Explaining the Meaning of Words: A Descriptive Study on Strategies

Table of Content

1. Introduction

2. Preliminaries: Theory and Terminology

3. Retrieving data for a collection of strategies

4. Isolating strategies, establishing relations
4.1 Strategies involving intension
a) giving a synonym
b) giving a hyperonym
c) naming the part of speech
d) syntagmatic or grammatical proximity
4.2 Strategies of Reference
a) giving an example
b) appealing to imagination
c) miming
d) telling an anecdote

5. Conclusion

6. Appendix
A. Transcripts
B. Dictionary Entries
C. Jabberwocky by Lewis Carroll

7. Works Cited

8. References

1. Introduction

Spoken words are the symbols of mental experience and written words are the symbols of spoken words. Just as all men have not the same writing so all men have not the same speech sounds, but the mental experiences, which these directly symbolize, are the same for all, as also are those things of which our experiences are the images. (Aristotle, De Int. I., 16a)

Words, things and mental experiences – these three, their nature and the relationships that hold between them, form the scope of lexical semantics. Aristotle, in this programmatic statement “On Interpretation”, explicitly distinguishes between spoken and written word. He assigns a clear priority to the former: rooted in a culture of still predominantly oral transmisson, the difference between writing and speech is more palpable for him than we are inclined to feel it today. We are quite used, if we think of a word, to imagine its written form – as if this is its origin and essence. In need for the meaning of a word, we usually go and look for its entry in a dictionary. “Words” and “books” seem inextricably linked to each other, also suggested by the etymology of terms such as “lexicon” (literally, “a collection of words”). Terminology in L2-learning seems to metaphorically extend this connection by suggesting that the lexical inventory of language is stored in memory like in some kind of dictionary or mental lexicon. Rieder, in her research-paper on incidental vocabulary acquisition, quite naturally talks about the role of written discourse in L2-learning: her approach is about finding the “principles and guidelines ^[that] characterize the process of (re)constructing the meaning of unknown words within the text comprehension process […].” (Rieder 2002: 53) Her object of study is the L2-learner, confronted with a written text containing unknown words. One of her objectives is to find out about the interplay of top-down (text-level) and bottom-up (word-level) processes in the effort of constructing on the one hand a mental representation of the text and meaning-hypotheses for the unknown words on the other (cp. Rieder 2002: 54).

In our study group, we have sketched out an approach that rests on similar assumptions and works on related objectives, but with two important differences. First, it works “the other way round”: instead of watching a recipient infer an unknown lexical item with the help of co-text (drawing on knowledge resources which allow him to create contextual coherence), we chose to have speakers of different competence levels explain a lexical item – in other words, build a coherent text whose purpose is to optimally (in their opinion) guide an inference process in a potential recipient ignorant of the word. The second difference is that our (potential) recipients are hearers rather than readers, that our texts are tied to persons rather than to letters. The decision to use oral material rests on the belief that producing a written text is different from speaking in one very relevant respect: it is much easier to follow the actual process of thinking in our subjects, “Die allmähliche Verfertigung der Gedanken”^[i], as Kleist put it, if we make them speak rather than read or write.

The material we observed and analyzed by means of a theoretical apparatus taken from lexical semantics and pragmatics, yields a variety of strategies for word-explanation. The purpose of this is to present our approach, describe its implementation, name and categorize the most salient strategies.

2. Preliminaries: Theory and Terminology

Our approach is all about getting people to talk about words: this talk is it that we want to analyze. In order to make coherent sense of quite a variety of such attempts, we feel the need to establish a portion of technical language^[ii] and think about the theoretical models underlying it. We will make use of a number of concepts from the field of linguistic semantics. We will concur with the idea that there are distinct types of meaning; we will distinguish between intensional lexical relations (between lexical items), extensional and referential relations; we will take up some of the implications of componential analysis and keep a close look at Snell-Hornby’s model of verb-descriptivity derived from it. Where possible, we will try to keep this consistent with a broader pragmatic and cognitive perspective on communication and understanding.

Lyons distinguishes a number of aspects of meaning: one set of distinctions he makes is “between lexical meaning and sentence meaning […]” (Lyons 1981: 139, emph. by auth.), grammatical meaning and utterance-meaning. “The meaning of a sentence depends upon the meaning of its constituent lexemes […]; and the meaning of some, if not all, lexemes depends upon the meaning of the sentences in which they occur.” (Lyons 1981: 140) There is also “grammatical meaning as a further component of sentence meaning.” (Lyons 1981: 140) More comprehensive (arguably outside the traditional scope of semantics and rather an object of pragmatics) is the notion of utterance-meaning: “The meaning of an utterance includes, but is not exhausted by, the meaning of the sentence that is uttered.” (Lyons 1981: 140) Utterance-meaning includes the contributions of situative context. The distinction between sentence- and utterance-meaning must remain blurry, however,

[…] because the notion of sentence-meaning is arguably dependent, […] upon the notion of utterance-meaning, so that one cannot give a full account of sentence-meaning without relating sentences, in principle, to their possible contexts of utterance. (Lyons 1981: 140)

For this reason, “[…] semantics in the narrow sense is not logically prior to pragmatics. The two are interdependent” (Lyons 1981: 168) In the realm of lexical meaning, which concerns us most immediately, Lyons discerns “two […] components [of meaning]: sense and denotation.” (Lyons 1981: 151)

It is obvious that some lexemes, if not all, are related both to other lexemes in the same language […] and to entities, properties, situations, relations, etc., in the outside world. […] A lexeme which is related to other lexemes is related to them in sense ; […] a lexeme which is related to the outside world is related by means of denotation. (Lyons 1981: 152)

The notion of reference is, much like denotation, “[…] a relation which holds between expressions and entities, properties or situations in the outside world.” (Lyons 1981: 168) The two are, however, not identical: “ […] there is an important difference between denotation and reference: the latter, unlike the former, is bound to the context of utterance.” (Lyons 1981: 168) Reference, in this view, is a pragmatic notion rather than just a semantic one: a lexical item in use refers to an entity “in the outside world” (to Aristotle’s “things”). A lexical item in the language system denotes something in the mind (it “symbolizes directly” Aristotles says, “mental experiences”). This “something in the mind”, then, bears some indirect relationship to things in “the outside world” (it is an “image” of it, Aristotle thought): this is it that we expect to discover in word-explanation – verbalized denotations. There have been attempts to describe these denotations by virtue of the systematic relations that hold between lexical items. Intensional or sense-relations between lexemes are subsumed under the headings of “substitutional” and “syntagmatic”: substitutional relations are either those of synonymy, hyponymy or antonymy. “[…] Syntagmatic relations hold between […] ‘eat’ and ‘food’; between ‘kick’ and ‘foot’; and so on.” (Lyons 1981: 155) Taken together, substitutional and syntagmatic relations between lexemes form so-called “semantic fields”. Componential analysis (at least chronologically^[iii]) emerged from semantic field theory. Its underlying assumption is that the denotation of a lexeme is determined by (and explicable in terms of) relationships within the language system. Componential analysis tackles meaning by utilizing the principles of contrastiveness and structure: analogous to the way phonetics makes use of “a bundle of contrastive features”, the meanings of a language are studied in terms of a set of (putative) universal atomic features or concepts^[iv]. “[…] For example the meaning of the word woman could be specified as +HUMAN, -MALE, +ADULT […].” (Leech 1981: 11) In Leech’s taxonomy, this constitutes “cognitive or denotative meaning”. Leech gives a broad overview of six other types of meaning a lexical item can possess, but calls them “[…] more peripheral categories.” (Leech 1981: 20) As most important one, after “conceptual meaning” (which is explicable in terms of componential analysis), Leech brings up “connotative meaning”. He defines it as “the communicative value an expression has by virtue of what it refers to, over and above its purely conceptual content”. Conceptually, he argues, “woman” is defined by the three abstract features (+HUMAN, -MALE, +ADULT);

“[…] these contrastive features, translated into ‘real world’ terms, become attributes of the referent […]. […] But there is a multitude of additional, non-criterial properties that we have learnt to expect a referent woman to possess. They […] may extend to features which are merely typical rather than invariable concomitants of woman-hood. (Leech 1981: 12, emph. by auth.)

The boundary between conceptual and connotative meaning, Leech maintains, “[…] is coincident with that nebulous but crucial distinction […] between ‘language’ and the ‘real world’.” (Leech 1981: 13) The deplorable nebulosity may, however, be suspected to arise, at least partly, from Leech’s decidedly structuralist view of language and semantics: trying to explicate meaning as a user-independent (cp. Leech 1981: 22) and use-independent system of structures (a justifiable theoretical decision), creative and intentional language use poses insurmountable theoretical problems. The “nebulous but crucial” distinction, thus, is probably rather that of “system” and “use.” Another complication has to do with polysemous lexical items, such as “bank1/bank2” (cp. Lyons 1981: 146). Neither Leech nor Lyons can account for how, in the process of understanding a particular piece of text, one of the word-meanings is favoured above the other. Leech goes on to talk about “Social and Affective Meaning”, “Reflected and Collocative Meaning”, “Associative” and “Thematic Meaning.” His approach frequently suggests that these are all properties of a lexeme. This isolated perspective on single lexemes entails incorporating into a lexeme its interdependence not only with sentence - but also with utterance -meaning, which might prove problematic. But there is another point of criticism. Jackendoff would much disagree with Leech’s and Lyon’s use of the terms ‘real world’or ‘outside world’:

What the information [that language conveys] is about – the reference of linguistic expressions – is not the real world. The referring expressions of natural language will be just those expressions that map into projectable expressions of conceptual structure. (Jackendoff 1983: 36, emph. by auth.)

Language, Jackendoff tells us, is not about “entities, properties, situations, relations, etc., in the outside world” – those exist only as internal representations of real-world experience each cognitive system has developed by itself and on the basis of unspecific real world stimuli. The observable isomorphisms in these representations, he goes on, are due to the fact that

[…] part of one’s genetic inheritance as a human being is a set of processes for constructing a projected world, and these processes are either largely independent of environmental input or else dependent on kinds of environmental input that a human being cannot help encountering. (Jackendoff 1983: 30)

This view elegantly preserves the possibility that abstract items such as “liberty” can indeed refer to something, without having to appeal to the Platonic concept of “ideas” floating somewhere in outer space (cp. Jackendoff 1983: 34): “In ^[this] view, the ontological presuppositions of natural language […] are linked to the nature of projected reality and thus to the structure that human beings impose on the world.” (Jackendoff 1983: 36, emph. by auth.) In order to ensure “semantic competence,” (cp. Leech 1981: 9) it seems reasonable that these distinct projected worlds are interpersonally negotiated and tuned, so that users share at least a number of “images”. Still, this does not resolve the question about polysemous lexical items, for the decision whether a particular “bank” is something you can sit on, something you can swim towards, or something you can bring your money to. Sperber/ Wilson suggest, that this is done on the basis of a situative computation of relevance:

The hearer treats the linguistically encoded word meaning […] as no more than a clue to the speaker’s meaning. Guided by his expectations of relevance, and using contextual assumptions made accessible by the encyclopedic entry of the linguistically encoded concept […], he starts deriving cognitive effects. When he has enough effects to satisfy his expectations of relevance, he stops.

Word-meaning, in this view, is inextricably linked with utterance-meaning, as the result of a process of explicating “[…] via decoding, disambiguation, reference resolution, and other pragmatic enrichment processes.” (Sperber/Wilson 2003) This can be taken to account for differences in “meaning”, arising from Leech’s peripheral meaning-categories, as well: they turn into differences in explicature which can be handled and explained by Relevance Theory. In the relevance-theoretic view, the rather static notion of projected world is thus modified into one that is, to a considerable extent, in flux: a set of contextual assumptions accessible for computing the greatest possible contextual effects. The projected world is dynamically changing shape with the contextual requirements under the dictate of the search for relevance.

The basic idea of componential analysis – “the view that senses of all lexemes in all languages are complexes of universal atomic concepts” (Lyons 1981: 154) – might well be one of those instances of Jackendoff’s “imposed structure”– but in this case, it is imposed by the linguist, and on the phenomenon “language” itself^[v]. Snell-Hornby derives from this a method of analysing a special kind of verbs:

There are verbs in German and English which themselves modify the action they express. […] They […] contain a […] semantic element that assumes the function of, and is expressed in the verb’s definition by one or more adjectives or manner adverbs. (Snell-Hornby 1983: 25)

[...]

^[i] „[...] l’idee vient en parlant. [...] Ein solches Reden ist wahrhaft lautes Denken. Die Reihen der Vorstellungen und ihre Bezeichnungen gehen nebeneinander fort [...]. Die Sprache ist alsdann keine Fessel, etwa wie ein Hemmschuh an dem Rade des Geistes, sondern ihm wie ein zweites mit ihm parallel fortlaufendes, Rad an seiner Achse.“ (von Kleist: 1805)

^[ii] we are aware that this technical language is a meta-communicative scientific construct; its purpose is to pin down “meaning” as effectively as possible. It should certainly not be seen as the “proper” or “standard” way to deal with word meaning in actual communication – it is just the scientific way, and thus one quite narrowly tied to the written word;

^[iii] Lyons (1977) claims that “[componential analysis] neither presupposes field-theory nor is it supposed by it.” (Lyons 1977: 326) Snell-Hornby, however, uses the term “semantic field” along with am method of componential analysis;

^[iv] the problematic question, whether these features must be definable “by relating them directly to entities outside language” – whether they are “atomic” and “basic” – or whether it suffices to view them as mere “algebraic factors” without any denotational meaning is not conclusively answered in Lyons (1977). He expresses the opinion that this question is not a crucial one and that, at any rate, “the factors would derive their linguistic significance from the fact that each of them enables the linguist to account for the acceptability […] of a set of sentences.” (Lyons 1977: 329)

^[v] if we accept that the forces yielding the individual representation of the world are analogous to those behind its scientific representation;