Vandenbroeck PDF
Vandenbroeck PDF
Vandenbroeck PDF
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=duke.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.
Duke University Press is collaborating with JSTOR to digitize, preserve and extend access to Poetics Today.
http://www.jstor.org
THE LIMITS OF TRANSLATABILITY
EXEMPLIFIED BY METAPHOR TRANSLATION *
I. INTRODUCTION
Although in view of its importance and frequency in language use metaphor
indubitably constitutes a pivotal issue in translation, it has hitherto received only
random attention on the part of translation theorists. Presumably one of the
main obstacles for a theory of translation to overcome is the intuitively
subscribed and generally accepted "inadequacy of any single generalization
about the translatability of metaphor." If however we accept that there is such a
thing as a theoretical constituent level on which translation phenomena can be
dealt with, we must also accept that it is the proper task of translation theory to
make generalizations about such phenomena. To admit the inadequacy of
generalizations about the translatability of metaphor is to admit that translation
theory as a whole is an absurd undertaking, since it then should be incapable of
accounting for the translation of one of the most frequent phenomena in
language use. Even if such generalizations must necessarily "fail to do justice to
the great complexity of the factors determining the ontology of metaphors
why certain metaphors are created and others not; why a metaphor that is
strikingly effective in one language becomes peculiar or even unintelligible if
transferred unchanged into another [...]; in short, why languages are
anisomorphic metaphorically" (Dagut, 1976: 32), it may content itself with the
more modest task of laying bare some of the hidden mechanisms governing the
translation of metaphors and their theoretic degree of translatability.
In this paper it is my intention to make such specifications as seem necessary to
provide a theoretical framework in which general statements about the
translation of metaphors can be made.
*
Paper presented at Synopsis 1: "Translation Theory and Intercultural Relations," held at the
Porter Institute for Poetics and Semiotics in collaboration with the M. Bernstein Chair of Translation
Theory, Tel Aviv University, 27 March-1 April 1978.
? Poetics Today, Vol. 2:4 (1981), 73-87
74 RAYMOND VAN DEN BROECK
2. THEORIZINGABOUT METAPHORTRANSLATION
In order that a phenomenon such as metaphor may be adequately dealt with in
translation theory, it is obvious that some preliminary work will have to be done.
All empirical phenomena are subjected to the rule that, if one wants to theorize
about them, they must be properly observed and described. The assumption
underlying any acquisition of scientific, i.e., intersubjective and systematic,
knowledge of a phenomenon is that certain relationships be laid open, that a
certain regularity be discovered. This regularity, in that it is not manifested by
the phenomena themselves, must be assumed, or constructed, by the student of
the discipline, whose proper task it is to state his assumptions about the
character, the relations, the causes and functions of the phenomenon observed,
by formulating them in the form of a hypothesis.
For a systematic discussion of the implications of metaphor in translation the
theorist should have the following at his disposal:
1) a suitable, i.e., operational, definition of metaphor;
2) a specification of what can actually be meant by "transferring a metaphor
from an SL to a TL," i.e., of possible modes of translating metaphors;
3) a specification of distinct contexts in which metaphors can occur, i.e., of the
various structural relationships into which metaphors in texts can enter, and
of the particular constraints which, according to their occurrences, are
imposed on them;
4) a specification of the constraints which can be imposed on the treatment of
metaphors by translation itself as a rule-governed activity, i.e., of the 'norms'
that are operative in the translation process.
Generalizations about metaphor translation, if ever they are possible, will at
least have to take these specifications as a starting-point.
become part of the established semantic stock (or 'lexicon') of the language.
They are the so-called lexicalized metaphors whose range may vary from mere
'formators' (such as in the face of, beforehand, everybody, already) to single
lexical items (such as to harbourevil thoughts, hard cash, a hard-boiledcharacter,
etc.) and idioms (such as have a lark, lay heads together, lay bare, lay a finger on,
etc.). In spite of the objections which can be raised against the use of the term
'dead' metaphor, the notion of 'deadness' may give insight in the process by
which a metaphor shifts from 'performance' to 'competence.' According to
modern semantics various stages can be distinguished in the process: "the first
stage of 'petrification,' almost inevitable in an institutionalized metaphor, is that
the reference and ground of the comparison becomes limited by convention; in
that, for example, a fox is 'a person who is like a fox in that he is cunning'[...] A
further stage is reached when the transferred definition loses its analogical
feeling, so that fox is felt to be virtually synonymous with a cunning man. But
even at this stage, a feeling of the link between the literal and transferred
meanings may persist. The stage of absolute 'deadness' is reached only when the
literal meaning has died out entirely [...]; or else, when the literal and transferred
meanings have diverged psychologically to the extent that no connection is felt
between them any more" (Leech, 1974: 227-228).
A second category is constituted by the large group of traditional, or
conventional metaphors, which are more or less 'institutionalized' in that they
are common to a literary school or generation. Instances of such 'shared poetic
metaphors' would be mere-hengest [sea-steed] (as a metaphor for 'ship'),
heofon-candle [the candle of heaven] (as a metaphor for 'sun,' 'moon,' 'star') or
heofon-ward [the warden of heaven] (as a metaphor for 'God') in Old English
poetry; Homer's 'fixed' metaphors such as rosy-fingereddawn; the Elizabethan's
pearly teeth, ruby lips, golden lads; or the Augustan's wateryplain, silverstreams,
etc. To modern readers some of these are bold and poetic, while most of the
others are faded and quaint. Yet it is beyond doubt that such metaphors can
clearly be distinguished from the more institutionalized patterns of the common
language. They belong to the restricted area of literature and are only
conventional within the period, school or generation to which they belong.
The third category is that of privatemetaphors, the so-called 'bold,' innovating
creations of individual poets. But here again it is not easy to draw strict
boundaries between the private and the more traditional. Many private systems
have large overlap with metaphorical traditions, so that it would be fallacious to
emphasize the 'uniqueness' of private metaphors.
Contrary to M.B. Dagut (1976) I would not insist, therefore, on making a basic
distinction between metaphor on the one hand, and the other three forms
(polysemes, idioms and formators) on the other hand. Dagut seems to miss the
point where he contends that "in the case of polysemes, idioms or formators [...]
translation is taking place between two different systems of
language
competence" and thus "depends essentially on the bilingual competence of the
translator" (1976: 24). Surely, "every metaphor is an entirely new and
unique
creation" (p. 23); yet the status of a metaphor is not a static but a
dynamic one. If
76 RAYMOND VAN DEN BROECK
obvious that metaphor may differ in function from text to text, from language to
language and from culture to culture (e.g., a primitive vs. a highly technological
culture).
4. MODES OF METAPHORTRANSLATION
Since the task of a theory is not to prescribe, but to describe and to explain, the
theory of translation cannot be expected to specify how metaphors should be
translated. What it can attempt, then, is to set up models according to which the
observable phenomena can properly be described. Possible applications of the
theory might be:
(1) tentative predictions of how, in regard to given circumstances and under
certain conditions, metaphors are most likely to be transferred into TL;
(2) specifications of how, from the viewpoint of a 'normative' theory,
metaphors are to be translated in order that 'optimal correspondence'
between SL text and TL text may be established (according to type of text,
function of metaphor, etc.).
By lack of thorough research in this area, however, it is impossible for the time
being to make such generalizations as might be of use to translation practice.
A tentative scheme of modes of metaphor translation would show the
following possibilities:
(1) Translation 'sensu stricto'. A metaphor is translated 'sensu stricto'
whenever both SL 'tenor' and SL 'vehicle' are transferred into the TL. For
lexicalized metaphors this mode of translating may give rise to two different
situations depending on whether or not the SL and the TL use corresponding
'vehicles':
a) If the 'vehicles' in SL and TL correspond, the resulting TL metaphor will
be idiomatic.
b) If the 'vehicles' in SL and TL differ, the resulting TL metaphor may be
either a semantic anomaly or a daring innovation.
(2) Substitution. This mode applies to those cases where the SL 'vehicle' is
replaced by a different TL 'vehicle' with more or less the same 'tenor.' Then the
SL and TL 'vehicles' may be considered translational equivalents in that they
share a common 'tenor.'
(3) Paraphrase. An SL metaphor is paraphrased whenever it is rendered by a
non-metaphorical expression in the TL. In fact this mode of translating
metaphors renders them into 'plain speech'; the resulting TL expression comes
up to the level of a commentary rather than of actual translation. The table on p.
78 may serve as an illustration.
corresponding
Le jour tombe Es wird Abend sense Paraphrase
Of course the modes described may be given other labels than the ones
provided by the scheme. For example, instead of three modes one might
distinguish between four modes of metaphor translation. Thus l(a) could be
labelled literal translation, l(b) onomasiological translation, (2) semasiological
translation and (3) discursive translation.
Nothing, so far, has been said about the practical applicability of the given
modes, about the desirability of their application to specific cases or the relative
chances of the resulting products in particular circumstances. Thus the given
modes stand for what they really are, i.e., mere theoretic possibilities. Neither
have I specified them with regard to the three main categories of metaphor
mentioned above, since the motives for applying them may not only be linked up
with the type of metaphor but also depend on the textual environment in which it
appears.
Of course science has its own metaphors. One may see formulas in science as
metaphors, e.g., E=mc2, which often relate hypothetical non-entities or 'forces';
but the only way a translator can deal with these conventional metaphors of
science is to leave them unchanged, since they belong to a universal repertory of
symbols.
The other extreme position is taken by the language of literature. It is now
commonly agreed that literary texts manifest structured information on at least
three levels: a) on the linguistic level, i.e., the level of the natural language in
which they are formulated (contextual information); b) on a situational
pragmatic level, in that they are situated in space and time and form part of a
system of socio-cultural norms and conventions (socio-cultural information); c)
on the literary-aesthetic level in as much as they are structured according to the
rules and standards of a literary tradition (intertextualinformation). Thereupon
in literary texts the material deriving from the linguistic code, the socio-cultural
elements, etc., are subordinated to a principle of structuration that belongs to a
hierarchically higher order, i.e., the principle of artistic structuration which is
superimposed on them (Lotman, 1972: 22 ff.). It is this principle according to
which characteristic deviations from the normal code are to be explained in
literary texts.
It is obvious that the adequate analysis of problems emerging from metaphor
translation in literature will have to confront the foregoing data. Therefore the
three basic types of metaphor will now be discussed within the theoretical
framework of artistic structure in literary texts.
a savage servility
slidesby on grease.
The difficulties involved in translating this stanza would be largely cross-cultural,
depending on whether in the target culture there are words having a range of
associations similar to those evoked by the English words savage, servility, and
grease. In this connection it seems worthwhile to quote the tentative hypothesis
put forward by Dagut (1976: 32) "that the framework of 'possible' metaphors for
any given language is determined by a combination of the accumulated cultural
experience of the members of that language-community and the
'institutionalized' semantic associations of the items in their lexicon."
A third kind of impediments for the translator of private metaphors to
overcome are those thrown up by aesthetic convention and tradition. Many
private systems (e.g., those of Blake and Yeats) are tradition-bound in that they
have "overlap with existing symbolical traditions" (Wellek and Warren, 1963:
190). Translatability will thus be high in the case of shared literary traditions, or
will at least depend on the availability in the target literary system of similar
symbolical traditions which can provide adequate substitutes. Another aspect of
this problem is that large differences in aesthetic and moral codes between the
SL and TL may impose certain constraints of a prohibitive nature on the
translation, e.g., when the target system is governed by rigid conventions such
that an SL metaphor is rejected for its boldness, lack of modesty, etc. This may
be a frequent cause of distorted, 'flattened' or 'purified' metaphors in
translations.
good illustration of the phenomenon ("As when for a lark/ gaily, one hoists up a
window/ shut many years"). Within the context of the poem both the literal and
the figurative meanings of the idiom are 'actualized'; the effect thus provoked is
akin to that of a real poetic metaphor.
Deliberate linguistic 'foregrounding' or aktualisace, to use the term that was
introduced by the pre-war Prague School of Linguistics, consists in using the
devices of language "in such a way that this use itself attracts attention and is
perceived as uncommon, as deprived of automatization, as deautomatized, such
as a live poetic metaphor (as opposed to a lexicalized one, which is
automatized)" (Havranek, in Garvin, 1958: 10). Foregrounding is of course not
confined to creative writing, but is also found in creative joking speech (punning)
and children's language-games. Literature, however, is characterized by the
"consistency and systematic character of foregrounding" (Mukarovsky, in
Garvin, 1958: 23).
In complex texts such as poems the structuring principle of artistic (or poetic)
organization to which the contextual patterns of ordinary language are
subordinated re-awakens the symbolic force of the dead metaphor, so that in a
sense it becomes a 'live' metaphor again. But at the same time it retains its
specific character of being a 'Bedeutungsmetapher' (as opposed to a real poetic
metaphor or 'Bezeichnungsmetapher'; cf. Hausmann, 1974: 117). The effect of
this simultaneous occurrence is that, as in puns, a contrast emerges by which the
'tenor' is played off against the 'vehicle.' The 'foregrounded' idiom manifests
structured information of at least three kinds: 1) contextual (although the
reference is ambiguous between the literal 'vehicle' and the figurative 'tenor'); 2)
poetic (in the sense given to it by Roman Jakobson when he characterizes the
"poetic function" as a highlighting of the message as such); 3) metalingual (in
that it gives information about the code as a semasiological system).
The real difficulty of rendering such 'deautomatized' idioms in the TL is that
"the functionally relevant features include some which are in fact formal
features of the language of the SL text. If the TL has no formally corresponding
features, the text, or the item, is (relatively) untranslatable" (Catford, 1965: 94).
Thus, foregrounded idioms or polysemes constitute the very limits of
translatability since the well-known anisomorphy of different languages will only
in rare cases allow corresponding idiomatic phrases or polysemes whose literal
and figurative meanings have total overlap.
6. GENERALIZATIONS:LAWS OF TRANSLATABILITY
From the previous discussion of the various categories of metaphor and the
functional relevancy of their occurrence in textual settings, certain hypotheses
concerning the relative success of translating them can be formulated. These
hypotheses will be presented here as basic principles (or 'laws') of translatability.
Since substitution of TL metaphors for SL metaphors (see 4, above) or
paraphrasing will in general not hamper translatability, it will be obvious that the
following principles basically relate to translation sensu stricto.
It goes without saying that translatability of metaphors does not stand
apart
84 RAYMOND VAN DEN BROECK
from translatability in general, but is only a special case or significant aspect of it.
This means that the basic law for general translatability also applies to metaphor
translation. This law as it has been formulated by Itamar Even-Zohar (1971: IX)
says that "the degree of translatability increases when the relational series which
produce information and rhetoric in the SL and TL grow closer." In the same
way the following particular laws of translatability in general apply to metaphor
translation:
(1) Translatability is high when a pair of languages are of a close basic 'type,'
provided that the conditions under (2) and (3) are fulfilled.
(2) Translatability is high when there is contact between SL and TL.
(3) Translatability is high when the general cultural evolution in SL and TL
proceeded on parallel lines.
(4) Translatability is high when translation involves no more than a single kind
of information. In other words, a text is more translatable if it displays
information of a sing-letype than if it is 'complex' in that various types, and
hence a greater quantity of information, are involved.
(Adapted from Even-Zohar)
The last principle is of special importance for metaphor translation, because
its further specification will allow us to distinguish between different degrees of
translatability according to particular categories of metaphor in different
contextual settings. It will therefore be reformulated as follows: translatability
keeps an inverse proportion with the quantity of information manifested by the
metaphor and the degree to which this information is structuredin a text. The less
the quantity of information conveyed by a metaphor and the less complex the
structural relations into which it enters in a text, the more translatable this
metaphor will be, and vice versa (BASIC LAW).
This basic law is now capable of further specification:
(1) Lexicalized metaphors in merely referential texts (where their'vehicles' as
such have no functional relevance), as they represent a single kind of
information, are of a high translatability.
(2) Foregrounded lexical metaphors in complex texts (poems, puns) are of a
very low translatability since they convey contextual (semantic as well as
pragmatic or cultural), poetic, and metalingual information simultaneously.
(3) Bold private metaphors in literary texts(and hence 'poetic' metaphors) will
be more translatable than conventional metaphors to the degree that they
are less culture-bound and are thus able to dispense with culture-specific
information.
(4) Finally it seems reasonable to assume that 'decorative' metaphors (as e.g., in
journalistic prose) will impose lower requirements on the translator than
'creative' ones (as in poetry); to the degree that they are less relevant for the
communicative function of the text - at least in so far as their'vehicles' are
concerned - they may often either be substituted by TL-specific
equivalents or paraphrased.
Thus the translatability of metaphor can stand as a model for translatability
with regard to different types of text. It will be obvious, however, that the
METAPHOR TRANSLATION 85
7. HIGHERCONSTRAINTSON TRANSLATION:TRANSLATIONALNORMS
A further step to be taken would involve a description of the various ways in
which metaphors are actually translated by different generations, in different
periods of literary history, between particular pairs of languages and literatures,
and by individual translators. Such a description would certainly allow the
student to discover regularities and recurrent patterns, which historical poetics
could account for. It is not my intention to make a start with this description, not
even to state the general lines along which it might be undertaken.
It seems however quite acceptable to assume that the actual ways in which
metaphors in literary texts are translated stand in close relation to rules and
norms. In this connection the concept of "initial norm" as developed by Gideon
Toury (1976) will be a useful tool for the distinction between 'basic' attitudes.
According to Toury the translator is always confronted with a "basic choice
between two polar alternatives deriving from the two major constituents of the
'value' in literary translation [...]: he either subjects himself to the original text,
with its textual relations and the norms expressed by it and contained in it, or to
the linguistic and literary norms active in the TL and in the target literary
polysystem, or a certain section of it."
Although the translator's practical decisions will rarely represent one or the
other alternative in its 'pure' form - that is to say that the decisions made will
thus generally be some combination of these two extremes - a description of
metaphor translation in terms of initial norms can start from the following basic
distinction:
(1) If the translation adheres to the SL norm, metaphors will tend to be
translated 'sensu stricto,'even if the resulting item in TL might prove to be
incompatible with the target linguistic and/or literary norms. This retentive
mode of translating is then responsible for the deviant (or 'alienating', i.e.,
exotic and/or archaic) character of translated metaphors.
(2) If the second position is adopted, SL metaphors are most likely to be
replaced by more or less corresponding (or equivalent) TL metaphors, or
will at least often be adapted. The prevailing mode here is substitution
by
which original metaphors are domesticated, i.e., adapted to the
prevailing
norms of the target system, which eventually determine the
acceptabilityof
translational equivalents.
The acceptance of this concept of translational norms has
fargoing
consequences for a description of actual metaphor translation. It may, for
example, account for the fact that sometimes bold SL metaphors are replaced by
plain or flat TL items, or that certain SL conventional metaphors are distorted or
even omitted in the translation. In such cases it is
important to realize that the
translation of metaphors, together with the other
literary devices, is subjected to
86 RAYMOND VAN DEN BROECK
8. CONCLUSION
One of the objections that could be raised to this discussion of metaphor in
translation is that it treats this poetic device as if it were something that can be
dealt with in isolation, i.e., apart from the other poetic means. It is clearly
understood, however, that the translatability of metaphor in concrete texts will
depend on the relations into which it enters with the other elements on various
levels (syntactic, prosodic, etc.). It is also obvious that analysis must necessarily
separate and treat in isolation what in reality presents itself combined and
unified. The few generalizations that follow from the discussion do not stand for
- to translation of
metaphor only. They may apply - mutatis mutandis
literature in general and must therefore be supplemented with further
generalizations concerning other poetic phenomena. What I have mainly
attempted to show is that certain generalizations can be made. Since the theory is
concerned with discovering regularities, it should not try to create them by
it
imposing rules or norms on translational practice. I hope that this paper, in that
brings a tentative description of how and to which extent metaphors can be
translated, has in a sense contributed to clarifying what a theory of translation
can actually try to achieve.
REFERENCES