





































Education, Language and Sociology Research 

ISSN 2690-3644 (Print) ISSN 2690-3652 (Online) 

Vol. 4, No. 3, 2023 

www.scholink.org/ojs/index.php/elsr 

132 
 

Original Paper 

Research on Cognitive Translation Process: An Overview 

Xu Yuemeng1 & Liu Yongjie2 

1 Institute of Modern History, Chinese Academy of Social Sciences—Beijing, China 

2 Research Institute for Global Value Chains, University of International Business and Economics—

Beijing, China 

 

Received: July 10, 2023        Accepted: July 31, 2023      Online Published: August 19, 2023 

doi:10.22158/elsr.v4n3p132        URL: http://dx.doi.org/10.22158/elsr.v4n3p132 

 

Abstract 

Translation process is conceived as a cognitive activity and has been received less attention in early 

research due to the difficulty of lacking direct access to translators’ thinking process. But researchers 

still made attempts to describe this process. And hence this article reviews the previous research on 

cognitive translation process, in particular the research model. The article concludes that a descriptive 

approach to the investigation of the translation process should be adopted rather than a prescriptive 

approach. 

Keywords 

cognitive translation process, research models, descriptive approaches 

 

1. Introduction 

Traditional research focuses on two domains of translation practice: translation as a product and 

translator’s competence. The research on translation as a product has been largely conducted from 

linguistic perspective by comparison with source language texts, for which a considerable amount of 

different and highly theoretical models has been formulated with equivalence being their primary concern 

and neglecting other text features such as cultural elements. As stated by Peter Newmark (1988): 

The source language text consists of words, that is all that is there on the page. Finally, all you have is 

words to translate and you have to account for each of them somewhere in your target language text. (p. 

193) 

Though Newmark (1988) suggests to ‘translate words that are more or less linguistically, referentially, 

culturally and subjectively influenced in meaning’ (p. 193) rather than isolated words, he is in an attempt 

to achieve equivalence at lexical level. 

As for Eugene Nida (1964), he was in pursuit of ‘dynamic equivalence’ (p. 159) through his extensive 

translation practice of Bible. He argued that reader’s response to the translated text should be taken into 



www.scholink.org/ojs/index.php/elsr              Education, Language and Sociology Research              Vol. 4, No. 3, 2023 

133 
Published by SCHOLINK INC. 

consideration of the translation process, in that Nida realized that people from different cultures might 

respond differently to the same text. What Nida tried to achieve was that the translated texts should appeal 

to and produce some kind of emotional response on the part of the reader as that of the reader of the 

source language and culture to the source texts. The approach adopted by Nida, however, was still 

product-oriented though having taken cultural elements into consideration. Nida’s approach is not the 

only case; in fact, most of product-oriented models, if not all, are more prescriptive rather than descriptive. 

Such models draw heavily attention on linguistic elements in translation, neglecting the subjectivity of 

translators in the translation processes. The product-oriented study focuses on the analysis of the 

translated texts, which indeed reflects the translators’ thinking activities to some extent, but far from 

enough because the processes are much more complicated than the translated texts could reveal. 

Therefore, other approaches should be employed to explore the translators’ thinking processes. 

Apart from the product, translation competence is another focus of translation research, as translation is 

regularly considered as a professional practice requiring specific competence, i.e., a translator must 

possess certain expertise to fulfill the translation task and reach a successful outcome. Therefore, as a 

consequence of translation theory being product- and competence-oriented, the research on translation 

process has hardly drawn any attention before the 1980s. In other words, the Translation Studies have 

been only focusing on one aspect of the translation proper, because the concept of translation entails at 

least two aspects: the products of translating and the processes of translating. Therefore, previous study 

on translation can be said to be incomplete. 

Such a situation was not changed until 1980 when Ericsson and Simon (1980) first introduced into 

Translation Studies the empirical approach of verbal reports, which was borrowed from the discipline of 

cognitive psychology. The verbal report is a data collection method, with which researchers can finally 

probe into the translation processes of translators that was considered to be inaccessible to direct 

observations which is due the very nature of human brain, the mysterious ‘black box’. Though the method 

is not without flaws—the method is challenged by many researchers who consider the data collected 

from verbal reports are subjective and incomplete, and hence are invalid, however, the verbal reports in 

fact can provide enough accurate information for analyzing and describing the translation processes. 

Hence, this part of the paper intends to give an overview of previous studies on translation processes. 

 

2. An Overview of Previous Studies on Cognitive Translation Processes 

Tapping into the cognitive translation processes has been a difficult task to accomplish, as it is constrained 

by such intrinsic difficulties as any kind of cognitive processing being not amendable to direction 

observation. Therefore, researchers have put forward models to investigate and analyze the translation 

process. These models explain the cognitive processes from theoretical point of view in terms of 

analyzing the products, i.e., the translated texts instead of actual translation processes, therefore less 

convincing. In order to better understand cognitive translation processes, researchers turn from 

theoretical studies to empirical studies, which up till now have proved to be an effective approach. This 



www.scholink.org/ojs/index.php/elsr              Education, Language and Sociology Research              Vol. 4, No. 3, 2023 

134 
Published by SCHOLINK INC. 

part is in an attempt to discuss the status quo of research on cognitive translation processes. To begin 

with, a detailed introduction to theoretical studies on translation process, including two models proposed 

by Seleskovitch and Lederer (1968; 1975; 1981) and Bell (1991), was outlined, pointing out the 

achievements they have made. The models put forward by such scholars, however, are purely theoretical 

and speculative, lacking empirical and concrete evidence to support themselves, and therefore less 

persuasive, which led to the introduction of empirical approaches to the study of translation process.  

Pioneering research on cognitive translation process was conducted by Seleskovitch and Lederer, 

employing the interpretive theory of translation (ITT). The ITT divided the translation/interpreting 

process into three interrelated phases: understanding, deserialization and re-expression. According to ITT, 

understanding, as the beginning of translation process, requires both linguistic and contextual/cultural 

knowledge, for linguistic knowledge itself doesn’t suffice to decode the message contained in source 

texts. The translator or interpreter needs to comprehend the ‘sense’ in its totality of source texts that is 

resulted from the interdependence of all linguistic and non-linguistic elements of the understanding 

process. The ‘sense’ translator or interpreter acquires during the understanding is the non-verbal synthesis, 

which requires the intermediate phase of deserialization. For ITT, the deserialization is an essential part, 

as the re-expression is achieved through deverbalized meaning rather than on the basis of linguistic form. 

Re-expression, as the reverse process of understanding, is similar to the process of natural language 

producing, in which the speaker expresses the end product of non-observable thinking activities through 

linguistic forms. For ITT, the translator or interpreter performs the role of ‘non-observable thinking 

activities’ in the context of translation through understanding and deserialization, as a speaker does in 

the process of natural language producing, and then formulates linguistic expressions of the intended 

meaning of the sender of the source texts through re-expression. At the end of the whole process, the end 

product, i.e., the translated texts, is produced and will be delivered to readers of the target language. 

Though ITT outlines a sketch for translation processes, it is a theoretical model based on linguistic theory, 

lacking of hard evidence. 

Bell (1991) argues that the study of translation is best served by the construction of models of the process 

of translating, for which he proposed an integrated model of translation process by drawing on linguistic 

and psychological knowledge. By challenging the notion that laid the ground work for early attempts to 

formulate a theory of translation, Bell argues that the process, during which a translation is produced, 

and the translator, who performs the act of translating, should be included in the description of a 

translation theory. Since the translation process is carried out by translators, knowledge and skills of 

translators, translator competence, should be dealt with first before the investigation of the translation 

process, because it is the question that needs to be answered—‘What is it that translators need to know 

and be able to do in order to translate?’ According to Bell, a translator should be familiar with the 

knowledge shared by people of both target and source languages, and that knowledge should at least 

contain five aspects: target language (TL) knowledge; text-type knowledge; source language (SL) 



www.scholink.org/ojs/index.php/elsr              Education, Language and Sociology Research              Vol. 4, No. 3, 2023 

135 
Published by SCHOLINK INC. 

knowledge; subject area (‘real world’) knowledge; and contrastive knowledge (Johnson & Whitelock, 

1987, p. 137). Lack of knowledge means that the translator would be unable to perform the translation. 

The concept of information processing is central to Bell’s model, for Bell divides translation process into 

analysis and synthesis, which require both short-term and long-term memory of the translator to 

collaborate in decoding the text in source language and encoding the text into the target language. The 

analysis/synthesis is achieved through operations at syntactic, semantic and pragmatic levels, which co-

occur with the five stages during the translation process—parsing, expression, development, ideation and 

planning.  

Decoding process begins with text processing which is preferable by concentrating on the clause, and 

proceeds in both a bottom-up and a top-down manner. As is suggested by Bell, a visual recognition 

system is required to distinguish words from other linguistic forms in the source text, and prepares the 

source text, through identifying distinctive features of letters, for syntactic processing of the clause. At 

this stage, the syntactic structure of the clause will be analyzed to acquire the lexical meaning, and when 

the syntactic analyzer fails to pass through the frequent lexis store and the frequent structure store, it will 

recourse to the parser for analyzing difficult clause structure, or the lexical search mechanism for tackling 

the linguistic items that it is not familiar with. When the syntactic analyzer finishes decoding the source 

text, the process can proceed to the next stage, the semantic analysis.  

The aim of the semantic analysis is to acquire the meaning of the clause through each syntactic 

component, and to retrieve ‘transitivity’ relations that underlie the syntactic structure of the clause. 

During the semantic analysis, the semantic analyzer tries to figure out ‘what the clause it about; what it 

represents; logical relationships between participants and processes; ideational meaning; semantic sense; 

and propositional content’ (Bell, 1991, p. 53). Once the logical form of the clause is identified, the task 

is handed over to the final phase—pragmatic analysis. 

In pragmatic analysis, the pragmatic processor has two tasks to fulfill—they are isolating the thematic 

structure and providing a register analysis. The thematic structure is concerned with the distribution of 

information and whether it is in a marked or unmarked order. The register analysis of the clause is 

performed in terms of tenor of discourse, mode of discourse and domain of discourse. Tenor and mode 

provide style information while domain provides purpose, suggesting a speech act. The information from 

thematic structure and register analysis form a semantic representation supported by the idea organizer 

and the planner.  

Synthesis is a reverse process of the analysis. The semantic representation of the source text is 

reprocessed by the pragmatic processor of the target language to find their equivalent style and ways to 

express the speech act in the target language. And then semantic processor restores the semantic 

representation received from previous stage to logical forms and passes on to syntactic processor, which 

searches for suitable lexical items to encode in the target language through the writing system. Finally, it 

gives rise to a target text. 

 



www.scholink.org/ojs/index.php/elsr              Education, Language and Sociology Research              Vol. 4, No. 3, 2023 

136 
Published by SCHOLINK INC. 

3. Conclusion 

The models of translation processes that are proposed by researchers so far are product-oriented, 

therefore theoretical and speculative rather than empirical and concentrated on idealizations rather than 

on actually occurring data. Researchers speculate the thinking activities of translators during the 

translation processes based on the text analysis, such as comparing the source texts and the translated 

texts to identify the inappropriate translations of the translators, which can provide the researchers certain 

information about the processes. Therefore, the results from such research cannot fully reflect the actual 

translation processes, and hence more rigid and robust approaches were needed. As proposed by 

Bassnett-McGuine, a descriptive approach to the investigation of the translation process should be 

adopted rather than a prescriptive approach. The ‘descriptive approach’ means that the research on 

translation processes should be conducted via empirical approaches rather than theoretical approaches. 

 

References 

Bassnett, S. (2002). Translation studies (3rd ed.). London and New York: Routledge. 

Bell, R. T. (1991). Translation and translating: Theory and practice. London and New York: Longman 

Group UK Limited. 

Lederer, M. (1981). La traduction simultanée. Paris: Minard. 

Newmark, P. (1988). A textbook of translation. New York and London: Prentice Hall International (UK) 

Ltd. 

Nida, E. A. (1964). Towards a Science of Translating. Shanghai: Shanghai Foreign Language Education 

Press. https://doi.org/10.1163/9789004495746 

Seleskovitch, D. (1968). L’interprète dans les conferences internatinales: Problèmes de langage et de 

communication. Paris: Minard. 

Seleskovitch, D. (1975). Langage, langues et mémoire: Etude de la prise de notes en interprétation 

consécutive. Paris: Minard. 

 

https://doi.org/10.1163/9789004495746

