|Publisher:||Duke University Press|
|Series:||Sound and Meaning: The Roman Jakobson Series in Linguistics and Poetics|
|Sold by:||Barnes & Noble|
|File size:||1 MB|
Read an Excerpt
The Union of Asymmetry and Semiosis in Language
By Edna Andrews
Duke University PressCopyright © 1990 Duke University Press
All rights reserved.
The Principles of Jakobsonian Markedness Theory
Gesamtbedeutung and a Definition of Markedness
There is no SIGNATUM without SIGNUM.—Roman Jakobson
The goal of any linguistic analysis is to describe or explain a particular phenomenon of language, and a multitude of approaches can be applied to achieve this goal. Whatever the approach, however, the linguist is faced with the problem of describing what is potentially infinite in a finite, meaningful manner. One means of linguistic analysis is based on the principle that there are certain invariants in language which recur at different levels within the structure of any given language and which generate a multitude of variations. Roman Jakobson, in the Saussurian tradition, is one of the greatest champions of explicating a theory of invariants by distinguishing between general versus contextual meaning (or Gesamtbedeutung and Grundbedeutung) in morphology and semantics.
Jakobson's major contribution to establishing a theory of general meaning came in his 1936 analysis of the Russian case system. In this pioneering work, Jakobson demonstrated that the various uses of each of the six Russian cases can be described in terms of three invariant, semantic features. Later, in his 1958 article, "Morfologiceskie nabljudenija nad slavjanskim skloneniem," Jakobson expanded his definition of these three semantic features, further demonstrating their relationship to the morphology of the case system. In these two major works and in other morphological studies, Jakobson combined the principle of Gesamtbedeutung (or general meaning) with semantic features, defining them in terms of the marked/unmarked relationship. In his article "Shifters, Verbal Categories, and the Russian Verb" (1957) Jakobson defined the relation marked/unmarked as follows: the marked term gives the statement of a property A; the unmarked term can be divided into two components: (1) a general meaning = nonstatement of A; (2) a specific meaning = statement of non-A. Gesamtbedeutung is given in every instance by means of a marking (or feature) or a bundle of features. Therefore, the definition of markedness in morphology includes both a general and a specific meaning, whereas the definition of markedness in phonology does not provide for a distinction in the unmarked term; only the "specific" definition occurs. Note the diagrams in figure 1.
Of course, Jakobson's theoretical approach has frequently been criticized on the premise that although such analyses are elegant, they are inherently unverifiable, are not self-explanatory, and have very limited predictive power, so limited that Wierzbicka, in The Case for Surface Case, argues that "the predictions they generate are very often incorrect" (1980:xv). Wierzbicka goes on to say that Jakobson's formulations are "too vague, too general to be empirically adequate" (xvi), and she also criticizes form-content linguistic analyses in the Diverian school (cf. Garcia) as suffering from the same lack of specificity, the same vagueness and unverifiability. As an alternative to looking for one invariant definition for each case category, Wierzbicka proposes an invariant meaning for each different "surface" manifestation of the Russian instrumental case. She gives seventeen invariant meanings for the instrumental and claims each to have full predictive power. Thus, she counters the validity of positing one invariant meaning for a given case and, in exchange, proposes seventeen so-called invariant meanings. Therefore, Wierzbicka does not abandon the notion that invariant meaning exists, but she redefines the line between invariance and variation, broadening the potential variation of invariance.
Wierzbicka's accusations would be valid complaints against the Jakobsonian approach if Jakobson, concluding his linguistic analysis with his 1936 case study, had failed to further develop his conceptualization and application of Gesamtbedeutung. However, this is not the case. Rather, these early articles represent only Jakobson's first step in working toward a cohesive theory of meaning using the principles of invariance and markedness.
Henrik Birnbaum has taken a slightly different approach in arguing against a theory of general meanings. Citing such eminent Slavic scholars as Potebnja and Kacnel'son, Birnbaum restates the question asked by Kacnel'son: Why should linguists, knowing very well that polysemy and synonymy exist in language, renounce these realities in search of some abstract belief in the isomorphism of form and content? Birnbaum concedes that Gesamtbedeutung is most likely "a genuine reality of language and not merely a construct invented by the linguist," but he ultimately decides that the concept is not terribly useful "when it actually comes to analyzing a grammatical form's overall meaning" ([in press]:7–8). As an example of the futility of working with general meaning, Birnbaum notes that Jakobson "compellingly illustrated his conception of general meaning as an invariant of lexical semantics," although he concludes his argument by pointing out that Jakobson did not develop "a systematic theory of general meaning of the lexicon" (21–22). However, C. H. van Schooneveld, working within a Jakobsonian framework, undertook precisely that task of developing a systematic theory of general meaning for the Russian lexicon (van Schooneveld: 1983,1985). Although van Schooneveld has not completed his analysis of the Russian lexicon, his work to date is convincing in its own right.
Unlike Wierzbicka and Birnbaum, I have argued that the quest for general meaning can be meaningful, can have descriptive power, and can provide a powerful pedagogical tool (Andrews 1984b, 1985a, 1986d, 1987a). The search for and application of general meaning can be a useful tool, even the essential one, in linguistic analysis. In the process of looking for general meanings, complex linguistic structures can be related and elucidated. Peircean semiotic theory, too, in particular Peirce's theory of interpretants, stresses the quest for general meaning. As Elizabeth Bruss states: "The meaning of a concept must be general, it cannot be exhausted on any single occasion of use but must extend to ever new instances. Peirce's logical interpretant is therefore a rule of interpretation which includes all possible uses and all possible users" (Bruss 1978:84). According to Peirce, the interpretant is the rule or law that permits the translation of the sign. Thus, we can derive a fundamental connection between Jakobson's definition of general meaning and Peirce's notion of interpretant.
The inadequacy of the theory of general meaning, as described by its critics, seems actually to reflect a criticism of the manner in which the concept has been applied, not the concept itself. To make Gesamtbedeutung a useful concept, the linguist is faced with the task of avoiding extreme abstraction, which indeed is not always easy. However, there are ways to avoid this problem. Whenever a particular morphological or semantic phenomenon is being analyzed, the linguist must never stray too far from the actual linguistic data itself, even for the sake of theoretical elegance.
If, in our quest for Gesamtbedeutung, we are better able to understand the languages we study, then such a quest is justified. However, we must at no time forsake detailed examination of language, which is our only true, measurable data, for vague abstractions. General meaning, if correctly applied, does not demand such a sacrifice.
Erica Garcia succinctly states what the goal of linguistic analysis should be:
A better understanding of that process which necessarily underlies the use of language, namely the creative synthesis of a coherent gestalt from a few abstract, well-defined clues. That knowledge will never be gained by solipsistic contemplation of elegant models. We must face up to the fact that language is used (and known!) by people, and consequently get our hands dirty with the nitty-gritty of controlled experimental research, (in press)
It is precisely this point of view that the following chapters will exemplify: a theoretical demonstration of the need for understanding the structure of form in order to determine meaning.
A Brief History of Markedness: From Trubetzkoy to the Present
Roman Jakobson first proposed the oppositional relation of markedness in 1921 (Holenstein 1976:126). The terms marked/unmarked (originally priznak in Russian, and secondly merkmal in German) were introduced and applied by Nikolai Trubetzkoy in his 1931 article on "Die phonologischen Systeme." The German term was "Merkmalhaltigkeit"—the property of containing a distinctive mark. However, Trubetzkoy's conception of markedness applied strictly to phonology.
As early as 1927 Jakobson had already formulated certain notions which were later to serve as the basis of his definition of morphological markedness. Specifically, Jakobson addressed the idea of "correlation" as binary opposition in his Proposition to the First International Congress of Linguists, which was cosigned by Karcevskij and Trubetzkoy (1928): "A phonological correlation is constituted by a series of binary oppositions defined by a common principle which can be conceived independently of each pair of opposite terms." When Jakobson brought his theory to the United States, he adopted the phrase "distinctive feature," which was used by Leonard Bloomfield, for the German term merkmal (Bloomfield 1933:79).
As a markedness theory began to take shape in those early days, the notion of binary opposition became very important in defining markedness relations. Yet, even in the early days of the Prague School, the two major figures of that school did not agree on what "markedness" meant. Specifically, Jakobson and Trubetzkoy defined the term "opposition" in very distinct ways.
Trubetzkoy, never a proponent of the theory of purely binary oppositions in phonology, developed three types of phonemic oppositions:
1. Privative—Two phonemes are identical except that one contains a "mark" which the other lacks (e.g., /b/ vs. /p/).
2. Gradual—Different degrees of some gradient property (e.g., /i/ ~ /e/ ~ /æ/).
3. Equipollent—Each member has a mark that the others lack (e.g., /p/ ~ /t/ ~ /k/) (Sampson 1980:108)
Jakobson, long a proponent of binary opposition, on the other hand, expanded and modified his theory of binariness to encompass morphology and semantics. However, the primacy of binary oppositions in linguistic systems is not universally accepted. Although Jakobson firmly adheres to the notion of asymmetrically defined binary opposition, where the unmarked term is not on a par with the marked term, the phonological distinctive features define the relationship marked/ unmarked differently from the morphological conceptual features. In phonology, a given feature (e.g., voicing) is either present or absent in a particular class of phonemes (or, possibly, indifferent to the feature, as the feature sharpness is not applicable to vowels). Therefore, the opposition in phonology is privative, whereas in morphology, marked means the necessary presence of an element, while unmarked means the element may or may not be present—it is simply not specified; therefore, unmarked does not merely imply negation. In short, the phonological features can be viewed as forming mutually exclusive sets (as shown in figure 1 under specific meaning), while the morphological features form an inclusion relationship.
In his theory of morphological markedness Jakobson clearly does not equate the unmarked element with the element that occurs most frequently in syntagmatic contexts. Rather, Jakobson bases his determination of the marked/unmarked pair on the presence of a property of meaning. These "properties of meaning" are defined as intrinsic properties given by the linguistic system itself, not by external reality (1967:671). The fact that Jakobson views the linguistic system as defining its own order of reality does not prohibit a potential relationship between linguistic and extra-linguistic phenomena. Yet, Jakobson does state that the intrinsically defined properties and relationships of language should be analyzed first before considering extrinsically defined properties and relationships, implying that the linguistic object is defined within the linguistic system proper and is not the same thing as an object in reality. An important concept related to Jakobson's distinction between intrinsic and extrinsic relationships is the notion of relative autonomy. Thus, linguistic signs within the linguistic system are defined in terms of oppositional relations, which are relative, while the linguistic object is necessarily autonomous of extralinguistic reality. In order to fully appreciate the importance of relative autonomy, Peirce's definition of the sign-object relationship becomes essential. Of particular interest is the notion of iconicity, the first of three sign-object relationships defined by Peirce.
For Peirce, an icon is "a sign which stands for its object because of some similarity to that object" (Fitzgerald 1966:55 [italics mine]). Furthermore, "the only way of directly communicating an idea is by means of an icon; and every indirect method of communicating an idea must depend for its establishment upon the use of an icon ..." (Peirce 2.278). Therefore, in language all forms of signified-signifier relationships, insofar as they convey ideas and make reference, must be iconic. The degree of iconicity inherent in a given signified-signifier relationship depends on the relative hierarchy, since icons do not occur alone and may only occur in combination with other types of signs, namely symbols and indices (Fitzgerald 1966:55). Examples of iconicity in language include the following: (1) plural versus singular nominal forms, where the plural, meaning "more than one," in many instances has an additional affix (cf. English "table/tables"); (2) positive, comparative, and superlative adjectives that show an increase in the number of phonemes in the affix (cf. English "big/bigger/biggest"). [For more examples of iconicity, see Jakobson 1971c:351–59.]
The above discussion would be incomplete, of course, without asking how "similarity" between the sign and its object is determined. According to Peirce, this similarity is not created by the interpreter but, rather, is a given within the ground of the relation. In other words, the interpreter does not establish or create the similarity—he merely uses it (Fitzgerald 1966:46–48). An iconic relationship unused remains a merely potential iconic sign, but nonetheless iconic it remains.
Those linguists and schools of linguistics which have adopted some form of markedness theory can be divided into two major groups: (1) those who work with meaning, and (2) those who attempt to describe language as a purely "formal" system without utilizing meaning. In particular, the Generative school redefined markedness relationships as properties derivable from a syntagmatic context in phonology, excluding any application of meaning. Noam Chomsky and Morris Halle, for example, redefined phonological markedness features for English based on distributional criteria in the form of general rules. The fundamental difference between Jakobson's work in phonology and that of Chomsky and Halle is the need, in Jakobson's theory, for a distinct phonemic level in phonology. Halle criticizes the phonemic level in The Sound Pattern of Russian on the grounds that its presence prevents the statement of certain linguistic generalizations on their simplest level. However, most of Chomsky's and Halle's objections to the difficulties involved in maintaining an autonomous phonemic level can be resolved by granting that the phoneme is not the basic phonological unit in language. Rather, it is the distinctive feature which is the minimal phonological unit in language. Jakobson has carefully defined distinctive features as invariant oppositions which are paradigmatically given and are fundamental to both speech perception and production: "Any distinctive feature exists only 'as a term of relation'" (1971a:641–42). Phonological features defined by syntagmatic properties (cf. tenseness and gravity in the Russian vowel system) are distinct, but not distinctive, units. Distinctive units require the paradigmatic axis (the axis of replacement) in order to find positions of contrastive distribution (Schane 1973:40).
Joan [Bybee] Hooper, working in the area of natural generative phonology, defines markedness in phonology by describing the unmarked term as a more "natural" and the marked term as a less "natural" phenomenon (1976:136). Hooper makes an additional point that "markedness" and "naturalness" are not two separate notions but are both derivable from a single set of rules (1976:135). In terms of rule development Hooper makes significant progress by including only phonological rules that have no exceptions with regard to surface forms. Leonard Babby's application of generative grammar to Russian uses two types of markedness—one having nothing to do with meaning and another which is loosely based on the Jakobsonian definition.
Excerpted from Markedness Theory by Edna Andrews. Copyright © 1990 Duke University Press. Excerpted by permission of Duke University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of ContentsContents
One The Principles of Jakobsonian Markedness Theory
Gesamtbedeutung and a Definition of Markedness
A Brief History of Markedness: From Trubetzkoy to the Present
Jakobson's and van Schooneveld's Morphological Features
Mark or Feature?
Deixis: A General Definition
Deixis and Reference
The Union of Conceptual Feature and Deictic Category: Shifters in the Russian Verb
Two Peirce and Jakobson Revisited: A Reconciliation
Peirce's Influence on Jakobson
Markedness Theory as a Theory of Interpretants
Sign, Object, and Interpretant
The Categories of Firstness, Secondness, and Thirdness
The Ultimate Interpretant
Meaning and Interpretants
Icon, Index, Symbol, and Artifice?
Dyads or Triads?
An Application of Peircean Interpretants to Linguistic Analysis
Three Markedness Theory as Mathematical Principle
Mathematics: Topology and Its Relevance to Jakobsonian Linguistics
The Role of Syllogisms and Different Types of Inference
Deductive, Inductive, and Abductive Reasoning
Continuity or Discontinuity?
Dynamics and Structural Stability
Catastrophe Theory and Linguistic Sign Theory
An Axiomatic Base for Markedness
The Morphological Conceptual Features and Group Theory
The Laws of Form and the Form of Laws
Four Myths About Markedness
Five The Category of Grammatical Gender in Russian, Serbo-Croatian, and Modern Greek
Modern Greek (Dhimotiki/Demotic) and the Role of the -s Marker
Jakobson's Analysis of the Category of Gender in Modern Russian
The Serbo-Croatian Gender Hierarchy
Grammatical Gender versus Declension: Syntagmatic Vs. Paradigmatic
Is Gender Meaningful?