nii-ti=

justIhere-toma-feel-mtn=around-gen:asp

`(It's so dark) I'm going to have to feel around (with my hands)' [Jean Ormsbee Charney 1997]

Location instrument (1) and (2) - the instrument, that is also used a location of the agent (1), or the instrument that uses the agent as a location (2). In the first sense, the instrument is often a means of transport, but different from the transport peripheral function because of a more direct instrumental usage. Basically, if one uses transport to move from point A to point B - it is a transport instrumental peripheral function, but if one uses transport to, for example, to plow the fields with a tractor or to run over someone or something in a car - it is a location instrument. In the second sense, the meaning is very close to possessed object close to the subject or physical proximity but I decided to differentiate these two meanings on the semantic map, because some language (e.g. Hebrew) marks these contexts differently (cf. `to read in glasses' and `to enter in dirty boots/with dirty hands' - more details in 4.1.1).

Hebrew

(2.23)ha'oved xarash et ha'sadot im traktor the-workerplough-pst.3sgaccthe-fieldcom-traktor `The worker plowed the fields on a tractor'

Russian

(2.24) On chital knigu v ochkah.

`He read the book in the glasses'

Action instrument - an abstract instrument, represented by an action, for example, jumping or running (Saint-Dizier et al. 2006).

Urdu

(2.25)Raam kuud kar ghar ke andar daakhil huaa

`Ram entered the house by jumping into it' [Saint-Dizier et al., 2006].

Abstract instrument - a non-physical instrument used by the agent and affecting the patient. This type of instruments is often represented by emotions, and emotive instruments often create ambiguous reading - I go into more details about this in the chapter 4.1.1.

Hebrew

(2.26)hu shalat b-a-ir im-paxad

he rule.pst.3msg in-def-city com-fear

`He ruled the city with fear' [Arsentyev, 2016].

Instrument source point - the instrument which has other means or objects inside.

Hebrew

(2.27)Ha'sporta'i tzadtzvi im rove the-sportsmanhunt-pst.3sgdeercom-rifle `The sportsman hunted deer with a rifle'

It is also important to notice that there is no particular tradition in using instrumental terms or deriving specific instrumental meanings. There is a number of works, which authors often refer to, such as (Apresyan 1974) and (Wierzbicka 1980) and (Fillmore 1981), but these works only partly cover the instrumental domain, and, therefore, later works often suggest different readings on the instrumental meanings.

The lack of the universal description of the notion instrumentality, and, therefore, of all the instrumental functions and different types of instrumental objects, leads to the necessity of creating my own classification, which was described in this chapter. I do not posit, however, that this type of framework is valid for all types of research - it could be shortened or expanded, depending on the purpose of the research. For this particular work, I believe, that this framework is sufficient and provides a good foundation for a cross-linguistic comparison.



3. Methodology

Compared to my previous work on the instrumental meanings (Arsentyev 2016), this paper exceeds it by far and provides both a rough descriptive analysis on a large sample of languages (more than 60 languages) and a more detailed analysis on a smaller number of languages. The former is completed by using various grammars and articles, which sometimes might be not very detailed considering the instrumental meanings. The latter is completed using extensive grammars with a decent amount of data about the instrumental grams and their functions, articles, online questionnaires, corpora, introspection and field work. The more detailed analysis includes the following languages:

· Permyak languages using the extensive grammars, online questionnaires and a number of papers by G.A. Nekrasova, who has been referring to the instrumental area in Permyak languages for more than twenty years;

· Hebrew using grammars and online questionnaires;

· Russian using grammars, online questionnaires, introspection and corpora;

· Adyghe using grammars and field work.

And the full list of the languages analyzed in the research is the following:

Table.1: The language sample, based on the competition type

Competition type	Language
The competition of adpositions	English
	Hebrew
	Kutenai
	Comanche
	Jakaltek
	Guarijio
	Spanish
	Yaqui
	Italian
	Luiseno
	Mayo
	German
	French
	Arabic
	Berber
	Thai
	Malay
The competition of cases	Finnish
	Udmurt
	Komi-Permyak
	Votic
	Estonian
	Komi-Zyrian
	Karelian
	Sami
	Lithuanian
	Livonian
	Mansi
	Khanty
The competition of cases and adpositions	Russian
	Turkish
	Kashmiri
	Urdu
	Hindi
	Hungarian
	Bengali
	Buryat
	Mari
	Erzya
	Chuvash
	Veps
The competition of applicatives and prepositions/postpositions	Ainu
	Abaza
	Indonesian
	Maybrat
	Tukang Besi
	Grebo
	Shona
	Swahili
	Sesotho
	ChichRewa
	Ndendeule
	Rwanda
	Serer
	Tuscarora (also a competition between two different instrumental applicatives)
	Kapampangan
	Oneida
The competition of applicatives and cases	Adyghe
	Dyirbal
	Greenlandic

The second table provides all the languages listed above, but takes into account not the type of the competition, but their geographical location. I use this table to illustrate my findings on the map in the chapter 6.

Table.2: The language sample, based on the location and language family

Region	Family	Language
America	Iroquoian	Oneida
		Tuscarora
	Mayan	Jakaltek
	Uto-Aztecan	Comanche
		Mayo
Africa	Niger-Congo	Serer
		Grebo
		Ndebele
		Chichewa
		Kinyarwanda
		Sesotho
		Swahili
		Shona
	Afroasiatic	Hebrew
		Arabic
Eurasia	Abkhazo-Adyghean	Abaza
		Adyghe
	Austronesian	Indonesian
		Kapampangan
		Maybrat
		Tukang-Besi
	Eskimo-Aleut	Greenlandic
	Uralic	Karelian
		Finnish
		Komi-Zyrian
		Komi-Permyak
		Khanty
		Mari
		Udmurt
		Veps
		Livonian
		Hungarian
		Mansi
		Votic
		Estonian
		Sami
		Erzya
	Indo-European	Russian
		English
		German
		French
		Spanish
		Italian
		Lithuanian
		Kashimiri
		Bengali
		Urdu
		Hindi
	Mongolic	Buryat
	Turkic	Chuvash
		Turkish
Isolates	America	Kutenai
	Eurasia	Ainu
Oceania	Pama-Nyungan	Dyirbal

I illustrate the table above on the four maps below. The first three maps are generated in the WALS application.



4. The competition of ways to express the instrumental meaning

4.1 Adpositions competition

In this category I put the following languages: English, Hebrew, Kutenai, Comanche, Jakaltek, Huarijio, Yaqui, Luiseno, Mayo, Thai, Arabic, Berber. Out of these languages, I most intensively studied Hebrew, and, therefore, the Hebrew discussion takes the biggest part of the chapter.

4.1.1 Hebrew discussion

Hebrew belongs to the Semitic languages and is spoken by around nine million people mainly in Israel. My data of Hebrew comes from (Neuman 2015) and extensive work with informants via online questionnaires and discussions. The data covers the usage of the locative and comitative prepositions be- and im-, checked on the instrumental main and peripheral functions and all types of instrumental objects, mentioned in the chapter 2.

Fig.10: The Hebrew instrumental functions semantic map



Fig.11: The Hebrew instrumental objects semantic map

So, in Hebrew the two main ways to express the prototypical instrumental meaning are the comitative preposition im- and the locative preposition be- (ba-), and according to informants, both ways are absolutely equal. However, it concerns only the prototypical instrumental meaning with a prototypical inanimate instrumental object, such as hammers, knives or a stick (4.1).

Hebrew

(4.1)hu xatax tapu'ax [im sakin | be-sakin] he cut.pst.3sg.m apple [com knife | loc knife] `He cut the apple with a knife'

When it comes to other instrumental functions and different instrumental objects, one of the instrumental grams becomes more preferable and sometimes the only possible way to express a particular function.

As the two instrumental grams are originally comitative and locative markers, the comitative and the locative meaning can only be expressed respectively, so by im- and be- only. The locative marker, however, has a wider range of functions when it comes to peripheral instrumental semantics, such as temporal, price and etc. (4.2-4.7).

Hebrew

(4.2)Ha'xayal met ba'emek the-soldierdie.3pst.sgloc-valley `The soldier died in the valley'(locative)

(4.3)hu hexlif et ha'sefer be'iparon he change.pst.3sg.m acc the-book loc-pencil `He exchanged the book for a pencil'(exchange)

(4.4)kaniti et ha'ca'acu'a haze be'љnei dolar buy.pst.1.sgacc the-toy this loc-two dollar `I bought this toy for two dollars'(price)

(4.5)bati elexa [im| be- magafayim] meluxlaxot come.pst.1g to-you com OR loc-boots dirty `I came to you with dirty boots'(physical proximity)

(4.6)hu nasa be-otobus hetravel.pst.3sg.m loc-bus `He traveled by bus'(means of transport)

(4.7)Peter gavo'ah mi-mary be-xamesh santimerim Peterto.be.3sgcomp-maryloc-fivecentimeter-pl `Peter is taller than Mary by five centimeters'(comparison)

(4.8)ani ozev be'od yoma'im Ileave-1sg.presloc-twoday-pl `I leave in two days'(temporal)

The comitative marker, on the other hand, covers more types of instrumental object usage, such as abstract or location instrumental objects (4.9-4.10)

Hebrew

(4.9)hu shalat b-a-ir im-paxad he rule.pst.3msg in-def-city com-fear `He ruled the city with fear'

(4.10)Hu yachol likro rak im-mishkafayim He can.m.pres read.inf only com-glasses.pl `He can only read in glasses'

Finished describing the semantic area of the comitative and locative preposition, now I go into details about the competitive functions and provide my thoughts and hypotheses on the factors influencing the choice of instrumental gram.

The first factor influencing the choice of the grams is the sentential semantics. Basically, by sentential semantics I mean the purpose of the instrumental object in the sentence. The most frequently encountered difference in the purpose is the instrumental semantics opposed to attributive semantics. For example, the body part as the instrumental object is usually marked with the locative preposition (except for the verb `to break'), but with the sentential attributive semantics both markers are possible (cf. (4.11) and (4.12)).

Hebrew

(4.11)hu heka oto be-egrofo he hit.pst.3msg him loc-fists.his `He hit him with his fists'

(4.12)bati elexa [im|be- yadayim] meluxlaxot come.pst.1g to-you com OR loc-hands dirty `I came to you with dirty hands'

The difference in these sentences is in the purpose of the instrumental object. In (4.11) the instrumental object acts as an instrument, but in (4.12) it is used as an attribute for the subject and does not have any instrumental purpose. This distinction also leads to the change of the instrumental function, that I mark as possessed object, close to subject or physical proximity in the semantic map.

The second distinction, based on the purpose of the instrumental object, can be shown on the examples (4.13) and (4.14).

Hebrew

(4.13)hu shalatb-a-ir im-paxad he rule.pst.3msgin-def-city com-fear `He ruled the city with fear' (using the fear as an instrument)

(4.14)hu shalatb-a-ir be-paxad he rule.pst.3msgin-def-city loc-fear `He ruled the city with fear' (being afraid of the process)

Comparing these examples, we can see that the same instrumental abstract object allows two different reading depending on the preposition it is applied to. In these examples, contrary to (4.11) and (4.12), the comitative marking gives an instrumental reading, while the locative marker represents an attributive reading.

The third distinction, based on the purpose, concerns the means of transport function and location instrument (2). Among all the mentioned examples, the contrast in (4.15) and (4.16) covers the subtlest difference.

Hebrew

(4.15)hu nasa be-otobus hetravel.pst.3sg.m loc-bus `He traveled by bus'(means of transport)

(4.16)Hu daras oto im-mexonit he drive.3msg.pst acc-3 com-car `He ran him over with a car'

In both examples, the instrumental object is a means of transport, however, they do serve different purposes. In (4.15) the transport is used to get from point A to point B - the typical meaning for the means of transport objects, however, in (4.16) the transport is used rather as a more typical instrument than a transport, so the instrumental object becomes just a means to achieve a particular goal (run over somebody in (4.16)). Such a distinction justifies the differentiation between means of transport as an instrumental function and a location instrument (2) in the semantic maps.

Based on these distinctions, the most basic conclusion would be that in Hebrew the instrumental functions are more important than the types of instrumental objects, as the choice of preposition is more determined by the sentential meaning rather than the instrumental object. Languages do differ in this aspect; for example, the English language marks the difference between the examples concerning the means of transport (4.15) and (4.16), but does not mark the difference between the physical proximity meaning (4.11) and (4.12). The Russian language, on the contrary, does not mark the difference when it comes to the transport instrumental object, but marks the difference in the physical proximity function. And Hebrew differentiates in all of these cases.

So, the opposition of instrumental functions and instrumental objects does not seem to be universal even in one language, as depending on the meaning Russian and English choose whether it is the instrumental function or the instrumental object is a more significant factor for the preposition usage. And this choice varies from one meaning to another. In Hebrew, it seems like the instrumental functions are more important, and I would say that it makes a lot of sense and is very logical. I cannot be sure in this, as it would require the neurolinguistics experiment to check, but I would hypothesize that the difference in preposition choice would signal the difference in the meaning, so it would require less cognitive effort for speakers to comprehend and produce such sentences.

However, going into deeper analysis, it becomes clear that Hebrew must use the difference in the prepositions choice when it comes to the means of transport meaning. So, the transmission from the `drive in a car/bus' to `run over in a car/bus' in Russian, English and Hebrew would require the following:

In Russian, it would require the change of the verbal morphology and sometimes the preposition depending on the particular means of transport;

In English, it would require the change of the predicate and sometimes the preposition;

In Hebrew, it does not require neither the change of the predicate nor the instrumental object, so the preposition has to change.

So, in Hebrew the change of the preposition plays the same role as the change of the verbs or verbal morphology in other languages. I believe, that it does not necessarily mean that the hypothesis about the significance of instrumental functions and objects is wrong, but rather implies a new (quite obvious) statement, that if the language does not have enough grammatical features, it would require to differentiate the two meanings with the help other grammatical features, such as prepositions in Hebrew.

I might call this type of distinction in (4.11-4.16) as the difference in the purpose of the instrumental object or the difference in the sentential semantics, but no matter the name, it leads to different instrumental functions, and depending on the purpose of the research, the list of these functions can be expanded quite easily. For example, based on Hebrew, it would be justified to add the attributive meaning (4.14) onto the semantic map of instrumental functions. But then, I think, the list might get expanded too heavily, so I would lose the illustrativity component I wanted to show with the semantic maps. So, I believe that sometimes it is better to go back to a descriptive method and provide the explanation of such cases as (4.13) and (4.14) in the text, rather than in the semantic maps.

Also, in Hebrew some verbs might restrict the usage of particular instrumental grams. For instance, the verb `to break' cannot be used in the clause with the locative preposition and, therefore, changes the compatibility of some instrumental functions with instrumental grams. As I mentioned above, the body part as an instrumental object usually requires the locative preposition, but with the verb `to break' it can only be used with the comitative preposition.

Hebrew

(4.17)Hu savar et ha-agartal im-ha'regel he break.3msg.pst acc the-vase com-the.foot

`He broke the vase with his leg'

(4.18)ha-birjon savar et ha-xalon im-even

the-thug break.3msg.pst accthe-window com-stone

`The thug broke the window with a stone'

(4.19)*ha-birjon savar et ha-xalon be-even

the-thug break.3msg.pstaccthe-window loc-stone

`The thug broke the window with a stone'

(4.20)ha-birjon heka et ha-xalon be-even

the-thug hit.3msg.pst accthe-window loc-stone

`The thug broke the window with a stone'

In the example (4.17), we can see that the prototypical instrumental function with body part as an instrumental object takes comitative as the instrumental gram, even though it is usually used with the locative preposition. In (4.18), we can see that the prototypical instrumental function with a typical instrumental object might takes comitative preposition as well, however, this type of object should be compatible with both comitative and locative preposition (4.20). But in (4.19) we can that it is not possible to use the locative preposition, and I believe that the only explanation of that is the restrictions determined by the verb.

In (Neuman 2015), the author posits another factor of determining between the locative and comitative prepositions, and it concerns the means of transport function. This factor is the education and the socioeconomic status of the speaker. Neuman suggests that the more educated the speaker is, the higher the chance that he or she would use the locative preposition.

So, the main factors of determining the choice of instrumental grams in Hebrew are the instrumental functions and the type of instrumental objects (which are typical factors for most languages). Other factors are the compatibility of instrumental grams with certain verbs, the socioeconomic status of speakers and the opposition of instrumental functions and instrumental objects.

4.1.2 Arabic discussion

My data of Arabic comes from (Saint-Dizier et al. 2006) - a multilingual analysis of the notion of instrumentality, and (Frolov 1977). The data covers the usage of the several prepositions such as bi-, its modifications bi-tarika, bi-istemali, bi-fadli and also the prepositions min khilal, min a, au, ala. Using the semantic maps from the chapter 2, I provide the following illustration.

Fig.12: The Arabic instrumental semantic map

However, additional research, specifically, discussion with native speakers, is necessary to provide more data and restrictions on the usage and competition of the instrumental prepositions in the language.



4.1.3 Berber discussion

The Berber languages comprise a large number of dialects, but a number of common elements can be selected. The data comes from (Saint-Dizier et al. 2006) and covers Berber from the Moroccan Rif, and the Algerian Kabyle.

Fig.13: The Berber instrumental semantic map

4.1.4. Thai discussion

The data of the Thai language comes from (Saint-Dizier et al. 2006). Thai has six prepositions to express the instrumental meanings with a little amount of overlaps or competition among them.

Fig.14: The Thai instrumental semantic map



Fig.15: The Thai instrumental object semantic map

4.2 Cases competition

In this category I put the following languages: Finnish, Udmurt, Komi-Permyak, Votic, Estonian, Sami, Karelian, Khanty, Livonian, Lithuanian, Mansi and Komi-Zyrian. All mentioned languages also have adpositions as grammatical morphemes, however, according to the data, the usage of postpositions is exclusively rare, compared to the usage of cases. And as in this chapter, I mostly discuss the cases competition, I decided to put all these languages in this section. It is opposed to the next chapter, where I talk about languages that also have cases and adpositions, but in those languages adpositions play a crucial role and sometimes might be the only way to express a particular instrumental meaning.



4.2.1 Permyak languages discussion

The data of Permyak languages, namely, Komi-Permyak, Komi-Zyrian and Udmurt, comes from (Nekrasova 1997), (Nekrasova 2011), (Nekrasova 2015), online discussion with native speakers and the usage of the Udmurt corpus. The competition occurs among the instrumental, accusative, elative, comitative, inessive cases and some postpositions. In all Permyak languages, the main grammatical morpheme to express the instrumental meanings is the instrumental case (unlike, for example, other Uralic languages, described in the next section). So, other cases and postpositions would be generally used rarer than the instrumental case, but with some instrumental functions the main instrumental case cannot be used.

Firstly, I show the common instrumental functions that all three languages share on the semantic map, and later I go into details considering each language separately.

Fig.16: The Permyak languages instrumental semantic map

As we can see, the instrumental case covers a wide variety of meanings, however, unlike many other languages in my sample, the instrumental case in Permyak languages does not cover the locative meanings. It is possible to express the prolative meanings in Komi-Zyrian, however, such usage is very rare.

Komi-Zyrian

(4.21)Вузасьысь морецн ыстiс караб ~ The trader sent the ship by sea[Nekrasova, 1997]

Inessive, on the other hand, is also a locative case presented in all three Permyak language, and its main semantic area is locative meanings. Both the instrumental and inessive cases seem to come from the Proto-Uralic language locative *-na/-nд, the intermediate step in the change from locative to instrumental was the associative meaning. I go a little bit more into details about grammaticalization patterns in chapter 6. As differentiated factors, inclusivity and contact seems to influence the choice of the instrumental gram. Basically, with the inessive case the instrumental object has to be in contact physically with the patient it is affecting, and with the instrumental case the instrumental object does not necessarily have to be in contact, so indirect instruments (such as rifles or computers), abstract instruments or action instruments are possible.

The Permyak languages also have other types of case competitions, such instrumental/accusative, instrumental/elative, instrumental/comitative and instrumental/postpositions. I do not show all of these competitions on the semantic map, because verbal semantics plays a crucial role in the choice of the instrumental gram, but in my sample this factor rarely influences the choice, so I do not have a specific semantic map for that. Therefore, I am going to describe this type of competition without semantic maps.