Analyzing Lexical Items and the Thematic Structures of Two Translations of to Have or to Be by Eric Fromm

Mohammad Amin Mozaheb ✉ Amir Ghajarieh and Heshmat Sepehriniya Assistant Professor, Department of Foreign Languages, Language Center, Imam Sadiq University, Tehran, Iran Assistant Professor, Translation Studies Department, Ershad Institute of Higher Education, Tehran, Iran MA in Translation Studies, Ershad Institute of Higher Education, Tehran, Iran ✉Corresponding Author: Mohammad Amin Mozaheb, E-mail: dr.mozaheb@isu.ac.ir


Introduction 1
In the globalization age with multi-voice nationalities and cultural differences, translation becomes a necessity more than ever. Translation plays different roles and purposes based on the requirements of society. There are different text types and genres in that each one specifies a specific approach to translation. Translation theories as what is expected by scholars most of the time are not in line with procedures applied in translation by translators and it seems that there are some gaps between translation theory and procedure. One of these gaps may be because of using different strategies used by translators. It is supposed that these gaps can be the result of a lack of consideration of different aspects of translation, cultural differences, and improper selection of translators (Vermes, 2001).
However, as Munday (2016) mentions, different models proved to have some drawbacks, and critics wrote some negative marks about them. Recently, another model has been proposed by Mona Baker in her influential work, In Other Words, in which she has proposed a typology for investigating the translation of marked and unmarked sentence structures (Baker, 2018). In the translation studies field, translation quality has always been a significant aspect (House, 2015). The question of "what is a good translation" has been investigated significantly through translation history. However, through some research, it is obvious that different approaches by different scholars have been introduced. The application of these models on translated texts is significantly important and can be used in fields such as translator training. As a result, the current study tried to apply House's model of TQA on two translations of technical terms chosen from the original book To Have or To Be and its two Persian translations.
Despite the importance of quality assurance in translating different literary, psychological, and philosophical texts, to the best of knowledge, there is little, if any, research conducted to evaluate the Persian translation of To Have or To Be in a quality assessment way. Consequently, the current study aimed at applying two models of Baker (2018) and House (2015) on two translations of To Have or To Be by Eric Fromm in order to investigate which of the translations have been able to translate the source text and its marked thematic structures better.

Statement of the Problem
The translation is a key to the new world and needs a reliable assessment tool, as its main issue is how to assess the quality. An important factor in the translation process is exploring the way and the methods used by translators for solving translation problems. In this context, even a number of approaches and techniques have been proposed (such as modulation, shifts, etc.). This shows that this topic is still an open question in translation studies and more and more studies must be conducted so that this issue can be examined in detail. As a result, this study also tried to use the framework proposed by Baker (2018) and House (2015) and applied it to two translations to examine the quality of translation as well as thematic structures in both source text and its target texts.

Research Questions
Considering the objectives above, this study attempts to answer the following questions: Q1: How has each translation dealt with technical terms in terms of covert and overt translation models? Q2: In terms of marked structures, how has each of the two translated versions approached the translation task?
Q3: To what extent do such translations affect the quality of the translation task?

Review of the Related Literature
In the field of translation quality and, among others comparative TQA models, House's (2015) model attracts the greatest attention. In her model, the overall focus is on source text and its analysis. She divided her model into three views; mentalist view, responsebased view, and discourse and text-based view which have been discussed earlier in this chapter. She also added three sorts of meaning which are semantic, pragmatic, and textual meaning. She believes the final target of the translation is functional equivalence and in order to achieve this target, situational dimension and linguistic material were defined. She also used three basic criteria to systematize different approaches including (i) the relationship between the original text and how it is perceived by the author, the translator, and the recipient; (ii) the relationship between the original text and its translation and (iii) the consequences which view about these relationships have when one wants to distinguish a translation from other types of multilingual text production (ibid, p.8).
She considers that in order to evaluate a translation, the assessor needs to distinguish mismatches between two texts, which are overtly erroneous errors and the other covertly erroneous error. If a translation text is to achieve the basic need of a dimensional, and as a result of this, a functional match, after that any mismatch among the dimensions is considered as an error. Dimensional errors like these were considered and referred to as covertly erroneous items. These kinds of errors were different from those overtly erroneous items that are caused either by a mismatch of the denotative and lexical meanings of source and translation text elements or by a breach from the target language system (House, 2015, p. 33).
House uses a socio-semiotic approach for the assessment of translation quality by classification of the product of translation into two kinds: Overt Translation and Covert Translation. Overt translations are those related to the text which includes the contexts of political texts, religious texts, literary texts, etc. while covert translation includes science, business, tourist information booklet, an economic Text, journalistic text, advertisement text, etc. (ibid, p.57).
According to Mehawesh (2014, pp. 92-93), House's assessment is based on the similarity between the ST and the TT in terms of register variables, the genre, and the ideational and interpersonal meanings. In her approach, she tries to equate the source and target context of the situation through the dimensions mentioned below: geographical origin, social class, time, participation, social role relationship, social attitude, medium, and province. The base of her model is to match function with function translators (like speakers) chosen from a paradigmatic network, their selections are constrained by different kinds of factors, and it can be said that most of them originate from House's dimensions. For not selecting wisely, there will be mismatches between ST and TT (ibid).

Word Order and Markedness
One of the most frequent issues within the realm of information structure, which has drawn a great deal of attention from researchers, is that of markedness. As Bloor and Bloor (1995) argue, "Markedness is a concept which is useful in the language study as a whole and not just with respect to the information and thematic structure" (p. 82). As one of the prominent figures in this field, Baker (1992) asserts that the markedness theory has been highly explored and used in different fields of linguistic studies like semantics, phonology, morphology, and syntax since the last century; however, in this study syntactic markedness is more emphasized rather than other types. The issue of syntactic markedness is interconnected with that of word order which "is very important in translation because it plays an important role in maintaining a coherent point of view and orienting messages at text level" (Baker, 1992, p. 110). Firbas (1992) argues that everybody feels instinctive that there is a kind word order that is like a default for declarative clauses, a word order which we use unless there is some logical reason in the context for using a different word-order known as the unmarked word order. In explaining his principle of emphasis, Firbas (1992) alludes to this addition by saying that the unusual order is done to show an additional communicative purpose not served by the usual order and is in this sense marked.
Regarding the explanation for markedness and unmarkedness, Greenberg (1966, p.34) assigns the words "marked" and "unmarked" to opposing structural entities that show a consistently asymmetric relationship in terms of distribution and/or paradigmatic complexity and/or syntagmatic structure, of these two entities the one that is always more widely distributed and/or it is better to say simpler is called "unmarked"; its complement is the " marked" member of the opposition. Schmid (1999) whose main concern in her book about translation called Translating the Elusive is the extent of translatability of different marked structures from English into German, delivers and delineates a comprehensive discussion of different constructions of non-canonical word order and examines both the structure as well as different discourse functions of these marked forms. She first tries to make a formal distinction between two main types of non-canonical structures: According to Schmid (1999), "There are two notions of word order variation: a marked constituent order that is different from SVO (Subject-Verb-Object), e.g. subject-verb inversion, and sentences that are made by constructional rather than just ordering operations under preservation of SVO e.g. left-dislocation"(p.49).
In a study done by Heidari-Tabrizi, Chalak, and Taherioun (2014) aiming to assess the quality of Persian translation of Orwell's (1949) Nineteen Eighty-Four by Balooch (2004) using the House's TQA model, they concluded that because of dimensional mismatches and a large number of main overt errors like omissions and substitutions, the translation was not in accordance with the House's view meaning that literary works need to be translated overtly.
In other studies similar to one aspect (translation quality assessment) of our study carried out by Yamini and Abedi (2010), they assessed the quality of Persian translation of William Shakespeare's Macbeth by Ala'uddin Pasargadi. The results of the used statistical procedure indicated a statistically significant difference between the overt and covert errors and among the five types of overtly erroneous items. Therefore, it can be said that this translation wants in compliance with the hypothesis suggested in House's Model that "a literary work has to be translated overtly and any deviation of it will be considered as an error". So, it can be concluded that this translation was a covert kind of translation, not an overt one.
Regarding the comparison of thematic structures in source texts and their translations in the target texts, Safaee, Saedi, and Rahim Pour (2003) investigated thematic structures and poems. The researchers wanted to show if the mishandling of marked themes in the translation of poetic texts can have any effect on the nature of the message to be negotiated or not. To examine this hypothesis, a test on some English poems containing marked themes was given to the students of English literature, and the translation of the same poems but with mishandled marked themes was given to the students of Persian literature. The statistical analysis on the results, unlike the analysis we did on the extracted poems, proved that mishandling of marked themes hardly had any significant effect on the reader's interpretation of poetic texts.

Methodology 3.1 Corpus
As was mentioned before, the aim of this study as corpus-based research is evaluation of Persian translation quality of To Have or To Be based on House (2015) TQA model. It also examines the use of marked thematic structures in source and target texts. Two Persian translations of To Have or To Be which were selected for this study are the one by Ahmad Sabouri and the other by Akbar Tabrizi.

Theoretical Framework of the Study
The theoretical framework of this study includes two parts. The first one is about translation quality assessment based on House's model (2015) and the second one is about thematic structures suggested by Baker (1992) that are explained in the following one by one.

House TQA Model
According to House's (2007) TQA model, there are two types of translation based on the discourse functional equivalence: Overt Translation and Covert Translation.
As House (1997) puts it, any mismatch between the ST and the TT is considered an error. Therefore, there are two kinds of mismatches: 'covertly erroneous items' including any dimensional mismatch like neglecting the features of the field, mode, and tenor during translating), and 'overtly erroneous items' (the mismatches of denotative or literal meanings of ST and TT like addition, deletion, mistranslation, breaches of target language scheme, etc.).
In House Model, first of all, the text needs to be classified into its relevant category. It means, is it overt or covert? Second, both the TT and ST text need to be scrutinized in order to distinguish Field, Tenor, and Mode. We need to assess this division from the sight of the syntactic, lexicon, and textual features, finally, the amount of mismatches between ST and TT distinguishes the quality; the more mismatches refer to lower quality of translation while fewer mismatches refer to high quality (House, 2007).
The most important elements in the House model of translation quality introduces as the register categories:

A. Field
Field refers to the nature of the social action that takes place; it captures "what is going on". As House defines, the dimension of Field captures the topic, the content of the text or its subject matter, with differentiation of degrees of generality, specificity, or 'granularity' in lexical items according to rubrics of specialized, general, and popular (ibid).

B. Tenor
Tenor points to the nature of the participants, the addressees, and the addresser, as well as the relationship between them in terms of social power and social distance, as well as the degree of 'emotional charge'. There are some other parts related to the tenor like the text producer's temporal, geographical, and social provenances as well as his intellectual, emotional, or affective stance (his 'personal viewpoint') (ibid). Under tenor and participation dimensions of language, there are other categories as below:

Social Role relationship:
House described this dimension as the analysis of the role relationship between addresser and addressees. This relation can be either symmetrical (marked by the existence of solidarity or equality) or asymmetrical (marked by the presence of some kind of authority) (House, 2007, p.29).

3.3.2.2.-Social attitudes:
As house described social attitudes are the degree of social distance or proximity resulting in relative formality or informality, in this way she adopted the distinction between different styles suggested by Martin Joos (1961), which consists of five different styles or levels of formality: frozen, formal, consultative, casual and intimate (ibid).

C. Mode
As summarized by House (2001) "Mode refers to both the channel-spoken or written, which can be "simple," i.e., "written to be read" or "complex," e.g. "written to be spoken as if not written", and the degree to which potential or real participation is allowed for between writer and reader "(p. 248).

Baker's Marked Thematic Structures
The model of analysis for thematic structures is the Hallidayan linguist's taxonomy of marked structures (marked themes) as proposed by Baker (1992). In her model, she has put forward three different major types of marked themes in English including fronted theme, predicated theme, and identifying theme (Baker, 1992, p.132).

Fronted Theme
In their definition of fronting, Greenbaum and Quirk (as cited in Baker, 1992, p. 132) state that it is the ability of marked theme to move into initial position an item which is otherwise unusual there". Furthermore, it can be divided into smaller subcategories like fronting of time or place adjunct, fronting of object or complement, fronting of predicator that depends on the item that is moved into clause-initial position. To clarify different kinds of fronting, consider the following sentence (p.146). Example: The book received a great deal of publicity.

Fronting of Time or Place Adjunct
As it is clear from its name, here, the adjunct is fronted or thematized. Of course, we should know that although this is a marked sentence, its degree of markedness is not very high since both time and place adjuncts in English have a high level of mobility within the clause and are common in theme position. E.g. In China, the book received a great deal of publicity.

1. B. Fronting of Object or Complement
Moving objects and complements into the beginning position is much more marked than fronted adjuncts of time and place in English because the position of object and complement is much more restricted. Moreover, as Baker (1992) asserts, when the object is brought to the sentence-initial position, it foregrounds the expressive meaning of the utterance as in the following examples: Object: A great deal of publicity the book received in China.
Complement: Well-publicized the book was.

C. Fronting of Predicator
According to Baker (2018, p.140), this kind of markedness is the least common and most marked theme in English. By fronting the predicator, a great deal of emphasis is conveyed as in the following example: They promised to publicize the book in China, and publicize it they did.

Predicated Theme
Considering predicating a theme, according to Baker (2018), this marked thematic structure is one of the most common ones in English. It involves using it -a structure that is also called a cleft structure to place an element near the beginning of the clause, as in It was the book that received a great deal of publicity in China.

Identifying Theme
This kind of thematic structure is the same as the previous one by a small difference. Here we make use of a Wh-structure instead of an it-structure. Identifying themes is also known as pseudo-cleft sentences. Halliday (1994, cited in Baker, 2018 argues that clefts and pseudo-clefts both are used to indicate a kind of contrast and to add "a semantic component of exclusiveness: the meaning is "this and this alone"(p.147). E.g. What the book received in China was a great deal of publicity

Design of the Study
This study has a quantitative design and the focus is to find two kinds of errors which are absolutely the result of the difference between the source language and target language. Also, this study tried to find out about marked thematic structures of the source text and their translation in the target texts. The applied models in this study are two: House TQA Model for assessing the quality of translation through covert and overt errors and Baker's model for comparing thematic structures. This thesis is generally corpus-based research. Primarily, the source text was read carefully, and then in order to find the types of errors (which are overtly or covertly erroneous) and thematic structures (marked or unmarked), the source text was compared to its translations. Since the study has used descriptive statistics of percentage and frequency, it is a kind of quantitative method.

Procedure
As the study covered just one book of two translations for this study, the researcher tried to check the entire text of both source text and target texts for finding enough examples to be compared. Doing the research, all pages of the original book and the translated texts first were read exactly. Next chosen examples, better to say the found examples, along with cases in their related target texts were written t o be examined carefully. It should be n o t e d that sampling was done based on English source text and not the translations. There were apparent irregularities between the partition and the number of pages in English compared to its Persian translation divisions.
In short, in order to evaluate Persian translations of To Have or To Be, in this research House's (2015) TQA model and Baker's (2018) thematic structure model were applied and the following steps were taken; For translation quality assessment (TQA) 1. A register analysis was done in order to be able to make an ST profile; 2. A complete description of the ST genre realized by the register was added; 3. A statement of function was made for the ST including the ideational and interpersonal component of that function, here the information and relationship between sender and receiver were included; 4. The same descriptive process was created for the TT as above; 5. A comparison between ST and TT profiles was made and a statement of mismatches of errors was produced and categorized based on the genre and situational dimension of genre and register; the founded errors were divided into covertly erroneous items and overtly erroneous items; 6. A statement of translated quality was presented; 7. Finally, based on House's (2025) TQA model, the translation was finally categorized into one of two types; overt translation or covert translation. If the translation has a lot of overt errors but no or few mismatches in tenor, field, or mode, it is considered to be more of a covert translation and if it has some mismatches along the field, tenor and mode but few or no overt errors it is more of an overt translation.
For thematic structure comparison 1. Both source and target texts were read exactly. 2. Thematic structure cases suggested in Baker's model in both ST and TT were extracted and written in tables to be compared.

Comparison between ST and TT profiles was done and a statement of mismatches of thematic errors was produced
and categorized based on the thematic structures mentioned in the model. 4. Extracted examples in both texts were described and explained. The translation results were finally classified as marked or unmarked thematic structures and their frequency and percentage in both ST and TT were tabulated for more clarification and further discussions.

Results
Considering all the data compared and examined in the tables above, comprehensive results are explained in two parts including results for covert and overt errors based on the House TQA model and marked thematic structures based on Baker's model for source text and its two translations using tables of frequency and percentage as well as bar charts.

Results of Comparison of ST and its TTs Based on House TQA Model
As it is shown in Regarding, the Breach of the target language system the number of all the errors found were 13 and 14 in TT1 and TT2 respectively. Considering omission error, the total number was respectively 2 and 15 for TT1 and TT2. The rarest errors in this study were respectively addition with a number of 1 and 10 in TT1 and TT2; a slight change in meaning with the number of 0 and 5 in TT1 and TT2; and phonetic transference with the number of 2 and 3 in TT1 and TT2. As it was mentioned before, four sentences extracted for comparison haven't been translated in Mr. Tabrizi's translation. In short, mistranslation, breach of the target language system, and omission were errors with the highest number of occurrences and addition, phonetic transference, and a slight change in meaning with the lowest number in both translations especially TT2. Considering items without any errors or error-free items found in this study, the number and percentage of these items for the translation done by Mr. Sabouri were respectively 76 and 76%. As it was shown in the above table and the following graphs, the number and percentage of error-free items for a translation done by Mr. Tabrizi is 27 and 27% respectively.  Comparing translation of To Have or To Be by two translators Mr. Sabouri and Mr. Tabrizi in the study, as illustrated in figures 1 and 2, the results of the data analysis showed that there was a considerably higher percentage of erroneous items compared to error-free items in the translation done by Mr. Tabrizi compared to his counterpart Mr. Sabouri. The total number of errors found in 100 sentence items compared in the study for TT1 translated by Mr. Sabouri was 24 whereas this number for TT2 translated by Mr. Tabrizi was 73 which is considerably higher compared to his counterpart, Mr. Sabouri. As it is clear in Table 4.20, 73 percent of all examined items in this study were erroneous items (27% error-free) for Mr. Tabrizi while the percentage of these items for Mr. Sabouri was 24 percent (76% error-free).To sum up, the number of erroneous items in TT2 is almost three times as erroneous items in TT1.

Results of Comparison of ST and its Two TTs based one Baker's Marked Thematic Structures
The Markedness can be explained better on a continuum compared to as a binary concept. For the classification of Markedness used here in this study, sentences were divided into two broad classes of Unmarked and Marked. Of course, as was mentioned in previous chapters, some forms of marked structures are more unusual and marked compared to some others. The researcher has tried to clarify marked sentences based on formal and functional properties. Here the classification suggested by Baker (2018) was used for analyzing data since it was thought to be effective for the present study's purpose. Considering this fact the following data was obtained: Collectively, as it is shown in Table 4.21, in the original text or source text 109 cases of five marked thematic structures suggested in the model were found. In the translation by Mr. Sabouri (TT1), the proportion of marked sentences to unmarked ones was found to be 103 to 6 while this proportion for Mr. Tabrizi's translation (TT2) was 85 to 24. This clearly shows that marked cases are considerably more frequent than their unmarked counterparts in both translated texts.   It would be useful to report the frequency of different marked structures in comparison with one another. As in ST, in both TT1 and TT2, the fronted time/place adjunct and identifying theme were the most frequent marked categories by percentages of 49.95% and 28.44%; 40.36% and 22.93% respectively. The least frequent marked categories were fronted object/complements and predicated themes. The percentage of these two marked structures for TT1 and TT2 was 11.92% and 9.17%; 10.09% and 4.58% respectively. About half of the marked sentences fall within the category of fronted time/place adjunct. For the fifth category of marked structures discussed in this study, or it is better to say the most unusual and marked one in English, no case was found in ST to be compared by its counterparts in TT1 and TT2. The following figures and tables illustrate the results more.

Conclusion
It's clear that all the calculations and analyses yielded on the study's data aim to give the researcher the answers to questions raised considering the topic of study.
To answer research question one, the findings of the study including dimensional mismatches and overt errors should be taken into account in the translation of technical terms in both TT1 and TT2. The investigation and analysis of data revealed just a onedimensional mismatch but significant overt errors in both translations especially TT2. Considering covert errors, the translation indicated a change at the level of Register. The mismatch was located in Tenor, with the discrepancy between the author's provenance and stance as the writer, and that of the translator that his job is translating.
As far as the overt errors are concerned, they were carried out at the six levels of mistranslation, addition, omission, a slight change in meaning, breach from the target system, and phonetic transference. Following a comparison of one hundred examples from ST and its TTs equivalents, the translation quality assessment of To Have or To Be not only recognized the problematic areas including mismatches and errors in this particular translation but also revealed some important results about the translation of technical terms in psychological works.
Considering all the examined items in the study, as it was mentioned before, error-free expressions in TT1 and TT2 were 78, 78%; and 33, 33% in frequency and percentage respectively. In other words, the translators' translation strategies could be revealed through the analysis of dimensional mismatches in general and analysis of overt errors in particular. According to the analysis, the most frequent errors were mistranslation, breach of TL system, omission and addition in TT2 with a larger frequency of errors, and mistranslation and breach of TL system in TT1. Other mentioned errors were not that important. In short, considering a large number of overt errors compared to covert one in this study for translation of technical terms, as house (2015)   As far as research question number three, dealing with the effect of such translations on the quality of translation task is concerned, the criteria used here (House's model and Baker's model) and feature needed for overt and covert translation and marked thematic structures should be borne in mind. In spite of not paying attention to denotative meaning in some parts of translated texts here especially in TT2 by Mr. Tabrizi, the translator should take account of the denotative meanings of the words of the source text.
While translating, a translator should manipulate the text in a way that sounds to be reasonable. Regarding this translation, it can be concluded that both overt and covert errors would be considered in translation quality assessment. It means a good translation is one that takes into account both dimensional matches of field, mode, and tenor; and a low number of overt textual errors. The findings of this study can be helpful for translators of different works especially socio-psychological ones to get familiar with problems in translation of these books as well as strategies used to manage such problems. Also, translation editors can also identify and take into account the above-mentioned errors in their translation and remove them when reviewing and editing the translation.
Also, for marked thematic structures of non-correspondence, as Givi and Anvari (1985) state, "the justification may lie under the free word-order of Persian language"(p.52). Similarly, as Bayati Dosti (2015) in his paper about markedness in thematic structures argues, "Due to the relatively unfixed word-order of Persian language, the translators don't feel bound to a fixed word-order when translating from a foreign language with a more rigid word-order like English and consequently fail to produce the same degree of markedness"(p.67). As Baker (1992, p.146) mentions "marked structures are often normalized in translation and this is a universal feature of translation language entitled third code". Therefore it can be said that in the study conducted here, the translator's tendency to translate marked thematic structures into the unmarked ones is a characteristic of translation from one language into next and it is predictable to see that a lot of translators have an inclination to normalize marked structures in translation. Normalization can be the reason that both translators tried to respect the markedness correspondence between ST and TTs. Of course, the analyses clarified that the TT items that failed to keep the thematic structure of ST items didn't have the same communicative meaning as ST.

Implications
Taking into account the findings of the current study, the researcher suggests the following pedagogical guidelines: Firstly, teachers, instructors, professors, and would-be translators can find the results of this study as a source in their learning or teaching of the translation process.
Secondly, translation editors can also recognize the above-mentioned most frequent errors of mistranslations, breach of TL system, and omissions in Persian translations of ST in the conducted study and remove them.
Regarding translation used in this study, this study's findings may be a means to extract and indicate the translators' problems and their use of wrong strategies in translation and in this way help other translators especially new ones, to avoid them. Also, other critics can use it as a framework to address these types of analysis.