The Grammaticalized Element Chans and Related Difficulties in Corpus Linguistics [1]

DOI: 10.55804/jtsuSPEKALI-17-17

 

1. Theoretical Postulates Defining Grammaticalization

According to the traditional definition, language is a system of arbitrary sound signs, in which each sign follows certain regularities. Linguistic signs create content and expression planes of the language and the exploration of the language system requires a thorough analysis of both planes [Gamkrelidze et al., 2008: 23].

Language, as a communication system, possesses various possibilities of reflecting, in surface structures, the information embedded in deep structures. The information embedded in the deep structures of the language system manifested in the grammatical categories and rules or regularities operating in the language, is studied within the framework of traditional linguistics. The surface structures of the language, specifically the area of its functioning, are considered the domain of functional grammar. The process of grammaticalization also belongs to the field of this research.

The primary function of language, its employment for communication, is carried out through linguistic inventory. The latter is formed by the unity of elements with grammatical and lexical meanings. The inventory of the language is divided into various paradigm classes. The language system is constantly changing; it changes and develops according to internal laws. Any change in the speech habit is related to a person's attempt to convey the thought more simply, clearly and accurately [Hopper & Traugott, 2008: 81]. Along with the changes in the language, the linguistic sign can change its function and move from one paradigmatic class of words to another. This type of change also includes the process of grammaticalization [Lehmann, 2015: 10]. Grammaticalization is also a well-known process in typology. It is a universal phenomenon and is considered a logical factor in language development.

During the process of grammaticalization, the linguistic sign carrying the lexical meaning loses its lexical meaning and can no longer make a reference, that is, it can no longer establish the usual relationship with the signifier. The delexicalized form acquires a specific grammatical function and becomes a functional element [Hopper, 2003: 27]. It should be noted that the process of grammaticalization is a qualitative, gradual change in the function of a linguistic unit and not a leap from one category to another. Accordingly, linguistic units in the process of grammaticalization are characterized by different degrees of grammaticalization [Lehmann, 2015: 45].

In the scientific literature dedicated to the issue, four stages of grammaticalization are distinguished. They are as follows:


  • Desemantization – when a linguistic unit loses its lexical meaning and this is determined by various factors, the main one being the semantic context. 
  • Extension – increasing the frequency of employment of a lexical unit with a new function. At this stage, the area of utilization of the decategorized element expands. 
  • Decategorization – a linguistic unit loses its paradigmatic (grammatical) markers, that is, a word that changes form gradually becomes a word that no longer changes its form. At this stage, the linguistic unit may preserve morphosyntactic features but it can no longer establish a morphosyntactic connection with other members of the sentence, and can no longer produce a paradigm. An independent word is dependent on other words and becomes a unit accompanying them. 
  • Erosion – a decategorized and desemantized linguistic unit which has moved from the form-changing word class to the non-form-changing word class, becomes simpler. The majority of particles in Georgian are considered to be derived in this way [Tandashvili, 2016: 95]. 


From the typological point of view, linguistic units of verbal origin create an interesting picture in the process of grammaticalization. What do we mean when we say that a language form is grammaticalized? Compare the cases of use of the form adga  (has got up) in the sentences below:

(1) Gogona dilit adre adga  da mzadebas sheudga (The girl got up early in the morning and started preparing). Cf. (2) netav, iseti ra utkhra, rom ert dgesats adga da martla sheusrula tkhvona?! (What did he say to him so that one day he got up and really fulfilled his request?!)

In the first sentence, adga performs the function of the syntactic nucleus of the sentence, of the predicate, whereas, in the second sentence, it does not have the function of the predicate. Predication is expressed not by adga, but by the verb sheusrula (fulfilled). As for adga, although it retains the main verb categories, such as a person, number, and tense, it loses the verb meaning and becomes a functional element conveying the meaning of the beginning of an action.

2. The Form “Chans” to Be Explored The process of grammaticalization in the natural language system is characteristic of both nouns and verbs. Typologically, a particularly interesting picture is made up of the grammaticalization of verb forms. This paper presents the process of grammaticalization of the verb form Chans. This issue is studied through the corpus-based research method. The research presented in the article was carried out on the basis of the National Corpus of the Georgian Language (GNC), and the contexts were confirmed in the corpus of old, middle and new Georgian language; the political texts corpus was taken as research material. Substitution, elimination, and scope methods are used to specify the functional meaning of the element. While clarifying the functional-semantic nuances, the paper relies on the language competence test conducted with the participants of the research.

The unit chans in Georgian is a one-person verb of the present screeve expressing predication, which agrees with the third subjective person in the nominative case. For example:

Panjridan araperi chans, mkholod shematsukhebeli khmauri ismis (Nothing can be seen from the window, only annoying noise is heard).

The unit chans does not always perform the function of the axis of the sentence of the verb predicate; devoid of lexical content, it retains the same form but no longer expresses predication. It becomes a functional element in the sentence. As an example, see the following sentences:

(1) chans, matareblis mosvlis drots kargad itsoda da amgvar sakmezec ar iko pirvelad (It seems that he knew the time of the arrival of the train very well, and was experienced in what he was doing)

(2) me morchi, chans, chemma magalitma skhvebi tsaakhalisa (I have finished; it seems that my example encouraged others).

(3) tskhovreba tsin aris da, rogorc chans, arts ormotsdaati tselia bevri (Life is ahead, and as it looks, years is not much either).

3. Research Methodology

When working on functional elements, various methods are used, among them, in this paper, I present the methods of substitution, elimination and scope.

When discussing the functions of a grammaticalized element, the method of substitution is particularly effective, which implies replacing a linguistic element with another element similar in lexical or functional meaning. If a valid variant is received as a result of substitution, the morphosyntactic function of the initial element is considered confirmed [Tandashvili, 2019: 12]. Syntagms used for replacement by participants of the research provide information about the semantic and functional meanings of the functional element, as well as about the syntactic connections of the syntagm in the sentence. For example, consider the following sentence:

Chans, matareblis mosvlis drots kargad itsoda da amgvar sakmezec ar iko pirvelad (It seems that he knew the time of the arrival of the train very well, and was experienced in what he was doing). Cf. It looks that he knew the time of the arrival of the train very well, and was not experienced in what he was doing.

In the case considered, It seems is replaced by it looks.  As a result of this replacement, both the original meaning of the sentence and the logical relations in the sentence, as well as the syntactic structure, are maintained. Both elements are functionally equivalent and express the speaker's assumption about the action conveyed by the verb.

Elimination is a syntactic rule that allows revealing the element or group of elements of the sentence sufficient and necessary for the sentence to convey the complete meaning. This test works especially effectively when exploring functional elements of verbal origin. Syntactically, the verb is the nucleus, the central part of the sentence in Georgian. The action/state is conveyed by the verb. If the element of the verbal origin or syntagm can be eliminated, this means that the verb function is lost, the predication is no longer expressed and we are dealing with a functional element.

Chans, matareblis mosvlis drots kargad itsoda da amgvar sakmezec ar iko pirvelad (It seems that he knew the time of the arrival of the train very well, and was experienced in what he was doing )Cf. It seems he knew the time of the arrival of the train very well, and was experienced in what he was doing.

It can be seen that the syntactic structure of the sentence is not violated as a result of the elimination of chans. The modality of the sentence does change; instead of conveying an assumption, the sentence expresses an assertion. Therefore, we can conclude that in the considered case, chans is devoid of the lexical meaning and has become a functional element.

When exploring functional elements, the scope method is employed to determine the area of functionality of a given element. The scope is the area of functioning of a functional element, i.e. the distance in the sentence up to which the semantic function of the given element extends. Defining the scope is important because the area of functional-semantic modification of the functional element is determined in the sentence.

(1) chans, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It seems that he knew the time of the arrival of the train very well, and was experienced in what he was doing).

(2) matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad, chans (He knew well the time of the train's arrival and was experienced in what he was doing, apparently.

(3) matareblis mosvlis drots kargad itsoda, chans, da amgvar sakmezets ar iko pirvelad (He knew well the time of the train's arrival, it seems, and was experienced in what he was doing).

(4) matareblis mosvlis drots kargad itsoda da, chans, amgvar sakmezets ar iko pirvelad (He knew well the time of the train's arrival, and it seems that he was experienced in what he was doing).

In the compound coordinated sentence presented above, the change of the place of the chans element neither breaks the syntactic construction of the sentence nor obscures its meaning. This confirms the fact that this element in the given example should not be qualified as lexical but as a functional element. The movement of the verb form within a sentence, especially between main and dependent clauses, is impossible when using it as a lexical element.

In the example discussed above, it can be seen that the use of the form chans as a functional element is indicated by the fact that it is separated by commas and can be eliminated easily.  The grammaticalized element chans performs the function of a modal element in a sentence. If it is removed, the sentence loses modality whereas its movement within the sentence causes a change in the modality of the sentence.

The change of the position of the grammaticalized element chans in a compound coordinated sentence is presented in the form of a diagram below:

(1) {[simple sentence] and [simple sentence]}.

(2) chans, {[simple sentence] and [simple sentence]}.

(3) {[simple sentence] and [simple sentence]}, chans.

(4) {[ simple sentence ]}, chans, and {[simple sentence]}.

(5) {[ simple sentence]} and, chans, {[simple sentence]}.

In the first sentence, by eliminating the element chans, we get a neutral proposition, the sentence loses the understanding of modality and it only describes the real situation without specifying the intention of the speaker. In the second and third sentences, the element chans  modifies the whole sentence. The sentence loses neutrality and acquires the meaning of modality: it conveys the speaker's intention, and his/her personal attitude, and expresses the speaker’s assumption. Thus, it expressed the semantics of assumption. The fourth and fifth sentences are particularly interesting. The compound coordinated sentence is connected by the conjunction and, the element chans is placed before the conjunction which leads to the change in the scope of the element chans:  the semantic function of the modality extends only to the left periphery of the sentence (to the simple sentence to the left of the element chans), while placing the element after the conjunction “and “; changes its semantic-functional area of operation. It moves to the right periphery and applies to the simple sentence placed after the conjunction. The fact is that in the compound sentence,  the connection “and “ locks the scope of the element chans.

As can be seen from the examples discussed above, the form chans is undergoing the process of grammaticalization. Out of the four discussed levels of grammaticalization, it should be evaluated as a unit at the extension level, which is confirmed by the cases of its use, on the one hand, as a predicate, and on the other hand, as a functional element.

4. Functional-Semantic Analysis

Functional-semantic analysis can be performed using a competency test. Therefore, in addition to the methods discussed above, when describing the functions of the grammaticalized element chans, we relied on the language competency test, conducted as follows:  10 informants were selected for the pilot study. They were given a list of sentences and the instruction to replace the underlined word in the sentence with a word of similar meaning and to rewrite the sentence without changing the content. We present the obtained results in the tables below.

 

Table 1

 

Chans, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It seems that he knew the train arrival time well and  was experienced in what he was doing).

Participant I

etkoboda, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It looked like he knew the train arrival time well and was experienced in what he was doing).

Participant II

Rogorts vkhdav, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (As I can see, he knew the train arrival time well and was experienced in what he was doing.

Participant III

etkoba, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad

(It looks that he knew the train arrival time well and was experienced in what he was doing).

Participant IV

albat, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad

(Perhaps he knew the train arrival time well and was experienced in what he was doing).

Participant V

sheidzleba, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (Possibly  he knew the train arrival time well and he was experienced in what he was doing).

Participant VI

albat, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (Probably he knew the train arrival time well and he was experienced in what he was doing).

Participant VII

ashkaraa, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (Perhaps he knew the train arrival time well and he was experienced in what he was doing).

Participant VIII

etkoba, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It looks like he knew the train arrival time well and he was experienced in what he was doing).

Participant IX

ashkaraa, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It is obvious that he knew the train arrival time well and he was experienced in what he was doing).

Participant X

Ashkara iko, matareblis mosvlis drots kargad itsoda da amgvar sakmezets ar iko pirvelad (It was obvious that he knew the train arrival time well and was experienced in what he was doing).

 

 

Table 2

 

Me movrchi, chans, chemma magalitma skhvebi tsaakhalisa ((I have finished; (It seems that my example encouraged others).

Participant I

Me movrchi, ashkaraa, chemma magalitma skhvebi tsaakhalisa ((I have finished; It’s obvious that my example encouraged others).

Participant II

Me movrchi, vxedav, chemma magalitma skhvebi tsaakhalisa ((I have finished; I can see that my example encouraged others).

Participant III

Me movrchi, da chemma magalitma skhvebi tsaakhalisa ((I have finished and my example encouraged others).

Participant IV

Me movrchi, vimedovneb, chemma magalitma skhvebi tsaakhalisa ((I have finished; I hope my example encouraged others).

Participant V

Me movrchi, shevamchnie, chemma magalitma skhvebi tsaakhalisa ((I have finished; I have noticed that my example encouraged others).

Participant VI

Me movrchi, tskhadia, chemma magalitma skhvebi tsaakhalisa ((I have finished; it is obvious that my example encouraged others).

Participant VII

Me movrchi, tvalsachinoa, chemma magalitma skhvebi tsaakhalisa ((I have finished; it is noticeable that my example encouraged others).

Participant VIII

Me movrchi, vpikrob, chemma magalitma skhvebi tsaakhalisa ((I have finished; I think that my example encouraged others).

Participant IX

Me movrchi, tvalsachinoa, chemma magalitma skhvebi tsaakhalisa ((I have finished; it is noticeable that my example encouraged others).

Participant X

Me movrchi, vpikrob, chemma magalitma skhvebi tsaakhalisa ((I have finished; I think that my example encouraged others).

 

Table  3

 

 

Tskhovreba tsin aris da, rogorts chans, arts ormotsdaati tselia bevri

(Life is ahead, and as it looks/seems fifty years is not much either).

Participant I

Tskhovreba tsin aris da, etkoba, arts ormotsdaati tselia bevri

(Life is ahead, and it looks like fifty years is not much either).

Participant II

Tskhovreba tsin aris da, vgrdznob, arts ormotsdaati tselia bevri

(Life is ahead, and I feel that fifty years is not much either).

Participant III

Tskhovreba tsin aris da, dartsmunebuli var, arts ormotsdaati tselia bevri

(Life is ahead, and I am sure that fifty years is not much either).

Participant IV

Tskhovreba tsin aris da, paktia, arts ormotsdaati tselia bevri

(Life is ahead, and it is a fact that fifty years is not much either).

Participant V

Ormotsdaati tseli arts ise bevria, radgan tskhovreba isev tsin aris

(Fifty years is not that much, as life is still ahead).

Participant VI

Ormotsdaati tseli arts ise bevria, vpikrob, tskhovreba isev tsin aris

(Fifty years is not that much, I think, life is still ahead).

Participant VII

Tskhovreba tsin aris da, rogorts aghmochnda, arts ormotsdaati tselia bevri

(Life is ahead and, as it has turned out, neither 50 years is much).

Participant VIII

Tskhovreba tsin aris da, rogorts ikveteba, arts ormotsdaati tselia bevri

(Life is ahead and, as it appears, neither 50 years is much)

Participant IX

Tskhovreba tsin aris da, rogorts aghmochnda, arts ormotsdaati tselia bevri

(Life is ahead and, as it has turned out, neither 50 years is much).

Participant X

Tskhovreba tsin aris da, mgoni, arts ormotsdaati tselia bevri

(Life is ahead and, as I believe, neither 50 years is much).

Based on the results of the language competence test, the following nuances can be distinguished: the main function of the grammaticalized element chans is to express the speaker's modality towards the action conveyed by the sentence. This is confirmed by both of the above-discussed methods and the lexical units chosen for replacement by the participants. The modalities related to the grammaticalized element chans are united into two large semantic categories, namely: the semantics of assumption and the semantics of certainty. In each group, smaller subgroups can be singled out and distinguished from each other by minor semantic nuances. It can be seen that the modalities related to the grammaticalized unit chans are presented in the form of a scheme:

Table  4

Grammaticalized element

Replacement element

Modality

Chans

etkoba (It looks like)

A logically conditioned assumption

Chans

rogorts vkhedav (As I can see)

A supposition based on visual perception

Chans

vkhedav (I see)

An assumption based on visual perception

Chans

 Sheidzleba (Probably)

Semantics of possibility

Chans

 albat (Perhaps)

Semantics of assumption

Chans

(vimedovneb (I hope)

Semantics of hope

Chans

mgoni (I believe)

Semantics of assumption

Chans

shevamchnie (I have noticed)

An assumption based on individual vision

Chans

vpikrob (I think)

An assumption based on personal judgement

Chans

vgrdznob (I feel)

An assumption based on personal judgement

Chans

rogorts aghmochnda (As it turned out)

Causal assumption

Chans

rogorts ikveteba (As it looks)

Causal assumption

Chans

ashkaraa (It is obvious)

Semantics of cognitive certainty

Chans

tskhadia (It is clear)

Semantics of logical certainty

Chans

paktia (It is a fact)

Semantics of logical certainty

Chans

tvalsachinoa (It is noticeable)

Semantics of visual certainty

Chans

dartsmunebuli var (I am sure)

Semantics of cognitive certainty

 

5. Grammaticalized Element Chans  in Corpus Linguistics

The process of grammaticalization takes place step by step, parallel forms co-exist in a certain period of time. Therefore, the issue of homonymy arises. This (the presence of homonymy) creates certain difficulties in corpus linguistics.

The leading direction of linguistics of the 21st century is computer linguistics, the creation of language corpora, which involves automatic language processing and placement in a language corpus. The smooth operation of the corpus requires the development of a sophisticated automatic analysis program-analyzer. When working on Georgian language material, we face difficulties in this direction. A necessary prerequisite for creating a sophisticated analyzer program is a complete analysis of the language system from the morphological, semantic, and functional points of view. Since the Georgian language material has not been properly studied from a functional point of view, correct annotation and removal of homonymy arising in the corpus become difficult.

As an example of homonymy, the form discussed below can be seen in this paper. The presented examples are taken from the National Corpus of the Georgian Language:

(1) Mtel tsaze ghrublebis erti ptilits ki arsad chans da zemodan mtvaris katsichamia sakhe dagvkatkatebs (There is not even a small cloud in the whole sky visible, and the man-eating face of the moon is staring at us from above).

The token chans in the corpus is annotated as follows - verb, medioactive voice, present screeve.

(2) Kantselariis gamgem, chans, droze veraperi moipikra da karga khnis shemdeg tkva (Apparently, the Head of the Chancellery could not think of anything in good time and spoke after a long time).

It can be seen that the token chans, like the token presented in the previous example, has been assigned a verb marker, although in the considered example it is devoid of predicate meaning.

6.  The Ways to Remove Homonymy

The pilot study conducted to find ways to remove the homonymy raised in the corpus is based on the National Corpus of the Georgian languag;e the contexts taken from the corpus of old, middle and new Georgian language and political texts are analyzed. The analysis material consists of 880 contexts. The material analyzed in the pilot study confirms the previously developed hypotheses about the grammaticalized element chans and yields the following combinatorial rules:

1. If the token rogorts is on the left of the token chans,  it must be assigned as a functional element marker;

Of the 138 similar collocations confirmed in the analyzed material, each of them is confirmed as a functional element chans.

2. If the token chans is separated by a comma on both sides, then it is assigned a functional element marker.

Of the 138 similar collocations confirmed in the analyzed material, chans is confirmed in each as a functional element.

3. If a token chans is placed at the beginning of a sentence and is separated from the next token by a comma, it should be assigned a functional element marker.

Of the 30 similar collocations confirmed in the analyzed material, the token chans is confirmed as a functional element in each of them.

We maintain that the proposed combinatorial rules, after being rewritten in the form of an algorithm, will help the automatic analyzer program of the National Corpus of the Georgian Language to recognize and correctly analyze the grammaticalized element chans, and reduce the cases of erroneous marking-annotation of homonymous forms.

References

გამყრელიძე თ. და სხვები...,
2008
თეორიული ენათმეცნიერების კურსი, თბილისის სახელმწიფო უნივერსიტეტის გამომცემლობა, თბილისი.
თანდაშვილი მ.,
2016
ქართული ენის ფუნქციური გრამატიკის საფუძვლები, ბათუმის შოთა რუსთაველის სახელობის უნივერსიტეტის გამომცემლობა, ბათუმი.
შარაშენიძე ნ. და სხვები...,
2020
მოდალობის კატეგორიის საკითხები ქართულში, გამომცემლობა უნივერსალი, თბილისი.
Frajzyngier Z.,
2008
Grammaticalization, Typology, and Semantics: Expanding the Agenda. In Rethinking Grammaticalization. New Perspectives; Amsterdam: Benjamins.
Heine B.,
2003
Grammaticalization; Oxford: Blackwell; In Janda &Joseph (eds.); The handbook of Historical Linguistics, Oxford.

2002
On the Role of Context in Grammaticalization, Amsterdam, Philadelphia: Benjamins.
Hopper P., Traugott J.,
2003
P. E.C. Grammaticalization; Cambridgr University Press, Second edition, Cambridge.
Lehman Ch.,
2015
Thoughts on Grammaticalization, Second edition, Seminar für Sprachwissenschaft der Universität, Erfurt.