Narrative intelligence represented in translational success: issues of coherence, gaps, complexity and aesthetics

*khoshsaligheh@um.ac.ir Abstract: Among the recent trends in researching translation and interpreting, the psychological aspects of translation process and the translator’s cognitive state are swiftly developing and thriving trends among the scholars in the field. This study sought to implement this psychological approach in translation studies and investigated the possible effect of the cognitive concept of Narrative Intelligence (NI) on successful performance in translation. Over one hundred Iranian undergraduate English translation students were invited and 104 volunteers participated in the study. This qualitative study traces and contrasts the narrative features in translation of two groups of high and low NI English translation students studying their senior year in Iranian universities. The results showed that the translation performance of high NI group reflected a closer congruency with narrative features than that of the low NI group. The findings of the study relating to a) narrative coherence, b) the degree of text complexity/simplicity, c) gaps in translation, and d) creative/aesthetic merit are presented and discussed.


INTRODUCTION
Admitting that the field of Translation Studies (TS) has undergone significant changes during the last five decades, one can accept Tymozcko's (2005Tymozcko's ( :1082) ) belief that translation has turned to be a cluster concept with an open definition that has allowed the concept to adapt to various cultural conditions and social functions.By defining the concept of translation as such, one can expect to trace the knowledge in other fields besides linguistics in translation, so that the processes of human cognition, identity formation, as well as cultural dimensions fall into sharp focus.
Although the knowledge of language and language use are essential elements for the translators, translation is cognitively a psycholinguistic formulation process by which the message in the source culture is reproduced in the target culture through the cognitive processes carried out in the translator's mind (Robinson, 2007;Wills, 1998).One key element of these mental efforts is intelligence in human cognition, the significance of which is frequently discussed in the literature (e.g., Bellgard, 2006;Crossly, 2007).In addition, the importance of different types of intelligence in translatorial success is also confirmed in various studies (Hubscher-Davidson, 2013;Pazhouhesh and Davoodi Rad, 2015;Zavalaa, 2012).Along with these studies, the role of Narrative Intelligence (NI) as a relatively novel construct in translation field has been probed very recently.In a recent study, Pishghadam et al. (2016) showed that NI, verbal intelligence and linguistic knowledge have significant effects on Translation Quality (TQ) in original and backtranslation.In another study (Heidari et al., 2015), the relation between TQ and overall NI and NI subscales were examined; the result of which revealed a significant correlation between TQs and NI; they also found that there is a meaningful difference between translation qualities of two groups of translation students with high versus low NI.
Therefore, acknowledging the fundamental role of narrative in the human behavior, it has been the subject of an increasing number of psychological studies such as in Artificial Intelligence (Mateas and Sengers, 1999;Sengers, 2000) that made an improvement in humancomputer interaction; Narrative Psychology (Bruner, 1987;1991); Developmental Psychology (Nelson, 1993;Oers, 2007) that highlights the central role of stories in the development of a social self in human beings, and in language education and teacher effectiveness (Black and Howard-Jones, 2000;Elizabeth et al., 2008;Pishghadam et al., 2013).

December 2016
Sri Lanka Journal of Social Sciences 39 (2) It is assumed that NI construes an important part of the cognitive abilities that are necessary for one to produce a good translation (Heidari et al., 2015), as it is believed that the humans make sense of the world around them through narratives (Bruner, 1987;Randall, 1999).Narratives are 'stories' that enable and guide people's behavior (Baker, 2007;Steele, 1986).
This study investigates the translation performance of two groups of participants, with high versus low NI.The authors seek to address the following question: What are the narrative differences between the translations of participants with highest versus low NI?
This study draws on the findings of a preliminary quantitative study (Heidari et al., 2015) in which empirical evidence indicated a significant correlation between the participants' translation quality and their NI scores (r = .370,p < .01).The inferential statistics also revealed a significant difference between the mean scores of the translation quality of the high NI group and the low NI group (t = 3.58, p < 0.01).

Defining narrative
Oxford Mini Dictionary defines narrative as "a spoken or written account of something" (Hawker, 2002: 406).Considering narrative, basically "an account of something", it is clear that narrative is inevitable, it is everywhere, and is key to human understanding, communication and social interaction (Mateas and Senger, 1999).
As Abbott (2002) puts it, narrative is exactly our understanding of the world around us in any moment.Bruner (1991) discusses that narratives provide links, connections, coherence, meaning and sense.Narrative is an instrument that incites active thinking and helps us work through problems.As Bamberg and McCabe (1998) state, narratives serve many purposes, to remember, to argue, to convince, to engage or to entertain the audience.

Narrative Intelligence
Cognitive and psychological research suggest that the use of narrative can be as a cognitive tool for situated understanding (Bruner, 1991), as a cornerstone of one's identity (Singer, 2004) and as a means of supporting early development of language (Johnston, 2008).It is this human ability to organise experience into narrative which is called NI (Blair and Meyer, 1997).In a very much similar point of view NI is considered as one's ability to organise and explain experiences in narrative terms (Mateas and Sengers, 1999), to comprehend and make inferences about narratives we are told and to produce affective responses such as empathy to narratives (Mar et al., 2011).Randall (1999: 11) expresses that "Without NI it can be argued, countless processes integral to human existence are impossible, perhaps inconceivable".Bruner (1996) argues that story making is central to creating an understanding of the world into which a person can feel they will fit.In a similar fashion, to Csorba (2012), NI can be defined as "capacity to think of the world in a narrative manner".To put it into a nut shell, narrative construction of whatever type is a cognitive ability that has its root in the human mindset.Randall (1999: 10) has defined this unique human faculty as "the capacity both to formulate (compose, narrate) and to follow (understand, read) the story of our life".Randall's categorisation of the dynamics of NI consists of intertwining sub-capacities, namely, the abilities to emplot (i.e., the ability to edit, to summarise, to cope with conflict, to prioritise the details of events, to perceive events, to connect events, to comprehend, and to fill in the blanks); to characterise (i.e., the ability to construct 'working pictures' of what we are like, based on a variety of cues, clues, and half perceived features reflected back to us in the reactions and opinions of others); to narrate (i.e., the ability to impose order on events, to summarise the central actions in a way that captures its core dynamic of development and denouement); to genre-ate (i.e., the ability to recognise 'narrative tone' both in situations and in lives, whether others' or our own); and to thematise (i.e., the ability to be aware of recurring patterns of meaning in particular events, situations, or lives, to identify symbols or motifs and entertain theories as to their application or relevance).

Narrative Competence
As we have believed that the narrative construction of meaning is an essential factor in how people act and experience their lives, one might need to have a clear understanding of the Narrative Competence and its nature.To Oers (2007) Narrative Competence is the ability to produce a comprehensive system of utterances that are coherent.Therefore, every narrative must have a degree of addressivity that is to be comprehended by another person and by oneself.He argues that narrative competence has its implication in the sociocultural milieu where an individual is willing to construct and clarify his meaning to others in a social context.According to Sri Lanka Journal of Social Sciences 39 (2) December 2016 Ganzevoort and Tromp (2006), Narrative Competence has four distinct levels; fluency, coherence, integration and purposefulness, respectively.They expect to find Narrative Competence in the sense that narrators would tell their stories more fluently and more completely.Dobson (2005) believes that Narrative Competence is an underlying skill which is based upon identifying, understanding, interpreting, creating and communicating.
To him this competence rests upon a more essential and fundamental competence, that of NI, which is our underlying form of literacy.

Narrative analysis in relation to linguistics
Narratives in a general view are the attempts of narrators to give meaning to their experiences.Narrative analysis attempts to systematically relate the narrative means utilised for the function of 'meaning making' to particular kinds of experiences (Bamberg, 2012).The study of the textual properties of narratives typically is concerned with the textual-structural properties as well as with content in terms of themes (Gubrium and Holstein, 2009 as cited in Bamberg, 2012).
There are some different models for narrative analysis that have been used in psychology and linguistics.Labov and Waletzky (1967) systematically evaluated narratives from a formal linguistic analysis viewpoint.They defined narrative as a choice of specific linguistic techniques that consists of six elements: Abstract (what the story is about); Orientation (the time, place, and the identity of characters for the reader/listener); Complication (events which are unusual, funny, scary, that make the story a story); Evaluation (comments on the events) or it is a clause that evaluates a narrative event; Resolution (how the events worked out); Coda (rounding off the story and bridging back to the present).
To Pavlenko (2007), a systematic analysis of autobiographic narratives should be done in terms of content, context and form on macro and micro levels.By content analysis, she means not only what was said or written, but also what was omitted and why.By context, she encourages researchers to consider both global and local contextual influences on narrative construction.The global or macro-level of analysis should focus on historic, political, economic and cultural conditions of narrative production.The local or micro-level should focus on the context of the interview or manuscript writing, and thus to the influence of language choice, audience, setting, modality, narrative functions, interactional concerns, and power relations on ways in which speakers and writers verbalise their experiences.An analysis of the written form of any narrative can highlight the linguistic, cultural, and genre aspects based on people's perception of their own life stories (macro-level); it is examining how storytellers achieve their interactional goals through particular narrative devices or lexical choice (microlevel) and illuminates individual creativity and agency in the presentation of self.Bamberg (2012) draws on two approaches: (i) a linguistic-based approach that works through the lexical and syntactic configurations of texts and follows their build-up into the topical organisation of the text, (ii) a cognitive-based approach that deems texts as cognitive structures whose segments are held together by an overarching structure of the plot organisation.
The first way of approaching narratives starts from the clause and its lexical-syntactic make-up as the basic analytic unit and assumes that tying clauses cohesively together follows the language-specific practices and norms of cohesion-building.The second approach is to describe the units that emerge in the course of narrative cohesion-building in terms of elements that ultimately result in some structural whole.This top-down fashion, gains its meaning from the narrative whole.The whole lends meaning to the components of the story and their sequential arrangement.And the way the components are arranged is to be viewed as a function of the whole.Consequently, the analysis within this analytic frame proceeds from the whole to its parts.

Translation and cognition
Empirical translation research emerged in the mid-1980s.It focuses on exploring what goes on in the translator's mind during the translation process.While translation process research has turned to be the core of Translation Psychology, in a general sense Translation Psychology deals largely with what is happening in the translating mind during the translation process.The scope of psychology as the field dealing with the workings of the human mind ranges from cognition (perception, memory, learning, and problem-solving) to affect or emotion (motivation, attitudes), as well as personality (Jääskeläinen, 2012) .
Based on scope of psychological research on translation, there are different models that explain the mental processes carried out by the translators.For instance, one can point to Bell's linguistic and psycholinguistic model that accounts for, "translation in terms of information processing and requires both short-term and long-term memories for the decoding of source language input and the encoding of target language output."(Albir and Alvers, 2009: 56).

December 2016
Sri Lanka Journal of Social Sciences 39 (2) Kiraly (1995) presents another model which shows that translation process consists of controlled and uncontrolled, non-observable processes.Gile's effort model defines the process of translation as a complex cognitive process which has an interactive and nonlinear nature, encompasses controlled and uncontrolled processes and requires processes of problem-solving, decision-making and the use of strategies (Albir and Alvers, 2009: 63).
Considering these models, there can be a more obvious picture of the role of translator's own active abilities in processing the source text.As Shangarffam and Abolsaba (2009: 104) state "the translator has his own feeling about language and his translation".They declare that this feeling consists of intelligence, sensitivity and knowledge and the combination of all these feelings comes into play in the task of translation.Obviously, many factors can affect the task of translation and translators are different from each other in many aspects.It can be the best answer to the question why some translators are more successful than others in the task of translation.

Participants
A total of 104 female and male Iranian final-year undergraduate students of English translation participated in this study.Based on a criterion of sampling, only those students who were native speakers of Persian and knew English as a foreign language at an advanced level were invited to contribute.The participants were selected from three universities in Mashhad, Iran.The participants' age ranged from 21 to 37 (M= 23.40,SD= 3.47).

Narrative Intelligence Scale
Narrative Intelligence Scale (NIS) (Pishghadam et al., 2011) was one of the instruments employed in this study.This instrument consists of 35 items.A score of 1 to 5 is allocated to each item, which ultimately gives a score range of minimum 35 to a maximum of 175.Pishghadam et al. (2011) developed this scale following the guidelines proposed by Randall (1999) to ensure the content validity and used Rasch model to establish the validity of the findings.The results of the Rasch model indicated that all items, except for six, meet the unidimentionality criterion.This measurement produced an item reliability of .99 and a reliability value of .98.In this research, Cronbach's alpha for the entire scale was over .83.

Translation production test
A short passage of approximately 130 words in English on everyday life matters was used for translation from the foreign language (L2) to mother language (L1).The text was subjected to the review of a number of TS professors for face validity and was approved appropriate.

Criteria for translation analysis in terms of a narrative text
Within the context of translation analysis in terms of a narrative text, the idea of establishing some criteria for evaluating such texts is highly problematic.Whatever criteria one might consider for the purposes of achieving this are open to further scrutiny and questioning (Bamberg, 2012).The following represents some tentative suggestions that could be offered to assist the attempts to establish criteria to be used in the evaluation of narrative approaches to Translation Studies (all of which will be elaborated in more details in Results and Discussion section).In very general terms, as the cognitive linguistics enterprise centralises the fact that language is in service of "constructing and communicating meaning" and it opens a window to our cognitive processes (Fauconnier, 2000 as cited in Harrell, 2007), these tentative criteria are considered as language at the service of narrative, all of which have been used separately in the seminal works of some key figures in the field of narratology: Narrative coherence (Bamberg, 2012;Bruner, 1991;Pavlenko, 2007) 1.1 Cohesion building (Bamberg, 2012;Newmark, 1998) 1.2 The degree of task completion/thematic completion ( Martinkova, 2013).

Data collection
In this study, data were collected using the following procedures: Session I: First, a picture was given to the participants individually and they were asked to look at the picture for a minute, then the picture was taken and they were required to tell the story in Persian.Second, the sentence "please tell the story of your first day of the New Year" was used to elicit students' memory of this memorable day.This task is chosen for the following reasons: First, the chance to remember the memory was high due to the closeness of time between the New Year and the time of the interview.Second, the event is common and almost equally important for all the Iranians.Third, all participants share almost the same customs regarding the first day of the New Year.Therefore, what to say by the subjects was almost predictable and there remained how to say the story.
To collect the data, the researcher met each volunteer student individually on the first session; they told their narratives and they were recorded and later transcribed by the researcher.These narratives were then rated using NIS by two raters and the inter-rater reliability was estimated .91.
Session II: The participants were instructed to translate an English passage into Persian which would typically take 40 minutes.They were allowed longer stretches of time when necessary.

Data analysis
The collected data on NI tasks were scored by two raters.The raters of this test were the researchers themselves who were familiar with narrative theory proposed by Randall (1999) and the process of scoring.Then, two groups were formed.The first 20 participants with the highest NI score and the last 20 participants with the lowest NI score were determined and the performance of the two groups were qualitatively examined to find out the possible effects that NI may have on translation.Each of these texts was carefully examined based on the suggested criteria to clarify their narrative organisation.Two separate diagrams were used to show the results of the findings between the translation qualities of the two groups of participants with high versus low NI.To improve reliability, two raters were used to evaluate the translations in terms of the set criteria, and resorting to intersubjectivity, they compromised and would agree on the most reasonable evaluation and results.Both of the raters were already familiar and well-informed of the research area.Scoring disagreements were also settled by discussion and consensus between the raters.

RESULTS AND DISCUSSION
The researchers compared high and low NI trainee translators in respect of narrative coherence and other criteria as specifed below.

Narrative coherence
For assessing the narrative coherence of the translated texts, two separate means, cohesion and thematic completion were measured.As Pavlenko (2006) puts it, for making coherent narratives we need contextappropriate structures as well as cohesive links between clauses and sentences.In line with Pavlenko (2006), other celebrated scholars in narratology (e.g., Bamberg, 2012;Bruner, 1991;Fisher, 1985 as cited in Baker, 2007) have the same opinions on the total rationality and cognitive makeup of the texts.For any narrative to be understood, there must be a meaningful whole that conveys the essential meaning of a story (Baker, 2007).It closely resembles what Bamberg (2012) considers as a criterion for assessing the linguistic structure of a narrative that ultimately result in some structural whole.
To assess cohesion building, Halliday and Hasan's (1976) categorisation of cohesive ties were followed.To assess the degree of task completion in terms of its content and thematic totality, following Martinkova (2013), the Source Text (ST) theme was deployed.
The results of our evaluations through counting the number of cohesive ties in each translated text showed that the participants with high NI used more cohesive ties in their translations (M = 5.5, SD = 1.3) than the low NI group (M = 4, SD = 0.9).In terms of task completion, there were not any cases of incomplete translation among high NI participants, but there were 6 cases of incomplete translations among low NI participants; they left in the end of or in the middle of the task without providing a full content for the text.Figure 1 represents a schematic view of the findings.
The use of cohesive ties among high NI participants is higher than the mean number of ties used by low NI participants.According to Halliday and Hasan (1976: 4) the more use of cohesive ties warrants a more coherent text which has texture.To them, it is the texture that defines the quality of a text and by texture they refer to "Cohesive relations between semantic elements provide texture to text", and it is the construction of cohesion that leads to the construction of meaning.
In terms of the completeness of the translated tasks, while the high NI group did not leave any part of their tasks incomplete, six participants in the low NI group did not finish their tasks.To Martinkova (2013:168), one means of coherence of a text is guaranteed by thematic development of a text.According to Bamberg (2012), it is by striving for linguistic-cohesive and cognitivecoherence approaches that stronger cohesive ties and better coherence become in the service of easier and better comprehension.Therefore, the outcome of any translation by doing so appears as fully encoded information and meaning.Generally, here it is worthy to mention Bamberg's (2012) idea that the words, sentences and cohesiveness produce a true cognitive make-up for a text in the form of coherence.

The degree of text complexity/simplicity
By assessing the degree of simplicity/complexity of the translated texts, we measure the use of simple and complex forms of language (Sikes and Gale, 2006).Bamberg (2012) states that assessing the lexical and syntactic configurations of texts are usually the tools to analyse the linguistic build-up of a narrative text.Therefore, we aim to scrutinise the simplicity/complexity orientations in the translations.In this exercise, three separate linguistic measures in each text were measured: (i) translation length (ii) the number of simple/independent sentences (iii) the number of compound/complex sentences .
The findings of our study highlighted the following points.Firstly, it was found that the number of words used to translate the English text into Persian was 115.9 (SD = 19.5)among high NI participants, while the same measure in low NI group was 92.5 (SD = 19).Secondly, the results showed that the number of independent, simple clauses in high NI group was 4.5 (SD = 0.8), whereas the same measure for low NI participants was 2.4 (SD = 0.8).Thirdly, the outcome of counting the number of compound sentences was 2.5 (SD = 0.7) for high NI participants, and 3.5 (SD = 0.6) for low NI participants.Figure 2 provides a schematic view of the findings regarding the number of simple/ compound sentences between the two groups.
As claimed by Bamberg (2012), the more use of words signals the more meaningful texts.This is also in line with Baumer et al. (2005) study which noticed that the average length of thorough and meaningful among high NI participants, while the same measure in low NI group was (M = 92.5,SD = 19).Secondly, the results showed that the number of independent, simple clauses in high NI group was (M = 4.5, SD = 0.8), whereas the same criterion for low NI participants was (M = 2.4, SD = 0.8).Thirdly, the outcome of counting the number of compound sentences was (M = 2.5, SD = 0.7) for high NI participants, and (M = 3.5, SD = 0.6) for low NI participants.Figure 4.2 provides a schematic view of the findings regarding the number of simple/ compound sentences between the two groups.As claimed by Bamberg (2012), the more use of words signals the more meaningful texts.This is also in line with Baumer's et al., (2005) study who noticed that the average length of thorough and meaningful stories produced by participants in experimental group was significantly greater than those produced by the participants in the control group.The use of more words in translating a text can also be explained from the "explicitation strategy" viewpoint.Thunes (2011, p. 270) explains that "it is now generally agreed among translation researchers that translations are more explicit than non-translations, and this typically causes stories produced by participants in experimental group was significantly greater than those produced by the participants in the control group.The use of more words in translating a text can also be explained from the 'explicitation strategy' viewpoint.Thunes (2011: 270) explains that, "it is now generally agreed among translation researchers that translations are more explicit than non-translations, and this typically causes translated text to contain a larger number of words than the corresponding original text." However, in narratology, it is generally accepted that 'complexity of linguistic forms' can be a measure that provides the narratives in any form (written or oral) with added impact (Sikes and Gale, 2006).It is what Greetz (1973 as cited in Sikes and Gale, 2006) calls complex or "thick" descriptions in linguistic forms.Baumer et al. (2005: 548) showed that there exists an increase in the use of subordinate clauses in narrative production among experimental group participants who were wellfamiliar with narrative structure as compared to the control group.With regard to these scholars' conception of linguistic complexity and in line with some translation scholars view of complexity in translation (e.g., Thunes, 2011), our assessment of translated texts in terms of their linguistic complexity was defined in terms of the use of two simple and compound/complex kinds of sentences.
Accordingly, considering our analysis at sentence level, our findings show that the average number of simple sentences used by high NI participants was remarkably higher than the low NI participants' (4.5 vs. 2.4 sentences); and the average number of compound sentences employed by high NI participants was remarkably lower than the low NI participants (2.5 vs. 3.5 sentences).This contradiction between the research findings of some previous studies in narratology (Baumer et al., 2005;Labov, 1997;Sikes and Gale, 2006) and our findings in framing narrative features in translation provides a subtle and novel stream of debate.Though among narratologists it is generally accepted that the more complex a narrative, the more competent the narrator, we found that simplicity is a favored and highlighted trend among translators with high NI ability.
The reason for such an orientation has been partly answered by translation scholars from the view point of some translation strategies.According to Thunes (2011), in order to measure the complexity of a translation task, we must compare how source and target expressions correspond syntactically and semantically in respect of linguistic properties.From this viewpoint, our comparison showed that high NI participants have favored explicitation strategy by dividing nearly all source language (SL) compound sentences into two or more Target Language (TL) sentences with a view of providing a simpler text.This could be however related to what Newmark (1988) calls Natural Translation.To him, a translator's last resource and touchstone of his linguistic sensitivity is to write in a manner that is natural to him, a manner that expresses a good style.By this he means a translation which is written by observing the norms of target language grammar, idiom and words.In line with him, Gutt (1999) argues that a good translation reads like a target language original, not like a translation.He also states that a translation should be so natural in its style that it is indistinguishable from an original in the target language.
Similarly, the exact trace of SL structures (forms) in most of the translations done by low NI participants created some unnatural texts.We found that low NI participants had reduced language structures to the extent that most of them did not belong to our own language norms at all.There is an agreement among theorists that the reason for such a trend can be firstly attributed to interference from source language.As Mollanazar ( 2001) argues, interference is a universal fact in translation and adjustment can somehow solve this problem.Toury (1995) also reminds us that interference happens whenever translator fails to adapt translation to target language.The remedy for this failure is applying shifts in translation; a solution that appeared to be quite forgotten among low NI participants who were mostly tracing the structure of SL in their conduct.
On the whole, as Sikes and Gale (2006) cogently argue, confusing, boring and turgid use of words, an overreliance on technical terms and unnecessarily complicated and complex structures all tend to militate against a good narrative.While high NI participants tended to produce plainer texts, the low NI participants resorted to the very compound structures of the ST and produced hard-to-read texts at the expense of ignoring the norms of target language.The narrators whose narrative styles diverge from mainstream standard are often perceived as lacking Narrative Competence.The competent narratives produced by second language learners "are distinguished by appropriate lexical, discursive and register choices", while the weaker narratives "display insufficient elaboration and evaluation" in different areas of language use (Pavlenko, 2006 : 15).

Gaps in translation
By narrative gaps in a translated text, it is meant those situations in which the translators fail to provide a reliable account of the original narrative by leaving out December 2016 Sri Lanka Journal of Social Sciences 39 (2) some (lexical and/or syntactical) parts of the original texts untranslated.As Oslen ( 2003) puts it, these kinds of gaps are due to the narrator's insufficient ability to follow the norms and to provide the readers with a real and meaningful story.
Translations were evaluated on two kinds of gaps that could be obviously found in participants' translations, namely lexical and syntactical gaps.By lexical gaps, it is meant those "absent hypothetical words which would seem to fit naturally into the pattern exhibited by existing words" (Trask, 1993: 157).Here the term 'syntactic gap' refers to the missing surface-structure representation of a clause and its possible arrangement in a text.This study examines such unlicensed deletion of syntactic materials that have not been recovered anywhere in the text.We examine possible reasons for these gaps in the discussion section.
The study showed that only 3 participants in high NI group had lexical gaps, and these gaps were simply missed lexical items.These missed lexical items or syntactic gaps do not represent any lack of understanding or deep unfamiliarity with translation.In contrast, the same measures among participants with low NI were as follows: 9 participants had lexical gaps in their tasks, and 8 cases had syntactic gaps in their translations.Figure 3 presents this comparison.
The amount of lexical and syntactic gaps in low the NI group was significantly higher than in the high NI group.The evaluation of the performance of low NI participants showed that the elided lexical items (untranslated vocabularies) had largely led to syntactic ambiguity too.It is opposite to what is generally believed about ellipsis analysis in Narratology.According to Kennedy (2003), ellipsis does not have any syntactic representation and this technique does not distort any part of the text.Unlike the types of gaps introduced in Narratology that are all licensed gaps and are used as a method to bring about inference and to activate the audiences' background knowledge to make them able to grasp the suspended information and to produce engaging narratives, the gaps identified in the assessed translations mostly hindered inference, and they demanded an effortful reading to make them comprehensible.Another point worth mentioning is that the failure of the participants in low NI group to translate some ST lexis had directly led to a syntactic gap.Conversely, the same story among high NI participants did not lead to a syntactic distortion, and they were well-compensated without creating any further gaps.
In evaluating the reasons for the lexical gaps, Bentivogli and Pianta (2000) argue that discrepancies between source and target language items may be due to one of the followings: Syntactic divergencies, Lexicalisation differences, Divergences in connotation, Denotation differences.The assessment of our findings based on their categories showed that nearly all of the lexical gaps were due to lexicalisation differences; and the syntactic gaps were due to the syntactic divergencies.
The conclusions regarding lexical gaps in translated texts could be summarised as follows.First, lexical gaps are mostly due the translator's insufficient familiarity with semantic field of a term; however, one should not ignore the gaps arising from cultural differences between languages.Then, there seems to be an increase in making lexical gaps when it comes to the translation of some technical words as it was witnessed in our own study.
By narrative gaps in a translated text, it is meant those situations in which the translators fail to provide a reliable account of their tasks by leaving out some (lexical and/or syntactical) parts of their texts untranslated.As Oslen (2003) puts it, these kinds of gaps are due to narrator's insufficient ability to follow the norms and to provide the readers with the real and meaningful story.
In its linguistic terms, translations were evaluated on two kinds of gaps that could be obviously found in participants' translations, namely lexical and syntactical gaps.By lexical gaps in this study, it is meant those "absent hypothetical words which would seem to fit naturally into the pattern exhibited by existing words" (Trask, 1992, p. 157).Here the term syntactic gap refers to the missing surface-structure representation of a clause and its possible arrangement in a text.This study aims at finding such unlicensed deletion of syntactic materials that have not been recovered anywhere in the text, and finding possible reasons for these gaps in the discussion section.
Our findings showed that the numbers of participants in high NI group that had lexical gaps in their translated texts were 3 who had totally 5 cases of missed lexical items in their translations, and the syntactic gaps were 3 cases.On the opposite, the same measures among participants with low NI were as follows: 9 participants had lexical gaps in their tasks, and 8 cases had syntactic gaps in their translations.As findings showed, the amount of both lexical and syntactic gaps in low NI group was significantly higher than high NI group.The evaluation of the performance of low NI participants showed that the elided lexical items (untranslated vocabularies) had largely led to syntactic ambiguity too.It is opposite to what is generally believed about ellipsis analysis in narratology.To Kennedy (2003) ellipsis does not have any syntactic representation and this technique does not distort any part of the text.Unlike the types of gaps introduced in narratology that are all licensed gaps and are used as a method to bring about inference and to activate the audiences' background knowledge to make them able to grasp the suspended information and to produce engaging narratives, the gaps identified in the assessed translations mostly hindered inference, and they demanded an effortful reading to make them comprehensible.Another point worth mentioning is that the failure of the participants in low NI group to translate some ST lexis had directly led to a syntactic gap.Conversely, the same Creative/aesthetic merit Morgan et al. (2009) state that the quantification of the creative character of a narrative may not be so apparent.Therefore, it may be difficult to determine when someone has truly said something in a creative fashion.That is why studying creativity poses some very real methodological problems.In line with their idea, Newman (2007) considers assessing creativity in any writing as a quite subjective task.
Sri Lanka Journal of Social Sciences 39 (2) December 2016 reactions from the reader.In this respect there is a need to examine our translated texts in terms of its creative voice i.e. choice of vocabulary and syntax to generate a creative image of source text (Mozaffari, 2013;Sikes and Gale, 2006); and also to examine its style of presentation according to its text type (Holst, 2010).
Based on the short description provided above, here we tried to assess the creative side of each of the translated texts on a four point Likert scale as poor, fair, good and excellent.The findings assessed by two raters showed that among participants with high NI the results were as following: 1) excellent translations= 8 cases; 2) good translations=10 cases; 3) fair translation= 2 cases; 4) poor translations= 0 cases.In contrast, the results of low NI group were as follows: 1) excellent translations= 0 cases; 2) good translations= 6 cases; 3) fair translation= 7 cases; 4) poor translations= 7 cases.Figure 4 summries these findings.
The linguistic changes through the use of some translation strategies namely, explicitation and addition, have produced a creative impact in the translations of high NI participants, while the translations of low NI participants were mostly a direct transfer of source text materials.However, the evaluation of the level of creativity cannot be done only by intratextual factors; extratextual factors are to be considered too (Nord, 2005, as cited in Holst, 2010).One prime extratextual factor here is the text type.The text type can be categorised as informative text type; the function of which is to communicate information, knowledge and opinions.In line with Liu (2004), even if a narrative is meant primarily to communicate information and knowledge, such as narratives that belong to the genre of news, a good narrative nevertheless obeys an aesthetic imperative, though an invisible one, governing aspects of the narrative such as, inter alia, selective articulation of details, consistency and coherence, dramatic storytelling technique and the perception of authenticity.Crafting In order to overcome and to maximise the objectivity of the study, some relevant criteria for assessing translations as narratives were set.According to Newmark (1988), creativity in translation involves interpreting the sense of words, not translating the words themselves, and he regards it as a touchstone of translators' linguistic sensitivity and creativity.In this regards, Jiang (2008) states that it is the task of the translator to use possible linguistic means to reconstruct the aesthetic aspect in the target language.
In narratological terms there can be a correlation between these two fields regarding creativity.Mills (2006: 12) maintains that "no one can write anything of significance, unless imagination is allowed to play a part in the process".Liu (2004) also claims that narratives are not only communicative; they also have aesthetic values to make them be heard and be appreciated by others.For this aesthetic value to be created, psychological procedures come into play and appear in the surface of the narrative.Jiang (2008) argues that in order to create aesthetics in translation, meaning making and experience upon meaning leads to the establishment of a mutual relationship between translator-reader.This translatorreader deal reminds what Holmes' (1988) map suggests in translation process while he says that, first the translator as a reader understands the source text and then develops another map to create an aesthetic image of the source language text.In a similar vein, Rosenblatt (1978) refers to the fact that when the reading transaction is fluid and coherent, the result is an aesthetically pleasing outcome which is also unique and personal to its reader.When a reader no longer struggles to decode meaning of common passages, he or she becomes mentally free to form a personal response to texts.Therefore, it is necessary to examine any narrative text in terms of its ability to invite easy responses and to elicit multidirectional, multi-voiced and multi-punctual sense, opportunities are created for a pedagogic practice that is in tune with the demands placed upon youth.Thus, schooling and training the translators on how to develop such abilities seems to be necessary for them to enhance their communicative capacity.
Second, in the context of translation education courses in universities, there seems to be a gap in the syllabus of translation training programs with regard to how to develop and boost cognitive abilities, including NI, among students.The main objection to the rational curriculum traditions in educational contexts is that training programs tend to be very inflexible over time and, therefore, have a limited capacity to adapt to possible changes in the context.
Further, teacher training courses on translation can also enjoy the fruitful results of applying the dynamics of NI in their education; and they can become more practice-based rather theory-based.The designers of teacher training programs shall be open to NI issues by using and building NI into virtual learning environments and shall aim to develop NI abilities in their trainees.Therefore, for successful communication, the translators will acquire an acceptable amount of NI.
In general, narratives play a crucial role in how young children become socially skilled individuals with an autobiography, being able to effectively communicate with others (Engel, 1995, cited in Dautenhahn, 2001;Nelson, 1993).Therefore, the ability to compose a coherent narrative comes before and predicts successful access to literacy at school.A poor oral narrative skill in pre-school periods is a predictor of difficulty with early literacy skills.On the other hand, one might claim that oral narrative skills in the early years predict academic progress.In consistent terms, Baker (2007: 169) argues that narratives are, "constructed -not discovered-by us in the process of making sense of reality, and they guide our behavior and our interaction with others".This means that this attribute of human mind can develop from childhood to adulthood and it can be boosted from early stages of language learning.Therefore, the better and the more improved Narrative Competence, the more successful an individual in conducting any sort of human communication such as translation.
For practical reasons, this study was limited to the evaluation of translation performance of Translation students in three universities in Mashhad, Iran.As a consequence, the conclusions made in the end of the these aesthetic aspects of narrative is not only the trade of the storyteller, but also her art.It was concluded that although the constraints of the text type have imposed some limitations on the aesthetics, the general concerns of meaningfulness, coherence and consistency, which are all overlapping concepts in narration, cannot be ignored among high NI participants' performance.
In comparison to informative text type, expressive text type seems to provide more room for creativity and it is more congruent with this narrative feature since it is more target language oriented (Munday, 2008), and the translator is bound to recreate the image and the abstract nature of a literary work.While nearly all of low NI participants had tried to translate the informative text quite formally, Koller (1995) states that it is the aesthetic value of a "literary text" that can be best produced using "formal-aesthetic equivalence".In a study by Safinezhad et al. (2015: 78), they point out the fact that literal translations do not afford to fill the gaps of a target text that were an inherent part of it and they do not tend to give any interpretation of the source text; therefore, to them, this method of translation is specially for literary texts by means of which the aesthetic value can be preserved.
On the whole, it can be stated that in order to produce an aesthetic/creative text, a translator is confronted with the parameters of choices made in the original text.These choices can serve as important criteria for making decisions about the translation, and serve as a standard for making appropriate choices and hence for reflection on the style of the work (Jiang;2008;868).Therefore, it is through this meaning comprehension that the aesthetic competence of mind flourishes to recreate an aesthetic experience.To Jiang (2008), the aesthetic quality originates in a creative process where the activity involved attempts to picture a novel scenario.The translator, as an author in his re-creative activity, has to articulate the life of the mind.However, in assessing creativity of translated texts, one should not neglect the purpose of the translation and text type.

CONCLUSIONS
The results of this study have important implications for prospective translators, educational policy makers, translation curriculum designers, translation material developers and translator trainers.First, the central role of narrative in cognitive and mental processes involved in translation became evident.Dobson (2005) argues for Narrative Competence as an underlying skill neglected in educational policy makers' call for enhanced literacy.He claims that when narrative is understood in paper are limited to the target population.However, they may be generalised to all the Translation students in the context of Iran because the students in these three universities come from different cities in Iran.The study also limited itself to assess the translation quality of the participants from English to Persian only and the opposite directionality was not covered in the present study.The study used delimitation to select a non-random sample of Iranian senior undergraduate students of English Translation, and as such, the conclusions might not apply to students at other levels or professional translators and for that matter those in other contexts and language combinations.

Figure 1 :Figure 2 :
Figure 1: Narrative coherence in translated texts Figure 4.3 gives an overview of this comparison.

Figure 4 . 3
Figure 4.3 Explored Gaps in Translated Texts in Linguistic Terms

Figure 3 :
Figure 3: Explored gaps in translated texts in linguistic terms

Figure 4 :
Figure 4: Narrative creativity/aesthetic merit in translated texts