trans-kom ISSN 1867-4844 http://www.trans-kom.

eu
trans-kom ist eine wissenschaftliche Zeitschrift für Translation und Fachkommunikation.

trans-kom 9 [1] (2016): 114-148

Seite 114

Ralph Krüger

Contextualising Computer-Assisted Translation Tools and

Modelling Their Usability

Abstract
In this paper, I attempt to contextualise computer-assisted translation (CAT) tools from a
theoretical and a professional perspective and to model the usability of these tools. The
theoretical contextualisation of CAT tools is based on Risku’s (2004) cognitive translational
theory of Situated Translation, which claims that cognition is not isolated in the translator’s head
but emerges in dynamic processes of interaction between the translator and his or her working
environment. The potential influence exercised by CAT tools on the translator’s cognitive per-
formance is illustrated by means of the Cologne Model of the Situated LSP Translator, which is
based on the theory of Situated Translation and which conceptualises CAT tools as important
environmental artefacts in the translational ecosystem. The professional contextualisation takes
the form of a survey of the role of CAT tools in the different phases of the computer-assisted
translation process. In this context, I discuss how the various sub-tasks of these process
phases can be supported by different CAT tools. Following this theoretical and professional
contextualisation, I propose a general model of CAT tool usability as well as a more specific
model of translation memory system (TMS) usability. These models are derived from ISO
standard 9241 “Ergonomics of Human-System Interaction” and attempt to capture all relevant
dimensions of CAT tool usability from a user-oriented perspective.

1 Introduction
In the last 15 to 20 years, translation technology has become an integral part of the
translation process and has changed the nature of this process in quite fundamental
ways (Christensen/Schjoldager 2010: 1; O’Brien 2012: 1). What has become very clear
over time is that translation technology has come to stay and that the good old days of
pen-and-paper translation are inevitably coming to an end. This is evidenced, for
example, by Gouadec (2007/2010: 156), who analysed 650 job advertisements for
translators and found that all of these advertisements require translators to have
knowledge of translation memory (TM) systems. Also, younger generations of translators
entering the translation market today will often have been educated in universities
featuring state-of-the-art computer infrastructure and translation technology and will

http://www.trans-kom.eu/bd09nr01/trans-kom_09_01_08_Krueger_CAT.20160705.pdf
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 115
and Modelling Their Usability

probably feel that computer-assisted translation (CAT) is the natural way to translate.1
Generally, the adoption of CAT tools is considered to have both advantages and
disadvantages. For example, the use of TM systems may entail, on the positive side, a
higher process standardisation, productivity and cross-document consistency and lower
translation costs and, on the negative side, a possibly higher focus on individual
translation segments (to the detriment of the textual perspective) or the uncritical adop-
tion of low-quality translation units from contaminated TMs (cf. Christensen/Schjoldager
2011: 128; Ehrensberger-Dow/Massey 2014a: 200). In this article, I will not attempt to
answer the question of whether the advantages of CAT tools outweigh their dis-
advantages or vice versa, particularly since – as discussed above – the alternative to
revert back to pen-and-paper translation is no longer a feasible option.2 What needs to
be emphasised, however, is that translation technology does not only have a profound
impact on the nature of the translation process but also on the translator’s cognition. This
is highlighted in the following quote by Pym:
[...] new translation technologies such as translation memories, data-based machine
translation, and collaborative translation management systems, far from being merely
added tools, are altering the very nature of the translator’s cognitive activity [...].
(Pym 2011: 1)
While this impact of CAT tools on the translator’s cognition has been recognised in
translation studies, there is still a scarcity of translation technology research which
does proper justice to this impact and which scrutinizes more closely the usability
dimension of translation technology, usability here being understood broadly as the
extent to which a user can achieve a certain goal with a given software tool (for a
detailed definition and discussion of usability, see section 4 below). The link should be
quite obvious: While CAT tools exhibiting a high usability should enhance the
translator’s cognitive performance, tools exhibiting a low usability will probably tend to
decrease it (a trivial observation with which most practicing translators will certainly
agree).3,4 However, as Dillon and Fraser (2006: 68) and Christensen and Schjoldager
(2010: 1) point out, the major share of the literature on translation technology focuses
on issues such as product reviews, product comparisons or aspects of workflow
management. And while the usability dimension of translation technology is often

1 In Dillon and Fraser’s (2006) investigation of translators’ perception of translation memory adoption,
the authors found, perhaps not surprisingly, that younger translators tend to adopt a more positive
stance towards the use of CAT tools than older translators, who were educated and socialised prior to
the ‘technological turn’ in the translation industry.
2 O’Brien (2012) even conceptualises translation as a form of “human-computer interaction”, which
stresses the high relevance of translation technology in the modern translation profession.
3 See also Abran, Khelifi and Suryn (2003: 325): “For the end-user, software usability is essential
because it is a determinant of performance: an application which features good usability will allow the
user to perform the expected task faster and more efficiently.”
4 In the words of Ehrensberger-Dow and O’Brien (2015: 102), low-usability CAT tools may lead to
“cognitive friction”, which can be understood as “the resistance encountered by a human intellect
when it engages with a complex system of rules that change as the problem changes” (Cooper
2004: 19).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 116
and Modelling Their Usability

implicitly reflected in these publications, it is rarely addressed in an explicit and/or

structured way.5 For example, in a very interesting and useful article for the German
Federal Association of Interpreters and Translators (BDÜ), Keller (2011) compares the
performance of nine TM systems in various phases of the translation process and
discusses in detail some relevant features of these systems. Although, at various
points in the article, the author highlights the benefits or shortcomings of specific
performance features of a given TM system, there is no detailed discussion of usability
aspects from the perspective or to the benefit of a practicing translator.
The lack of consideration of the actual interaction of translators with translation
technology in practical contexts (which will inevitably raise the question of usability) as
well as the lack of corresponding empirical research is also highlighted by Christensen
and Schjoldager:
Empirically documented knowledge about the nature and applications of TM systems and
translators’ interactions with them is both scarce and fragmented. In particular, more
research is needed on how translators interact with TM technology and on how it influences
translators’ cognitive processes. (Christensen/Schjoldager 2010: 11)
In the light of the supposedly strong influence of CAT tools on the translator’s cognition
and the relative lack of research on the translator’s interaction with CAT tools, it seems
desirable, within translation technology research, to place more emphasis on CAT tool
usability and to take into account the perspective of practicing translators in real-world
translation settings. The need to shift the focus from isolated investigations of translation
technology to joint considerations of this technology and its intended users is underlined
by Lagoudaki, who points out, quite rightly in my opinion, that “many of the existing
commercial TM systems are technology-driven applications (e.g. with an abundance of
useless features and a complex, impractical and difficult to learn user interface) rather
than user-driven applications.” (Lagoudaki 2006: 2, italics added). This view is mirrored
by O’Brien, who claims that “there is little evidence to suggest that tools that are
proposed as aids to the translation process have been designed from the point of view of
the humans wo have to use them.” (O’Brien 2012: 115).
The aim of this article is twofold. Firstly, it aims to contextualise CAT tools from a
theoretical and a professional perspective. To do so, the conceptualisation of CAT tools
in the cognitive translational theory of Situated Translation and the practical role of
these tools in the computer-assisted translation process are discussed. This is inten-
ded to underline the theoretical justification and the practical relevance of more
exhaustive and intensive CAT tool usability research and to lay the groundwork for the
discussion of CAT tool usability as the second focal point of this article. In this context, I
propose a general model of CAT tool usability as well as a more specific model of

5 Campbell et al. assume that it is due to the highly specialised nature of translation technology and the
fact that the widespread adoption of this technology is still in progress that “the usability of translation
technology has not been thoroughly evaluated” (Campbell et al. 2013: 2042). However, as discussed
briefly at the beginning of this article, the widespread adoption translation technology has become a
fact by now, which makes the need for larger-scale CAT tool usability research all the more pressing.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 117
and Modelling Their Usability

translation memory system (TMS) usability and I discuss the various usability dimensions
of these models in detail. The models can be used as foundations for structured
approaches to CAT tool usability testing.

2 Theoretical Contextualisation: CAT Tools in the Framework of

Situated Translation
In this section, I attempt a theoretical contextualisation of CAT tools. To do so, I will
sketch, very briefly, the development of cognitive translation studies and demonstrate
that modern cognitive translational approaches such as Situated Translation offer
adequate concepts and models which allow us to capture the importance of CAT tools
for the translator’s cognitive performance in sound theoretical terms.
Since its inception in the middle of the 20th century, modern translation studies has
undergone a development which led from a contrastive perspective on linguistic
systems via a contrastive perspective on texts to a holistic perspective on translational
action and the person of the translator (Stolze 1994/2011). This shift of epistemic focus
from the product to the process and the agent of translation eventually led to the
‘cognitive turn’ in translation studies (cf. Siever 2010: 341) and to the central question
of ‘what happens in the minds of translators’ (cf. Krings 1986). Within cognitive
translation studies, a similar development can be observed. Starting from an ‘ideal
translator’ (cf. Kade 1968: 62), whose mind served as some kind of storage box for
various linguistically-oriented ‘translation algorithms’, the epistemic interest has
widened gradually to include the individual knowledge requirements of the translator
and hence the translator as a real person.6 At first, this real translator was investigated
in relative isolation from the various influences in his/her professional environment, but
gradually these factors became more important in constructing cognitively-oriented
theories of translation. Within cognitive translation studies, this development has
culminated in the above-mentioned theory of Situated Translation (Risku 2004), which
assigns central importance to the individual situational factors of the translator and
his/her working environment. Situated Translation is based on the cognitive-scientific
paradigm of situated cognition, which claims that humans and their environment form a
cognitive system, more precisely, a cognitive ecosystem (Strohner 1995: 56). This
means that human cognition is not isolated in the human brain but emerges from the
specific relationship between humans and their environment (Strohner 1995: 54). From
a translational perspective, the concept of a cognitive ecosystem allows describing the
translator’s cognition by taking into account all the situational factors that were largely
ignored in the early stages of cognitive translation studies. The need for such a macro-

6 These two early stages of cognitive translation studies correspond to the two cognitive scientific
paradigms of symbol manipulation and connectionism. For a more detailed discussion of the different
paradigms in cognitive science and cognitive translation studies, see, for example, Risku (2004,
2005) and Krüger (2015).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 118
and Modelling Their Usability

scopic and holistic perspective on the translator’s mind and his/her environment is
stressed by Risku:
The mind is only one part of the story. We need to find out not only what happens in a
translator’s mind, but also what happens elsewhere, e.g. in their hands, in their computers,
on their desks, in their languages or in their dialogues. Translation is not done solely by the
mind, but by complex systems. These systems include people, their specific social and
physical environments and all their cultural artefacts. (Risku 2010: 103)
The notion of artefacts is particularly important in the context of the present article.
Artefacts are understood, quite generally, as objects made or used by humans for a
particular purpose and range from calendars and spectacles to languages and objects of
art to the products of modern information technology (Risku 2004: 20). Artefacts with high
relevance to translation practice and to the present article are, obviously, “the many tools
that are part of modern translation work (text processors, online research tools, translation
memories)” (Risku 2013: 36). From the perspective of Situated Translation, it becomes
clear why CAT tools, in the words of Pym (2011: 1, see quote above), alter “the very nature
of the translator’s cognitive activity” – these tools form an integral part of the translational
ecosystem. And since situated cognition claims that cognition emerges from the interaction
between humans and the various artefacts in their ecosystems, CAT tools can also be
conceptualised as an integral part of the translator’s cognition.7 While such an extended
view of the human mind and human cognition is still not unanimously accepted in cognitive
science (cf. Adams/Aizawa 2010), it provides a plausible basis for understanding and
analysing human-machine interaction (cf. Ehrensberger-Dow/Massey 2014b: 63).

2.1 From Theoretical to Professional Contextualisation: CAT Tools in the

Cologne Model of the Situated LSP Translator
Based on Risku’s theory of Situated Translation and Serrano Piqueras’ (2011) models of
the tectonics and the dynamics of the translation process, I have developed the Cologne
Model of the Situated LSP Translator,8 which aims at a holistic description of the LSP
translator and the relevant factors influencing his/her cognition in real-world translation
environments.9 In this section, I will briefly discuss the role of CAT tools in the Cologne
model since this model provides a link between the theoretical and the professional con-
textualisation of these tools. The model is depicted in figure 1 on the following page:

7 On the conceptualisation of CAT tools – in this case translation memories – as artefacts in the sense
of Situated Translation, see also Christensen: “A TM is to be considered a material artifact involved in
the human translator’s process of organizing the functional skill of memorizing translation decisions
made during earlier translations into cognitive functional skills.” (Christensen 2011: 140).
8 LSP = Language for Special Purposes.
9 In addition to the works of Risku (2004) and Serrano Piqueras (2011), the Cologne model is explicitly
informed by Holz-Mänttäri’s (1984) theory of translational action, Reinke’s (2004) model of the work
phases of the translation process and Schubert’s (2007) integrated model of mono- and multilingual
specialised communication.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 119
and Modelling Their Usability

Fig. 1: The Cologne Model of the Situated LSP Translator

Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 120
and Modelling Their Usability

Since the model aims at an exhaustive description of the situated translator’s pro-
fessional working environment, it is too complex to be illustrated here in full (for a
detailed discussion, see Krüger 2015 and Krüger forthcoming). In this article, I will
merely give a brief overview of and elaborate on the components which are relevant in
the present context. The model is divided into two levels. The lower level depicts the
translation process as a sub-process of the ‘process chain of specialised communi-
cation’ (cf. Schubert 2007: 132). This process chain forms the superordinate structure
in which the translational ecosystem is embedded. The translation process, which
corresponds to the transfer phase of Schubert’s process chain, is divided into several
work phases as identified by Reinke (2004: 102) and Serrano Piqueras (2011: 44).10 In
the model, I have depicted the maximum number of work phases of the translator.
Which of these phases actually apply and how prominent these phases are in the
translation process depends on the respective translation setting.11,12 These work
phases will serve to structure the professional contextualisation of CAT tools in section
3 below.
The upper level of the model depicts the actual constituents of the translational
ecosystem, i.e. the situated translator and his/her working environment. The relevant
components of this working environment are the translator’s cooperation partners,
various social, physical and psychological factors as well as different artefacts, which I
assigned to specific artefact groups. The group with specific relevance to the present
article is the artefact group concerned with translation technology in a narrow sense,
which could also be called CAT tools. The artefacts listed in this group – translation
memory systems, terminology management systems, alignment tools, machine trans-
lation (MT) systems13 and project management (PM) components – merely serve for

10 The work phases illustrated in the model also show a high degree of similarity to the translation
workflow depicted in ISO standard “Translation Services – Requirements for Translation Services”
(ISO 17100 2015).
11 For example, the phase of project initiation, which involves the acquisition of customers and translation
jobs, is a very important phase for freelance translators, whereas in-house translators in translation
agencies or language services will not usually be required to acquire new customers/jobs or to
negotiate fees for translation assignments.
12 Reinke (2004: 101-103) identifies three primary translation settings: (1) internal translation in the
language service of a company or a public authority; (2) external translation (a) by a translation
agency (in-house or by means of outsourcing), (b) by freelance translators (directly for a
company/public authority or via a translation agency), or (c) by project-specific teams of freelance
translators (directly for a company/public authority or via a translation agency), and (3) a combination
of internal and external translation or a combination of various forms of external translation.
13 It is debatable whether MT systems can be classified as a subset of computer-assisted translation
tools since these systems usually do not assist the process of human translation but perform the
translation automatically, with the translator’s role being reduced to the tasks of pre- and post-editing
(cf. Schubert 2007: 171; Reinke 2013a: 34). However, Reinke points out that it is possible to integrate
TM and MT technology in a batch processing or an interactive processing scenario. In a batch
processing scenario, “all segments of the source text that do not produce an exact or high percentage
‘fuzzy match’ when being compared with the TM database may be exported for processing by MT”. In
an interactive processing scenario, “translators can invoke the MT system each time there is no
match with the TM database” (both quotations Reinke 2013a: 35). Particularly the scenario of
interactive processing illustrates that MT systems can indeed be used to assist (instead of replace)
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 121
and Modelling Their Usability

illustration purposes, i.e. I did not attempt to include an exhaustive list of CAT tools in
the model. However, the artefact group translation technology in a narrow sense – or
CAT tools – includes only those tools that are (relatively) specific to the translation
process. The artefact group translation technology in a wider sense – in Risku’s
(2013: 36) words above, “the many tools that are part of modern translation work” –, would
also include more general tools such as text processing software, web concordancers
(cf. Krüger 2012) or concept mapping software (cf. Austermühl 2012), which are also
applicable and relevant in other professional fields beyond translation. While these
tools also play an important role in assisting the human translation process, they are
excluded from consideration in the present article.
As mentioned above, Situated Translation claims that translation technology as
well as other environmental artefacts form an integral part of the translator’s cognition.
Hence, the Cologne model of the Situated LSP Translator incorporates translation tools
as important constituents of the translational ecosystem. By acknowledging the potentially
high cognitive relevance of translation technology, Situated Translation provides a
sound justification for integrating CAT tools in translation theory building and for
assigning the usability of these tools a more central place in translation technology
research.

3 Professional Contextualisation: CAT Tools in the Translation

Process
In this section, I will attempt a professional contextualisation of CAT tools by surveying
the role of these tools in the computer-assisted translation process. In this context, I will
zoom in on the translation process depicted in the Cologne Model of the Situated LSP
Translator and attempt to establish a link between the various sub-tasks of the
individual work phases of this process and different CAT tools (both autonomous
software and integrated components of TM systems) which can be used to support
these sub-tasks (this is informed by Reinke 2004: 109). The survey of CAT tool support
in the translation process is not yet specifically concerned with the usability of these
tools. Rather, it is intended to lay the foundation for the structured discussion of CAT
tool usability in section 4, which will, at various points, make reference to the present
survey. The computer-assisted translation process is depicted in figure 2 on the next
page:14

the process of human translation. Bundgaard, Christensen and Schjoldager (2016: 108) call this
scenario “MT-assisted TM translation”.
14 Most examples discussed in the context of the computer-assisted translation process in the present
section and in the context of the usability of CAT tools in section 4 are based on SDL Trados Studio in
its current version (2015). This is due to the fact that Trados Studio is one of the most widely used
CAT tools in the translation industry and offers a plethora of functions and characteristics which are
relevant to the present article. However, the prominence of examples from Trados Studio should not
be understood as an endorsement of this TM system at the expense of other systems.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 122
and Modelling Their Usability

Fig. 2: The computer-assisted translation process

Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 123
and Modelling Their Usability

The translation process starts with the project initiation phase, which involves the
acquisition of new customers and translation jobs.15 Due to the nature of this phase, it
is not directly supported by any specific CAT tools. However, the workflows of
individual customers may require translations in specific file formats and/or translation
memory systems – for example because the customer has purchased a specific TM
system/developed a proprietary in-house system or because a given TM system is
optimally integrated with the customer’s content management system. If the translator
owns the TM system required by the customer, this may present a competitive
advantage and may prompt the customer to assign the translation job to this trans-
lator.16 Also, various TM providers (such as Across) maintain a translator database
which contains translation service providers (TSPs) working with the company’s TM
system. Customers who also work with this TM system or who have a workflow
compatible with this system may then search this database for suitable TSPs. From the
translator’s perspective, then, there seems to exist a potential link between the
acquisition of new customers/translation jobs and the ownership of a given TM system.
In the general preparation phase, the translation job is registered and (in the case
of a freelancer network, a translation agency or a language service) assigned to one or
more translators, the customer files are placed in suitable folder structures, and the
capacity for handling the translation job is planned. Today, most commercially available
TM systems feature a specific project management component which supports
management and handling of files (among other things, the PM component can assist
the translator in creating specific folder structures17), management of customer/
translator data and workflow management (cf. Reinke 2013a: 30). Also, TM systems
generally feature a statistics or analysis component, which establishes the amount of
new matches, fuzzy matches, repetitions, 100 % matches and perfect matches in a
given translation job and which thereby supports capacity planning.
The translation preparation phase involves reviewing the various files provided by
the customer, configuring the TM system, importing the source text into the TM system,
aligning legacy texts (if available), creating or importing TMs or terminology databases
(termbases), conducting preparatory terminology work and searching for parallel/
explanatory texts for documenting the translation job. Reviewing the customer files (for
completeness, suitability, etc.) can be supported by the TM component of the TM
system or by a terminology management tool, in case the customer provides one or

15 Again, whether this phase and other phases of the translation process are actually relevant for a
particular translator depends on his/her specific translation setting (see the discussion in section 2.1).
16 However, there is the danger that customers base their translator selection process solely on the
availability of specific TM systems and on the translator’s competent handling of these systems and
not on other relevant sub-competencies of the translator’s overall translation competence (for
example, his or her subject-matter competence).
17 For example, when working with the project package workflow in Trados Studio, the software
automatically creates folders for source and target files, translation memories, etc. Also, when
performing batch tasks such as Analyze Files or Export for Bilingual Review, Trados Studio
automatically creates corresponding folders and places the relevant files in these folders.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 124
and Modelling Their Usability

more TMs and/or termbases as reference for the translation job. Configuring the TM
system involves general preparatory tasks such as activating or deactivating the
AutoCorrect function, activating or (in the case of highly contaminated TMs)
deactivating the Auto-propagation function, configuring keyboard shortcuts, configuring
AutoSuggest dictionaries, etc. The settings or options panel of current TM systems
usually allows a very fine-grained configuration of the system according to specific user
needs.18 How a given source text in a given file format is imported and presented in the
TM system can be adjusted in the import/filter component of the system. For example,
when importing Microsoft Word files, Trados Studio offers functions for treating special
characters as inline placeholder tags or for extracting comments as translatable text or
as Studio comments. When importing PowerPoint files, notes can be extracted and
presented after each slide or at the end of all slides or can be excluded from the ST
import, etc. Most current TM systems convert the translatable file into the XML-based
format XLIFF (XML Localization Interchange File Format) or into a similar format (such
as SDLXLIFF or MQXLIFF) after the import. XLIFF has become the standard format for
exchanging localisable content between different TM systems. Another step in the
translation preparation phase is the alignment19 of legacy texts. Before switching to the
integrated TM system approach with Trados Studio 2009, SDL offered the autonomous
tool WinAlign for aligning source and target texts. The alignment could then be
exported into various exchange formats and imported into a TM. In Trados Studio, the
alignment function is integrated directly into the TM system, with no additional export
and import processes required. TMs can be created and imported via the (nowadays
usually integrated) TM component of the TM system. Relevant questions in the context
of this sub-task are, for example, whether the TM system supports common inter-
change formats for TM data (particularly the XML-based format Translation Memory
eXchange, TMX), whether it supports the selection of multiple TMs, the parallel
selection of TMs with different language combinations (for example, SDLs AnyTM
function), etc. The creation and import of terminology databases and the preparatory
terminology work (which involves populating these termbases prior to the actual
translation) are usually done in autonomous terminology management systems such as
MultiTerm (SDL), crossTerm (Across) or qTerm (Kilgray). With respect to termbase
creation, a relevant question would be whether the respective terminology management
system supports data structures according to the meta-model specified in ISO 16642
(2003) “Computer Applications in Terminology – Terminological Markup Framework”
(cf. Arntz/Picht/Schmitz 1989/2014: 242). The seamless import (and export) of term-
bases depends on whether the respective terminology management system supports

18
The configurability of TM systems will be taken up again in the usability discussion in section 4.2.2
below.
19
In the translation process, an alignment creates “TM databases from previously translated documents
that are only available as separate source and target text files by comparing a source text and its
translation, matching the corresponding segments, and binding them together as units in a TM”
(Reinke 2013a: 30).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 125
and Modelling Their Usability

common interchange formats for terminology data (particularly the XML/SGML-based

formats TermBase eXchange, TBX, Machine-Readable Terminology Interchange
Format, MARTIF, and Open Lexicon Interchange Format, OLIF; cf. Drewer/Ziegler
2011: 181). The search for parallel or explanatory texts is usually not supported by any
CAT tools understood in a narrow sense. In applied corpus-based translation studies,
various researchers propose a structured approach to compiling corpora of parallel/
explanatory texts from the internet (the so-called Web for Corpus or WfC approach;
see, for example, Krüger 2012: 515-518). It is certainly true that the internet has
become the most important research source in professional translation. At the same
time, electronic corpora are gaining in importance both in translation research and in
translation practice. However, it is debatable whether the internet and electronic
corpora can be considered prototypical CAT tools or translation technology in a narrow
sense as understood in this article. In the Cologne Model of the Situated LSP Trans-
lator, these tools are included in the artefact group digital research and communication
resources and are therefore excluded from the present discussion of the computer-
assisted translation process.20
The actual translation phase involves the (segment-based) source text reception
and target text production, editing 100 % and fuzzy matches, storing new and/or
revised translations in the translation memory, storing new terms in the termbase,
querying TM(s) and termbase(s) and querying parallel/explanatory texts and
glossaries. The (segment-based) ST reception and TT production are supported by the
editor component of the TM system and, in the case of TT production, also by the
machine translation component of the TM system (if available, see the discussion of
the interactive processing of translation segments by MT systems in footnote 13).
Functions related to the ST reception in the editor include the presentation of
placeables (for example, the options No Tag Text, Partial Tag Text and Full Tag Text in
Trados Studio), the adaptation of font sizes used for displaying the translatable text
(see the usability discussion in section 4.2.2 below), etc. The TT production can be
supported, among other things, by functions simplifying the writing process (for
example, the AutoCorrect, AutoSuggest and QuickPlace functions of Trados Studio),
by keyboard shortcuts for commonly used functions such as Confirm Segment or Move
to Next Unconfirmed Segment, etc. 100 % and fuzzy matches can usually be edited
directly in the editor or in the TM component of the translation memory system. The
same holds for storing new or revised translations in the TM. The termbase can usually

20 At the European Master’s in Translation (EMT) Network Meeting in Riga in November 2015, the
Working Group “Tools and Technology” discussed a research proposal entitled “Use of Corpora as
CAT Tools”. The participants were indecisive whether electronic corpora should be treated as
prototypical CAT tools or not. However, there was unanimous agreement that these resources offer
enormous potential for translation teaching and practice and that they should be assigned a more
prominent place in translation technology research. In the light this, future versions of the Cologne
Model of the Situated LSP Translator may well include electronic corpora in the artefact group
translation technology in a narrow sense, which would make them an integral component of the
computer-assisted translation process.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 126
and Modelling Their Usability

be populated with new terms or queried directly from the editor of the TM system or in
the external terminology management tool. Most current TM systems also include
functions for automated term lookup, i.e., if the source segment includes a term from
the specified termbase, this term, along with available additional information, will be
presented automatically in the term recognition window. The translation memory can
usually be queried directly from the editor or from the TM component of the translation
memory system. The concordance feature can be used to retrieve instances of specific
search strings below match level from the TM and to display these strings in their
immediate context (cf. Reinke 2013a: 20). Querying parallel/explanatory texts or
glossaries (in electronic form) is usually not supported by any CAT tools in the narrow
sense. However, in this context, translators can use specific web concordancers such
as WebCorp Live (2015) in order to access the internet as a corpus (the so-called Web
as Corpus or WaC approach; see, for example, Krüger 2012: 518-520). Although these
web concordancers are very useful tools both in translation teaching and practice, they
are assigned to the artefact group translation technology in a wider sense in the
Cologne model and are hence excluded from the present discussion of the computer-
assisted translation process.
The quality control phase involves checking the target text for correctness of
content, for stylistic correctness, for terminological correctness/consistency and for
correct grammar/spelling/punctuation as well as checking the placeable elements in the
target text, checking the TT layout, sending the TT for external review and exchanging
queries with the customer. Checking the TT for correctness of content is mainly a
manual task which is not supported by any specific CAT tools. Some tools offer
functions for comparing the length of ST and TT segments (significant differences
between ST and TT segment length may be an indicator of inadmissible omissions or
additions in the target text), and terminology checkers can be configured to alert the
user when a TT segment does not include an equivalent for a particular ST term
present in the corresponding ST segment. However, these functions operate merely at
the formal level and usually cannot contribute in any significant way to this quality
control task.21 On the other hand, checking the target text for stylistic correctness can
be supported by quality assurance (QA) tools featuring a specific style checker. An
example for this would be the authoring support tool Acrolinx (cf. Siegel/Lieske
2015: 62). Checking the TT for terminological correctness/consistency can be supported
by specific terminology checkers, which can be part of external QA tools, such as
Acrolinx, D.O.G. Error Spy or ApSIC Xbench, or integrated components of TM systems
(for example, the integrated Terminology Verifier of Trados Studio). The same holds for
checking the grammar, spelling and punctuation of the target text. This task can also

21 In fact, these functions often lead to false positives, i.e. alleged errors which, upon closer
investigation, turn out to be no errors at all (see, for example, Reinke 2013b: 42-43). Depending on
the amount of such false positives, these functions will actually impede (instead of support) the
human translation process.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 127
and Modelling Their Usability

be supported by external QA tools such as the ones listed above or by internal QA

components of most TM systems. Also, most external QA tools and TM systems offer
functions for checking the placeable elements in the target text (for example, the
integrated Tag Verifier of Trados Studio). The layout of the target text can be checked
by converting the preliminary TT version into the final format or – and this is usually the
more economic approach – by using the integrated preview function of the TM system
(if available). The external review of the target text can often be supported by the PM
component or the review workflow of the TM system. Trados Studio, for example,
includes the batch task Export for Bilingual Review. This exports the translated
document into a bilingual Microsoft Word file with the option Track Changes activated.
After review, the translator can reimport this Word file into Studio using the batch task
Update from Bilingual Review and can accept or reject the reviewer's changes in line
with the respective workflow in Microsoft Word.22 The task of exchanging queries with
the customer is not normally supported by specific CAT tools or functions. However,
with the proliferation of cloud-based translation memory systems (such as GlobalSight
or Memsource), it seems feasible to integrate this information exchange process into
the overall PM workflow of such systems.23
The phase of final administrative tasks involves converting the target text into the
final format (if requested by the client), delivering the translation (for example, via email
or FTP server), archiving the translation project, invoicing the project and keeping
accounts. The TT conversion into the final format is supported by the export/filter
component of the TM system (see the discussion of the import/filter component in the
translation preparation phase). The delivery of the translation can be supported by the
PM component of the TM system. For example, when creating a return package within
the translation package workflow of Trados Studio, the programme offers the function
Send Package by Email, which creates a draft email (in Microsoft Outlook) with the
project package attached. The TM system memoQ, on the other hand, features an
online project workflow. Here, the translator can deliver the translation directly from
memoQ by uploading it to the designated server. The PM component of a TM system
may also support archiving the translation project (at least within the TM system). For
example, in the Projects view of Trados Studio, finished projects can be set to the
status Completed, and the view offers filter options for displaying or hiding completed
projects. Of course, from the translator’s perspective, archiving translation projects
goes beyond this tool-internal sub-step and includes, for example, grouping projects
completed in previous years in corresponding folders, storing completed projects on
external hard drives, etc. Invoicing the translation project again is supported by the
statistics/analysis component of the TM system. Since, nowadays, most translation

22 This workflow is particularly useful if the reviewer does not work with the same translation memory
system as the translator. When the translation is exported into a bilingual Word file, the review
process becomes independent of a specific TM system.
23 For an up-to-date overview of cloud-based translation memory systems, see Mieth (2015).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 128
and Modelling Their Usability

projects are invoiced based on the number of translated words – broken down into
different match categories such as perfect match, 100 % match, repetition, fuzzy match
(usually various subcategories) and no match –, the initial analysis of the translatable
files will generally serve as basis for calculating the translation fee.
Finally, the phase of follow-up work involves tasks such as incorporating customer
feedback, dealing with customer complaints and maintaining contact with the customer
in order to acquire future translation jobs. The only task of this phase which can be
supported by CAT tools is the incorporation of customer feedback. This task can be
supported by the PM component/review workflow of the TM system (see the discussion
of the task of sending the target text for external review in the quality control phase)
and by the editor/TM component of the system. Ideally, the TM system should allow the
user to assign attributes such as From Customer Review or Approved by Customer to
the revised translation units so that the customer's preferences can be clearly identified
when querying the TM in future translation scenarios.
This concludes the overview of the computer-assisted translation process and the
contextualisation of CAT tools from a professional perspective. Although I tried to
discuss the different phases of the computer-assisted translation process and the
respective sub-tasks in some detail, the overview is by no means exhaustive. This is
due to the potentially very complex nature and wide scope of this process. In this
overview, I have attempted to find a balance between completeness of description and
ease of presentation. In any case, what was still missing from the present discussion is
the usability dimension of the various external tools or internal components/functions
illustrated above. This usability dimension of CAT tools will be the main concern of the
following sections.

4 Modelling the Usability of CAT Tools

As mentioned in the introduction of this article, translation studies publications specifically
and exhaustively concerned with the usability of CAT tools from a user-oriented
perspective still seem to be relatively scarce.24 A cursory search in relevant databases
such as Translation Studies Abstracts Online and BITRA as well as a corresponding
internet search yielded three PhD theses (Höge 2002; Dragsted 2004; Lagoudaki
2008), one MA thesis (Guillardeau 2009) and several academic articles (e.g. Dillon/
Fraser 2006; Dragsted 2006; Lagoudaki 2006; Colominas 2008; Christensen/
Schjoldager 2011; Campbell et al. 2013). Most of these publications only seem to focus
on particular aspects of overall CAT tool usability (for example, Dragsted 2004 and

24 That the usability dimension is indeed underrepresented in translation technology research is also
evidenced by the Routledge Encyclopedia of Translation Technology (Chan 2015). This publication
which, by nature, should aim at a more or less exhaustive overview of the field of translation
technology, contains no proper entry on usability (although the term is included in the index and
appears several times in the publication).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 129
and Modelling Their Usability

2006 focus on the cognitive impact of segmentation, and Colominas 2008 is concerned
with the precision of sub-sentential segmentation), on specific types of CAT tools
(Guillardeau 2009 is concerned with open-source TM systems) or they do not derive
their usability definition from the most current or most widely used usability standards
(for example, Höge 2002: 92 works with the usability definition of ISO/IEC 9126,
whereas most current usability models are based on ISO 9241; see the discussion
below). Also, there is a relatively new research strand concerned with the ergonomics
of the translation process (O’Brien 2012; Ehrensberger-Dow/Massey 2014a,b;
Ehrensberger-Dow/O’Brien 2015). Within this research strand, there are three research
foci, i.e. cognitive, physical and organizational ergonomics. Issues pertaining to trans-
lation software (and hardware) are investigated in the context of cognitive ergonomics
(cf. Ehrensberger-Dow/O’Brien 2015: 106), which is concerned with “mental processes,
such as perception, memory, reasoning, and motor response, as they affect interactions
among humans and other elements of a system” (International Ergonomics Association
n. d.). However, the issue of CAT tool usability is not addressed specifically in cognitive
ergonomics research, and the investigation of translation technology remains at a
rather coarse-grained level, being mostly concerned with shortcomings of user inter-
faces and the possibility to customise tool settings according to individual preferences
(Ehrensberger-Dow/Massey 2014a: 204; ErgoTrans 2015: 3-4). Therefore, it seems
that translational ergonomics may provide a high-level theoretical framework for CAT
tool usability research (as does Situated Translation) but – probably due to the very
wide scope of this approach – there have not been any detailed theoretical or empirical
contributions to CAT tool usability research yet. In the present article, I will present a
draft of a general model of CAT tool usability which is based on current usability
standards and which attempts to capture all relevant usability aspects of CAT tools
from a user-oriented perspective. Based on this general model of CAT tool usability, I
propose a draft of a more specific model of translation memory system (TMS) usability.
But first, I will discuss in some detail the term usability, which was evoked already at
several points in this article but not yet properly defined.
In an article for the Software Quality Journal, Abran, Khelifi and Suryn (2003: 324)
point out that the term usability “has not been defined homogeneously, either by the
researchers or by the standardization bodies” and that the issue can be approached
from different perspectives. In this context, the authors make a distinction between
product-oriented standards such as ISO/IEC 9126 (2001) “Software Engineering –
Product Quality”25 and process-oriented standards such as ISO 9241 (2011) “Ergo-
nomics of Human-System Interaction”26 (Abran/Khelifi/Suryn 2003: 324). From a
product perspective, usability is defined as “[t]he capability of the software product to
25 In the meantime, the standard has been replaced by the standard ISO/IEC 25010 (2011) “Systems
and Software Engineering – Systems and Software Quality Requirements and Evaluation (SQuaRE)
– System and Software Quality Models”.
26 Previously, ISO 9241 was called “Ergonomics Requirements for Office Work with Visual Display
Terminals (VDTs)”.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 130
and Modelling Their Usability

be understood, learned, used and attractive to the user when used under specific
conditions” (ISO/IEC 9126-1 2001). In ISO/IEC 9126 (2001), software usability is
constituted by the five sub-dimensions of understandability, learnability, operability,
attractiveness and usability compliance (cf. Abran/Khelifi/Suryn 2003: 326). From a
process perspective, on the other hand, usability is defined as “[t]he extent to which a
product can be used by specified users to achieve specified goals with effectiveness,
efficiency and satisfaction in a specific context of use” (ISO 9241-11). According to
Abran and Khelifi and Suryn (2006: 330), the process-oriented standard ISO 9241
takes a broader perspective on usability than the product-oriented standard ISO/IEC
9126. If we compare the definitions of usability in the two standards, it becomes
obvious that the definition in ISO 9241 places more emphasis on the fact that usability
is dependent on specific goals, specific users and specific contexts of use. This is in
line with Rudlof’s (2006: 15) observation that usability is not an in vitro quality but can
only be established relative to a particular context of use.27 This in vivo quality of the
usability concept is probably better reflected by the process-oriented approach than by
the product-oriented approach to this topic.28 However, Abran and Khelifi and Suryn
(2003: 330) point out that the product-based approach and the process-based
approach to usability do not exclude each other but that they are in fact complimentary
perspectives on the same topic. Accordingly, different authors (such as Dix et al.
1993/2004; Nielsen 1994; Abran/Khelifi/Suryn 2003) used the usability definition of IS0
9241 (2011) as a starting point for their usability models and complemented these
models with elements from ISO/IEC 9126 (2001) and/or other usability standards. The
CAT Tool Usability Model which I propose in figure 3 below is also derived primarily
from the usability definition of ISO 9241 (2011) but, as the models mentioned above, its
scope goes beyond this definition to include other usability dimensions.

27 For the purpose of the present article, the context of use is understood to include the users of a
specific software and their specific goals, which are stated separately in the ISO 9241 (2011)
definition of usability. A context of use with this broader extension is reminiscent of the functional
linguistic notion of context of situation (cf. Baker 2006: 324).
28 On the context-dependence of software usability, see also Abran, Khelifi and Suryn: “To verify
whether or not the required level of usability is achieved, it is necessary to measure the performance
and the satisfaction of users working with the product. The measurement of usability is a complex
interaction between users and context of use; this might produce different levels of usability per-
formance for the same product when it is used in different contexts.” (Abran/Khelifi/Suryn 2003: 329).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 131
and Modelling Their Usability

4.1 A General Model of CAT Tool Usability

Fig. 3: A general model of CAT tool usability

The three usability dimensions of effectiveness, efficiency and satisfaction discernible

in the CAT Tool Usability Model in figure 3 are derived from the ISO 9241 definition of
usability (see section 4 above). Effectiveness is primarily a qualitative dimension, which
describes how well a given user can achieve specified goals or perform specified tasks
using the software. Efficiency, on the other hand, is a quantitative dimension, which is
concerned with the effort required to achieve these goals or to perform these tasks.
Satisfaction, in turn, is a subjective dimension, which captures the way the users feel
about working with the software (cf. Rudlof 2006: 15). In the general model of CAT tool
usability above, I complemented these three dimensions with the dimension of
learnability, which constitutes one of the five sub-dimensions of usability in ISO/IEC
9126 (2001) (see section 4 above) and which various authors, such as Dix et al.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 132
and Modelling Their Usability

(1993/2004), Nielsen (1994) and Abran, Khelifi and Suryn (2003), integrated into their
own usability models. Learnability, as understood in the general model of CAT tool
usability, is concerned with how easily new users can familiarise themselves with a
given software system and which resources they can draw on in this process.
As can be seen in figure 3, the general model of CAT tool usability is embedded in
a specific context of use. This is intended to highlight the in vivo nature of usability, i.e.
the fact that usability can only be established relative to specific users pursuing specific
goals in a specific context (which is in line with the usability definition in ISO 9241). For
example, the suitability of the standard configuration of a given CAT tool cannot be
determined in isolation but only if we specify which task the configuration is to be
suitable for and in which situation the CAT tool is to be used.29 This emphasis on the
context of use or the situation in which a given CAT tool is employed is also in line with
the basic tenets of Situated Translation, which stresses the strict situation-dependence
of the translator’s cognitive performance (see section 2 above).30
I will discuss the CAT tool usability model depicted in figure 3 at a rather abstract
level since this is a very general model which basically serves as a coarse-grained
template for more explicit or finer-grained models of specific CAT tools such as
terminology management systems, machine translation systems or translation memory
systems. The TMS Usability Model depicted in figure 4 is such a fleshed out or
expanded version of the general CAT Tool Usability Model and will therefore be
discussed at a more detailed level.

4.1.1 The CAT Tool Usability Dimension of Effectiveness

As mentioned above, effectiveness is a qualitative dimension, which is concerned with
the quality of the task(s) performed with a given CAT tool (for example, the quality of a
terminology database, of an alignment, or of a human translation). If we want to make
judgements about the effectiveness of a given CAT tool, we must therefore establish
various quality parameters against which the task performed with the CAT tool can be
checked. I will address this dimension in more detail in the discussion of the TMS
Usability Model below.

29 In their evaluation of translation memory software for specific multilingual working environments,
Campbell et al. also stress the importance of taking into account the specific context of use of a TM
system: “We found that it was crucial to understand the organization’s context and workflow prior to
attempting to evaluate a specific piece of software. This step may seem obvious, but the contextual
inquiry provided the necessary context for understanding which of the issues uncovered in this
evaluation were specific to the software tool being evaluated and which issues were more intrinsic to
the overall workflow.” (Campbell et al. 2013: 2045-2046).
30 In this context, see also Ehrensberger-Dow and Massey: “Presumably, […] translation performance is
affected not only by what happens in the translator’s mind or on the computer screen, but also by how
translators interact with their technological, physical, and organizational environment. That interaction
may involve translators adapting environmental factors like tools and technology to suit their working
needs, or adjusting their own cognitive processes and physical actions to fit those same factors.”
(Ehrensberger-Dow and Massey 2014b: 59).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 133
and Modelling Their Usability

4.1.2 The CAT Tool Usability Dimension of Efficiency

The efficiency dimension is finer-grained than the dimension of effectiveness. One
important aspect of this dimension is the time required to perform a given task using
the CAT tool in question. This temporal aspect is covered by most of the major usability
models (see, for example, Abran/Khelifi/Suryn 2003: 334) and can be evaluated quite
easily by means of time measurements. Another important aspect of the efficiency
dimension is the reliability of the CAT tool, i.e. the number of errors occurring during
the task and the recovery performance in the event of an error (cf. Guillardeau
2009: 87). Since errors are usually more or less time-consuming, this usability aspect
also has a temporal dimension to it. The frequency of use of documentation is also an
efficiency aspect which is covered in most of the major usability models (cf.
Abran/Khelifi/Suryn 2003: 332, 334). The basic idea is that the more frequently the user
has to consult the documentation, the less intuitive and efficient working with the
software will be. The configurability/expandability of the CAT tool refers to how well the
tool can be customised according to specific work or user requirements and whether
the user can add new functions to the tool, and the suitability of the standard
configuration is concerned with how well this configuration is suited to perform a given
task in a given context of use. I will also discuss these aspects in more detail in the
context of the TMS Usability Model below. The aspect of workflow adaptability/integra-
tion with other software is the last aspect of the efficiency dimension. This aspect is
concerned with how flexibly the CAT tool can be adapted to different workflows and/or
how well it is integrated with other software. Examples would be the compatibility
between a given terminology management system and a given translation memory
system or the number of data exchange formats supported by these systems. With
reference to the research strand of translational ergonomics discussed in section 4
above, this workflow adaptability/integration of CAT tools can be considered as an
aspect of organizational ergonomics, which is concerned with “the optimization of
sociotechnical systems, including their organizational structures, policies, and pro-
cesses” (International Ergonomics Association n. d.).

4.1.3 The CAT Tool Usability Dimension of Satisfaction

The dimension of satisfaction is solely concerned with the user's satisfaction with the
functions and characteristics offered by the CAT tool. At a very general level, this
satisfaction can be established using standardised rating scales such as the Software
Usability Scale (SUS, Brooke 1996).31 When using this scale to measure the
satisfaction of users with a given piece of software, the participants are asked to rate
the following 10 aspects with answers ranging from Strongly Agree to Strongly
Disagree:

31 For an application of this scale in the context of translation technology research, see Campbell et al.
(2013). The scale can be accessed online under Usability.gov (2016).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 134
and Modelling Their Usability

1. I think that I would like to use this system frequently.

2. I found the system unnecessarily complex.
3. I thought the system was easy to use.
4. I think that I would need the support of a technical person to be able to use this system.
5. I found the various functions in this system were well integrated.
6. I thought there was too much inconsistency in this system.
7. I would imagine that most people would learn to use this system very quickly.
8. I found the system very cumbersome to use.
9. I felt very confident using the system.
10. I needed to learn a lot of things before I could get going with this system.
As mentioned above, this is a very general scale which can be used to measure user
satisfaction with a wide range of potentially very different systems. If the exact nature of
the system to be evaluated is determined, a rating scale which is tailored to the
characteristics of this system could be used. The usability dimension of satisfaction will
also be discussed in some more detail in the context of the TMS Usability Model below.

4.1.4 The CAT Tool Usability Dimension of Learnability

The first aspect of the learnability dimension is concerned with the effort the user has to
invest in order to become familiar with the CAT tool. From a product-oriented
perspective, this aspect would cover the intuitiveness of a given CAT tool. In most of
the major software usability models, the learnability dimension is reduced to the “time
to learn” which has to be invested before a user can handle a software product with
confidence (cf. Abran/Khelifi/Suryn 2003: 332, 334). However, Guillardeau (2009: 84-86)
points out, rightly in my opinion, that learnability is also dependent on the availability of
product documentation. In the general model of CAT tool usability, I specify further that
it is not only the availability of such documentation but also its scope, level of detail and
accuracy that will eventually contribute to the learnability of the software product. Also,
the support offered by the developer and/or vendor of the CAT tool is an important
contributor to learnability. Commercial vendors such as SDL, Across, Kilgray, etc.
usually offer a high level of support (which will not always be free of charge), whereas
the support level for open-source software such as OmegaT and Anaphraseus may
vary (cf. Guillardeau 2009: 85-86).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 135
and Modelling Their Usability

4.2 A Model of Translation Memory System (TMS) Usability

Fig. 4: A model of translation memory system (TMS) usability

In this section, I will be concerned with the more specific model of translation memory
system usability – TM systems being the central tools used in the computer-assisted
translation process. Compared to the general model of CAT tool usability discussed in
section 4.1, the four usability dimensions have been fleshed out in more detail in this
model. The TMS Usability Model is also embedded in a specific context of use in order
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 136
and Modelling Their Usability

to do justice to the in vivo nature of TMS usability (see the discussion in section 4.1
above).

4.2.1 The TMS Usability Dimension of Effectiveness

In the TMS Usability Model, we can address the effectiveness dimension in more detail
since we know the nature of the product to be produced using the CAT tool, i.e. we are
concerned with the quality of a translation product. The topic of translation quality has
been addressed in detail both in translation studies (for example, House’s 1997 model
of translation quality assessment) and in the translation industry (for example, the
standard SAE J2450 2005 “Translation Quality Metric”), so there should be enough
frameworks available for measuring the quality of the translation product produced with
a given TM system.32 In the TMS Usability Model above, I have depicted three tool-
internal aspects which I consider to be of particular relevance in producing a high-
quality translation.
The first tool-internal aspect probably contributing to translation quality is the
representation of the translation units in their wider and/or their original context. The
wider context refers to the immediate co-text of a given translation unit, i.e. the text
immediately preceding and following the translation unit. Serrano Piqueras (2011: 123)
assumes that the (usually sentence-based) segmentation enforced by TM systems
may affect the translator’s text-pragmatic competence, for example their use of trans-
phrastic cohesive devices (such as pronouns or pronominal adverbs used to establish
anaphoric or cataphoric reference). If the translation memory system makes it easy for
the translator to keep the immediate co-text of a given translation unit in focus, this may
ultimately improve the cohesion of the target text (provided the translator’s text-
pragmatic competence is sufficiently developed).33 Both the presentation of translation
units in their wider context and the presentation of these units in their original context
can be supported by the preview function of the TM system (if available). For example,
memoQ features a live preview of the translation in the original format at the bottom of
the screen and therefore provides the translator with (potentially) valuable information
on formatting and co-text. Using an eye tracker, we could investigate if and how often
the translator consults this live preview and how often the translator fixates the co-text
of the present translation unit (in the editor or the preview). Also, if we focus on the
issue of sentence-based segmentation and TT cohesion, we could check the final
translation product for an adequate use of cohesive devices, such as pronouns,
pronominal adverbs, etc. If these data were compared with the data obtained from
another TM system (ideally used by the same translator to translate a similar source

32 For an up-to-date and exhaustive discussion of translation criticism, see Reinart (2014).
33 At the EMT Network Meeting in Riga in November 2015, the Working Group “Tools and Technology”
initiated a research project entitled “Use of CAT Tools and Translators’ Perception of Document
Coherence and Cohesion”. Basically, this research project sets out to investigate the segmentation
aspect of the TMS effectiveness dimension outlined above.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 137
and Modelling Their Usability

text), we could establish potential differences between the two data sets and check
whether these differences can be linked to any specific characteristics of the two TM
systems.
The second aspect of translation quality immediately affected by the TM system is
the completeness and precision of TM/concordance/termbase match retrieval (cf.
Lagoudaki 2006: 23). If the search results obtained from the translation memory or
from the termbase are not complete and adequately precise, this may affect, for
example, the internal consistency of the target text (at word/term, phrase, sentence or
paragraph level) or the external consistency of the TT with previous translations
produced for the customer. The retrieval performance of TM systems can be impacted,
among other things, by the occurrence of placeable and localizable elements in the
translation segments (cf. Azzano 2011). The general view in translation technology
research is that the retrieval performance of commercial TM systems still offers room
for improvement. For example, Reinke claims that “[a]lthough commercial TM systems
have been available for over two decades, their retrieval performance has not improved
considerably in terms of quality and quantity” (Reinke 2013a: 41).
The third and last tool-internal contributor to translation quality is the availability
and performance of integrated QA tools and the compatibility of the TM system with ex-
ternal QA tools. For example, if the translator performs the translation in a TM system
offering no integrated spellchecker, he or she will have to copy the translation into a
text-processing software such as Microsoft Word, probably remove formatting informat-
ion which should be excluded from the spell check, run the check, and correct the iden-
tified errors in the TM system. This process is not only rather time-consuming (which
would actually affect the efficiency dimension of TMS usability) but also very error-prone.
TM systems such as Trados Studio offer an integrated spell check using the Microsoft
Word spell checker. This allows the translator to use the Microsoft Word dictionary which
he or she may have customized during previous spell checks in Microsoft Word (or in
other TM systems supporting the Word spell checker). Also, Trados Studio can be
configured to ignore certain match types such as 100 %, context and perfect matches.
This is particularly useful if the text to translate is embedded in large amounts of
context/perfect matches and the customer instructs the translator to translate and
proofread only the new segments and leave the other segments alone (which are usually
not paid for in these kinds of assignments). The internal spell checker of Trados Studio
could then be configured to check only the newly translated matches; again, there is a
temporal aspect to this function, which would also affect the efficiency dimension of TMS
usability. The compatibility with external QA tools is mostly dependent on the file format
into which the TM system converts the files to translate. Ideally, the TM system should
use standardised formats such as XLIFF or formats derived from this format (such as the
Trados Studio format SDLXLIFF, see the discussion of the translation preparation phase
in section 3), which can be checked by most of the commercially available QA tools.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 138
and Modelling Their Usability

4.2.2 The TMS Usability Dimension of Efficiency

I now turn to the efficiency dimension of the TMS Usability Model. The time required to
complete the translation can be measured quite easily in respective usability
experiments. However, this parameter should probably not be interpreted in isolation
but rather be correlated with the quality of the translation product in order to account for
the fact that high-quality translations cannot be done in a rush but simply require a
certain amount of time. If we want to make comparative statements about this temporal
efficiency aspect of two TM systems, we would therefore have to check whether the
two translations produced with these TM systems are roughly similar in quality.
The reliability of a translation memory system is concerned with the number of
errors occurring during the translation and the recovery performance of the system in
the case of an error. One relevant question which could be asked in this context is
whether the TM system features a function for automatically saving the translation and
whether this function can be configured by the user. For example, Trados Studio
features a specific AutoSave function which can be activated or deactivated and for
which the user can specify a certain interval. If the TM system crashes during trans-
lation, the file can be restored to the last (auto)saved version.
The frequency of use of documentation was discussed briefly in the context of the
general model of CAT tool usability. I will address the issue of documentation in some
more detail in the discussion of the learnability dimension of TMS usability.
The configurability of the TM system refers to how well the tool can be customised
according to specific work and user requirements. In this context, we could also
evaluate the suitability of the standard configuration of the system. While the question
of suitability is usually dependent on the actual context of use (see the discussion in
4.1 above), there are also some relatively context-independent aspects to it. For
example, in the standard configuration of Trados Studio, the minimum ST and TT font
sizes are set to 8 (this can be changed in the editor options). Depending on the size of
the monitor the translator works with and the size of the editor window, text displayed in
font size 8 is hardly discernible to the eye. Regardless of the actual context of use, this
standard setting may therefore violate the principles of ISO 9241-12 (2011) “Presenta-
tion of Information”, which requires that on-screen information must be displayed in a
suitable font type and font size (cf. Rudlof 2006: 25-26). Another aspect that could be
mentioned in this context is the standard assignment of keyboard shortcuts to
frequently used functions such as the concordance function.34 For example, Trados
Studio features a concordance function for both source and target segments of the TM.
In most cases, the translator will probably want to search for unknown strings in the
source segments of the translation memory in order to check whether the TM contains
any suitable translation solutions. If the cursor is placed in the source segment of the

34 For a study on the general usefulness of the concordance function of TM systems, see O’Brien,
O’Hagan and Flanagan (2010).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 139
and Modelling Their Usability

editor, this function can be accessed quite easily via the function key F3; if the cursor is
placed in the target segment of the editor, the more complex key combination Ctrl+F3
has to be used. It is common practice for translators to copy the source-segment text
into the target segment and then overwrite this text with their translation (cf. Ehrensberger-
Dow/Massey 2014b: 71; this is particularly the case when the source segments contains
non-translatable elements such as proper names, numbers, email addresses, etc.).
Therefore, when the need to perform a concordance search (usually in the source
segments of the TM) arises, the cursor will often be placed in the target segment of the
editor. However, the key combination required to access the source segments of the
TM from a target segment in the editor view (Ctrl+F3) is quite inconvenient since it
forces the translator to remove his or her hands from their normal writing position on
the keyboard. Since this use of the concordance feature can be claimed to be the
prototypical one, the easier shortcut F3 (or another ergonomic shortcut) should be
assigned to it.35 In both cases, the standard configuration of the TM system can be
claimed to be suboptimal (regardless of any specific context of use), but at the same
time, the configurability of the tool allows the user to actually optimise the standard
configuration.36
If the TM system features a good expandability, it will allow the user to add new
functions to the system. The suitability of the standard functionality is concerned with
the question of whether the original functions of the system are sufficient to perform a
given translation task. Again, the issue of expandability seems to be strongly
dependent on the actual context of use (for example, on the individual user preferences
or the nature of the translation task to be performed with the TM system). On
TranslationZone.com, SDL offers a wide variety of apps and plugins (most of them free
of charge), which can be downloaded and integrated into Trados Studio to expand its
standard functionality (for example, apps for configuring customised menus, for
integrating machine translation services such as SYSTRAN into Studio, for performing
advanced number checks, etc.). Translators can use this wide variety of tools to
expand and fine-tune the functionality of Trados Studio according to their individual
work preferences. However, it should not be disregarded that both configuring and
expanding a TM system according to individual preferences or requirements is often a
rather time-consuming task which many practicing translators may have to skip
because of the (often enormous) time pressure involved in their work.37 This observa-
tion is confirmed by the results of the Capturing Translation Processes (CTP) research

35 The keyboard shortcuts can also be changed in the options of Trados Studio.
36 In this context, Krüger and Serrano Piqueras (2015: 14-15) point out that translators should attempt to
optimise the psycho-motor routines involved in the translation process (which would involve, among
other things, optimising the keyboard shortcuts used in the writing phase of the translation process).
The intensive use of peripheral devices in computer-assisted translation is illustrated in a correspond-
ing ergonomics study by Ehrensberger-Dow and Massey (2014a: 204).
37 In the Cologne Model of the Situated LSP Translator depicted in figure 1 above, this potential time
pressure is covered by the category physical and psychological factors.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 140
and Modelling Their Usability

project conducted at Zurich University of Applied Sciences. The results show that
translators, instead of adjusting CAT tool settings according to their personal pre-
ferences, “might be adapting their performance to software rather than the converse”
(Ehrensberger-Dow/Massey 2014b: 80).
If the TM system features a good workflow adaptability, it can be adapted flexibly
and easily to different workflows (as discussed in section 4.1.2, this workflow
adaptability can be considered as an aspect of organizational ergonomics). An
example of good workflow adaptability of Trados Studio is that the system offers
different translation workflows (the single document workflow, the package workflow
and the GroupShare workflow) for different translation scenarios. If the translator is
assigned a rather straightforward translation job (for example, a single file to translate
using a TM which the customer provided already for an earlier assignment), the
economic and quick single document workflow can be used. On the other hand,
complex assignments including multiple files and new translation memories will usually
be handled via the more extensive package workflow. An example of suboptimal
workflow adaptability of Trados Studio is that the tool creates fixed folder structures for
specific batch tasks such as Analyze Files, Export for Bilingual Review, etc. (see the
discussion of the computer-assisted translation process in section 3). These folder
structures, which cannot be changed in the options of the system, may be incompatible
with the translator’s preferred folder structures (which may be independent of specific
CAT tools). In this case, the file management workflow specified by the TM system is
imposed on the translator’ file management workflow. The integration of the TM system
with other software is concerned, for example, with the compatibility between a given
TM system and a given terminology management system. While systems provided by
the same vendor (for example, SDL Trados Studio and SDL MultiTerm) will usually be
optimally compatible, inter-vendor compatibility cannot be taken for granted.38
The speed and ease of use of TM/concordance/termbase match retrieval is con-
cerned with how quickly the user can perform these retrieval tasks using the TM
system (cf. Lagoudaki 2006: 23-24) – which will affect the overall time required to
complete the translation. For example, MultiTerm termbases can be queried directly
from the editor component of Trados Studio, relieving the translator from accessing
MultiTerm every time he or she wants to search for a specific term in the termbase.
This function is due to the good integration of Trados Studio with MultiTerm (see the
discussion of the previous efficiency category) and adds to the speed and ease of use
of termbase match retrieval. The concordance shortcuts in Trados Studio discussed in
the context of the configurability of TM systems could also be addressed in the present
category of the efficiency dimension. In this case, the shortcut of the prototypical
concordance search function is not very ergonomic or user friendly, which adversely
affects the overall speed and ease of use with which this task can be performed.

38 Although this inter-vendor compatibility has improved considerably with the introduction of standar-
dised XML-based data exchange formats such as XLIFF, TMX and TBX.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 141
and Modelling Their Usability

The last aspect of the efficiency dimension of TMS usability is the treatment of
placeable and/or localizable elements in the translation memory system. Questions
which are relevant in this context are, for example, how tags are displayed, whether the
system allows reordering tags or deleting unnecessary tags in the target text, how
hyperlinks or email addresses are displayed and localized, etc. Placeable and
localizable elements are a vast topic and certainly influence the overall usability of the
TM system but they cannot be addressed here in detail. For an exhaustive discussion
and investigation of these elements, see Azzano (2011).

4.2.3 The TMS Usability Dimension of Satisfaction

This dimension is concerned with the user’s subjective satisfaction with the functions
and characteristics offered by the TM system. In section 4.1.3 above, I discussed the
System Usability Scale (SUS), which is a rather coarse-grained scale that can be used
to measure user satisfaction with a wide range of potentially very different software
systems. If we want to measure user satisfaction with specific software systems such
as TM systems and if we also want to take into consideration the specific context of
use of these systems, this scale could be specified and expanded accordingly. For
example, question 1 of the SUS could be reformulated as “I think that I would like to
use this system frequently in context of use/translation scenario xy.”39 Also, question 3
of this scale could be tailored to specific functions of the system, for example, by
reformulating it as follows: “I found the concordance feature/the tag verifier/the PM
component, etc. of the TM system easy to use”. The System Usability Scale in its
original form should therefore be regarded primarily as a template which can be
adapted to measure user satisfaction with specific systems.40 Also, user satisfaction
could be correlated with various elements from the other TMS usability dimensions. For
example, we could measure the satisfaction of users working with the standard
configuration/functionality of the TM system and the satisfaction of users who
configured/expanded the system according to their individual work preferences. Such a
differentiated and context-sensitive TM System Usability Scale seems to be a
necessary supplement to the TM Usability Model. However, developing such a scale is
an ambitious undertaking, which is beyond the bounds of the present article.

39 An example of such a context of use or translation scenario would be a freelance setting, where a
freelance translator works regularly for translation agencies providing him/her with preprocessed
XLIFF files to translate.
40 Dillon and Fraser (2006: 77) developed a questionnaire in order to measure translators’ perception of
TM adoption. However, the questionnaire is mostly concerned with the overall impact of using TM
systems (for example, users can agree or disagree with statements such as “I would use out on work
if I did not have TM software”, or “Using TM makes translation easier”). Still, several elements of this
questionnaire could be used to develop a specific TM System Usability Scale.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 142
and Modelling Their Usability

4.2.4 The TMS Usability Dimension of Learnability

As discussed in the context of the general model of CAT tool usability (see section
4.1.4 above), the learnability dimension consists of three elements: 1) the effort that the
user has to invest in order to become familiar with the TM system; 2) the scope, level of
detail and accuracy of the documentation provided for the system; and 3) the support
offered by the provider of the TM system. While the effort to become familiar with the
TMS is partly dependent on strictly tool-internal aspects (i.e., the intuitiveness of the
TM system, for example, the self-descriptiveness of dialog boxes, menus, etc.; cf.
Rudlof 2006: 54), it will also depend on the quality of the documentation and the
support offered by the provider. If the documentation of the TMS system is comprehen-
sive, adequately detailed and free of errors and if the support offered by the TMS
provider is targeted to the user’s needs, this should support the translator’s process of
becoming familiar with the TM system. There is a link here between the learnability
dimension of TMS usability and the frequency of use of documentation as one aspect
of the efficiency dimension. If the TM system features a good intuitiveness, the need to
consult the documentation should be relatively low. Also, if the translator consults the
documentation and finds the required information right away and this information is
sufficiently detailed, the need for future consultation of the documentation may
decrease. Documentation and support services for a given TM system should be
understood in a wide sense, ranging from traditional product briefs, user guides and
technical data sheets (both in printed and electronic form) to online help and FAQ
pages to moderated forums hosted on the website of the TMS provider, YouTube
channels, webinars, etc. These resources, if available and of adequate quality, will
certainly add to the learnability and ultimately to the overall usability of the TM system.

5 Conclusion and Outlook

In this article, I attempted to contextualise CAT tools from a theoretical and a professional
perspective and to model the usability of these tools. The theoretical contextualisation
linked CAT tools to the cognitive translational theory of Situated Translation, which
claims that the translator’s cognition is not isolated in the head but emerges from the
translator’s interaction with his or her working environment.41 This link underlined the
general relevance of CAT tools for the translator’s cognitive performance. The Cologne
Model of the Situated LSP Translator served as a bridge between the theoretical and
the professional contextualisation of CAT tools. In this model, CAT tools are concep-
tualised as relevant artefacts in the translation environment and can therefore be
viewed as an integral part of the translational ecosystem and hence of the translator’s

41 Such a link is also proposed by Christensen: “As regards a theoretical frame for future TM research,
the paradigm of situated, embodied cognition seems highly useful, because it takes into account the
fact that translation as a distributed activity does not take place only within the brain of an individual
human.” (Christensen 2011: 155).
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 143
and Modelling Their Usability

cognition. The professional contextualisation of CAT tools took the form of an overview
of these tools in the different work phases of the computer-assisted translation process.
This overview was derived from the Cologne Model of the Situated LSP Translator and
illustrated which work phases and sub-tasks of the translation process can be
supported in which way by CAT tools. Then, the usability of CAT tools was moved into
focus. In this context, I first discussed the notion of usability from a general software
perspective, before proposing a coarse-grained general model of CAT tool usability
and a finer-grained model of translation memory system usability, which are both
structured along the four usability dimensions of effectiveness, efficiency, satisfaction
and learnability and which are embedded in a variable context of use.42 These models
attempt to capture all relevant aspects of CAT tool usability from a user-oriented
perspective and may be used as foundations for structured approaches to testing the
usability of CAT tools. Such structured approaches to CAT tool usability testing are a
prerequisite if we want to elicit meaningful empirical data on the different components
of the two usability models and to obtain a differentiated and holistic picture of CAT tool
usability. Also, if the two models are to be applied in order to measure the usability of
specific CAT tools, we need to define the specific contexts of use of these tools. At
various points in this article, I stressed the fact that usability is an in vivo concept which
can only be established for specific users, specific tasks and specific work situations (in
other words, for specific contexts of use).43 Both the different translation settings
identified by Reinke (2004: 101-103, see footnote 12 above) and the work phases of
the translation process depicted in the Cologne Model of the Situated LSP Translator
may serve as a template for specifying such contexts of use.
It is a desideratum that the results of well-structured CAT tool usability research be
fed back to the developers of translation technology, which eventually may lead these
developers to optimise the usability of their tools to the benefit of their users. Cognitive
translational theories such as Situated Translation stress the relevance of such high-
usability technology for the professional translator, a fact which is also underlined by
Ehrensberger-Dow and O’Brien in the context of translational ergonomics:
If we take seriously the notion of translation being a situated activity, then we have to
provide opportunities for translators to explain what their personal preferences are and to
have their voices heard by designers of translation workspaces and language technology.
(Ehrensberger-Dow/O’Brien 2015: 112)

42 In the two models, there are some interdependencies between the different dimensions, particularly
between the effectiveness and efficiency dimensions (see the corresponding discussions above). For
the sake of simplicity, I refrained from making explicit these interdependencies by means of corres-
ponding arrows in the models.
43 In the discussion of the two models, I pointed out that some components of these models are more
context-dependent (for example, the suitability of the standard configuration/functionality of the CAT
tool/TMS) while others are more context-independent (for example, the standard font size and
keyboard shortcuts in TM systems such as Trados Studio). Again, for the sake of simplicity, I did not
reflect this relative context-(in)dependence of the various components explicitly in the two models.
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 144
and Modelling Their Usability

Ideally, then, the efforts of translational ergonomics and CAT tool usability research will
lead to the development of more “caring” translation technology (O’Brien 2012: 103),
which effectively supports the situated translator as the central agent of the computer-
assisted translation process.

References
Abran, Alan; Adel Khelifi, Witold Suryn (2003): “Usability Meanings and Interpretations in ISO
Standards.” Software Quality Journal [11]: 323-336
Adams, Fred; Ken Aizawa (2010): “Defending the Bounds of Cognition.” Richard Menary (ed.):
The Extended Mind. Cambridge: MIT Press, 67-80
Arntz, Reiner; Heribert Picht, Klaus-Dirk Schmitz (1989): Einführung in die Terminologiearbeit.
7th ed. 2014. Hildesheim: Olms
Austermühl, Frank (2012): “Using Concept Mapping and the Web as Corpus to Develop
Terminological Competence among Translators and Interpreters.” Translation Spaces [1]:
54-80 – https://benjamins.com/#catalog/journals/ts.1.09aus/fulltext (10.11.2015)
Azzano, Dino (2011): Placeable and Localizable Elements in Translation Memory Systems.
PhD Thesis Munich: Ludwig Maximilian University –
https://edoc.ub.uni-muenchen.de/13841/2/Azzano_Dino.pdf (10.11.2015)
Baker, Mona (2006): “Contextualization in Translator- and Interpreter-mediated Events.” Journal
of Pragmatics, Special Issue: Translation and Context 38 [3]: 321-337
Brooke, John (1996): “SUS – A Quick and Dirty Usability Scale.” Patrick W. Jordan, Bruce
Thomas, Bernard A. Weerdmeester, Ian L. McClelland (eds): Usability Evaluation in
Industry. London: Taylor & Francis, 189-194
Bundgaard, Kristine; Tina Paulsen Christensen, Anne Schjoldager (2016): “Translator-computer
Interaction in Action – an Observational Process Study of Computer-aided Translation.”
JoSTrans – The Journal of Specialised Translation [25]: 106-130 –
http://www.jostrans.org/issue25/art_bundgaard.pdf (05.02.2016)
Campbell, Susan G.; Sarah C. Wayland, Alina Goldman, Sergey Blok (2013): “Speaking the
User’s Language: Evaluating Translation Memory Software for a Linguistically Diverse
Workplace.” Proceedings of the 57th Annual Meeting of the Human Factors and
Ergonomics Society, 2042-2046
Chan, Sin-Wai (ed.) (2015): The Routledge Encyclopedia of Translation Technology. London/
New York: Routledge
Christensen, Tina Paulsen (2011): “Studies on the Mental Processes in Translation Memory-
assisted Translation – the State of the Art.” trans-kom 4 [2]: 137-160 – http://www.trans-
kom.eu/bd04nr02/trans-kom_04_02_02_Christensen_Translation_Memory.20111205.pdf
(05.02.2016)
Christensen, Tina Paulsen; Anne Schjoldager (2010): “Translation-Memory (TM) Research:
What Do We Know and How Do We Know It?” Hermes – The Journal of Language and
Communication [44]: 1-13 –
http://pure.au.dk/portal/files/10113/Hermes-44paulsen_christensenschjoldager.pdf
(10.11.2015)
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 145
and Modelling Their Usability

Christensen, Tina Paulsen; Anne Schjoldager (2011): “The Impact of Translation-Memory (TM)
Technology on Cognitive Processes: Student-Translators' Retrospective Comments in an
Online Questionnaire.” Bernadette Sharp, Michael Zock, Michael Carl, Arnt Lykke Jakobsen
(eds): Proceedings of the 8th International NLPCS Workshop: Special Theme: Human-
Machine Interaction in Translation. 20.-21. August 2011, Copenhagen, Denmark, 119-130
Colominas, Carme (2008): “Towards Chunk-based Translation Memories.” Babel 54 [4]: 343-
354
Cooper, Alan (2004): The Inmates Are Running the Asylum: Why Hi-Tech Products Drive Us
Crazy and How to Restore the Sanity. Indianapolis: Sams
Dillon, Sarah; Janet Fraser (2006): “Translators and TM: An Investigation of Translators’
Perceptions of Translation Memory Adoption.” Machine Translation 20 [2]: 67-79 –
http://link.springer.com/article/10.1007%2Fs10590-006-9004-8#/page-1 (10.11.2015)
Dix, Alan; Janet Finlay, Gregory D. Abowd, Russel Beale (1993): Human-Computer Interaction.
3rd ed. 2004. Upper Saddle River: Prentice Hall
Dragsted, Barbara (2004): Segmentation in Translation and Translation Memory Systems – An
Empirical Investigation of Cognitive Segmentation and Effects of Integrating a TM System
into the Translation Process. Copenhagen: Samfundslitteratur
Dragsted, Barbara (2006): “Computer-Aided Translation as a Distributed Cognitive Task.”
Pragmatics & Cognition 14 [2]: 443-464
Drewer, Petra; Wolfgang Ziegler (2011): Technische Dokumentation. Eine Einführung in die
übersetzungsgerechte Texterstellung und das Content-Management. Würzburg: Vogel
Ehrensberger-Dow, Maureen; Gary Massey (2014a): “Translators and Machines: Working
Together.” Wolfram Baur, Brigitte Eichner, Sylvia Kalina, Norma Keßler, Felix Mayer,
Jeanette Ørsted (eds): Man vs. Machine? The Future of Translators, Interpreters and
Terminologists. Proceedings of the XXth FIT World Congress. Berlin: BDÜ Fachverlag,
199-207

trans-kom ISSN 1867-4844

trans-kom ist eine wissenschaftliche Zeitschrift für Translation und Fachkommunikation.
trans-kom veröffentlicht Forschungsergebnisse und wissenschaftliche Diskussionsbeiträge zu Themen
des Übersetzens und Dolmetschens, der Fachkommunikation, der Technikkommunikation, der Fach-
sprachen, der Terminologie und verwandter Gebiete.
Beiträge können in deutscher, englischer, französischer oder spanischer Sprache eingereicht werden.
Sie müssen nach den Publikationsrichtlinien der Zeitschrift gestaltet sein. Diese Richtlinien können von
der trans-kom-Website heruntergeladen werden. Alle Beiträge werden vor der Veröffentlichung
anonym begutachtet.
trans-kom wird ausschließlich im Internet publiziert: http://www.trans-kom.eu

Redaktion
Leona Van Vaerenbergh Klaus Schubert
University of Antwerp Universität Hildesheim
Arts and Philosophy Institut für Übersetzungswissenschaft
Applied Linguistics / Translation and Interpreting und Fachkommunikation
Schilderstraat 41 Universitätsplatz 1
B-2000 Antwerpen D-31141 Hildesheim
Belgien Deutschland
[email protected] [email protected]
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 146
and Modelling Their Usability

Ehrensberger-Dow, Maureen; Gary Massey (2014b): “Cognitive Ergonomic Issues in

Professional Translation.” John W. Schwieter, Aline Ferreira (eds): The Development of
Translation Competence: Theories and Methodologies from Psycholinguistics and
Cognitive Science. Cambridge: Cambridge Scholars, 58-86
Ehrensberger-Dow, Maureen; Sharon O’Brien (2015): “Ergonomics of the Translation Work-
place – Potential for Cognitive Friction.” Translation Spaces 4 [1]: 98-118
ErgoTrans (2015): Cognitive and Physical Ergonomics of Translation – Final Report –
https://www.zhaw.ch/storage/linguistik/forschung/uebersetzungswissenschaft/ergotrans_fin
al_report.pdf (20.04.2016)
Gouadec, Daniel (2007): Translation as a Profession. 2nd ed. 2010. Amsterdam/Philadelphia:
Benjamins
Guillardeau, Sébastien (2009): Freie Translation Memory Systeme für die Übersetzungspraxis:
Ein kritischer Vergleich. MA Thesis Wien: Universität Wien –
http://www.omegat.org/resources/DA_Guillardeau.pdf (10.11.2015)
Höge, Monika (2002): Towards a Framework for the Evaluation of Translators’ Aids’ Systems.
PhD Thesis, Department of Translation Studies, University of Helsinki –
https://helda.helsinki.fi/bitstream/handle/10138/19236/towardsa.pdf?sequence=2
(10.11.2015)
Holz-Mänttäri, Justa (1984): Translatorisches Handeln. Theorie und Methode. Helsinki: Suoma-
lainen Tiedeakatemia
House, Juliane (1997): Translation Quality Assessment. A Model Revisited. Tübingen: Narr
International Ergonomics Association (n.d.): “What Is Ergonomics? Definitions and Domains of
Ergonomics.” – http://www.iea.cc/whats/index.html (20.04.2016)
ISO 16642 (2003): Computer Applications in Terminology – Terminological Markup Framework.
Geneva: International Organization for Standardization
ISO 17100 (2015): Translation Services – Requirements for Translation Services. Geneva:
International Organization for Standardization
ISO 9241 (2011): Ergonomics of Human-System Interaction. Geneva: International Organiza-
tion for Standardization
ISO/IEC 25010 (2011): Systems and Software Engineering – Systems and Software Quality
Requirements and Evaluation (SQuaRE) – System and Software Quality Models. Geneva:
International Organization for Standardization
ISO/IEC 9126 (2001): Software Engineering – Product Quality. Geneva: International Organiza-
tion for Standardization
Kade, Otto (1968): Zufall und Gesetzmäßigkeit in der Übersetzung. Leipzig: Enzyklopädie
Keller, Nicole (2011): “Neun auf einen Blick.” MDÜ [5] –
http://www.bdue-fachverlag.de/download/mdue/1090 (10.11.2015)
Krings, Hans P. (1986): Was in den Köpfen von Übersetzern vorgeht. Eine empirische Unter-
suchung zur Struktur des Übersetzungsprozesses an fortgeschrittenen Französischlernern.
Tübingen: Narr
Krüger, Ralph (2012): “Working with Corpora in the Translation Classroom.” Languages in the
Globalised World: Interpreting Multilingual Literacy and Cross-Cultural Encounters, Special
Issue: Studies in Second Language Learning and Teaching 2 [4]: 505-525 –
https://repozytorium.amu.edu.pl/jspui/bitstream/10593/5724/1/SSLLT%202%284%29%205
05-525%20Kruger.pdf (10.11.2015)
Krüger, Ralph (2015): “Fachübersetzen aus kognitionstranslatologischer Perspektive – das
Kölner Modell des situierten Fachübersetzers.” trans-kom 8 [2]: 273-313 – http://www.trans-
kom.eu/bd08nr02/trans-kom_08_02_01_Krueger_Modell.20151211.pdf (11.12.2015)
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 147
and Modelling Their Usability

Krüger, Ralph; Jesús Serrano Piqueras (2015): “Situated Translation in the Translation Class-
room.” Current Trends in Translation Teaching and Learning [7], 5-30 –
http://www.cttl.org/uploads/5/2/4/3/5243866/chapter_1_situated_translation_in_the_translati
on_classroom_revised.pdf (07.12.2015)
Krüger, Ralph (forthc.): “Situated LSP Translation from a Cognitive Translational Perspective.”
Lebende Sprachen 61 [2]
Lagoudaki, Elina (2006): “Translation Memories Survey 2006: Users‘ Perception Around TM
Use.” Translating and the Computer [28]: 1-29 – http://www.mt-archive.info/Aslib-2006-
Lagoudaki.pdf (10.11.2015)
Lagoudaki, Elina (2008): Expanding the Possibilities of Translation Memory Systems: From the
Translator’s Wishlist to the Developer’s Design. PhD Thesis London: Centre for Translation
Studies, University College London
Mieth, Christiane (2015): “Vernetzt übersetzt.” technische kommunikation 37 [6]: 29-33
Nielsen, Jakob (1994): Usability Engineering. Boston: Academic Press
O’Brien, Sharon (2012): “Translation as Human-Computer Interaction.” Translation Spaces 1
[1]: 101-122
O’Brien, Sharon; Minako O’Hagan, Marian Flanagan (2010): “Keeping an Eye on the UI Design
of Translation Memory: How Do Translators Use the ‘Concordance’ Feature?” European
Conference on Cognitive Ergonomics. Delft, 1-4 –
http://doras.dcu.ie/16693/1/ECCE_2010.pdf (05.02.2016)
Pym, Anthony (2011): “What Technology Does to Translation.” The International Journal for
Translation & Interpreting Research 3 [1]: 1-9 –
http://trans-int.org/index.php/transint/article/viewFile/121/81 (10.11.2015)
Reinart, Sylvia (2014): Lost in Translation (Criticism)? Auf dem Weg zu einer konstruktiven
Übersetzungskritik. Berlin: Frank & Timme
Reinke, Uwe (2004): Translation Memories. Systeme, Konzepte, Linguistische Optimierung.
Frankfurt/M: Lang
Reinke, Uwe (2013a): “State of the Art in Translation Memory Technology.” Translation:
Computation, Corpora, Cognition, Special Issue: Language Technologies for a Multilingual
Europe 3 [1]: 27-48 – http://www.t-c3.org/index.php/t-c3/article/view/25/29 (10.11.2015)
Reinke, Uwe (2013b): “Computergestützte Qualitätssicherung in der Fachtextübersetzung.”
Monika Krein-Kühle, Ursula Wienen, Ralph Krüger (eds): Kölner Konferenz zur Fachtext-
übersetzung (2010). Frankfurt/M.: Lang, 33-49
Risku, Hanna (2004): Translationsmanagement. Interkulturelle Fachkommunikation im Kommu-
nikationszeitalter. Tübingen: Narr
Risku, Hanna (2005): “Translations- und kognitionswissenschaftliche Paradigmen – Der
Mensch im Mittelpunkt.” Lew N. Zybatow (ed.): Translationswissenschaft im inter-
disziplinären Dialog. (Innsbrucker Ringvorlesungen zur Translationswissenschaft III.)
Frankfurt/M.: Lang, 55-69
Risku, Hanna (2010): “A Cognitive Scientific View on Technical Communication and Trans-
lation. Do Embodiment and Situatedness Really Make a Difference?” Target 22 [1]: 94-111
Risku, Hanna (2013): “Extended Translation. A Sociocognitive Research Agenda.” Target 25
[1]: 33-45
Rudlof, Christiane (2006): Handbuch Software-Ergonomie – Usability Engineering. Tübingen:
Unfallkasse Post und Telekom – http://www.ukpt.de/pages/dateien/software-ergonomie.pdf
(10.11.2015)
SAE J2450 (2005): Translation Quality Metric. Warrendale: Society of Automotive Engineers
Ralph Krüger trans-kom 9 [1] (2016): 114-148
Contextualising Computer-Assisted Translation Tools Seite 148
and Modelling Their Usability

Schubert, Klaus (2007): Wissen, Sprache, Medium, Arbeit. Ein integratives Modell der ein- und
mehrsprachigen Fachkommunikation. Tübingen: Narr – http://d-nb.info/1045615382/34
(10.11.2015)
Serrano Piqueras, Jesús (2011): Überlegungen zur Untersuchung des Einflusses von
Translation-Memory-Systemen auf die Übersetzungskompetenz. MA Thesis Köln: Institut
für Translation und Mehrsprachige Kommunikation, Fachhochschule Köln
Siegel, Melanie; Christian Lieske (2015): “Beitrag der Sprachtechnologie zur Barrierefreiheit:
Unterstützung für Leichte Sprache.” trans-kom 8 [1]: 40-78 – http://www.trans-
kom.eu/bd08nr01/trans-kom_08_01_03_Siegel_Lieske_Barrierefrei.20150717.pdf
(10.11.2015)
Siever, Holger (2010): Übersetzen und Interpretation. Die Herausbildung der Übersetzungs-
wissenschaft als eigenständige wissenschaftliche Disziplin im deutschen Sprachraum von
1960 bis 2000. Frankfurt/M.: Lang
Stolze, Radegundis (1994): Übersetzungstheorien. Eine Einführung. 6th ed. 2011. Tübingen:
Narr
Strohner, Hans (1995): Kognitive Systeme. Eine Einführung in die Kognitionswissenschaft.
Opladen: Westdeutscher Verlag
Usability.gov (2016): System Usability Scale (SUS) – http://www.usability.gov/how-to-and-
tools/methods/system-usability-scale.html (02.06.2016)
WebCorp Live (2015): WebCorp Live – http://www.webcorp.org.uk/live/ (02.06.2016)

Author
Ralph Krüger is a lecturer in translation studies, specialised translation and translation tech-
nology at the Institute of Translation and Multilingual Communication at Technische Hochschule
Köln (Cologne University of Applied Sciences), Germany. He holds a PhD in translation studies
from the University of Salford, UK. His main research interests include translation technology,
cognitive translation studies, the interface between scientific and technical translation and
cognitive linguistics, the didactics of specialised translation and the application of electronic
corpora in the translation classroom. Prior to joining academia, Ralph worked as LSP translator
for a major German translation agency.
Email: [email protected]
Web page: https://www.th-koeln.de/personen/ralph.krueger/
Neu bei Frank & Timme

TransÜD. Arbeiten zur TTT: Transkulturalität –

Theorie und Praxis des Translation – Transfer
Übersetzens und Dolmetschens
Herausgegeben von
Herausgegeben von Prof. Dr. Dörte Andres, Dr. Martina Behr,
Prof. Dr. Klaus-Dieter Baumann, Prof. Dr. Larisa Schippel,
Dr. Susanne Hagemann, Dr. Cornelia Zwischenberger
Prof. Dr. Dr. h.c. Hartwig Kalverkämper,
Prof. Dr. Klaus Schubert Marc Orlando: Training 21st century translators
and interpreters: At the crossroads of practice,
research and pedagogy. ISBN 978-3-7329-0245-3
Susanne Hagemann: Einführung in das trans-
lationswissenschaftliche Arbeiten. Ein Lehr-
Christian Trollmann: Nationalsozialismus auf
und Übungsbuch. ISBN 978-3-7329-0125-8
Japanisch? Deutsch-Japanische Beziehungen
1933–1945 aus translationssoziologischer
Anja Maibaum: Spielfilm-Synchronisation.
Sicht. ISBN 978-3-7329-0281-1
Eine translationskritische Analyse am Beispiel
amerikanischer Historienfilme über den
Zweiten Weltkrieg. ISBN 978-3-7329-0220-0

Franziska Heidrich: Kommunikationsoptimie-

rung im Fachübersetzungsprozess.
ISBN 978-3-7329-0262-0

FFF: Forum für

Fachsprachen-Forschung
Herausgegeben von
Prof. Dr. Dr. h.c. Hartwig Kalverkämper

Eva Martha Eckkrammer: Medizin für den

Laien: Vom Pesttraktat zum digitalen Rat­
gebertext. 2 Bände im Schuber.
ISBN 978-3-86596-312-3

Kommunikation – Partizipation –
Inklusion
Herausgegeben von
Dr. Bettina M. Bock,
Prof. Dr. Ulla Fix,
Daisy Lange

Nathalie Mälzer (Hg.): Barrierefreie Kommuni-

kation – Perspektiven aus Theorie und Praxis.
ISBN 978-3-7329-0231-6

Wittelsbacherstraße 27a, D-10707 Berlin

Frank & Timme Telefon (030) 88 66 79 11, Fax (030) 88 39 87 31
[email protected], www.frank-timme.de
Verlag für wissenschaftliche Literatur