A Cognitive Approach to Translating Metaphors^*

1. Introduction

   egardless of its popularity and mechanism of operation, metaphor as a linguistic device exists in all human languages. The word 'metaphor' comes from Greek metapherein, meaning 'to transfer' or to 'carry over.' Reference to this universal linguistic phenomenon can be found in the writings of Greek philosophers and rhetoricians, as well as of contemporary linguists (cf. Richards, 1965; Leech, 1969; Dagut, 1976; and Maalej, 2002). The earliest definition of metaphor—quoted from Aristotle's The Poetics by Richard (1965: 89) is "a shift carrying over a word from its normal use to a new one." Under this quite broad definition, all other instances of semantic extension such as allegory, synecdoche, metonymy, etc. might be categorized as being metaphoric. Whichever term is used for labeling these expressions, they all exhibit some kind of semantic and logical violation to the referential components of their lexical constituents. Hence they are studied as instances of figurative (as opposite to literal) language, where words gain extra features over their referential ones. Therefore, the meaning of any of these lexical constituents cannot be predicted from their referential meaning. Unfortunately, the translator has to suffer twice when he approaches these metaphoric expressions. First, s/he has to work out their figurative meaning intralingually (i.e. in the language in which a metaphor is recorded). Second, s/he has to find out equivalent meanings and similar functions of these expressions in the TL.

   Studies of metaphor have been largely dedicated to issues such as the meaning, forms, components, typology, and the role of metaphors as speech ornaments and meaning-enhancing analogies. These studies shy away from the exploration of the continuous connection of metaphors as mental or picturesque representations of the real world and the language used to realize these pictures in words. Despite the large amount of literature available on the literary aspects of this linguistic phenomenon, very little research has been done on the cogno-cultural translation of metaphors. This paper intends to show how metaphors reflect cognitive and cultural human experiences encoded by language as a means of recording human experience and how culture models and constrains this cognition. In particular this paper is an argument in favor of a cognitive approach in the translation of metaphors, especially between culturally distinct languages, e.g. English and Arabic. The study of the metaphoric expressions of a given culture would, hopefully, give us a chance to see how the members of that culture structure or map their experience of the world and record it into their native language. Since one of the basic assumptions is that culture influences metaphor in an important way, the following section attempts to clarify how metaphor is a cultural object. For cognitive conceptualization of metaphor, the present research draws on Mandelblit's (1995) "Cognitive Translation Hypothesis," which is the subject of Section 3.

   
   

2. Cultural Conceptualization of Metaphor

   One of the oldest definitions of culture, which is used by the Encyclopedia Britannica (1983, vol. 4:657) is "that complex whole, which includes knowledge, belief, art, morals, law, customs, and other capabilities and habits acquired by the man as a member of society."

   The following section will illustrate how the metaphoric choices available to a user are filtered by the value and belief systems prevailing in the cultural community the text is translated into. Following Lakoff and Johnson (1980: 12), "a culture may be thought of as providing, among other things, a pool of available metaphors for making sense of reality"; "to live by a metaphor is to have your reality structured by that metaphor and to base your perceptions and actions upon that structuring of reality"(ibid). This is related to the fact that people of a given culture use language to reflect their attitudes towards the world in general and the life of the community they live in particular. This in turn gives rise to the reason for our argument in favor of a cognitive approach in translating metaphors, which takes into account cultural beliefs and values especially between culturally distinct languages, e.g. English and Arabic. To put it differently, since different cultures classify the world's complexities in different ways, translations from one language to another are often very difficult. This difficulty would increase a lot when translating between two distant cultures where all traditions, symbols, life conditions and methods of experience representation are different. For example, if you say "a man has a 'big head' in English, it means 'he is arrogant,' whereas in Italian 'he is clever.'

   This also explains the ease of translating some universal metaphors denoting similar ideas in different cultures. Metaphors related to the parts of human body are examples of the case. Consider the following English metaphors of the human body:

   1. a. 'To give someone a hand,' meaning 'help someone.'
   2. a. 'To keep an eye on something,' meaning 'watch or pay attention to something.'

   The Arabic translation of the above metaphoric expressions means the same and reads as follows, respectively:

   1. b. yamuddu yada ?almusaa'adah
   2. b. yaDa'u 'ayynahu 'alaa

   But the question is how many of these instances can be found among human languages? Unfortunately, very few exist. In this regard, Chitoran (1973: 69-70) states that

   "the differences in environment, climate, cultural development, etc., among various communities may be extremely significant, but basically, human societies are linked by a common biological history. The objective reality in which they live is definitely not identical but it is by and large similar."

   However, the universe we are living in is made up of things, and we are constantly confronted with them, obliged to communicate about them, and to define ourselves in relation to them. This is a characteristic of all human societies, and due to this fact various language systems are not easily translatable. Therefore, because different cultures conceptualize the world in different ways, metaphors are characterized as being culture-specific. This is in line with Dagut's (1976: 32) argument that there is no simplistic general rule for the translation of metaphor, but the translatability of any given SL metaphor depends on (1) the particular cultural experiences and semantic associations exploited by it, and (2) the extent to which these can, or cannot, be reproduced non-anomalously into the TL, depending on the degree of overlap in each particular case.

   As he goes further, Dagut (ibid. 28) says that "what determines the translatability of a SL metaphor is not its 'boldness' or 'originality,' but rather the extent to which the cultural experience and semantic associations on which it draws are shared by speakers of the particular TL." We would like to go even further to state that the inherent difficulty of metaphor translation is not the absence of an equivalent lexical item in the TL, but rather the diversity of cultural conceptualization of even identical objects or worlds in both communities whose languages are involved in translation. Snell-Hornby ( 1995: 41) expresses the same idea as he states that "the extent to which a text is translatable varies with the degree to which it is embedded in its own specific culture, also with the distance that separates the cultural background of source text and target audience in terms of time and place."

   
   

3. Metaphor and Cognitive Equivalence in Translation

   Katan (1999) suggests that a cognitive approach to the study of culture can be seen in terms of the form of things that people have in mind, their models for perceiving, relating to, and interpreting them. This view of culture suggests that, when translating a text to a SL of any other culture, one needs to be aware not only of the patterns of thinking, and acting in one's own culture, but also of the TL's cultural models of reality. Nida (1964) described the 'best' translation as the one capable of evoking in the TL reader the same response as the SL text does to the SL reader. Although we find this a rather unreachable objective, we still believe that some of it can be achieved provided that the following two conditions are satisfied: First, the translator must understand the way in which receptive readers perceive the world and structure their experience. Second, he must also try his best to find a way to accommodate his text to the experience of the target-language reader, and to the way it is recoded in the TL. Our argument in favor of a cognitive approach to the translation of metaphors derives from the notion of 'cognitive equivalence,' where metaphors can be translated from one language to another with a minimum degree of loss. For this reason, I think that metaphors must be looked at as cognitive constructs rather than mere linguistic entities or rhetorical phenomena. In other words, metaphors represent instances of how people conceptualize their experience and how they record it. Hence, it is believed that the cognitive approach will work for this purpose.

   In cognitive linguistics, metaphor is often given a cognitive function in which human beings draw upon the experience of each other or non-human surroundings or even other concepts or images. Lakoff and Johnson (1980) define metaphor as a means to understand one domain of experience (the target domain) in terms of another, a familiar one (source domain). This usually takes the form of analogy or comparison between two existent entities or one existent entity and another one assumed to exist. To say that someone is a 'lion,' for example, reveals that a link has been established between that individual (tenor) and the 'lion' (vehicle) as a symbol of bravery or strength. Therefore, metaphors are 'conceptual' phenomena in which the source domain is mapped onto the target domain. To put it differently, the structural components of the source conceptual schema are transferred to the target domain. Here one should deter the crucial role of culture in this process of symbolization and conceptualization. In the Arab world, for instance, an 'owl' is often conceptualized as a sign of bad omen. Surprisingly, it is a symbol of wisdom in the Western culture.

   In the cognitive study of metaphor an emphasis is made on the psychological as well as on the sociocultural and linguistic aspects of metaphor. Furthermore, metaphors are associated with 'indirectness' (Green, 1989: 124; and Maleej, 1990); this is possibly why they are common as a special mode of expression in politics and public speeches where direct expressions are censured. To those who studied metaphor within the scope of cognitive linguistics (e.g. Lakoff & Johnson, 1980; and Goatly, 1997), metaphor is 'pervasive in everyday life, not just in language but in thought and action,' and that our 'ordinary conceptual system is fundamentally metaphorical in nature' (Lakoff & Johnson 1980: 3).

   For the translation of metaphor, I would like to incorporate Mandelblit's (1995) 'Cognitive Translation Hypothesis,' but this time for a different purpose and in a different framework. Mandelblit's proposed two schemes of cognitive mapping conditions (i.e. Similar Mapping Condition (SMC) and Different Mapping Condition (DMC)). While Mandelblit intended to show that 'the difference in reaction time is due to a conceptual shift that the translator is required to make between the conceptual mapping systems of the source and target languages' (p. 493), we are more interested in the outcome of his research than in its methodology and objectives. He found out that metaphorical expressions take more time and are more difficult to translate if they exploit a cognitive domain different from that of the target language equivalent expression. According to the hypothesis, the reason for this delay, difficulty and uncertainty in the translation of different domain metaphors is the search for another conceptual mapping (i.e. another cognitive domain). That is to say the fact that metaphors almost always exploit such different cognitive domains implies the search for a cognitive equivalence for SL metaphors in the TL. In other words, the translator is called upon to play the role of a proxy agent doing the act of conceptual mapping on behalf of the TL reader. If he can touch upon a similar TL cognitive domain, then his task will be fulfilled quite successfully and easily. If not, he has to look for the cognitive domain that fits in the TL as the SL one does. The result of the first action is often an equivalent TL metaphor or—under the worst conditions—a TL simile. The result of the second action, however, is open to many possibilities, of which rendering the SL metaphor into a TL one is the least likely. Thus a metaphor might be rendered into a simile, a paraphrase, a footnote, an explanation or—as a last resort—it can be omitted.

   Therefore, I believe that attempts of literal rendering or mere linguistic meaning transference of the metaphoric expressions from one language to another are deemed to result in a noticeably bad product, especially when these expressions draw on culture-specific methods of thinking rather than on shared or universal notions or schemata.

   Referring to cultural aspects and drawing on the general guidelines of the cognitive framework (i.e. the cognitive equivalent hypothesis) for metaphor translation, we utilized three sets of authentic English and Arabic examples of metaphors. The first set comprises metaphors of similar mapping conditions reflecting shared ideas which are expressed by identical metaphors in both languages, and other instances of metaphors which are realized by different lexical items in the TL. The third contains metaphors of different mapping conditions, and which lack equivalents in the TL.

   
   3.1. a. Metaphors of similar mapping conditions

   This category represents metaphors expressing a small number of ideas shared by the two languages and hence expressed, roughly speaking, by similar expressions. Anthropologists call theses shared ideas 'cultural universals.' Comprising many diverse sub-cultures, a universal culture can be thought of as a constellation of common core attitudes and values reflected by practices common to most of the sub-cultures. Similarities in mapping conditions across diverse cultures could be labeled as 'pancultural metaphorical expression,' which derives from 'panhuman sharedness of basic experience' Emanatian (1995: 165). Consider the following almost similar English and Arabic metaphors; most of them are proverbs reflecting the wisdom of many sub-cultures. Having a didactic function, these metaphors figure human philosophical insights, logic, wisdom, and instructions in ways which reinforce universal conventional images and attitudes, and therefore both reflect and reproduce those conventions. In other words these metaphors are a reflection of human experience; they can contribute to exposing the way such conventions are embedded in language.

   3. SL/ History repeats itself.
   TL/ ?ttaariikhu yu'iidu nafsah
   Lit. The history repeats itself
   
   4. SL/ Necessity is the mother of invention.
   TL/ ?alHaajah um ?al?ikhtiraa'
   Lit. The need is the mother of invention.
   
   5. SL/ Actions speak louder than words.
   TL/ ?al?af'aal ?ablagh min ?alaqwaal
   Lit. Actions more rhetorical- than sayings
   
   6. SL/ Birds of a feather flock together.
   TL/ ?aTuyyuru 'alaa ?ashkaalihaa taqa'
   Lit. Birds shapes- their on fall
   
   7. SL/ A drowning man will clutch at a straw.
   TL/ ?alghareeq yata'alaq fii qashah
   Lit. A drowning clutches in a straw
   
   8. SL/ This is a mouth-watering opportunity.
   TL/ ?innaha lafurSah tusiilu ?allu'aab
   Lit. This opportunity causes saliva come out (of my mouth).
   9. SL/ You make my blood boil.
   TL/ ? innaka taj'alu ?adama yaghlii fii 'uruuqii
   Lit. You make blood boils in my veins
   
   10. SL/ This issue is the cornerstone.
   TL/ haathihi ?almas?alatu hiyya Hajaru ?azzawiyyatu
   Lit. This issue it the cornerstone
   
   11. SL/ To throw dust in the eyes.
   TL/ tharu ?aramaadi fii ?al'uun
   Lit. Spread off ash in the eyes
   
   12. SL/ A bird in the hand is worth two in the bush.
   TL/ 'uSfuurun fii ?alyadi khayrun min 'asharah 'ala ?ashajarah
   Lit. One bird in hand is better than ten on the tree.
   
   13. SL/ A cat has nine lives.
   TL/ qiTTah bisab' ?arwaaH
   Lit. a cat with seven lives
   
   14. SL/ Time is money.
   TL/ ?alwaqru min thahab
   Lit. The time from gold.
   
   15. SL/Cleanliness is next to godliness.
   TL/ ?annaDafatu min ?alliyman
   Lit. Cleanliness from faith (in Allah)

   Although examples 3-11 above represent metaphors expressing a small number of ideas shared by the two languages, hence expressed, roughly speaking, by similar expressions, examples (12), (13), (14), and (15) above reflect values, and beliefs peculiar to each particular culture (i.e. English and Arabic). Notice how users of each language conceptualize the concept of number in (12) and (13) to reflect the same idea. English employs 'two' whereas Arabic is only satisfied with 'ten.' However, in (13), English uses number 'nine' while Arabic employs 'seven.' Notice also, in example (14), the variation in 'value' conceptualization in each language; English refers to 'money' (i.e. the monetary value) while in Arabic, 'time' is likened to 'gold' in superiority (i.e. superiority value). Further, religious affiliations affect the lexical choice to express the same idea in each language, as it is the case in (15). However, examples, such as (4) above, embody a conceptual metaphor, where the SL (i.e. English) concept or experience is borrowed and loan-translated into Arabic.

   
   3.1. b. Metaphors having similar mapping conditions but lexically realized differently

   As stated above, beliefs and religion are aspects of culture that play a very significant role in translation. As is shown in the following examples, although the English examples and their Arabic counterpart metaphors are related to the same conceptual domain, the religion or ethical system in the TL has led to major differences in lexical choice.

   16. SL/ A fox is not taken twice in the same snare.
   TL/ laa yuldaghu ?alm?uminu min ?aljuHri marraryin
   Lit. No believer (in Allah) stung from a hole twice
   
   17. SL/ Every cloud has a silver lining.
   TL/ ?inna ma-'al- 'usri yusraa
   Lit. Verily, with every difficulty there is relief.
   
   18. SL/Many hands make light work.
   TL/ yadu ?allahi ma' aljamaa'ah
   Lit. Hand of Allah with the group

   In the three examples above, the only plausible justification for this variation in the use of metaphoric expressions is the fact that the users of each language map the particular conceptual domain of their own world differently. That is to say, the Arabic translation is quite consonant with those of Islamic beliefs because the equivalent Arabic translation is either a saying from Prophet Mohammed's sayings as in example (16), or a verse from the Holy Qur'an as it is the case in (17), which is the sixth verse of Al InshiraaH. This verse roughly means 'whatever difficulties or troubles are encountered by me, Allah always provides a solution or a relief if we only follow His Path.'

   
   3.2. Metaphors of different mapping conditions

   Examples of this category generate when working on culture-bound SL metaphors that are mapped into a domain different from that of the TL. Since 'languages are the best mirror of human cultures,' and 'it is through the vocabulary of human languages that we can discover and identify the culture-specific conceptual configurations characteristic of different peoples of the world' Wierzbicka (1992: 22), different cultures conceptualize experiences in varying ways. Therefore, following Dagut (1972:32), 'the translatability of any given SL metaphor depends on (1) the particular cultural experience and semantic associations exploited by it, and (2) the extent to which these can, or not, be reproduced non-anomalously in TL, depending on the degree of overlap in each particular case.' This is typically the case when working on metaphors mapped in the religious and political domains. Such metaphors are called root metaphors underlying people's views or attachments and shaping their understanding of a situation. Religion is considered the most common root metaphor since birth, marriage, death and other life experiences can convey different meanings to different people depending on their religious beliefs. Below are examples of Arabic religious (mostly Qur'anic) conceptual metaphors the image of each of which cannot be reproduced in the TL. Therefore, the translator has no choice other than replacing the SL image with a TL image that does not clash with the target culture. This can only be done by resorting to the strategy of different cognitive mapping in search for cognitive equivalence. As mentioned before, the product of this process might be a TL simile, paraphrase, explanatory remark, or even a footnote. The reader is invited to see how inadequate the translation is due to the absence of identical cognitive mapping of the SL expressions in the TL on behalf of the translator. The Holy Qur'an constitutes a rich source of such metaphors which pose a serious problem even to the most experienced professional translators. To shed more light on this subtle aspect of Arabic metaphor, let us consider some authentic exemplary metaphors cited from the Holy Qur'an along with their English translation by Ali (1989):

   19. SL/ ?uHilla lakum laylataS-Siyaamir-rafathu ?ilaa nisaa?ikum Hunna libaasul-lakum wa ?antum libaasul-lahunn [Surat Al Bagarah, verse 187]
   TL: On the night of the fasts, you are allowed to approach your wives. They are your garments and you are their garment. [Surat Al Bagarah, verse 187]
   
   20. SL/ nisaa?ukum Harthun lakum fa?tuu Harthakum ?annaa shi?tum
   [Surat Al Bagarah, verse 223]
   TL/ Your wives are as a tilth for you, so approach your tilth how you will.
   [Surat Al Bagarah, verse 223]
   
   
   21. SL/ qaalat ?annaa yakuunu lii ghulaamun walam yamsasnii basharun walam ?aku baghiyyaa [Surat Maryam, verse 20]
   TL/ She said: "How shall I have a son, given that no man has touched me, and I am not unchaste?" [Surat Maryam, verse 20]
   
   22. SL/ ?almaalu walbanuuna ziinatu ?alHayaati ?adunyaa
   [Surat Al Kahf, verse 46]
   TL/ Wealth and children are the adornment of the life of this world.
   [Surat Al Kahf, verse 46]
   
   23. SL/ walaa yaghtab ba'Dukum ba'Daa. 'ayuHibbu ?an ya?kula laHma ?akhiihi maytan [Surat Al Hujuraat, verse 12]
   TL/ Nor speak ill of each other behind their backs. Would any of you like to eat the flesh of his dead brother? [Surat Al Hujuraat, verse 12]

   As can be seen, the literal English translations for the above verses do not work as equivalents for the Arabic euphemistic areas of sex and related matters; that is because the question of sexual intercourse is always delicate to handle in the Arabic culture. Witness the Qur'anic euphemisms and their English counterparts in (19), (20), and (21) above. In (20), 'sexual intercourse' in the SL is compared to a husband's tilth, where a husband sows the seed in order to reap the harvest; he chooses his own time and mode of cultivation. He does not sow out of season nor cultivate in a manner which will exhaust the soil. However, this image cannot be reproduced in the TL. Likewise, 'sexual intercourse' is euphemistically referred to as 'approaching' in (19) and 'touching' in (21), but this translation is inadequate due to the absence of identical cognitive mapping of the SL expressions in the TL on behalf of the translator. In a similar way, in the Arabic SL text in (22), the back-biting is likened to eating the flesh of a dead brother in terms of abomination; however, the English translation does not seem to reproduce plausible counterpart metaphors when the general meaning of the Arabic words is given.

   It is apparent that the attempts to maintain these metaphors in English translation have communicatively failed. To solve this problem, the translator of these verses provides footnotes to explain the meanings of these Arabic metaphors. However, a better policy in such cases, according to El-Hassan and Al-Said (1989), is to provide a brief explanation in the main body of the text provided that it does not unduly interrupt the flow of the text.

   
   

4. Conclusions

   Since metaphor is shaped by the socio-cultural beliefs and attitudes of a specific culture, our translation of this linguistic phenomenon is based on the 'cognitive equivalence,' where metaphors must be looked at as cognitive constructs representing instances of how people conceptualize their experiences, attitudes and practices, and record them. Then operationally, we have drawn a distinction between the individual linguistic culture having its own set of metaphors related to a range of ideas, conventions, and beliefs, and a proposed 'universal culture' comprising many individual cultures (i.e. sub-cultures) sharing a set of metaphors reflecting the core values and practices common to most of the individual cultures.

   Since metaphors are related to different cultural domains, this implies that the translator has to do the job of conceptual mapping on behalf of the TL reader; he has to look for a TL similar cognitive equivalence in the target culture. The more the SL and TL cultures in question conceptualize experience in a similar way, the easier the task of translation will be. But since human real-world experiences are not always similar, and metaphors record these experiences, the task of the translator becomes more difficult when translating these metaphors across languages related to different cultures. The difficulty of metaphor rendition lies not in the assumption that languages cannot provide equivalent expressions for their metaphors, but in the fact that they lack counterpart metaphors related to the same conceptual domain or area. Therefore, in search for cognitive equivalence to replace the SL image with a TL image that does not clash with the target culture, we have differentiated between three cognitive mapping conditions to the translation of metaphors: (1) metaphors of similar mapping conditions, (2) metaphors having similar mapping conditions but lexically implemented differently, and (3) metaphors of different mapping conditions. The difference between these three can be represented as a cline or continuum, with the set of metaphors of similar mapping conditions at one end, and those of different mapping conditions at the other end of the continuum, and those of similar mapping conditions but lexically realized differently as an intermediate set in between the polar opposites. Examples of the first category generate when working on cultural universal SL metaphors derived from shared human experience (Emanatian 199); the second set is related to the same conceptual domain in the SL and the TL, but the ethical system in the TL or the SL has led to major differences in lexical choice; whereas the third set includes the culture-bound SL metaphors that are mapped into a domain different from that of the TL.

   It can be concluded that translators, whose task is to produce a TL text that bears a close resemblance to the SL text, should be aware of cognitive and cultural issues when translating from Arabic into English or vice-versa. Therefore, it is not enough for translators to be bilingual, but they should be bicultural as well. Because translators suffer twice when approaching some metaphors which are cultur-bound and due to their figurative meaning intralingualy, it is recommended that translators be trained in coping with metaphor translation not only in foreign-language programs, but also in their native language. Sometimes, even native speakers are not always able to comprehend the figurative meaning of messages in their own language (Al-Ali 2004).