<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Lost in Translation: Enterprise Architecture in e- Government Projects</article-title>
      </title-group>
      <contrib-group>
        <aff id="aff0">
          <label>0</label>
          <institution>Czech University of Life Sciences Prague, Faculty of Economics and Management, Department of Information Technologies</institution>
          ,
          <addr-line>Kamýcká 129, 165 00 Praha 6 - Suchdol</addr-line>
          ,
          <country country="CZ">Czech Republic</country>
        </aff>
      </contrib-group>
      <pub-date>
        <year>2015</year>
      </pub-date>
      <fpage>279</fpage>
      <lpage>287</lpage>
      <abstract>
        <p>e-Government projects in public administration are inherently complex and prone to delays and failures. Enterprise architecture (EA) is increasingly used as a method of practice to tackle the organization complexity and offers a common communication platform between various EA stakeholders. However, the attempts to translate EA terminology to audience who uses English as the second language (ESL) often result in misunderstandings and ambiguity. We present preliminary results of a survey among various stakeholders, ESL speakers, who are using ArchiMate, an EA modeling language, in e-Government projects. We found that using an EA standard created in English speaking environment and then translated or mixed with original language version creates issues with understanding among ESL speakers. Although, the agreement on meaning and usage of elements at the pragmatic level was prevailing, experts disagreed more often on the elements use at the semantic level. We suggest selecting a minimum viable set of elements and examples of their use for EA projects in the given domain in order to reduce ambiguity.</p>
      </abstract>
      <kwd-group>
        <kwd>Enterprise architecture</kwd>
        <kwd>public administration</kwd>
        <kwd>ambiguity</kwd>
        <kwd>ArchiMate</kwd>
        <kwd>ESL speakers</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>-</title>
      <p>Acknowledgement: This work was supported by the Faculty of Economics and Management, Czech
University of Life Sciences Prague grant no.2019MEZ0009</p>
    </sec>
    <sec id="sec-2">
      <title>1. Introduction</title>
      <p>
        e-Government projects in public administration are inherently complex and prone to delays and
failures. Many public agencies promote enterprise architecture (EA) as a method of managing
organizational complexity
        <xref ref-type="bibr" rid="ref1 ref3">(Hiekkanen et al., 2013; Lankhorst, 2017)</xref>
        and more efficient
implementation of e-Government and ICT projects
        <xref ref-type="bibr" rid="ref8">(Tamm, Seddon, Shanks, &amp; Reynolds, 2011)</xref>
        .
Besides being a time-consuming exercise, the common pitfall of EA creation is bias owing to expert's
subjective perception of the organization goals, strategy and operations
        <xref ref-type="bibr" rid="ref6">(Perez-Castillo,
RuizGonzalez, Genero, &amp; Piattini, 2019)</xref>
        . The problem is then multiplied by a lack of mutual
understanding between subject matter experts and a number of other involved stakeholders namely
project portolio managers, and enterprise architects who are often hired as external consultants. The
misunderstandings stem from a gap between perception and communication about real world
objects (Gustas &amp; Gustiené, 2004). Additionally, the threat of ambiguity and misunderstanding
rapidly increases when the enterprise architects translate the EA terminology from English to
another language in an attempt to convey the message to other stakeholders who use English as the
second language (ESL). Surprisingly, this topic is underresearched in the literature.
      </p>
      <p>
        The more complex the area for which the enterprise architecture should be developed, the bigger
the gap, and also the bigger ambiguity of perception and interpretation of real word objects can
grow. This is especially intensified in public administration with its complex, multifaceted and
rulebased nature
        <xref ref-type="bibr" rid="ref1">(Hiekkanen et al., 2013)</xref>
        . The subject matter experts in public administration are mostly
struggling with a lack of systematic and architecture approaches while portfolio project managers
and enterprise architects lack deep knowledge of the public administration domain. A deficit of
communication and collaboration then results in unwanted project and program delays and running
off the scope and budget. We assume that the gaps in conception between experts working on large
e-Government projects can be closed by using the visual and easy understandable elements. These
elements might stimulate willingness, mutual cooperation and knowledge sharing among teams
(Banaeianjahromi &amp; Smolander, 2019).
      </p>
      <p>The objective of our research is to focus on closing the gap in the way of thinking of subject matter
experts and portfolio project managers as well as enterprise architects in the public sector. We use
the ArchiMate modeling language to develop a high-level view of the enterprise architecture for the
State Land Office of the Czech Republic. The purpose of the proposed enterprise architecture is to
align e-Government initiative with the agency's efficient ICT operation and development. Based on
the collected data we formulate following research questions:</p>
      <p>RQ 1: What is the common understanding of enterprise architecture between ESL experts in
e-Government projects?
RQ 2: How can the enterprise architecture elements assist with closing the gap in
understanding between ESL experts in e-Government projects?</p>
      <p>This paper presents an overview of literature on the enterprise architecture levels and its creation,
and ArchiMate modeling language. Then we describe our research methods, present preliminary
findings of the survey and the proposal of EA for the public agency. Finally, we discuss the
implications of the findings and outline future research.</p>
    </sec>
    <sec id="sec-3">
      <title>2. Literature Review</title>
      <p>
        Enterprise architecture (EA) is a practice method for describing and managing an organization's
structure, infrastructure and processes in order to reduce the complexity of doing so. Enterprise
architecture core purpose is to describe altogether business, IT and evolution of an organization, i.e.
essentials that are more stable than particular information systems or software
        <xref ref-type="bibr" rid="ref3">(Lankhorst, 2017)</xref>
        .
The process of EA creation and management starts with description of the current state and future
state, then development of a transition plan and its implementation
        <xref ref-type="bibr" rid="ref2">(Kotusev, 2017)</xref>
        . EA creation is
realized at three levels: (1) the pragmatic level that concentrates on a strategic description of a
longterm in ; (2) semantic level that describes static and
dynamic structures of business processes across organization and technical system boundaries; and
(3) syntactic level which defines implementation details needed for the data processing in a specific
application or software component. (Abraham, Aier, &amp; Winter, 2015; Gustas &amp; Gustiené, 2004).
Shared understanding of all stakeholders involved in the EA creation process is a precondition for
success (Abraham et al., 2015).
      </p>
      <p>
        The Open Group ArchiMate standard provides an independent modeling language that aids
enterprise architecture
        <xref ref-type="bibr" rid="ref5">(Open Group, 2019)</xref>
        . ArchiMate provides graphical representation for all
three EA creation levels formulated by Gustas &amp; Gustiené (2004). The strategic aspects are covered
in motivation and strategic layers, static and dynamic structures of business processes are included
in the business layer, implementation details are part of the application and technology layers, while
the transition from a current to future state is described by the implementation layer. The standard
is supposed to model the enterprise architecture but also to serve as a communication tool between
various EA stakeholders. Due to the subjective bias each stakeholder often comes up with his or her
own interpretation of ArchiMate graphical elements
        <xref ref-type="bibr" rid="ref6">(Perez-Castillo et al., 2019)</xref>
        which hinders the
modeling effort (Chiprianov, Kermarrec, Rouvrais, &amp; Simonin, 2014) and undermines the ArchiMate
intended purpose. Therefore, the agreement between the experts on the definition of what each
element means in a specific area of public administration, especially in order to deliver projects and
programs in expected quality, is critically important .
      </p>
    </sec>
    <sec id="sec-4">
      <title>3. Methodology</title>
      <p>In order to indicate the level of EA elements understanding among ESL speakers we formed a
convenient sample of subject matter experts, portfolio project managers and IT experts working on
e-Government projects in the Czech Republic. We administered a survey via email including a short
explanatory letter and then followed up over the phone. As this is an ongoing research, we further
present only preliminary results drawn from 9 responses.</p>
      <p>The instrument consisted of 19 questions divided in four sections: demographics, enterprise
architecture practices, understanding of architecture elements, and their usage. We asked whether
the expert and his organization use English or Czech or a combination of both when speaking about
and using EA elements. The answers could inform about potential causes of ambiguity. In the
section about the architecture elements, we presented two motivation layer elements (stakeholder,
driver), three strategy level elements (course of action, capability), two business layer elements
(business service, business role) and two implementation layer elements (work package and
deliverable). This allowed us to learn about the perceived understanding of the elements. In the
practical usage section, respondents answered how they use various elements in different scenarios
within the public administration domain. This would allow us to compare the answers with the
previous section and measure the gaps between perceived semantics and actual use.</p>
      <p>In order to demonstrate how enterprise architecture elements can assist with closing the gap in
understanding between various experts, we proposed a high-level enterprise architecture view for
the State Land Office of the Czech Republic, a recently established Czech government agency. The</p>
    </sec>
    <sec id="sec-5">
      <title>4. Results</title>
      <p>first author of this paper has been involved as a consultant in the agency EA development project.
The enterprise architecture serves as a common communication framework for experts with the aim
to support achieving both operational and developmental goals of the agency.</p>
      <p>The results presented in this section are preliminary as the survey is still ongoing. The respondent
demographics is described in Table 1.</p>
      <p>While being ESL speakers, more than 55 % of the respondents used combination of both Czech
and English and 45 % used only Czech or English in the EA communication within the team.
However, the communication with the client occurred in all three ways equally by one third. For
each element, four possible answers were offered; and we measured the frequency of equal answers.
The respondents in majority understood the elements meaning equally except for the Course of
Action where answers altered between 'direction' (5), 'approach' (2) and 'procedure' (2) (Table 2).</p>
      <p>While the respondents were quite confident in indicating the EA elements meaning, the answers
about the practical usage of the elements were more diverse. The Resource and Goal elements usage
were presented as multiple-choice items, and the other three (Business Service, Business Role and
Deliverable) were described as illustrative cases with options to agree or disagree. The most uniform
answers were given to the Goal element which all respondents marked as 'goal' (9) but also as
'ambition' (4) or 'path' (1). The least uniformity recorded answers to the item Deliverable where only
two respondents answered about the meaning and five agreed upon its use in a practical example.
Business Service and Business Role were understood and used almost equally with the least
discrepancies. On the contrary, the use of the Resource element was perceived as 'workers' (7), or
'machines or vehicles' (7), or 'energy' (4) (Table 2).</p>
      <p>Business Service</p>
      <p>Business Role
Implementation</p>
      <p>Work package
Deliverable
8
8
8
5
8
6
6
2
9
7
7
6
5
4.1.</p>
      <sec id="sec-5-1">
        <title>A Case Study of the State Land Office of the Czech Republic</title>
        <p>
          The State Land Office of the Czech Republic has been established by the Act on the State Land Office
in 2013 in order to administer state property. The agency has replaced former land authorities and
has been given competences to manage and transfer agricultural land, consolidate land, settle
restitution and property claims, and privatize state property
          <xref ref-type="bibr" rid="ref7">(SLO, 2019)</xref>
          . The management of the
Office decided to actively apply the enterprise architecture principles in order to fulfill the long-term
e-Government initiative promoted by a nation-wide Strategy of Coordinated and Complex
Digitization of the Czech Republic 2018+
          <xref ref-type="bibr" rid="ref4">(Ministry of Interior, 2018)</xref>
          .
        </p>
        <p>The objective of the agency is to improve efficiency of the ICT management both for meeting the
operational and developmental objectives. This should be done by creating the enterprise
architecture that would allow to continuously analyze the current state, identify and close the gaps
by implementing projects. To meet the set objective, the agency launched the enterprise architecture
initiative and invited consultants to create a high-level schema describing objectives, capabilities and
ambitions (Figure 1). The first author of this paper created the schema of the agency EA initiative.</p>
        <p>The State Land Office strategic goal is to support the ambitions of e-Government and the Digital
Strategy of the Czech Republic and, at the same time, to make its own ICT efficient and accessible.
This goal breaks down to a number of partial goals: effective management of ICT operation, effective
management of ICT development, and compliance with the Digital Strategy of Czech Republic. Each
of the partial goals are assessed from the current state and future state viewpoints. It means that a
continual and repeatable evaluation of the current state of ICT operations at the Department of
Administration and ICT infrastructure and development at the Department of project management
and internal development must be ongoing in order to effectively innovate ICT operations and
improve ICT developments. The gaps between the current and future state should be discussed,
consulted and mitigated through meetings at all managerial levels, i.e. board meetings, EA board
meetings, project steering committee meetings, team meetings and also subject matter expert
meetings. SPARX is an enterprise architecture tool adopted by the agency. The enterprise
architecture framework depicted on Figure 1 proposes a single enterprise architecture
communication platform reflecting all strategic aspects and requirements enforced by the Digital
Strategy of Czech Republic and also has potential to reduce the misunderstandings about the
nationwide e-Government goals as well as the ambiguity of perception of enterprise architecture elements.</p>
      </sec>
    </sec>
    <sec id="sec-6">
      <title>5. Discussion</title>
      <p>Although results of the survey are not yet conclusive, we can observe that the motivation and
strategic layer that describe "the why" of an organization are less ambiguous than the business layers
that represent the semantic aspect of the enterprise architecture. Particularly, the implementation
layer seems to be challenging both for the meaning perception and the usage of the elements. The
ambiguity of elements meaning across EA layers has been reported in a number of EA projects and
initiatives particularly in the public sector (Banaeianjahromi &amp; Smolander, 2019). One of the factors
causing the ambiguity is the mixed usage of English and another language versions of EA standards
in communication between EA and other experts, ESL speakers. Apart from the lost meaning due to
translations, the misperceptions of elements may be a root cause of the mixed use of various
elements which further hampers smooth and flawless communication (Gustas &amp; Gustiené, 2004).
Despite being non-conclusive and unfinished, our survey findings were similar.</p>
      <p>
        We argue that although the holistic approach of the ArchiMate language was used in the
proposed EA, the target audience perceived the ArchiMate elements and their graphical symbols
differently which would result in further problems in the enterprise architecture management and
consequent projects
        <xref ref-type="bibr" rid="ref1 ref6">(Hiekkanen et al., 2013; Perez-Castillo et al., 2019)</xref>
        . Each element might have at
least more than one essential interpretation in the real world, while each person might use a
particular element in a different context which adds to the ambiguity. By gathering more empirical
evidence both from Czech and other language speakers, we would be able to prove or disapprove
whether the EA elements can clearly contribute to less ambiguity, better communication and
efficient ICT management in public organizations in ESL environments. Consistent with other
research , we argue that instead of assuming that ArchiMate is a language that
allows to communicate the enterprise architecture in an unambiguous way, it should be rather used
to unify the meanings of elements and their projections onto real world objects.
      </p>
    </sec>
    <sec id="sec-7">
      <title>6. Conclusion</title>
      <p>The views of the various stakeholders involved in enterprise architecture creation in public
administration are different due to language barriers, diverse domain knowledge and EA
experience. Therefore, EA modeling language such as ArchiMate is often used in contradiction to its
original purpose to express, analyze and model structures that are important for an organization.
This ambiguity can be mitigated by reducing the freedom of expression of the language and
providing use cases of ArchiMate elements application in e-Government projects.</p>
      <p>By completing the survey and result analysis, we shall be able to make several contributions.
First, we found that using a standard created in English speaking environment impedes
understanding among ESL audience. This is a practical implication for all stakeholders working on
an EA project especially in public sector. Due to a small sample consisting of Czech speakers only
the findings cannot be yet generalized to all ESL speakers. Second, as ArchiMate has become the
most used EA modelling language with many types of elements it has also broadened the gap
between their meaning and usage. This in practice will require to devise a minimum viable set of
elements and examples of their use that could be used in the given domain for EA projects in order
to reduce ambiguity. As the research on ambiguity in EA projects due to language barriers is scarce,
we aim to gather a large sample of EA professionals both from public and private sector and
continue are investigation.
Banaeianjahromi, N., &amp; Smolander, K. (2019). Lack of Communication and Collaboration in Enterprise
Architecture Development. Information Systems Frontiers, 21(4), 877 908.
https://doi.org/10.1007/s10796-017-9779-6</p>
      <p>0). Organization Modeling and Simulation Using BORM
Approach BT - Enterprise and Organizational Modeling and Simulation. In J. Barjis (Ed.) (pp. 27 40).</p>
      <p>Berlin, Heidelberg: Springer Berlin Heidelberg.</p>
      <p>Chiprianov, V., Kermarrec, Y., Rouvrais, S., &amp; Simonin, J. (2014). Extending enterprise architecture modeling
languages for domain specificity and collaboration: application to telecommunication service design.
Software and Systems Modeling, 13(3), 963 974.
https://doi.org/http://dx.doi.org/10.1007/s10270012-0298-0
Gustas, R., &amp; Gustiené, P. (2004). Towards the Enterprise Engineering Approach for Information System
Modelling Across Organisational and Technical Boundaries. Enterprise Information Systems V.</p>
      <p>Dordrecht: Springer Netherlands. https://doi.org/10.1007/1-4020-2673-0_24</p>
      <sec id="sec-7-1">
        <title>About the Authors</title>
        <p>Marti
As an Adjunct Professor, he is teaching Enterprise Architecture at the Faculty of Economics and
Management, Czech University of Life Sciences Prague. His research focuses on information management in
public administration, enterprise architecture and business processes.</p>
        <p>Czech
University of Life Sciences. His research topics are e-governance, information systems ethics, and big data
in decision making.</p>
      </sec>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          <string-name>
            <surname>Hiekkanen</surname>
            ,
            <given-names>K.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Korhonen</surname>
            ,
            <given-names>J. J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Collin</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Patricio</surname>
            ,
            <given-names>E.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Helenius</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Mykkänen</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          (
          <year>2013</year>
          ).
          <article-title>Perceptions on EA Use -- An Empirical Study</article-title>
          .
          <source>In 2013 IEEE 15th Conference on Business Informatics</source>
          (pp.
          <fpage>292</fpage>
          <lpage>297</lpage>
          ). https://doi.org/10.1109/CBI.
          <year>2013</year>
          .48
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          <string-name>
            <surname>Kotusev</surname>
            ,
            <given-names>S.</given-names>
          </string-name>
          (
          <year>2017</year>
          ).
          <article-title>Conceptual model of enterprise architecture management</article-title>
          .
          <source>International Journal of Cooperative Information Systems</source>
          ,
          <volume>26</volume>
          (
          <issue>03</issue>
          ),
          <fpage>1</fpage>
          <lpage>36</lpage>
          . https://doi.org/10.1142/S0218843017300017
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          <string-name>
            <surname>Lankhorst</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          (
          <year>2017</year>
          ). Enterprise Architecture at Work: Modelling, Communication and
          <string-name>
            <surname>Analysis (The Enterprise Engineering Series). (J. L. G. Dietz</surname>
          </string-name>
          , E. Proper, &amp; J. Tribolet, Eds.)
          <article-title>(Fourth Edi)</article-title>
          . Berlin: Springer-Verlag GmbH Germany.
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          <source>Ministry of Interior</source>
          . (
          <year>2018</year>
          ).
          <article-title>Government Council for the Information Society</article-title>
          .
          <source>Retrieved March</source>
          <volume>20</volume>
          ,
          <year>2020</year>
          , from https://www.mvcr.cz/clanek/rada-vlady
          <article-title>-pro-informacni-spolecnost.aspx?q=Y2hudW09Ng%3D%3D</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          <string-name>
            <given-names>Open</given-names>
            <surname>Group</surname>
          </string-name>
          . (
          <year>2019</year>
          ).
          <source>ArchiMate 3</source>
          .1 Specification: The Open Group Standard. Van Haren Publishing.
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          <string-name>
            <surname>Perez-Castillo</surname>
            ,
            <given-names>R.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Ruiz-Gonzalez</surname>
            ,
            <given-names>F.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Genero</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Piattini</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          (
          <year>2019</year>
          ).
          <article-title>A systematic mapping study on enterprise architecture mining</article-title>
          .
          <source>Enterprise Information Systems</source>
          . Taylor &amp; Francis. https://doi.org/10.1080/17517575.
          <year>2019</year>
          .1590859
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          <string-name>
            <surname>SLO.</surname>
          </string-name>
          (
          <year>2019</year>
          ).
          <source>State Land Office. Retrieved March</source>
          <volume>20</volume>
          ,
          <year>2020</year>
          , from https://www.spucr.cz
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          <string-name>
            <surname>Tamm</surname>
            ,
            <given-names>T.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Seddon</surname>
            ,
            <given-names>P. B.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Shanks</surname>
            ,
            <given-names>G.</given-names>
          </string-name>
          , &amp;
          <string-name>
            <surname>Reynolds</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          (
          <year>2011</year>
          ).
          <article-title>How does enterprise architecture add value to organisations? Communications of the Association for Information Systems</article-title>
          ,
          <volume>28</volume>
          (
          <issue>1</issue>
          ),
          <fpage>10</fpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>