Филологические науки / 3. Теоретические и методологические проблемы исследования языка

Ph.d. S. V. Kiyko

Chernivtsi National University, Ukraine

HOMONYMY AND THE category of markedness

Many linguists view homonymy as a negative phenomenon which interferes with communication, complicates the perception of information, and decreases the effectiveness of the language as a means of communication. Homonymy is characterized by L. Novikov as an unnatural phenomenon that complicates communication [6: 209]. The same idea is expressed by M.P. Kolesnikov: homonymy erases the formal differences between the signs with different content, and distorts information [5: 7]. The experimental research in the area of text perception shows that when given the sentence with the ambiguous elements, the time of the recipient’s reaction to the message is considerably increased [1].

At the same time some researches think that homonym contributes to the compactness of the language and allows to economize the units of the plain of content [7: 52]. Investigating the mutual influence of the word’s form and meaning, O. Ducháček came to the conclusion that homonymy is not harmful for the language, which is confirmed by the existence of homonyms in practically every language. Moreover, the use of homonyms in literature serves various purposes, for instance, to form puns [2: 11].

However, the quoted authors solve the problem of homonymy’s benefits or drawbacks mostly in theory, citing only several most vivid examples, without the processing of the sufficient volume of material, in particular, without the consecutive analysis of homonyms in the unilingual dictionaries.

The objective of our research is to review the categoric statements according to which homonymy causes interference in the process of communication, and to determine the factors that differentiate the meaning of homonymic units, based on the broad factual material. The study is based on the consecutive analysis of homonyms selected from the Dictionary of the German Language Duden [Duden 2000]. The object of the investigation is the homonymy of the Modern German nouns. The total number of studied homonyms is 2128 lexical units combined into 1018 homonymic rows. Most homonymic rows have two components, e.g. die Mutter¹ “mother”, die Mutter² “nut”; the total number of such pairs is 937 (1874 homonyms). We selected 72 three-component rows (216 homonyms), e.g. die Messe¹ “mass” (religious), die Messe² “fair”, die Messe³ “wardroom”; 8 four-component rows (32 homonyms), e.g. die Note¹ “note” (musical), die Note² “academic grade”, die Note³ “diplomatic note”, die Note⁴ “hue, undertone” etc. There is also one six-component row: Atlas¹ “one of the Titans”, der Atlas² “geographic atlas”, der Atlas³ “neck vertebra”, der Atlas⁴ “satin”, der Atlas⁵ “telamon”, der Atlas⁶ “the mountain in Africa”.

It is a known fact that in the plane of content any homonymic group is characterized by the absence of the interlexemic semantic ties. It means that in most cases homonyms must belong to different lexico-semantic groups (LSGs), e.g.:

1) names of people referring to their age, gender, nationality, relations → names of dishes: Pinkel¹ (coll.) “man” vs. Pinkel² (Northern German) “a sort of sausage”, Knacker¹ “an old man” vs. Knacker² “smoke-dried sausage”; Tatar¹ “Tatar” (nationality) vs. Tatar² “raw steak”;

2) names of animals → names of diseases: Krebs¹ “crayfish” vs. Krebs² “cancer”, Star¹ “starling” vs. Star² “cataract“, Wolf¹ “Wolf” – Wolf² “lupus“;

3) names of cloth/fabric → types of clothing: Trikot¹ “knitted fabric“ vs. Trikot¹ “tights”, Reversible¹ “two-sided fabric” vs. Reversible² “two-sided clothes” etc.

In our research 86% of all homonymic rows belong to different LSGs. It means that more than two-thirds of homonymic nouns are semantically differentiated based on the fact that they belong to different LSGs. 14% of homonymic nouns (298 homonyms, 138 homonymic rows) belong to the same LSG. They are differentiated in most cases with the help of grammatical gender, e.g. der Assi¹ “assistant” (he) – die Assi² “assistant” (she), das Band¹ “strip, band” – der Band² “book volume” – die Band³ “band“. If the nouns belong to the same grammatical gender, different form of plural may be used for their differentiation (5 homonymic pairs), e.g. das Wort¹ “word” (plural Wörter) – das Wort² (plural Worte) “cue, catchword”, die Bildung¹ (plural die Bildungen) “formation” – die Bildung² (only singular) “education” etc. In other cases the decisive role in the differentiation of the complete homonymic nouns belongs to sociological, areal, stylistic or chronological aspects, i.e. the homonyms are differentiated via their belonging to different subsystems of lexis. The above cited criteria for homonymic differentiation are presented in Chart 1:

Chart 1

Criteria of Differentiation of the Homonymic Nouns

№	Criteria of differentiation	Number of homonymic rows	Examples
1	Belong to different LSGs	876	der Hahn^1“rooster” – der Hahn^2 “water tap”
2	Areally marked	16	der Flaum^1 (areal) “lard” – der Flaum^2 “fluff”
3	Socially marked	11	der Riemen^1“belt” – der Riemen^2 (nav.) “oar”
4	Chronologically marked	5	die Schelle^1 (arch.) “handcuffs” – die Schelle^2 “bell”
5	Stylistically marked	4	der Skater^1 (coll.) “skater” (on skates) – der Skater^2 “skater” (on a skateboard)
6	Used in set expressions	3	der Plan^1 “action” – der Plan^2 “plan”

As we can see, the fact that most homonymic nouns belong to different LSGs, and that those belonging to the same LSG can be differentiated with the help of various grammatical indices and stylistic markings, allows to quite accurately differentiate their meaning. Regarding this the components of the homonymic row can be differentiated based on the category of markedness, which correlates with the cognitive operator of norm / deviation. It is a known fact that the linguistic notion of markedness is applied to various components of the language structure; it has high explanatory potential and cognitive value. The notion of markedness was derived from phonology and gained special value in the typological description of the asymmetry of grammatical parameters in the works of G. Greenberg [3]. In grammar the marked (strong) member of the opposition has some formally expressed feature (e.g. plural of nouns) and narrower and more precise meaning than the unmarked one.

In the homonymic row, the homonym registered in the dictionary under №1 is as a rule unmarked, while the others are marked, i.e. they are limited in their use stylistically, chronologically, arelly or socially. This means that the homonymic row consists of marked and unmarked members. Strong members of the homonymic rows can be characterized as functionally limited lexis, which is opposed in their differrential features to the active, generally used, neutral nominative language content. Moreover, the notion of “marked lexis” is much wider than that of “stylistically marked lexis”: marked lexemes bear any additional (to their lexical meaning) information about the grammatical meaning, sphere of use, temporal reference, emotional and expressive coloring or functional stylistic use of the lexical units.

The fact that one of the elements of the homonymic row (the first one as a rule) is unmarked was proved by a number of psycholinguistic experiments, where we offered the German speakers to suggest the first association word which occurred to them referring the homonyms in the list. This research was based on 200 homonymic pairs from our selection, chosen based on different semantic, stylistic, chronological, areal or social reference of one of the components of the pair. The questionnaire included both homogenic homonymic pairs whose homonyms have common origin, and heterogenic pairs, which coincide in sounding due to borrowing, incidental phonetic coincidence etc. Both homogenic and heterogenic pairs had 100 examples.

The experiment was carried out in a group of students from the Institute of Historical Studies, Friedrich-Schiller-University Jena (Germany), aged 21-25, whose native language is German. The procedure was as follows: the instructions were given orally. It was required to write down the first word that occurred to the informants as connected with the stimulus word. The data obtained were grouped and the frequency of use of each reaction word to the given stimulus word was calculated (see Chart 2, where some examples are cited). In the Chart 2 the 3^rd and 5^th columns contain dictionary definitions of every homonym to compare the obtained data. The frequency of occurrence of associate words is given in parentheses after the words.

Chart 2

The List of Associations with the Various Components of the Homonymic Rows

№	Homonymic row	The meaning of the 1^st component of the homonymic row	The list of associations	The meaning of the 2^nd component	The list of associations
1	Mutter	Mutter^1, die; -, Mütter 1. a) Frau, die Kind(er) geboren hat; b) Frau, als Mutter Kinder erzieht; c) Vorsteherin eines Klosters; 2. Tier, das Junge geworfen hat; 3. (Technik) Matrize; 4. (Jargon) Mutterfirma	Kind(er) (14), Frau (9), Tochter (3), Liebe (2), Pflege, Zuhause Σ 30	Mutter^2, die; -, -n kurz für Schraubenmutter	Σ 0
2	Geschoss	Geschoss^1 das; -es, -e; das, was aus einer (Feuer)Waffe abgeschossen wurde	Kugel (2), Waffe, schießen, Ziel Σ 5	Geschoss^2 das; -es, -e Etage, Stockwerk	Etage (15), Stockwerk (7), Aufzug, hoch, Haus Σ 25
3	Land	Land^1 das; -(e)s; nur Sg 1. der Teil der Erde, der nicht vom Wasser bedeckt ist; 2. ein bestimmtes Gebiet, wo man besonders Pflanzen anbaut	Erde (7), Acker (3), Boden (2), Festland, bearbeiten, Fläche Σ 15	Land^2 das; Länder 1. das politisch selbstständige Gebiet; 2. Teil eines Landes, der eigene Regierung hat	Staat (9), Bundesland (3), Sachsen (2), Bayern Σ 15

The obtained results show that the homonym cited in the dictionary under №2 is mostly marked (150 examples, 75%). In 200 homonymic pairs only 22% (43 examples) show that the first component of the pair is marked, e.g. der Kuli¹ “cheap worker” (0 associates) and der Kuli² “ballpoint pen” (30 associates), der Rauch¹ “thick fur” (0 associates) and der Rauch² “smoke” (30 associates), der/das Scharlach¹ “purple colour” (0 associates) and der Scharlach² ”scarlet fever” (30 associates), das Geschoss¹ ”shot” (5 associates) and das Geschoss² “floor, storey” (25 associates) etc. Pairs heterogenic in their origin prevail among the homonymic pairs with the first marked component (26 heterogenic pairs as opposed to 15 homogenic), which is apparently conditioned by the peculiarities of lexicographic practice. It is a known fact that in the homogenic pairs the most frequent component is cited first, which is not observed for the heterogenic pairs.

In seven cases (3% of examples) the number of associations with the first and second component of the pair is approximately equal, e.g. das Land¹ “dry land” and das Land² “country” (15 associates), das Pflaster¹ “cobblestone” (14 associates) and das Pflaster² “plaster” (16 associates), der Mittag¹ “afternoon” (16 associates) and der Mittag² “dinner” (14associates), der Bogen¹ “arc” or “bow” (16 associates) and der Bogen² “a sheet of paper” (14 associates), die Fessel¹ “chain” (16 associates) and die Fessel² “ankles” (14associates), der Leiter¹ “leader, guide” (16 associates) and die Leiter² “ladder” (14 associates), die Raserei¹ “fury, wrath” (14 associates) and die Raserei² “fast ride, race” (16 associates).

We grouped the homonymic pairs according to the markedness of one of the components in the subgroup, and in each subgroup we calculated the relative number of the pairs with the marked components (when the ratio between the associates of the marked and unmarked component is 30 to 0). Thus, among the homonymic pairs with one socially marked component the relative portion of such homonymic pairs makes 75%, i.e. we registered 25 homonymic pairs with the most marked component out of 33 homonymic pairs analyzed (see Chart 3).

Chart 3

Homonymic Rows with Socially Marked Components

Homonymic row	The markedness of the 1^st component	Number of associations	The markedness of the 2^nd component	Number of associations
Gast, Fall, Glas, Riemen, Pink		30	navy	0
Drossel, Schütze		30	engineering	0
Stuhl, Galle		30	medicine	0
Schiff		30	architecture	0
Dom		30	geology	0
Flucht		30	construction	0
Jäger		30	military	0
Set		30	printing	0
Kraut		30	soldiers language	0
Locke		30	hunting	0
Schmiere		30	criminal	0
Karre		30	geology	0
Post		30	basketball	0
Riff		30	music	0
Standard		30	Jazz jargon	0
Popper		30	Jargon	0
Stift		28	christl. church	2
Stab		27	military	3
Spannung		26	physics	4
Hyazinth		26	greek mythology	4
Lob		26	tennis, badminton	4
Gesellschaft		19	economy	11
Kreuzer		11	military	19
Raute		4	geometry	26
Neptun	roman mythology	27		3
Venus	roman mythology	24	.	6
Viola	botany	18		12
Zettel	textile industry	0		30
Rauch	technical	0		30
Stern	navy	0		30

Here belong homonymic units used in the scientific (general scientific and branch terms, scientific and technical professionalisms), official (administrative, diplomatic and law terminology), publicistic (publicistic terms, socio-political lexis and terms), religious (religious lexis and terms) styles of the German language, as well as bookish lexical units, e.g. der Gast² (navy) “sailor”, die Locke² (hunting) “decoy”, die Schmiere² (criminal) “lookout” etc. The common feature of these groups of words is their use as a means of communication of the separate social, professional and age groups of people. The rest of the groups may be ranged as follows:

1) homonymic groups where one of the components is only used in set expressions: 100% (5:5). Here belong such examples as die Lampe² (in the expression Meister Lampe) “Master Hare”, der Onkel² (großer/dicker Onkel) “toe”, der August² (dummer August) “clown”, Hummel (Hummel, Hummel!) “hey, hello” etc.

2) homonymic groups where one component is shortened: 100% (1:1). We only found one homonymic pair of this type: die Birne¹ “pear” and die Birne² (short of Glühbirne) “light bulb”.

3) homonymic groups with one chronologically marked component: 83% (5:6). Chronologically marked lexis includes obsolete words (archaic and historic), e.g. der Zelt² (arch.) “pass”, (arch.) die Hecke² “mating season” etc.

5) homonymic groups with one areally marked component: 52% (11:21). This group is represented by the words with clear dialectal reference, e.g. die Beige² (Southern German, Swiss) “pile”, das Heck² (Northern German) “enclosure, pen” etc.

6) homonymic groups with one stylistically marked component: 25% (5:17). This group is represented by the words which, apart from their objective notional meaning, have components of subjective character: emotion, expression, imagery, evaluation. According to the emotional and expressive coloring connotatively marked lexis is divided into positively and negatively colored. Elevated, rhetorical words, lexical units with the emotional approval, and some joke lexemes bear positive emotional charge, e.g. die Hochzeit² (humorous) “flourish”, der Reif² “wedding ring” etc. Negative evaluation is characteristic for colloquial words, which are differenttiated according to the level of pejoration – from humorously ironic and familiar to rude and vulgar (expressive colloquialism), e.g. der Schwindel² (coll. vulgar) “fraud”, die Raserei² (coll. vulgar) “races”, die Horde² (coll. vulgar) “gaggle” etc.

7) homonymic groups with components without marking: 29% (28:96). In this group the number of associates is influenced by the frequency of the word’s use, i.e. the words with greater frequency get more associates, e.g. die Mutter¹ “mother” (30 associates) and die Mutter² “nut” (0 associates), der Bart¹ “”beard” (30 associates) and der Bart² “key bit” (0 associates), die Lippe¹ “lip” (30 associates) and die Lippe² “the tributary of the Rhein” (0 associates) etc.

The results of the psycholinguistic experiment show that the components of the homonymic pairs that are used in set expressions, shortened or chronologically marked, are the most marked. The least marked are those components that are cited in the dictionary without any stylistic marking, or they belong to the connotatively marked lexis, i.e. have emotional and expressive coloring.

The psycholinguistic analysis we carried out shows that the predominate number of homonymic pairs (97% of our selection) have marked and unmarked components. This allows to explain homonymy from the point of view of the cognitive-language correlation of “markedness / unmarkedness”, and wider – “norm / deviation”. From the cognitive point of view language markedness is derived from cognitive markedness, i.e. the unmarked language meaning corresponds to the cognitively normal (natural, expected) state of things, and the marked language meaning corresponds to cognitive deviation, i.e. unnatural, unexpected state of things. As stated by O.E. Kibrik, normal state of things belongs to the cognitive image of human experience, and is conceptualized with the minimal mental calculating effort, i.e. is activated automatically; and deviations from this image require additional calculating resources for their activation [4: 62]. Thus, language markedness reflects cognitive operators of norm / deviation in the specific language means in language structures, including homonymic pairs and homonymic rows. The presence of marked and unmarked elements in the homonymic row in its turn demonstrates the synergetic potency of homonymy.

The prospects of the further research lie in the studies of the influence of cognitive-language correlation “norm/deviation” based on the consecutive analysis of homonyms of other parts of speech, primarily verbs and adjectives.

Acknowledgements

This project was made possible by a generous research grant from the Erasmus Mundus Research Program. I am heartily thankful to my supervisor, Prof. Dr. Christine Römer, Friedrich-Schiller-Universität Jena, whose support enabled me to develop an understanding of the subject.

References

1. Cairns H. S. Effects of bias on processing and reprocessing of lexically ambiguous sentences / H. S. Cairns // Journal of Experimental Psychology. – 1973. – Nr. 97. – P. 337-343.

2. Ducháček O. O vzájemném vlivu tvaru a významu slov / Otto Ducháček. – Praha : Státní pedagogické nakladatelství, 1953. – 191 s.

3. Greenberg J. Language universals, with spezial reference to feature hierarchies / J. Greenberg. – The Hague : Mouton amd Company, 1966. – 89 p.

4. Kibrik А.Е. Lingvističeskaja rekonstrukcija kognitivnoj struktury / А.Е. Kibrik // Voprosy jazykoznanija. – 2008. – № 4. – S. 51–77.

5. Kolesnikov N.P. Omonimija v predloshenii i voprosy jeje ustranenija : av. diss. … dokt. filol. nauk : 10.02.01 „Russkij jazyk“. – Tbilisi, 1968. – 22 s.

6. Novikov L.А. Semantika russkogo jazyka / L.A. Novikov. – М. : Vys. shkola, 1983. – 272 s.

7. Osmanova R.А. О javlenii omonii v lesginskom literaturnom jazyke / R.А. Оsmanova // Učenyje zapiski Azerb. gos. un-ta. – Baku : AGU, 1962. – S. 57–64.