Cluster and derived equivalence in collaborative corpus-informed translation education

By Barbara Lewandowska-Tomaszczyk (State University of Applied Sciences in Konin, Poland)

Abstract & Keywords

This paper presents corpus-informed teaching practice and effects of online collaborative Native - Non-native (NS-NNS) translation tasks (Lewandowska-Tomaszczyk & Bogucki 2016; Lewandowska-Tomaszczyk & Slomski 2016) in the identification of clustering equivalence English-to-Polish and Polish-to-English patterns. The paper elaborates on the concept of cluster equivalence (Lewandowska-Tomaszczyk 2017), making reference to large monolingual and parallel examples and presents sets of translational clusters of equivalence patterns and collocational patterning, in which the phenomena of re-conceptualization, meaning approximation and displacement of senses are observed (Lewandowska-Tomaszczyk 2010).

Keywords: translator education, corpus-assisted translation, cluster equivalence, student collaboration, meaning approximation

©inTRAlinea & Barbara Lewandowska-Tomaszczyk (2019).
"Cluster and derived equivalence in collaborative corpus-informed translation education"
inTRAlinea Special Issue: New Insights into Translator Training
Edited by: Paulina Pietrzak
This article can be freely reproduced under Creative Commons License.
Stable URL:

1. Introduction

The paper presents a description of the processes and effects of online collaborative writing and translation tasks, originally discussed in Lewandowska-Tomaszczyk & Bogucki (2016) and Lewandowska-Tomaszczyk & Slomski (2016) in Native - Non-native (NS-NNS) student pairs for improving the identification of clustering equivalence patterns in practice. The analysis and discussion of the learner data aim “to create a corpus-informed teaching approach, develop adequate didactic materials, and foster discussion between practitioners and theorists in the field of translational education” (Lewandowska-Tomaszczyk 2016). 

The paper develops the concept of cluster equivalence introduced by the author in previous publications (Lewandowska-Tomaszczyk 2017), exemplified on selected, mostly law-related terms in English-Polish parallel materials and  subject to frequent clarification requests in NS-NNS contacts.  The argument makes reference to comparable and parallel examples and presents sets of translational clusters of equivalence patterns and collocational patterning, identified both in general language and LSP, in which the phenomena of re-conceptualization, meaning approximation and displacement of senses are observed (Lewandowska-Tomaszczyk 2010). 

The present study derives the language materials from English and Polish monolingual corpora, the British National Corpus and the National Corpus of Polish, and English-Polish and Polish-English comparable and parallel corpora, developed at the Department of English Language and Applied Linguistics of Lodz University, accompanied by Polish and English monitor corpora with daily updates and freely available collocation tools. The materials of student cooperative tasks are drawn from the Trans-Atlantic Project of cooperation (TAPP, see Maylath et al. 2013) between native American students of North Dakota State University and the Concordia College, Minnesota in the USA on the one hand and Polish students of two universities – the University of Lodz and the State University of Applied Sciences in Konin on the other (Lewandowska-Tomaszczyk & Slomski 2016) in the years 2015-2017.

2. Collaborative corpus-informed teaching

The concept of cross-linguistic cluster equivalence patterns to be developed more fully in the second part of the paper, appears to be useful in collaborative corpus-informed teaching tasks at MA translation classes between students of two USA universities (North Dakota State University and Concordia College, Minnesota) and two universities in Poland (University of Lodz and State University of Applied Sciences in Konin). The cooperation involved NS-NNS peer collaborative writing/translation tasks in the period of 4 semesters in the years 2015-2017. Apart from mutual peer correction tasks the Polish students performed analyses of the concordance materials in English and Polish monolingual corpora (British National Corpus and National Corpus of Polish) and Polish-to-English and English-to-Polish parallel corpora (Paralela). Furthermore, they generated and analysed collocational profiles of relevant items and performed a study on the syntactic/semantic construal (Langacker 1987) and re-concepualization types (qualitative analysis) in the selected classes of examples (Lewandowska-Tomaszczyk 2010).

The cooperation involved regular online collaborative tasks between Polish MA students of English and translation and American students of technical and engineering subjects and was helpful in meaning clarification and fuller comprehension both of the direct meanings as well as intended messages.

Out of the properties envisaged to be part and parcel of translational competence (Pietrzak 2013) such as those originally proposed by Neubert (2000), i.e.,  language competence, textual competence, subject competence, cultural competence and transfer competence, it is transfer competence, involving pattern-matching competence between a SL and a TL, accompanied by a decision-taking competence  as well as performance competence, i.e., ability to perform in consequence of pattern-matching ability and conscious decision making that are the most significant criterial properties. The social-psychological competences leading to development of higher criticality and  self-authorship (inter alia self-regulation) are also particularly conducive to the development of translator collaborative competence.

3. Peer corrective feedback

As one of their major tasks the Polish students in the collaborative TAPP project were to choose quality press articles, translate them into English and send to US pairs appointed by the lecturers. American students sent their feedback and upon correction, the Polish students sent their modified versions back to their American pairs. There were several, typically 2-3 turns, of consecutive revised and modified drafts and Question-Answer turns  between US and Polish  students.  The final versions of the English texts were sent to the American students and both teachers. During the same period of time the American students prepared their authored texts and sent them off to the Polish students, who commented on the texts and asked questions. The American students modified their texts and/or provided additional clarification.

All students got additional feedback in person from their University tutors either during meetings or at the consultation periods. All Polish students’ English language proficiency level, monitored at the seminars and classes of practical English grammar, speaking and writing/integrated skills, ranged between B1 and C1 levels (by The Common European Framework of Reference for Languages), i.e., between so-called independent user’s threshold or intermediate proficiency or vantage or intermediate proficiency in English on the one hand and effective operational or advanced proficiency on the other. All of the American students are native speakers of English, first or second generation college students, generally of middle class background and they are most commonly majors in engineering, architecture, and nursing.

The highest frequencies of Polish student feedback referred to terminological clarification requests. In the section to follow we will first develop terminological definitions of the theoretical concepts used in the paper and present instances of requests for the clarification of legal terminology and their effects, aided by corpus-informed teaching classes for full understanding of the direct and implied meanings of the texts.

4. Meaning approximation versus sense selection

The focus of this section is translational clusters of equivalence patterns and collocational patterning, in which the phenomena of re-conceptualization, meaning approximation and displacement of senses are observed (Lewandowska-Tomaszczyk 2010). Concepts of approximation of meanings, re-conceptualization cycles (Lewandowska-Tomaszczyk 2010, 2016), and inter-language displacements constitute the foundation of the development of cluster equivalence patterns, observed at the interface of Source Language and Target Language. However, it must be observed that cross-language relations constitute a double-faceted phenomenon.

On the one hand, clustering  involves the option of looking for equivalence in clusters of similar, approximate forms across languages rather than in word-for-word equivalents (Lewandowska-Tomaszczyk 2017, 2017a). And yet, when needed, required or wished for, such clusters enable the translator to search for specific wording in the TL text and to employ a selective rather than cumulative technique from the cluster in order to take care of the, subtle and yet meaningful, differences between various options of construal of the same scene or event.

5. Language and extralinguistic world

It is a truism to say that language is not directly linked with the extralinguistic world. However, the question remains with regard to the types and forms of relationship between language and outside reality. Following cognitive linguistics I assume a relation between the world and language in terms of the process of mental imagery construction, which combines both language users’ linguistic and encyclopaedic knowledge,  their intentions,  expectations and preferences.

Meaning approximation

Communication, as was mentioned in Section 3., relies on approximate (and not identical) meaning sharing (Lewandowska-Tomaszczyk 2012b), both in the process of monolingual  discourse and, even more clearly, in translation. A Source Language system is never fully equivalent to a Target Language, so translating necessarily involves a number of cycles of re-conceptualization of the SL message (Lewandowska-Tomaszczyk 2010). This means, that the (language-specific) construal (syntactic packaging) of an event (Langacker 1987), i.e., a particular point of view, a vantage, in terms of which the original scene is conceptualized, contributes to a large extent to a modified L2 scene construal. In other words, a particular syntactic structure, construes the event scene and contributes to the meaning of the relevant linguistic expression. There are obviously other reasons for the reconceptualization phenomena. Speakers’ language proficiency, their volitional and deontic purposes (i.e., their will and obligations what to say and how to phrase it) as well as large contextual and interactional factors play a role here (for details see Lewandowska-Tomaszczyk 2017a). 


As argued above, languages portray different pictures of outside reality. In their structure, languages do not possess identically structured categories to ensure complete representational identity in meaning transfer. George Lakoff (1987:322-323) introduces the concept of commensurability and cross-language commensurability criteria to measure cross-linguistic similarities and distance, which cover the following:

  1. truth-conditional criteria [classical translatability],
  2. criteria of use, [acceptability]
  3. framing criteria, [knowledge frames]
  4. conceptual organization criteria – a hierarchy of concepts - [distinct categorization of the same objects]

Truth-conditional criteria are most likely, though obviously not in each case, to be satisfied in translation. They refer to classical formal translatability criteria and are traditionally considered to express one of the properties of adequate, ‘true’ translation. Criteria of use pertain to the acceptability criteria, referring to style, genre and speaker’s idiosyncratic preferences. Knowledge frames are captured by the framing criteria, which are dynamic, shaped by cultural preferences and shared, conventional patterns. Conceptual organizational criteria are responsible for particular language- and culture- specific categorization of concepts, most likely to be characteristic for a particular language system.  Languages may share some of the criteria but they would never share all of them. Therefore, the transfer of all nuances of meaning in their totality is simply not possible.

Displacement of senses (Lewandowska-Tomaszczyk 1987)

Relevant to the present discussion is also the phenomenon of the displacement of senses, which accounts for the asymmetry between cross-linguistic meanings both in synchronic and diachronic perspectives.  In such cases as e.g., the superordinate verbal concept go in English, the Polish equivalence patters is displaced towards a lower categorization hierarchy level, i.e., its subordinate categorization. Thus, the English-Polish equivalence structure is asymmetrical and displaced towards Polish obligatory subcategories – iść ‘walk’ versus jechać ‘move by using a vehicle’, each of which would again have an asymmetric equivalent in English e.g., drive, ride, etc. Such meaning spaces as I call them, provide larger cluster equivalence areas between SLs and TLs.

6. Translational equivalence

It may be clear from the above discussion that the concept of translational (and in fact more general cross-linguistic) equivalence is of a cluster rather than single unit nature. This is not a new idea in lexicography and translation studies. And yet, a more overall corpus-generated  model on the one hand and a detailed cognitive process-based typology of such phenomena had not been put forward before.  I propose in Lewandowska-Tomaszczyk (2010)  that both translation and monolingual types of communication can be symmetric or asymmetric. Fully symmetric, i.e., closely aligned, or parallel, communication, is not frequent, or perhaps not possible as argued above, similarly to non-ubiquity of fully equivalent translation. Asymmetric types of interaction with two opposing poles, one -  coarse-grained approximate correspondence of  generalized alignment, on the one hand, and asymmetric fine-grained alignment (correspondence), i.e., a particularization relation, on the other, are much more frequent and form larger bulks of authentic texts. To put it in a more diagrammatic mode, translational equivalence can be

  • loose, Gestalt in nature (more-or-less, i.e., an approximation of an exact solution, in the cases when full meanings are accessible (e.g., I made  a hundred Euro on that  vs almost structures, where no exact solution exists for a given case e.g., vague expressions e.g., several cases)
  • parallel (more) aligned (usually domain-specific as in engineering, medical sciences, etc.)
  • particulate, fine-grained communication (explanatory as in the cases of detailed definitions)

And yet, there seem to exist some context-bound constraints on meaning modulation substantiated in terms of the semantic approximation  tolerance threshold (Lewandowska-Tomaszczyk 2012b), the boundaries  of which are delineated  in terms of the degree of resemblance of particular language signs used in cross-linguistic communication. Resemblance is a culture-, context- and speaker-specific notion, in which no one of the variables can be fully predicted or determined outside the cultural or situational context. Nevertheless, the original linguistic sign is to a large extent, as proposed by Gärdenfors (2000) a determinant of the width of the possible tolerance spaces with respect to particular attribute values.

The degree of linguistic unit resemblance can be established in terms of their resemblance criteria (Lewandowska 2012a, 2012b):

  • qualitative features [material/physical and functional properties, shapes, topology, etc., axiological value, evaluative charge, cultural function, etc.]
  • construal properties [syntactic constraints on meaning and pressure of syntactic selection]
  • quantitative features [near sets] – number of object feature values in common; number of shared contexts & frequency of occurrence

Each shift in the feature brings about a degree of its reconceptualization in a Target Language system. 

6.1. Cluster equivalence and derived equivalence

As argued above cross-language equivalence is not of a word-for-word but rather of a cluster-for-cluster type. Translational cluster equivalence is built from more loosely constrained units that are related by a degree of resemblance according to  a number of the criteria identified in section 6. I call part of the cluster, occupied by a more constrained class of similar concepts/forms derived equivalence. For example, the English concept compassion exhibits a cluster of derived, semantically closely related, more constrained and predictable, members such as sympathy, empathy, concern, pity, although not fully identical in meaning (Lewandowska-Tomaszczyk and Wilson 2016).

The Polish derived equivalence pattern of współczucie ‘compassion’ will embrace a somewhat distinct space. The Polish item sympatia ‘sympathy, fondness’ on the other hand, as indicated by its association numbers, is more loosely associated with współczucie ‘compassion’ than the pair sympathy / compassion in English, or in some senses it can even be considered its false friend.  Furthermore, while there appears a similar derived cluster litość ‘pity’ in both languages, its Polish associated cluster member politowanie ‘pity expressed with a feeling of superiority’  makes the cluster more negatively evaluated as a whole. The members of these clusters are identified and used as equivalence units in Polish-to-English and English-to-Polish parallel data, e.g.,:

Eng. that would have moved me to pity
Pol. budziła we mnie współczucie ‘compassion’
Pol. umiał obudzić w sobie współczucie ‘compassion’ [dla niego]
Eng. was so sorry ‘przykrość’ for him

Extended cluster equivalence on the other hand implies a more distant resemblance relation, much less predictable, less constrained and occasionally extended beyond the conventional semantic space threshold. Polish współczucie can trigger English love as an extended inter-cluster equivalent, or can surface as an element signifying causality of the compassion act, representing the modulation tolerance threshold on the proposed equivalence patterns (Lewandowska-Tomaszczyk and Wilson 2016 for a detailed description of such processes). The latter signify objects that might satisfy the particular meaning conditions: causes, presuppositions, results, or even more emergent outside knowledge-based equivalents, i.e., objects that are not likely but not impossible to satisfy the conditions of meaning approximation (Gärdenfors 2000) as e.g., in the English original (3), in which the character experiences ‘a moving effect’ in himself, translated into the more interactive, co-temporal feeling of współczując ‘feeling sympathy/compassion to’ :

Eng. and that her brother always took that for a received truth on such occasions, and that it always had a moving effect upon him
Pol. Wiedział o tym brat Peggotty, współczując niedoli wdowy. lit. ‘It was known to Peggotty’s brother,  feeling sympathy to the widow’.

The topmost Polish equivalents of Eng. compassion presented in Table 1. below, which are generated from the PARALELA materials, include Polish equivalents of the nominal (współczucie ‘compassion’),  verbal (współczuć ‘to show sympathy, compassion’, okazać ‘to show’) and prepositional (nad ‘for somebody, something’) character. The most frequent translational clustering with other emotion concepts is also present in the forms litość ‘pity’, miłosierdzie ‘mercy’ and miłość ‘love’.

PARALELA, Dice’s similarity coefficient
A, B, C – components of Dice’s score
A - total frequencies of both wordforms in SL and TL
B - presence of wordform in SL only
C - presence of wordform only in TL

Table 1. Top Polish cluster equivalents of Eng. compassion

The conclusions with respect to translational equivalence then are not fully synonymous with what e.g., Krzeszowski (2016)  proposes as a ‘translational equivalence delusion’. They rather support the postulate for translational equivalence to involve four basic processes, in which the phenomenon of clustering involves  various possible cluster equivalence types discussed (see also Lewandowska-Tomaszczyk 2017):

  1. cluster-for-cluster mapping
  2. higher schematization of the Target as contrasted with the Source version
  3. POS re-categorization  (e.g. Eng. disgust, aversion, nausea, distaste, horror, despise, to loathe, to shrink (from something), dirt, reluctance, translated as  Pol. wstręt ‘repulsion’, brzydzić się ‘to loathe’)
  4. contrual reconceptualization

In the sections to follow a more detailed description will be presented of the Polish students’ practical application and use of the theoretical cluster equivalence instruments in in their translation tasks.

7. NS-NNS collaborative practice in clarification of law-related terms

7.1. Specialist terminology

The American students collaborating with the Polish ones in the TAPP project specialize in non-humanities fields and their knowledge of legal, scientific and technological terminology seems to exceed that characterizing the Polish students. Below there is an example of sets of questions asked by the NNS and answers given by their American pairs.

 7.1.1. law, legislation, legislature

(4) Questions: NNS > NS

(a) when you write “legislation [...] need to approve the transfer” you mean that there should be some new regulations introduced in the Dakota, Minnesota and Canada's legislation or that those two states and Canada should sign a kind of tripartite agreement or pact?

(b) By “legislature” that operates on a biennum [biennium] you mean the group of people that decide about rules and regulations, am I right? That they're elected every two years?

Parallel to their cooperative tasks the Polish students are exposed to the data they generate from Polish and English large monolingual corpora – the National Corpus of Polish and the British National Corpus on the one hand and Polish-to-English and English-to-Polish parallel corpora (Paralela) to  elaborate on and comprehend meanings they accessed from the concordances and identify series of derived and cluster equivalence patterns.

7.1.2. legislation/prawodawstwo

Parallel corpora

Sources: Acquis communautaire, Community Research and Development Information Service, Cordis Focus Newsletter, The European Union budget at a glance, Inforegio news

(5) Eng legislation >Pol clustering  equivalence pattern

legislation (1187 segments)

  • prawodawstwo
  • przepisy
  • prawo/a
  • produkty (prawne)
  • siła wnioskodawcza
  • legislacja
  • ustawodawstwo/a 

Parallel corpus concordances


(Eng) The current legislation on FP6 

(Pol) Obecne prawodawstwo dotyczące 6. Programu Ramowego 

In both languages the corresponding forms legislation/prawodawstwo can be used as plural nouns, although in English the relevant term corresponds to a more obviously countable and considered common plural noun (e.g. laws) in such cases:


(Eng) Approximation of the laws of Member States

(Pol) Zbliżenie ustawodawstw Państw Członkowskich

7.1.3. legislation/przepisy clusters


(Eng) Communication from the Commission concerning the fraud-proofing of legislation and contract management

(Pol) Komunikat Komisji odnośnie odporności na oszustwa przepisów [regulations] oraz zarządzania umowami


(Eng) the decision to withdraw from the joint declaration would not affect national legislation

(Pol) decyzja o wycofaniu się ze wspólnej deklaracji nie będzie miała wpływu na prawo  [law] krajowe

A comment on the quality of the corpus materials presented in sections 7.1.2. and 7.1.3. might be useful at this point, e.g. the phrase “odporność na oszustwa przepisów” (example (7)) calls for a more adequate equivalent phrase such as e.g.,  ‘instrumenty sprawdzające nadużycia w prawodawstwie / stanowieniu prawa / uchwalaniu prawa’; dostosowanie in example (11) is not necessarily a fitting equivalent of “adopt”  (cf. “adapt”); and the phrase “państwa członkowskie” should not be capitalized in Polish. Such inadequacies are one of the most important reasons why the next step in LSP translation didactics, i.e., consulting professional terminological databases and thesauri (referred to in example (20) below) is a necessary stage in translator education. This is also a reason why the idea of joint projects of building terminological databases of closed-domain threshold concepts proposed in the conclusions, should be seriously considered.

7.2.  Collocations

Steady word combinations, i.e., collocations, of the adjectival, verbal, and prepositional nature, have also been generated by Polish students by means of the available collocating HASK tools (Pęzik 2014):


(Eng) The panel recognised that the processes have pre-empted future EU legislation in this field.

(Pol) Panel uznał, że metoda [method] znajdzie duże zastosowanie w innych produktach


(Eng) There are many examples of crossborder projects that result in changes to legislation and policy

(Pol) Trzeba przyznać, że często pojawiają się projekty transgraniczne, które przyczyniają się do postępu w ustanawianiu prawodawstwa [establishing legislation]


(Eng) The EGTC regulation requires Member States to adopt national legislation to facilitate the implementation of EGTCs.

(Pol) Rozporządzenie o EGTC wymaga od państw członkowskich dostosowania legislacji krajowej w celu ułatwienia wdrożenia EGTC.


(Eng) if the total length of the insurance periods completed under the legislations of all the Member States

(Pol) jeżeli całkowita długość okresów ubezpieczenia, ukończonych na podstawie ustawodawstw wszystkich zainteresowanych Państw Członkowskich

Pol>Eng clusters

prawodawstwo (wspólnotowe) (1022 segments)


(Pol) Prawodawstwo wspólnotowe

(Eng) The Community legislation


(Pol) Prawodawstwo i rynki komunikacji elektronicznej w Europie w 2004 r

(Eng) European Electronic Communications Regulation and Markets 2004


(Pol) Prawodawstwo wspólnotowe zobowiązuje państwa członkowskie

(Eng) Community law requires Member States

Further instances of the sense displacement are observed in identifying other cluster members (legislation, legislature) to act as an equivalent of Polish prawodawstwo:

Pol. prawodawstwo – Eng. legislature


(Pol) Zignorował pan i złamał prawodawstwo Sądu Najwyższego .

(Eng) The legislature and the supreme court [were ignored…]

Access to complete sets of relevant law-related collocations and their frequencies in the tools is also provided either in the Excel or visualized forms, combining new options of collocability of the term, accompanied by relevant statistics  (t-test TTEST and the overall frequency A):

(17) Adj collocates of Eng. law


































































(18) VERBAL collocates of Eng. law






















































































(19) NOUN collocates of law  (accessed and generated on 14 March 2017)

The collocation patterns possess some new terms, unfamiliar to the Polish students, which encourage them to consult a terminological thesaurus and access historical background  information, helpful to explain the term in more detail as e.g., in the case of the phrase Sus law, sixteenth on the collocation list:

Sus law (16th on the collocate list)


In England and Wales, the sus law (from "suspected person") was the informal name for a stop and search law that permitted a police officer to stop, search and potentially arrest people on suspicion of them being in breach of section 4 of the Vagrancy Act 1824. 

The Vagrancy Act 1824 (5 Geo. 4. c. 83) is an Act of Parliament of the United Kingdom that makes it an offence to sleep rough or beg.

7.2.1. federal nexus

A procedure of consultation with NSs, definitions they propose as well as their extensions in terminological data bases and exploring their historical background – were all used in the case of the identification of another of the law-related terms, namely, federal nexus. Two different pairs of NNS – NS peer students were discussing the term:

(21) NNS1>NS1>NNS1 interaction

Q: What do you mean by "various federal nexuses" and what is EPA - just expand the abbreviation, please

A: I had to look this up as well after Interviewed Michael. The term “federal nexus” applies when a project involves federal funding, federal permit or approval, use of federal lands, or a federal program. The existence of a federal nexus often triggers the need for federal approvals.  EPA stands for Environmental Protection Agency.

(22) NNS 2 > NS 2 > NNS2 interaction

Q: What is a “federal nexus” and why is it important in permitting?

A: The term “federal nexus” applies when a WSDOT project involves federal funding, federal permit or approval, use of federal lands, or a federal program. The existence of a federal nexus often triggers the need for federal approvals under certain statutes, including NEPA, Section 106 of the National Historic Preservation Act, and the Endangered Species Act

In order to cumulate more extensive information, the Polish students consulted legal source documentation available online:


federal nexus  [source documentation]

  • What is the difference between a “permit” and an “approval”?
  • A “permit” is a document required by law that authorizes a specific type of activity under certain conditions. An example is a Section 404 permit issued by the U.S. Army Corps of Engineers (Corps).

An “approval” means any document or process other than a permit that needs a signature by someone in authority at an agency having jurisdiction or control over an activity. An approval may also include documentation, certification, concurrence, easement, or license. For example, Section 106 of the National Historic Preservation Act, requires no permit, but does require concurrence by the State Historic Preservation Office.

Further meaning elaboration strategies presented below, which involve cooperative tasks and are supported by the corpus data, focus on the identification of the context-induced senses of particular terms and general-language phraseology.

7.3. Contextual clarification

(24) NNS>NS Clarification requests, attempts at self-clarification [paraphrase]

Just a few questions more:

Q1 sustainable source  - do you mean that this source is stable/balanced/reliable or rather renewable?

Q2 About the appropriation issue - I understand that appropriation is giving permission to use the water, but does it involve constructing a new installation, I mean like some pipes to supply water from the main/municipal pipeline to the building/allotment?

Q3  senior users – are they old users?


A1 I'm referring to the natural source (the Red River). That sentence wasn’t very clear. Sorry.

A2 Appropriation means the state government is granting permission to use natural resources.

A3 Senior users are communities and individuals who have been taking water from a source for the longest time. This is an important factor because it determines how fairly we can divide the water resources. When you can regulate how much upstream users are going to use, you know how much water is left over for communities downstream. Our system at the moment is just "first come, first serve", regardless of the impact to others.

To obtain more elaborate meaning contextualization, the explanation of the term appropriation and others proposed by the NS was confronted with the Monitor corpus data (25), which contain materials from selected press titles and internet fora with daily updates ( They provided a more elaborate contextually-bound sense of the term, putting the term in the practical use in reporting on works of state committees:


The House and Senate appropriations committees heard proposals Wednesday , Jan . 4, for a new legislative forecast for the coming two-year budget cycle that includes nearly $ 1.9 billion less in general fund revenues than what 's projected for the 2015-17 biennium, which ends June 30. 4/1/2017

8. Conclusions

Although translation students in general typically come across numerous SL – TL asymmetry cases, the parallel corpus practice makes it possible for them to realize the richness and complexity of both Source Language and Target Language clusters considered equivalent by translators. The exposure to the corpus data rather than the use of bilingual dictionaries alone, makes it possible for them to appreciate the variability of the source data to the extent unnoticed before. This is typically not the case in bilingual dictionaries, where there is one or few headwords on the SL side and clusters are revealed only in the Target Language. In the approach I’ve been using it is both the richness of the Source Language ideas and forms that impose, to a considerable degree, the complex variability of the Target language lexical forms and structures. Both meaning approximation and translational, and more generally, communicative, re-conceptualization processes that I consider the core processes in perception, cognitive conception and linguistic expression (Lewandowska-Tomaszczyk 2017a), that condition and shape translational equivalence.

The groups of Polish students participating in the collaborative tasks, aided by different types of corpus data and tools and stimulated by discussions concerning translation processes,  acquire this line of translation understanding and try to implement it in their practical translation tasks. After completion of the cooperative tasks, the students were asked to fill in a post-task survey, which revealed benefits brought by the peer contacts. Both US and Polish students identified significantly more advantages than disadvantages from the cooperation.  American students stated that their Polish pairs offered very good suggestions toward helping them to be clear and more articulate in their writing and they also learned that they needed to cooperate with people rooted in different cultures and linguistic systems.

In their  writing, US NSs learned to develop  more consideration towards people from other cultures and their expectations – this experience developed their sense of criticality towards their own (native) language. The Polish students learned about terminology so that they could understand what they were “to look for in Polish as the equivalents” as they phrased it. They also acquired a number of less formal phrases and strategies when contacting people of their own age and background. They learned the strategies of acquiring more sophisticated and fuller meanings of the terms they needed for their writing and translation tasks by consulting corpus data and terminological databases. They also learned to recognize asymmetries and rich clustering between both Source and Target languages they make attempts to  solve, resorting to corpus-derived cluster equivalence correspondences.

There are other weaknesses in the non-native students’ translations that will be analyzed in a forthcoming publication, less closely connected with the LSP character of the examples given above. First of all, what is observed is the use of an excessive number of words, verbosity, in the students’ English translations when compared with corresponding native English texts. This is primarily instantiated by a lower frequency, or complete avoidance in some cases of B1-level students, of the use of phrasal verbs at the advantage of more descriptive and less idiomatic expressions. What can also be noticed is the students’ weak metalinguistic competence, which was expressed in the frequent use of an inadequate genre or style in the email letters the Polish students exchanged with the American ones. In the English-to-Polish translation tasks a particularly strong English language interference on the native Polish language results in the Polish students’ use of inadequate and clumsy syntactic structures and unconventional idiomaticity in their native Polish.

Let me conclude with pointing to some gaps which might additionally need to be bridged in further research and practice in this field, particularly with reference to translation trainees’ terminology acquisition. What would be considered a desirable development there would be domain-specific translational threshold concepts reference materials for NS & NNS. Threshold concepts of a particular domain are constitutive elements of ‘core knowledge’ of the field, which can secure its fundamental understanding. A threshold concept is ”akin to a portal, opening up a new and previously inaccessible way of thinking about something. It represents a transformed way of understanding, or interpreting, or viewing something without which the learner cannot progress” (Meyer and Land 2006: 3). Research, preparation and development of sets of foreign-language domain-specific threshold concept repositories, should engage specialists in the field, corpus linguists, terminologists, FL teachers and LSP specialists.

The development of collaborative corpus-informed didactic materials and methodology, involving larger studies of cluster equivalence patterns and, when needed, the threshold concept repositories across pairs of languages, seems one of the most urgent tasks in translator education.


I wish to thank two anonymous reviewers for the constructive comments on the first draft of the paper.


Gärdenfors, Peter. 2000. Conceptual Spaces – The Geometry of Thought. Cambridge, Mass.: The MIT Press. A Bradford Book.

Krzeszowski, T.P. (2016).   The Translation Equivalence Delusion: Meaning and Translation. Frankfurt am Main: Peter Lang.

Lakoff, G. (1987). Women, Fire and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press.

Langacker, R. W. (1987). Foundations of Cognitive Grammar vol. 1, Stanford: Stanford University Press.

Lewandowska-Tomaszczyk, B. (1987). Conceptual Analysis, Linguistic Meaning, and Verbal Interaction. Łódź: Łódź University Press.

----- (2010). “Re-conceptualization and the emergence of discourse meaning as a theory of translation”. In: Lewandowska-Tomaszczyk, B. and M. Thelen (eds.) Meaning in Translation. Frankfurt a. Main: Peter Lang.

----- (2012a). “Explicit and tacit - An interplay of the qualitative and quantitative approaches to translation" In: M. Oakes and M. Ji (eds).Quantitative Methods in Corpus-Based Translation Studies. Amsterdam: Benjamins. 3-33.

----- (2012b). “Approximative spaces and the tolerance  threshold in communication”. International Journal of Cognitive Linguistics  volume 2, number 2. Nova Science publishers, inc. 

----- (2016). “Collaborative Native - Non-native Student Translation Tasks in CMC: A Corpus-Based Study”. Abstract and a plenary lecture at TALC 2016 Conference at the University of Giessen. Available at

----- (2017). “Cluster Equivalence, General Language, and Language for Specific Purposes“. In: M. Grygiel (ed.) Cognitive Approaches to Specialist Language. Newcastle upon Tyne: Cambridge Scholars Publishing. 384 – 418.

----- (2017a). Partial perception and approximate understanding. Research in Language 15.2. 129-152.  

Lewandowska-Tomaszczyk, B., & Ł. Bogucki (2016). “Volunteer translation, collaborative knowledge acquisition and what is likely to follow”. In: Ł. Bogucki, B. Lewandowska-Tomaszczyk, M. Thelen (eds.) Translation and Meaning. New Series. Vol. 2, Pt 2. Frankfurt a. Main: Peter Lang. 37-46.

Lewandowska-Tomaszczyk, B. & H. Slomski  (2016). “Collaboration in language development between American and Polish university students”. In: Lewandowska-Tomaszczyk, B. (ed.). Konin Language Studies volume Collaborative Practices in Language ed.  Volume 4 No.3/2016. 305-330.

Lewandowska-Tomaszczyk, B. and Wilson, P.A. (2016). “Compassion, empathy and sympathy expression features in affective robotics”. In: Peter Baranyi et al. (eds.) IEEE: CogInfoCom Cognitive Infocommunication. Wrocław: Wrocław Technical University.  65-70. Appeared as a modified paper  (2018) Compassion cluster expression features in affective robotics from a cross-cultural perspective. In Cognitive Infocommunications, theory and applications, P. Baranyi, R. Klempous & J. Nikodem (eds.). Springer. 202-227.

Maylath, B., T. King & E. Arnó Macià. (2013). “Linking engineering students in Spain and technical writing students in the us as coauthors - The challenges and outcomes of
subject-matter experts and language specialists collaborating internationally.” Connexions. 1(2): 159-185.

Meyer, J.H.F. and Land, R. (2003), Threshold concepts and troublesome knowledge: linkages to ways of thinking and practicing. In: Rust, C. (ed.), Improving Student Learning - Theory and Practice Ten Years On. Oxford: Oxford Centre for Staff and Learning Development (OCSLD), pp 412-424. [See also for an on-line version: Enhancing Teaching-Learning Environments in Undergraduate Courses, ETL Project, Occasional Report 4, May 2003   last accessed 25 June 2008].

Neubert, A. (2000). “Competence in Language, in Languages, and in Translation”. In: Ch.

Schäffner & B. Adab (Eds.) Developing Translation Competence. Amsterdam/Philadelphia: Benjamins, 3-18.

Pęzik, P. (2014). Graph-Based Analysis of Collocational Profiles. In: V. Jesenšek & P.

Grzybek (eds.) Phraseologie im Wörterbuch und Korpus (Phraseology in Dictionaries and Corpora), ZORA 97. Maribor, Bielsko‑Biała, Budapest, Kansas, Praha: Filozofska fakuteta. 227–43.

Pietrzak, P. (2013). “Teaching, training, educating? Terminological ambiguity in translator education theory”. In: B. Lewandowska-Tomaszczyk & M. Thelen (eds.) Translation and meaning part 10. Maastricht: Maastricht School of Translation and Interpreting. Zuyd University of Applied Sciences. 245-251.


British National Corpus:

National Corpus of Polish:

Monitor Polish and English Corpora:;

Parallel English-to-Polish and Polish-to-English Corpora PARALELA:

About the author(s)

Barbara Lewandowska-Tomaszczyk, Dr habil., full professor of English and Linguistics at the State University of Applied Sciences in Konin, head of the Department of Research in Language, Literature and Translation, for many years served as head of the Department of English Language and Applied Linguistics at the University of Lodz. Author and editor of numerous books and papers in cognitive and corpus linguistics, collaborative knowledge acquisition and translation, invited to read papers at conferences and give workshops at European, American and Asian Universities.

Email: [please login or register to view author's email address]

©inTRAlinea & Barbara Lewandowska-Tomaszczyk (2019).
"Cluster and derived equivalence in collaborative corpus-informed translation education"
inTRAlinea Special Issue: New Insights into Translator Training
Edited by: Paulina Pietrzak
This article can be freely reproduced under Creative Commons License.
Stable URL:

Go to top of page