Content uploaded by Christian Stary
Author content
All content in this area was uploaded by Christian Stary on Aug 25, 2021
Content may be subject to copyright.
Informing Work Interaction Design by 3rd
Generation Activity Theory
Jeannette Hemmecke1, Christian Stary2
1 Occupational Psychology Coaching, Hagenberg Austria
2 Business Informatics-Communications Engineering, University of Linz, Austria
Jeannette@Hemmecke.com, Christian.Stary@JKU.at
Abstract. In this paper, we reflect on stakeholder interaction design supported
by task behavior specifications and stakeholder benefits when framing design
with (i) eliciting task knowledge and understanding work activities and activity
systems, (ii) means of interactional representation for design, and (iii) capturing
the dynamics of activity systems. Thereby, complimentary inputs to theory
development and work interaction design techniques become evident. In
particular, eliciting implicit knowledge on human task accomplishment and
work processes helps understanding and representing activities as design-
relevant behavior entities. Eliciting implicit knowledge influences the
representation of work knowledge and the subsequent design process of socio-
technical systems. We elaborate on some methodological interventions for
creating stakeholder task behavior models, including patterns of information
exchange for collaborative task accomplishment.
Keywords: Activity Theory, workplace design, stakeholder interaction
1 Introduction
Today’s organizations are continuously changing. Resilience requirements have been
followed by dynamic capability development, underpinned by digital transformation
processes. All of them require organizational commitment, bringing upfront
integrated or at least adjusted design of work and IT systems to support human
workforce and organizational development. Thereby, an organization’s knowledge
acquisition capabilities in relation to work activities play a crucial role [1]. They
affect stakeholders as knowledge holders and (re-)designers of socio-technical
systems. Adopting the notion of ‘knowledge as knowing’ and activity theory (AT) as
framework help eliciting knowledge of organizations and their people, and finally
conducting business process improvements (ibid.). In particular, AT is of benefit
when engineers design artifacts [2]. In this paper, we provide methodological input by
addressing the interface between work knowledge acquisition and the design of task-
based and stakeholder-oriented socio-technical systems.
While modeling-based designs of various kind should facilitate knowledge
elicitation and transforming it to technological artifacts, we argue for tackling implicit
work knowledge. Framing this social process by AT facilitates acquiring task- and
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
100
stakeholder-relevant knowledge for further modeling and design processes.
Clemmensen et al. [3] have shown in an in-depth analysis of AT contributions in the
field of Human-Computer Interaction over the last 20 years that AT is used
successfully in various ways ranging from developing new analytical tools,
facilitating empirical analysis to informing the design of concrete interactive systems
like learning environments, groupware systems or knowledge management support
systems. In this paper we will investigate the contribution of AT concepts for
developing a methodology to acquire knowledge within interaction design in digitally
supported work processes, both in terms of relevant behavior, and task-based system
development support. The relevant AT concepts range from object-directedness of
activities and personal sense-making of knowledge to re-organize tasks according to
their context.
We build upon seminal work in this context as provided by Bonnie Nardi and
Victor Kaptelinin [4-8]. We follow the understanding that people act with technology,
which is always designed and utilized in the context of people (cf. [9], [10]). We also
follow the need of understanding people’s desires and intentions (ibid.) that drive
their behavior. This is why we argue for eliciting implicit knowledge, as, according to
our experience [11,12], it contributes to deeper understand the needs and capabilities
of users in interactive system design. This information needs to be captured in a
structured while systemic way, e.g., by connecting task behavior models.
In section 2 we recapture organizations in terms of activity systems (in the sense
of AT) and conclude for eliciting task-relevant stakeholder knowledge in a situated
and context-based way, recognizing the flow of knowledge in terms of change
(creativity) and mutual exchange of information and objects of work. In Section 3 we
propose methodological interventions according to activity-oriented knowledge
acquisition (as presented in section 2), in order to facilitate activity-(system) based
interaction design. Its implementation enables dynamic changes and prototyping.
Section 4 concludes the paper, wrapping up the presented findings and sketching
future research activities.
For illustration purposes, we will use a scenario stemming from current healthcare
developments throughout the paper. It is a domain heavily debated in socio-technical
research and development communities (cf. [13]). We focus on home healthcare, as it
requires cooperation among several stakeholders that needs to be rearranged due to
remote operations and novel digital communication capabilities, such as edge
computing (cf. [14]) or IoT [15]. Consider a patient at home coordinating several
digital healthcare support systems ensuring individual wellbeing in a self-organized
way while being connected on demand to medical experts and system vendors of
his/her medical equipment. It is a scenario re-occurring in transformations to digitized
environments. Such types of scenarios are driven by customization activities to
individualize task activity chains, e.g., by means of apps, and include user control of
devices and work processes novel for concerned users.
Setting up such systems does not only require architectures that allow networking
or composing systems in a modular while in an effective and efficient way [16] but
rather conceptual understanding of adjusting features according to specific
stakeholder roles and their task-related behavior [14]. For instance, home healthcare
users or patients have become users in different roles. Once a medical device, such as
an instrument for measuring the blood pressure, is delivered from a provider or
101
vendor, users start in the role ‘configurator’, continue as ‘operator’ when measuring
their blood pressure, and act in a third role when maintaining the device. In addition
to configuration, operation, and maintenance, users are confronted with networking
capabilities of digital devices, enabling coordination with other digital systems in
their (home) environment, communication with peers and medical care takers, and
adaptation to fine-tune a device’s use. This complexity requires not only several
technical skills, but also awareness and transparency, that using medical devices in
home healthcare does not only affect the patients themselves, but also other
stakeholder behaviors, such as medical caretakers and infrastructure providers.
Designing such socio-technical systems can be structured by task-based stakeholder
orientation including corresponding service abstractions [17,18].
We consider a home healthcare scenario with several support tasks (for an
overview see also the grey boxes in Figure 4). For the patient, a Personal Scheduler
coordinates all home healthcare tasks. Personal Scheduler denotes a behavior
abstraction of a task in the sense that it is not specified, whether it is performed by
humans or digital devices or services. Since design abstracts from actual
implementation, such an abstraction facilitates task-specific specifications. Due to the
use of these task specifications throughout the paper we label them with capital
letters.
The Personal Scheduler handles all patient tasks to be set in a certain period of
time, and can be set public to other concerned stakeholders, such as relatives, medical
experts, and social services. A Medication Handler takes care of providing the correct
medication at any time and location and is collaboratively controlled by medical care
takers and the patient. Healthcare instruments support specific services, e.g., Blood
Pressure Measurement sensing the medical condition of the patient. They are linked to
their provider companies or producers, ensuring proper operation. In addition, a
Shopping Collector serves as container for all items to be provided for medication and
wellbeing. These sets of home healthcare tasks include the mutual interaction of
various stakeholders and their interactions with digital systems, such as apps on
smartphones. Since they need to follow a certain domain logic, e.g., identifying the
demands from existing medications when making a shopping list, workflows can be
defined when digitizing home healthcare settings.
2 Activity-theory-based Acquisition of Work Knowledge
Acquiring work knowledge is a complex endeavor because of the multiple facets of
and different perspectives on one and the same work process, the interrelatedness of
subjective, objective as well as organizational factors, plus the – often tacit –
knowledge that is needed to engage in certain processes and achieve certain
outcomes. Therefore, the acquisition and representation of work knowledge has to
take into account the heterogeneous elements, relations and processes of that
knowledge as well as to understand that the acquisition process itself becomes part of
the whole endeavor. Thereby, the context-sensitive elicitation of stakeholder-specific
knowledge is of immanent importance, as various mental models, either consciously
or unconsciously, shape design and thus, user acceptance (cf. [19]). Referring to our
home healthcare scenario, the mental models of how a patient envisions medical care
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
102
at home might significantly differ from how a home health nurse conceives it. Both
will probably be different from the understanding of medical experts working in
hospitals.
What makes up the multiple facets of human work knowledge elicitation?
Individual stakeholders handle work tasks in their own specific way, they do not only
pursue organizational goals, but individual ones as well. They act according to their
individual mental models, knowledge, skills, self-restraints, personal needs and
intentions, etc. In today's knowledge society, in many cases, teams instead of single
individuals work on a task together in order to bundle their competences. By that, the
complexity of acquiring work knowledge is increasing. It is not only the knowledge
of single actors that must be elicited but group processes, interpersonal relationships,
patterns of communication, collective mental models etc. have to be taken into
account. In the addressed home healthcare scenarios at least the following groups are
involved: healthcare specialists, patients, service and product providers from various
fields, including home, delivery, medical equipment service.
Certainly, acting on a work task is not only influenced by its actors (individual or
team), but by the requirements, drawbacks and potential of other stakeholders like
those working on subsequent or precedent tasks. Moreover, each work task is
essentially formed by the organization or the organizational setting itself with its
culture, structures, processes, technical tools and physical surroundings. Work tasks
are embedded in a broader organizational context, they (should) contribute to the
organizational goal. Last but not least, the content/object of the task itself is
contributing essentially to what a work task is made up of.
A task produces a certain outcome that is the starting point of or a means in some
other tasks in the work process. Moreover, the task is transforming some sort of
objects (material or immaterial) and, by that, it is dependent on the characteristics of
the object itself as well as of the domain the task belongs to. Such a situation is given
in the scenario of home healthcare, when some health indicators, such as the blood
pressure, require further consultations of further stakeholders, for instance when the
patient’s measurement of blood pressure indicates a value above average. For deeper
analysis, the patient contacts medical experts. From the point of view of medical
experts like nurses or physicians the object is the physical condition of a patient
measured through indicators like blood pressure, body temperature or blood sugar
level aiming at a healthy functioning of the body. From the point of view of the
patient the outcome of a healthy body might be much more dependent on the feeling
of personal wellbeing, subjective performance capability, and absence of pain.
However, patient and healthcare workers cooperate on the shared object of the
patient’s physical condition, apparently having different perspectives on the same
object, but sharing the same data (e.g., the measured blood pressure value).
The more expertise is needed in order to perform a work task, which is typically
the case in complex and unforeseeable situations, the more the tacit knowledge plays
a crucial role [20]; for an empirical investigation [cf. 21]. Consider the behavior of a
patient not being used to digital system support. He/she is likely to have developed a
certain behavior pattern of acquiring medical advice, e.g., visiting the local doctor. It
is also very likely, the activity of visiting a local doctor is triggered by a set of tacit
knowledge, e.g., feeling poorly, strange differences in blood measurement results (the
patient might already use medical technological tools but to the point not digitally
103
connected). This existing knowledge including the patterns of behavior needs to be
elicited in case of transforming to digital healthcare support. The same is true for the
tacit knowledge of the local doctor: The doctors knows the patient in context: his/her
medical history, his/her needs of re-assuring, his/her worries concerning side-effects
of pharmaceuticals, the family situation etc. All this work expertise and tacit
knowledge are a challenge to elicit (ibid.).
Recognizing shifts of organizational settings, as it happens in healthcare from
medical environments to home settings, makes it even more reasonable to elicit
implicit knowledge. At least the following issues need to be tackled in that context.
First, it needs to be acquired, whether a home infrastructure allows locating and
operating medical equipment in a way that a hospital or clinic does. Secondly, the
delegation of roles is evident: the patient as a user of medical devices at home
becomes an essential part of medical workflows, namely collecting fundamental data
for further processing. In order to design work tasks given such a change of settings
requires activating of tacit knowledge, as this shift can be considered as an
organizational learning step (cf. [22-24]) embodying the patient in novel roles
(cf.[13]).
Recognizing such multi-faceted complexity we suggest framing task knowledge
acquisition by AT. AT enables us to make graspable the complex interrelations in
which work knowledge is embedded. The interrelations range from those between
different actors, teams and the organization as an institution with its structures,
functional roles and processes to objects, mediating tools, and the socio-cultural
context. AT meets the challenge of comprising the complexity of organizational work
as well as to regard the dynamic nature of knowledge oscillating from the conscious
to the unconscious level according to its respective context. Moreover, AT allows us
to consider the elicitation process as part of the context and the future technology as
changing (mediating) the activity itself [25]. Beyond that, we back Karanasios’ [25]
argument that the design of information systems informed by AT should contribute to
subjects’ emancipation instead of control, restriction or prediction.
2.1 Organizations as Activity Systems
Organizations can be viewed under the “lens” of AT [25-27]. AT itself was originally
formulated as an encompassing theory of psychology [28], with roots in Vygotsky's
cultural-historical psychology [29] and emphasis on the crucial role of language and
signs in mediation, and with underpinnings in the German philosophy of the 18th and
19th century, above all deriving Marx' concept of activity as crucial entity that
mediates between subject (human) and object (the German “Gegenstand”). Today's
AT has grown into a meta-theory that has applications in many different fields, above
all in Human-Computer Interaction e.g., [3,5,9,30-33], knowledge management
[24,34-39], organizational learning and research [26 40-44].
A special characteristic of AT that makes it so powerful for the purpose of
knowledge elicitation is that it takes into account the dynamics of knowledge [6] as
well as its immanent connectedness to its personal, social, cultural as well as physical
context [34] or like Nardi ([6], p.76) puts it: “the activity itself is the context.” Within
the frame of AT, the unit of analysis for (work) knowledge is neither a person nor a
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
104
small snippet of behavior like in other psychological theories, but the activity (in the
sense of an activity system) itself. Activity can be seen as an integral unit of
consciousness and behavior [28] enabling to grasp the complex interrelations of
actions, actors, institutions, tools in use, social rules, products and the society within a
certain context, in a systematic way.
The concept of activity enables us to dissolve the artificial separation of body and
mind because activity is the joining element between subjective and objective
processes (cf. Figure 1). “In activity there does take place a transfer of an object into
its subjective form, into an image; also in activity a transfer of activity into its
objective results, into its products, is brought about.” ([28], p. 50). Thus, the process
of knowledge elicitation has to consider subjective views, meaning and motives as
well as objective conditions, means and products – in its reciprocity. Moreover, AT
helps to understand that each knowledge elicitation process becomes part of the
context wherein it is applied and thereby might change the activity itself. Hence, for
home health care system design we not only need to incorporate the medical expert’s
cause-and-effects-knowledge in the context of the respective illness and additional
parameters (age, sex, other health indicators etc.), but also the patient’s values about
what is important to him/her (simply feeling well or being able to practice a serious
sport). Moreover, context data is needed to interpret every single measurement result
(e.g., room temperature, medication, activity level during measurement process, day
time, former results etc.).
Fig 1. Activity as mediating unit between subject and object
In line with the above mentioned shift of organizational context, a patient having to
identify the next step in home healthcare, e.g., after measuring his/her blood pressure,
has different background and knowledge than a medical expert. For instance, guiding
a patient to find out whether to consult medical experts due to the result of current
measurements, requires understanding of the individual situation. It becomes even
more essential when scheduling medication and ordering further drugs by putting
them on the shopping list. Thereby, the course of previous activities and the current
wellbeing of the patient are likely to be additional context parameters.
For the analysis of an activity, the general internal structure of activity helps to
understand the dynamics within an activity system when the context changes (cf.
Figure 2): Each activity, motivated by an object that meets an actor's need, consists of
actions directed toward concrete goals and of operations, i.e., by what means the goal
is reached under specific circumstances. The activity holds the actions and operations
together in the sense that it gives them its meaning in the respective context [28]. For
105
instance, the further shopping of the prescribed drugs depends on whether the state of
wellbeing and health could be reached in the past by taking those drugs. Therefore
current subjective and objective data on the state of wellbeing and health of the
patient need to be incorporated and decided upon when scheduling the next
medication and shopping trip.
The dynamics within an activity can be described as follows: An activity might
convert into an action by losing its original motive (the motive turns into a goal). In
return, an action might become an activity through transformation of a goal into a
motive. Furthermore, an action can become an operation through regular practice, or,
if conditions change, an operation might (re)turn into an action. Operations can be
crystallized in form of external means like software programs. However, if the
conditions change, the operational content of the action changes too [28]. Means
should always be considered in relation to their goal, operations in relation to their
action. In our home healthcare scenarios, a typical goal is to identify the daily status
of wellbeing based on the current blood pressure. It is motivated by keeping a certain
level of wellbeing which leads to activating a certain app and the corresponding
device. Measuring the blood pressure is one of the required actions that leads to
operating a corresponding instrument at a certain point in time, e.g., the morning after
getting up.
Fig 2. General internal structure of an activity with its three main components activity, actions
and operations
Although “[h]uman activity does not exist except in the form of action or a chain
of actions” ([28], p. 64), it is the activity and its initiating motive that is responsible
for personal sense making. This implies that all observed or recorded behavior can
only be understood within the context of the respective activity of which it is part.
The same behavior can contribute to different activities and, by that, can have a
completely different meaning. How an activity is accomplished depends upon the
why it is accomplished. As a person facilitating knowledge elicitation, one has to ask
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
106
why a person (or a team) is engaged in what s/he does (they do) in order to be aware
of the context in which a certain task is accomplished in this or that way.
However, AT does not formulate hierarchies of needs or motives but takes into
account that (new) needs and motives can come into existence through (new)
activities [28]. In that context, Leont'ev [28] emphasizes the difference between
objective and subjective significance: Whereas objective significance is the objective
meaning of certain objects, like examination marks developing in a society over time,
subjective significance is the personal sense, it is the role that an object plays in an
individual's life. The objective meaning can only be understood through analyzing it
as a product of society and historical development. In contrast, the personal sense
reveals only by figuring out the role that an object plays in a person's life, by
regarding this person's needs and motives. Personal sense is therefore bound to
emotions, personal history, values and aspirations, i.e. a system of individual
constructs [45]. Still, the personal sense is not independent from the objective
meaning and by that not independent from societal history. In that sense, knowledge
can be seen as “… representations exist inseparable from the subject’s activity, and
they fill it with the riches accumulated in them and make it alive and creative...” ([28],
p. 76). Though, a holistic knowledge elicitation process can be done through an
analysis of activity and its internal structure and dynamics.
Looking from a task perspective, e.g., scheduling medication, supports focused
elicitation of task-specific behavior on all levels of activities. An activity represents a
unit of work with a certain objective, e.g., ensuring wellbeing every day. It requires a
set of actions, such as the patient configuring a digital blood pressure measurement
system and operating it for measuring blood pressure every day. For elicitation, all
required actions, including those not having been explicit so far need to be part of
contextual inquiry. In the addressed case, the shift of context corresponds to assigning
new technical roles to users related to crucial tasks, namely collecting reliable data for
further processing. Eliciting this information could lead to some design effort, e.g.,
configuration support for a blood pressure device through a wizard app linked to
second level customer support of the device vendor.
When we started to apply ideas from AT to knowledge elicitation it soon became
evident that AT is not one uniform theory, but a slightly diverse pool of concepts and
models which are more or less related to Vygotsky's cultural-historical school [29]
and/or Leont'ev's [28] theory of activity. Central notions are not always used with the
same meaning (e.g., the terms “motive”, “need”, and “object”), but are interpreted in
different ways. This is especially true when different concepts are mingled without
clearly stating their definition. Already Leont'ev himself observed this:
“I have had occasion to encounter, distressingly often, the expression
activity approach and other terms about activity, not always in a sufficiently
distinct and defined meaning, and situated somewhere in a broad space of
meaning and concepts. [...] Now when I see the phrase 'from the point of
view of the activity approach', I must state, sincerely, it disturbs me” ([46],
p. 31).
We therefore went back to the historical roots of AT, above all to Leont'ev's
“Activity, consciousness, and personality” ([28]; comparing it with the Russian
original “Deyatelnost, soznanie, lichnost”, 1975, as well as its German translation
“Tätigkeit, Bewußtsein und Persönlichkeit”, 1982 [47]), to Vygotsky's concept of
107
mediation [29] and the philosophical influences from the German dialectical
materialism according to Marx and Engels.
Then, we searched for AT approaches of the second and third generation2 choosing
only those for analysis that described their reception in sufficient detail and were
related to the field of knowledge theory or knowledge generation in organizations.
Therefore, in addition to 1) Leont'ev's theory of activity [28], the following
approaches were included in our analysis: 2) Engeström's theory of expansive
learning [42,43], 3) Raeithel's activity theoretical design theory [32,48], 4) the
occupational psychological cooperation model of Wehner et al. [49], 5) Clases' model
of cooperative knowledge production [35,36], 6) the model of situatedness of
knowledge sharing of Boer, van Baalen and Kumar [34], 7) Hasan and Gould's
cultural-historical activity theory (CHAT [37,50], and 8) Blackler's theory of
organizations as activity systems [26]. We compared them systematically to find the
differences as well as the common ground. These steps enabled us to finally develop
an activity-theoretical understanding of knowledge and knowledge elicitation (cf.
[51]).
It has become evident, that Engeström's [42,43] and Raeithel's [32] work can be
seen as second generation approaches of AT that ground deeply in the original work
of Leont'ev [28] and in Vygotsky's concept of mediation [29]. However, they go
beyond by incorporating other theories (above all systemic and constructivist
concepts of Maturana, Varela, Luhmann, Mead, and others) and broaden the horizon
of AT from focusing on single subject activity systems to collective subject activity
systems. Whereas Leont'ev himself was interested in individuals' activities in order to
understand their development in the context of their biographies and the socio-
historical circumstances, especially Engeström calls for teams as collective subjects of
activities in order to understand the development of new forms of activity.
Apparently, also Leont'ev himself emphasized that (work) activity is always a
social process because of the historically grown division of labor, the mediating
means (language above all) and the influence of the society on each activity [52].
Both Raeithel and Engeström put a stronger focus than Leont'ev on the interaction
between activity systems in order to understand each activity system in interrelation
with other activity systems and to draw conclusions for design (Engeström designed
learning, whereas Raeithel designed software). However, both came up with different
words for quite similar ideas. The major connection of interacting activity systems
that Raeithel [32], later Wehner at al. [49], Clases & Wehner [36], suggests has the
form of a process: The result (transformed object into an outcome) of one activity is
becoming the object motivating another subject's activity. Engeström [42,43] himself
emphasizes a shared object as major connection between two or more interacting
activity systems. The idea of multivoicedness of activity systems emerges, i.e.,
different view on the same object are the drivers of conflict as well as innovation [43].
The interrelation between single actors and whole groups or organizations becomes
essential in the second generation approaches. Therefore, they come to the conclusion
that there are two main ingredients for the development: reflection (not only
2We term Leont'ev's work the first generation. This is not in accordance with Engeström's
classification of first, second and third generation [42], because for our purpose, it makes sense
to group Engeström's and Raeithel's approach into a category of its own as we will show.#
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
108
individual but also collective) and communication/cooperation. Development means
new forms of activity are coming into being; this is called innovation or knowledge
generation. In order to grasp work activities in organizations the process of
knowledge elicitation has to focus on both, on the needs, capabilities and activities of
single individuals, and on the role of teams, collective activities, as well as the
interrelationship between individual and collective activities.
The five remaining approaches (4 to 8) are mainly influenced by either Raeithel or
Engeström, but seem not to take any additional concepts from Leont'ev or Vygotsky.
We call those approaches third generation approaches, because, again, they
incorporate other theories (the works of Weick, Suchman, Lave & Wenger and others
are repeatedly merged into those AT approaches) or accentuate certain aspects like
the interaction between activity systems on different context levels like Boer et al.
[34] do.
When we describe here what all these approaches have in common (in accordance
with the philosophical roots and the original works of Leont'ev and Vygotsky), we
build a picture of constituents and processes that are immanent in an activity
theoretical understanding of the world:
1) Activity is the mediated relationship between a subject and an object; there is
no direct/unmediated relationship between subject and object.
2) Mediation takes place through societal forms like social and organizational
structure or formal and informal rules as well as through operative forms like
artifacts, tools, physical as well as symbolic means and media.
3) Activity is seen as the smallest unit of analysis that is needed to understand
individual, group, and organizational behavior; only on the level of activity
personal sense making can be grasped.
4) Context is important, mostly the activity itself with its internal structure and
dynamics is seen as the context. When groups or organizations are
considered, the interrelationship between several interacting activity systems
as well as the multiple views of different stakeholders (multi-voicedness) are
crucial for understanding the more complex knowledge processes within
organizations.
5) The history (and in some approaches the anticipation of future) of an activity
system has to be taken into account in analysis (and design) in order to
understand present behavior and draw possible future paths; that can be seen
as the timely interdependency of former, present and future activity systems
as Boer et al. [34] put it.
6) Every activity is directed towards an object (except of Blackler's approach
where activity is directed towards organizational routines meaning people act
because they always acted like this); the term “object” is mainly used in a
broad sense as a material or mental object that is aimed to be transformed
somehow through the activity and the belonging actions and operations.
7) All approaches emphasize the dynamics of activity: the role of
conditions/context for possible transformations of activity – action –
operation etc. as well as the future possibilities of new forms of activities
with new motives coming into existence.
8) Contradictions, dilemmas, crises, and conflicts within an activity system and
between activity systems are seen as the motor of development. Therefore,
109
reflective, communicative and cooperative processes are especially focused
on and fostered.
Figure 3 sketches the general structure of an activity system in interaction with
other activity systems. An activity system consists of the following main
components: a) an actor (be it an individual, a team or a whole organization,
depending on the level of abstraction), b) an object that serves as a motive for the
actor in question, can be described as follows, c) the mediating means (e.g., tools,
symbols, language, a knowledge base etc.), d) the community (e.g., co-workers), e)
explicit and implicit rules that mediate the relationship between the actor and the
community, f) the organizational and role structure (in other terms: the division of
labor).
Fig 3. Structure of an activity system, embodied in its context
Figure 3 illustrates three exemplary types of interaction between activity systems:
1) Interaction between activity systems can have the form of a process, i.e., the object
transformed by the activity of actor A serves a motive for actor B who engages in
another activity transforming the object into an outcome. 2) Interacting activity
systems can be bound together via an object-means relationship, i.e., the result of the
activity of actor D might become part of the means in actor A's activity; alternatively,
actor E's outcome might serve as new set of rules for communicating and cooperating
with the members of the community (object-rules relationship). 3) More than one
activity system is working on the same object (actor A and actor C) and interact via
the transforming the same object. They might have different views on the same object
and might be grounded in different group and professional cultures (rules) in their
respective communities.
With respect to the home healthcare scenario, activity systems could capture the
patient’s home setting and a medical care taking institutions, such as a healthcare
center for cardiological risk patients:
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
110
• Actor A could be a patient knowledgeable about running his/her home
healthcare (explicit/implicit rules in Fig. 3), and motivated to ensure a high
level of his/her wellbeing. The patient is provided with knowledge and
equipment to achieve this objective (Means in Fig. 3). His/her activities
concern his/her wellbeing (Object in Fig. 3). Being a certain type of patient,
e.g., assigned to a cardiological risk group (referring to Community in Fig.
3), he/she needs to follow a certain procedure regularly (referring to
Organizational/role structure in Fig. 3), including operating his/her blood
pressure measurement device. The results of this procedure allows him/her to
decide whether further action needs to be taken or not (referring to Outcome
in Fig. 3) to ensure his/her wellbeing. This outcome represents the interface
to Actor B, in case further action needs to be taken.
• Actor B could be a medical care center knowledgeable about treating
cardiological risk patients including handling requests from home healthcare
patients (explicit/implicit rules in Fig. 3). Its medical staff is motivated to
provide high quality services to provide support and care about the wellbeing
of cardiological risk patients. The staff members are qualified and skilled to
achieve these objectives (Means in Fig. 3). Their activities concern patient
treatment and in-situ support (Object in Fig. 3). The care center is
specialized to cardiological risk patients (referring to Community in Fig. 3),
and the institutional help to be provided, ranging from advising patients to
emergency visits (referring to Organizational/role structure in Fig. 3). The
outcome of the center is high quality provision of services to patients leading
to a minimal set of cases that cannot be handled by its staff (referring to
Outcome in Fig. 3). This outcome represents the interface to further actors,
e.g., for reporting to official health authorities.
The model of activity is useful for work knowledge elicitation because it is a model
of structural relations and content-free in its nature (cf. [53]). It shows relations
between elements that hold or are considered true for every (work) activity and can be
applied to every (work) domain (cf. [3]). Whereas knowledge elicitation is often done
either through the lens of a tool language (by which the tool sets the limits, but not
human experience), as for example in the case of traditional knowledge engineering
techniques, such as CommonKADS [54], or through the subjective lens of knowledge
elicitation facilitator (by which the mental model of the facilitating person sets the
limits, but not the people who are engaged in the work tasks) knowledge elicitation
with the help of AT is done through the lens of human experience. Each work
activity, regardless of the (knowledge) domain it belongs to, can be viewed under the
lens of this structure. The components and their relations enable to look in specific
directions, e.g., to collect information about the availability and usage of means, about
implicit and explicit rules and how they foster or impede communication and
cooperative learning, about organizational structure that is either enabling or
hindering the activity of a subject, etc. With the help of this content-free structural
model, the knowledge elicitation facilitator is able to elicit task and stakeholder
relevant knowledge for further modeling and design processes from the genuine
perspectives of the actors in the activity system(s) with only little biases through
domain dependent or personal (pre)conceptions.
111
Knowledge is seen as something dynamic being part of every activity. Knowledge
generation emerges through the interaction between subject and object that is by
activity. The dynamic character of knowledge can be explained by borrowing
Raeithel's metaphor of knowledge being both at the same time, the river as well as the
river bed. The river flows and forms the river bed, at the same time the river bed is
directing the river flow. Knowledge, especially tacit knowledge, is part of both and, in
that sense, changes its state (from liquid to solid and vice versa). This has major
implications for the process of knowledge elicitation: knowledge is never fix, it is
always changing and always dependent on the societal, organizational, personal and
objective circumstances – at the same time building and changing those circumstances
itself. Knowledge from the point of view of AT is therefore situated and contextual, it
is activity itself [26,37]. In that sense, knowledge is individual as well as collective. It
can be in a more fluid state (expressed in behavior or communication patterns) or in a
more solid state (when means, rules or structures are analyzed).
What are the consequences that follow from AT and its understanding of work
knowledge for the knowledge elicitation process and for designing software? In
accordance with Raeithel [32], Engeström [42,43] and Wehner et al. [49] and their
design suggestions we draw the following conclusions. The knowledge modeling and
software design should allow for a high amount of openness and flexibility. From the
point of view of AT, (new) software should enable the development of new forms of
activity by supporting individual and collective reflection, by encouraging individual
as well as collective experimentation in single situations to solve problems, by
facilitating multiple forms of communication between different stakeholders, and by
making visible contradictions, crises or conflicts as initial drivers of the development
of innovations.
The development of artifacts should contribute to learning. Of course, development
and learning are not only a matter of developing and applying artifacts like software
but also of nurturing an open organizational culture where people respect each other,
are encouraged to discuss and learn from mistakes, are animated to try out new
procedures, experiment with new tools and have enough freedom to decide how they
accomplish their tasks. AT can serve as a framework for thinking about
methodological approaches to knowledge elicitation in human work settings in order
to foster sound analysis and design.
2.2 Eliciting Human Work Knowledge
In the previous section we have shown the common ground of different approaches
and generations of AT. In this section we want to focus on what can be deduced from
this analysis for the elicitation of human work knowledge. We suggest twelve
overarching activity theoretical principles for knowledge elicitation (cf. [51]) based
on a conceptual analysis on the one hand, however, triggered by our experience
(teaching knowledge elicitation) but flawed by the practical facilitation process of
knowledge elicitation on the other hand. The conceptual analysis was done by
comparing not only the core notions of AT (e.g., activity, subject, object, context) and
their usage in different AT approaches, but we were looking especially for an AT
understanding of knowledge-related processes (e.g., what is knowledge in terms of
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
112
AT, where is it represented, how can the process of knowledge generation be
described in terms of AT). The twelve principles we came up with can be understood
as an activity theoretically informed framework for the process of knowledge
elicitation.
The benefit of such a theory-driven conceptual approach allows various
implementations from a method perspective which is of particular importance in
actual development settings. AT principles need to be operationalized, i.e. put into
practice by applying different methods, e.g., Critical Incident Technique or Repertory
Grids [55,56], depending on the situation, capabilities, and project constraints. Hence,
the AT framework provides constitutive elements for work knowledge elicitation
from stakeholders that can be addressed by specific elicitation methods in different
ways. From an AT point of view there is quite an openness towards methods from
different traditions (cf. [3]), as far as they are not prescribing but exploring.
The AT principles allow a context-sensitive knowledge elicitation of work
processes in organizations. They show how AT meets the challenges of today’s
knowledge society by enabling to take into account the dynamic, ever changing nature
of human work knowledge as well as the social, cultural-historical, and temporal
intertwining of work activities. Since AT presents us a domain unspecific inner
structure of activity systems using a rich vocabulary (cf. [3]), it allows us to model
work processes in a flexible as well as structured way without losing either the whole
picture or important details, considering both personal sense making as well as
interrelations and multiple perspectives, and moreover, being expandable in its nature
(AT sees itself as natural subject to change and development, cf. [9]).
In the following we elaborate on the twelve principles and their meaning for
eliciting human work knowledge. Since each principle is derived from findings
detailed in the previous sections, we do not replicate the respective references. The
meaning of the mentioned terms and concepts corresponds to those introduced above.
(1) The object-orientedness of activity: Work activity can only be understood
when analyzed according to both points of view: a) from within the view of
the actor (subject) – personal knowledge, personal sense making, tacit
personal needs – as well as b) from outside (object) – domain knowledge,
„objective” context, organizational needs. Only by understanding work
activity in its dynamic nature expressive (software) design is possible. Since
each activity is directed towards an object („Gegenstand“) which is
transformed through actions by one or more actors into an (organizational)
outcome, the search for the common object is crucial for detecting the
starting point of the elicitation process.
(2) The mediated character of activity: All activity is mediated through
knowledge, means (tools) as well as social rules, i.e., each activity is
changing according to the knowledge, means and rules that are applied
throughout the working process. Every new software – moreover every new
design, product or service – changes the original activity since applying a
new means starts a new mediation process that cannot be predicted. New
means can affect the activity system as a whole by possibly touching the
gentle balance of organizational structures, business processes, functional
roles, cooperation patterns, human needs, knowledge and sense making, as
well as technical tools and instruments. The elicitation process therefore does
113
not end with modeling an existing activity, but starts (again) with applying
(prototypical) new means.
(3) Activity systems as units of analysis: Since activity systems with their
universal inner structure and interrelations of element categories (cf. Figure 3
above) reproduce or come into life through conscious actions and
unconscious automated operations conducted by actors, the activity itself can
be understood as the background against which concrete actions, knowledge
or human/organizational needs have to be viewed as a Gestalt. Only by
putting the elicited work knowledge in relation to all other elements of one
or more interacting activity systems, sound design choices might be made.
(4) Contradictions as driving forces: Contradictions are integral, historically
grown tensions within and between activity systems which are seen as an
important source of development and learning. During the elicitation
process, contradictions might show as reported conflicts, problems or
difficulties, or as subtly observable uncertainty, confusion or ambivalence.
By focusing those tensions throughout the elicitation process, new practices
and possible design solutions might become evident. Since activity systems
are open systems they can incorporate new elements like a new tool. This
new tool should solve an existing contradiction and by that contributes to an
organizational learning step –, but might also induce new contradictions
itself. In that sense, learning is an organization’s as well as humans’ life-long
process.
(5) The dynamics of activity systems: Activity systems are dynamically
changing, its components transform into one another according to (inner and
outer) context changes: Actions might become operations through practice
and automation or, vice versa, operations might become conscious actions
through a change of environmental factors. Mediating rules might become
the object of an activity, as well as an outcome of an activity might become
the mediating tool for another activity. Every time a component changes into
another or something new is incorporated into an activity system, the
meaning of the activity system and its components might change. This has
consequences for the knowledge elicitation process: The elicitation process
itself becomes part of the context. The elicited knowledge is only
interpretable within the frame of reference of the respecting activity system.
Methods used for knowledge elicitation must be sensible to changes in
knowledge, as well as enable to make the changes graspable in order to draw
conclusions for software design.
(6) Interdependency in time (culture-historicality and future anticipation):
Activities change over time and inhere their own history and their own
possible future paths. For the elicitation of work knowledge, comparisons of
different states (of the activity system or its components) in time can help to
view similarities and differences and by that understand what the key
concepts or driving forces in an activity are. In order to draw conclusions for
software design, comparisons of the old and the new as well as the past, the
present and possible future options are useful for understanding possible
developmental options.
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
114
(7) Social character of activity: All activities are intrinsically social. That does
not only refer to visible cooperative activities such as team work, but to all
activities, even the apparently loneliest one like writing a book. Our thinking
is seen as existing only as a social activity because language, tools and other
means are only created in interaction between humans and then, one day,
become internalized as some sort of cognitive models. Thus, all work
knowledge can be seen as social in its nature. Metaphorically spoken, the
individual knowledge is the river that flows in a collective riverbed. The
riverbed makes certain paths more likely than others. Domain knowledge
(which is a sort of collective knowledge) is important to understand certain
individual ways of doing things. At the same time, the river (that is the
personal knowledge) shapes the riverbed itself. Individual and collective
knowledge have to be brought together and to be viewed in context of each
other.
(8) Multivoicedness of activity: Organizational activity systems encompass
multiple points of view of different stakeholders, i.e., the system is multi-
voiced. The multiple perspectives are both a source of conflicts and a source
of innovation and learning. The elicitation of work knowledge has to take
into account the multiple viewpoints of different stakeholders as well as their
interrelations. Elicitation methods that help to see things from different
perspectives (like circular questions do [57]) can help to bring diverse
perspectives together.
(9) Interdependency of different context levels: Every work activity in
organizations can be viewed as an activity system on a more abstract or a
more concrete context level. Every organization can be analyzed as one
(abstract) activity system or as many (concrete) systems like that of its teams
or individuals. The context levels influence each other. Depending of which
context level is focused, different things become evident. Moreover, for the
elicitation of relevant work knowledge it is not only necessary to pay
attention to the local activity system(s) of the organization, but to the broader
global activity system(s) of the relevant domain including the tools, concepts
etc.
(10) Personal sense making of knowledge: Knowledge is bound to a personal
sense. Therefore, knowledge elicitation should always start from an
individual activity system (motivated by a certain need) in order to
understand the knowledge in its personal sense making. Actions or
unconscious operations are part of the activity system and are surely part of
the knowledge, but do not suffice as units of analysis. The personal sense
making (only to be understood when an activity is the unit of analysis) is the
context of meaning of the (elicited) knowledge and important for the transfer
of the elicited knowledge to abstract models as well as to other people. Often
the most critical step in knowledge transfer is that new actors have to embed
the acquired new knowledge into their own sense making.
(11) Elicitation of tacit knowledge as a form of externalization of activities: The
elicitation of knowledge as a process can be viewed as a process of
externalization, i.e., our thinking, mental models, personal constructs etc.
shape all practical activities and the creation of material tools, structures and
115
objects. The counter process is that of internalization, i.e., how practical
activity and the usage of tools influences our thinking and the creation of
mental representations of things and activities. Since the philosophical roots
of AT is dialectical materialism meaning the practical activity is seen as
primary to thinking, the elicitation of work knowledge should connect to
concrete experience in order to evoke mental models.
(12) Organizational learning as development of new forms of activities:
(Organizational) Learning in an AT sense means that individuals, teams or
organizations create new activities (above all new objects and motives).
Learning that eventually comes up with new activities or even new forms of
activities can take several forms: from experimenting with alternative
solutions for ongoing problems via “misusing” tools or concepts from other
activities in new ways, up to creating new tools, prototypes and new mental
models. The elicitation of work knowledge itself is part of this learning
process since it encourages reflection and by that possibly opens up new
future pathways.
Through these principles, AT is presented as a framework for knowledge elicitation,
especially for the elicitation of human work knowledge. By that, AT can be
understood as a sort of glasses through which work knowledge can be located,
important aspects of interrelated knowledge are considered and the elicitation process
can be seen. These principles are meant to help the elicitation of work knowledge in a
holistic and context-sensitive way taking into account the dynamic, changing nature
of knowledge as well as the impossibility of knowing in advance how new tools (like
a software) will change the context, the activity, and the knowledge itself. Moreover,
these principles emphasize that the elicitation process itself becomes part of the
context.
3 Work Interaction Development Support
In this section, we study how models and specification approaches can support
activity-based design of socio-technical systems, as they aim capturing work
knowledge in terms of how tasks can be accomplished. In addition, specification
should allow (i) representing context, and (ii) refining behavior-relevant knowledge
from elicited representation to implementation-relevant execution schemes. The latter
supports prototyping and generating artifacts. We build upon findings from model-
based design approaches as they aim to incorporate context information including task
and user characteristics while taking an implementation-oriented engineering
perspective on development (cf. [58-60]).
In the following sub sections we introduce a workflow-oriented development
approach and demonstrate how activities encoded in behavior representations can be
used to construct socio-technical facilities. The starting point is contextual inquiry
enabling to shape behavior entities that finally could be refined for role-specific
execution of dynamically evolving systems.
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
116
3.1 Contextual Inquiry and Activity-system-based Behavior
Refinement
In this section, we discuss the application of the overarching AT principles to task
acquisition and analysis techniques with respect to task- or role-based development.
In particular, we reveal consequences to
(i) the setting and procedure, and
(ii) fundamental entities and relationships designers need to tackle.
Both aspects have an impact on how designs develop and allow representing a
work setting from a socio-technical perspective. The setting and procedure provides
the scope of acquisition and frames the activities and their sequence to be set for
eliciting task knowledge. The fundamental entities and relationships address the
content, in particular its structure that designers utilize when acquiring task-relevant
knowledge. In the following, we refer to fundamental properties of the AT and the
selected principles when discussing each of the aspects (i) and (ii).
Setting and Procedure. According to the principles derived from AT considerations
in section 2, designers capture work processes in and of organizations in a context-
sensitive way. More specifically, AT advises designers to account for the dynamic
nature of these processes. They need to recognize human work knowledge in its
continuously changing nature according to its specific social and cultural-historical
context. For task-oriented interaction design framed by AT, designers needs to
consider an organization as an activity system, where design work is embodied, while
AT defines the setting for eliciting, representing, and processing work knowledge, but
does not prescribe domain-specific or inner structures of this system. Hence,
respective development support should allow developers generating models of work
processes
• in a flexible as well as structured way
• keeping the ‘big picture’ of task accomplishment while
• recognizing relevant perspectives and elements (including their
relations) that
• make sense to members of an organization as they understand the
organization and the work processes they are part of.
The object-orientedness of activity (1) has an effect on the setting as both
perspectives, the actor’s and domain expert’s of the organization need to be captured.
Activity systems as units of analysis (3) contain both conscious actions and
unconscious automated operations conducted by actors. As activities represent the
background against which concrete actions, knowledge, its elicitation, as well as
human/organizational needs have to be put in relation to all other elements of one or
more interacting activity systems. They finally also represent or enable developing
design alternatives framed by relevant context elements.
Both principles have an effect on the procedure, as each activity is part of an
activity system and directed towards a work object. This object is transformed
through actions by one or more actors into an (organizational) outcome. Hence, the
starting point of the elicitation process needs to be the common object. Although
117
thereby the data (exchange) play a crucial role, the object is considered in the context
of a role accomplishing a task involving the object, and thus, in line with the setting
given by AT.
The dynamics of activity systems (5) raises the awareness that every analysis and
design have to be considered as snapshots in time. Not only that activity systems are
dynamically changing, each intervention (including acquisition and design)
transforms components and relations. These changes might have an impact to the
meaning of the activity system and individual sense-making processes. The major
challenge in knowledge acquisition therefore is recognizing it as part of the current
context. Modeling approaches used for knowledge elicitation have not only to capture
the relations being correct at a certain point in time, but also adaptable for future
designs. The models themselves provide the structural means that developers activate
at a certain point in time.
Consequently, adaptability needs to take into account the interdependency in time
(6) as changes of activities have their specific history and trigger specific
developments. In order to preserve acquired knowledge of work processes,
development repositories or design memories keep development context, such as
stakeholder data and time, besides the actual content, e.g., task models. They help to
avoid running through similar development cycles without being effective in design
and development. Consequently, organizational learning as development of new
forms of activities (12) has to become integral part of development methods. It
enables individuals and collectives to proceed or initiate development activities (cf.
double loop) while operating in their work activity system. New activities (above all
new objects and motives) need to be reflected before being embodied into new socio-
technical work practice [61]. From a method perspective, knowledge claims referring
to underlying problems or assumptions, and design rationales need to be acquired and
documented for each learning step.
The mediated character of activity (2) has essential impact on how to acquire and
process knowledge. In this context the problems of as-it-is and as-it-should-be pops
up. Once design is understood as inherent to the activity system, it is an intervention
on-the-fly. This holds for acquisition as reflection as-it-is as well as for brainstorming
as-it-could/should-be. Each activity in that system is changing according to the
knowledge, means and rules that become evident when analyzing work processes.
New tools or services are an intervention, starting a new mediation process that
becomes part of design and development. This dynamic (and non-predictable) nature
of design is rarely reflected in usability life cycle considerations. In particular,
domino-effects beyond individual work tasks or procedure up to reflecting sense-
making are rarely part of methodological designs.
Contradictions as driving forces (4) play a crucial role in that context, since, as
grown tensions within and between activity systems, analysis and design can be a
learning endeavor for all stakeholders of the system. These tensions can reveal
practices and designs that have not been considered so far. They lay ground for social,
technical, and organizational innovations and change, e.g., in terms of coordinating
actions, IT tool support, and role assignments, respectively. Even if technical artifacts,
such as novel technologies trigger change and interventions in an activity system, the
social character of activities (7) needs to be recognized.
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
118
According to the analysis in section 2, the social character also holds for individual
task accomplishment, since a work result is considered as some form of expression
from (in terms of a cognitive model) and in a social context (we produce a result for
other humans). Traditionally, work knowledge is not considered as social in its nature,
as acquisition, analysis, and design consider tasks as goal oriented activity that needs
to be evaluated in terms of its objective results, e.g., a good or bad service. It is even
de-humanized.
In that context we have to consider the multi-voicedness of activity (8). Accepting
individual mental models and thus cognitive representations of human work and
socio-technical systems, developers need to respond to different perspectives resulting
from each stakeholder's way handling tasks, tools, and organizational settings. As part
of contextual inquiry acquisition techniques need to elicit viewpoints of stakeholders
as well as their interrelations.
Not only are various perspectives a methodological concern, context is represented
at certain levels of abstraction. As each of the elicited work activities can be
considered an activity system on a more abstract or a more concrete context level, the
interdependency of different context levels (9) can influence analysis and design.
Specific items, such as role behavior or tool characteristic, can have specific meaning
according to the considered level of context, e.g., individual task accomplishment and
department. In terms of contingency theory, each activity system encapsulates
another. As such, it might influence sense-making processes of stakeholders. Personal
sense-making of knowledge (10) is likely to influence knowledge elicitation, as it
reveals the personal meaning and the needs that an activity system fulfills for an
individual. As many actions might be performed unconsciously, the elicitation of tacit
knowledge as a form of externalization of activities (11) requires methodological
consideration.
Acquisition techniques should comprise externalization of mental models.
Underlying motives and constructs triggering practical activities influence the way we
deal with tasks, goods, services, material, tools, and people. Externalization also
captures previously internalized activities, as due to routines or task complexity they
have become encoded into mental models. According to AT, the activity should be
the unit of elicitation and analysis, as it allows asking for the personal sense of task
accomplishment. Sense-making finally represents the context of meaning of the
(elicited) knowledge, i.e. how acquired information needs to be understood. It is an
inevitable source for the non-reductionist transfer of elicited knowledge to formal or
diagrammatic models, and subsequently to technical system features. In this way, this
knowledge is communicated to other people, either directly through representation, or
indirectly, through user behavior when interacting with technology.
Fundamental entities and relationships. The setting and procedure provide the
frame for content that designers need to address from an AT perspective. In the
following, we apply to task- and stakeholder-centered design the developed principles
given in section 2, as elicitation entities and their relationships are essential for
developers when creating task-effective digital support systems or interactive
artifacts.
Activity systems as units of analysis (3) defines the scope of elicitation and
representation, as an activity provides the background against which concrete actions
119
become their form (Gestalt). Models need to capture the inner structure and
interrelations of element categories besides the categories themselves, in particular the
actions performed by actors. All elements of a particular activity system designers
have to address also from a relational perspective, as they could provide the context to
other elements of one or more interacting activity or activity systems.
For elicitation, representation and execution of work knowledge for design and
implementation, the object-orientedness of activities (1) plays a crucial role. Work
activities serve as context element for actions from an actor (subject) perspective (role
– functional, formal, semi-…), and from external (object) perspective according to the
domain, sector or type of organization at hand. Since each activity is directed towards
an object that is transformed through actions by one or more actors into an
(organizational) outcome designers need to ask for the effect on the activity system
and also specify required input and output data. However, the search for the common
object is crucial for detecting the starting point of the elicitation process. It influences
the perspective, level of abstraction, and scope an acquisition procedure takes.
As all activity is mediated through knowledge, developers need to elicit means
(tools) as well as social rules valid for a working process (mediated character of
activity (2)). They also need to represent design knowledge in a way that the effect of
new software or design, can be evaluated before implementation, since it changes the
original activity. From the dynamics of activity systems (5) a new process of
mediation is started. Although, then the elicitation process does not end with
modeling an (existing) activity, it starts (again) with applying (prototypical) new
means.
For implementing interdependency in time (culture-historicality and future
anticipation) (6) - when activities change over time and inhere their own history and
their own possible future path - changes need to be captured on a meta-level. It would
allow observing all developments over time, and thus, facilitate collective learning
processes which brings organizational learning as development of new forms of
activities (12) into play. It is also a question of system dynamics. It requires direct
addressing of underlying drivers or mediators of work. In that context (4)
contradictions as driving forces have a special quality, as it is not very likely that all
members of an organization agree on proposed changes nor that a learning step is free
of conflicting elements. Hence, contradictions need to be kept as an input pool for
further analysis or design.
When the social character of activity (7) is addressed explicitly, notations used for
elicitation and specification need to capture respective elements and relations.
Although from an AT perspective all activities are intrinsically social, social
interaction with other actors could be represented and modeled explicitly, e.g., as
done in subject-oriented business process modeling [62]. The latter allows
encapsulating actors, actions, and objects/tools through corresponding abstraction.
The granularity hereby could be identified from an individual perspective as it
facilitates sense-making of represented knowledge (Personal sense-making of
knowledge (10)) – see next section.
The other principles, in particular (8) Multivoicedness of activity and (9)
Interdependency of different context levels have been considered as part of the setting
and procedure in the previous subsection, as they are not part of elicitation and
specification process directly. For (11) Elicitation of tacit knowledge as a form of
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
120
externalization of activities, elements could be generated through using specific
methods, such as Repertory Grids (cf. [12]). Asking for knowledge categories, such as
values, that become evident on tasks or actors, could lead to novel content items and
design elements (cf. Stary, 2014).
3.2 Behavior Encapsulation and Interaction Design
In the following we model activity systems as set of behavior encapsulations
according to the idea of implementing organizations as system of (social) actors. Once
an activity system has been elicited it can be refined in a non-disruptive way [63].
Each activity system consists of active elements or actors (modeled as action systems
in AT) and their relations for communication and interaction.
We follow the approach of Subject-oriented Business Process Management (S-
BPM) [62]. It allows representing simultaneously acting components in any type of
organization. They operate as active, autonomous, concurrent behavior entities. Such
entities can be a human, a piece of software, a machine (e.g., a robot), a device (e.g., a
sensor), or a combination of these, such as intelligent sensor systems. Action systems
are termed subjects according to the S-BPM approach. For the sake of intelligibility
we denote them S-BPM subjects in the following. S-BPM subjects can execute local
actions that do not involve interacting with other S-BPM subjects, as well as
communicative actions that are concerned with exchanging messages between S-BPM
subjects, i.e., sending and receiving messages. Messages also can contain business
objects, and thus, are carriers of data manipulated by the actions of an activity system.
S-BPM involves all stakeholders concerned by the actions of an activity system in
terms of their task-specific roles. They are represented by S-BPM subjects and are
one of five core symbols used for specifying S-BPM models. Based on these symbols,
two types of diagrams exist in S-BPM which can be produced to conjointly represent
an integrated system: S-BPM Subject Interaction Diagrams (SIDs) and S-BPM
Subject Behavior Diagrams (SBDs):
• Subject Interaction Diagrams (SIDs) provide an integrated view of an
activity system, comprising the S-BPM subjects involved and the messages
they exchange.
• Subject Behavior Diagrams (SBDs) provide a local view of the process from
the perspective of individual S-BPM subjects. As they refine actions, they
capture AT operations. They include sequences of states representing local
actions and communicative actions including sending messages and
receiving messages. Since SBDs also implement rules through the flow of
actions, the conditions under which operations are set can be captured in an
integrated way. Dedicated handler can be added to handle complex events.
An activity system or situation can be structured in S-BPM as an interacting set of
action systems, encoded in S-BPM diagrams according to their communicating with
each other. When these action systems need to communicate directly via message
(containing business objects) with another action system, as e.g., required in case of
Personal Scheduler, a S-BPM subject-behavior diagram also encodes this link. It is
executed during runtime once being implemented. On the modeling layer, the
121
corresponding activity is a request sent to another S-BPM subject. The sending S-
BPM subject waits until it receives an answer. Then, it processes the received answer
– see Figure 4 for that pattern. The rectangles denote the messages that the action
systems exchange.
Fig. 4. Detailed activity system ‘Patient’ refined as set of AT actions based on interacting S-
BPM subjects for personal assistance in home healthcare
Figure 4 shows a Subject Interaction Diagram (SID). SIDs provide a global view of
an activity system, comprising the S-BPM subjects (shaded rectangles) involved and
the messages (white rectangles) they exchange. They represent activity-system
specific actions as actors (S-BPM subjects) involved in communication when
considering a patient at home coordinating several digital healthcare support systems
as activity system:
• Personal Scheduler coordinates all home healthcare tasks. As a subject
Personal Scheduler is a behavior abstraction of an action system and can
either be performed by humans, digital devices or services. In design we
abstract from actual implementations. The Personal Scheduler handles all
patient tasks to be set in a certain period of time, and can be set public to
other concerned stakeholders, such as relatives, medical experts, and social
services.
• Medication Handler takes care of providing the correct medication at any
time and location and is collaboratively controlled by medical care takers
and the patient.
• Blood Pressure Measurement is a healthcare instrument supporting services
to recognize the medical condition of the patient. They are linked to their
provider companies or producers, ensuring proper operation.
• Shopping Collector serves as container for all items to be provided for
medication and wellbeing.
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
122
Since each of the action systems need to follow a certain task logic, e.g.,
identifying the demands from existing medications when making a shopping list, each
of them needs to be refined accordingly through Subject Behavior Diagrams (SBDs).
The refined view on individual subjects represents sequences of operations, including
communication with other action systems (sending and receiving messages).
Sequences are represented as arrows, with labels indicating the outcome of the
preceding operation (see Figure 5). The part shown in the figure represents a
scheduling request to the Personal Scheduler subject from the Medication Handler
subject. It also reveals the choreographic synchronization of behavior abstractions,
allowing to represent action systems in parallel while being synchronized through
operationally required message passing.
Fig. 5. Scheduling request to the Personal Scheduler subject from the Medication Handler
subject on the operation level when refining action systems
Given these modeling capabilities, subject-oriented designs (S-BPM models)
capture work interactions through simple communication protocols (using SIDs for an
overview) and (ii) standardized behavior structures (enabled by send-receive pairs
between SBDs), which (iii) scale in terms of complexity and scope. In addition, all
required data are represented along the interaction paths. In this way, subject-oriented
process integration is adaptive, as it allows meeting ad-hoc and domain-specific
requirements, once a corresponding interaction behavior can be identified.
As validated behavior specifications can be executed without further model
transformation, stakeholders can control the entire process, ranging from elicitation to
specifications of executable domain-specific work flows, and even can make ad-hoc
changes by replacing individual S-BPM subjects during runtime. As long as the
interaction interface between S-BPM subjects hold, their internal behavior (i.e.
operations and actions being part of an activity system) can be modified. Figure 6
gives an example for handling a monitoring task when connecting a dedicated device,
123
such as Blood Pressure Measurement, to a Personal Scheduler on request, or when
incorporating intelligent sensor systems in ambient environments.
The subject-oriented system structures, such as integrated blood pressure
measurement devices in personal health scheduling support systems, according to
their communicating with each other. When these devices need to communicate
directly with the cloud, e.g., as required in case of maintenance, or calling a specialist
for medication, this link is encoded in the subject behavior diagram, and executed
during runtime after technical implementation. On the modeling layer the activity is a
request sent to another subject, waiting until an answer is received, and processing the
received answer (see also Figure 6 for that pattern).
Fig. 6. Monitoring request processing (SID)
Once an SBD, e.g., the Blood Pressure Measurement subject is instantiated, it has to
be decided (i) whether a human or a digital device (organizational implementation)
and (ii) which actual device is assigned to the subject, acting as technical subject
carrier (technical implementation). Typical subjects as edge devices are smart devices
which can have Internet connectivity, including smart phones, tablets, laptops,
healthcare devices, etc.
Figure 7 provides a schematic visualization of this constellation, as it can be used
for implementing the sample home healthcare support system (see also [14]).
Infrastructure nodes are subject carriers representing resources including hardware
(compute, networking and storage) capabilities. They provide ‘local’ real-time data
processing capabilities, and can, despite multi-tenancy, execute applications in
isolation to prevent unwanted interference from other action systems. Policies to
control service orchestration, filtering, and for adding security can be implemented
dedicating a specific control subject. In case the open source engine UeberFlow [64]
(download at: http://www.i2pm.net/interest-groups/ueber-flow/home) is utilized for
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
124
execution, operations and thus, actions are ordered in the sequence as defined through
S-BPM models.
Fig. 7. Execution Computing Architecture
The approach takes into account the structured findings revealing that perspectives
on the situation trigger
1. actions of activity systems encapsulating behavior by focusing on actions needed to
be performed to achieve an objective or implement an intention (usually referring
to some task), and thereby, establishing some functional role
2. operations as refinements of actions allowing to detail functional/technical services
for task accomplishment
3. communication acts identifying which action system(s) needs to be interacted with
4. the mutually adjustment of encapsulated behavior specifications, as it plays a
crucial role not only for acting as a collective in a specific situation but also to
complete work tasks or reach intended goals.
Subject-oriented design models contain interactions in a flexible as well as structured
way – SIDs keep the ‘big picture’ of task accomplishment while recognizing relevant
perspectives and elements (including their relations) that make sense to members of
an organization as they understand the organization of work tasks and related
behaviors. When using subject-oriented representations throughout elicitation
implementing the 12 principles detailed in section 2.2 can be actively supported – set
the relevant text to italics: A SID is able to represent an action system as units of
analysis for a specific actor, e.g., patient. When defining the scope of work task
interaction, S-BPM also implements the principle of object-orientedness of work
activities. New means can be captured by changing S-BPM models, probing them as
novel work practice. Hereby the social and mediating character of activities is
captured through exchanging messages and business objects. Contradiction as driving
forces can be documented on the collective and individual layer through SIDs and
SBDs.
125
Since each stakeholder can generate individual subject behavior (personal
sensemaking of knowledge) framed by identical communication patterns, S-BPM
models, multivoicedness of activities can be documented, even their interdepencence
in time. Subjects allow behavior encapsulation on various levels of abstraction,
enabling interdependency of different context levels. Finally, the dynamics of activity
systems can be managed by adapting settings in a subject-oriented way. The
adaptation become essential when externalizing activities stemming from tacit
knowledge is supported, and learning cycles at the organizational level lead to new
forms activities.
4 Conclusions
In our analysis we presented AT as a framework for acquiring knowledge and
representing human work tasks and their arrangement for meeting specific goals. The
defined principles allowed us to understand AT as a context-sensitive way to locate
work knowledge and ask for important aspects of interrelated items. Development
techniques have to take into account the dynamic, changing nature of knowledge,
modifying the context of tasks and their accomplishment on-the-fly.
The developed principles help to reflect on current elicitation and explore
capabilities of development techniques. We have investigated the setting, procedure,
and representational elements. Taking into account the coupling of actors, actions,
and objects is facilitated by an activity-system-centered development approach.
Behavior encapsulations represent task-relevant roles and work tasks, while business
objects are part of interactions between actors or technological artifacts. Finally,
change and learning require meta-level representations and processes, in order to
sustain and keep already achieved insights in those processes.
However, further research is required, in particular coupling elicitation and
acquisition procedures implementing an activity-system perspective to design patterns
and development languages. To that respect modeling skills on various levels of
abstraction are required. User support could stem from tangible digital media as
already prototyped for work modeling (cf. [65]). The refinement of activity systems
into actions and operational procedures may require deeper analysis and novel
instruments for articulation support (cf. [66]). Finally, for understanding motives
underlying individual value systems should be elicited and aligned with already
externalized knowledge integrating explicit with tacit knowledge, e.g., through Value
networks (cf. [56]).
References
1. White, G. R. T., Cicmil, S. (2016). Knowledge acquisition through process mapping: factors affecting
the performance of work-based activity. International Journal of Productivity and Performance
Management, 65(3).
2. Cash, P., Hicks, B., Culley, S. (2015). Activity Theory as a means for multi-scale analysis of the
engineering design process: A protocol study of design in practice. Design Studies, 38, 1-32.
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
126
3. Clemmensen, T., Kaptelinin, V., Nardi, B. (2016). Making HCI theory work: an analysis of the use of
activity theory in HCI research. Behaviour & Information Technology, 35 (8), 608-627
4. Nardi, B. A. (1996a). Activity theory and human-computer interaction. In B. A. Nardi (Ed.), Context
and consciousness: Activity theory and human-computer interaction (pp. 7-16). London: MIT Press.
5. Nardi, B. A. (1996b, Ed.). Context and consciousness: Activity theory and human-computer
interaction. London: MIT Press.
6. Nardi, B. A. (1996c). Studying context: A comparison of activity theory, situated action models, and
distributed cognition. In B. A. Nardi (Ed.), Context and consciousness: Activity theory and human-
computer interaction (pp. 69-102). London: MIT Press..
7. Kaptelinin, V. (1996). Activity theory: Implications for human-computer interaction. In B. A. Nardi
(ed.), Context and consciousness: Activity theory and human-computer interaction (pp. 103-116).
London: MIT Press.
8. Kaptelinin, V., Nardi, B. (2012). Activity theory in HCI: Fundamentals and Reflections. Synthesis
Lectures Human-Centered Informatics, 5(1), 1-105.
9. Kaptelinin, V., Nardi, B. A. (2006). Acting with technology: Activity theory and interaction design.
MIT Press.
10. Barab, S., Schatz, S., Scheckler, R. (2004). Using activity theory to conceptualize online community
and using online community to conceptualize activity theory. Mind, Culture, and Activity, 11(1), 25-
47.
11. Hemmecke, J., Stary, C. (2004). A framework for the externalization of tacit knowledge embedding
repertory grids. In Proceedings OKLC-2004, 5th European Conference on Organizational Knowledge,
Learning and Capabilities, Innsbruck (Vol. 56).
12. Hemmecke, J., Stary, C. (2006). The tacit dimension of user tasks: elicitation and contextual
representation. In Task Models and Diagrams for Users Interface Design (pp. 308-323). Berlin,
Heidelberg: Springer.
13. Ackerman, M., Goggins, S., Herrmann, T., Prilla, M., Stary, C. (2017). Designing healthcare that
works: a sociotechnical approach. Academic Press.
14. Stary, C., Fleischmann, A., & Schmidt, W. (2018). Subject-oriented fog computing: Enabling
stakeholder participation in development. In Proceedings 2018 IEEE 4th World Forum on Internet of
Things (WF-IoT), IEEE, 7-12.
15. Zhang, W., Kumar, M., Yu, J., & Yang, J. (2018). Medical long-distance monitoring system based on
internet of things. EURASIP Journal on Wireless Communications and Networking, 2018(1), 176.
16. Yi, S., Li, C., Li, Q. (2015). A survey of fog computing: concepts, applications and issues.
In Proceedings of the 2015 Workshop on Mobile Big Data, ACM, pp. 37-42. Blackler, F. (1993).
Knowledge and the theory of organizations: organizations as activity systems and the reframing of
management. Journal of Management Studies, 30 (6), 863-884.
17. Huang, D., T. Xing, H. Wu (2013). Mobile cloud computing service models: A user-centric approach.
In IEEE Netw., 27 (5), 6–11.
18. Shi, Y., Ding, G., Wang, H., Roman, H. E., Lu, S. (2015). The fog computing service for healthcare.
In 2nd International Symposium on Future Information and Communication Technologies for
Ubiquitous HealthCare (Ubi-HealthTech), IEEE, pp. 1-5.
19. Hackos, J. T., & Redish, J. (1998). User and task analysis for interface design. Wiley & Sons, New
York.
20. Hacker, W. (1992). Expertenkönnen: Erkennen und Vermitteln. Göttingen: Verlag für Angewandte
Psychologie.
21. Büssing, A., Herbig, B., Ewert, T. (1999). Implizites Wissen und erfahrungsgeleitetes Arbeitshandeln:
Konzeptualisierung und Methodenentwicklung. Berichte aus dem Lehrstuhl für Psychologie der TU
München Nr. 48. München: Lehrstuhl für Psychologie der TU München. Available at
http://www.psy.wi.tum.de/LS-Berichte/Bericht-48.pdf (May 30, 2016).
22. Senge, P. (1990) The fifth discipline: The art and science of the learning organization. New York:
Currency Doubleday.
23. Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization Science, 5
(1), 14-37.
127
24. Kim, Y. J., Chaudhury, A., Rao, H. R. (2002). A knowledge management perspectice to evaluation of
enterprise information portals. Knowledge and Process Management, 9 (2), 57-71.
25. Karanasios, S. (2018). Toward a unified view of technology and activity: The contribution of activity
theory to information systems research. Information Technology & People, 31 (1), 134-155.
26. Blackler, F. (1993). Knowledge and the theory of organizations: organizations as activity systems and
the reframing of management. Journal of Management Studies, 30 (6), 863-884.
27. Jonassen, D. H., Tessmer, M., Hannum, W. H. (1999). Task analysis methods for instructional design.
Mahwah, New Jersey: Lawrence Erlbaum Associates.
28. Leont'ev, A. N. (1978). Activity, consciousness, and personality. Englewood Cliffs, New Jersey:
Prentice-Hall. Available at http://lchc.ucsd.edu/mca/Paper/leontev/ (May 30, 2016).
29. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes.
Cambridge, Massachusetts: Harvard University Press. (edited by Michael Cole, Vera John-Steiner,
Sylvia Scribner, Ellen Souberman).
30. Bourguin, G., Derycke, A., Tarby, J. C. (2001). Beyond the interface: Co-evolution inside interactive
systems—A proposal founded on activity theory. In People and Computers XV—Interaction without
Frontiers (pp. 297-310). London: Springer.
31. Dahme, C., Raeithel, A. (1997). Ein tätigkeitstheoretischer Ansatz zur Entwicklung von brauchbarer
Software. Informatik-Spektrum, 20, 5-12.
32. Raeithel, A. (1992). Activity theory as a foundation for design. In C. Floyd; H. Züllighoven; R. Budde;
R. Keil-Slawik (eds.), Software development and reality construction (p. 391-415). Berlin: Springer.
33. Uden, L., Willis, N. (2001). Designing user interfaces using activity theory. In Proceedings of the 34th
Hawaii International Conference on System Sciences. IEEE Computer Society Press.
34. Boer, N.-I., van Baalen, P. J., Kumar, K. (2002). An activity theory approach for studying the
situatedness of knowledge sharing. In Proceedings of the 35th Hawaii International Conference on
System Sciences. IEEE Computer Society Press.
35. Clases, C. (2003). Eine arbeitspsychologische Perspektive auf soziale Dynamiken kooperativer
Wissensproduktion. In H.-W. Franz, J. Howaldt, H. Jacobsen, R. Kopp (eds.), Forschern lernen
beraten: Der Wandel von Wissensproduktion und -transfer in den Sozialwissenschaften (pp. 303-324).
Berlin: edition sigma.
36. Clases, C., Wehner, T. (2002). Steps across the border: Cooperation, knowledge production and
systems design. Computer Supported Cooperative Work, 11, 39-54.
37. Hasan, H., Gould, E. (2001). Support for the sense-making activity of managers. Decision Support
Systems, 31, 71-86.
38. Sukumaran S., Chandran K., Chandran K. (2014). Knowledge Management Strategy Using Activity
Theory for a Law Firm. In Uden L., Wang L., Corchado Rodríguez J., Yang HC., Ting IH. (eds.), The
8th International Conference on Knowledge Management in Organizations. Springer Proceedings in
Complexity (pp. 521-531). Springer, Dordrecht.
39. Uden, L. (2009). Activity Theory for Knowledge Management in Organisations. In A. Zilli, E.
Damiani, P. Ceravolo, A. Corallo, G. Elia (eds.), Semantic Knowledge Management: An Ontology-
Based Framework (pp. 201-215). IGI Global.
40. Blackler, F. (1995). Knowledge, knowledge work and organizations: An overview and interpretation.
Organization Studies, 16 (6), 1021-1046.
41. Blackler, F., Crump, N., McDonald, S. (2000). Organizing processes in complex activity networks.
Organization, 7 (2), 277-300.
42. Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental
research. Helsinki: Orienta-Konsulit.
43. Engeström, Y. (2001). Expansive learning at work: Toward an activity theoretical reconceptualization.
Journal of Education and Work, 14 (1), 133-156.
44. Virkkunen, J., Kuutti, K. (2000). Understanding organizational learning by focusing on „activity
systems“. Accounting, Management and Information Technology, 10, 291-319.
45. Kelly G. A. (1991). The Psychology of Personal Constructs. London: Routledge (Reprint, New York:
Norton, 1955).
46. Leont'ev, A. N. (1989). The problem of activity in the history of soviet psychology. Soviet Psychology,
27 (1), 22-39. (Originally published in Russian in Vopr. Psikhol. 4/1986, 109-120.)
Interaction Design and Architecture(s) Journal - IxD&A, N.37, 2018, pp. 100-129
128
47. Leontjew, A. N. (1982). Tätigkeit, Bewußtsein, Persönlichkeit. Köln: Pahl-Rugenstein. Available at
http://www.kritische-psychologie.de/1982/taetigkeit-bewusstsein-persoenlichkeit (May 30,
2016).
48. Raeithel, A. (1983). Tätigkeit, Arbeit und Praxis: Grundbegriffe für eine praktische Psychologie.
Frankfurt am Main: Campus Verlag.
49. Wehner, T., Raeithel, A., Clases, C., Endres, E. (1996). Von der Mühe und den Wegen der
Zusammenarbeit: Theorie und Empirie eines arbeitspsychologischen Kooperationsmodells. In
E. Endres; T. Wehner (eds.), Zwischenbetriebliche Kooperation: Die Gestaltung von
Lieferbeziehungen (pp. 31-58). Weinheim: Beltz Psychologie Verlags Union.
50. Hasan, H. (2003). An activity-based model of collective knowledge. In Proceedings of the 36th Hawaii
International Conference on System Sciences. IEEE Computer Society Press, New York.
51. Hemmecke, J. (2012). Repertory Grids als Methode zum Explizieren impliziten Wissens in
Organisationen: Ein Beitrag zur Methodenentwicklung im Wissensmanagement. PhD thesis, Vienna
University. Available at http://othes.univie.ac.at/27576/ (May 31, 2016).
52. Leont'ev, A. N. (1975). Deyatelnost, soznanie, lichnost. Available at
http://www.psy.msu.ru/people/leontiev/dsl/index.html (May 30, 2016).
53. Jonassen, D. H., Beissner, K., & Yacci, M. (1993). Structural knowledge: Techniques for representing,
conveying, and acquiring structural knowledge. Psychology Press.
54. Schreiber, G. (2000). Knowledge engineering and management: the CommonKADS methodology. MIT
press, Boston, MA.
55. Stary, C., Maroscher, M., Stary, E. (2012). Wissensmanagement in der Praxis:-Methoden-Werkzeuge-
Beispiele. Carl Hanser, München.
56. Stary, C. (2018) Transactional Value Analytics in Organizational Development. In Analytics and
Knowledge Management, eds. S. Hawamdeh, H.-C., Chang, CRC Press, Boca Raton, 241-270.
57. Schlippe, A. von, Schweitzer, J. (2006). Lehrbuch der systemischen Therapie und Beratung.
Göttingen: Vandenhoeck & Ruprecht.
58. Paternò, F. (1999). Model-Based Design and Evaluation of Interactive Applications. Berlin: Springer.
59. Dittmar, A., Forbrig, P., Heftberger, S., Stary, C. (2004). Support for Task Modeling–A “Constructive”
Exploration. In Engineering Human Computer Interaction and Interactive Systems (pp. 59-76). Berlin,
Heidelberg: Springer.
60. Stary, C. (2009) Model-based Tools: A User-Centered Design-for-All Approach. In C. Stephanidis
(ed.), The Universal Access Handbook (ch. 25). London: Taylor & Francis.
61. Firestone, J. M., McElroy, M. W. (2003). Key issues in the new knowledge management. Routledge.
62. Fleischmann, A., Schmidt, W., Stary, C., Obermeier, S., Börger, E. (2012). Subject-oriented business
process management. Berlin: Springer.
63. Stary, C. (2014). Non-disruptive knowledge and business processing in knowledge life cycles–aligning
value network analysis to process management. Journal of Knowledge Management, 18(4), 651-686.
64. Krenn, F., & C. Stary (2016). Exploring the potential of dynamic perspective taking on business
processes, Complex Systems Informatics and Modeling Quarterly 8, 15-27
65. Oppl, S., & Stary, C. (2014) Facilitating shared understanding of work situations using a tangible
tabletop interface. Behaviour & Information Technology, 33(6), 619-635.
66. Oppl, S., & Stary, C. (2019) Designing Digital Work - Concepts and Methods for Human-Centered
Digitaliization, Palgrave Macmillan.
129
