IEEE Technology and Society Magazine - September 2017 - 79

design principles for self-governing institutions that support sustainable common-pool resource management
[32] through an axiomatization in (executable) computational logic.

Social Relations
Weber [40] claimed that an action is "social" if the acting
individual takes account of the behavior of others and is
thereby oriented in its course. Further, Sztompka [37]
proposed a hierarchy of social interactions and relations,
shown in Table 1. The hierarchy makes clear how several
social concepts already familiar to computer scientists
relate to each other. For example, social behavior is
commonly analyzed in ant-based systems (e.g., [14], [16]),
where actions (e.g., leaving pheromone) are done for the
benefit of other ants, and in doing so for the benefit of
the colony as a whole. Similarly, regular interactions
describe the kinds of interactions occurring in repeated
games such as the iterated prisoners' dilemma [1]. Here,
knowledge of future interactions with the same opponent
is crucial to determining future behavior.
However, Sztompka's hierarchy demonstrates that
there are many forms of social interaction, which require
varying forms of knowledge (concerning oneself, other
individuals, and the environment) as well as cognitive
capabilities. In developing this tenet of our framework,
our intention is to map Sztompka's framework to computational systems, and extend as necessary. In particular, the various forms and distinct characteristics of
interaction and relation in the socio-technical complex
[9] need to be considered for the design of sociallysensitive systems.

Social Organization
Organization is perhaps the most familiar feature of the
social nature of technical systems. It is concerned with
network structure, roles of individuals and sub-groups
within the system, as well as the processes by which
such organization arises.
Self-organization is a concept that originated in
physics and chemistry, where it was used to describe
how processes and interactions on microscopic level
can lead to the emergence of organization, such as patterns and structures, at a macroscopic level [6]. In biology it can refer to the phenomena that global order can
be created through local interactions between molecules or cells [20] and it has been used in the higher
level investigations of emergent behaviors of groups of
animals such as fish, birds, or social insects, e.g. [9].
Self-organization is now investigated in a variety of
fields, not only the natural sciences, but also life sciences (where it has been studied at various levels ranging
from viral particle assembly [36] to flocks and schools
of animals [9]), in social sciences [29], [31], [30], and in
SEPTEMBER 2017

∕

TABLE 1. Sztompka's Sociological Hierarchy
of Interactions
Type

Requires

Behavior

Physical Movement

Action

Meaning

Social Behavior

Directed towards others

Social Action

Await response

Social Contact

Unique/rare interaction

Social Interaction

Interactions

Repeated Interaction

Accidental, not planned, but
repeated interaction

Regular Interaction

Regularity

Regulated Interaction

Interactions described by law,
custom or tradition

Social Relation

A scheme of social interactions

more technological fields such as computing sciences
and artificial intelligence [7].
Computer scientists [14], [21] have drawn on work
from theoretical biology [2], [11], [12] to exploit the benefits of decentralized approaches that rely on the interaction between many agents instead of one central
control mechanism. The foundations provided from
work on social insects behavior have been used by
researchers to design distributed control and optimization algorithms [7].
There is now a substantial literature on the organization and self-organization of (socio-) technical systems,
and these aspects will underlie many of the other social
concepts that a system may explicitly possess. Indeed,
the organization might be thought of as the platform, or
set of constraints, upon which social relations play out,
and social values are observed and propagated.

Formalizing Social Concepts
In summary, we propose that for socio-technical systems
to possess, be aware of, and act in accordance with
social concepts, these social concepts will need to be formalized and made explicit. Further, we argue that they
will need to be designed in. We describe such systems
and their design process as socially-sensitive systems
design. A key benefit of socially-sensitive systems will be
better positioning, through increased social potential. We
have proposed here a framework for socially-sensitive
systems design, based on the three core tenets of social
organization, social values, and social relations.
Ultimately, we anticipate the benefit to technical systems to include increased robustness, increased empathy with humans, and reduced pathologies of digital
communities. Further, there is the potential for insights

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2017