[Home] [Inhalt / Table of contents / Table des matières]

 [PDF]

Killing Time: Metaphors and their implications in lexicon and grammar

Fernando Balbachan, Indiana (fbalbach@indiana.edu)

desde la tierna cuna a la tumba enlutada

 from the tender crib to the mourning grave

(Francisco de Quevedo y Villegas, Metaphysical Poems)

Abstract

Metaphors have recently acquired a fairly vindicated relevance in the field of linguistics, not as mere figures of speech but as important semantic processes pertinent to both psycholinguistic and lexicology. Metaphors in the broad sense may be viewed as a means of categorizing our world, through broad-range phenomena such as metaphors in the strict sense, metonymy, and the so called “metaphtonymy” (Goossens 1990). From the point of view of lexical productive processes in word and phrase formations, their role as meaning extension in the lexicon is incontrovertible. The latter raises the question about the very status of metaphors inside the lexicon. In order to discuss that issue, this paper will closely analyze from a cross-linguistic point of view the interaction of metaphors with morphology and syntax, illustrating the discussion with Dutch examples (Dirven 1985). In particular, we will argue around the problematic case of the metaphorical use of Spanish “matar el tiempo” (lit. killing time). It is intended to demonstrate how this expression, even implying both a selectional constraint violation and a syntactic anomaly (the absence of the preposition ‘a’), is strictly a “dead” metaphor and involves a lexicalization entry at the level of the lexicon. Therefore, in spite of our initial intuition, we will claim that there is no such interaction with respect to syntax level in this case.

In der jüngeren kognitiven Linguistik wurde die große Relevanz der Metapher herausgestellt, die nicht mehr nur als reine Stilfigur betrachtet sondern als ein wichtiger semantischer Prozess z.B. in psycholinguistischen und lexikologischen Fragestellungen erörtert wird. Metaphern im weitesten Sinne dienen dazu die uns umgebende Welt zu kategorisieren; unter so verstandener Metaphorik subsumiert man Metaphern im engeren Sinne, Metonymien und sog. „Metaphtonymien“ (Goossens 1990). Die Rolle der Metapher als ein produktives Verfahren zur Bedeutungserweiterung von Lexemen ist unumstritten. Es stellt sich aber die Frage nach dem genauen Status der Metapher im Lexikon. Der vorliegende Aufsatz hat daher zum Ziel u.a. anhand niederländischer Beispiele (Dirven 1985) die übereinzelsprachlich festzustellende Interaktion zwischen Metaphern, Morphologie und Syntax analysierend darzustellen. Besonders diskutiert wird der spanische metaphorische Ausdruck „matar el tiempo“ (wörtl. „die Zeit töten“). Es soll gezeigt werden, dass dieser Ausdruck eine „tote Metapher“ und somit das Ergebnis eines Lexikalisierungsprozesses ist, sichtbar an einer Selektionsrestriktionsverletzung und einer syntaktischen Anomalie (Nicht-Setzung der Präposition „a“). Das Ergebnis der Analyse führt zur Verwerfung der anfänglich aufgestellten Hypothese einer bestehenden Interaktion zwischen dieser konkreten Metapher und der Syntax.

1.      Introduction

All through many centuries of a traditional approach, metaphors were considered a device of the poetic imagination and the rhetorical flourish – a matter of extraordinary rather than ordinary language (Lakoff & Johnson 1980). Carrying on this heavy burden from the Greek tropoi on, the very role of metaphors in categorizing our world and their vital importance as underlying semantic processes in everyday language were surprisingly neglected until the appearance of Cognitive Linguistics. We must concede that prior to Cognitive Linguistics there were some others approaches (Cassirer 1923, Jakobson 1960); but they failed to narrow down the object of study around the specifically linguistic phenomena. That is not the case, though, of other proposals such as Systemic Functional Grammar’s notion of “ideational grammatical metaphor” (Halliday & Matthiessen 1999) and Dirven’s comprehensive lexicological study on the metaphor (Dirven 1985, Dirven 2002).

Therefore, this paper is intended to vindicate the immanently linguistic importance of metaphorical processes, taken in the broad sense, in the meaning extension within the lexicon as well as their interaction with the levels of linguistic structure (morphology and specially syntax). We shall begin by explaining the basic mapping process inherent to metaphors according to Cognitive Linguistics (section 2). Then we will introduce a basic insight to classify metaphorical processes (section 3) like metaphor in the strict sense, metonymy or even metaphtonymy (Goossens 1990). In section 4 we will talk about the ubiquity of metaphor in ordinary language from a lexicological point of view, its role in productivity, creativity and semantic change. Finally, in section 5 we will discuss the implications of metaphors regarding access to lexicon and grammar interaction. We will illustrate the situation with two interesting examples from Dutch and Spanish to promote an exhaustive analysis on how metaphors are supposed to be represented within the lexicon, how they interact with syntax and what place they are entitled to take in lexicology.

2.      Not figuratively speaking, what is a metaphor?

2.1       The cognitive model

Following Lakoff (Lakoff & Johnson 1980), we can state that “our conceptual system, in terms of which we both think and act, is fundamentally metaphorical in nature […]. Since communication is based on the same conceptual system, language is an important source of evidence for what that system is like” (Lakoff 1980, p.3). Metaphor plays an important role indeed in our ordinary language. Cognitive Linguistics’ foundations are based on the previous assumption. Basically, a metaphor (from Greek metaforein, ‘to transfer’) is viewed as an experientially-based mapping from an ICM (Idealized Cognitive Model) in one domain onto an ICM of another domain (Lakoff 1982):

“[…] metaphoric mapping involves a source domain and a target domain […]. The mapping is typically partial. It maps the structure in the source domain onto a corresponding structure in the target domain”(Lakoff 1987, p.288).

Figure 1: The mapping process in metaphor

The ICM accounts for a complex structured whole, a gestalt, which involves four structuring principles:

·        Propositional structure (cf. ‘frames’ in Fillmore 1982)

·        Image-schematic structure (Langacker 1986)

·        Metaphoric mappings

·        Metonymic mappings

For the sake of simplicity we can consider ICMs as systematic and idealized cognitive models that fit one’s understanding of world for a myriad of human activities and facts of experience and that are used in forming conceptual categories and reasoning. For the purposes of what follows we simply posit the existence of discrete cognitive domains (Langacker 1986 and Goossens 1990), articulated throughout language, by which we categorize our experience. Moreover, it is the very idea of mapping onto different domains that essentially distinguishes metaphor from metonymy (see section 0 : 3.1   Basic distinction between metaphor and metonymy ).

2.2       The linguistic vehicle

The canonical view of the linguistic support for a metaphor (Richards 1936 and Leech 1969) may be ultimately considered an instance that the cognitive mapping process adopts inside language, grounding the categories of two different domains, as follows:

Figure 2: Canonical linguistic form for metaphors

The cognitive mapping is always a partial transposition of categories from a more concrete domain onto a less concrete one in order to categorize facts of experience in terms of features of already familiar experience. Then, the mapping process will necessarily hide aspects of the ground that are inconsistent with the current metaphor.

For example, a common metaphor in English is ARGUMENT IS WAR. This metaphor may be tested in many occurrences in English (eventually also in other languages):

Your claims are indefensible

He attacked every weak point in my argument

His criticisms were right on target

I demolished his argument

I’ve never won an argument with him

You disagree? Okay, shoot!

If you use that strategy, he’ll wipe you out

He shot down all of my arguments

The latter can be verified even in the case of the least frequent occurrences within the semantic field of war:

“impregnable, adj. 1. strong enough to resist or withstand attack; not to be taken by force, unconquerable: an impregnable fort. 2. not to be overcome or overthrown: an impregnable argument [1400-50; late ME]” (source: Webster’s RH).

It is not the case that arguments are subspecies of war. The language by which we talk about arguments is not poetic, fanciful, or rhetorical; it is literal. We talk about arguments that way because we conceive of them that way.

However, as we stated before the mapping is partial, since arguing is also a cooperative activity: someone who is arguing with you can be viewed as giving you his time in an effort at mutual understanding. Therefore, these cooperative aspects are to be hidden by the mapping process, highlighting the battle ground.

3.      Metaphorical processes: metaphor, metonymy and metaphtonymy

3.1       Basic distinction between metaphor and metonymy

Starting out with the Aristotelian poetic, there has been a whole and traditional general agreement on both the tight connection between metaphor and metonymy and, at the same time, the basic distinction between them. It is commonly said that while metonymies are based on a relationship of contiguity between the concept (X) and the vehicle (Y) (cf. Figure 2 ), elements of metaphor come from different conceptual fields. Thus, many theories have been stating this basic distinction by means of more or less the same terms (Ullmann 1962, Leech 1969, Halliday 1985), opposing contiguity (in the case of metonymy) to resemblance or similarity (in the case of metaphor). But from our cognitive treatment of the matter, we must point out the crucial role that the notion of domains plays in both conceptual processes:

“…metaphoric mapping involves a source domain and a target domain from two different discrete domains, while a metonymic mapping occurs within a single conceptual domain which is structured by an ICM” (Lakoff 1987, p.288).

Nevertheless, as we shall demonstrate in the following sections, the boundary lines between domains are often fuzzy, causing metaphor and metonymy to merge into a striking phenomenon coined “metaphtonymy”. In what follows we will also go over two different treatments of the metaphorical processes: Dirven’s (1985 & 2002) and Lipka’s (1990a & 1990b).

3.2       Metaphor in the broad sense or in the strict sense?

Dirven (1985) identifies three major metaphorical processes, all of them sharing a common feature: they are all associative processes which eliminate or cancel the first or literal interpretation, so that another figurative interpretation must be looked for. On the contrary, in other instances of two interpretations such as ambiguity, paradox, oxymoron, etc. both interpretations can be valid. Dirven provides the broadest treatment for the metaphorical processes, including: metonymy (with the subcase of synecdoche), metaphor (properly speaking) and synaesthesia (figure 3).

Figure 3: Dirven’s (1985) conception of metaphorical processes:

In turn, he follows Ullmann’s model (Ullmann 1962) for the linguistic structure in order to demonstrate how metaphors spread through all levels of language, which he calls “the ubiquity of metaphor”. That model consists of four formal levels and two omnipresent elements (morphology and semantics) as “part and parcel [in greater or lesser proportion for each] of these four levels of structure” (Dirven 1985, p.88. The italics are ours].

Figure 4: Ubiquity of metaphor all throughout the linguistic structure

However, the so called ubiquity of metaphor seems not to withstand after a rigorous analysis. On one hand, postulating sequences of phonemes carrying meaning looks doubtful. We might be willing to concede a relationship of metaphorical onomatopoeia between the sounds and the concept of ‘curved, fast motion’; but even so, what linguistic status should this phenomenon receive other than a conventionally fixed plain archimorpheme? On the other hand, there seems not to be much difference between the so called discourse metaphor and allegory (which Dirven himself had already discarded as non metaphorical process). Finally, sentence metaphors are not relevant in terms of productivity. They belong to what Lyons (1977) named creativity: they are stylistic devices that produce contextual meaning, interpretable by Grice’s maxims (Grice 1975) with no possibility of lexicalization. All this criticism leads us to focus on word metaphors and phrase metaphors as evidence pertinent to lexicology.

More recently Dirven (2002) has slightly changed his mind regarding the prevalence of metaphor as the ubiquitous conceptual process and its foundations on the mere concept of similarity:

“Precisely in this respect we find one of the strongest differences between metonymy and metaphor. In a sense, metonymy is a much more ubiquitous process than metaphor. In order to become operative, the strategy of metaphor needs the possibility of seeing contrast(s), which is needed as much as the presence of conceived similarity or similarities. Since this is not always and not immediately given, metaphor requires far more conditions to be fulfilled in order to become operative than metonymy. Metonymy is extremely active everywhere and all of the time” (Dirven 2002, p.339).

As a matter of fact, Dirven’s hesitation on the prevalence of either metaphor or metonymy as the central conceptual process could become a good introduction to the problem exposed in the section 3.3 where we will develop a new approach to this issue. Let this brief preview suffice for now.

Coming up with a new turn on the topic, Lipka (1990a and 1990b) also contributes to the distinction between metaphor and metonymy. Although more interested in both devices of semantic transfer from a lexicological perspective about productivity, he tries to shed some light upon the discussion. Lipka invokes the neurolinguistic criterion of Tournier (1985) in the tradition of Roman Jakobson, regarding different psychological realities for metaphor and metonymy, as proven from aphasic disorders (Stachowiak 1985). However, as this approach to the phenomena seems not to be fully contributive to our lexicological purposes, in what follows we will keep on handling Dirven’s exhaustive classification of the metaphorical processes and we will return to Lipka in section 4 when we will face the productive processes at the level of the lexicon.

3.3       Metaphtonymy: interaction metaphor-metonymy

So far we have emphasized the differences between metaphor and metonymy. Going further in our goal of exploring metaphorical phenomena, we will now introduce a very interesting position that demonstrates how these two cognitive processes are not mutually exclusive. Louis Goossens (1990) explored the interaction between metaphor and metonymy – coining the neologism metaphtonymy – in conventionalized expressions where linguistic action is the target domain. Although his corpus is mainly based on British occurrences we cannot but point out the striking accuracy of his conclusions beyond dialectical constraints.

Goossens begins by identifying three donor domains for his corpus of word and phrase metaphors of linguistic interaction:

1) Sound: divided in turn into human sound, animal sound, natural sound, artificial sound by means of an instrument and artificial sound by other means

2) Body parts: body parts fit into a more complex domain or scene which has to be processed with reference to linguistic action in its own right

3) Physical violent action

Case 0 in table 1 shows the basic typical occurrence for a purely metaphorical expression. The remaining cases illustrate the phenomenon of metaphtonymy. Goossens argues that in those cases where the sound hangs together with a human activity that can naturally co-occur with linguistic action, we are in presence of a metaphor from metonymy case. For example, in giggle even if we interpret the utterance as “she said that as if giggling” – what assumes a metaphorical-prevalence reading – we will still be able to detect the conceptual link with the metonymic reading:

“For all them it is possible to use them metonymically, that is with reference to a scene where both the non-linguistic and the linguistic action reading are relevant, and it is that metonymic reading which is the basis for the metaphorical use” (Goossens 1990, p.332).

“The main point here is that underlying the metaphor there is an awareness that the donor domain and the target domain can be joined together naturally in one complex scene, in which case they produce a metonymy, of course. The actual contexts into which these items fit will be decisive for the interpretation as either a metonymy or a metaphor from metonymy” (Goossens 1990, p.336).

That is also the case for “beat one’s breast”, where the metonymic basis is the religious practice of beating one’s breast while one publicly confesses one’s own sins.

On the other hand, Goossens also posits another pattern of metaphtonymy: metonymy within metaphor, when a metonymically used entity is embedded in a complex metaphorical expression making the metonymy function within the target domain (cases 4 and 5). For example, in case 4 tongue can be processed literally in the donor scene. Mapping this onto the linguistic action we get something like “depriving oneself of one’s ability to speak”, where the metonymy between tongue and ‘the speech faculty’ remains regardless the metaphorical mapping. That explains why tongue is a better donor element rather than bite one’s finger to map self-punishment onto the target domain of linguistic interaction.

Finally, Goossens proposes two more patterns for metaphtonymy: metaphor within metonymy and demetonymization in a metaphorical context. However, these two patterns are exceptionally rare cases and their implications for lexicological purposes, unlike the previous two patterns of metaphtonymy, seem not to be relevant.

4. Productive processes at the level of the lexicon

4.1 Meaning extensions

The metaphorical processes operate at the level of lexicon as means for extending the meanings of existing lexemes. Dirven considers them boosters of semantic transfer:

“The three processes can be distinguished by a number of criteria, one of which is that both metonymy and synaesthesia apply the transfer process starting from a derived meaning of a given lexical item, whereas with metaphor the transfer often applies to the basic meaning of it” (Dirven 1985, p.114).

The latter is exactly the case for the meaning extensions of the word cup (see figure 5).

Figure 5: Example of meaning extensions as in Dirven (1985)

After having analyzed the distribution pattern of meaning extensions among the three metaphorical processes, one may wonder why metonymy has been less productive than metaphor. A possible explanation may be found in the nature of the concept denoted by cup: the feature ‘related to drinking’ seems to be transferable to fewer domains, whereas the more general features of cup such as its shape or function or other features allow a transference to numerous domains. More generally, it can be stated that the specific metaphorical process that is selected in the extension of the meanings of a given lexical item is, to a large extent, determined by the features of the lexical item in question.

4.2       Productivity and lexicalization

Following Leech’s (Leech 1981) concept of lexical rule, Lipka (1990a&b) approaches the relationship between productivity and metaphorical processes at the level of the lexicon from a different point of view. He states that metaphor and metonymy are both borderline cases between derivation and purely semantic changes. Thus, they play an important role as highly productive devices of semantic transfer, characterized by:

·                    productivity: Although they may appear as basically unrestricted specific-language-based processes, there seem to be some patterns of productivity. For example, metaphors of personification, animation and anthropomorphic transfers spread all over natural languages.

·                    degree of acceptability: In this respect Lipka makes two remarks: It is certainly true that there are degrees of acceptability, since salient attributes, responsible for the ground in metaphor, may at times be difficult to find. And as a logical consequence of the first statement, he also remarks that the degree of acceptability is correlated with our encyclopedic knowledge as users of a language. That explains why it would probably be fairly difficult to find metaphors grounded, let us say, on the mating behavior of Tyrannosaurus Rex

·                    possibility of institutionalization (or lexicalization): Metaphors and metonymies are tied to the norm of individual natural languages, in terms of underlying interaction with the linguistic structure. Thus, while it is common that a metaphorical occurrence be lexicalized from a diachronic point of view, within this very process of lexicalization some interaction between morphology and semantics may certainly be verified. This remarkable feature of tight interaction inside the linguistic structure leads us to our next main section and to our most important forum of discussion: The implications of metaphors in lexicon and grammar.

Lipka’s concept of productivity for meaning extensions is partially embedded in a diachronic perspective. Basically, Lipka identifies two typical processes where metaphors and metonymy take place, showing a general schema as a lexical rule for semantic shift or transfer: radial shift and chaining shift:

Figure 6: Radial shift

Figure 7: Chaining shift

5.      Metaphors and their implications in lexicon and grammar

5.1      The interaction with morphology and syntax

What follows from Dirven’s and Lipka’s vision of these metaphorical processes is that they are not isolated from the rest of the rules of a language, but on the contrary, they seem to be interwoven with these rules in a very intricate way. Hence, we will first explore the interaction of metaphors with respect to morphology and syntax adopting a cross-linguistic perspective. Then, we will analyze a problematic example that will raise the issue of how metaphors are encoded within lexicon.

It is common to many languages that diminutive morphemes carry additional meanings: diminutive (in the strict sense) but also emotional (affective or derogatory) and intensifying meanings. On the other hand, the suffixation of nouns bearing a diminutive particle may boost one or another meaning depending on the meaning of the noun they are being added to. For example, in Dutch the diminutive is built by the suffix –je. However, the metonymical extensions for ‘hart-je’ (little heart) (i.e. ‘courage’, ‘tenderness’ and ‘love’) can only have the emotional meaning (Dirven 1985):

Met een klein hartje = (with little courage)

But more interesting is the fact that the metaphorical locative reading (‘inmost and central part’) can be rendered only by the diminutive form with intensifying interpretation:

In het hartje van de stad

In the (very) heart of the city

*In het hart van de stad

The latter basically demonstrates that the metaphorisation of ‘heart’ through meaning extension triggers the selection of the diminutive form as a grammatical rule and not as a stylistic choice of the speaker.

Staying with Dutch, Dirven provides us another example of the interaction between metaphor and linguistic structure. In the following case, we will explore the interaction with syntax, taking the verb kruipen (to climb). If the subject is animate, the verb accepts an optional prepositional particle (op) either in the front or the rear part of the verbal phrase:

Hij kroop op de berg / Hij kroop de berg op

He climbed up the mountain

However, if the subject is inanimate, although metaphorically personified, the verb requires the prepositional particle (op) always to follow the direct object noun phrase:

De trein kroop de berg op

The train climbed up the mountain

* De trein kroop op de berg

One possible explanation is that the metaphorical context causes the verb to be seen as detached from any means used for climbing (e.g. human legs) and as exclusively limited to movement. In that sense a greater emphasis upon the direction of the movement is achieved by the construction V+DirectObject+particle. In short, what is remarkable in both Dutch cases is how the metaphorical process interacts with morphology as well as with syntactic rules.

Another very interesting example of interaction between the linguistic structure and metaphors is Rio de la Plata Spanish metaphor of “estirar la pata” (to stretch the leg), meaning to die by virtue of metonymy with respect to the perception of rigor mortis. The remarkable phenomenon here is how appealing to the metaphorical meaning of “estirar” is ostensively violating restrictional selections for the feature [+human] in the subject, when selecting “pata” [one animal leg] instead of “piernas” [the two human legs] for the productive metaphorical process:

Spanish ->  Todos                               vamos a estirar   la            pata                                        algún     día

Gloss ->     Everybody[+human]      going to                stretch   the          leg [-plural –human]           some      day

English ->   Everybody is going to die some day

Unlike the previous example:

Spanish ->   Todos                              vamos a estirar las     piernas                        después de un largo vuelo

Gloss ->     Everybody[+human]      going to                stretch  the     legs [+ plural, +human] after          a    long  flight

English ->   Everybody is going to stretch the legs after a long flight

This ostensive violation of semantic features is an uncontroversial proof of how the metaphorical meaning is encoded as a new entry in the lexicon with its own semantic features, separated from the literal meaning of estirar verb.

5.2       A problematic case: killing time

It is time for us to get the final destination of this current paper. So far we have examined metaphors and have been handling evidence of their importance not only for a cognitive model of linguistics but also for the structure of language itself. In addition, we have reviewed their importance for lexicological productivity. Now, we must direct our efforts towards the core – metaphorically speaking – from where all these phenomena emerge: the lexicon. In order to do so, we will analyze one problematic case of metaphorical personification in Spanish, which will shed some light upon relevant issues such as metaphorisation, lexicalization and access to lexicon.

We have found a case similar to the Dutch verb kruipen with respect to interaction between syntax and metaphor, although not as straightforward as the case of “estirar la pata”. Let us start analyzing the metaphor matar el tiempo (lit. killing time) by clarifying some basic syntactic rules of Spanish:

In Spanish the use of the preposition ‘a’ preceding animate direct object noun phrases is mandatory:

This occurrence poses a problem for our view of interaction syntax-metaphor, as the linguistic form of the verbal expression does not observe the mandatory use of ‘a’ for animate DO NPs. In other words, how this metaphor of personification is encoded within the lexicon, so that it does not show in surface the same structure as any other occurrence of the DO for the verb matar (prepositional phrase with ‘a’ preceding a categorial selection of animate noun phrase). Is it just a whole unit, a mere lexicalization, or is the metaphorical process still productive enough to interact with syntax and morphology? All these issues will be addressed in the following sections.

5.2.1    A metaphor of personification?

With respect to our initial cognitive model, it is evident that this latter metaphor entails an incompliant case. Time here is not conceptualized as a commodity (cf. section 0 : 2.2          The linguistic vehicle ). In The Invariance Hypothesis, Lakoff (1990) explores the so called generic-level metaphors, among which he posits the metaphors of personification. In his view, the personification schema fits a single pattern:

“Events (like death) are understood in terms of actions by some agent. It is the agent that is personified” (Lakoff 1990, p.68).

Thus, for example in the DEATH IS DEPARTURE metaphor:

“Departure is an event. If we understand this event as an action on the part of some causal agent – someone who brings about, or helps to bring about departure – then we can account for figures like drivers, coachmen, footmen, etc,” (Lakoff 1990, p.69).

However, this simple formula lacks specific mapping details. Why not understand death in terms of teaching? Therefore, Lakoff constrains his generic-level hypothesis that EVENTS ARE ACTIONS: the action must have the same overall event-shape as the event. What is preserved across the mapping is the causal structure, the aspectual structure, and the persistence of entities.

Although this explanation looks accurate enough for many common metaphors (DEATH AS REAPING PEOPLE LIKE PLANTS, and others), it certainly does not fit in our problematic Spanish example. What is personified there is not the agent of the action but the patient, and, on the other hand, what kind of event does the metaphor deal with, so that its causal structure may be projected towards the target domain of the action ‘to kill’? As a matter of fact the very meaning of the expression in Spanish lies in the complete absence of events to metaphorize: this weird idea of «beating the running of time by killing it, but with an expectable event to come afterwards in mind, rather than passing the immortal desire of wholly stopping it».

With all our recently acquired background on the topic, we may offer an explanation for the semantic selection. If we think of death as the natural consequence of the passing of time upon the human condition, it is not surprisingly striking considering that ‘time kills people’ is the prototypical framework – cf. the metaphysical categorization of time in neostoicism and other philosophical schools, also represented in Quevedo’s epigraph for this current paper – as a mapped action from the event DEATH. Therefore, the reciprocal of the metaphor would result in a sort of also metaphorical sweet revenge on the part of the human race against its mortal condition. Nevertheless, this hypothesis would not be compliant with the contradictory fact that the intended effect in killing time is not to refrain but rather to booster the passing of time.

5.2.2    An ideational grammatical metaphor?

Much more promising for the approach to this example, the perspective of the Systemic Functional Grammar, evolved from the foundations of Halliday’s work, basically claims that language has three main uses or functions: the interpersonal (to linguistically interact with other persons), the ideational (to linguistically encode our conceptualization of the world), and the textual one (to organize significant units of new and already known information into large discourse units). All these functions simultaneously apply to the clause as unit of analysis.

It is clear that it is the ideational function where metaphorical processes may take place, specifically in its transitive system. Nevertheless, Steiner (2003) also posits interpersonal metaphors in addition to the ideational ones, although he denies the existence of textual metaphors. This is an interesting counter-argument to Dirven’s and Ullmann’s ubiquity of metaphorical processes all along the linguistic structure (see section 3.2: Metaphor in the broad sense or in the strict sense?). As a matter of fact, Halliday explicitly defines an ideational grammatical metaphor as that in which some unit of experiential (ideational) or logical meaning receives different encodings. This definition posits the ideational grammatical metaphor in the middle of the semogenic resources such as rankshift and transcategorization (Halliday & Matthiessen 1999). Steiner provides us with the following characterization of the direct encoding (also known as congruent variant), the logogenetical base -in the sense of related to lexical productive processes in word or phrase formations- upon which metaphors operate:

“Within the Systemic Functional Grammar, a direct way of encoding would involve the one-to-one mapping of, for instance, a doing figure onto a material process, a sensing figure onto a mental process, a saying figure onto a verbal process, a being and having figure onto a relational process, etc.” [Steiner 2003 pp.142-143; However, it is remarkable how language-dependent this concept of directness or congruence is, resulting in several degrees of metaphoricity across languages].

 “Rankshift preserves the categorial status of a constituent, while changing its function, whereas transcategorization represents a complete moving over of some category into another, say a verb into a noun, without rankshift necessarily associated with it. [Ideational] Grammatical metaphor is the highly interesting middle case, in which a tension remains between the grammar and the semantics of a construction, in the sense that the grammatical realization preserves features both of its congruent and of its metaphorical variant” (Steiner 2003 pp.140-141).

In order to illustrate the above mentioned definition of ideational grammatical metaphor let us analyze the following example

The sudden closing of the door was followed by complete darkness

The process of closing is grammatically encoded as a nominalized construction, which, in addition to acquiring nominal properties, preserves several verbal features (patterns of complementation, potential morphology, etc.). Furthermore, the verb follow does not pre-select a process, so that closing is a process treated as a thing – and in this sense there is this tension or juncture between different categorical meanings. In other words, the direct encoding of the clause according to the transitive system

(Somebody) closed the door and then (everything) got dark

yields to the metaphorical encoding, where the agent (somebody) and the patient role are missing inside the entire encoding of the process into an event.

But all this approach yet provides us with a new fresh perspective on our previous problematic Spanish example:

Firstly, if we are positing the ideational grammatical metaphor as a more likely solution, we should find out the direct encoding clause from which the metaphorical process took place, as well as the intercategorial mapping (in terms of the ideational function) between the congruent clause and the metaphorised one. In that sense, we may want to appeal to the ergative system as opposed to the transitive system within the ideational function of language. Hence, within the ergative system we deal with broadly-used English middle clauses

Two mechanism of survival have evolved in plants

that combine certain non-realized categories (no potential of direct object, impossibility of adding purpose encoding features like purpose phrases) with a non-agentive subject. Also ergative, it is frequent to find the following structure:

The sergeant trained the soldiers to kill

where we identify an instigator (sergeant) that makes an agent to perform the action, so that soldiers is patient and agent at the same time. That is the reason why the ergative system is tightly connected with the concept of external causality of a process.

One probable first attempt to identify the basic clause is interpreting the meaning of the Spanish metaphor with regards to the ergative system:

This interpretation allows us to identify the metaphorical process by moving on from the transitive system towards the ergative system, undoing the path of the metaphorisation. The original agent (he) is encoded as an instigator, and the original patient (time) is actually the agent/patient within the ergative system direct clause. Moreover, the fact that certain grammatical constraints remain after performing the mapping, typical of the ideational grammatical metaphor (cf. above the example ‘the sudden closing of the door’), may account for the absence of the preposition ‘a’ in the metaphor (our original problem). Since the original role for time was agent, once mapped into the transitive structure of matar (to kill), the grammatical constraint, still carried on, would make it reluctant to adopt the place of a typical animated patient which would certainly require the preposition ‘a’.

However, all the benefits of the present analysis seem to vanish when contrasted with one main flaw. This theory still does not explain the very reason in having chosen matar (to kill) –logogenetically speaking – instead of any other verb to metaphorically encode this idea of ‘making time go faster’.

5.2.3    Is it a metaphor at all or just a lexicalization?

So far we have been exploring the implications of metaphors in grammar, but there is a whole fertile perspective to explode: the psycholinguistic approach to the phenomenon, i.e. the encoding of metaphors inside lexicon and the access to them while processing language. The psycholinguistic turn is pertinent to our lexicological main aim, as it may provide us with more solid and empirical arguments to hypothesize the status of metaphors either as units of more complex processing or lexicalized units like idioms.

In his excellent work, Blank (1988) introduces this conjunction approach between psycholinguistics and lexicology by explaining the two traditional ways regarding how metaphors were supposed to be encoded inside lexicon:

“One strategy is idiom addition, where the metaphorical sense is simply added to the lexicon as a new entry. Another strategy is abstraction, where a literal sense is weakened so that it covers either the literal or metaphorical sense. The problem with these approaches […] is that they don’t account for systematicity of families of familiar metaphors” (Blank 1988, pp.21-22).

These traditional approaches are not completely wrong. Blank distinguishes lexicalized or familiar metaphors – they are said to become “frozen” or “dead” – as those that receive predictable interpretations, wholly from the local context of predication. In those cases, the traditional strategies may yet apply. But, however, how to account for more complex metaphors, where selectional constraints are clearly violated, without falling into a doubtful hypothesis of an endless list of new entries in the lexicon, one for each non familiar metaphorical occurrence? Thus, Blank claims a rules–based failure-driven stage model of processing, as follows:

1. There is more than one processing strategy involved. Lexicalization may be local, by adding or modifying an individual entry (idiom addition or abstraction), or it may be global, by learning rules that apply to the lexicon as a whole (sense or meaning extension, as in section 4.2: Productivity and lexicalization). Notice the recursion of this idea of lexical rule tightly related to metaphors (Leech 1981).

2. The processes are staged. First look it up in the lexicon, then try to apply a sense extension rule. Local lexicalization should render a familiar metaphor indistinguishable from literal language with respect to access, but checking global rules implies an extra step (and extra time).

3. Both stages are automatic. The second stage is failure-driven and obligatory (people cannot volitionally ignore the figurative sense).

4. Rather than waiting until completion of a sentence, transfer of control may be triggered at the point of the failed process (e.g. a selectional constraint violation). This immediacy strategy would try to assign a semantic interpretation as soon as possible.

The outcome that Blank offers after his naming-task-access experiment is incontrovertible:

“The data strongly suggest that we can distinguish two kinds of familiar metaphors. Very familiar metaphors such as those that follow the TIME IS MONEY theme are accessed in virtually the same as literal senses, suggesting that they are stored the same way. Fairly familiar metaphors such as those that follow the FERTILE IDEAS theme take slightly longer (on order of 30 msec) to access. Thus the hypothesis appears to be confirmed – there are two stages of processing. The first stage is likely an exhaustive search for fully lexicalized senses. The very familiar metaphors are here processed right alongside literal senses. If the first stage fails – that is, none of these senses meets the predicative constraints of the sentence – then a second-stage search for extended senses begins.” (Blank 1988 pp.29-30).

How to apply all this wisdom onto our problematic case? Apparently the distinction between lexicalized metaphors and not familiar metaphors is a promising beginning. As a matter of fact, we may propose two tests to check the encoding status of the metaphor killing time in Spanish:

These tests suggest the plausibility of killing time as a lexicalized metaphor in Spanish, rather than a form built by virtue of rules, at least under its synchronic representation inside lexicon. In short, the metaphorical meaning is not a constructo over the literal sense of the verb matar, but another brand new meaning entry. That would explain the absence of the preposition ‘a’, as the metaphor is considered a whole and fixed unit, lexicalized in the lexicon, as conventional as any other idiom entry – although logogenetically speaking, it might have been motivated as a particular literal occurrence of the verb matar (to kill) still carrying on the preposition. Yet, some counterarguments seem to persist, supporting the interaction between the syntactic level and the metaphorical process over the lexicon entry of matar Spanish verb. For example, it is still possible to syntactically operate right in the middle of the metaphorical construction as a meaning extension of the original entry for matar. Thus, the following is a broadly used occurrence for the participle form of matar verb as an adjective in the noun phrase:

Spanish ->     tiempo muerto

Gloss ->         time     killed

Interestingly, this occurrence wholly preserves the meaning of the original metaphor, as a period of time while you are doing nothing but wait for something else.

If the above mentioned hypothesis is true -the idea of the metaphor as a new brand entry in the lexicon-, then the access to this metaphor in Spanish should imply no extra time (and no extra stage) with respect to any other literal occurrence of the verb matar; and, on the contrary, the semantic interpretation should be assigned at the early first stage of lexicalized-senses search. In addition, that would involve no interaction at all with syntax level. On the other hand, the syntactic operation in the middle of the metaphorical meaning (“tiempo muerto”) is raising the issue towards another direction: maybe the metaphorical use of “matar el tiempo” is just an extension of the same entry matar at the lexicon, through interaction between syntax and the metaphorical process. Unfortunately, the ultimate confirmation of my hypothesis (perhaps by means of a true-false assertion access task) falls out of the scope of this current paper, although it seems to be the most convincing way of shedding definitive light upon the issue.

6.      Conclusions

In this paper we have pointed out the importance of metaphorical processes in the everyday use of language, not as mere figure of poetic language but as semantic underlying processes that determine our categorization of the world. In that sense, following Lakoff’s proposal, we have covered the process of mapping and their cognitive implications.

We have also demonstrated how intricately related these processes are. So much so, that they rely on intersected psychological realities (aphasia disorders). In turn, we have seen some occurrences of interaction between metaphors and metonymies (metaphtonymy).

In addition, we have approached the topic from a strictly lexicological perspective, accounting for the greater part of meaning extensions of lexical items that are based on these phenomena. Therefore, they must be considered as an integral part of the rules of the lexicon.

Then we have led the discussion around the interaction between metaphors and levels of the linguistic structure, adopting examples from Dutch and Spanish. In the particular case of the metaphorical use of Spanish “matar el tiempo” (killing time), we have explored several positions: the cognitive model, the systemic functional grammar and the psycholinguistic approach. Despite the need for further research on the topic, we have gathered strong evidence of the way this particular expression is encoded in the lexicon, as we have distinguished between a lexicalized metaphor and a non familiar metaphor. Instead of disregarding theories it is highly convenient to fusion them into a more explicative perspective. Hence, the coverage of the cognitive approach has accurately provided us with an interesting clue on the semantic word choice in the metaphor, whereas the systemic functional grammar has allowed us to follow its ergative underlying interpretation. And finally, the exhaustive analysis at the lexicon level has revealed the very lexicological status of the expression as a lexicalized or dead metaphor.

In summary, this paper has been intended to vindicate the immanently linguistic importance of metaphorical processes, pointing toward a promising road for further research and helping shed some valuable light upon the discussion.

References

Blank, Glenn (1988): “Metaphors in the Lexicon”, in: Metaphor and Symbolic Activity 3(1), 21-36.

Cassirer, Ernest (1923): Philosophie der symbolischen Formen: I. Die Sprache. (translation by New Haven: Yale University Press 1968).

Dirven, René (1985): “Metaphor as a basic means for extending the lexicon”, in: Wolf Parotté & René Dirven (eds.): The Ubiquity of Metaphor (Current Issues in Linguistic Theory 29). Amsterdam: Benjamins, 85-119.

Dirven, René (2002): “Structuring of word meaning III: Figurative use of language”, in: Cruse, D. Alan, Franz Hundsnurscher, Michael Job and Peter Rolf Lutzeier (eds.): Lexicology: An International Handbook on the Nature and Structure of Words and Vocabularies, vol. 1. Berlin and New York: Walter de Gruyter, 37-342.

Fillmore, Charles (1982): “Frame Semantics”, in: Linguistic Society of Korea, (ed.): Linguistics in the Morning Calm. Seoul: Hanshin, 111-138.

Gerrig, Richard & Gibbs, Raymond (1988): “Beyond the Lexicon: Creativity in Language Production”, in: Metaphor and Symbolic Activity 3(1), 1-19.

Goossens, Louis (1990): “Metaphtonymy. The interaction of metaphor and metonymy in linguistic action”, in: Cognitive Linguistics 1 (4), 323-340.

Grice, Herbert (1975): “Logic and Conversation”, in: Cole, P. and Morgan, J., (eds.): Syntax and Semantics. Vol 3: Speech acts. New York, Academic Press.

Halliday, Michel (1985): An Introduction to Functional Grammar. London, Edward Arnold.

Halliday, Michel & Matthiessen, Christian (1999): Construing Experience Through Meaning. A language-based Approach to Cognition. London: Cassell.

Jakobson, Roman (1960): “Closing Statement: Linguistics and Poetics”, in: Sebeok, Thomas A. (ed.), Style in Language. Cambridge, MA: The MIT Press, 350-377.

Lakoff, George (1982): Categories and Cognitive Models. Cognitive Science Report, n°19. Institute for Cognitive Studies, University of California, Berkeley.

Lakoff, George (1987): Women, Fire, and Dangerous Things. Chicago: The University of Chicago Press.

Lakoff, George (1990): “The Invariance Hypothesis: Is Abstract Reason Based on Image-Schemas?”, in: Cognitive Linguistics, 1 (1), 39-74.

Lakoff, George & Johnson, Mark (1980): Metaphors we live by. Chicago: The University of Chicago Press.

Langacker, Ronald (1986): Foundations of Cognitive Grammar, vol. 1. Stanford: Stanford University Press.

Leech, Geoffrey (1969): A Linguistic Guide to English Poetry, London, Longman.

Leech, Geoffrey (1981): Semantics. The Study of Meaning, 2^nd ed. Harmondsworth: Penguin.

Lipka, Leonhard (1990a): An Outline of English Lexicology. Tübingen: Niemeyer.

Lipka, Leonhard (1990b): „Metaphor and Metonymy as Productive Processes on the Level of the Lexicon”, in: W. Bahner, J.Schildt & D. Viehweger (eds.), Proceedings of the XIVth International Congress of Linguistics, Berlin: Akademie-Verlag.

Lyons, John (1977): Semantics, 2 vols., Cambridge: Cambridge University Press.

Richards, Ivor (1936): Metaphor. The Philosophy of Rhetoric. London, Oxford University Press.

Steiner, Erich (2003): “Ideational grammatical metaphor: Exploring some implications for the overall model”, in: Languages in Contrast 4 (1), 137-164.

Stachowiak, Franz (1985): “Metaphor production and comprehension in aphasia”, in: Wolf Parotté & René Dirven (eds.), The Ubiquity of Metaphor (Current Issues in Linguistic Theory 29). Amsterdam: Benjamins, 559-599.

Tournier, Jean (1985): Introduction Descriptive à la Lexicogénétique de l’Anglais Contemporain. Paris-Genève: Champion-Slatkine.

Ullmann, Stephen (1962): The Principles of Semantics. Glasglow: Jackson.

[PDF]

[Home] [Inhalt / Table of contents / Table des matières]

ISSN 1618-2006