Changing

Things
 Changing Things

The Future of Objects in a Digital World

Johan Redström and Heather Wiltse

 Contents

1 Introduction
2 What Is Going On with Things?
3 Just Press Play, Please
4 Fluid Assemblages
5 Things for Us
6 Things in Themselves
7 A Conceptual Toolkit
8 Assembling an Analytic Playlist
9 Making Concepts

References
Index
 1

Introduction

We are surrounded by things. It is hard to think of any moment of life in which

we are not using or at least around artifacts designed and produced by humans.
Acquiring, configuring, arranging, repurposing, and eventually disposing of
things constitute a cycle that provides a kind of ongoing material narrative that
continues throughout life. In at least industrialized regions, most humans arrive
in the world to find a full assortment of things already present around them; and
going through people’s things, deciding what to keep and what to dispose, often
marks the last phase of settling their affairs after they have passed. Since its
beginnings in mid-eighteenth-century Great Britain, the Industrial Revolution
and corresponding development of consumer culture shifted the design,
production, and acquisition of things into high gear and almost uninterrupted
acceleration, such that some people and parts of the world are now often rather
drowning in excess or discarded stuff, while many others suffer from the
consequences. We are technological beings through and through and since our
earliest beginnings, and gradually we have come to the point where it is now the
artificial rather than what could be perhaps called “nature” that constitutes the
ultimate horizon of our existence.
And now, again, the ways things are made and used have fundamentally
changed. Many of the things we now live with do not take only physical form:
smartphones and watches, laptops and game consoles, wearable health and
fitness trackers, and similar are different from our things of the past. For reasons
to be explored in this book, we will call these things fluid assemblages.
Whereas your mobile phone might look rather similar to any other sleek
appliance at home, you only need to disconnect it from the network to realize
just how different it is. The mobile phone, and what you do with it, is in no sense
delimited or defined by its physical presence. Sure, we can touch it, feel it, look
at it—but it is what we do with it, or perhaps rather through it, that defines its
role in our life: searching, talking, watching, browsing, sharing, shopping,
listening, tracking, posting, liking, and so on. It is nearly impossible to assess
and understand what these things do based on physical appearance alone. Unlike
our old stable and predictable physical things, these digital things are networked,
dynamic, and contextually configured. Indeed, they can at times also be
changeable and unpredictable things, even inscrutable when it comes to what
they actually do and whom they really serve.
Just as we can recognize the very real actions and effects of these networked
computational things, it is also in many cases becoming ever more difficult to
clearly identify and pin them down. It is almost as if they emerge from some
kind of virtual ether when activated, only to withdraw again when the
applications are closed or the devices switched off. Moreover, they show up in
different ways depending on the context, responding to variables such as user
account, location, time, and so forth. They can also change over time as the code
is updated, while the code that generates unique instances of things can travel
across devices. But when they are up and running, these things—compositions
of devices, code, and networked resources and connections—are remarkably
solid in terms of providing certain kinds of functionality and enabling certain
actions.
In what follows, we will explore the idea that the changes in what things are,
and how we live with them, currently underway are in some ways as drastic as
the shift from handmade to industrially manufactured things. And while this time
it is also a change driven by technological developments and economic interests,
to think of it as a matter of the digital and computational would be much like
reducing the Industrial Revolution to a matter of iron and steel, steam engines
and electricity. Rather, it is in the reconfigurations of relations between people
and things, between the design of and life with, that the most significant
transformations are taking place.
While the first decades of industrialized production were primarily about
mimicking forms familiar from earlier forms of production, there was a gradual
realization that this change actually called for a different understanding of how
something becomes meaningful in someone’s life. This insight sparked the birth
of industrial design as we now know it: the realization that the new industrial
opportunities called for a new unity of art and technology, a new sociopolitical
program, a new aesthetics, and above all a new vision of the “good life.” Or, as
seen from another perspective, industrialized capitalism had to call a new
discipline into being to realize its full potential—and so it did.
We will argue that we now face a challenge somewhat similar to this situation:
a fundamentally different way of “making things” made possible by rapid
technological development and further accelerated by new economical ideas that
are currently radically outperforming the pace in which design develops. This is
causing an increasingly problematic gap between what these new “things” are
and the ways we are meant to incorporate them into the fabric of our everyday
lives. Of course, there is already massive benefit present in our lives, just as also
the first years of industrially produced copies of previously manmade objects
were of great value. Indeed, one might even be inclined to think we are already
doing so well that this should not be of too much concern. But there is also a
slightly uncanny feeling that these new things bring about something that is not
necessarily what we first thought. The fact that the business model of some of
the now largest corporations is built on trading detailed information about what
we do, with whom, where, and how, is at least sometimes causing concern,
leaving us with a feeling that perhaps we do not quite understand precisely what
it is that these things do. And while we’re at it, why do we always have to make
do with “beta” these days, why can’t things just work?
At the same time, they are wonderful things. Things we want to have with us
all the time. Things we want to keep in our hands. Some of them are the first
things we pick up in the morning and the last things we put down when going to
bed. They are things we use to stay in touch with loved ones. They offer safety,
excitement, entertainment, inspiration, and much more. This complex situation
cannot be reduced to a matter of good or bad. Instead we have to ask to what
extent do our existing ways of relating to things, be it in use, design, practice, or
theory, actually allow us to grasp what these new things are? And most
important, what sense and sensibility do we now need to cultivate in order to
design and live well with them?
Our purpose with this book is to investigate and begin to articulate some of
the key characteristics of these things—these fluid assemblages—so that we
might be able to better understand and care for their character and consequences.
In doing this, we are attuned to the constant and in many ways inevitable
interplay between theoretical lenses and empirical observations, many of which
have originated with our own experiences of interacting with and through these
things. The theoretical frameworks we are accustomed to thinking with,
explicitly or implicitly, individually and within discursive communities, shape
what we can see, connect, conceptualize, visualize, and articulate; and it is when
these are no longer adequate to account for what we are seeing and break down
that we must do the challenging and often distinctly uncomfortable work of
pushing beyond them, even as the only way beyond is through.
In this spirit, we have worked explicitly and intentionally with an interplay
between case studies and theoretical frameworks, using each to push back
against and expand the other. Although we at times foreground one or the other,
the two are always intertwined. Although we do our best to respect the areas of
scholarship with which we engage, our faithfulness is primarily to the account
we try to develop. Our aim is to develop a perspective that is incisive, rich, and
connected to key foundational issues in philosophy, design, and related areas, not
least because of the many strains of thought we bring into conversation with
each other. We do not claim to do this comprehensively or rigorously, but hope
that we can through our (unfaithful) articulations at least sketch the outlines of
territories worth investigating—with the hope that others may join us there and
improve on our efforts.
Even as we engage with things that currently exist, our goal is not merely to
give an account of the present, but to identify the underlying dynamics,
trajectories, and potentialities that may guide future developments. Moreover,
while present configurations, technologies, and sociotechnical norms may
provide the environment and resources that scaffold and shape ongoing
processes of design and development, there is always the possibility to engage
with these matters more explicitly and intentionally. This is the kind of reflective
and critical engagement that we hope to enable and prototype here, inquiring
into what is but always with an orientation to what things could (and perhaps
should or should not) be assembled. In short, our project is about changing
things.

About this book

Over time, a variety of ways of thinking and talking about contemporary things
have been developed in order to attempt to understand them and the roles they
play in the world. New technological developments require new theory, while (at
least ideally) incisive critical theory can inform responsible and innovative
design practice and, thus, the building of the artificial world. This interplay
between things and “thing theory” has been ongoing at least since the Industrial
Revolution, and has been made explicit in design theory, philosophy of
technology, and other fields concerned with the character of contemporary life
and society.
There are a few key assumptions and commitments that guide our
investigation. The first is that when we refer to “things,” we mean designed
technical artifacts, not things in general. Certainly, all the material things in the
world are significant and worthy of philosophical consideration. But here we are
more specifically concerned with things that are the objects of industrial design
—things that are designed with at least some degree of professional design
expertise and in service of particular client and user groups, goals, functions, and
interests.
The second is that we are interested in how these things actually exist in the
world and what they do. In this sense, our approach could be considered
materialist in that we believe, to use the overused expression, matter matters and
is not reducible to experience, concepts, semantics, discourse, social
constructions, subjectivities, and the like. While these aspects are absolutely
significant, and in very concrete ways, they also provide only a partial
perspective. This is especially important to recognize as we consider things that
outstrip our usual sense-making capabilities, thus requiring us, somewhat
paradoxically, to move outside the frame of human experience.
At the same time, human experience is our motivation for addressing these
things. We care about the ways in which things mediate human experience and
engagement with the world, and how experience and engagement are shaped by
the character of things and the variety of mediations they can enable. In the
context of contemporary challenges such as sustainable development, such a
human-centric view might seem quite insufficient as the things we are
addressing, their production and their use, have an impact on the world far
beyond human experience. The reason for starting with human experience
instead lies with the need to understand and articulate how we perceive and act
upon this new complexity. Or in other words, if the categories—such as
“things”—we form and use as the basis for the way we understand and reason
about our actions do not quite match what is actually there, then chances are we
will not be very successful in changing things in the direction we intend.
Finally, while our focus here is on the things we call fluid assemblages, our
investigation is in some ways broader. We find that in order to understand what
is going on with fluid assemblages we need to bring in a number of different
perspectives and work across a variety of scales, from the level of local
interactions to that of larger social structures and dynamics. This is a matter of
properly accounting for the different kinds of entities that are gathered together
in things that guide their continual unfolding, modes of presence, appropriations,
figurations, and configurations in social worlds in which they act in ways that
are not neutral. As such, it turns out that a large extent of our investigation is as
much about designed things in general as it is for fluid assemblages in particular,
even as fluid assemblages represent gatherings of more striking dynamism and
scale. While we point to fluid assemblages as a new and distinctive type of thing,
we also recognize continuity with previous kinds of things rather than total
rupture. In fact, many of the dynamics entailed in fluid assemblages are simply
the realization of tendencies and forces at work long before technical
developments enabled them to be expressed in this particular form.

Overview of the book

This book has three main parts. The first serves to introduce the notion of fluid
assemblages and the design issues they entail. Following this introductory
chapter, we ask the question of “What is going on with things?” In Chapter 3,
“Just Press Play, Please,” we move deeper into a specific example of how these
developments unfold, and we then conclude this first part with our initial
account of what is a fluid assemblage.
The second part of the book engages more broadly in concepts and cases,
taking a journey through a range of different domains and disciplines to enable a
multifaceted view on things. In Chapter 4 we develop the concept of fluid
assemblages in more detail. In the next two chapters we consider existing
possibilities for considering things as they exist for us and things in themselves,
respectively. Then in Chapter 7 we develop a conceptual toolkit for fluid
assemblages, drawing on and applying concepts developed in other contexts. We
try out the toolkit in Chapter 8, assembling an analytic playlist in an extended
analysis of Spotify as one particular case.
With a set of cases and associated conceptions in place, we are then better
equipped to return to the main question raised here, namely that of how to design
and live well with fluid assemblages. The third and final part of the book
consists of the concluding chapter in which we further articulate issues in the
designing of and the living with fluid assemblages. In particular, we will discuss
how the notion of an assemblage helps us explain why current aesthetics and
approaches to design are seemingly not able to address and resolve certain
critical issues in how these things are made part of our lives and how we use
them. This opens up, or so we will argue, for rethinking how industrial design
has come to configure design and use in relation to each other, and how this now
needs to be replaced by other configurations to form a different social contract
(much like how mass manufactured goods could not rest on the same social
contract between maker and user as was the case when making was bespoke and
a craftsmanship). Indeed, whereas things used to be passive resources for us to
make part of our lives, we are perhaps now closer to a more symmetric set-up:
we are quickly becoming as much part of the doings of things as they are a part
of ours. We may even ask: if the primary human “product” of the industrially
manufactured object was the “consumer,” what would be the corresponding
entity brought about by the fluid assemblage? This book is an attempt to find
out.
 2

What Is Going On with Things?

To ask “what is a thing?” is to dwell on a question at least as old as philosophy.

Our relation to things, in the wide sense, is so significant for our existence that
perhaps only our relation to other people matters more to us. As such, these
questions have formed the basis for much intellectual effort over the course of
history. Even one of the central questions asked here—how things that change
can be perceived as both the same and not the same at the same time—has
fascinated people for thousands of years. In the story about the ship of Theseus,
a thought experiment used by Heraclitus, Plato, and onward, the materials of a
ship are gradually replaced as the wooden planks and other materials decay,
eventually resulting in a ship that retains none of its original parts. Is it still the
same ship? In the seventeenth century, the philosopher Thomas Hobbes would
complicate matters further by asking us to consider what would happen if all the
discarded pieces were used to build another ship, so that in the end we have two
ships: one that has continuously retained a certain form over time and another
that is made from the original materials. Which one is now Theseus’s ship? Are
there now two of them? What is it that makes it possible for a thing to have an
identity over time: form, material, or something else?
Across a wide range of academic disciplines, there are enormous resources to
draw upon if one wants to understand more about things and how we relate to
them: with respect to material, aesthetics, culture, politics, economics, and so on.
And yet, the reason we keep asking the same questions over and over again is
that things change, urging us to add yet other pieces to the by now gigantic
picture we collectively try to paint of what it means to make and live with things.
And thus we are trying to add only a very small piece here.
In order to bring into focus what is now changing when it comes to our
everyday things, it may be helpful to consider briefly a few aspects of things that
have been possible to more or less take for granted up until now.

Taking things for granted

We can start at the very beginning, with our perceptual experience of things. A
“thing” is something quite extraordinary. It is something that stands out,
something that in our experience emerges as a meaningful whole that can be
understood as some form of unity distinct from its surrounding. We could think
of perceptual phenomena such as the ability to distinguish figure–ground
relationships. Precisely what it is that grounds this experience of a meaningful
unity is far from trivial to understand, but we do know that it begins already in
basic perception and that aspects of these abilities may even be hardwired in our
perceptual systems. For instance, how an infant develops the perception of
object permanence, such as being able to follow an object and to understand also
that things that are out of sight can still exist, was called out by development
psychologist Jean Piaget as critical to how our intellectual abilities develop. Not
to mention it enables us to play Peekaboo.
The fact that artifacts have fairly stable physical properties is what allows us
to make sense of and relate to them in other ways as well. For example, it is a
hammer’s solidity, in many senses, that allows for appropriating it for various
purposes. Regardless of whether that use is the most typical one of striking
another object or another perhaps more unusual one, such as using it for a
doorstop, it is the hammer’s physical form and properties that allow for these
actions. Ecological psychologist J. J. Gibson even argued that our perceptual
systems have evolved to intimately connect the sensorial, and primarily visual,
presence of things with an immediate perception of what we can do with them,
and coined the term “affordance” to describe what it is that we directly see we
can do with a thing (Gibson 1979). Indeed, Gibson’s theory of direct perception
is partly grounded in an analysis of variants and invariants in the ambient optic
array that the animal perceives as it moves around in an environment. For such
perceptual systems to evolve, rich aspects of the animal–environment
relationship must remain constant beyond more rudimentary forms of object
consistency.
Certainly, things also change, but they do so in certain ways and not others. A
wooden handle may perhaps become worn and smooth over time and use, but
even this kind of process of aging is itself fairly straightforward and predictable.
Or, we can think of a raincoat that can be used for weather protection but also as
a way to block the sun, a pillow, a cover, a container, a way to carry things by
using the pockets, and certainly many other things as well. The capability of
coming into multiple relations to things through use is based on creative
appropriation of a generally stable physical form.1
Even less physically solid things can still be stable with respect to their
character. Take, for example, the classic children’s toy “silly putty,” a pliable,
somewhat sticky material that can be molded, stretched, torn apart, and smashed
back together into a single piece again. It flows slowly if left to sit, and when
pressed against newsprint will lift off the ink. It is a remarkable substance (that is
great fun to play with), and this is largely due to its complex set of properties and
resulting variety of possible behaviors. Yet even with this variety, these core
properties themselves remain stable and thus, at least at a basic level, predictable
and understandable.
Most things are not only relatively stable over time, but also across instances
of the same type of thing. Each individual artifact is in some ways unique due to
its particular existence in time and space and its own history; yet at the same
time, and especially when it comes to things that are mass-produced, it is
generally reasonable to expect that similar types of things will be consistent
across different specific instantiations. So different kinds of hammers tend to still
be quite similar to each other, and hammers of the exact same model will likely
be all but indistinguishable, especially when brand new. Indeed, this example
would not work as an illustration if this was not the case: that one can evoke an
experience of what a thing is and what can be done with it by just saying a word
to someone. The same goes for toasters, chairs, lamps, dry erase markers,
raincoats, flower pots, bicycles, and many other things that populate our world.
In addition to stability, another factor that helps in making sense of things is
the fact that the connections between forms and functions are often readily
apparent. Indeed, making these connections clear has been considered a central
objective in many industrial designs. For instance, the term “affordance”
mentioned earlier has made its way into design as well, and although prevalent
interpretations of affordances in design differ substantially from Gibson’s initial
theory, the term is still used to address how to make the intended use of
something easy to perceive and understand.
Thus, there are many reasons for us to take things for granted. Indeed, most of
the time it makes sense to think of them as stuff, simply there to be used.2 While
we recognize the difference between natural and artificial, there are actually
quite a few aspects of experience suggesting that the two categories are not that
different from each other: they are both there for us to be used when and how we
want to. In theory, literally, this attitude translates into a treatment of things as
already there, and it is only in certain forms of design theory and philosophy that
we actually see an account of how they come into being (and continue to
evolve). But, as we will argue, it is actually this “coming into being” that is the
perhaps most critical aspect of contemporary things. Currently obscured by our
persistence in taking things for granted, contemporary things have become
dynamic and active, constantly responding and changing over time. While recent
technological developments such as an “Internet of Things (IoT)” may have
accelerated this development, this story begins much earlier.

Things have changed

The first significant milestone in the change in what things are (at least from the
perspective of our present investigation) was the introduction of mechanics, and
its miniaturization that eventually followed. While simple mechanical
constructions certainly expand the complexity of what a “thing” can be, they
maintain some kind of relation between internal and external complexity. That is,
a more complex or powerful machine has also tended to present itself as more
complex and powerful. The reason is simple: because of the way mechanical
constructions work, more complicated ones typically require more space as well.
The same holds for power; for instance, a huge steam engine is typically more
powerful than a small one.
Over time, something intriguing is starting to happen in these relations,
between internal and external complexity. As mechanics began to be
miniaturized, our immediate perception of external versus internal complexity
started to evolve in new directions. An interesting illustration is the historical
fascination with automata. In Europe, these evolved significantly during the
renaissance, thanks to the development of more advanced and miniature
mechanical constructions and clockworks. Often, the intention was to mimic
living creatures, to give the illusion of something mechanical being alive; and
although we would not mistake any such design for an actual living creature,
some of these designs are distinctively uncanny. Still, although certainly very
complex, key aspects of the basic relation between visible and actual complexity
were still present.
Digital technologies introduce a fundamental break with this condition as
there is no longer any perceivable relation between internal and external
structure, complexity included. Looking at an early digital device such as a
pocket calculator, we can of course inspect its interior and, knowing something
about electronics, we might be able to locate the different parts, trace
connections between the keypad and analog/digital converters, processor, etc.—
but we cannot really “see” anything of what it does when it is operating besides
the information that has been designed to be presented on its screen.
Nevertheless, the pocket calculator is still possible to understand as a “thing,” as
perhaps especially its exterior possesses a certain degree of physical
permanence. Although mechanics to some extent introduced a more complex
temporal complexity to what a “thing” can be, even the most complex automata
are next to nothing when compared to the temporal flexibility opened up by
computational technologies, and the gradual separation of machine and program.
After computational technologies came networked computational
technologies. While it might at first glance seem that this development amounted
to more or less adding communication as another function, the significance of
networking capabilities was much more profound in terms of changing what
things are and what they do. This meant that the scope of possible actions and
effects, as well as resources available, was no longer limited to the immediate
physical vicinity of the thing but was rather determined by the topology of the
network of which it was part. To be sure, many of the actions supported can be
seen as acts of communication. Yet, there is also something else going on when
it becomes possible to essentially change the state of another thing or cause it to
do something remotely. This is what happens every time a sent message appears
in an inbox, sets off a notification, or causes a smartphone to vibrate in a pocket,
to take a simple example. We can also think here of all the many financial
transactions that now take place via computational systems: everything from
personal banking to automated high-frequency stock market trading that now
plays a huge role in the world of global finance. We can see that (computational)
things are involved here, but it is the connections between them that enable these
sometimes extraordinarily complex actions that have very real consequences in
the world. An even more extreme and clear example is that of drone warfare.
In addition to dramatically expanding the scope of actions that can be
accomplished through use of a networked thing, the addition of network
capabilities also brought the ability for things to make use of networked
resources during their operations. In other words, instead of being more or less
self-contained and with functions determined by certain local resources, digital
networked things can be composed by components brought together from
elsewhere in the network. This can also happen in dynamic, contextually
dependent, and evolving ways, such that a particular “thing” is constantly
changing both over time and across instances in relation to specific contexts.
Things are often now composed on the fly from a variety of networked
resources, both physical and digital.
However, even if they seem quite straightforward in terms of user experience,
it is not possible to see how they really function beneath their user-facing
surfaces. This is due to both the fact that they are constituted largely by code and
also that even the code is generally not accessible to end users. The notable
exception here is open-source software; yet even in this case, seeing the
underlying code is different from actually watching the processes that are in
operation, connections made, data sent, and so on as a programmed thing runs.
We can see these dynamics at work when we consider something like a
smartphone. It is in a certain sense clearly an object-like, perhaps even beautiful,
product of industrial design that we can hold in our hands. But what we actually
do with the device relies only minimally on its physical presence as a hunk of
metal and glass. Rather, in using a smartphone we make use of the apps and
services that it runs locally or, probably more often than not, that connect to
other systems and platforms. Also, although smartphones of the same model are
all but identical out of the box, customization for a particular user begins as soon
as one is turned on for the first time. This involves setting language and location,
and creating or signing in to an account that is used across devices. If a person
already has such an account that has been used on another similar device part of
the same “ecology” (such as Apple’s iCloud), the settings and apps can be
carried over directly, essentially replicating the forms and functions of an old
device on a new one (even as the new one no doubt has more advanced technical
capabilities that provide new kinds of functionality as well). The device will also
evolve over time as new software updates become available; apps are added,
updated, and deleted; and media content is generated, modified, and removed.
Moreover, many of these changes are much more than cosmetic, since they can
substantively change what a smartphone can do. It is also important to note that
much of this functionality relies on network connections (cellular, wireless
internet, GPS) that are activated as soon as the device is turned on. The extent of
functionality that is enabled by these connections can be brought into sharp relief
when, for example, flying or traveling in a country where one does not have
cellular service. What we think of as this particular smartphone, then, in terms of
how it is present to us as a thing available for use is dynamic and evolving, and
relies on resources not contained within the device itself. And unlike our earlier
things, these things are made to “know” who is using them and to make intensive
use of this information.
Taking a step back, it seems that the basic ways things are made have now
fundamentally changed. This has happened before. Indeed, entirely new areas,
such as industrial design, were called into being to care for the new relations
between design and use that became necessary to develop as industrial
production drove a change from everyday things being handmade (typically by a
craftsman for a known customer—as in how a tailor would make clothes to
measure) to being mass-produced for mass consumption. And while aiming for
the one optimal design that would be the best for all seemed to work well during
the first decades of industrial design, not the least also commercially, the need to
cater for individual identity increased in importance; and thus phenomena such
as “streamlining,” “branding,” and so on that applied varying cultural and
aesthetic expressions to functionally nearly identical products were inevitable.
What is important here, however, is that while such moves certainly changed the
ways we relate to things, and to consumption as an expression of identity, the
basic forms of production still produced reasonably stable objects.
Certainly, industry has changed significantly over the past century in all sorts
of ways, globalization included; but much of this has not necessarily been that
present at the surface level of interacting with products. Taking a new piece of
clothing made by a brand we have been using for a long time, we might notice
that while it used to be produced in a certain place it is now made somewhere
else, that the tactile quality is slightly different as certain materials have been
changed, etc.; but typically we see little that challenges our perception of this
new thing as yet another instance of, for example, a certain kind of raincoat. At
least we do not see any change amounting to the shift from made to measure by
someone for someone, to the industrial mass production for mass consumption.
And so although production has changed in a number of ways, it has not
changed in ways that fundamentally challenge how we basically perceive what
the thing is and how it is made and used the way that the shift from craft to
industrial production once did. Networked computational things, however,
constitute a radical shift from this condition.
During the investigations and explorations that led to this book, we have come
to think that the change from things to fluid assemblages actually is a shift in
some ways as fundamental as industrialization when it comes to challenging the
way we understand what things are. And again, it is how the things are made—
the assembling of things—that is transforming the relations between production
and consumption, between design and use.
Just as radical is the shift from things that are more or less defined by their
physical presence to things that are dynamically defined on the basis of both
physical and digital components and in relation to similarly dynamic and
interlocking networked infrastructures. What we experience as things are
therefore now not so much like the metaphorical tips of icebergs with underlying
structures that we cannot see, as might have been said of something like an early
pocket calculator, but rather like the top layers of lava flows that span continents
and are constantly moving, changing shape, connecting to other flows, and
eventually creating new formations.
We need to account for the assembling of things in order to understand their
temporal forms and their associated relations and consequences, and we need to
account for the structure of the resulting assemblages in order to understand their
networked, spatial, topological forms and their associated behaviors and
capabilities.
There are, therefore, compelling reasons not to take things for granted. While
prevalent dichotomies between production and consumption, between design
and use, enforce perspectives that as a person you are either making a thing or
experiencing something already made, the character of contemporary things cut
across such simplifications. If we remain with such divisions between making
and using, design theory will stay with the former, philosophy of technology
with the latter. But as we will argue in this book, these basic dichotomies need
revisiting if we are to design and live well with these things. And so, this is
where we begin: at the intersections between making and using, trying to
articulate and understand things that are unfolding, assembled, and dynamic.

Toward a new account of things

Starting to articulate this shift, we suggest that the technical artifacts that now
exist in our world are more like fluid assemblages than what we traditionally
think of as things: assemblages because they are made out of a diverse range of
material and immaterial resources both contained within the object as it appears
in front of us as well as located elsewhere in the network; fluid because their
precise forms are assembled dynamically and thus change continuously. These
are quite different from the stable, predictable things that are the traditional
objects of industrial design and focus of theorizing about the artificial. In the
ways they are manifested, the functions they serve, and the “user experiences”
they support, these fluid assemblages can appear to be quite thing-like. And yet,
the interfaces that are engaged in use are only the surfaces that conceal the many
layers of code, platforms, systems, and interconnections between them that allow
specific instances to come into being as things available for use.
Moreover, in many cases it is not at all straightforward to determine what
constitutes “use” of a thing, and by whom. We can see this, for example, when
we consider the many popular web applications in which people essentially pay
for the service through having their personal data and attention sold to
advertisers. In fact, one of the more notable shifts in moving from stable things
to fluid assemblages is that, rather than ending at a single, fixed point of sale,
relations between producers and consumers often continue for as long as use of
the product continues. In fact, it now seems like this relationship is even more
stable than the things themselves, as terms of service often state that these may
change over time.
From a commercial perspective, considerable effort is put into telling
compelling stories about shifts from “products” to “services” to help customers
understand the new forms of business. Just think about what the notion of a
personal “music collection” refers to now, with services such as Spotify and
Apple Music, in contrast to earlier practices of collecting records. Overall, we
can begin to see a range of new practices with quite different dynamics than
those that have generally governed design and use so far. We can see, among
other things, use feeding into continual design and dynamic customization;
multiple stakeholders extracting different kinds of value through ongoing use
(e.g., value for both users and those who aggregate data regarding use for
various purposes); and different arrangements around ownership, with
differences also in associated rights, regulations, and restrictions regarding
access and use. To give an example: few people anticipated that access to, and
aggregation of, people’s internet searches would become the perhaps most
important asset in the advertisement business, and that a business based on a
“search engine” thus would reshape the economic reality of newspapers and by
extension that of journalism too.
In parallel to these technological developments, there has been across a
number of discourses a “return to things.” With the earlier emergence of the
post-industrial society and the introduction of the new technologies we associate
with it in everyday life, there came an interest in and also significant orientation
toward information and communication, the immaterial, virtual and symbolic.
Perhaps it is therefore no surprise that it was followed by a “return to things”
that displaces the postmodern emphasis on subjectivities and semiotics with
attention to objects and materiality. While to some extent certainly a backlash
against the general excesses and limitations of the linguistic turn in some areas,
and the visions of technological futures turning to tunnel visions of immaterial
information societies in others, these new perspectives in many cases also have
deeply political and practical motivations. Ranging from pressing issues of
sustainability and how to live within the constraints of finite resources to the
realization that also computation and communication are things literally taking
place, the “real” world continuously calls for our attention, as does the very
“real” experience of being a body in the world. In many ways, the synthesis of
the thesis and its antithesis in ideas such as that of “an Internet of Things” was
therefore inevitable. And so we return to “things,” this stuff that we create and
that according to some is what defines us and the time we live in: the artificial as
the effective horizon of human existence, the Anthropocene.
And yet, as we can see by even the brief survey of technological
developments earlier, we are now dealing with a rather new set of dynamics and
logics around things and their production, use, and functions. This state of affairs
leaves us with a complexity and a rupture that we do not yet fully understand but
that we, as a start, need to try to articulate. Or in other words, what is now
needed is to move beyond thinking about the present discourse as a “return to
things,” since in fact what we now encounter is something we have never lived
together with before. This is certainly not a return to something we were once
comfortable with. We need a new vocabulary and theoretical/analytical approach
that will allow us to speak to the multifaceted, dynamic, constantly emerging and
evolving forms and functions of contemporary things. The aim of this book is to
sketch out such an approach, and in doing so to prototype a discourse around
things as fluid assemblages.

Notes
1 This ability to come into multiple distinct relations to things is referred to as multistability within the
subfield of philosophy of technology known as postphenomenology (e.g., Ihde 1990). In this context, it
serves to counter technological determinism and more substantive theories of technology by
 highlighting the capacity for human agency and creativity in relation to things.
2 This also aligns with philosopher Martin Heidegger’s diagnosis of how our relation to technology
causes us to see the world (including possibly other people and even ourselves) as bestand, or standing
reserve, simply there as a resource to be used (Heidegger 1977).
 3

Just Press Play, Please

In what follows we will take another look at the historical developments outlined
in the previous chapter, but this time as seen through an example: the very
simple and mundane act of pressing a play button to listen to music.1
While techniques for music playback have quite a long history, there has been
in just the past decade or so an explosion in the development of technologies and
services for listening to music. Yet, even with such a diverse array of music-
playing technologies historically and in terms of currently available options,
some things remain constant. We approach a music-playing thing because we
want to hear music; and whether we find it on the plastic button of a tape or CD
player, the click wheel of a classic iPod, or in the interface of a digital app, we
know to look for the familiar right-facing triangle icon. To hear music, we press
play.
This simple act of pressing play has remained quite consistent, even as the
complexity of the underlying systems that make the playing possible has
increased tremendously. Indeed, many digital music players are now only one
component of vast ecosystems including digital service providers, musicians,
record labels, advertisers, and other digital platforms. Moreover, they participate
in these ecosystems in much more complex and dynamic ways than their simpler
historical predecessors.
We will here consider simply the assembly or activation of a music-playing
thing such that its functionality of pressing play is made available to us. We will
trace the nature of this assembly through a variety of cases, beginning with older
analog technologies and watching for significant changes as we move on to
address more contemporary technologies and (eco)systems.
Through these illustrations, we will introduce another key idea in this book:
while our new massively data-driven and networked computational things are
inherently different from previous “things,” it seems that what designers aim to
express and users expect to experience in use are not at all this radical disruption
but rather something familiar and in line with what we have lived with before. In
other words, what we tend to search for and expect in use is something of a
paradox, namely that the complexity of technology is resolved through a
simplicity of expression. This is a design paradox we need to unpack as it sits at
the center of what designing and living with fluid assemblages is like at the
moment.
To understand the background of this tension between technical complexity
and keeping things simple, we need to turn back to the developments in
industrial design in the early 1950s and the search for clear and functional design
of (technical) objects on one hand, and a growing interest in increasingly
complex products and systems on the other hand. At the design school HfG Ulm,
we see early examples of both the kind of industrial design that is still today
highly influential in the technical domain (e.g., the HfG–Braun collaboration
about the design of radios and other music systems for the home) and the first
steps toward making design an interdisciplinary project set up to deal with
increasingly “wicked problems” (e.g., diagrams and graphs inspired by
mathematics used to visualize design and its processes).
While the orientation toward simplicity stems from a concern for usefulness
and utility, it is more than anything also an aesthetic orientation. Indeed, if there
is one well-known phrase that captures the Modernist aesthetic, it would
probably be “form follows function.” To say that it is an aesthetic orientation is
not to diminish its scope, but to suggest that it is a central part of a “design
worldview” inherently tied to a certain way of thinking and creating design
expression. Of course, there are many other aesthetic orientations, but in
industrial design as practiced in the parts of the world where these new devices
are often conceived, this worldview is still something much deeper than just one
out of many orientations one may choose between while designing. And while it
largely has given way to other orientations, and especially more diversity, within
many domains of design, it still exercises a massive influence in the domain of
tools and technology. Indeed, notions such as “ease of use” and “usability” are,
within the domain of computational things, still central qualities sought.
The concern for complexity also has its roots in early Modernism, and the idea
that there are important relations between design and society, between individual
objects and industrial systems. Even in early examples of industrial design, we
see an explicit interest in how objects are related to each other in systematic
ways, and how design can help address and make sense of the resulting
complexity. Indeed, it is almost as if the orientation toward simplicity in
expression coevolves with the increasing complexity of function and the systems
the things exist within: the more complex the functions and the relations between
things, the more important to keep things simple. From a design point of view,
this may seem like an, or even the most, obvious response, which makes it even
more important to unpack it as the aesthetic orientation it actually is. Because
there is nothing given, necessary or “natural” in the domain of design. The
artificial is, through and through, inherently something made by someone—and
thus something that always could be made differently.
Now, let us trace the simple act of pressing play to see how things have indeed
changed.

A short history of music players

Beginning with pre-digital examples of music-playing technology, we can think
of classic record and tape players. These devices are quite respectable design
objects in the classic sense: things that are mass-produced in factories and then
purchased by users who then own and can do whatever they like with them.
Playing music on them requires plugging in the power cord and perhaps pushing
a power button, loading a record or tape, and then pressing play. Especially in
the case of a record player, it is typically possible to manually intervene at more
or less all stages of the mechanical process as well, such as manually lifting and
positioning the arm with the needle onto the record, stopping or slowing down
the rotation with the hand, etc. All of the assembling of these things occurs on
the manufacturer’s side before they reach the end users, and unless he or she
decides to physically modify the device it will remain the same. Although the
mechanics involved are somewhat harder to inspect, this logic also applies in the
case of stand-alone CD players.
However, when we move to considering a CD played in a computer rather
than a stand-alone CD player, we notice some different dynamics emerging.
First, and most obviously, computers do much more than play CDs; this is only
one of many functions they have which are managed by the underlying operating
system and installed software. And software is in fact needed to play the CD.
The physical CD drive may not seem much different than that of a classic CD
player, but the fact that it is now operated by software marks a key shift. There
are now multiple software options that can be used in conjunction with the same
CD drive—ones which can be updated and configured independently of the
underlying hardware, and thereby change overall functionality. Also, and in
contrast to the elegant simplicity and functional transparency of a record player
arm lifting up and over a record, loading a CD in a computer launches countless
computational processes that are not generally visible (although it might be
possible to use system monitoring tools in order to see some of what goes on).
This would often include a query to a music database service such as Gracenote
in order to retrieve the track names for the CD, revealing another (networked)
component of the assemblage.
CDs can also be ripped and stored in a computer hard drive, marking another
key development when music can be stored in digital formats and played without
the need to load an external storage device. And of course they can also just start
in digital format and be distributed without the need to ever involve physical
storage media other than computer hard drives. As music-playing things, digital
music player applications appeared as a rather new kind of animal. They are
“assembled” from a variety of components when the application is launched.
These include the code for the application itself and the underlying operating
system that manages its processes, including sound output (which might be
internal or peripheral speakers). The computer itself is the component that seems
most object-like in a traditional sense; yet even with this simple example we can
see that only a small part of its functionality is determined through its assembly
in a factory. Much of what a computer does is rather determined by its operating
system and applications, which can be updated and configured in ways that can
greatly change functionality without changing the underlying hardware.
In addition to the assembling that happens through basic software, it is also
possible to further modify an application’s “assembling” as a thing available for
use through configuring its settings. These can change how it behaves and how it
looks. One such example are various forms of automatically generated playlists,
ranging from “shuffle” functions first known as “random” playback order in CD
players that mix up the predefined playback order to more elaborate algorithms
based on categorizations, tags, and other kinds of metadata attached to the song.
Another example might be how the classic WinAmp player allowed for
customization through “skins.” Moreover, anyone could develop these skins and
make them available for others; and branding them with one’s logo could
become a point of pride for their creators, serving as a visual reminder that it was
another person (and not WinAmp) who created this particular component of the
user’s personal WinAmp assemblage.
At the same time that the MP3 file format and players such as WinAmp gave
people much more freedom in terms of how they could play and distribute
music, other trajectories sought to restrict the ways in which people could
acquire, listen to, and distribute music, even as they also capitalized on the
possibilities of the digital. The most significant player in this regard is arguably
iTunes, with its “walled garden” approach to providing a coherent and seamless
user experience while also ensuring that only certain kinds of “acceptable” use
are possible. It is well known that the possibility to make infinite duplicate
copies of music files without loss of quality and to easily and widely distribute
them via the internet posed a significant challenge to existing structures in the
music industry, and led to the emergence of new sociotechnical configurations
that is still ongoing. However, even as these dynamics have driven the
development of many more contemporary music-playing systems, our concern
here is with the ways in which these are assembled and appear as things
available for use.

Music and metadata

A key development associated with iTunes and the iPod music player was in
structured metadata associated with media files. This was clearly visible in the
iPod in particular, where music could be accessed in multiple ways, through
artist, album, genre, playlist, etc. Significantly, iTunes also included metadata
reflecting usage, such as play count, skip count, and date last played, as well as
data about when the file was added to the iTunes library and last modified. This
arguably marks the beginning of the evolution of music players in which usage
affects the future constitution and behavior of the system. This can be seen in a
single track itself that has updated metadata, and in the resulting ways in which
tracks are displayed when sorting according to these variables. However, it also
works in a more subtle way by affecting the frequency with which tracks are
played on “shuffle” mode in both iTunes and synchronized devices (such as the
various iPod models and now the iPhone).
Both this personalization of the iTunes data and experience and the
enforcement of certain usage restrictions are accomplished through accounts.
Accounts have now become quite common and effectively extend the
relationship between producer and consumer for as long as use of the product
continues. Even web-based music players that do not require accounts track
users and customize the offerings in fairly sophisticated ways. For example, the
Last.fm music player web page (http://www.last.fm/listen) loads a variety of
trackers, beacons, and analytics that, as of this writing and as revealed by the
Ghostery browser extension, include ones for Audience Science, BlueKai,
ClickTale, DoubleClick, Google Analytics, Omniture (Adobe Analytics),
Qualtrics, Spotify Embed, and Yahoo Analytics. Refreshing the page or
connecting from different locations also updates the musical suggestions
provided. However, the extensive and fluid assemblage of Last.fm is rather
disguised by an interface that invites the user to simply “type in an artist or genre
and press play.”

From playback to runtime assembly

A significant aspect of this continuing relationship between providers and users
is that “use” can be precisely scripted and either enabled or limited in dynamic
ways. For example, use can be customized or restricted based on location. On a
basic level, detecting the country from which a person is connecting to a web-
based service allows for language customization and for presenting what is most
popular in that country. But it can also be used to restrict access to certain
content or prevent access entirely, as in the case of Songza that could not be
accessed through internet connections coming from outside the United States or
Canada (and has now as of this writing been incorporated into Google Play
Music). These restrictions can sometimes, however, be bypassed by connecting
through a VPN service—another component that can be brought into the
assemblage on the side of “use.”
It is interesting to note that when we reach this situation of dynamic
customization there is no longer any single, stable “object” that can be viewed
“objectively.” Instead, what is stable across users is the set of rules and processes
governing the ways in which the product is constituted at runtime for specific
accounts connecting from certain locations at certain times—although even these
rules themselves change over time. Indeed, one of the most prominent aspects of
modern web-based music players, such as Deezer, Slacker, and Pandora, is how
they adapt their music recommendations over time based on what individuals
listen to and indicate that they like.
Importantly, these “things” can also be continuously disassembled. For
instance, streaming content providers may stop making certain content available.
Starting to use Spotify on one device will stop playback on another device.
Content may also stop being available because of changes in the governing legal
contracts, as when the music of an artist from one day to another is no longer
available as a new commercial agreement could not be reached. One can also
experience the geographical specificity of such legal agreements when traveling,
as some content is available in some countries but not in others.
A major dynamic in this runtime production and customization of music-
playing things is that not only are they assembled dynamically, but the
components assembled come from a variety of sources. One way this can be
seen is in the many examples of services that load ads in conjunction with the
application. These ads themselves represent the complex and extensive
assemblages of advertising services, such as Google ads or Apple’s iAd
program. From a slightly different angle, many services now allow for
authentication through social media accounts (such as those of Facebook or
Twitter) and also connect to the functionality of these accounts in other ways
(e.g., loading Facebook friends into a “friends” list, or enabling the sharing of
one’s activity).
The assembling of the music-playing things and their functionality in these
instances is enabled and constituted partially through these other services.
Indeed, quite a few reasons behind the particular design of some of these
assemblages are related to the shift from a focus on consumer purchases to
selling user data in many business models: since what is “sold” is not a “thing,”
but data about the user that can be used to, for instance, customize advertisement
and direct users to certain other services, gathering as much such data as possible
becomes a key driver. This is a major reason for the increasing importance of
accounts to access music, but it can also be seen in the extensive user profiling
and tracking in services not necessarily requiring a login. For example,
SoundCloud’s cookie policy (https://soundcloud.com/pages/cookies) describes
how, in addition to their own cookies used for managing sessions, they use a
number of third party services (from Google, AdsWizz, Global Radio Services
Limited, Quantcast, and Scorecard Research) that provide analytic and
advertising functionality. They also use the “similar technologies” of Clear GIFs,
Flash cookies, HTML5 local storage, activity tracking (“in-app analytics and
messaging service provided by Appboy, Inc.”), app performance tracking
(“adjust” service provided by Adjust GmbH), and bug reporting (“Crashlytics”
service provided by Crashlytics Inc).2
Another general source of input for runtime customization is users
themselves. One way this works is through application settings, but there are
also a number of other means by which use of a thing can later feed back into
how it is assembled. As previously noted, simply recording which music tracks
are listened to can affect how music can be sorted and displayed. This allows for
features that display the artists, tracks, etc. that a person has listened to the most.
Recent listening was placed front and center in the Rdio3 online application,
turning activity into the main content of the site in the form of a collection of
album artwork representing a timeline of recent listens. And of course listening
activity also feeds back into the recommendations provided later on.
Finally, it is interesting to note the extent of the shift from buying something
and then really and truly owning it to using things to which one has access only
provisionally. Systems can be upgraded or, more neutrally, modified without
users’ consent. Content and features can be added and removed. Use is regulated
through a mutually reinforcing combination of system architecture and law
(Lessig 2006) such that, for example, customization and restriction of a web-
based service based on location are reinforced by stipulations that users must not
try to circumvent them. Instead of operating manuals, users are now faced with
sometimes staggeringly extensive terms of service (which are perhaps even less
likely to be read); and they must accept these before gaining access to the
system, thereby entering into standing legal agreements of which most typical
users have only the faintest understanding. Yet, these terms of service sometimes
contain dire warnings and regulations regarding use, such as the Google Play
terms of service that states in part (https://play.google.com/intl/en/about/play-
terms.html):
NONE OF THE PRODUCTS ARE INTENDED FOR USE IN THE OPERATION OF NUCLEAR
FACILITIES, LIFE SUPPORT SYSTEMS, EMERGENCY COMMUNICATIONS, AIRCRAFT
NAVIGATION OR COMMUNICATION SYSTEMS, AIR TRAFFIC CONTROL SYSTEMS, OR
ANY OTHER SUCH ACTIVITIES IN WHICH CASE THE FAILURE OF THE PRODUCTS
COULD LEAD TO DEATH, PERSONAL INJURY, OR SEVERE PHYSICAL OR
ENVIRONMENTAL DAMAGE.4

— Pressing play has become serious business indeed.

Compositions and breach of contract

In earlier work, acts of defining what a given thing is were discussed based on a
distinction between acts of design and acts of use (Redström 2008). Consider a
glass bottle for instance. Acts of designing—of making as craft—a glass bottle
would be acts such as preparing the material, heating the glass, blowing and
shaping it, and cooling it. Acts of designing a bottle for industrial production
would instead entail acts of producing a prototype that can be mass-produced,
through sketches, models, etc. While the process of making the bottle as such
can differ, there is still a clear distinction between such acts of defining what the
“bottle” is, and what then happens as we use it. Use is still a matter of defining
what the bottle “is” (to us), but these acts will be based on the fact that the bottle
is there for us in its physical form. And so I may use it to contain fluid that I can
drink, thus defining it as a drinking vessel, but I can also use it to express my
feelings by throwing it to the wall, thus (re)defining it as a kind of prop in a
performance of sorts. It can be used as a small window in a cottage I am
building, thus defining the bottle as a kind of building material. Eventually, it
will perhaps be used as material for making a new bottle, thus closing the loop.
In general, we might say that there are potentially a range of different acts
defining what this thing is, but that they basically fall into two categories: one of
“design” causing the thing to come into being, and one of “use” bringing the
thing into a practice for some purpose. This distinction has been the stable basis
for a kind of social contract established between design and use that, on one
hand, allows “designers” to create objects for intended forms of use and intended
users, and for “users” to acquire, interpret, and make use of these objects for
their own purposes based on the typically predictable and stable properties of the
objects as present physical things in their lives.
When it comes to the fluid assemblages we now design and use, this basic
picture is breaking down. And importantly, the basic social contract between
design and use is becoming increasingly problematic as the underlying premises
for that contract are being replaced by new forms of making and using. This
causes a rupture that we do not yet know how to address, but that we, as a start,
need to try to articulate. The basic cause for this change is that the fluid
assemblage is never really made, at least not in the common sense that a bottle is
made. The fluid assemblage is continuously in the making, intertwining acts of
design and acts of use over time in ways that traditional mechanical objects
certainly cannot. Yet, as we tried to show with the examples earlier, many
“things” do their best in keeping up appearances, maintaining that the basic
contract is still valid, and that the basic relations between designing and using
are still in place.
Looking at contemporary design, there is in many cases no single, uniform,
consistent, stable thing when it comes to design objects (Wiltse, Stolterman, and
Redström 2015). Rather, as we use computational and other digital materials, the
composition of things is determined on the fly according to a potentially infinite
array of constantly shifting parameters and operations, many of which are
hidden. This new kind of complexity we are now facing is something rather
different, even as it recasts a classic and related tension between simplicity and
complexity in new ways. The complexity that stems from dynamics of use in a
social context was in a way external to the things themselves. The composition
of a thing and the composition of the systems in which it was embedded were
closely related, but also possible to separate. Now we are in a situation in which
the composition of a thing at any given moment is determined in non-trivial
ways by variables that are external to the thing itself.
It appears almost as if we are heading toward a blind spot, where certain
issues are occluded by our prevalent perspectives, and where we therefore need
to develop new accounts of the basic “what” it is that we design. This leaves us
with a problematic gap between existing frameworks and emerging design
issues.

Notes
1 This chapter is based on the paper “Press Play: Acts of Defining (in) Fluid Assemblages” presented by
the authors at the Nordes 2015 conference.
2 https://soundcloud.com/pages/cookies, accessed December 20, 2017.
3 Rdio has now been acquired by Pandora, and is no longer available.
4 The all-caps formatting has been removed since we first noticed this text, but the content remains
otherwise the same.
 4

Fluid Assemblages

In the previous two chapters, we have explored aspects of why and how our
computational networked artifacts are different from what we have encountered
in the past, although they still appear to be “things” to us. Indeed, we have
argued that how we currently design and live with them is something of a
paradox, as we on one hand persist in calling these out as “things” while at the
same time acknowledging that they are not at all like our things of the past. In
what follows, we will try to unpack this seemingly contradictory position and
how it might have come about.

Assemblages
The notion of assemblage used here stems from the work of Gilles Deleuze and
Félix Guattari (Deleuze and Guattari 1987). This is a difficult concept to capture
(and it was probably never meant to live a life in intellectual captivity).
Addressing the agency of assemblages, Jane Bennett offers a very good first
presentation of what these things are like:
Assemblages are ad hoc groupings of diverse elements, of vibrant materials of all sorts. Assemblages
are living, throbbing confederations that are able to function despite the persistent presence of energies
that confound them from within. They have uneven topographies, because some of the points at which
the various affects and bodies cross paths are more heavily trafficked than others, and so power is not
distributed equally across its surface. Assemblages are not governed by any central head: no one
materiality or type of material has sufficient competence to determine consistently the trajectory or
impact of the group. … And precisely because each member-actant maintains an energetic pulse
slightly “off” from that of the assemblage, an assemblage is never a stolid block but an open-ended
collective, a “non-totalizable sum.” (Bennett 2010, 23–24)

There are a couple of key points to make before turning to our things that
change. Looking at different ways that components can be drawn together, the
notion of assemblage sits between collection and totality. In contrast to how a
collection is a gathering of objects that does not have properties beyond the parts
included in the collection, an assemblage is characterized by emergent
properties. An assemblage cannot be reduced to its parts. But whereas a totality
also is irreducible, the assemblage differs from it in that it can still be taken
apart. An assemblage is decomposable. In other words, whereas the collection
does not achieve any additional emergent properties or capacities, and a totality
cannot be taken apart, an assemblage is characterized by both emergent
properties and that its components retain their identity.
To better understand what makes emergent properties come about, and thus
what creates the difference between totality and assemblage, we need to look
into how the components are related to each other. In Deleuze and Guattari’s
account, a totality is characterized by interiority relations; that is, the
components change as they become part of the totality. This implies that they,
typically, cannot turn back into what they were before becoming part of the
whole. The emergent properties of the totality are therefore “transcendent”: a
totality moves above and beyond its constituent parts. An assemblage, in
contrast, is made out of exteriority relations between components, that is,
relations between the components that do not change their individual properties
or otherwise transform them. This makes it possible for an assemblage to be
taken apart (or otherwise transformed), and for the components to retain their
properties afterward. Thus, the emergent properties of assemblages are
“immanent,” a result of continuously interacting components that will disappear
should those interactions come to an end.
If we turn back to the things addressed in this book, even this very brief
introduction of assemblages illustrates the effectiveness of the concept in this
context. The examples of what happens as components start to interact with each
other have shown that we are talking about much more than collections here.
The smartphone is not just a collection of parts; it is an assemblage of apps,
hardware, network connectivity and so on and so forth with capacities way
beyond the sum of its parts. At the same time, however, it is clear that we can
take this assemblage apart, at least to some extent. If our phone had been a
“seamless totality,” for instance, removing an app or changing networks would
not have been possible as there is no reverse. Indeed, the notion of immanence is
a most present aspect of their character: as soon as the interactions between
components do not work, the emergent properties of the assemblage change or
disappear completely. Just consider what happens when connectivity drops, the
battery runs out or the device makes itself unavailable for further interaction
because of a software update.
From a more philosophical point of view, the addition of “fluid” to
assemblage is completely superfluous. An assemblage is in its very nature fluid
in one way or another; this is just another way of saying that its emergent
properties are immanent. The reason for talking about “fluidity” here is to put an
emphasis on how speed has become such a significant factor. While assemblages
such as a community, an organization, or a machine are also constantly in
becoming, they typically change at the pace of human perception or in many
cases even slower. Manually building a machine takes time, as does taking it
apart. These computational networked machines, however, create their particular
instantiations in runtime. The conditions for their coming into being have of
course been created over longer periods of time, but the assembling as such is
literally happening during use—and in response to how use unfolds down to
events happening at the level of microseconds. They have a fluidity that extends
way beyond what we have seen in terms of speed before—and not only in terms
of processor clock speed or network transfer capacity but increasingly also in
terms of “time to market,” frequency of updates, design methodology, and so on
and so forth.

Design methodology
The fact that these things are already a significant part of our everyday lives
seems to suggest that design (in the wide sense of word) has already resolved
much of what is involved in making them. Indeed, in terms of technology their
coming into being has been underway for quite some time, with inventions such
as object-oriented programming and massively networked computing providing
not only the means but importantly also the conceptual frameworks necessary to
conceive them. As industrial design started out with an ambition to create a new
unity between art and technology, we could perhaps think of design as engaged
in a spectrum ranging from the more technical (toward engineering) to the
experiential (toward the artistic). If we look at how design has responded to the
condition brought about by fluid assemblages across this spectrum, we can see
that we are pushing toward development at one end, whereas we seem to be
holding back at the other end.
If we start in the more technical end of this spectrum, we find several rather
recent developments that respond to the design condition posed by fluid
assemblages. Relevant examples include the methodologies for “lean” and
“agile” development. While their origins can be found in the lean engineering
pioneered in Japan by Toyota and frameworks such as Kansei engineering, these
ideas have more recently come to a different expression in software development
and associated industries. These methodologies depart dramatically from earlier
models for product and systems development in which things would only come
together toward the end, after often long processes of development and
optimization. Instead, these approaches rely on short sprints where each and
every effort is meant to produce results that can be shipped as products. Thus,
instead of producing one major “new” product every year or so, the product is
updated continuously, sometimes even on a daily basis.
Design methodology has relied on iterative prototyping to resolve complexity
and uncertainty since its very beginning. To make prototypes is for design a
central part of most learning processes, not just in the general sense of learning
by doing, but more specifically because there are very few ways of learning how
a certain “whole” comes together without actually working its tensions and
conflicting elements out. In this respect, lean and agile are not that different from
earlier practice, and can even be celebrated as a learning process that allows for
failing early and often. However, they also differ substantially from how design
has been done before. Traditional design methodology is completely conditioned
by the requirements of mass production, and as such its learning process is
directed toward optimizsation, toward understanding what needs to be done, and
then resolving all the issues involved in getting there.
The logic behind this process is quite obvious: when you are about to mass-
produce something, you better make sure it is the right thing you replicate over
and over.
The logic of lean and agile processes is quite different. While they too are
about resolving uncertainty by means of iteration, they are not at all oriented
toward this kind of optimization. Instead they orient toward notions such as the
Minimum Viable Product. It is no longer about first finding the best possible
solution and then moving to market, but rather to make the entire development
process into a dialogue with market: by constantly releasing updates and making
extensive use of user data, progress is tracked and measured, to form the basis
for decisions about how to make the next move. On the user side, there is for
instance frequent use of A/B testing: by releasing two different versions (hence
the A and B) to different groups of users and then tracking which version yields
the best response or result, design decisions can be based on actual data. And by
rolling out such tests extremely frequently, perhaps even several times a day, the
distance between designing and using can be shortened dramatically compared
to when in the past it might be even years between when the designing first
happened and people could actually use the thing.
To see the vast difference between the design philosophy behind notions such
as the Minimum Viable Product and what ideas have historically governed
industrial design, compare the following two reflections: the first from Eric Ries,
one of the main proponents of lean methods for start-ups; the second from Max
Bill, first student at the Bauhaus and then later Headmaster at the HfG Ulm, the
school that introduced the industrial design aesthetics that still dominates much
technology design:
At this point in our careers, my cofounders and I are determined to make new mistakes. We do
everything wrong: instead of spending years perfecting our technology, we build a minimum viable
product, an early product that is terrible, full of bugs and crash-your-computer-yes-really stability
problems. Then we ship it to customers before it's ready. And we charge money for it. After securing
initial customers, we change the product constantly–much too fast by traditional standards–shipping
new versions of our product dozens of times every single day.

We really did have customers in those early days–true visionary early adopters–and we often talked to
them and asked for their feedback. But we emphatically did not do what they said. We viewed their
input as only one source of information about our product and overall vision. In fact, we were much
more likely to run experiments on our customers than we were to cater to their whims. (Ries 2011, 3–
4)

For around a hundred years now the call to action has sounded in successive ways throughout the
world: we have a duty to make useful, ethical products that are true to materials and manufactured
under socially responsible conditions, using the best means available to us. Inherent within this call is
a sense of moral responsibility, a social understanding.

/ … /

This makes us realise that what we’re actually striving for is something quite different—namely, an
extreme utilisation of materials, where the maximum effect is achieved with the minimum of materials.
For example, we can construct a tower 300 metres high (the Eiffel Tower) and make it so light (as
Eiffel did) that if its height were reduced by a factor of one thousandth, ie by 30 centimetres, then its
weight would drop by just seven grams—the weight of a pencil. This is a shining exemplar of the
extreme exploitation of materials, an emblem for the technical age and the rational use of materials, as
well as the germ of a new ideal of beauty. (Bill 2010a, 32–33)

These reflections come from different places and points in time, and one should
be careful when comparing them. Still, they illustrate a couple of key points
made here regarding design’s relation to the technologies it is working with.
While the notion of making the most out of existing resources is a central
concern, how to actually achieve that has fundamentally changed as people are
now trying to cope with a different kind of complexity. While actual practice
may not be quite as extreme as Ries’s account mentioned previously (after all,
many companies that work with lean processes also have to comply with other
requirements and certifications that prevent them from releasing things that are
not functional), it points to a significant shift from refining something until it
becomes the optimal design solution to instead trying to expose minimal
products to a market as quickly as possible. In a sense, this means that the
former so important difference between prototype and product is disappearing.
As such, the shifts in perspectives, not to mention values and objectives, also
illustrate the radically different configuration of production and consumption
that these new technologies and associated industries have introduced.
Compared to how things used to be, it may seem as if we live in a constant
“beta,” but more than anything, what we see here is a completely different
configuration of the relations between designing and using.

Design aesthetics
As we turn to how design addresses human experience, to how these things
present themselves to us and become part of our practices, it is less obvious that
we have developed design to care for this condition. On the contrary, much of
what is going on over here seems to be about reassuring us that things have not
changed that much after all. Indeed, we may ask: how is it possible for these
complex assemblages to be experienced as “things,” even as simple things that
we do not hesitate to let into our lives? To get started, we have a working
hypothesis regarding how fluid assemblages come to present themselves to us as
stable, contained, and predictable things: Fluid assemblages make use of
mimicry: to appear as “things,” they camouflage themselves as totalities.
The distinction between exteriority and interiority relations allows us to probe
deeper into how the “thingness” of these assemblages is achieved. To see this,
we first need to attend to how design typically has come to approach the relation
between intended use and aesthetics. The perhaps primary task of industrial
design is to bring about a meaningful whole. Design is, in this sense, the
opposite of analysis: whereas analysis treats complexity by means of taking
apart into ever more manageable parts, design is fundamentally propositional in
how it aims to resolve conflicts by iteratively prototyping and proposing how a
unified whole could be brought together. Much of what is treated in design
aesthetics (as distinct from in Aesthetics in the philosophical sense) is about how
to create such totalities, often articulated in terms such as consistency and
coherence, as gestalt or as a unitary and consistent use experience.
While it may be obvious that design is about creating the conditions for
emergence in that sense—of making the sum greater than the parts (and the
greater the difference, the greater the value!)—we may still ask whether this is
really about seeking totality. Certainly, this question is one of nuance and
difference, but if we turn to the present and pressing issues of sustainable
development we get an indication. Had design’s notion of a whole been
fundamentally oriented toward assemblages, then we would expect there to be an
awareness of how the constituent components retain integrity after the
assemblage is taken apart. Should this have been the case, new notions such as
“recycling” would not have been needed, as taking things apart and using them
for other assemblages would have been an obvious consideration.1
On the contrary, it seems that as technology matures, it moves toward
totalities. Working with the first computers, users had to know how to replace a
radio tube and other parts that might break; users of contemporary computers
may not even be able to replace the battery. An old car will expose most parts of
its engine, mechanics, etc. as you open the hood; in a contemporary car most of
it is likely to be covered by plastic to indicate that this is the domain of the
experts in the workshop (who in turn typically rely on complex computational
diagnostic equipment to uncover what is not working).
Certainly, there is modular design, open design, and a range of other
approaches that genuinely engage with design as assemblage—but overall, it is
probably fair to say that design aesthetics have a basic orientation toward
meaningful wholes understood as totalities rather than assemblages. To a
significant degree, this design principle put forward by Max Bill in “The basis
and aim of aesthetics in the machine age” from 1953 (thus around the time when
he took on the position as headmaster of HfG Ulm) still holds for much
industrial design:
The basis of all production should be to fulfil, as a unity, the totality of all functions, including the
aesthetic functions of an object. (Bill 2010b, 70)

It is worth a short reflection upon why this was, for the industrial technology
of the time, a very reasonable and effective principle. The emergence of
industrial design is completely conditioned by the shift from manual to machine
production. Giving form to things prior to this shift was a matter of craft, most of
the time working with just one or a few materials, and with objects that typically
would have one or maybe a few key functions. Importantly, these things were
also most of the time made by just one or a few people who would be able to
lead the process from material to finished object, thus able to make sure that the
outcome was a “whole,” and not just a collection of parts coming out of
otherwise separate processes. With industrial production, this changed. Indeed, it
is called assembly line for a reason. And further, industrial design involves
working with many different materials, larger sets of functions, and over time
therefore also increasingly complex things. In this context, to achieve something
that truly comes together in the sense that the traditional object is given form and
made by craft is far from trivial. Thus, how to achieve such a seamless totality
became the perhaps most central aesthetical issue when trying to find ways for
the “new” to not only (unsuccessfully) mimic the “old,” but to offer something
that would, literally, transcend it.
The once intended effect, but now increasingly problematic consequence, is
therefore that industrial design typically lets interiority relations completely
dominate over exteriority relations when thinking about how form defines the
way material builds things. In this way, both designers and users become
oriented toward things not only as “meaningful wholes,” but as totalities. Not
only in the sense of how they are made and built, but perhaps even more so in
terms of what kind of use experience designers aim for and what kind of
experience users typically expect and appreciate.
This disposition has three significant implications with respect to following
established aesthetics when designing fluid assemblages. First, it means that we
as designers will seek to, literally, transcend the experience of an assemblage and
push it toward the experience of a totality. For instance, this means we will try to
more or less mask exteriority relations, and instead emphasize how the
components have fused into a new seamless whole. Second, it means that we as
users feel most confident when we only need to attend to a whole. Most
certainly, we do not want to pay too much attention to all the interactions taking
place between various components. While we might realize it is an assemblage,
we certainly do not want a huge part in its immanence. On the contrary, we
prefer if it is stable and unified, and it is only rarely that we will be interested in
committing to the effort required to maintain and manage complex interactions
between constituent parts. These two implications bring about the third, and
most critical one, namely that we therefore come unprepared and unaware of just
how different a fluid assemblage is from a traditional object—and that our
current ways of working with design expressions to a significant degree keeps
this hidden from us. In other words, there is presently a rapidly growing design
space that designers do not quite know how to deal with in terms of aesthetics, or
perhaps even care about from an aesthetical point of view.
So, we find ourselves in a situation where we on one hand have technologies
and the beginnings of new design methodologies for making fluid assemblages,
but where we at the same time persist in making them appear as normal “things.”
Like Theseus’s ship, our new things are both the same and something very
different.
 Note
1 Just consider open source software as a counterexample: certainly one could think of it as a large
recycling center as pieces of code are made available to be used to other things, but you rarely see this
articulated as a matter of “recycling”—it is quite simply about sharing and contributing code for also
others to use, and this relation to the code produced is an inherent part of its making.
 5

Things for Us

There can be seen two main orientations to considering things.1 The first, and
most common, is to look at them in relation to human activities in which they are
involved. In other words, the basic concern here is what a particular thing is for
humans as they use or otherwise relate to or are affected by it. The second is to
inquire into what things are “in themselves.” Of course we can never really
escape our own situated perspectives as humans, but we can try to get to the
bottom of things through an investigation more oriented toward basic ontology.
These orientations have through history generated a vast range of
contributions that can be productively brought to bear in making sense of
contemporary things. While we may realize that we need to build on work from
both strands, as one of the things that we discover when exploring fluid
assemblages is that a framing of human (or “user”) experience is not sufficient
for understanding what they really are and do, we may still believe that such a
combination of multiple perspectives will, in the end, result in a quite thorough
coverage. Yet this is a problematic assumption. Different analytic frames have
conceptions of their objects of study that can actually be incommensurable,
resting on different foundational assumptions and orientations. As Deleuze and
Guattari state: “A concept always has components that can prevent the
appearance of another concept or, on the contrary, that can themselves appear
only at the cost of the disappearance of other concepts” (Deleuze and Guattari
1994, 31). Foregrounding certain aspects inevitably means backgrounding
others. Another concern is that perspectives honed on more traditional objects of
study will miss or not properly account for much of what is relevant about
contemporary ones.
There are also more fundamental reasons to care about the consequences of
the conceptual frames used. Building on the “philosophy-physics” of Niels Bohr,
Karen Barad goes even further in her agential realist account. As she states:
Discursive practices are specific material (re)configurings of the world through which local
determinations of boundaries, properties, and meanings are differentially enacted. That is, discursive
practices are ongoing agential intra-actions of the world through which local determinacy is enacted
within the phenomena produced. (Barad 2007, 820–821)

In other words, discursive practices do not merely describe phenomena, but

participate in their production. This foundational role of conceptual and
discursive frameworks thus entails a corresponding accountability for the ways
in which we conceptualize, analyze, and discuss the things and phenomena with
which we engage.
Proceeding with care, then, there are two main themes we need to pay
attention to in considering existing and potential ways of trying to understand
changing things. The first is to see what existing bodies of literature give us:
tools, perspectives, sensitivities, and so on. The second is that we need to be
careful to notice aspects of contemporary things that existing perspectives do not
handle well: significant characteristics and dynamics that are not accounted for,
conceptual tools that are now clunky rather than incisive, underlying
assumptions that are no longer valid, black boxes that have become too large and
conceal too much, and basic orientations that point away from where much of
the action is.
In this chapter, we will focus on general orientations to things as they are
involved in human activities: action and perception, communication, and
networked computation. The latter, however, starts to move into a more explicit
focus on things, and is essential for making sense of fluid assemblages. This
includes scaling, looking at the range of aspects that can be considered from the
level of artifacts to those of systems and social implications. In the following
chapter, we will then turn to issues of conceptualizing things in themselves that
are relevant for understanding fluid assemblages, how they are both similar to
and different from other kinds of things.
The purpose of the overview presented here is twofold. First, we want to
illuminate various aspects of things by reading them through activities in which
they are implicated—in other words, what they are for humans in particular
kinds of use. This also requires engaging with scholarship in different disciplines
that investigate these domains of human activity. In doing so, there will always
be a sort of tension or torque between the primary concerns of various bodies of
scholarship and the ways in which we engage them, since we are interested in
what they have to say about things in particular. For some areas, particularly
those in the social sciences, this actually goes against some of their primary
commitments related to respecting the primacy of lived human experience and
praxis. While we are sympathetic to these commitments, and share them in many
ways, we also see the need to get to a more neutral ground that does not rely on
human experience and existing social practice as analytic frame, since there is
much of concern that is not accessible experientially during normal use. Thus,
while we recognize the fundamental entanglement of experience, practice, and
things (or materiality), we approach this entanglement through the lens of things
—what they are and what they do.
The second purpose is in a sense the reverse: to point to the multiple complex
dynamics in which things are implicated, and to expand the types of issues that
are brought into focus when considering their design and configuration. The first
purpose can be seen as a spotlight shone on particular aspects of things from
perspectives that consider many more moving parts, such that they account for
only a small part of what things are and, conversely, things are only a small part
of what these perspectives consider. This second purpose can be seen as
beginning with things as wholes, and recognizing the broader dynamics and
issues in which they are implicated. Rather than reflecting limited applicability,
these touchpoints provide openings onto much larger sets of concerns and serve
as connections that can be followed to explore those vantage points. These are
two sides of the same coin, and reflect the holistic designerly perspective we
take here. Rather than reducing or selecting piecemeal from rich areas of
scholarship, we mean to expand things (specifically fluid assemblages) as both a
lens of its own and—crucially—a site for inquiry and intervention. Moreover,
addressing multiple sets of dynamics and their interactions is necessary when
they are all present in things.
It should also be noted that this investigation in many ways reverses the usual
order when it comes to considering things and human activities. Typically, a
certain domain or aspect of human activity is under investigation, and things are
considered in the sense and to the extent that they are involved or implicated in
that activity. The activity or matter of concern is in focus, and things are one of
the moving parts that must be accounted for. This also means that things are
accounted for in terms of the human activity that is in question. The illumination
of things that is achieved comes from a spotlight shone from the perspective of
one particular set of concerns. However, things clearly can have multiple roles,
identities, and functions. Understanding their character and consequences in the
world in general, then, and in a more holistic way requires considering the
multiple roles and identities they can have. Thus, the notion of a “general”
understanding of things that we aim for here is of the Aristotelian rather than
Platonic kind: we aim to reach toward the general by considering many different
particulars, like understanding what a chair in general is by encountering and
sitting in many particular instances of chairs, as there is not one universal
instance, expression, account, or articulation present.
This survey will inevitably be reductive in some ways, and will not quite do
justice to the rich areas of scholarship referenced. Yet it is also necessary to get
out of the thick of that richness and zoom out to a higher level in order to see
something else that is more to do with general orientations, assumptions, and
perspectives. Another key aspect of our investigation is that it is about
connections, synergies, and tensions among different conceptual frameworks.
So, our often necessarily cursory overviews can be seen as pointers toward paths
leading to other areas that are potentially worth exploring in more depth, while
our primary mission here is to create a higher-level map of a larger territory and
to trace a web of connections among some of its main features.

Tool—Action and perception

Perhaps one of the most basic ways of relating intentionally to things is to see
them as tools that can be used for some specific purpose. Tool use might even be
one of the things that makes us human (Nelson and Stolterman 2012), although
other animals have also been observed using tools. From the time a human
picked up a bone and turned it into a weapon or other implement, humans have
been appropriating things for their own ends and using the materials they can
find to build better tools. In other words, they have been designing. In the craft
tradition that preceded professional industrial design, there was a close
interrelationship of designing, building, and using, and organic connection
among materials, places, and contexts of use. This can be seen, for example, in
Jones’s (1992) discussion of craft methods, and example of the craft process in
the evolution of wagon making. Fast forwarding from this early tool use, the
Industrial Revolution called industrial design into being in service of mass
production (Dilnot 2014). Whereas things had previously been crafted for
particular uses, the challenge became one of designing for use in a more general
way in order to support mass production of uniform products. The perspective of
things as tools is, then, quite old and quite common, found in history,
philosophies, and practical technological development. It also generally makes
sense, resonating with many of our own ordinary experiences of relating to
things. Tools operate in the domain of the practical.
The activities most prominently associated with this perspective of thing as
tools are action and perception: what people can do, achieve, and perceive. As
Peter-Paul Verbeek, a philosopher working in the postphenomenology subfield
of philosophy of technology, states: “Technology mediates our behavior and our
perception, and thereby actively shapes subjectivity and objectivity: the ways in
which we are present in our world and the world is present to us” (Verbeek 2005,
203). In Verbeek’s development of postphenomenology in particular there is an
emphasis on not only how the world comes to us, but also how humans are
themselves constituted through relations. Yet at the same time, in
postphenomenology and more generally within what might be referred to
broadly as the “tool” perspective, there tend to be basic assumptions about
human agency. Humans are at the center of the picture, determining what
matters in relation to their needs and desires and generally driving the action
even as their relations with things affect them as well. It is still humans who
choose whether or not to look at the ultrasound image (Verbeek 2008), to use the
system, or to purchase the product and “domesticate” it such that it becomes a
meaningful part of their lives.
Corresponding to this focus on human agency is a parallel focus on human
subjectivity. This is to say that it is human experience that matters, and the
aspects of things that are considered relevant are those that enter into human
experience. This orientation is indeed at the very foundations of
phenomenological method, which considers the world as it appears to us.
Something that exists in the world but does not appear or relate to us does not
enter the picture. This might seem logical in some ways, but when it comes to
contemporary computational things it gets to be more problematic as there is no
inherent or necessary connection between a thing’s function and how it appears
during use. This is another issue to which we will return later on.
Another basic orientation within this perspective is that technological tools
somehow deal with the real. Of course this cannot be the real in any direct or
unproblematic sense, especially from a philosophical perspective. Yet in
postphenomenology in particular there is a sense, particularly through the
connection to scientific instruments and praxis, that what is at stake has to do
with how people can come to connect to and make sense of what is real. And
how could the world and one’s connection to it feel any more real or substantive
than it does when hitting something with a hammer? A hammer evokes a
reassuring solidity and simplicity of action and reaction that helps anchor even
Heidegger’s lofty ontological discussion (Heidegger [1927] 2010). Tools, at
bottom, are imminently practical, handy, goal-oriented things, and this heritage
can be seen in even much more sophisticated technological tools. While this
might seem completely straightforward and unproblematic, it poses a contrast to
some of the basic assumptions and orientations around things as media for
communication that will be discussed later.
Tools are used for specific purposes based on the particular functions that they
have. To design things as tools means to think about intended use and purpose. It
also involves considering people who will do the using—the “users”—and
contexts of use. Indeed, understanding eventual users and contexts of use is the
focus of user-centered design, which seeks to design products (and now also
increasingly services) based on human needs, desires, preferences, and ways of
working. In this perspective, things have their identity in relation to the goals
they help humans to achieve. The field of human–computer interaction (HCI)
began with attempts to configure the interfaces of computational systems such
that it would be possible for humans (more specifically, human brains, as bodies
were not so much in the picture at this stage) to understand their functions in
order to interact with them successfully in order to accomplish given goals. Lucy
Suchman’s (2007) groundbreaking work showed that actual in situ human use is
typically less logically planned than systems developers tended to assume, yet it
also maintained a focus on instrumental use of technology. In short, there seems
to be a common understanding that technological tools exist to help humans get
stuff done.
Ensuring a good fit between product and use practices means conceiving in
quite fine detail and making assumptions about the scenarios in which it will be
used. These intended scenarios of use have been referred to as “scripts.”
However, these can be “de-scripted” in the real world of actual use, where
people can use things in quite different ways than their designers originally
intended (Akrich 1991). However, even as it has been recognized that users
frequently do not behave in relation to products in ways that designers intend,
the general framing of how people relate to things is still in terms of what they
do with them. There are very significant semiotic aspects to be sure, and
(especially high-end) products often try to both sell and signify a certain lifestyle
more than the products themselves. Yet at the same time the reason this is
possible is to a large extent that the product implies certain practices of use,
which in turn are part of a certain kind of lifestyle.
Technological tools might be commonly thought of as relatively
straightforward, with any notable complexity coming in the form of
technological advances that render them ever more sophisticated. Yet, as the
more detailed concepts and examples in the following chapters will hopefully
show, there is much more to tools than engineering, making, and use. Tools that
help us perceive and act in the world inevitably shape the character of that action
and perception. The ways in which we take up with the world are often mediated
by technologies, a human–world interface at least as important as the interface
that enables interaction with a device. Technologies are often manifestations of
larger systems and ecosystems that govern the techniques used in administration
in a society at larger-than-individual scales, leading to particular configurations
of power and collective forms of life. While things that are tools operate in the
domain of the practical, it is important to remember that the practical is about
everyday praxis—about the ways in which things are actually done. And
questioning regarding the ways in which things are—and perhaps should be—
done cuts to the very heart of issues related to that most basic of political
questions: what it means to live a good life, together.

Media—Communication
Many of the contemporary digital networked things that now pervade everyday
life can be seen as media technologies. That is to say that they are involved in
the creation, modification, sharing, and distribution of information, content, and
messages that are part of social practices, rituals of communication, and the
many data-intensive and information-based processes that run and govern our
contemporary societies from local to global levels. This lens of information and
communication and perspectives from the disciplinary fields of information,
communication, and media studies bring into focus important aspects of
contemporary things. Although information and communication technologies
can frequently be seen as tools as well, and certainly have a material dimension
that can put them also in the category of stuff, this lens brings into focus their
operation in the domain of the symbolic.
Media are involved in processes of information and communication. Although
closely related, there are, at least for our purposes, a couple of main foundational
issues associated with each. Information deals with underlying representational
practices and units of meaning. It is the lifeblood that flows through digital
networked technologies that are used for information and communication. Thus,
while issues of representation and information are relevant in consideration of
any kind of communication and information processing activity, they are
absolutely key when it comes to considering contemporary technologies and
their uses. Different technological configurations represent, transmit, store, and
process data and information in ways that vary, but share commonalities through
underlying elements and infrastructures. We thus need to consider what
information is and does in these systems.2 Communication is a social practice,
and communication technologies with different capabilities enable new types of
communication practices that develop around them and in relation to existing
practices. It is here that we run squarely up against the social, and it is typically
the perspective of social and cultural practices that is used when considering
communication technologies. Thus, this is one of the places where the torque
between our perspective and that of the work that we consider might be most
keenly felt.
First, we need to consider what is involved in (mediated) communication in
terms of its most basic components. Perhaps the most obvious place to start a
consideration of things in relation to communication is to think of the invention
of key media technologies—actual technical apparatuses. Beginning with
language as the most fundamental communication technology, relying on the
bodily apparatus of vocal chords and so on, we can then think of the inventions
of written language and recording media (clay, papyrus, paper), printing press,
telegraph, radio, phonograph, film, telephone, broadcast television, and others.
More recently the internet and World Wide Web have brought many more
possibilities, from email and chat to virtual environments to many forms of
social and collaborative media.
Now, each of these media technologies involves slightly different
technological capabilities and communication protocols. Importantly, each
involves some particular form of representational practice by which meaning
takes on a symbolic and material form that requires interpretation on the part of
the receiver in order to render it meaningful. This is a hermeneutic practice that
involves making sense out of both the content and the technical and social
protocols and norms involved with a particular medium.
Furthermore, different media forms contain other media. Or, as media theorist
Marshall McLuhan famously stated: “The ‘content’ of any medium is always
another medium. The content of writing is speech, just as the written word is the
content of print, and print is the content of the telegraph” (McLuhan 1964, 8).
More recently, Bolter and Grusin have argued that the representation of one
medium in another—what they call remediation—is a “defining characteristic of
new digital media” (Bolter and Grusin 2000, 45). They also identify two
associated characteristics: immediacy, or the desire for the medium to disappear
altogether to allow for the feeling of direct access to what is mediated; and at the
other extreme hypermediacy, which instead draws attention to the medium and
its multiplicity. Indeed, especially at the time their book was written during the
late 1990s, the fashion did seem to push toward the excesses of possibilities
enabled by new media. This phenomenon is perhaps most easily called to mind
through the form of a television news program in which one frame commonly
contains multiple windows, scrolling tickers, animated graphics, and more.
Returning to communication as such, it can be said that there is never any
such thing as completely direct, unproblematic, or unmediated communication.
At a fundamental level, communication can be seen as an attempt at reconciling
self and other (Peters 2012), a project that is never perfectly achievable. There
will always, even in the “simplest” face-to-face interaction, be glitches,
misunderstandings, and breakdowns that must be negotiated. Representation,
interpretation, and the impossibility of perfect understanding are basic conditions
of communication—even those that do not involve advanced technologies. And
of course when technologies are involved there are additional related
complexities, and sources of breakdown.
There have been different positions regarding the role and significance of
media technologies themselves. Outlining first two extremes, the mathematical
model of communication developed by Claude Shannon and Warren Weaver in
the heyday of cybernetics during the time of the Cold War saw technology as the
channel for information transmission (Shannon and Weaver 1959). In their
influential model, a signal is encoded by a sender into a message that is sent
through a channel and decoded on the other end by a receiver. In this view the
channel is relatively neutral as the message, at least ideally, simply passes
through it, although it is always threatened by noise that could degrade it. On the
other hand, Marshall McLuhan famously asserted that “the medium is the
message” (McLuhan 1964); in other words, a medium is not a neutral channel
that just conveys a message more or less effectively, but is rather itself a
significant (or even the primary) message.
Concerns about the role and effectiveness of the medium also dominated early
work in the field of computer-mediated communication (CMC), which
considered whether the text-heavy medium of the computer was “rich” enough
to enable effective human (or, perhaps better stated, humane) communication.
On the other hand, it was also the disconnect between information that could be
transmitted over the internet and the “richness” of a person’s physical body and
context that enabled the free play of identity performance that characterized
much early online interaction (Turkle 1997; Hayles 2008).
As is typically the case with sweeping claims tinged to greater or lesser
degrees with technological determinism, the reality is more complex than any of
the previously mentioned perspectives, even as they have each helped to
articulate certain issues and dynamics. Media technologies and the capabilities
they provide play a significant role, but communicative practices in which they
are used are always contingent, culturally and historically embedded affairs.
Pointers to various dynamics involved can be found in Lisa Gitelman’s
comprehensive definition of media as “socially realized structures of
communication, where structures include both technological forms and their
associated protocols, and where communication is a cultural practice, a ritualized
collocation of different people on the same mental map, sharing or engaged with
popular ontologies of representation” (Gitelman 2006, 7).
One key aspect of communication as a cultural practice is that it involves
performance of self. This is the case also in face-to-face communication, in
which the persona that one presents to others is a performance that does not
necessarily match what is going on “backstage,” to use Goffman’s (1959)
influential dramaturgical metaphor. Different media technologies allow for
different types of performances. The early online environments allowed for very
carefully crafted performances of self that could be very different from those “in
real life” (Turkle 1997). Now in the age of social media there are generally much
stronger connections between “real life” and “virtual” interactions, with
interactions crossing over between various media, and between online and
offline. In fact, an analytic and practical distinction between online and offline
has become untenable in a situation of constant crossover between online and
offline, “virtual” and “real” interactions that Coleman (Coleman 2011) has
termed “X-reality.” Connection between online profiles and offline identity has
also come to be legally required through terms of service agreements, such as
those of Facebook, that explicitly prohibit using a false identity. Or conversely, a
verified connection to an offline identity can be rewarded, as in the case of
Twitter’s “verified identity” system and associated badge prominently displayed
along with verified accounts. Rather than playing with different identities online,
it is now seemingly more common to curate one’s personal brand in a way that
maintains some level of consistency across social media channels. And brands in
the more traditional corporate sense have very carefully managed social media
presence, and run sophisticated search engine and social media marketing
campaigns.
With this very broad background in place, we can now already point to a few
key differences between the concepts of media and tools (discussed in the first
section). Although these perspectives may be used in relation to the same
ostensive objects of study, they come with different kinds of underlying
assumptions. While tools are generally considered in terms of more or less
practical goals and effects in relation to practices of use, media are involved in
culturally situated practices of communication and practices of making and
interpreting meanings that can never be singular or conclusive. There are always
gaps between sender and receiver, world and representation, reality and model
that cannot be perfectly bridged. These gaps thus become key sites for
investigation, particularly as more and more of our social realities and access to
them are brought into the realm of the computational.

Computer—Data processing
Computation can in one sense perhaps be seen as quite different from media,
having more affinity with mathematical logic, science, and engineering than
culturally situated practices of communication and meaning-making. Yet the
ways in which computation exists in the world are very much based not only on
these types of practices, involving issues of representation, interpretation, and
culture in its many applications, but also on its most fundamental levels
involving data, algorithms, and protocols. Computation is a tool for data
processing, a medium of information.
The roots of computation are contained in the word itself, from a time in
which “computers” were people who performed computations. The use of
computers to perform complex calculations for scientific purposes is certainly
still present, as is their use in increasingly sophisticated military applications
(think, for example, of drone warfare or battlefield robots). Another major theme
of information management and access was eloquently prefigured by Vannevar
Bush, who had in his capacity as director of the Office of Scientific Research
and Development in the United States coordinated the effort to apply scientific
knowledge to warfare during the Second World War. Writing in an article in The
Atlantic magazine at the end of that war in 1945, he suggested that the energies
and achievements that had been marshaled so effectively for the cause of war
should now be channeled toward other worthy objectives during peacetime. The
problem he identified was that of an overwhelming amount of information that
was becoming increasingly difficult for scientists to stay on top of, particularly
as scientific specialization increased. And there was a need not only to store
information, but also, importantly, to make it accessible in a way that supports
how people think, namely, by association. What he envisioned as one possible
solution seems remarkably prescient:
Consider a future device for individual use, which is a sort of mechanized private file and library. It
needs a name, and, to coin one at random, “memex” will do. A memex is a device in which an
individual stores all his books, records, and communications, and which is mechanized so that it may
be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.
(Bush 1945)

Since storing files of all kinds on computers, external storage devices, and
networked storage services is now a normal part of life, as is googling for all
imaginable types of information, it is perhaps difficult for many of us to imagine
a time in which these types of structures were not in place (or were much more
rudimentary). Yet it is possible to see how computational systems are still being
developed in order to serve this purpose, from automated organization
techniques to ever “smarter” access to contextually relevant information.
One particular arena in which information became increasingly important was
in military intelligence. During the Cold War, information came to be seen as an
end in itself (Ceruzzi 2003), foreshadowing more contemporary post-2001
programs of massive surveillance that have operated on the general principle of
collecting as much data as possible. It is thus always worth being mindful of this
when considering computational things that generate information that could be
seen as “intelligence” and/or used for control in some context.
Another trajectory has its roots in the American counterculture of the 1960s
and 1970s that came to expression in the burgeoning cyberculture of the 1990s
as a group of hippie entrepreneurs (including the founders of the Apple computer
company and Wired magazine) reimagined computers as tools that could be used
to serve the utopian ideals of personal liberation, expression, and community
building rather than warfare (Turner 2010). The first commercial sites on the
internet had some of the communal spirit of the early online communities, seen
in features such as reviews (and reviews of reviews) that allowed for policing of
sites (such as Amazon and eBay) that would not have been possible in a top-
down way; and this community element in fact became essential for such sites to
survive (Ceruzzi 2003). Now, it seems that every computational device and
application is in some way “social.” In this sense, the invitation or even
imperative to share with others can be seen as a heritage of both the communal
spirit of early cyberculture and the drive of both governments and businesses to
be able to find out what people are up to with ever-increasing speed and
precision.
It is also worth considering some of the basic aspects of computational
architecture, in which we can see that computers have always been assemblages
in a fundamental sense. Early computer models consisted of a central mainframe
computer with “dumb” terminals that were used to access it and perform
computations using a time-sharing model. The processor and display were
clearly separated. In fact, computer historian Paul Ceruzzi states that a “systems
approach” to studying the history of computing is not particularly helpful
because a modern computer “is a system: an arrangement of hardware and
software in hierarchical layers” (Ceruzzi 2003, 4). And again:
 The word “software” suggests that there is a single entity, separate from the computer’s hardware, that
works with the hardware to solve a problem. In fact, there is no such single entity. A computer system
is like an onion, with many distinct layers of software over a hardware core. Even at the center—the
level of the central processor—there is no clear distinction: computer chips carrying “microcode”
direct other chips to perform the processor’s most basic operations. Engineers call these codes
“firmware,” a term that suggests the blurred distinction. (Ceruzzi 2003, 80)

Another way in which the assemblage character of computers can be seen is in

the very concept of an “assembler,” an early program that assembled another
program into machine instructions that the computer could execute. It was also
actually a milestone for computation when instructions and data were stored in
the same place in stored programs (Ceruzzi 2003). This was also the milestone
that turned computers into things with agency of their own, not just sitting
waiting for instructions from elsewhere but potentially running programs that
allowed them to respond to external events. The programmability of computers
is what allows the flexibility for them to be turned into such a vast and diverse
array of devices, and it is the characteristic that effectively defines modern
computing (Ensmenger 2012). Yet even with this flexibility and assemblage
character, computational systems have also always needed to appear solid and
unified during use. One basic example of this was file storage systems developed
in early personal computers, where files were stored in fragments on the disk but
represented as single files for the user; the operating system was responsible for
storing the data and retrieving and reassembling the files when needed (Ceruzzi
2003). This was even to some extent visible in consumer defragmentation
software, which provided a visual representation of system memory as it was
optimized by the program.
From a design perspective, working with what might be called digital
materials implies a quite different set of actions and implications than working
with and giving form to physical materials. Using digital materials does not use
them up, as with physical ones, but rather establishes relations and conditions
under which they are activated. These are relations that persist and need to be
maintained throughout the life of a digital product. Starting with electricity, this
extends to the connections between hardware and software components, pieces
of code (particularly apparent in object-oriented programming languages), and
network connections to other resources and devices through the internet and
World Wide Web or other more near-range protocols such as Bluetooth or RFID.
These relations can also extend to those between devices in what can be referred
to as ubiquitous computing, leading to complexity and emergent effects of a
different order (Coyne 2010; Ekman et al. 2015).
 Roles, relations, agencies
At this point we have covered quite a bit of ground in considering things as
tools, media, and computers. Although this has already been at a quite high
level, especially in relation to the depth at which all of the issues and dynamics
we have referenced can be explored, we can already at this point extend the
previous analysis by recognizing the possibilities for variation in the distribution
of roles, relations, and agencies that are implied in relation to tools, media, and
computers.
To begin, it is possible to note a simple yet significant distinction between
what might be called active and passive roles. When a thing is used as a tool to
mediate action or perception, a person may be on the receiving end of that action
or perception just as well as at the active end. This can be referred to simply as a
distinction between subject and object (with no intention whatsoever of
returning to the tired distinction between these in an absolute sense!). A
telescope or binoculars can be trained on a distant human target rather than the
heavens; medical imaging equipment can be used to reveal the inside of a
person’s body; a read receipt can reveal that someone else opened a message;
and text created with technology (as in fact all are) can be interpreted in terms of
what it says on behalf of and/or about its author. This mediational structure in
which human activity is in focus as what is revealed in fact happens all the time
whenever someone accesses a web page or uses really any kind of fluid
assemblage.
It is possible to make a similar distinction between sender and receiver of
messages and content, or between producer and consumer, author or reader of
media content. The leveling of opportunities to be in both of these roles is in fact
one of the most hyped, discussed, and debated characteristics of media that have
been variously described as new (Bolter and Grusin 2000; Manovich 2001),
connective (van Dijck 2013), convergent (Jenkins 2006; Meikle and Young
2012), collaborative (Löwgren and Reimer 2013), social, and so on.
Whereas a few decades ago the media landscape was dominated by a few big
industry players that controlled the means of production and distribution, and
were thus the big producers for mass consumption, now anyone with a
smartphone and internet connection can be a producer and distributor. So far so
good, as there is already a quite high level of awareness of the various roles
people can take in relation to media, even if the social implications of this state
of affairs are not as obvious or straightforward all the time.
However, this situation is very different when it comes to producing and
consuming data, or rather, it is perhaps more similar to the dynamics of media a
few decades ago. That is to say that while all of us end up producing data
through our ordinary activities, the consolidation, processing, and use of all of
the data generated happens more at a government and industry level rather than
at an individual one. This is not to say that such large-scale data is not made
available to individuals, since it often is in various ways. Even something as
simple as hashtags that are trending on Twitter is one such example (even as the
exact mechanisms behind this are opaque). Yet there remains a significant
disparity between the capabilities of large-scale corporate or governmental
entities who control the means of producing data versus those about whom the
data is produced and those who wish to study it (such as academics).
Now we have looked at some lenses on things in relation to the roles they play
in human experience and affairs. While each of them highlights significant
aspects, it is also possible to see that none of them can fully account for what
things are. Since we are trying to get to the bottom of what (digital, networked,
computational) things are, the next logical question is something like: Well
alright then, what are these things really, if we try to get away from these more
focused lenses and instead look at what is actually there, in some kind of more
original sense? This question takes us into the realm of ontology.

Notes
1 An earlier version of part of this argument appeared in Wiltse (2017).
2 It should perhaps be emphasized that we will focus here only on data/information as it exists practically
in these systems, and not on the more foundational and also important aspects of these concepts in
themselves. For more on the philosophy of information, see, for example, the work of philosopher
Luciano Floridi (Floridi 2011).
 6

Things in Themselves

Turning now to more basic ontological issues, we are not concerned here with
ontology and the nature of being in general, but rather the nature of being of
things that are fluid assemblages. However, as we loosen our grip on existing
lenses in order to try to see what is there in a more fundamental sense, there are a
few considerations of more basic ontologies that will be useful. Because we are,
after all, still inquiring into what is, even if only a particular subset.
Two of the most basic things that need to be accounted for when it comes to
the being of things are identity and change. In other words, how can something
be delineated as a this thing rather than as part of an undifferentiated whole?
What things are crucial for making that differentiation, and not just “accidental”?
And on the other hand, how is it possible to account for change, the fact that
things are not pure, stable, essential forms but rather changing their forms and
relations over time? Also, when addressing fluid assemblages we explicitly need
to account for these: on the one hand we need to account for their solidity, the
fact that they are actual entities that exist and do things in the world; on the other
hand, we need to account for the fact that they are in flux, and that any solidity is
only temporary.
Things can also be investigated at different scales, bringing into focus
dynamics from the very local and concrete up to the level of systems. Just as we
suggested in the earlier section that a basic understanding and awareness of
differences between general orientations to considering the role of things in
human activities can be useful, especially when working with multiple
perspectives, we here suggest that sensitivity to scale is important for similar
reasons.
Issues of structure at different scales can be seen in the interplay of two key
orientations that are foundational for a design approach: concern with
particulars and systemics (Nelson and Stolterman 2012), and the ways in which
they relate to each other. Much of design deals with the details of particulars,
both in the concrete formal and aesthetic expressions of things and in their
integration in specific contexts and practices of use. Some types of design also
deal explicitly with systems, and there are systems-oriented perspectives that
will be discussed in each section. However, it is important to recognize from the
outset the inherent interconnectedness and complexity of the real world. Thus,
our purpose in identifying so many different perspectives and issues here is not
in order to analytically break down our object of study into many different
component parts; rather, it is to provide fine-grained conceptual lenses that can
be used to develop sensitivities and analytic capacities that can be combined in
order to gain a better and more holistic understanding. This holistic and systemic
orientation is indeed fundamental for design, since every design “is either an
element of a system or a system itself and is part of ensuing causal
entanglements” (Nelson and Stolterman 2012, 57). In all designed things—and
especially in fluid assemblages—there is a constant interplay between the
particular and the systemic, and a temporary focus on one should not cause us to
lose sight of the other. In fact, we argue that fluid assemblages require a new
level of agility when it comes to moving between the two.
A consideration of properties and interactions addresses what it means to be
that kind of a thing at a basic level, and in which types of interactions they are
generally involved. But the significance of things is in many cases not due to
their existence in one instance, but rather in many—the fact that they are
ubiquitous enough to be implicated in larger-scale societal patterns and
dynamics. Indeed, things—whether tools, media, or computers—collectively
constitute a significant part of contemporary environments. Consideration of
different kinds of things in terms of systemics and scale thus addresses a more or
less quantitative question—what is it like to have not just one but many,
sometimes vastly many, of this kind of thing in the world? What are the
dynamics and consequences of things being interconnected, and of scaling up
and out? The qualitative question that is the counterpart of systemics and scale
is: What are the social implications and forms of life that are supported or
precluded, encouraged or foreclosed, in a world where these things hold a
significant sway?
In the following sections we will consider three basic orientations to
considering what things are: as stuff, assemblages, and objects or machines.
Each of these perspectives can be seen in multiple and vast bodies of work,
which can of course not be completely addressed here. The aim of this section is
also the opposite of being encyclopedic: rather than accounting for the full depth
and breadth of existing work, we want to get to the bottom of the basic elements
and contours. The references we give are indicative rather than comprehensive,
but they can also serve as entrance points for further investigation.

Stuff
Viewing things as “stuff” is perhaps the most basic way of relating to some other
entity in the world with a physical presence in our lives. The very word itself is
in fact telling, as “stuff” is such a general reference that it does not really say
much at all about the character of what is referred to. As such it can say
something about the way that people relate to things though. Think, for example,
of phrases such as “I have so much stuff” or “I need to get my stuff”; the
reference is not individual but aggregate, a particular selection of the total
accumulation of the various life accessories, basic necessities, prize possessions,
curios, and detritus that are our companions at various (and variously lengthy)
stages of life.1 The scale of this situation can be encountered with some force
when moving and needing to sort through one’s things that must be either
packed and moved or else somehow disposed of. This can be a time for making
quite particular kinds of calculations around the value that particular things bring
to one’s life in relation to the cost of keeping them, involving perhaps
considerations of utility, history, emotional attachment, expected future value,
and more. Stuff is a collection of things that have physical presence and are there
for us in the ordinary dealings of our lives, variously supportive and frustrating,
meaningful and mundane, broken down and repaired, treasured and discarded.
They have their own temporal trajectories and narratives that are anchored in a
teleological terminus of their eventually becoming waste (Viney 2014).
Our stuff can also say a lot about who we are, individually and collectively,
how we order our lives, what we value, what privileges we have or do not have,
and so on (Menzel and Mann 1994). The field of material culture takes stuff as
its object of study, showing that stuff not only reflects but also helps create who
we are and how we live our lives (Shove 2007; Miller 2013). The field of design
anthropology considers similar issues but in a way that is more active,
interventionist, and future-oriented (Gunn 2013). The practice turn in social
theory also places stuff in a central role, considering the ways in which things
become meaningful over time in everyday life (Cetina, Schatzki, and Savigny
2005). Diving even deeper, Martin Heidegger famously saw relations with things
as fundamentally constitutive of our dwelling in the world (Heidegger 1993).
 At this everyday level of relations with stuff, it is not strictly rational
functionality that determines the development and uses of objects. Jean
Baudrillard demonstrated this at length in his book The System of Objects,
arguing that “it is the whole system of needs, socialised or unconscious, cultural
or practical—in short, a whole inessential system, directly experienced—which
surges back on to the essential technical order and threatens the objective status
of the object itself” (Baudrillard 2005, 7). Rather than being essential to it, form
becomes a connotation for function, as in the tail fins of mid-twentieth-century
American cars that signified speed but were functionally counterproductive
(Baudrillard 2005). Baudrillard’s remarks summarizing this situation seem now
remarkably prescient and relevant to fluid assemblages (Baudrillard 2005, 67):
Every object claims to be functional, just as every regime claims to be democratic. The term evokes all
the virtues of modernity, yet it is perfectly ambiguous. With its reference to “function” it suggests that
the object fulfils itself in the precision of its relationship to the real world and to human needs. But as
our analysis has shown, “functional” in no way qualifies what is adapted to a goal, merely what is
adapted to an order or system: functionality is the ability to become integrated into an overall scheme.
An object’s functionality is the very thing that enables it to transcend its main “function” in the
direction of a secondary one, to play a part, to become a combining element, an adjustable item, within
a universal system of signs.

Although Baudrillard (2005) framed consumption in terms of sign values

operating within discourses, it is worth returning briefly to the more literal
aspects of consumption and its roots in commodity capitalism. As Karl Marx
famously argued, one of the defining characteristics of commodities is that they
conceal their conditions of production. A commodity does not reveal the labor,
and laborers, involved in its production, but rather appears as a more or less free
floating good available for exchange. This situation still maintains as a default
today, which is why it is possible to feel shock when learning through news
reports of dreadful conditions for workers who make our smartphones or
clothing. It can also be seen in efforts to be more transparent about conditions of
production such that consumers can make more ethically sound purchasing
decisions, as in certification and labeling schemes for fair trade, sustainability,
not using animal testing, and so on.
In the more high-tech realm, the efforts of Fairphone to produce “the world’s
first ethical, modular smartphone”2 are illuminating. They state that their phones
“hold a complex story of the hundreds of people who helped make it,” and that
they “want to open up that story, so we can make a positive impact in how
phones are made, used and recycled.”3 Their efforts that they report do indeed
reveal the complexities and challenges of ensuring that something is ethically
produced and with longevity in mind, in everything from sourcing raw materials
to production to repair and recycling.
A large part of the motivation for initiatives such as these is the general
recognition of accelerating climate collapse and the need to reorient present
systems of production and consumption toward ones that are compatible with
ongoing sustainment for our species and other life forms on earth. As Timothy
Morton forcefully describes the current situation:
The end of the world has already occurred. We can be uncannily precise about the date on which the
world ended. Convenience is not readily associated with historiography, nor indeed with geological
time. But in this case, it is uncannily clear. It was April 1784, when James Watt patented the steam
engine, an act that commenced the depositing of carbon in Earth’s crust–namely, the inception of
humanity as a geophysical force on a planetary scale. Since for something to happen it often needs to
happen twice, the world also ended in 1945, in Trinity, New Mexico, where the Manhattan Project
tested the Gadget, the first of the atom bombs, and later that year when two nuclear bombs were
dropped on Hiroshima and Nagasaki. (Morton 2013, 7)

While these kinds of pressing environmental concerns and corresponding human

responsibilities in the Anthropocene have been significant factors, they have not
been the only ones in a “turn to things” and matter that have spanned (at least)
the social sciences and humanities. This (re)turn has in various ways felt around
for the more solid moorings of theory, ideas, and signifiers, even if things seem
to always somehow arrive after “as the alternative to ideas, the limit to theory,
victims of the word” (Brown 2001, 16). There are many very rich strands of
thought that can be seen as falling under the heading of a return to things and
new materialisms (Dolphijn and van der Tuin 2012), which we cannot cover
here; but it is worth highlighting at least a couple key points. One is a
recognition of what can be the active, forceful agency of things—what political
theorist Jane Bennett calls Vibrant Matter in her book of the same name, in
which she uses examples such as debris, a power grid, food, and stem cells
(Bennett 2010). Rather than seeing matter as simply inert and lying around until
humans do something with it, this perspective calls attention to the internal
dynamism and latent capacities that things possess, and their abilities to affect
(and be affected by) other things.
Another key characteristic of new materialist perspectives is a focus on
process and change over stability and identity. A feminist new materialism sees
this possibility for dynamism, complexity, and movement as a welcome and
indeed essential liberation from universalist and masculine ideals (Braidotti
2003). Moreover, this ongoing becoming is characterized by a fundamental
entanglement among entities and agencies that Karen Barad articulates most
forcefully in her philosophy-physics, a perspective she terms agential realism.
She uses the neologism “intra-action” to refer to the “mutual constitution of
entangled agencies” (Barad 2007, 33). This is a fundamentally different
conception of agency than has been the norm, seeing agency not as residing in
things themselves as internal properties but rather constituted through their intra-
actions. Here the basic ontological unit is not things but rather phenomena, in
which intra-acting agencies participate in differential mattering (Barad 2007).
According to Barad, things do not pre-exist, but are rather “agentially enacted
and become determinately bounded and propertied within phenomena” (Barad
2007, 150). Now what she is referring to here are not any particular class of
things, but all things at a fundamental level. However, there is a remarkable
parallel between this description and fluid assemblages, where these dynamics
can be seen more clearly than in the case of other kinds of things. Fluid
assemblages do not pre-exist, but are rather enacted and become determinately
formed and propertied as things available for use within particular contexts.

Assemblages
Unlike notions such as “stuff” or “things,” most of us do not use the notion
“assemblage” to relate to entities we encounter. Thus, it is not very
straightforward to account for how relating to things as assemblages turns out in
everyday life. The concept does, however, offer significant potential when it
comes to analysis. The notion of assemblage was introduced in Chapter 3, thus
here we will only mention a few additional aspects as we position the different
ontological orientations in relation to each other.
The key reason we make use of assemblage is that it allows us to be rather
precise with regards to how a certain whole comes into being—even as we
engage with a rather naïve understanding of this idea.4 As we try to discuss in
this book, the perhaps most critical aspect of these networked computational
things is how they draw matter and matters together in often very complex ways.
And to better see how this is done, Manuel DeLanda’s key characteristics of
assemblages are quite useful:

1. Assemblages have a fully contingent historical identity, and each of them is

therefore an individual entity: an individual person, an individual
community, an individual organization, an individual city (DeLanda 2016,
19).
2. Assemblages are always composed of heterogeneous components (DeLanda
2016, 20).
3. Assemblages can become component parts of larger assemblages (DeLanda
2016, 20).
4. Assemblages emerge from the interactions between their parts, but once an
assemblage is in place it immediately starts acting as a source of limitations
and opportunities for its components (downward causality) (DeLanda 2016,
21).

With respect to such characteristics of assemblages and assembling, not only

do our networked computational things combine a range of technologies,
computation, communication, sensors, interfaces, etc. to become what they are,
they also make use of resources in a variety of places, both geographically and
thematically, and on top of all these they very much operate on basis of the
social fabric of everyday life. But whereas our traditional things are mostly
totalities—stable wholes that cannot be taken apart as they are made of parts that
have given up their individuality—these computational things are to significant
extent built (in runtime!) using components that retain theirs: that can be
recombined, reused, and thanks to digital technology even be made active in
many places at the same time (such as when a certain piece of software is
executed in multiple instances).
There is a relevant relation to the notion of form to be made here, especially if
we think of design as a matter of giving form to things. In Aristotle’s philosophy,
form is the way matter builds things. For instance, it is the “form” that we see
when looking a mountain—for how could we have the material of a mountain
inside our head? (His theory of perception might be outdated, but it is still
fascinating how the concepts of form and matter allow him to explain what is
going on, such as how perception then is a matter of in-form-ation).
Furthermore, form is not only about what is already real, or actual, but also
about what may become. For instance, within the form of the acorn there is also
the potential to become an oak. The idea of form as a matter of both what has
already been realized and what has the potential to become is of special
importance to the things considered here—just think about how to understand
the form of computation, the code as such and the specific “machine” it
instantiates when it is executed.
With respect to such issues, Gilles Deleuze makes an important distinction:
Now, the process of realization is subject to two essential rules, one of resemblance and another of
limitation. For the real is supposed to be in the image of the possible that it realizes. (It simply has
existence or reality added to it, which is translated by saying that, from the point of view of the
concept, there is no difference between the possible and the real.) … The virtual, on the other hand,
does not have to be realized, but rather actualized; and the rules for actualization are not those of
 resemblance and limitation, but those of difference or divergence and of creation. (Deleuze 1988, 97)

While this is part of building an ontology based on difference in itself

(metaphorically somewhat like when Carl Friedrich Gauss developed a
differential geometry that in contrast to earlier algebraic solutions to how to
describe curved surfaces does not depend on a reference to an external space), it
actually applies also to the things considered in this book: that these things are
not characterized by the distinction real versus possible, but importantly by the
distinction actual versus virtual—and where the virtual is very much part of what
makes these things “real.”
This is important as we are here talking about things that have been designed:
these are not just objects given in one sense or another, they are made. They
have been made real by someone, and what Deleuze’s distinction tells us is that
not only was something “actual” made real, so was something at this point
“virtual.” We might even in the light of changing relations between design and
use, and how these things literally come together only as we actually use them,
push ourselves to define design in the context of fluid assemblages to be largely
about the domain of the “virtual” along with the lines of differentiation,
divergence, and creation that will turn virtual into actual. That is, designing as
such deals not only with the virtual and how to make something actual—so does
in this case the design output: the “things” made are not only “real” in terms of
what is now after design made actual—but importantly also in terms of what is
now made real in the virtual sense. Even if this would be only partially correct, it
becomes very clear how insufficient it is to understand this kind of design in the
same way as we worked with industrially manufactured objects: “totalities” that
are “finished” as they are made, and with a prototype serving as mold for mass
multiplication, the key challenge being to turn what is possible into reality. Or, to
put it differently: the basic responsibility of designers lies not with what is made
actual through design, but with what is made real. And in the case of these
assemblages, what is made real is very much also a matter of the virtual: that
which has not yet become actual, but could.

Objects/machines
It is at this point following the other two sections that it in some ways becomes
particularly difficult to make clear-cut distinctions among these different ways of
conceptualizing things, since there is much in object-oriented perspectives that
includes ideas about materialism and assemblages. However, there are also a
number of other key insights that are particular to the object-oriented
philosophical movement that has emerged within the last decade or so.
The overall umbrella of this movement has been called the speculative turn,
developing as a counter to the linguistic turn that had made it impossible to talk
about anything other than language that can never escape its own webs of
significations to reach any reality outside of itself (Bryant, Srnicek, and Harman
2011). The movement was ignited by the philosopher Quentin Meillassoux’s
diagnosis of correlationism as the condition that has haunted all of philosophy
since Kant: a concern with correlation, “the idea according to which we only
ever have access to the correlation between thinking and being, and never to
either term considered apart from the other” (Meillassoux 2009, 5). Thinking
about the world had turned into thinking about only the relation between human
and world, and speaking of anything as existing “in-itself” had become the
height of philosophical naiveté. But in accepting these foundational terms for
inquiry, Meillassoux suggests:
Contemporary philosophers have lost the great outdoors, the absolute outside of pre-critical thinkers:
that outside which was not relative to us, and which was given as indifferent to its own givenness to be
what it is, existing in itself regardless of whether we are thinking of it or not; that outside which
thought could explore with the legitimate feeling of being on foreign territory—of being entirely
elsewhere. (Meillassoux 2009, 7)

While this rediscovery of reality might seem rather amusing to non-

philosophers, it is worth considering just how deep these correlationist
assumptions run in shaping also other forms of inquiry. One example that is
quite close to our concerns here is the focus on user experience in design.
Certainly things themselves have been carefully shaped and examined in design,
but this has been done with a view toward the meanings and experiences that
they are assumed to enable. The underlying questions concern not so much what
things are and do in themselves, but what they are and do for humans who use
them. The linguistic turn has also been felt in design as a turn toward product
semantics, what objects signify. The split between thing-for-a-human and thing-
in-itself only deepens when it comes to computational things.
But things in themselves also have ways of reasserting themselves. Now we
realize, for example, that plastic consumer goods and packaging continue to
have a life of their own after humans have disposed of them, in landfills and
oceans where they affect other life-forms and ecologies in ways that have
nothing to do with user experience as it is considered in design. The need to
account for multiple roles of things and nonhuman agencies is made more
pressing by our ecological crisis. Morton articulates this need in terms of
hyperobjects, a term he coined to refer to “things that are massively distributed
in time and space relative to humans” (Morton 2013, 1)—the agencies and
ecologies of the Anthropocene and a history that is no longer exclusively human.
In the digital realm, we have also learned, for example, that social media are not
only means for sharing what we are up to, but also, among other things,
organization and recruitment tools for terrorists, the nexus of economic
ecosystems, large-scale influence and dis-information machines, and tools that
foreign governments can use to hack elections. The things we have created are in
some ways taking on lives of their own. They may not be exactly like
Frankenstein’s monster, but there is still something uncanny in the recognition
that they are not entirely under our control or even open for our understanding.
Returning to philosophy, one of the key characteristics of object-oriented
philosophy is of course its focus on objects as the building blocks of reality. This
is a conception of objects in a much more expansive and inclusive sense than
might be our ordinary habit of mind. It tends to work with a flat ontology that
sees all kinds of entities as equally existing, and does not discriminate on the
basis of scale, whether something is considered real or imaginary, corporeal or
incorporeal. Graham Harman (2011) explains the reasons for this in his diagnosis
of the traditional “undermining” and “overmining” of objects. Undermining
involves always breaking objects down into smaller constituent pieces that are
seen as more fundamental in some way, whereas overmining sees objects only in
terms of their perception or the effects that they have. Undermining cannot
account for the emergence of objects as independent entities, whereas
overmining does not recognize the reality of objects that exceeds the effects that
they have in the world (Harman 2016). But Harman reserves his strongest
criticism for the combination of the two that he calls “duomining,” a position he
associates with materialism as “the hereditary enemy of any object-oriented
philosophy” in that it both reduces down to ultimate elements while also treating
them as bundles of qualities (Harman 2011, 13). In opposition to materialism, he
proposes the concept of immaterialism. While materialism emphasizes fluidity
and change, action and intra-action, practice, contingency, multiplicity, and
immanence, the principles of immaterialism focus on stability, substances and
essences, interactions, singularity, and withdrawn reality over pure immanence
(Harman 2016). The case study he uses to illustrate this object-oriented approach
to social theory is of the Dutch East India Company and the various stages of its
existence over time.
Now, the purpose here is not to follow Harman in waging assault on
materialism, or to provide an account of the retorts from the other side, for
example, Braidotti and Vermeulen (2014). But we can at least acknowledge the
truth in the somewhat milder formulation of Levi Bryant:
Materialism has come to mean simply that something is historical, socially constructed, involves
cultural practices, and is contingent. It has nothing to do with processes that take place in the heart of
stars, suffering from cancer, or transforming fossil fuels into greenhouse gases. We wonder where the
materialism in materialism is. (Bryant 2014, 2)

In fact, we see the presence and tension between dynamics associated with both
materialism and immaterialism in the case of fluid assemblages: the fact that
they are in flux, but also stabilize as particular things with particular capabilities
and agencies. The concept of object is helpful here, referring to “any entity that
cannot be paraphrased in terms of either its components or its effects” (Harman
2016, 3). This helps in recognizing entities that exist as such across different
scales.
Another term that has been used and that is also quite relevant for our project
is machine. The distinction between tools and machines that Lewis Mumford
made early in the twentieth century is still remarkably relevant:
The essential distinction between a machine and a tool lies in the degree of independence in the
operation from the skill and motive power of the operator: the tool lends itself to manipulation, the
machine to automatic action. (Mumford 2010, 10)

This “automatic action” is indeed a highly significant difference, even as we

might now more likely consider the implications of machine learning and
artificial intelligence as animating agencies of machines rather than steam or
electricity. Another more recent variation on this concept that is also attuned to
computational things is Ian Bogost’s unit operations, a term he uses to refer to
“the logics by which objects perceive and engage their worlds” (Bogost 2012,
29). Bryant develops his ontology using the concept of machine to refer to “any
entity, material or immaterial, corporeal or incorporeal, that exists”; and one of
his reasons for using this term is the fact that it evokes this sense in which
entities operate and function (Bryant 2014, 15).
One of the things that objects/machines do is relate to each other. This is not
to say that they establish relationships with each other, but rather that they can
affect each other in various ways, according to their character. One of the main
insights that Harman (2002, 2011) develops is that there is always an excess of
reality to objects that is not exhausted by interactions with them, that is
withdrawn in a fundamental sense. Objects are able to access only certain
aspects of other objects that are accessible to them, based on their particular
character. Harman illustrates this point using the example of fire burning cotton,
in which the fire interacts with only the flammability of the cotton and not its
other properties that may be accessible to other entities (such as humans who can
appreciate its fuzzy softness).
One of the ways that machines can relate to each other is through a structural
coupling, a configuration that Bryant (2014) describes as one machine becoming
a medium for another. It is in this way that machines come to have a purpose or
use. As he states: “A machine functions as a medium for another machine not
only when it amplifies or extends a sense-organ, but also whenever it modifies
the activity or becoming of any other machine” (Bryant 2014, 33). This could be
machines in the more industrial sense we might immediately associate with the
term, but also includes what could otherwise be described as sociotechnical or
even purely social assemblages. They are the entities that come to exist and to
have effects in the world. Bryant suggests a practice of onto-cartography as a
way of mapping the machines in a given world in order to better understand and
—especially—to change it. As he describes it:
In its initial formulation, onto-cartography is the investigation of structural couplings between
machines and how they modify the becomings, activities, movements, and ways in which the coupled
machines relate to the world about them. It is a mapping (cartography) of these couplings between
machines (onta) and their vectors of becoming, movement and activity. (Bryant 2014, 35)

A key method for this mapping activity among entities that are fundamentally
foreign to us is alien phenomenology. In his book on the subject, Bogost
describes the kind of speculative practice that is needed:
Speculative realism really does require speculation: benighted meandering in an exotic world of utterly
incomprehensible objects. As philosophers, our job is to amplify the black noise of objects to make the
resonant frequencies of the stuffs inside them hum in credibly satisfying ways. Our job is to write the
speculative fictions of their processes, of their unit operations. Our job is to get our hands dirty with
grease, juice, gunpowder, and gypsum. Our job is to go where everyone has gone before, but where
few have bothered to linger. (Bogost 2012, 34)

One of the objects to which Bogost has devoted the most attention is the
computer, an object in its own right that can clearly be seen as having its own
withdrawn reality and agencies. He also argues for a practice of “philosophical
carpentry,” building things as “philosophical lab equipment” (Bogost 2012, 100).
This resonates remarkably well with practices of constructive design research,
which in turn would need to work with an orientation of alien phenomenology in
order to adequately investigate the character and possibilities of fluid
assemblages.
There is in fact some work already along these lines, as in work on “thing
ethnography” in which things serve as participants in inquiry in capturing the
world around them (Giaccardi et al. 2016); and in the “Morse Things” project
that investigates what it could be like to live together with things that are
networked and communicate with each other (Wakkary et al. 2017). This kind of
material speculation is indeed becoming increasingly important as we are now
living with entities that interact with each other not only on the basis of their
“natural” physical properties (as in the case of fire and cotton), but that have
designed properties that allow them to interact with each other. Designing these
kinds of things means, then, not only designing things, but designing ecosystems
and the connections among entities that are their mechanisms of evolution.

Notes
1 A beautiful art/photography project illustrating this is Material World: A Global Family, portrait by
Peter Menzel and Charles Mann (1994). It shows portraits of families outside their houses with all their
possessions laid out and listed.
2 https://www.fairphone.com/en/, accessed October 23, 2017.
3 https://www.fairphone.com/en/our-goals/, accessed October 23, 2017.
4 Indeed, as Manuel DeLanda points out, not only does the English translation of the French term
“agencement” imply that we lose aspects of its original meaning, pinning down the concept is made
further complicated by Deleuze and Guattari offering several different definitions so that “each
definition connects the concept to a separate aspect of their philosophy, using the terms that are relevant
for that aspect, so when taken in isolation the different definitions do not seem to yield a coherent
notion” (DeLanda 2016, 1).
 7

A Conceptual Toolkit

While the kind of fluid assemblages addressed in this book are in many ways
relatively new, they also connect to existing dynamics and trajectories and there
is already much to work with in terms of conceptual tools that can help us better
articulate and make sense of what is going on. The aim of this chapter is to
identify some basic components of such a conceptual toolkit, and the kinds of
analytic uses to which they can be put.
The perhaps most obvious approach to accounting for existing scholarship
regarding a particular set of concerns would be in terms of disciplinary
trajectories, and the related historical developments in their objects of study.
However, making disciplinary distinctions can be a fraught enterprise. Rather
than being driven by a desire to add another installment to one particular
disciplinary narrative, we want to work in a more transdisciplinary manner that
is motivated by what we have in front of us and need to account for. For this
reason we are much more interested in the kinds of analytic distinctions that can
be useful in making sense of various aspects of things. What we are after here is
conceptual precision, nuance, and utility in analytic tools that can help us see
and account for key elements of things in a multifaceted, holistic way. This
means that in our investigation of existing scholarship we are on the lookout for
aspects of things that they can bring into focus.
There is a fine line between aiming for transdisciplinarity and simply being
undisciplined. In some ways it may seem that we stray toward the latter as we
interpret and misinterpret, use and misuse the conceptual tools we now borrow
from contexts to which we do not exactly belong. But what we aim to do is to
work in a way resembling the structure of the objects we study: “things” coming
together by assembling technologies and transcending concepts that were
initially developed for different places and purposes. This said, we also
intentionally make quite extensive use of quotations, inviting other voices to join
our assemblage in their own words in order to keep crucial aspects of their
original context.
We believe that just as these “things that change” need to be approached as
fluid assemblages, so does our account of them. The perhaps not so surprising
but still equally important image that this yields is one not that different from
looking at something in a shattered mirror: some things are clearly seen in some
places, but overall the image is fragmentary and it is truly difficult to trace
continuous contours. The insight dawns upon us that this might not only be an
effect of the limited reach and concern of these perspectives, but perhaps more
fundamentally because the phenomena at hand will not be contained within one
account, that there might not be a single complete description, no unified theory.
Any notion of one perspective being more “foundational” than another is thus
quite misleading, as there is no bottom to be reached, no firm ground to be
found. Rather, what we face here in terms of theory is structurally not that
different from the things we study: layers and layers of connections between
concepts, descriptions, and perspectives that, while each making sense on their
own, still require us to be mobile in order to understand how a “bigger” picture
is emerging. Indeed, while there are dimensions also in a rhizomatic structure
like this, there are no given directions. It is not particularly important to
determine if one term is more foundational than another, but it is crucial to
understand the relation between a perspective and what can be seen from that
point of view. In this we in some ways still heed the mantra echoing in
philosophy since Kant’s own Copernican Revolution: to understand our world,
we need to turn to the things themselves—however problematic and even
contradictory that at times may seem. Again, we therefore ask: what is this thing
we’re making and using?

Experiential
As a starting point, we begin at the place where we first encounter things in
ordinary life: as they enter into experience in personally meaningful ways.

Equipment

Arguably, the most famous and influential philosophical analysis of tools was
put forward by Martin Heidegger. Before considering it directly though, we need
to look at its context. The project in his major work Being and Time (Heidegger
[1927] 2010) was to formulate the question of the meaning of being. The
primary element of this investigation concerned the kind of being who asks this
question about the nature of being as such—which is to say, human beings. This
kind of being that is fundamentally concerned about its own being he terms
Dasein. Following in the tradition of phenomenology and his teacher, Edmund
Husserl, the structure of being of Dasein for Heidegger is always being-in-the-
world. It is this structure that he sets out to investigate.
The phenomenological method he uses takes phenomena as its unit of
analysis, which he understands as something “showing itself in itself.” This
“self-showing” of phenomena is always the basis for appearances, which, in
fulfilling their function, refer back to the phenomena. This is a fundamental
relationality in which we encounter things as they exist for us. Moreover, the
“being toward the world” of Dasein “is essentially taking care” (Heidegger 2010,
57). This is care in an ontological sense, not in the sense of affection or distress
or similar; we take care about the things we encounter because of the basic fact
that they are caught up together with us as the world in which we have our
being. It is only “when we put ourselves in the place of taking care in the world”
(Heidegger 2010, 67) that other beings become accessible.
This is (an extremely condensed version of) the path that leads Heidegger to
his consideration of tools, and our relation to them. He states that the beings we
encounter when taking care in the world are not just things, but rather useful
things, or equipment1—things that we can use in order to do something. This
structure of “in order to” that is essential in our relation to useful things also
contains other references: useful things are related to other useful things. It is
this totality of useful things that is discovered “always already … before the
individual useful thing” (Heidegger 2010, 68). Moreover, when we deal with a
useful thing, we are not dealing with or even considering it as such and in itself,
but rather in the sense and to the extent that we are able to adequately
appropriate it for something.
Here we arrive at the breakthrough insight in Heidegger’s analysis of our
relations to things, in which he has been discussing the example of hammering
with a hammer:
In such useful dealings, taking care subordinates itself to the in-order-to constitutive for the particular
utensil in our dealings; the less we just stare at the thing called hammer, the more we take hold of it
and use it, the more original our relation to it becomes and the more undisguisedly it is encountered as
what it is, as a useful thing. The act of hammering itself discovers the specific “handiness”
[“Handlichkeit”] of the hammer. We shall call the useful thing’s kind of being in which it reveals itself
 by itself handiness [Zuhandenheit]. It is only because useful things have this “being-in-themselves”
[“An-sich-sein”], and do not merely occur, that they are handy in the broadest sense and are at our
disposal. (Heidegger [1927] 2010, 69)

And further:
What is peculiar to what is initially at hand is that it withdraws, so to speak, in its character of
handiness in order to be really handy. What everyday dealings are initially busy with is not tools
themselves, but the work. What is to be produced in each case is what is primarily taken care of and is
thus also what is at hand. The work bears the totality of references in which useful things are
encountered. (Heidegger [1927] 2010, 69)

The paradox here is that the more closely we engage with something and it is
thus revealed to us as what it is, the less we are actually aware of it. It
“withdraws” as we focus on what it is that we are using the thing to do. It comes
into objective presence only when its “handiness” turns into “unhandiness,”
when our use of it breaks down for some reason as it becomes damaged or is not
working well for our purpose. Rather than being withdrawn in smooth and
effective use, it becomes conspicuous and obtrusive as an obstinate, broken tool.
Under normal circumstances, then, when things are serving their function as
equipment for us, we are caught up in our activities and the things are, in a
fundamental sense, not present for us. Heidegger summarizes this interpretation
by saying that “being-in-the-world signifies the unthematic, circumspect
absorption in the references constitutive for the handiness of the totality of useful
things” (Heidegger 2010, 75).
Although Heidegger’s concern was ontology at a fundamental level, this
insight about things withdrawing from awareness during effective use can be
applied in quite practical ways. It has, for example, been used in relation to
understanding interactions with computational things in the context of human–
computer interaction, which originated as a project of designing interfaces that
would make it possible for humans to use complex computational systems (e.g.,
Winograd and Flores 1986; Dourish 2001). Withdrawal in handy use can even be
seen as a goal for design: when something works well, it should disappear from
the user’s awareness and allow for focus on the project at hand. When computers
and graphical user interfaces were relatively new this might have seemed like an
ambitious goal indeed. However, now computational devices are commonplace
and use a number of familiar standard interaction conventions (e.g., dragging,
swiping, opening and closing windows, and so on); so rather than being clunky
and frustrating to use, they are now often (even if not always!) easy to use at the
level of basic interaction, even working seemingly “automagically” or being (by
careful design) addictively pleasurable to use. This sentiment can also be seen
explicitly in the advertising for the iPhone X, which is described as having a
screen “so immersive the device itself disappears into the experience.”2
This has significant implications as “use” of a thing is now increasingly
distinctly different for different kinds of users versus for the owners of the
broader system; what appears to us as a “useful thing” is in many cases just one
of many manifestations of a fluid assemblage. Consider Facebook for instance:
to the typical user it is a social platform for staying in touch, for sharing updates,
photos, and more; to the typical advertiser, it is a highly structured marketplace
where the data gathered allows advertising to be targeted to very specific groups
based not only on general demographics or distribution patterns, but on personal
data, interests, mobility, properties of their social network, and so on. Indeed,
entering the Facebook website as an average user, one is greeted by statements
like: “Connect with friends and the world around you on Facebook” and “It’s
free and always will be.” Entering as a potential customer on Facebook business,
on the other hand, one is informed that “Advertising on Facebook makes it easy
to find the right people, capture their attention and get results” along with
examples of success stories of “how businesses similar to yours are growing
with Facebook marketing.” That different stakeholders, and in particular
producers and consumers, have different relations to things is certainly nothing
new, nor is it particularly surprising that such groups have different reasons for
attending to a given thing. But what has happened in cases such as Facebook and
Google is they are now quite different things to different people: on one hand a
way of staying in touch with friends, on the other a marketing platform; on one
hand a tool for finding information, on the other a data harvesting machine to
optimize targeted advertising.
It is crucial that we do not dismiss this complexity in simple terms such as
whether this is good or bad, right or wrong, but rather actually acknowledge that
what seems to be a fairly straightforward relation to technology as something
that is there for us as useful things is really much more complex. These fluid
assemblages are in fact anything but straightforward when it comes to how, and
to whom, they come to appear as “useful things”—and what those uses are.
Engagement

Another aspect of things that was a focus in Heidegger’s later work is the way in
which they entail a gathering of different elements. Distinguishing things from
mere objects, he asserts that things stand on their own and are not dependent on
our representations of them as objects are. He goes on to describe how things
rather gather together in themselves the “fourfold” of earth and sky, divinities
and mortals. Somewhat less poetically, he also connects to a more originary
meaning of the word thing:
The Old High German word thing means gathering and indeed a gathering for the negotiation of an
affair under discussion, a disputed case. Consequently the Old High German words thing and dinc
become the name for an affair; they name what concernfully approaches the human in some way, what
accordingly is under discussion. (Heidegger [1994] 2012, 12)

This understanding of things as material gatherings around shared matters of

concern has been picked up quite productively in design theory (Binder et al.
2011) and science and technology studies (Latour 2005), and is something to
which we will return later on.
Keeping for now though quite close to actual interactions with things, we can
note that there is a certain basic ordering and connective function of things. Even
if we do not follow Heidegger all the way to seeing earth and sky, divinities and
mortals, in the things around us, it is possible to recognize that they are in some
real sense about more than just the material object that is objectively present for
us. Trying out this way of seeing with a simple exercise, we can think of the
toothbrush that typically sits by the sink in the bathroom at home. It is not just an
isolated object but a thing for us as we reach for it in the morning and evening
(and a toothbrush in particular is even more literally a “thing for us” since its
relations with humans tend to be quite strictly monogamous throughout its useful
life). It is typically something that withdraws during use as we go through our
daily rituals of brushing our teeth without giving it much thought (unless,
perhaps, we have just been to the dentist and received remedial instructions in
brushing). It is also something that does not exist on its own but also entails
relations to other things: the tube of toothpaste, the sink and its running water,
the cup it is stored in, etc. More broadly, it also connects to other aspects of life;
this could be memories of reminders from parents to brush early in life, everyday
feelings of being prepared for the day ahead or a night of sleep, desire to rid
one’s mouth of garlicky aromas that could negatively impact certain kinds of
interpersonal relations, and so on. Expanding our scope of consideration beyond
ourselves, we might also think of what will happen to a toothbrush after we
dispose of it, when the plastic it is made of will exist in some form, in some
place on Earth, for hundreds of years. A humble toothbrush can thus gather
together in itself our everyday rituals, memories, social norms, interpersonal
relations, planetary impacts and responsibilities, and more.
Things, then, seem to invoke and lend themselves to certain ideas about how
to carry out the affairs with which they are involved. The fact that our lives are
thoroughly textured by technological things thus calls for consideration
regarding the kinds of ideas and patterns that they support or discourage. This
has been the project of neo-Heideggerian philosopher of technology Albert
Borgmann, who develops what he refers to as the “device paradigm” to describe
a pattern he identifies in the use of technology in contemporary life (Borgmann
1984). Whereas “focal things unify and gather, devices divide and scatter”
(Strong and Higgs 2000, 32). He illustrates this with his famous example of the
wood burning stove in contrast to modern (particularly North American) central
heating systems. The wood stove gathers together a set of focal things and
practices involving chopping and fetching the wood, building and tending the
fire, and so on that can correspond to particular roles in a family. It serves as a
focal point in the home that people gather around, perhaps cooking, warming
oneself after coming in from the cold, or simply sitting around it. In short, using
a wood stove entails engagement with one’s world and other people in it. In
contrast, a central heating system disburdens users from the significant labor
involved in using a wood stove for heating, instead providing heat as an
effortless commodity in response to a setting on the control unit on the wall.3
The means of producing the heat are separated from the ends, and very minimal
engagement and effort are required from users in order for it to function.
Indeed, our contemporary computational devices exhibit more complex
patterns of engagement, of presence and withdrawal, than we have seen before.
In relation to the toothbrush discussed earlier, let us consider the development of
another personal tool we use on an everyday basis: a wristwatch. Although
always a site for expression of technological progress and its increasing
miniaturization, it has throughout also been a means of personal expression.
Being part technology, part fashion, it is an interesting expression on what and
how we think about and make use of technology. For instance, we may look
toward the change from wearing it as part of the clothing, or on a chain in a
pocket, to wearing it on the arm—and the associated differences in the acts of
use. Or, we could think of the introduction of digital wristwatches, some even
with computational abilities in the form of very tiny calculators, as digital
technology started to diffuse into everyday life in the 1970s. More recent
developments include the increasing use of sensors and communication
capabilities, such as GPS to track position, accelerometers to track motion, or air
pressure sensors to track altitude and barometric pressure. The use of sensors for
heart rate, breathing, and other bodily measures that was first introduced for
athletes and amateurs to track their training and progression has over time
become part of setting and measuring an exercise or training regime also for
everyday use. For instance, most GPS sports watches now connect to various
other services to exploit the data collected.
A key example is services such as Strava (strava.com), perhaps most known
for turning every ride or run into a potential competition using a feature called
“segments.” A segment is a part of a trail or road, such as a climb or a technical
descent, and the ingenuity of Strava is to publish leaderboards for all segments.
Thus, every run or ride can be turned into a local competition, the individual
either aiming for a personal record on the segment or a better position on the
leaderboard including winning trophies called “KOM” or “QOM” (King/Queen
Of the Mountain, a term borrowed from Tour de France). To illustrate, consider
how a new verb—“Strava’d”—is used to describe implications for road use:
In 2011, David Millar, the British pro now riding in the Tour de France, smashed the KOM on a circuit
of Richmond Park in south-west London, among the most Strava’d roads in Britain. Using the bike
he’d ridden to victory in the time trial at the Tour of Italy earlier that year, he completed the 6.7-mile
loop in 13 minutes and 35 seconds, an average speed of 30mph.

When it was pointed out the speed limit in the park is 20mph, his record was removed, the BBC took
down a video of the ride, and Millar apologised for his “naivety.” But the current record involves an
average speed of 28mph and is still on the site, begging to be beaten. (Usborne 2013)

To understand what this becomes, we cannot just look at the different parts—the
GPS watch, the sensors, the data collection, the services, the connections
between them, what content users upload, etc.—we need to look at how they
come together in the “things” we use. Until recently, such devices have been
considered specialized sports equipment rather than general watches, but the
tendency is clear: this kind of “bringing together” is quickly becoming part of
things intended also for more general use. Consider the Apple Watch for
instance: while it lacks some of the sensors mentioned earlier, it makes use of its
continuous connection to the iPhone to provide additional functionalities.
Although in some ways an extension of the interaction with the iPhone, it also
adds important new aspects of wear such as the continuous tracking of certain
bodily data such as heart rate and movement. It might not track your
performance the way the sports watches do during a race, but there is certainly
an element of performance and progress also here: how much do you move
around, what is your average heart rate, how many stairs did you climb today,
what was your sleep like, etc. But unlike the previous sports watches, the Apple
Watch combines this new and extended functionality with an exterior design that
instead strongly enforces our perception of the thing as a “watch”: ranging from
its sleek and minimalistic form to the traditional faces it is often presented with,
it is at the same time a statement of innovation and a strong statement of
continuity.
With such technological advances in mind, it is not difficult to imagine also
the toothbrush as a site of bringing together, of making a wider range of
relations, objects, data and more, present through the engagement with this
“useful thing.” Given its place in our daily routines, it could become a means for
presenting what lies ahead during the day or summarizing what has passed,
equipped with sensors it could become a tool for tracking health (saliva is a
promising biofluid for early detection of a range of diseases), and so on and so
forth. But what is here hidden, and what is revealed? And more than anything,
what is through technology transformed from one useful thing into another, as
when the paved road is transformed from surface for easy transportation to race
track, along with our relations to other people, the different social rules we
assume apply, and much more?
Borgmann’s framework has been rightly criticized for its extreme pessimism
about the merits and effects of devices, and for the nebulousness of his concept
of engagement. Verbeek (2005), for example, argues that things that Borgmann
would see as devices that disburden users in ways that threaten focal things and
practices, such as a CD player or television, do not necessarily do so. While it
might be possible to recognize a general pattern of disengaged consumption as a
diagnosis that could be appropriate in many cases, it is less clear cut to apply it
in blanket terms to one or another particular kind of thing. However, the concept
of engagement is still a useful one for considering the character of interactions
with and through things, what and how much they require of us, and how they
connect us to or disconnect us from our material and social worlds.
Similarly to Heidegger, we can also see here in Borgmann’s analysis a
dynamic of presence and withdrawal, a contrast between a condition where
things are characterized by a certain openness in their function and where they
light up and activate networks of rich relations in the world around them, and a
condition in which they hide the details of their operation from view and enable
effortless, disconnected, and distracted consumption of commodities. The latter
situation in which operation is hidden from view and does not require user effort
or awareness seems to be particularly characteristic of modern computational
technologies, and indeed increasingly the other kinds of things that now often
have computational components embedded in them. Think of vehicles, for
example; it is from these things that we get the expression of “looking under the
hood” as a way to open something up in order to try to figure out how it
functions or, more typically, why it is malfunctioning and what might be done to
repair it. Now, however, it seems vehicle components are being gradually
engulfed in various coverings that shield them from observation and intervention
by the “unauthorized,” while their computational components are visually and
practically impenetrable in more fundamental ways.

Technological mediation

One of the basic insights developed in philosophy of technology and related

fields is that technological tools are involved in and shape human activity in
non-neutral ways. As political theorist Langdon Winner states: “Technologies
are not merely aids to human activity, but also powerful forces acting to reshape
that activity and its meaning” (Winner 1986, 6). This makes it important to
examine what technologies actually do as they are taken up and used by humans,
and how they mediate human actions and interactions.
This technological mediation is an explicit focus in the subfield of philosophy
of technology known as postphenomenology. Postphenomenology is a
philosophical approach initiated by Don Ihde around the 1980s through bringing
insights and orientations from pragmatism and science studies into
phenomenology (Ihde 2008). Postphenomenology attempts to get out of the
subjectivity of phenomenology by analyzing the structure of relations between
human and world. Variations in possible structures are emphasized in order to
avoid reductionism, and variational analyses are one of the primary
postphenomenological methods. One of the key moves that Ihde made was in
foregrounding the role of technologies in frequently mediating human–world
relations, such that human–world relations became instead human–technology–
world relations. This structure is particularly characteristic of scientific practice,
and since Ihde was heavily influenced by science studies this was one of his key
concerns. As he states: “Inter-relational phenomenology not only provides a
rigorous analytic process for understanding technologies, but also shifts and
complements the understanding of science praxis as being technologically
embodied and entailing human perception and action” (Ihde 2008, 7).
In his phenomenology of technics, Ihde (1990) develops a basic set of
relational structures that have provided the primary analytic framework for the
field of postphenomenology. The general intentionality relation he schematized
as human-technology-world. The other initial set of relations elaborated on this,
and consisted of embodiment relations, hermeneutic relations, alterity relations,
and background relations. Embodiment relations [(I-technology) → world]
describe the relation where a technology is taken up as part of one’s own bodily
apparatus in orienting toward the world, as is the case with a pair of eyeglasses.
A person wearing eyeglasses experiences them as transparent, both literally in
the case of eyeglasses but also phenomenologically. Hermeneutic relations [I →
(technology–world)], on the other hand, involve a technology that converts some
part of the world into a text that can be read and interpreted. This is the case with
a thermometer that registers the level of heat in the environment and produces a
numeric display. A person can then interpret the number displayed based on
previous experience of how a given temperature scale relates to how hot or cold
it feels. Alterity relations [I → technology–(-world)] involve technologies that
are related to as, such as a robot. These relations are becoming increasingly
common in voice-activated personal assistants in smartphones or tabletop digital
assistants. Background relations concern technologies such as those used for
climate control, which tend to always be running in the background but receive
little direct attention (unless they break down). Other formal relations have also
been added to this initial set that Ihde developed, including cyborg relations
[(human/technology) → world] (Verbeek 2008) in which human and technology
are fused into a new entity, and digital material mediation [I → ([trace |
substrate] → world)] (Wiltse 2014) in which digital technologies with multiple
functional component parts mediate perception and action by making activities
visible in more complex ways.
Another key concept from postphenomenology is that of multistability. This
points to the fact that it is possible to relate to each artifact in multiple ways.
Some of these ways are at the micro-perceptual level, as with illusion drawings
where one typically sees one figure in the drawing initially but can then see a
completely different second figure through shifting focus (Ihde 1990).
Multistability is also possible at the macro-perceptual level where culture shapes
representational and perceptual practices. Variational analysis is done by
identifying or brainstorming different possible multistabilities.
Postphenomenology, in accordance with its roots in phenomenology, is
oriented toward subjectivity and the technological mediation of perception that
structures and enables it. It can help us recognize the many and non-neutral ways
in which technologies mediate engagement with the world. However, when it
comes to fluid assemblages, this mediation becomes substantially more complex
than is the case with other more traditional technologies. In Chapter 9, we will
work with a couple elements in the postphenomenological toolkit to illuminate
these dynamics.

Technospheres

Even as it is possible to consider social implications of technological tools at the

rather high levels of forms of life and the administrative systems of industrial
society, it is also important to at the same time remain attentive to more intimate
technological relations at the level of individuals. For after all, humans are not
just hapless dupes of “the system,” even though the systems in which we live
and work do exert powerful forms of influence over our activities and forms of
life. Rather, we are active agents and subjects of our own lives. Yet our very
subjectivity in the world is shaped to a large extent by the technologies all
around us and by the connections and interactions that they mediate. In fact, as
media theorist Mark Deuze has argued, we can’t get out of media (Deuze 2014).
In the contemporary globalized world, individual subjectivities have a
distinctly global horizon. In the context of a discussion of the disjunctures of the
landscapes constituted through global cultural flows, anthropologist Arjun
Appadurai (1996) proposed the concept of technoscapes as one dimension for
exploration (the others being ethnoscapes, mediascapes, financescapes, and
ideoscapes). He defines technoscapes as “the global configuration, also ever
fluid, of technology and the fact that technology, both high and low, both
mechanical and informational, now moves at high speeds across various kinds of
previously impervious boundaries” (Appadurai 1996, 34). Appadurai’s other
“scapes” also imply technological mediation in various ways. In addition to the
global interactions and imaginaries enabled by technologies, the mere presence
of certain technologies and their affordances shapes our experience of ourselves
and our possibilities (Kiran 2012).
Another term for the technologies that constitute our environments and texture
our worlds is technosphere, proposed by philosopher Michel Puech in his
discussion of “the ethics of ordinary technology” (Puech 2016). Many of our
decisions about how to live our lives, relate to others, configure our material
conditions (to the extent possible), and go about our ordinary activities involve
decisions about how to relate to technologies that shape and mediate our actions
and interactions. This more intimate layer that serves as the interface between
subject and world Puech terms the proximal technosphere, and it is in this
existentially significant site he finds also a space of ethical significance. A key
example is the mobile phone, something that now mediates many everyday
activities and interactions. He calls for conscious development of ordinary,
practical wisdom and associated virtues and skills that are geared to life in the
technosphere.

Technical
Now that we have briefly considered the human side of technological mediaion,
we need to turn to what happens at the level of the technologies. And here we
also need to make a very important shift from thinking in terms of human agency
to thinking in terms of also technological agency. To consider what these things
actually do, it is necessary to look not only at what humans do with them but
what they do themselves—both in response to human action and more
independently.
 However, at the same time it is important to remember that all technologies
are, by definition, part of the artificial world that is created, configured, and
maintained by humans. They are, in very important ways, always sociotechnical.
Our focus on the technical here is then meant to be just that: a focus, but one that
is not meant to in any way remove technologies from their social contexts of
production and use. However, we also need to be able to examine their
properties and character, their durable and consequential presence in the world.
And it turns out that in doing this we always end up running back into the social
anyway.

Protocols and interfaces

Just as we considered how technologies can mediate human interactions, we

need to dig down one more level in order to look at how technologies
communicate, interact, and affect each other—at their protocols and interfaces.
A discussion of protocols and interfaces could be quite technical, and indeed
technical details can be important to examine in order to understand how things
really work. However, our interest here is more at the conceptual level of trying
to understand broadly how things connect, both at the level of facilitating human
use (as in when a message is sent across a network) and at the more abstract
level of possibilities and openness to connections. While the former shapes what
people are able to do with technologies during more or less normal and intended
acts of use, the latter shapes what is actually possible in terms of how things are
able to be accessed and affected, and eventually how ecosystems evolve. Our
concern here is with the forms that computational things can take, and
particularly those aspects of form that determine their possibilities for
interaction.
The two terms used here as markers for these matters are quite similar, but
also reference slightly different dynamics. Protocol, in a general sense, refers to
formal rules governing communication or procedures in particular settings. For
our purposes here we are most interested in the computing protocols that govern
the exchange of information across networks, as in the TCP/IP protocol
governing information exchange over the internet, or the DNS protocol for
converting names into web addresses. However, it is worth bearing in mind also
this more general sense of protocols not just as existing technical standards, but
rules governing interaction and exchange. These in turn also determine the
characteristics of interfaces for specific computational technologies, which
includes their user-facing interfaces but also those that let them interact with
each other. There are also both standard interfaces, such as the basic interface for
a particular operating system (Mac OS, Android, etc.), and more local interfaces
that are designed for particular applications. We might think of protocols as rules
and standards that are in effect more or less everywhere, and interfaces as the
external structures and interactive capabilities that actually exist in particular
instances of computational things. As media and cultural theorist Alexander
Galloway argues:
Protocol is a language that regulates flow, directs netspace, codes relationships, and connects life-
forms. Protocol does not produce or causally effect objects, but rather is a structuring agent that
appears as the result of a set of object dispositions. Protocol is the reason that the Internet works and
performs work. In the same way that computer fonts regulate the representation of text, protocol may
be defined as a set of instructions for the compilation and interaction of objects. Protocol is always a
second-order process; it governs the architecture of the architecture of objects. Protocol is how control
exists after distribution achieves hegemony as a formal diagram. It is etiquette for autonomous agents.
It is the chivalry of the object. (Galloway 2004, 74–75)

Galloway’s larger argument, referenced in the preceding quote, is that protocol is

the predominant form of control that exists in society after decentralization, a
form of organization for which bureaucracy was then the corresponding form of
control. The protocols of networked computing fall into two types: those that are
radically distributed (as in TCP/IP for data exchange) and those that are rigidly
hierarchical (as in DNS addressing information that uses a hierarchical inverted-
tree structure). Galloway argues that this contradiction at the heart of networked
protocol is what enables its generativity, which requires universalization and
homogeneity, standardization and openness, control and freedom (Galloway
2004).
In formal terms, protocol provides basic common structure for the connective
possibilities of things. Those possibilities can be realized in concrete interfaces,
while shaping the forms that they can take. This dynamic has been illustrated
nicely by interaction designer Timo Arnall, who explored the protocol of RFID
(Radio Frequency Identification) as an “immaterial” that provides certain types
of possibilities for interface design (Arnall 2014). RFID technology consists of a
reader and a small tag that can be embedded in all kinds of objects (such as a
metro or library card) and store a small amount of data that it transmits
wirelessly when the reader activates the tag. However, the fact that the interface
is invisible means that it can be difficult for designers to work with it as a design
material and for users to understand how it works (and does not work). For
example, people have been concerned about the possibility for criminals to
easily collect personal information without detection, say by walking by
someone with an RFID reader that is able to scan the RFID-enabled cards in
someone’s pocket. In order to address this, Arnall worked in the mode of
discursive design to visualize technical explorations of RFID as a design
material. For example, for one of his projects, “Ghost in the Field,” he used the
technique of light painting to create a visualization of the three-dimensional
physical space in which a tag and reader are able to interact. This provides a
clearer picture of what is possible (e.g., scanning the top of the tag from a few
centimeters away) and what is not (e.g., scanning a tag from a meter away).
While the zone of interaction between RFID tag and reader exists in a clearly
delimited physical space, other types of interactions get more complex and
difficult to visualize. Natural language and gesture-based interaction are prime
examples. This interaction complexity at the interface promises to only increase,
a situation that Lars-Erik Janlert and Erik Stolterman (2014; 2017) trace to the
combination of increasing technological complexity and device miniaturization.
They suggest that this situation will lead to an increase in surface-free interaction
modalities such as gesture, sound, heat, smell, wind, movement, and so on, and
possibly new approaches to interaction in terms of ecosystems and fields.
In such an interactive environment, “users” are immersed in an ecology in which they are not
conceived as interacting with particular, targeted objects one at a time but rather are moving in
situational and interactional “force fields,” causing minor or major perturbations in their environment
by their moves and actions while being guided, buffeted, seduced, or affected in any which way by
constant movements and changes in their environment taken as a whole. (Janlert and Stolterman 2014,
531)

These are examples of an underlying trajectory set in motion by technology

leaving the scale of human perception (Maeda 2000; Redström 2005). As
discussed in Chapter 2, there used to be a rather direct relationship between
surface and technological complexity, but when “surface” was replaced by
“interface,” this relation became completely arbitrary. Whereas the surfaces of
earlier technologies were partly determined by their actual workings, interfaces
are designed with respect to their intended functionality, thus enforcing rather
than revealing their agency.
This “ecological” mode of interacting with things is quite new, in line with our
experiences in the physical environment perhaps but not as much when it comes
to technological things. However, it also brings into focus the two senses of
interface that we want to highlight here: the interface as the interaction
mechanism between person and thing, and the interface as the interaction
mechanism between things. It seems that these interfaces will become
increasingly difficult to clearly identify and disentangle, but this does not mean
that they are not there. On the contrary, it is the multiplicity of interfaces, in the
formal sense of connection points between entities, that enables expanding
ecosystems and interaction complexity.
One good example of increasing device complexity leading to increasingly
complicated interfaces is the remote control. Remote controls, particularly those
for television and other media devices, are notoriously frustrating to use due to
their very large number of buttons. Or, the minimalist Apple TV remote provides
an example of another direction that can be taken in managing complexity,
where all kinds of media come through one box and the complexity is managed
through nested on-screen interactive menus rather than individual buttons on a
remote. However, the remote control is a good example in another sense as well,
as pointed out by Caetlin Benson-Allott (2015) in her study of the remote
control: remote controls offer control rather than power. Their marketing often
contains connotations of the latter, but what they actually offer is control based
on a pre-defined match between certain actions and effects. This is important to
keep in mind as interfaces perform disappearing acts and become more
“intelligent” and “intuitive”—the possibilities are programmed, and not
unlimited. They are enabled but also constrained by pre-existing, if complex and
responsive, structures.
The increasing connectivity and complexity of devices is characteristic of a
state of affairs commonly referred to as the Internet of Things (IoT), in which all
kinds of things are connected to the internet and able to interact with each other
and plug into a variety of data-intensive cloud computing and social media
platforms and services. IoT has of course been described as bringing
“revolutionary” changes to everyday life and industry, while also raising new
kinds of concerns regarding security, privacy, distraction, and so on (Greengard
2015). These risks actually highlight another structural aspect of protocols and
interfaces: the parameters of a thing’s openness to connection. While these are
designed for certain types of intended connections, they can also enable others
that are subversive or even hostile. It is in this context that hacking and
cyberwarfare are our new realities, with vulnerabilities also networked and
distributed.
However, the technical novelty and surrounding hype can distract from
underlying dynamics that have historical precedent. Architect and writer Keller
Easterling (2012) points to the experiments of the architects Cedric Price and
Christopher Alexander with active form (in contrast to object form) as early
rehearsals of an IoT. This is a way of seeing the form of space itself as carrying
information through activity. It is a disposition in which the action is embodied
as potential in relations and relative positions, unfolding as performance over
time. Architect and theorist Lars Spuybroek makes a similar case in his reading
of John Ruskin’s study of the Gothic, which he updates as an ontology of the
digital (Spuybroek 2016). Gothic form was vital and active, harboring a reserve
of changefulness in its internal logic of growth and variation. Spuybroek sees
here a logic and aesthetic for the digital, relevant now more than ever.
Finally, this last point touches on one other noteworthy aspect of interfaces:
the aesthetic. Indeed, this is the subject of much work in design, a concern with
form ranging from the fit or interface between user and thing (ergonomics in the
physical sense, human–computer interaction in the more cognitive) to the
semiotic and the webs of associations and meanings that a form activates. And of
course the aesthetic is political, and computer software can be seen as having
parallels to ideology (Galloway 2012). Software and interfaces mediate (always
imperfectly) between the external and the internal, and operate in the realm of
simulation. They both reflect and obfuscate what goes on. Moreover, they are
functional: code is language that is executable. Discussing the tensions and
translations between code as language and machine, Galloway also offers
another observation that resonates intriguingly with our conception of fluid
assemblages:
The dialectical movement between fluidity and fixity, seen in the internal workings of software where
states and state changes carry the day, is precisely the same political problem posed by ideology …
Software might not be narrative in the strict sense of the word, but it still might have a beginning,
middle, and end—to paraphrase Aristotle—even if those narrative moments are recast as mere
variables inside the larger world of software simulation. Thus too might ideology be recast in digital
format. (Galloway 2012, 72–73)

Interfaces tell stories about what is going on in machines and in the aspects of
the world that they mediate, but they and the software underlying them also do
things themselves.
Protocols and interfaces are, then, rules and zones of connection, sites of
affecting and being affected. They are the connective tissue of systems and
ecosystems, the trading zones that enable flows and exchanges and cascading
effects. In (at least) computational systems, they are activated by algorithms.

Algorithms

Underlying the user-facing surfaces and interfaces of digital networked things is

the running code that makes them tick. There are computational processes
happening at the level of the device itself, and also at the level of interconnected
components, services, and infrastructures. So, we turn now to consider
algorithms from a perspective of considering the mediations and cultural work
that these computational entities perform.
 First, what are algorithms in terms of how they show up in the world? Ed Finn
begins his superb study of algorithms with a good roundup:
The word algorithm frequently encompasses a range of computational processes including close
surveillance of user behaviors, “big data” aggregation of the resulting information, analytics engines
that combine multiple forms of statistical calculation to parse that data, and finally a set of human-
facing actions, recommendations, and interfaces that generally reflect only a small part of the cultural
processing going on behind the scenes. Computation comes to have a kind of presence in the world,
becoming a “thing” that both obscures and highlights particular forms of what Wendy Hui Kyong
Chun calls “programmability.” (Finn 2017, 16)

Here Finn points directly at one of the big puzzles of fluid assemblages: how can
computation become a “thing” with a presence in the world? It is here at the
level of algorithms that we find a key part of the answer. It is also worth noting
how quotidian our interactions with algorithms have become. While many
operate outside of our awareness and understanding, we are also now
accustomed to algorithmic recommendations (of the form “since you enjoyed
this, you might also like … ”) and wondering about how the Facebook
algorithms determine what should be shown in our newsfeeds (Bucher 2012).
In considering algorithms, we need to first go even further back than the
algorithms at work in a single thing to consider the historical logic and
assumptions underlying them that guide what they do. Finn (2017) points to
magic and sorcery as forerunners of the kind of logic in which words actually do
things in the world, and to cyberpunk novels as explorations of how this logic
plays out in the (imaginary) cultural space of computation. In addition to this
mythological origin story is also the pragmatic engineering conception of
algorithms in terms of problems and solutions, which is now the main logic
governing computational development. This logic is also underpinned by what
David Golumbia (2009) refers to in his critique of “computationalism”: a belief
in the power of computing that tends to be associated with instrumental reason,
the “essential politics of the algorithm” (Finn 2017, 18). The assumptions about
the nature of the universe and how it can be dealt with that underlie the
pragmatic conception of algorithms Finn (2017) refers to as “effective
computability.” He traces it back to the Turing machine and its suggestion of
universal computation, and to attempts at mathematical proofs of effective
computability that seem to indicate a movement toward universal truth, but
always with a remainder that does not fit left over. Algorithms have, then,
always “encoded a particular kind of abstraction, the abstraction of the desire for
an answer” (Finn 2017, 25).
One key early site where notions of effective computability were developed
was in cybernetics. Cybernetics was developed after the Second World War as a
field concerned with studying messages as a means of communication and
control (Wiener 1968). It claimed to be a universal science, applying to and
explaining social as well as physical reality (Bowker 1993). Importantly, it was
founded on a probabilistic theory of messages, reflecting the shift from
Newtonian mechanics to probabilistic models of the physical universe earlier in
the century. This meant that completely accurate and certain models were never
a goal in cybernetics, since they were then known to not be achievable. Rather,
the goal was information and models that were adequate for some particular
purpose. This paved the way for computer simulations being seen as reasonable
objects of study in relation to all sorts of physical, social, and biological
phenomena (Finn 2017).
In addition to the “good enough” logic of probabilistic simulations, one other
significant legacy of cybernetics for our purposes is its focus on how entities use
information to adjust to their environments in order to maintain their integrity
(homeostasis) and resist decay (entropy). This includes both living and non-
living organisms. In the words of cybernetics founder Norbert Wiener (1968,
27):
It is my thesis that the physical functioning of the living individual and the operation of some of the
newer communication machines are precisely parallel in their analogous attempts to control entropy
through feedback. Both of them have sensory receptors as one stage in their cycle of operation: that is,
in both of them there exists a special apparatus for collecting information from the outer world at low
energy levels, and for making it available in the operation of the individual or of the machine. In both
cases, these external messages are not taken neat, but through the internal transforming powers of the
apparatus, whether it be alive or dead. The information is then turned into a new form available for the
further stages of performance. In both the animal and the machine this performance is made to be
effective on the outer world. In both of them, their performed action on the outer world, and not merely
their intended action, is reported back to the central regulatory apparatus.

Collecting information and feeding it back into ongoing operation is also a

central function of fluid assemblages. It is also worth highlighting some
characteristics of this kind of apparatus that Wiener points or alludes to in this
context. First, the fact that an organism collects information means that it is
equipped (we could say programmed) to do so, and moreover that the
information it is interested in collecting is that which helps it with furthering its
goals as an organism. Second, this information is not “neat,” but is rather
transformed by the apparatus. Third, this information is then fed back into
ongoing operations, and in a way that has an effect. We can then formulate each
of these as questions regarding things to look for in specific instances of fluid
assemblages (which can be seen as organisms in the language of cybernetics):
(1) what information is it equipped to collect; (2) in what ways does it transform
the information it collects; (3) how and where is the information fed back into
ongoing operations; and (4) what effect does this have in other parts of the
assemblage and its larger environment?
We will also return to consider algorithms in more detail in the following
sections as we scale up and out. But for now it is worth extending one more
plank to help us make the bridge to consideration at these higher levels.
Returning to Finn (2017), he suggests that the algorithm serves as the alchemical
realm where the material and symbolic “operate in productive indeterminacy”
(p. 34). He uses the ubiquitous progress bar as an example, since it both maps
“progress” in a way that we do not believe completely reflects underlying reality,
and yet does reflect the functional reality that the software installation is not yet
complete and nothing else is going to happen until it is. Turning then to the
implications, he states:
As our generally unthinking acceptance of the progress bar demonstrates, we are primed to accept
these magical calculations on multiple levels. We believe in the power of code as a set of magical
symbols linking the invisible and visible, echoing our long cultural tradition of logos, or language as
an underlying system of order and reason, and its power as a kind of sourcery. We believe in the
elegant abstractions of cybernetics and, ultimately, the computational universe—that algorithms
embody and reproduce the mathematical substrate of reality in culturally readable ways. This is what it
means to say that an algorithm is a culture machine: it operates both within and beyond the reflexive
barrier of effective computability, producing culture at a macro-social level at the same time as it
produces cultural objects, processes, and experiences. (Finn 2017, 34)

We will circle back to these aspects of algorithmic culture later on. But first,
we need to dive down one more level. As protocols and interfaces are activated
by algorithms, algorithms in turn rely on data and information as the material
that they process.

Data and information

At a basic level, computational technologies deal with data and information.

This is their raw material, the stuff that they process in various ways that enable
the complex behaviors they perform. It is also associated with communication,
as in the common use of the term “information and communication
technologies” (ICTs). This is also one point at which a distinction between a
technology seen as a tool and one seen as a medium for information and
communication becomes particularly blurry, as ICTs have been at the heart of
changes in organizational practices, information flows, and management of
personal affairs in what has been termed the “network” or “information” society
(Castells 2000; Kallinikos 2007). In short, they help get stuff done, and they help
people do things together. But the particular angle that we are concerned with
here is more to do with the role and character of information in itself within
computational systems.
As mentioned earlier, information emerged as a key concept of cybernetic
thought that also underlies computational technologies. Data is a similar concept
that typically signifies something like the smallest unit of meaning, whereas
information implies something that has been in some way rendered meaningful
—that can inform. Data requires abstraction. In order for the analog world to be
turned into data, details must be left out according to some kind of ordering and
representational scheme. The map cannot be coextensive with the territory, yet
through leaving out details the map becomes much more useful in other ways.
However obvious this observation may be, it is very important to keep this
dynamic in mind and to inquire critically into the origins of data. This is
especially the case when data shows tendencies of growing into big data,
expanding outside its initially intended scope of use, or traveling between
different contexts of interpretation.
For insight into some of these issues, we can turn to the now classic account
by science studies scholar Bruno Latour (1999) of how “circulating references”
are created in scientific practice. It is noteworthy that he does not really use the
term data—except to say that one “should never speak of data” (Latour 1999,
42)! Rather than considering data as what is given, he suggests speaking instead
in terms of “achievements.” Because data that is produced and that can lead to
better understanding is an achievement that involves a good deal of work.
Latour (1999) shows how this works in his study of scientists investigating a
border between savanna and forest in the Amazon to try to figure out whether it
is the savanna or the forest that is advancing while the other is on the retreat. In
order to collect the kind of data that will be helpful in answering this question
(plants, soil samples, and so on), they must first construct a grid to impose on the
area to be studied, literally unfurling lengths of string at the site in order to do so.
This allows for establishing correspondence between samples and sites from
which they originated. Another lovely ordering device is the pedocomparator, a
box for soil samples containing a grid of smaller cardboard boxes that can fit in a
drawer for storage but also turn into a suitcase for transporting samples between
field site and laboratory. Through this ordering, the scientists are able to both
label and order the samples and also make sense of them visually, being now
able to perceive in this condensed format a shift from forest to savanna in the
changing color.
Here there is more at stake for Latour than cataloging the practical activities
and helpful technological accessories of scientific practice. It is rather a much
more fundamental question regarding the connection between world and
language, matter and form. Whereas the “canonical view,” following Kant, sees
an unbridgeable gap between them in phenomena, Latour sees a chain of
mediations operating as circulating reference. The dirt from the Amazon can be
labeled with a coordinate, placed in a box in a grid, and become part of other
successively higher-level representations about what is going on at the border of
forest and savanna, some of which are published in a final report. While details
of the local context are lost and reduced in these transformations, others are
amplified as it becomes possible to compare many such references that are also
much better able to travel than the box full of dirt. And importantly, it is possible
to travel both ways, both up and down across this chain of mediations: from the
dirt to the dense charts in the report, but also from the chart back down to the dirt
samples and the location in the Amazon from which they came. These
mediations are thus each phenomena, a combination of transformation and
connection; and these phenomena circulate all along the chain of references.
While this chain of references may exist in followable forms in scientific
practice (at least ideally), things become more difficult outside of science. But it
is worth remembering that any data that somehow represents some aspect of the
world comes with this chain of mediations, which could, at least in theory, be
followed back to the point at which it bumps up against the phenomenon in
which it originated.
One of the primary mechanisms involved in these mediations between
phenomena and data is categories. Categories often serve as rather blunt
instruments for culling away the particularities of local phenomena in order to
make them fit larger ordering schemes in which they can be classified, collected,
and compared. Categories can also become standards that ensure data
compatibility and are used for regulation. They are also social, often messy,
contested, and political in both enactment and consequences. Having a category
for something enables collection of data that can make something visible, but
this also allows it to be more easily regulated and administered. These matters
have been investigated in depth by Geoffrey Bowker and Susan Leigh Star in
their important book Sorting Things Out: Classification and Its Consequences
(1999). In it, they examine work practices and consequences around
classifications such as the international classification of diseases (ICD), racial
classification by the apartheid government of South Africa, and classification of
nursing work.
There are a number of reasons why classifications are important, but one that
is particularly significant for our project here is the ways in which they can
become embedded in information infrastructures that are in turn embedded in
computational infrastructures. In this way, political and social struggles can be
rendered as technical and eventually come to be naturalized (Bowker and Star
1999). Standards and categories valorize some points of view while silencing
others. This is indeed necessary and inescapable, and not a bad thing in itself;
but it is always an ethical choice. The same dynamic could be seen in Latour’s
conception of circulating references discussed earlier, in which there is both
amplification and reduction in moving along the chain of mediations between
local particular and more abstract and universal reference. Even the concept of
data serves as a frame delimiting what will be included for consideration and
what will be left out, something that has significant consequences for
approaching inquiry (Boyd and Crawford 2012; Markham 2013).
Data and information are thus produced in ways that are not “natural” or
given, but rather reflect particular points of view and agendas in relation to
phenomena that change over time. Raw data is indeed an oxymoron (Gitelman
2013). It is always “cooked” in some way. It is (now stretching the metaphor to a
perhaps inadvisable extent) prepared by particular chefs working in and for
particular contexts, for consumption by particular entities with their own
preferences and budgets and temporal horizons.
We can also note that data collection now commonly occurs outside of
scientific and work practice, encompassing many aspects of our lives. Some
people even collect large amounts of information about themselves and their
activities through self-tracking tools, which has become known as the
“quantified self” movement. And then of course there is all the data that is
collected about us without our explicit (or fully understanding) consent when we
use things that are connected to the internet. It is now mightily challenging or
even impossible to escape dataveillance, what digital media scholar Rita Raley
describes as “the disciplinary and control practice of monitoring, aggregating,
and sorting data” (Raley 2013). This is especially the case with fluid
assemblages, which rely on data collection not only to customize their forms for
particular users, but also to inform their larger networks of other interested
parties that can later use it to more precisely target advertising, conduct
surveillance, gather information about a particular population for scientific or
regulatory purposes, and so on. Each function, each button or command, is thus
potentially a data point reflecting a particular classification. For example, the
architecture of Facebook now has a plethora of mechanisms by which users
indicate how they relate to each other and to content; particularly interesting in
this context is the addition of other “reactions” to the original “like” button,
which allow for more nuanced responses to content but also much more precise
information for Facebook. Links between things are also data—a significant
dynamic when it comes to online environments and networked things.
 Now, if these complexities are inherent to data as such, what happens as we
turn to “big data” where aspects of aggregation and thus to some extent practices
of filling the gaps with projections start to have a significant impact? To begin to
appreciate that quantity in this case also gets new qualities, consider the
following remark from an article in The Economist in 2017:
Uber, for its part, is best known for its cheap taxi rides. But if the firm is worth an estimated $68bn, it
is in part because it owns the biggest pool of data about supply (drivers) and demand (passengers) for
personal transportation. Similarly, for most people Tesla is a maker of fancy electric cars. But its latest
models collect mountains of data, which allow the firm to optimise its self-driving algorithms and then
update the software accordingly. By the end of last year, the firm had gathered 1.3bn miles-worth of
driving data—orders of magnitude more than Waymo, Alphabet’s self-driving-car division. (2017a)

In another article in the same issue, it is further argued that:

Access to data also protects companies from rivals in another way. The case for being sanguine about
competition in the tech industry rests on the potential for incumbents to be blindsided by a startup in a
garage or an unexpected technological shift. But both are less likely in the data age. The giants’
surveillance systems span the entire economy: Google can see what people search for, Facebook what
they share, Amazon what they buy. They own app stores and operating systems, and rent out
computing power to startups. They have a “God’s eye view” of activities in their own markets and
beyond. They can see when a new product or service gains traction, allowing them to copy it or simply
buy the upstart before it becomes too great a threat. Many think Facebook’s $22bn purchase in 2014 of
WhatsApp, a messaging app with fewer than 60 employees, falls into this category of “shoot-out
acquisitions” that eliminate potential rivals. By providing barriers to entry and early-warning systems,
data can stifle competition. (The Economist 2017b)

The gaps that can be filled by “big data,” and increasingly also by the new forms
of intelligence surrounding what is called “meta-data,” thus not only connect the
dots when it comes to potential customers, their needs and desires, but
increasingly also to the inherent logic of the market in question. While examples
such as the ones mentioned earlier will soon be dated and others added, there
seems to be underlying movement here that certainly points in a direction but
that is also hard to predict—indeed, that it would be a company emerging from
the plethora of software start-ups that eventually would shake the financial
ground of newspapers worldwide by offering a different way of advertising was
not something many anticipated. Perhaps it is therefore not so surprising that
obtaining a “God’s eye view” using “big data” also becomes big business.
To understand how such advantages are made and maintained using big data,
let us consider an example from the opposite end of the spectrum, that of people
forming also intimate relations using technological tools. In an article in The
Guardian, Judith Duportail describes the experience and results of requesting
and reading her data as registered by Tinder:
 I asked Tinder for my data. It sent me 800 pages of my deepest, darkest secrets. … At 9.24pm (and one
second) on the night of Wednesday 18 December 2013, from the second arrondissement of Paris, I
wrote “Hello!” to my first ever Tinder match. Since that day I’ve fired up the app 920 times and
matched with 870 different people. I recall a few of them very well: the ones who either became
lovers, friends or terrible first dates. I’ve forgotten all the others. But Tinder has not. … Some 800
pages came back containing information such as my Facebook “likes,” my photos from Instagram
(even after I deleted the associated account), my education, the age-rank of men I was interested in,
how many times I connected, when and where every online conversation with every single one of my
matches happened … the list goes on. (Duportail 2017)

Such combinations of data, however, do not only come together for the purpose
of the “perfect match,” but extend beyond this particular application:
The trouble is these 800 pages of my most intimate data are actually just the tip of the iceberg. “Your
personal data affects who you see first on Tinder, yes,” says Dehaye [privacy activist Paul-Olivier
Dehaye from personaldata.io]. “But also what job offers you have access to on LinkedIn, how much
you will pay for insuring your car, which ad you will see in the tube and if you can subscribe to a
loan.” (Duportail 2017)

Indeed, the notion of a “God’s eye view” based on big data might be suggestive,
but there is something quite real about what new possibilities are being opened
up through abstraction and aggregation here: how new structures spanning
multiple contexts not only can be registered but can actually be put in use in
runtime. Big data is still also just data that constantly feeds various algorithms
and feedback mechanisms—scenarios such as the one involving Tinder
discussed earlier is not the result of months of human analysis and intelligence,
of ethical and other considerations leading to decisions; it is the outcome of
machines running algorithms fed with data. It is also from this perspective we
need to consider what is now emerging as “artificial intelligence”—intelligence
understood not only in the sense of a capacity to learn or to reason, but
increasingly instead in the sense of intelligence agencies gathering information
to predict and control.
An interesting example of how big data is used to form new assemblages is
Netflix’s investments in producing original content. The origin of one of its early
success stories, “House of Cards,” is said to stem from data harvested pointing to
an effective intersection of viewing interests between political thrillers, director
David Fincher, actor Kevin Spacey and the older British TV series. In an
interview with The New York Times in early 2013, the company’s chief
communications officer Jonathan Friedland said, “Because we have a direct
relationship with consumers, we know what people like to watch and that helps
us understand how big the interest is going to be for a given show. It gave us
some confidence that we could find an audience for a show like ‘House of
Cards’” (2013). The article further comments that “Netflix has always used data
to decide which shows to license, and now that expertise is extended to the first-
run. And there was not one trailer for ‘House of Cards,’ there were many. Fans
of Mr. Spacey saw trailers featuring him, women watching ‘Thelma and Louise’
saw trailers featuring the show’s female characters and serious film buffs saw
trailers that reflected Mr. Fincher’s touch.”
With respect to just how detailed such decisions based on data might be, Phil
Simon in an article in Wired, the year after, uses the example of how the “House
of Cards” cover was designed down to the proportions of color hues:
At Netflix, comparing the hues of similar pictures isn’t a one-time experiment conducted by an
employee with far too much time on his hands. It’s a regular occurrence. Netflix recognizes that there
is tremendous potential value in these discoveries. To that end, the company has created the tools to
unlock that value. At the Hadoop Summit, Magnusson and Smith talked about how data on titles,
colors, and covers helps Netflix in many ways. For one, analyzing colors allows the company to
measure the distance between customers. It can also determine, in Smith’s words, the “average color of
titles for each customer in a 216-degree vector over the last N days.” (Simon 2014)

Indeed, this is a significant extension of the conceptual connection between

the “programming” of computers and the “programming” of television as data is
used to draw things together to form new assemblages and even define the
spaces in which acts of creativity are meant to, or have to, take place. In many
ways, this is an extension of what market, consumer, and user research has been
about since its inception, and how designers make heavy use of “user-centered
design methodology” to narrow down what might and might not work for a
given target group. And still, from the wider perspective on fluid assemblages
discussed here, this is moving beyond relations between forms of
“programming” toward creating a close conceptual relation also between
ensemble and assemblage.

Technological ecologies

Some technological systems are designed from the top down, perhaps most
commonly organizational systems where use and associated routines can be
enforced bureaucratically. But in more general terms, things always exist in
some kind of relation to other things, and they can be pulled together into ad-hoc
“systems” in use. These evolving relations and practices can then in turn feed
into more explicit design to support these interconnections in better ways. This
recognition has been labeled with terms like “product ecology” (Forlizzi 2008)
and “device landscapes” (Stolterman et al. 2013) in the field of human–computer
interaction.
 However, there are already existing examples of intentionally designed and
curated technological ecologies. Perhaps the most well known is Apple. Its
ecosystem is certainly designed (and controlled) in a quite top-down fashion, yet
this has also been done by implementing connections among its devices. This
has been accomplished in large part through its user accounts, currently on the
iCloud platform. This system not only enables the syncing and transfer of data
and preferences across devices in current use, but also between old and new
devices. It is a system that allows evolution both in terms of Apple’s own
collection of devices and also in terms of an individual user’s collection as it is
expanded and upgraded over time. And indeed this constant upgrading and lack
of attachment to any particular device is very much part of Apple’s business
model, and the rapid outdating of devices accompanied by much-hyped and
“best ever” updates part of every Apple device owner’s experience.4 This
dynamic can also be seen in similar platforms, such as Google and the suite of
interconnected applications and devices it supports.
Another mechanism for supporting development in technological ecosystems
is the application programming interface, or API. These open protocols basically
allow apps to become services for other apps, and to plug in to certain kinds of
functionality that they offer. One notable early example is Twitter, whose API
allowed an ecosystem of associated apps and services to grow up around it.
Other types of mechanisms that are even further beneath the surface support
things like cross-platform authentication and content serving and customization.
This ecological structure of artifacts evolving in relation to each other and
their use can be seen at a more foundational level in philosopher of technology
Gilbert Simondon’s structural analysis of technology (Simondon [1958] 2017).
He looks at technical objects in terms of their genesis and emergence of
consistency and convergence at three levels: the element, the individual, and the
ensemble. He sees technologies (such as a car engine) starting as rather abstract
ideas with each component conceived more or less independently and strictly in
terms of its function. However, when technologies are actually brought into
existence these structures have reciprocal influences on each other during
operation (through generated heat or motion and so on, in the case of
thermodynamic machines). This means that in order for it to function properly
these components need to work together in harmony, and pressure for good and
efficient operation can lead to a convergence of functions in components that
might have originally been conceived of as doing one thing in the more
“abstract” initial version of the technology. The end result of this evolution
through a series of technical objects is “a system that is entirely coherent within
itself and entirely unified” (Simondon [1958] 2017, 29).
 Simondon thus saw this internal structural rationality of technical evolution as
something leading toward the most essential and concrete expression of a
particular technical idea (as opposed to the competing “inessential” and
superficial demands of customers for particular kinds of superficial styling).
However, addressing similar but somewhat more contemporary dynamics,
virtual reality and digital media pioneer Jaron Lanier (2010) laments what he
refers to as technological “lock-in” through software, where existing structures
guide further development in ways that effectively cut off other possible paths.
Lock-in is, he says, “like a wave gradually washing over the rulebook of life,
culling the ambiguities of flexible thoughts as more and more thought structures
are solidified into effectively permanent reality” (Lanier 2010, 9). An example
he uses is the MIDI format, a rigid structure for representing musical notes that
previously “transcended absolute definition” (Lanier 2010, 9).
The particular dynamics of technological ecosystems vary, and there are
certainly significant differences between the industrial technical objects
Simondon considered and the digital ones that Lanier did. However, for present
purposes, it will suffice to note in summary that technical artifacts interact with
each other both directly and indirectly in complex ecosystems; and that these
interactions influence both configurations at particular moments in time in
particular contexts and the ways in which both artifacts and ecosystems develop
over time.

Platforms and infrastructures

Other concepts that are often used in relation to systems and ecologies are
platforms and infrastructures. These terms refer to the relatively stable structures
that provide common resources for the operations of other things. For example,
classic examples of industrial infrastructures would include water and sewer
systems, transportation networks, electrical grids, and telecommunications
systems. The internet and World Wide Web have now effectively become
infrastructures in the sense that they are common resources that are used for a
vast array of processes and activities across sectors and industries, work and
everyday life.
However, the common sense understanding of infrastructure as inert substrate
is not quite accurate. Research in science and technology studies has shown that
infrastructure is fundamentally relational, becoming infrastructure in relation to
organized practices (Star and Ruhleder 1996; Star 1999). In a seminal article,
Star and Ruhleder (1996) proposed a set of characteristics of infrastructure: that
it is embedded in other structures and sociotechnical arrangements, transparent
in use, having a reach or scope beyond one particular site or event, learned as
part of membership in a group or organization, linked with conventions in
particular communities of practice, embodying standards that allow it to connect
to other infrastructures and tools, built on an installed base with a certain inertia
that resists change, and becoming visible on breakdown.5
The concept of platforms is somewhat similar, but has its roots in media
studies and focuses on a somewhat different set of issues. While infrastructure
studies “has highlighted key features of infrastructure such as ubiquity,
reliability, invisibility, gateways, and breakdown,” key features in platform
studies “include programmability, affordances and constraints, connection of
heterogeneous actors, and accessibility of data and logic through application
programming interfaces (APIs)” (Plantin et al. 2017, 2). The term “platform”
itself has also served a key role in the discourse around networked computational
systems and services, as technology companies use the term flexibly to position
their offerings in relation to various contexts and concerns (Gillespie 2010). A
social media platform can be a “platform” for self-expression and even
democracy, but then also “just a platform” when questions about regulation
applying to media content providers come to the fore.
As Plantin et al. (2017) have argued, both concepts are needed in order to
understand entities such as Google and Facebook. While they have many
characteristics of platforms, being highly programmable and supporting a range
of apps, platforms and platform ecologies have also become in practice like
large-scale infrastructures. For example, in the case of Facebook:
Viewed simultaneously as an infrastructure and platform, Facebook presents a disturbing image. As an
infrastructure, Facebook is progressively expanding and embedding itself in our daily existence, taking
over more and more functions formerly provided by other, less restrictive means. The API, as a
gateway, transforms Facebook from a centrally controlled system into something more like a network
of independently developed, yet seamlessly interconnected systems and services. As a proprietary,
largely opaque platform, Facebook filters our daily communicative acts through a profit-extracting
sieve, deploying its intimate view of users’ activities and relationships for the benefit of advertisers and
others, who in turn provide further data (via the API) for the Facebook social graph. As a result, its
power to shape our communication behavior for its own ends increases. (Plantin et al. 2017, 12)

The means by which something like Facebook acts as a “profit-extracting

sieve” are particularly relevant from a design perspective, since they are
embodied in the interactive mechanisms that define the thing as it is available for
use. Alaimo and Kallinikos refer to this as encoding, the organizing of “user
platform participation along specific activity corridors (such as sharing,
following, or tagging) that heavily stylize and shape user interaction” (Alaimo
and Kallinikos 2017, 176). Other familiar examples of encoding are emoji
characters, standardized representations of emotion that both facilitate affective
communication and commoditize affective labor for the market (Stark and
Crawford 2015). After being encoded, data about user activities on a platform
are then aggregated into larger entities, which are subjected to further
computation that leads to higher-level descriptions of platform activity (and of
users and their relationships, as operationally defined by their platform activity)
(Alaimo and Kallinikos 2017).
Another view of things is from the perspective of “big data,” which seems to
have become the contemporary instantiation of the information-driven “network
society” (Castells 2000). Specifically, this data that is the new fuel for all kinds
of societal processes needs to come from somewhere. This creates a drive for an
increasing number and penetration of (mostly passive) data-collecting sensors in
everyday life—a key means by which big data becomes embedded infrastructure
(Burdon and Andrejevic 2016). The passive nature of much of this data
collection provides a direct challenge to design perspectives focused on user
experience, since in this case things enter (and can ultimately affect) people’s
lives while—often by design—not entering into or even being hidden from
conscious experience. Given these dynamics, design perspectives turning toward
connecting products within systems in order to provide services (e.g., Dubberly
2017) are a step in the right direction, but still fall remarkably short of
adequately accounting for the (politico-economic) role and function of
connected things in contemporary (post)industrial contexts.

Societal
We have already in the previous sections encountered societal dimensions of
things as they are part of our lifeworlds and as they exist as technologies that do
particular things in the world. Now we turn to face these dimensions squarely,
looking at the character and implications of things as both products of and actors
in society.

Politics and actor-networks

Critiquing the conventional view of things that considers them only in terms of
their making and use, political theorist and philosopher of technology Langdon
Winner has argued that, rather than being only neutral aids to human activity,
they are also “powerful forces acting to reshape that activity and its meaning”
(Winner 1986, 6). Moreover, he asserts that as “they become woven into the
texture of everyday existence, the devices, techniques, and systems we adopt
shed their tool-like qualities to become part of our very humanity” (Winner
1986, 12). Seen in this light, technological design becomes a historical process
of world-making, a perspective much in line with current scholarship in design
studies (Yelavich and Adams 2014; Fry, Dilnot, and Stewart 2015). It also points
to a different set of questions than those that are typically asked about
technology. As Winner puts it:
From this point of view, the important question about technology becomes, As we “make things
work,” what kind of world are we making. This suggests that we pay attention not only to the making
of physical instruments and processes, although that certainly remains important, but also to the
production of psychological, social, and political conditions as a part of any significant technical
change. Are we going to design and build circumstances that enlarge possibilities for growth in human
freedom, sociability, intelligence, creativity, and self-government? Or are we headed in an altogether
different direction? (Winner 1986, 17)

Winner (1986) also addresses head-on the fraught question of whether artifacts
“have politics.” Although it is not terribly difficult to see that they can be used
for political ends, the idea that there can be something inherently political about
certain technologies is much more controversial. As one example of how this
might be the case, he offers the now-classic case of the low-hanging bridges in
New York.6 These were part of the significant developments in the city’s
infrastructure that were led by master city planner Robert Moses. One effect of
these low bridges was to prevent buses from using the parkways. Moses was,
according to his biographer, a racist and a classist, and through the design of his
bridges he was able to effectively limit the mobility of poor people and blacks
who most relied on public transit, in particular discouraging their use of his prize
development, Jones Beach. This was especially significant in that the bridges
enforced the prejudices of their designer even after he was no longer around.
The powerful role of technologies in effectively deciding particular states of
affairs suggests that they be brought under the same kinds of conditions for
public decision making as is the case with other forms of legislation. As
philosopher of technology Andrew Feenberg states (building on Winner):
The technical codes that shape our lives reflect particular social interests to which we have delegated
the power to decide where and how we live, what kinds of food we eat, how we communicate, are
entertained, healed and so on. The legislative authority of technology increases constantly as it
becomes more and more pervasive. But if technology is so powerful, why don’t we apply the same
democratic standards to it we apply to other political institutions? By those standards the design
process as it now exists is clearly illegitimate. (Feenberg 1999, 131)
Now the kinds of design that Winner and Feenberg discuss tend to be more along
the lines of public infrastructure than what might be labeled as industrial design
of individual products for mass production. Yet at the same time the differences
might be more of scale than of kind, as even individual products can embody
and enforce particular ideas about how to do things, and also connections to
other elements within systems of varying levels of intentional design and
formality.
It is also the case that categories, discussed earlier as mechanisms used for
producing data, can have politics, encoding particular models of the world in
persistent technical forms that have consequences (Suchman 1994). As Susan
Leigh Star states, explicitly referencing Winner’s low-hanging bridge example:
“There are millions of tiny bridges built into large-scale information
infrastructures, and millions of (literal and metaphoric) public buses that cannot
pass through them” (Star 1999, 389). And as addressed in the previous section,
the information infrastructures of contemporary data-driven society involve
encoding activity and steering interactions through particular platform channels.
This means that when designing a user interface for interactions that connect to
and through these kinds of platforms, it really is politics all the way down.
In summary, looking at technological things in terms of their politics means
considering the ways in which they are implicated in distributions of power and
authority, and in supporting or foreclosing certain forms of life. It does not mean
politics in the sense of governments, elections, political parties, or similar—
although of course these particular mechanisms of politics are also vitally
important, and indeed all involve technologies. But there is also a more
foundational sense of the political as the collective project of figuring out how to
live together well. This involves decisions about how to provide for common
needs, to ensure opportunities to engage in meaningful work, to support health
and wellbeing, and generally to create the kind of conditions that are conducive
to the collective thriving of life on the planet. These are issues in which designed
technological things are very deeply implicated, and a proper consideration of
their social implications means inquiring into their political dimensions also in
these larger senses.
One way of approaching this is in terms of actor-network-theory (ANT), a
sociological approach that, like (and in response to) social constructivism, was
developed in science and technology studies. In contrast to much sociology, and
as described by leading proponent Bruno Latour (2007), it refuses to use the
abstract notion of “the social” or broad concepts such as “power” as explanations
for certain states of affairs. Rather, actor-network theory views particular forms
of more or less durable social order as remarkable accomplishments to be
explained. It does this by looking for actors that have particular effects in the
world and serve as mediators between other actors. In identifying actors it is
careful to dissolve any pre-existing conceptual distinctions between human and
nonhuman, science and culture, subject and object, and so on. It is relentlessly
descriptive rather than explanatory, at least in terms of explanation done by
“adding” or “framing” analyses through some additional theoretical apparatus
that does not emerge through following the actors themselves and accounting for
their own sense-making processes (Latour 2007). Rather than “society,” ANT
sees imbroglios and collectives of humans and nonhumans that exchange their
properties in distinctly non-modern ways (Latour 1999).
One of the distinctive aspects of ANT that makes it particularly relevant for
our purposes here is that it considers nonhumans to be actors as well as humans.
For example, a speed bump is an actor that has a particular kind of agency when
it comes to getting drivers to slow down (Latour 1999). City officials who want
drivers to obey a certain speed limit can use other mechanisms to achieve this
result, such as posting speed limit signs that are backed up by law and the
possibility of fines and social censure if caught not obeying them. But the speed
bump, or “sleeping policeman” as the French call it, enforces this limit in a
material way that enlists drivers’ compliance not only through a morally
grounded appeal to obey the law, but by addressing their selfish desire to
maintain the integrity of their cars’ suspensions. In this case the city officials
who want to achieve this speed regulation delegate the job of enforcement to the
speed bump.
Through work in science studies that showed the construction (and for ANT,
not only social construction) of scientific facts, it came to be caught up in the
“science wars” of the 1980s and accusations that it was attempting to tear down
the credibility of science and reduce it to nothing more than “social
construction.” Reflecting on and responding to this situation, Latour articulates a
role for a different kind of critique and critic. Drawing on Heidegger’s notion of
things as gatherings, and also in the old political sense of the Thing as
parliament, he suggests that “things have become Things again, objects have
reentered the arena, the Thing, in which they have to be gathered first in order to
exist later as what stands apart” (Latour 2004, 236). Rather than dogmatic
matters of fact, he argues for a renewed attention to matters of concern (Latour
2004) and thing-centered political assemblies for making matters of concern
public (Latour 2005). This notion of things as gatherers and representations of
matters of concern has also been picked up in design as a way to describe and
envision its political role in enabling democratic participation and the actual
doing of politics (Binder et al. 2015; DiSalvo 2012).
 ANT can provide a useful methodological orientation and sensitivity for our
enterprise here. It points to the gatherings that enable and are manifested in
things that are political matters of concern—and if this the case for all kinds of
things, it is especially so for fluid assemblages. For example, we might try it out
with Uber, following the various actors that are involved in holding this
particular assemblage together. An initial roundup might include the company
and its employees, drivers, passengers, apps for drivers and passengers,
regulatory frameworks, advertisements for Uber, journalistic coverage,
lawmakers and regulators, citizens, Google maps that includes Uber as a
transportation option, and so on.

Techniques

A recognition of these political dimensions of technologies is hardly new. Many

who can be seen as early theorists of technology were highly critical of it, seeing
it as a force that impacted society and even humanity in deleterious ways.
Moreover, they typically saw it as an overpowering monolith, Technology (with
a capital T) rather than specific technologies. Perhaps the classic expression in
this vein was that of the late Heidegger. He challenged what he called the
“anthropological view” that has an instrumental conception of technology as
something only used as a means in human activity. Rather, he asserted:
The manufacture and utilisation of equipment, tools, and machines, the manufactured and used things
themselves, and the needs and ends that they serve, all belong to what technology is. The whole
complex of these contrivances is technology. (Heidegger 1977, 4–5)

Heidegger provocatively proclaimed, moreover, that the essence of technology

is not technological, but rather a particular mode of revealing of the world
—enframing—such that it comes to be seen only as a standing reserve of
resources to be used. He uses the example of a hydroelectric plant built on the
Rhine that turns the river into merely a source of energy, which he sees as vastly
different from the old wooden bridge that previously connected the two banks
(Heidegger 1977). Heidegger’s nostalgic tendencies are on full display in this
example, which can be (and has been) easily critiqued in its particulars. Yet his
forceful portrayal of technology as something more than technological remains
compelling, and particularly useful in setting us out on a path toward not-only-
instrumental considerations of technologies.
At a more down-to-earth level, it is also understandable to some extent how
those living in the wake of the Industrial Revolution and the rapid
industrialization of the twentieth century might regard technological
development with some degree of suspicion and trepidation, given the massive
reordering of society that it entailed and that was due in large part to new
technologies and associated forms of production. However, even if it is plain,
particularly after the “empirical turn” in philosophy of technology (Crease 2001;
Brey 2010), that the details of particular technologies and contexts are highly
significant, there are still insights we can glean from these more general, and
generally pessimistic, critiques. One such concept we can use as a marker for
some of these issues, and also a pointer to more concrete matters to investigate,
is technique.
The sociologist Jacques Ellul, writing in the middle of the twentieth century,
defined technique as “the totality of methods rationally arrived at and having
absolute efficiency (for a given stage of development) in every field of human
activity” (Ellul 1964, p. xxv). Discussing specifically the machine as the most
striking manifestation of technology of the time, Ellul pointed to technique as
that which “integrates the machine into society” (Ellul 1964, 5) and in so doing
imposes the efficiency and rationality of the machine in all domains of human
life. In other words, a technology is not just the physical object itself, but also all
of the associated procedures and social arrangements that are involved in its
functioning.
Writing a few decades earlier, Lewis Mumford described the “cultural
preparation” that precedes actual technological developments and makes society
amenable to the techniques of a particular machine (Mumford 2010). He points
to the clock as the key machine of the industrial age, and monasteries as sites
where the necessary cultural preparation for it occurred. For the monasteries
“helped to give human enterprise the regular collective beat and rhythm of the
machine; for the clock is not merely a means of keeping track of the hours, but
of synchronizing the actions of men” (Mumford 2010, 13–14). He suggests that
the Church’s “contempt for the body” (Mumford 2010, 35) and attempt to
mortify and subdue it was another key factor in preparing society to submit to
subjugation by machines.
Perhaps the most thoroughly developed and incisive critique of modern
industrial society was written by Herbert Marcuse in his magnum opus One
Dimensional Man (1964). Working in the tradition of the Frankfurt school of
social theory, he integrated philosophy, social theory, and politics in order to
develop a critical account of tendencies in highly developed societies. He
diagnosed modern society and thought as “one-dimensional” due to the all-
encompassing technical rationality that both served and produced man’s needs
while at the same time incorporating or rendering impossible any meaningful
critique. Rather than being rational, Marcuse argued that the technical apparatus
of production and the associated ordering of the social world was actually at
bottom highly irrational; and that rather than leading to individual liberty, it
effectively enslaved people as instruments in the machine. Although technical
rationality produces a relatively high standard of living (for some), it comes at
the price of a loss of liberty, autonomy, and possibility to realize the highest
human possibilities. Moreover, this system precludes the possibility of any
meaningful alternative, such that even apparent oppositions only serve to
reinforce the existing arrangement. For Marcuse, the only hope (and it is a small
one) comes through making the “Great Refusal” of the current order of things.
Then, perhaps, other alternatives might be imagined.
The essentialist, substantive theories of technology of which perhaps
especially Ellul and the late Heidegger are typical are unabating in their
pessimism, the heart of which is the conception of technology (in the broad
sense including techniques and so on) where the technical means are inextricably
linked to ends and particular (undesirable) forms of life. The only hope, in this
view, is for some kind of awakening and revealing of another ground from which
to imagine other possibilities, perhaps coming from art. However, this portrayal
of the almost total lack of human agency is not adequate for envisioning a
program of technological reformation. This is the diagnosis and project of
philosopher of technology, and student of Marcuse, Andrew Feenberg, who
places agency front and center in his Critical Theory of Technology (Feenberg
1999, 2002).
Feenberg follows earlier critical theorists in pointing to the prevalence of
technocratic forms of administration that are legitimated through reference to
scientific expertise, but also rejects essentialism in his conceptualization of the
ambivalence of technologies—the possibilities for them to be developed and
used in a variety of different ways and with different social consequences. His
concern is the “subordinate position” that humans typically have in relation to
the “technical systems that enroll us,” and he argues that we need to “begin to
intervene in the design process in the defence of the conditions of a meaningful
life and a livable environment” (Feenberg 1999, p. xiv). This process of
democratic participation in the design of technical systems he refers to as
democratic rationalization, which involves people who are “engaged in
technically mediated activities and able to actualize ambivalent potentialities
suppressed by the prevailing technological rationality” (Feenberg 1999, 105). By
this means technical systems could be grasped and changed from within by those
who understand and are affected by them, and thus able to envision different
possibilities.
 While the techniques of the early and mid-twentieth century might be most
readily associated with large industrial machines, these are not the most
prominent currently. Rather, we might think of large data-driven organizational
and administrative systems and even social media platforms that provide the
frames within which much work and human interactions of all kinds take place.
As one particular example, we might think of the rise of what has become
known as the “gig economy,” where digital platforms enable efficient matching
between tasks and humans available to do them. These platforms make human
labor available outside of the confines of stable and regulated employment
relations, with predictable consequences for worker rights and economic
security. One dynamic that has become characteristic of these arrangements are
information asymmetries between those who own these kinds of platforms and
those who do the work. This was highlighted in the case of Amazon’s
Mechanical Turk by the Turkopticon project, which allowed workers who
perform human computation tasks in the system to share information with each
other regarding their experiences with particular employers (Irani and Silberman
2013). Another example is Uber, which leverages intentionally designed
information asymmetries to extract a profit from the difference between the
routes and “upfront” fare prices it shows passengers and the most efficient routes
shown to the drivers that are used to calculate their pay (Kravets 2017).
More generally, our data-driven economy is now fueled by digital labor that is
only beginning to be conceptualized as such (The Economist 2017a). The
dominant techniques of our age are arguably those involving the collection,
processing, and leveraging of massive quantities of data, the creation of which is
mediated in large part by fluid assemblages manifested as things available for
use.

Political economy

Another trajectory that we can follow in approaching especially the political and
economic aspects of our current sociotechnical landscape comes from media and
cultural studies. Of course there is very much more involved in terms of the role
of media and culture in society, but it is the political economy of media that
seems particularly salient in relation to fluid assemblages.
As suggested by the term, political economy refers to both political and
economic dynamics at play in the production and distribution of media, as well
as consumption and use. And “at the heart of studies of political economy lies
the question of how social resources are controlled and by whom” (Kellner and
Durham 2012, 15).
What is especially significant in the context of fluid assemblages is the fact
that the core product of media industries is not the media content itself, but
rather the audience for that content that is then sold to advertisers as a
commodity (Smythe 2012). In the US context at least, the ratings industry that
developed to report on the number of viewers of network television shows
mediated the production and market for the commodity audience. This market
also came to define “the audience” (meaning the one attractive to advertisers) as
white men aged 18 to 34, reflecting entrenched (but inaccurate) stereotypes
about economic and societal power (Meehan 2012). This structure thus
determined the types of content that were prioritized.
The contemporary media landscape in which the audience commodity is
produced is much more diverse than television, and its development has been
characterized by processes of convergence among “old” and “new” media forms
and increasing participation by people making up the group formerly known as
the audience (Jenkins 2006). However, the underlying model has remained
remarkably consistent; what has changed are the increasingly fine-tuned
mechanisms and diversity of channels through which attention is aggressively
bought and sold by what legal scholar Tim Wu calls the “attention merchants”
(Wu 2016). The contemporary media landscape involves advertising platforms
that are built on top of search engines (especially Google) and social networks
(especially Facebook) and apps. In this sense we are also back in a situation in
which a few big players dominate the market and the way it is structured, as well
as playing a major role in society in other ways. For example, media scholar
Siva Vaidhyanathan (2012) has warned about the consequences of Google’s
dominance over the internet and involvement in so many of our personal
activities and societal functions.
One other thread that is worth picking up on here is the way in which the role
and agency of the “audience” or “users” are seen. Scholars associated with the
“Frankfurt school” of social research, working during and in the aftermath of the
Second World War, saw media and other cultural institutions as mechanisms of
mass control that reinforce existing social hierarchies and state authority. One of
the strongest statements in this vein was by key figures Theodor Adorno and
Max Horkheimer, who excoriated the “culture industry” as “mass deception”
(Adorno and Horkheimer 1997). In reaction to this view of people as passive
receivers of messaging, audience studies emerged as a way to open up for
audience agency. Messages were not only “encoded” into media content by
producers, but also “decoded” by the audience who could provide their own
alternate readings (Hall 2012). It is interesting to note this as one framing of
agency: one in which the content remains the same and freedom comes from
positioning oneself in relation to it. This seems not entirely dissimilar to the
agency of being able to accept or reject terms of service (and thus use) for an
application, or choosing which content to post and share within the structures of
a social media platform.
A key political economy question is: Who controls the means of production
and consumption of content? We now can and need to ask new kinds of
questions about who controls the means of production in this new media
landscape. The “democratization” of media participation means not only that
private citizens can generate and share media content, but also that, for example,
the Russian government was able to expertly leverage and manipulate social
media platforms in a sophisticated attempt to influence the outcome of the 2016
American presidential election.7
The internet itself is also a basic underlying “means of production” at this
point, but it too can be reconfigured—as evidenced in persistent attempts to
undermine or eliminate the principle of equal access to the internet known as net
neutrality. These platforms are also involved in much of the basic functioning of
fluid assemblages. In this sense, it is no longer only information, advertising, and
cultural texts that are implicated, but also things themselves. The developmental
trajectories of media industries and industrial design are thus forming a new and
significant point of convergence.

Algorithmic culture

Algorithms—the animating logic of computational systems—are becoming ever

more significant in shaping societal processes. Although they have until now
been considered primarily within computer science contexts and applications,
their cultural significance is now hard to overlook. So just how do algorithms
participate in the production of culture?
One way in which they participate, as argued by media scholar Tarleton
Gillespie (2014), is in the production and certification of knowledge,
representing a particular “knowledge logic” with particular assumptions about
what counts as knowledge and what its most relevant components are. Gillespie
calls these public relevance algorithms, and he outlines six dimensions that
“have political valence” (Gillespie 2014, 168):

1. Patterns of inclusion: the choices behind what makes it into an index in the
first place, what is excluded, and how data is made algorithm ready.
2. Cycles of anticipation: the implications of algorithm providers’ attempts to
thoroughly know and predict their users, and how the conclusions they draw
can matter.
3. The evaluation of relevance: the criteria by which algorithms determine
what is relevant, how those criteria are obscured from us, and how they
enact political choices about appropriate and legitimate knowledge.
4. The promise of algorithmic objectivity: the way the technical character of the
algorithm is positioned as an assurance of impartiality, and how that claim is
maintained in the face of controversy.
5. Entanglement with practice: how users reshape their practices to suit the
algorithms they depend on, and how they can turn algorithms into terrains
for political context, sometimes even to interrogate the politics of the
algorithm itself.
6. The production of calculated publics: how the algorithmic presentation of
publics back to themselves shape a public’s sense of itself, and who is best
positioned to benefit from that knowledge.

Each of these dimensions offer inroads into investigating the cultural

consequences of algorithms. Most of them also correspond to design decisions
about the ways in which fluid assemblages can be manifested as things available
for use, and how these things reveal or conceal the algorithmic functioning of the
larger assemblage. However, it is worth remembering that concealment of
algorithms in “black boxes” has so far been the norm (Pasquale 2015).
Algorithms are also shaping the production of even more traditional cultural
forms. One noteworthy example, already discussed earlier, is Netflix’s
production of “House of Cards.” It was made based on a prediction of relevance
for fans of certain types of content, but also, and significantly, because Netflix
knew that it would be able to distribute the content precisely to, and packaged
precisely for, these audiences. This was enabled by the shift Netflix made from
sending out DVDs in the mail to streaming content, which meant that they were
able to track user behavior in real time. As Finn puts it:
Now Netflix can track precisely how their customers watch particular shows, how long they hesitate
between options, and perhaps even how much pausing, fast-forwarding, or rewinding goes on. The
instant gratification of streaming creates a different kind of rating relationship—not the evaluation of a
film I watched last week or ten years ago, but right now. Netflix is no longer constructing a model of
abstract relationships between movies based on ratings, but a model of live user behavior in their
various apps. (Finn 2017, 91)

This is what Finn (2017) calls the aesthetics of abstraction: simplifying complex
and messy details in order to provide reliable personalized services. At least
Netflix lets users in on the ways in which their experience is being customized,
using tags like “because you liked ….”
At the heart of these new forms of interaction is what Finn calls algorithmic
arbitrage, exploiting the gap between computation and culture as Netflix did in
order to precisely identify the market gaps it could fill with House of Cards and
its other series. But the important thing here is not just the success of this
particular show, but the underlying trajectories it signals.
As computational systems become more efficient, and the patina of personal data we leave behind us
grows thicker, the presence of this arbitrage in our cultural lives is rapidly expanding, and beginning to
reinvent what the eternal consumer present, the moment of “now,” actually means. After all, there are
billions of dollars changing hands over the question of who gets to construct the present for you. When
you access a website, perhaps to find out what is happening in the world “right now,” hundreds of
servers are involved in auctions lasting fractions of a second to determine which advertisements will
appear on the page, and maybe even organize its content according to models predicting your interest
in different topics. (Finn 2017, 110–111)

In the language we have been developing here around fluid assemblages, we

could say that Netflix and other things that are made available for use are
literally made in a particular moment for a particular user, as defined by a
particular constellation of data points. Finn (2017) argues that we need to
cultivate algorithmic imagination as a practice of trying to grasp these existing
currents and to imagine what kinds of generative human and machine
collaboration could be possible. Something very similar could be said for design.
What types of creative design practices might be appropriate for working with
algorithmic processes as co-creator and co-producer? How might we understand
algorithmic form and aesthetics in ways that maximize their best possibilities,
while avoiding being swept into service of their worst tendencies and most
powerful stakeholders?

Notes
1 The original German word Zeug has been translated into English as both “equipment” and “useful
things.” The Stambaugh translation revised by Schmidt (Heidegger 2010) that is used here refers to
useful things, but we here work with equipment as the overarching term for referring to the issues
Heidegger raises. This is in keeping with how it has been used in relevant discourses, and it also seems
to be the most precise and evocative word for our purposes here.
2 https://www.apple.com/iphone-x/, accessed November 17, 2017.
3 It is standard in North America to have a central heating system in which the machine that does the
heating is located in a utility room or basement. It blows warm air through air ducts and out of vents into
the living space. The desired temperature is set at a small dial or control panel on the wall.
4 For the uninitiated, a helpful and quite accurate primer on the experience of owning Apple products can
be found at http://theoatmeal.com/comics/apple.
5 Here can be seen the Heideggerian framing of a thing withdrawing and becoming invisible during use
and becoming present and visible on breakdown, discussed earlier in the section on Equipment.
6 The accuracy and validity of this account have been disputed, but since it provides a compelling and
feasible case of technology serving a political function it still makes a good example.
7 A collection of relevant coverage can be found, among other places, at https://nyti.ms/2kbHS5m. Some
specific examples can be seen in Frenkel (2017) and White (2017).
 8

Assembling an Analytic Playlist

Earlier in this book, we took a historical journey through technological

developments using the simple act of pressing play to listen to music in order to
see what is becoming of things. In what follows we will return to the example of
using a “thing” to listen to music. This time, however, we will just look into one
particular example of a fluid assemblage: Spotify. Indeed, wordings such as
“particular example” are in a way misleading, since what “Spotify” actually
refers to is perhaps more the central organizing node in the fluid assemblage that
it instantiates. But at the same time, a conversation like this makes perfect sense
to us:

– Do you use a record or cassette player to listen to music?

– No, I use Spotify.

And in this way, it actually stands forth as a kind of “thing” to us. Now, let us
take a look at how.
Already when launching the application it is possible to identify elements that
are assembled: the fact that it is an application running on a particular device and
operating system, that an internet connection is required in order for it to
function, and that it is necessary to log in to an account before it is possible to
use it. There is also an option to log in using Facebook account credentials, thus
already establishing another connection in the assemblage. The next step after
logging in the first time, and periodically thereafter, is to accept (updated) terms
of service and privacy policy, which is necessary for and implied by continuing
to use the service. This means that one has entered into a standing legal
agreement before even listening to music.
On the first main screen one then sees a number of featured playlists, some
labeled as “sponsored” by a particular entity. In addition to the sponsored
playlists are ones selected for a particular day and time and activity, for example,
commuting home on a Monday evening or relaxing on a Saturday morning. This
is followed by sections for top lists, new releases, personalized
recommendations, and various genres. The genres include what one might
typically think of as music genres, such as pop, rock, metal, soul, and so on, but
also others that reference a particular mood or activity, such as sleeping,
traveling, dinner, or gaming. There is also a category called “Fresh Finds” that
includes playlists with selections “surfaced by Spotify’s tastemakers.” The
“Trending” category contains playlists that are currently popular. All of these
things combined give a strong impression that Spotify is continually updated,
that it is not merely a reliable tool for playing music but also a conduit through
which one is presented with fresh content. This content is both actively selected
and curated by Spotify and their “tastemakers,” and assembled, presumably
algorithmically to at least some extent, on the basis of current popularity.
Now it is of course possible to view this in terms of media, to see the Spotify
shell as the channel and what is inside of it as the content that is transmitted
through that channel. With a traditional media framing, one might typically
consider the ways in which media are produced and consumed, and the role of
the technology itself in shaping these dynamics of production and consumption.
With the advent of new media and associated practices, other possibilities and
practices entered the picture, such as the convergence of media forms and
industries; rise of amateur and fan production and remixing; and the
collaborative creation, curation, and consumption of media (Bolter and Grusin
2000; Jenkins 2006; Löwgren and Reimer 2013). We can certainly see these
dynamics in Spotify in, for instance, the ability to create and share playlists,
including collaborative playlists. It is also relatively easy for artists to get their
own music on Spotify; although it is necessary to use a label or distributor,
Spotify has deals with a number of companies that can serve these functions for
artists who do not already have a label (https://www.spotifyartists.com/guides/).
This also points to the network of actors that is brought together in order to
create and maintain the ecosystem that keeps Spotify running.
But what about Spotify itself as a thing? What is it that is actually present, that
exists in the world in some form, and what does it do? This is where things
become more complicated, because even though we can see an instance of
Spotify running on a particular device, that is clearly not either Spotify “in itself”
or all the components that are necessary for it to work. And just as it is not
obvious or easy to see what is actually going on in order to make Spotify appear
as a thing available for use, it is also not straightforward to see or understand
how the things we do are tracked, made visible, and have consequences in
various ways—although if we note things such as currently trending music or
personalized recommendations, we can definitely understand that this is going
on somehow.
To begin somewhere, we may try to understand just the Spotify application
itself as it runs on a computer. To try to get down to basics, we can see what
happens if the application is started without a network connection, to get a sense
for what happens locally and what relies on networked resources and content.
What we get is a sidebar “container” with main categories for browsing Spotify
content and one’s own music, but the main window contains a message stating:
“This view is not available offline! Please go online to load.” Now although this
type of message is now quite common and unremarkable, it is actually very
interesting when it comes to considering Spotify as a thing. Because the main
section is really more than a view: it contains the primary functional components
that enable playing music. Even the play, backward and forward buttons that
control music playback are dimmed and inactive. A banner message at the top of
the window reinforces the necessity of a network connection: “No internet
connection detected. Spotify will automatically try to reconnect when it detects
an internet connection (error code: 4).” In the bottom right corner of the window
is another indicator that it is in offline mode.
Offline mode is clearly an abnormal, inoperative state for the Spotify
application. Without a network connection, it just does not do much. Local
application and networked resources typically work together to produce Spotify
as the thing we use, but this hybridity becomes noticeable only when there is a
breakdown. It is one crucial connection in the Spotify assemblage.
Of course there is also quite a lot going on even if we consider only the
Spotify app, which also relies on various resources and processes provided by
the underlying operating system. This is perhaps even more difficult to see and
conceptualize, for non-programmers in particular. But it can become visible
when, for instance, the application crashes and one has the opportunity to send a
crash report. Viewing the details of one such report, it is possible to see hundreds
of lines of diagnostic information that includes local system information and the
processes that were running at the time of the crash.
There is a lot of information contained in even this first bit of the crash report,
ranging from straightforward (such as path to the local Spotify application) to
downright bizarre (what is a “EXC_CORPSE_NOTIFY” exception?). We might
also note the very long version code that seems to indicate that there have been
many versions and frequent updates. So, this one particular crash report one of
us generated is specific not just to the Spotify application, but to version
1.0.28.87.g8f9312a4 running on Mac OS X 10.11.6 on a computer that has been
awake for 1,200 seconds since boot and 33 seconds since wake.
Returning to normal operations and establishing a network connection, the
content that is the main focus and ostensible purpose of Spotify comes flooding
in and is foregrounded. If we were to follow the content, we would reach
networks of music distribution, licensing, curating, promoting, and so on, and of
course also the musicians who make the music. We could look at the role of
Spotify in reshaping business models in the music industry, and the implications
of this shifting landscape for the ways in which music is produced and
consumed. But let us now use the tools presented in the previous chapter to take
a tour through Spotify as a fluid assemblage.

Equipment for listening

When using Spotify as a “tool” for listening to music, it tends to withdraw into
background awareness when it is playing music as intended. When something
goes wrong—perhaps a loss of network connection, or an application crash—it
comes to presence as a malfunctioning thing, and the sources of breakdown must
be remedied in order to return to listening.
However, as seen in the crash report earlier, figuring out exactly what went
wrong is often difficult or impossible. A typical strategy for dealing with both
malfunctioning computational devices and applications is to restart them and
hope that takes care of the problem somehow. It is both practical tactic and
quasi-ritual, a way of entreating the processors and bits and bytes to sort
themselves out and begin working again.

Engagement?
Devices that provide music as a commodity without the focused effort and
communal aspects of playing musical instruments receive condemnation within
philosopher Albert Borgmann’s device paradigm (Borgmann 1984). But what
about all the “social” aspects, such as the fact that you can see live updates about
what other people are listening to after connecting to one’s Facebook account, or
“scrobble” one’s own activity to a Last.fm account? Indeed, the settings options
within the Spotify desktop application seem to be mostly about making
connections—to other devices, to one’s Spotify and other social profiles, and to
various sources of local music. More technical settings, such as cache location
and proxy server settings, are initially hidden within “Advanced” settings.
A user is aggressively encouraged to “engage” socially, but not to engage with
the workings of Spotify itself. While we might maintain a critical stance toward
blanket diagnoses of disengaged music consumption, Borgmann’s description of
the tendency for things to hide their workings and provide effortless
commodities seems to hold remarkably well. One can use Spotify heavily, but
not become a “prodigious” or terribly creative user in the sense of someone
playing a musical instrument, or indeed a hip-hop artist re-appropriating the
music playback technology of a turntable as a music-making thing itself. And
one of the key attractions of Spotify, which characterized moments of wonder at
using it for the first time when it first became available, is just how easy it is to
instantly play almost any music one would like to hear.
“Engagement” is also a common term of art in marketing, as in how Spotify’s
brand-facing page boasts that “your audience” is “extremely engaged,” spending
an average of 148 minutes listening to music on Spotify across platforms. And
further, because the audience is growing and becoming increasingly mobile,
brands have “more ways to reach them” through “moments-based marketing.”1
As Spotify describes the potential for leveraging the intimacy of audio:
Streaming opens up an entirely new set of addressable moments for marketers. The music streaming ad
revenue opportunity is worth $1.5 billion today, and it’s expected to reach at least $7 billion by 2030.
Audio’s unique ability to flex to consumers’ need states makes it an especially powerful marketing
tool. The mobile moments “at work” and “working out” alone have opened up $220M in ad revenue
opportunity. Leverage audio to reach your audience when they’re most engaged, with messaging that
matches their moment.2

Many mediations
Spotify is perhaps most associated with the act of listening to music: it is the
technology that mediates access to that content. But as seen earlier, it is also a
way for advertisers to reach “consumers” (all music listeners naturally being
subsumed in this ontological category). It is also a way for artists to reach an
audience, and to get detailed data about how people listen to their music. This
includes data about what tracks were streamed when and how many times, how
many “followers” the artist has, and more. Users are automatically notified of
new content from artists they follow, and are sent concert recommendation
emails when the artist comes through their town.3 It is also possible for people to
see what their friends are listening to through the social feed.
 Following good postphenomenological practice, we might look for
multistabilities—different kinds of relations that users can achieve to Spotify.
And this is certainly possible. But what is also striking is how many more
possibilities for varying relations seem to lie with Spotify itself (both Spotify as
thing for use and Spotify the company). It actively manages different kinds of
relations to categories of listeners, artists, brands, and developers, presenting
itself quite differently to each. And in fact, much of this presentation is about
different kinds of mediation: Spotify as a way to access music, to access an
audience, to access consumers, to access other music fans, and so on. Through
its “charts” features and year-end report on global listening behavior, it also
provides a window on what others are listening to in different parts of the world.

An audio technosphere
Spotify’s description of “moments-based marketing” for brands might just as
well be a description of its role in people’s proximal technosphere:
Our lives are shaped by what we hear—and each one of us has a unique soundtrack. Ever since
streaming gave us instant, on-demand access to all the world’s music and podcasts, it’s become easier
than ever to personalize that soundtrack.4

And Spotify excels at being involved in every moment of life, especially through
its mood- and activity-themed playlists: for studying, working out, partying,
relaxing, celebrating holidays, being just so done with winter, and many more.
Of course, these also serve as mechanisms for allowing advertisers to reach users
at precisely the right moments, and in the right mental and emotional states to be
receptive to particular messages.
Of course, users also have significant agency in determining the role that
Spotify will play in their lives. And it is interesting to note the possible
(intended) role of music in this context of curated content for particular moments
and moods—that one might turn to Spotify not just to play music (which tends to
be an emotionally rich and evocative experience anyway), but to feel something,
and to access and assess one’s feelings and current situation as part of the
process of selecting what to listen to.
Addressing the possibilities of future developments, and specifically voice-
activated assistants and devices, for delivering “deeper personalization,” Spotify
tells marketers:
Streaming already gives us insight into who people are, what they’re doing and how they’re feeling in
 the moment. As audio innovation grows, that consumer understanding will enable deeper
personalization than ever before.

The experts we spoke to talked about the potential of “dynamic audio,” or the ability to offer mood-
based targeting and creative that can adapt to your real-time context. As devices become more
connected, they’ll be able to serve up increasingly relevant content. On Spotify, for example, listeners
are 100% logged in with a persistent ID across devices—and since they’re listening all day long,
streaming provides deep intelligence about real-time context and emotional state.5

The text goes on to mention the possibility to “re-engage” listeners, “to keep the
conversation going with visual reminders or sequential messages.”
On this description of life in the technosphere, one might wonder if something
has been lost in the understanding of what a conversation is and can be. And it
seems increasingly true that we indeed cannot get out or escape.

Connection mechanisms
As mentioned in the quote earlier, it is necessary to be logged in all the time in
order to use Spotify. It simply does not exist as a music-playing thing until
logging in. This being logged in is thus one of the primary “protocols,” or modes
of connection. And the precise modes of connection available vary, with
corresponding interfaces, for regular users, artists, developers, and so on.
It is indeed the information for the latter that one might expect to be most
revealing when it comes to protocols, and Spotify’s developer site6 does not
disappoint. Here one can find information about the Spotify Web API, Web API
Console, the Spotify Play Button, Spotify iOS SDK, Spotify Android SDK, and
ways to integrate Spotify into other products. The play button7 is particularly
interesting, especially in light of our investigation of “pressing play” in Chapter
3. It is a widget for playing a specific album, track, artist, or playlist that can be
embedded in a website. It both utilizes and connects back to the Spotify service,
requiring users to log in or create an account after 30 seconds if not already
logged in.

Spotify’s algorithms
One of the aspects of algorithms highlighted earlier is how they come with
embedded assumptions about the effective computability of the world—
assumptions that even if the model is not totally accurate, it at least gestures
toward something like universal truth. With that in mind, consider Spotify’s
comments on their “Year in Music” analysis, “Wrapped”:
With billions and billions of data activities taking place on Spotify every single day, we understand
people. Every shuffle, share and song tells us a story about them. Now, we can share those stories back
to every listener — and to the world.8

And further on the same page:

We’re able to build experiences like Wrapped with a little bit of magic … and a lot of data. We call it
our streaming intelligence. Since our audience of 140 million listeners streams in moments throughout
the day, across devices, we’re able to get a real understanding of who they are, how they stream, and
what context they’re in.

This streaming intelligence lets us create hyper-personalised experiences for our listeners, like Your
Time Capsule, which takes them back to the music they loved in their teenage years. Or Release Radar,
a customised weekly playlist featuring new music from their favourite artists.

Spotify asserts that they “understand people” on the basis of their activities
within the platform; that their activities on the platform tell “stories” about them;
that they, Spotify, “build experiences”; and that they get a “real understanding”
of who users are and what “context” they are in (context presumably defined on
the basis of their device type, music selection, and so on). These are some rather
audacious claims, all founded on a belief in the power of lots of data and
computation to reveal reality. One might question the ways in which “people,”
“stories,” “experiences,” “understanding,” “context,” “engagement,” and so on
are operationally defined in texts such as this, and the algorithms they describe;
but then again, how can one challenge magic and data?
One way to begin to make sense of a fluid assemblage like this is to ask the
questions outlined in Chapter 7 on the basis of Wiener’s description of
cybernetic organisms: (1) what information is it equipped to collect; (2) in what
ways does it transform the information it collects; (3) how and where is the
information fed back into ongoing operations; and (4) what effect does this have
in other parts of the assemblage and its larger environment? We will not here go
through each of these in detail, although even the short quotes here provide a
good start in finding answers that help illuminate the character and functioning
of the Spotify assemblage.

Data and information (and understanding?)

As is made even more obvious in earlier quotes, Spotify both produces and
utilizes lots and lots of data about activities on its platform. It is on the basis of
this data that it can make claims to advertisers about how well it knows and
understands its users. But we can also interrogate just what these claims are
founded on, working in the way of Latour (1999) following the data presented in
a table in a journal article back to the soil samples in a lab back to the box in
which they were sorted back to the location from which they were extracted in
the Amazon. Take the metric of “engagement,” for example. This is defined as
the number of minutes per day that people spend playing content on the Spotify
platform across devices. All good categories cull away detail in order to produce
uniform data that can be compared, and this is no exception. After all, how could
each of these minutes of played content reflect equal levels of engagement?
Surely some moments reflect more attentiveness than others, and some probably
reflect a device playing to an empty room or to headphones that have been
removed from ears (although perhaps headphones in the future will be able to
detect proximity to ears, and thus more reliably deliver advertising content?).
The messy details of everyday music listening practices are smoothed over in a
system in which every minute that content is played is considered equal. This is
not simply good or bad—again, data that has been generated and rendered
comparable by categories in a standard system can be extremely useful.
However, when consuming and utilizing data, it is important to keep in mind that
it is never raw, and to be curious and critical about the methods that have been
used to prepare it.
We can also notice as users how action possibilities look like data waiting to
be created. For example, when seeing a playlist with a follower count
prominently displayed, one can assume that following this playlist will cause
that number to increment by one. Playing a certain kind of playlist will also
affect the kind of advertising someone on a free plan receives. Playing tracks,
following artists, creating playlists, and everything else one can do within the
Spotify application similarly generates data.
The quite precise data-fueled view that Spotify has on user activities is
brought into sharp relief in their comments on their “Year in Music” review for
2017, which they phrase as a list of 2018 goals:
Like as many things as the person who streamed “That’s What I Like” over 4,653 times this year.
Sweat more than the 27,914 people who put “Despacito” on their running playlists.
Have fewer issues than the person who streamed “Issues” over 3,152 times this year.9

To this list, perhaps one might add: understand myself better than Spotify’s
dataveillance systems do.
 The Spotify ecology
As already mentioned, Spotify works across a number of devices, including
desktop computers, smartphones, and other devices such as home speakers. It
also connects to other services, such as Facebook and Last.fm. Options to
purchase merchandise within Spotify link to third-party vendors, artist
information is provided by Rovi, and of course ads come from marketers for
brands. And surely more components and actors in the Spotify ecology could be
added to this initial list.
What is particularly interesting, though, is Spotify’s cookie policy.10 Now,
while this type of document might not usually make a short list for exciting
reading, it is quite revealing in terms of identifying different components of the
Spotify ecology. There are several different cookie types listed: those for
essential operation, performance/analytics, functionality, targeting/advertising,
third party, and Spotify ads. The third-party cookies even collect information
about activity across different websites. And the description of Spotify ads
identifies several other components and actors:
We work with web publishers, advertising networks, and service providers to deliver Spotify ads on
other web sites and services. Cookies may be used to serve you with advertisements that may be
relevant to you and your interests on other web sites and services and to regulate the advertisements
you receive and measure their effectiveness.

In the following section on managing cookies and other preferences, the same
page also states:
 There is no accepted standard on how to respond to Do Not Track signals, and we do not respond to
such signals. We use the AdChoices Icon on our website, and it may also appear on certain targeted
advertisements that we (or a service provider acting on our behalf) send to you based on information
about your online activities. You may click on the AdChoices icon or visit aboutads.info to receive
more information about the collection and use of information about your online activities for online
behavioral advertising or to learn how to opt out of having your data used for online behavioral
advertising by Digital Advertising Alliance (DAA) participating companies. Canadian users can also
visit youradchoices.ca. European users can also visit youronlinechoices.com to learn how to opt out of
having their data used for online behavioral advertising by European Interactive Digital Advertising
Alliance (EDAA) member companies.

First, it is interesting to note the explicit position taken in relation to Do Not

Track signals, which is one mechanism for configuring relations online:
specifically, Spotify does not honor this kind of request. Second, there is a
striking array of different entities for different parts of the world that allow one
to opt out of behavioral advertising—and even these can represent an
assemblage of “member companies.”
Indeed, this kind of analysis of all the components that are somehow
connected to the Spotify ecology could continue and become much more
extensive; but hopefully this brief look is enough to indicate some potential
directions that could be used for further investigation.

Spotify as infrastructural platform

Since it is now so easy to use Spotify or other similar services to instantly stream
music, it is now almost difficult to remember the time when it was necessary to
go to the store to purchase a physical CD or tape or record. But this is what made
the capabilities that the internet brought so remarkable: that it became possible to
make copies of digital music, without loss of quality, and easily share them with
anyone via the internet. Of course, these behaviors were dubbed “piracy” by the
music industry, which launched into full-on assault against them. Yet the internet
had already shown its enormous promise for being a new kind of infrastructure
for a great many things, music distribution and consumption very much
included.
Spotify entered the scene as a way to leverage this infrastructure and the
practices that had grown up around it, but to also bring the “pirates” back within
a legal framework—paying for the service either with their attention on the ad-
supported free version or with a monthly fee for the premium version. Artists
were encouraged to not care so much about exactly how much they made
through making their music available on the service, since it was for the greater
good of establishing this model as a viable framework for music distribution in
the internet era. In other words, the Spotify platform has been set up from the
beginning to be the infrastructure for distributing and listening to music.
Spotify and similar services continue to be the focus of strong criticism on the
basis of how little artists actually make through using them. In Spotify’s artist
guide,11 it is truly remarkable that virtually no information is given about how
much artists are paid—this seems to now be left entirely to music distributors.
Rather, the emphasis is on the ability to find out about listeners, and how well an
artist’s music is performing. The number of monthly listeners is prominently
displayed for each artist in Spotify, follower counts are shown for playlists,
number of listens for tracks—reflecting the ways in which Spotify platform
participation is encoded, data about user activities aggregated, and that data
further computationally processed to lead to higher-level descriptions of what
goes on in the platform (the year in review feature being a good example).

The politics of Spotify

How is Spotify involved in distributions of power and authority? As mentioned
in the previous section, it emerged in response to the music industry losing
control in the face of peer-to-peer online music sharing. Their goal was to move
users back within a legal framework—in other words, it was explicitly political,
setting up a framework for how things should be done in contradistinction to
another model.
We might also look for the various actors that are gathered together around the
both material and political thing that is Spotify. As mentioned throughout the
previous sections, there are many. While we are not able here to go through a full
analysis, it is perhaps at least worth noting that what might on the surface seem
to be two of the most fundamental kinds of human actors—artists and listeners—
often seem to be themselves used, rather than, or in addition to, being served by
it.

Spot the techniques

Spotify quite clearly utilizes the data-driven techniques that increasingly
characterize administrative, economic, and sociocultural processes. It produces
vast quantities of data about activities on its platform that it feeds into its own
operations and also provides as part of the service. This includes precisely
tailoring how Spotify shows up for specific users, ideally enabling a better
experience for users who see things they are interested in while also enabling
precise targeting of advertisements. We might also consider what have become
techniques around listening to music, such as creating playlists that include
music from different sources, and listening to music as a service rather than
owning a physical copy. Within fluid assemblages such as Spotify, these
techniques also become forms of digital labor that provide value for others in the
assemblage.

Spotify’s political economy

We have already seen how one of the main functions of Spotify is to produce an
audience that can be sold to advertisers in quite sophisticated ways. Other
questions we can ask in relation to Spotify’s political economy concern who
controls the means of production and consumption. And here we can remember
that Spotify has positioned itself as a means for controlling the distribution and
consumption of music, also providing much more detailed data about
consumption than was possible in previous models that relied on physical media
and radio distribution. It also opened up a way for artists not signed to a major
label to distribute their music—although it is noteworthy that an artist needs to
use a label, distributor, or aggregator to deliver their music to Spotify, in order to
“make sure everything on Spotify is properly licensed.”12 This is different from
a service such as SoundCloud, which allows artists to upload tracks directly,
without these intervening layers of actors and mechanisms of control.13

Algorithmic culture and Spotify

Spotify is driven by algorithms that produce and process enormous quantities of
data. It does this as a matter of course during normal use, but also as part of A/B
testing of different versions at massive scale that is used to optimize
development of the platform for certain metrics. Algorithmic processes are used
as co-creator and co-producer, leading to new types of content (such as branded
playlists and new kinds of advertising formats14) and new kinds of possibilities
to discover and play music and see what others around the world are listening to.
 Spotify is not just another media format and associated device that makes it
possible to press play—it is a fluid assemblage that incorporates many different
actors and entities in its data-fueled functioning and continuous development.

An open playlist
The notion of partial perspectives has probably never been more appropriate
than in the case of fluid assemblages. For instance, when reading our story
earlier it is crucial to remember that we here do not speak from the privileged
position of actually knowing how Spotify works. We can only articulate what we
can access, and here we only see what we are meant to see. Thus, this is just one
piece of the overall story: the one from the perspective of those who are using it.
Another, and equally important, part of the story is completely missing: the story
told from the perspective of those who are making it. But then again, this is
unfortunately quite often the case when it comes to these things.
And so we here present our story not as a final complete album, but as an
open playlist—with the invitation for others to add to it.

Notes
1 https://spotifyforbrands.com/uk/, accessed December 19, 2017.
2 https://spotifyforbrands.com/uk/feature/power-of-audio/, accessed December 19, 2017.
3 https://artists.spotify.com/blog/nerd-out-on-your-data, accessed December 19, 2017.
4 https://spotifyforbrands.com/uk/feature/power-of-audio/chapter-1, accessed December 19, 2017.
5 https://spotifyforbrands.com/uk/feature/power-of-audio/chapter-3, accessed December 19, 2017.
6 https://developer.spotify.com/, accessed December 19, 2017.
7 https://developer.spotify.com/technologies/widgets/spotify-play-button/, accessed December 19, 2017.
8 https://spotifyforbrands.com/uk/insight/2017-wrapped-data/, accessed December 19, 2017.
9 https://spotifyforbrands.com/uk/insight/2017-wrapped-data/, accessed December 19, 2017.
10 https://www.spotify.com/uk/legal/privacy-policy/#s13, accessed December 19, 2017.
11 https://artists.spotify.com/guide, accessed December 20, 2017
12 https://artists.spotify.com/faq/music#how-do-i-get-my-music-on-spotify, accessed December 20, 2017.
13 https://soundcloud.com/upload, accessed December 20, 2017.
14 https://spotifyforbrands.com/uk/gallery/, accessed December 20, 2017.
 9

Making Concepts

When we address these new technologies as “things,” we do not mean that fluid
assemblages after all are still just things. They are not. Whereas the traditional
thing has a surface, these assemblages have nested structures of interfaces.
Whereas the traditional thing is stable, these are fluid and their properties
immanent. And whereas typical things sport a rather sharp distinction between
making and using them, between producing and consuming, to use a fluid
assemblage is to become part of something created in runtime. Indeed, the
differences are so many and so substantial that it is at first difficult to see why
one would think of these assemblages as things in the first place.
As we turn to how they present themselves to us, however, and how we come
to incorporate them into our lives, the resemblance becomes more obvious. The
original meaning of the word “thing” (“þing” in old Norse, “Ding” in German,
“ting” in Swedish) is “assembly,” a gathering of the governing people to settle
matters of shared concern. A related heritage from the Vikings is how the
Swedish word for a thing (an object) is “sak” (“Sache” in German), now present
in English as “sake” (as end, purpose), thus also having this double meaning
related to both object and issue (cf., also “res” in Latin). Things are always
complex entanglements of the social and the material, but we seem to have
entered a new phase in the development of “thinging” and “thingness” where
networked computational technologies make entirely new kinds of assemblies
and assemblages possible. And while they may not be typical things, they are
certainly part of a long history of stuff standing forth as “things” to us.
Indeed, it is precisely this “standing forth” that is the basic reason for
approaching the examples discussed here as things. To be a “thing” is not a
physical property, but a state of, or coming into, being. Reasoning around what
makes a thing a thing, Heidegger remarks:
 The jug is not a vessel because it was made; rather, the jug had to be made because it is this holding
vessel. The making, it is true, lets the jug come into its own. But that which in the jug’s nature is its
own is never brought about by its making. (Heidegger 1971, 166)1

And further:
But what the vessel of this aspect is as this jug, what and how the jug is as this jug-thing, something we
can never learn—let alone think properly—by looking at the outward appearance, the idea. … Instead
of “object”—as that which stands before, over against, opposite us—we use the more precise
expression “what stands forth.” In the full nature of what stands forth, a twofold standing prevails.
First, standing forth has the sense of stemming from somewhere, whether this be a process of self-
making or of being made by another. Secondly, standing forth has the sense of the made thing’s
standing forth into the unconcealedness of what is already present. (Heidegger 1971, 166)

Here, we have tried to understand and articulate this twofold using a range of
lenses and concepts. With respect to the first, the stemming from, we have
primarily used the notion of assemblage to explore how something is being
gathered, or made, to stand forth as thing. Whereas the jug in Heidegger’s
example has a distinct maker, and a kind of stable presence as jug-Thing once
made, the fluid assemblages are constantly in the making. In a sense, their
presence as thing is immanent, but they are nevertheless able to stand forth as
“things” when they, literally, “do their thing.” With respect to the second, to
stand forth into unconcealedness, we have discussed how assemblages come to
present themselves through interfaces. But we have also tried to show that there
are significant limitations to what is actually made to “stand forth into
unconcealedness.”
And so, when we get to the question of whether the fluid assemblages are
really things or not we face both aesthetical and ethical design dilemmas:
aesthetical ones because of how current aesthetics attuned to seamless “wholes,”
or totalities, conceal much of what is actually going on; ethical ones because we
may doubt that these things are “genuine” in any reasonable sense of that term.
To Heidegger, “Only what conjoins itself out of world becomes a thing”
(Heidegger 1971, 180). Certainly, these things come into being by means of a
gathering, a bringing together that moves above and beyond what it came from.
At the same time, their forms of disclosure are at times deceptive. They may, or
may not, genuinely bring about nearness as they are equally about the
annihilation of distance. And with respect to gathering, to assembling, they are
definitely assemblages, not totalities. Because of the exteriority relations, all
their components retain their integrity, and the assemblage makes full use of this
fact—not the least in the way we as users do not really perceive that we are now
part of a large machine as we keep all of our individual properties.
What we see here is, therefore, a related but still different form of “thinging.”
While still a way of standing forth as a “thing” to us, the condition of being a
thing has largely been transformed by this process of constant becoming. This
new form of becoming lacks much of the continuity of earlier forms, and what
may appear as near, as genuinely close to us as beings, might as well be just the
absence of distance. We do not use these things, we become part of their coming
into being. This thinging is, therefore, mirrored by processes of objectification.
In what follows, we present a set of concepts formed to articulate such
aspects. They are by no means an exhaustive list of what new notions we might
need to account for the thingness of fluid assemblages. Rather, they are meant as
a kind of design examples of what we think could be created as part of an
emerging conversation about these things: a set of hopefully inspirational
prototypes of concepts we might add to the long list of notions discussed earlier
in this book.

Present-as-particular
In design, as well as when it comes to experience, we are not only interested in
what things in general are. Rather, what we really care about are particulars.
When designing, one does not make a chair in general—one makes a chair.
When sitting down in a chair, one cannot sit on a chair in general—one sits on a
chair. And when you use a mobile phone, you do not use a phone in general—
you use that particular phone. It is not technology in general that makes itself
present to us, it is some particular expression and instance of it. In fact, we
cannot even produce a chair in general, nor sit on one, should we want to. Plato
suffered substantially from this insight.
While it is true for all things artificial, as well as most other kinds and
categories, that we get to know them through encounters with particulars, the
fluid assemblages discussed here take aspects of particularity much further.
Whereas the mass-produced object may become unique through imperfections of
the production process, through wear and tear, and so on and so forth, fluid
assemblages may receive their particular expression and presence only at the
moment of use. Rather than being “things,” a more accurate description of what
happens is that they make themselves present as things as we use them.
To understand this constant becoming, we need to find complementary ways
of discussing more precisely how these things make active use of what becomes
part of our experience and frame of reference, and what does not. For a fluid
assemblage to work in this way, the mutual presence of both parts is critical: that
we, on one hand, honestly think we are using the mobile phone that we hold in
our hand, but on the other hand somehow also know that this phone without its
connectivity and constant use of resources elsewhere in the world would not at
all be the phone as we know it. The expression “bricked,” used to describe an
utterly non-responsive state of a computational device, was coined for a reason
—but importantly, we use it to describe the state of the thing, the device as such,
not the system or structure it is part of or the relation between these parts.
Now, of course we can withdraw from each experience and instead try to seek
that which is invariant across all these encounters and from these draw insights
into what they in general might be. But we can also do the opposite, and instead
of analyzing invariance of instances ask the question of how such technologies
can be made present to a user, how the particular instance of a technology comes
about—how they make themselves present-as-particulars.
Already in the basic question “what is this thing?,” we can see a trace of how
something becomes present to us as a “thing.” In this question, the “this” reveals
several aspects of what is going on. First, “this” positions the thing as a
particular, as something that can be distinguished from a “that.” Second, “this”
connects the expression to the act of expressing it: while we may not exactly
point our finger toward the thing, “this” assumes the presence of some
concurrent act of pointing out what particular it is we orient toward. Indeed, if
we do not clearly “see” this, our first response is likely to be “which one do you
mean?”. This act of pointing out what “this” refers to, in turn, tells us that to be a
thing, it needs to be distinct in terms of place and time. Without either place or
time, we cannot differentiate between this and that thing and we lose track of this
particular.
Consider for instance what happens when we encounter a number of mass-
produced and completely identical copies of something. Here, “this” may still
refer to a particular, to an individual thing. Although we may question the
relevance of the remark, we still understand what a person—such as a child
choosing between many seemingly identical plush toys—means when saying
“no, I don’t want that one, I want this one.” Had our reference “this” been to the
essence or general character of the thing, this would not have worked. But, as it
refers to “here and now” (or “here and then,” or “there and now,” etc.) it does not
at all depend on essence or character, only a specific place and time. In other
words, the way that the thing presents itself to us as a particular is largely based
on the possibility of differentiation: that we can separate it from the background
or from other instances. That our distinctions do not depend on essence but rely
on difference opens up enormous possibilities when it comes to bringing
something new into someone’s life: a person does not need to understand what
the thing actually is at all in order to be able to still understand that it is a “thing”
and allow it to take presence as such. Had our understanding of what makes a
thing a thing been based on grasping also what it does, this would not have been
possible.
From this we may learn that what makes something a thing to us to a
significant extent depends on how it presents itself to us: to be a thing, it must be
able to be present for us in certain ways. And so, to unpack how we come to
understand these massive technological systems composed of distributed
components communicating with each other as “things” present to us here and
now, we perhaps should not ask what they in general are or how they in general
work, but rather ask how they make themselves present-as-particulars. This is
why it is so important to turn to actual examples of how these technologies come
to appear, or stand forth, to us, and not just toward how they in principle
function, which is otherwise a typical way of approaching technology. Unlike
more traditional technologies now since long part of the philosophical canon,
like the hammer or the microscope, these new technologies lack continuity
between their presence-as-particulars and what they are capable of doing, most
notably because the here-and-now “this” that we encounter not even remotely
represents (or, “makes present”) what they actually entail. As technologies, they
are not, in any sense of word, unconcealed through the interfaces we encounter.
Indeed, they are particular about what they make present in use.

Multiinstability
One of the key features of a post phenomenological account is its attention to the
multistability of artifacts and the many variations that are possible in use; that is,
any given artifact can be perceived and appropriated in a vast number of ways.
Recognition and analysis of these multistabilities counter technologically
deterministic readings of things that might see them in the more narrow light of
intended use cases and dominant narratives around particular kinds of
(technological) progress. It thus helps ward off reductive approaches that can
risk misunderstanding or not seeing what technologies actually do in the world
in various contexts, and it does this by pointing to the rich variety of
predispositions and possibilities that creative human beings bring to their
engagements with the world.
The concept of multistability points to the instabilities surrounding artifacts
that are introduced on the side of what might be called use or appropriation—
that is to say, on the human side of human–technology relations. It signals the
varying relations that can be achieved to any given technology, where the
technology does in fact remain more or less given, while it is the human that
introduces an element of potentially significant variation. Of course we might
recognize that each unique technology will be in some ways different to others
of its kind, but we also generally assume a quite high level of stability with
respect to the character and constitution of the technology itself as an artifact.
Indeed, we can recognize this stability in our common usage of words such as
“hammer” or “telescope” or “park bench,” where just the term itself, as a sort of
category descriptor, evokes a quite clear picture of a certain kind of physical
form and associated possible uses (both typical and potentially more creative).
However, if we consider more technologically advanced examples, such as
those that might fall under the general categories of “computers” or
“smartphones,” we find that significant sources of instability in human–
technology relations in many cases actually come from the side of the
technologies themselves. Many of the technologies we now live with are capable
of not only multistability with respect to the relations into which they enter, but
also with respect to their very constitution as things available for use. We might
even say that just as variability can be introduced by humans in deciding how to
relate to things, we now also have variability that is introduced from the side of
technologies themselves—what we might call a multiinstability.
Consider, for example, a computer. One of the first things you do when
starting up a new computer for the first time is to create a user account. Initial
basic configuration choices such as language, region, and time zone will be set
for that account, and in the case of Apple products one is prompted to create
and/or log in to an iCloud account that, among many other things, will sync
preferences and data across devices. Even at this point the new computer is no
longer a generic thing, but rather one customized for a particular person in a
particular place. The customization continues as apps are installed and set up,
which now typically also includes more processes of creating and/or signing into
accounts and syncing associated preferences and data. This introduces a kind of
variability that comes from the side of the “thing” itself. As discussed
extensively in the examples, with such dynamically customized things different
users will see different things—not only because of the different experiences,
predispositions, cultural heritages, and so on that they bring as different humans,
but also because they are literally seeing and interacting with different things.
A relevant example is the “bttn,” a big red physical button that can be
configured to connect to a wide variety of web services such as IFTTT,
Facebook, Twitter, SMS, various “smart home” systems, and more
(https://bt.tn/). Pushing a bttn might thus have no effect if it has not been
configured, or it may initiate a quite complex chain of events including
everything from communication to controlling physical devices. This is an
object that is almost like a blank slate. With the exact same physical form, it can
be configured to be many different things that enable as many different relations.
While the action that a user takes with the bttn remains exactly the same (one of
pressing), the mediating relations it enables can vary dramatically. If we were to
look for multistability on the side of use, we would thus almost entirely miss the
multiinstabilities inherent in the bttn. Furthermore, such instability or variation
also means that such things have quite distinct “sides.” The bttn has a physical
form, but this outward form clearly does not fully reveal what it actually is in
any given configuration. Understanding its character requires looking also at its
“back end” in order to see the various services and other things to which it
connects. Its physical presence not only does not exhaust its capabilities, it does
not even come close to revealing them through its outward physical form.
We can see this dynamic in a somewhat different way when we consider
something like Google, which has a business model of profiling users in order to
sell their attention to advertisers. What ordinary users experience in terms of
their interactions and the use value they get from Google are only a small part of
what these interactions and the overall system functionality actually entail. And
even as a site like Google has information pages that face different types of users
(“ordinary” users, advertisers, and businesses), this is not just a case of a
company providing different types of offerings to different types of clients: in
this case, they are all inextricably linked components of the same system. What
any given user sees as search results are the product of a complex interplay of
past use of the system, contextual variables like location, dynamic ad markets,
and ongoing tuning of the Google algorithms. This is a complexity that is once
again hidden behind Google’s iconically simple search page.
Interestingly, when the functionality and scope of use change significantly in
relation to what is already familiar, we may or may not be aware of such
changes. In some cases, such changes are presented and broadcasted as
narratives of innovation and technological progression; in others, they are much
less explicit and even hidden from use. A prime example can be seen in the case
of Google, which, partly thanks to its iconically simple interface, is still
primarily perceived as a search engine; however, it equally accurately could be
described as a customized information experience used to create a marketplace
for selling user attention for advertisement.
In light of examples such as the ones discussed in this book where it is
obvious that the intentions of the user cannot be said to exhaust the actual
working of the thing, but where a range of other activities are taking place
simultaneously, we think philosophical accounts countering the kind of
technological determinism evident in some of these designs are more important
than ever. As discussed earlier, much of this could be approached using other
terms than “things,” instead working with notions such as systems, media, etc.
But since many of these designs explicitly build on the ways we relate to, and
experience, things through use, there is a need to account for these encounters
also from the perspective of “thingness.” In our view, the use of simple
interfaces mimicking aspects of their physical and mechanical predecessors is
quickly becoming not primarily a convenient design solution to a complex
interaction issue, but increasingly a pervasive strategy necessary to unpack in the
light of technological determinism as packaged in current consumption practices.
These fluid assemblages also need to be understood by their multiinstabilities
—all the ways in which these things change dynamically to configure their
relations to particular humans, even as particular humans also configure their
relations to the things. Not only that, these fluid assemblages also come into
relations with other things. This happens during the course of normal operation
and may not even be visible to users. This can lead to shock when we find out,
for example, that our mobile phones are constantly reporting our locations and
other data about us to third parties (Goodin 2015; Örstadius and Larsson 2015).
One of the main implications is that we cannot take things for granted,
assuming that they can be easily identified and will just “sit still” while we try to
figure out how they come into relations with humans. This is, after all, a basic
assumption in much analysis of what is going on at the human side of the
equation, evident in notions such as “multistability.” But these things are
increasingly having their own kinds of agencies, complexities, and variabilities.
There are thus tensions between stabilities and instabilities that play out over
time in ways that are not necessarily obvious or straightforward. Treating such
things as stable objects only ever reinforces them as “given,” as presented to us
ready to be used, whereas we in fact need to shift our attitudes toward them as
being “designed,” as being made and thus also open for intervention and change.
This, we believe, is a most pressing issue in countering technological
determinism: our understanding of design versus use, of making versus
appropriating and interpreting, needs to become less of a dichotomy and more of
a continuum. To “use” a fluid assemblage is to literally become part of its
making. As a consequence, we need theory that crosses this divide, addressing
the continuity between making and using by breaking with the still prevalent
habit and tradition of considering things as stable things ready for use and
analysis. It is as if we have to account for not only being present- or ready-at-
hand, but for becoming-in-hands.

Multiintentionality
Whereas the notion of multistability inquires into a certain kind of variability on
the side of the human in human–technology relations, it still assumes certain
invariability when it comes to perspective. Of course, different people may
engage differently with a tool, thus exploring its multistability, but we retain the
integrity of these individuals when it comes to defining what the thing is through
use. In other words, while the “user” may come up with different ideas and
interpretations of what a thing should be, we never really leave the perspective
of the user. This is a heritage from Heidegger and the idea that the most
primordial encounter with a tool is through use. Fluid assemblages, however,
introduce a different complexity into the matter of perspectives as any given
system might be quite different things to different users: what appears to
someone appears as a map and navigation service is to someone else a tool for
mapping people’s movements and how much time they spend in different places.
Also, a traditional tool could be transformed into something making such
multiple intentions possible. Taking a hammer for instance, one could equip it
with sensors and more so that while the carpenter makes use of it as a hammer, a
researcher is able to use it as an instrument to study muscular forces and strain
during work. But in comparison, the diversity in basic perspective when it comes
to these networked computational things is staggering. To exemplify, when
entering Facebook the headline reads “Connect with friends and the world
around you on Facebook”; when entering Facebook blueprint the headline reads
“The tools you need to learn how Facebook can help grow your business.” These
differences are not just about something becoming picked up through use in
varying ways, this is about an assemblage presenting very different views on
what it is, what it does, and why. This is not just a matter of multistability that
different users interpret and make use of the thing in different ways. Rather, this
is about an assemblage becoming present as fundamentally different “things”
framed by different intentions. What we’re looking at here is something we may
call “multiintentionality.”
While we may recognize this, we are clearly also reluctant to acknowledge
just how far the implications of this multiintentionality reaches. Indeed, it is
precisely because these different perspectives, or interfaces, have been made
separate that this multiintentionality can evolve. Had all these possible ways of
accessing or using the assemblage been available through the same interface, the
situation would have been very different as this would open up as a more or less
continuous field of possible interpretations. Their design, however, is based on
the opposite strategy: to offer an interface as simple as possible with respect to
the identified and intended forms of use in question: to listen to music, to find a
restaurant, to play a game, to send a message to your loved one, to buy a piece of
clothing, to track user locations, and so on and so forth. This is the primary
design program of the “app” genre. Thus, you access the assemblage not through
an open question about what it can do for you, but via carefully designed
interfaces that offer not only a specific functionality but also a careful framing of
the assemblage as a particular thing.
If we bring this observation back to the phenomenological notion of
bracketing, and the method of eidetic reduction, we get another view on the
significant difference between surface and interface. When Edmund Husserl
formed a method to seek the essence of things as they emerge through
experience, he looked for something that would transcend the particularities of
each individual experience and allow us to grasp what is invariant. At the risk of
overly reducing these extensive philosophical investigations, there are a couple
of moves that stand out as crucial in trying to understand the “essence” of
something in this way. First, he made the seemingly paradoxical claim that in
order to “turn to the things themselves” we need to bracket the “natural” world
(epoché). This means that we turn to how the things present themselves to our
consciousness, but that we at the same time suspend all judgment regarding what
may or may not actually be there independent of human experience. The second
move is to see what is invariant by actively performing a kind of systematic
variation upon these phenomena. To get an idea of how this method, called
eidetic variation, works, imagine that you are trying to form an idea of what the
essence of “tree” is: rather than just standing at some point looking at a tree, you
would walk around it, look at it from different angles and distances. You would
then turn to another tree and do the same, and so on and so forth. And
throughout this process, your attention is tuned toward what stays invariant
across all these experiences of trees; you try to transcend them and form an
understanding beyond what is present in any given experience. Eventually you
understand what “tree” is. Eidetic variation, or reduction, is somewhat similar to
this only that it operates not (only) on sensory experience, but on how things, or
phenomena, appear in consciousness.
Obviously, this does not come close to capturing the depth of
phenomenological method, but even this brief glimpse of it allows us to reflect
upon the significant differences between surface and interface. This method is
about tracing the outlines of something assuming that its metaphorical “surface”
(as that which is, again metaphorically, presented to us) is continuous—that the
different variations presented through consciousness are not only connected but
come to express specific kinds of continuity across instances, namely one that is
directly related to its essence. The notion of “shape/form” is more metaphorical
here, but that there is a basic relation between form and idea is a notion at least
as old as Plato, and here it comes to expression as a way of understanding
essence through what is invariant. Indeed, the origin of the word “eidos” holds
meanings related to both “form” and “shape” as well as being the start of our
word “idea.”
Here, however, we must also note the importance of intentionality. According
to phenomenology, the human consciousness is always directed toward
something; it has an inherent directionality. Thus, when we try to grasp the
essence of a hammer, the way we approach its variations to seek what is
invariant will not only depend on how the thing presents itself in some kind of,
on our side, completely passive act of mere registration, but importantly through
how we actively relate to it. This is where Heidegger makes his move,
emphasizing that it is through use that a thing like a hammer will present its
most significant characteristics to us. This is because, unlike the tree, the
hammer is primarily a tool, and thus “the less we just stare at the thing called
hammer, the more we take hold of it and use it, the more original our relation to
it becomes and the more undisguisedly it is encountered as what it is, as a useful
thing” (Heidegger 2010, 69).
Turning now to the notion of interfaces, precisely this matter of what it means
to turn to something transforms. First, because we do not face a surface in a
general sense, but an already from the start defined framing of how interior and
exterior are functionally related to each other. And second, because this framing
is not just a matter of our own intentions as we approach the thing as a tool, but
importantly also something based on the intentions framing the interface as such.
While a surface can be functionally specific, an interface has to be. We can talk
about general interfaces, but they are still somehow based on functionality, on a
certain connection between interior and exterior. In other words, whereas a
surface is open for continuous variation across instances, an interface is not.
Compare performing a kind of eidetic variation on a tree versus trying to do
the same with a fluid assemblage. Moving around trees, looking at them from
different perspectives, the experience of “tree” is continuous in a certain sense.
Turning to our fluid assemblage, we can of course make it present itself to us in
a number of different ways; we can compare use across different devices, we can
use it in different countries or log in as different users with different histories and
preferences and register the differences in results and perspective this brings
about, and so on. But there are distinct limitations to what variations we can
experience, and we know that there is much about this thing that falls outside
what we actually experience through this particular interface.
This means that the act of “bracketing” starts to perform in a very different
way: instead of suspending judgment about the natural world, like we did when
we did not engage in, say, the genetic biology of plants when trying to grasp the
phenomenological essence of “tree,” what we instead get here is a bracketing of
what falls outside what this particular interface gives us access to. And while this
may capture my understanding of what this thing is and does, it will not include
aspects I was not meant to see through this interface.
 To get to a more complete and continuous outline of what a fluid assemblage
actually is, we therefore cannot stay with our own intentional directionality but
must include also other, and sometimes completely different, trajectories. We
must turn to other interfaces, and even try to get to what exists between
interfaces. It might even be that no one actually has a complete grasp on the
assemblage. We may even question whether there is any essence to be found at
all, as these things might not have such. What is more clear, however, is that we
now have to relate to some kind of “multiintentionality” where the
transcendence we need to aim for more significantly than before needs to
embrace diversity, and of acknowledging that there is no one point from which
one can form a more “complete” conceptual framing of something like a fluid
assemblage.
The observation made earlier, that to “use” an assemblage is to become part of
it, therefore takes on very profound meanings. While a term such as “consumer
data harvesting” may be a description of what an analyst is up to, it also
highlights how being a consumer in a very practical way is also to perform a
function, to be a component in an assemblage, that among other things generates
such data.2 Thus, what seems to us to be a situation where it is our intention and
objective that determine what is going on and what is going to happen next, we
might at the very same time be acting as a component without really knowing
precisely what function we perform.

Tuning formations
Use of things that are fluid assemblages entails participating in these larger
structures and processes; and designing them means configuring the terms of this
participation for the entities involved—humans and things, users and usees.
Thus, there are extensive ongoing feedback loops in which things are updated
and customized, and data about use is collected in order to inform future
customization and, perhaps even more important, to turn our everyday practices
into data that feeds the processes of “cybernetic capitalism” (Raley 2013).
Dataveillance—“the disciplinary and control practice of monitoring,
aggregating, and sorting data”—now not only monitors but also predicts and
prescribes (Raley 2013, 124). Big data analysis is structured by non-transparent
and self-reinforcing mathematical models that can have very real consequences
for, among many other things, whether someone is able to get a job, get
insurance, or get parole (O’Neil 2016).
 It is possible to see these intertwined processes of ongoing design, use, and
assembling as processes of tuning.3 The forms that things take are tuned, as are
the structures of the larger formations of which they are part. The processes of
formation themselves are also subject to this tuning; it is perhaps even these
processes that are the main site for design intervention, taking precedence over
the honing of physical forms that has so far been the dominant mode of
industrial design. Fluid assemblages form, operate, and evolve through ongoing
processes of tuning within larger formations. These formations typically orient
tuning processes toward extracting the most useful data, and thus to tuning
human behaviors in order to enable that extraction.
It is perhaps worth emphasizing again at this point that fluid assemblages are
not strictly good or bad. The fact that a thing that is a fluid assemblage is a thing
for use and also a thing in a larger assemblage can be the source of fantastic
possibilities as well as the most misdirected uses of big data and systems of
dataveillance. The personalization of things can be delightful, even as it can also
border on creepy. However, it is also important to recognize the degree to which
things have well and truly changed. The contract between design and use that
was inherent in industrial systems of production has been replaced by terms of
service, licensing agreements, privacy policies, forced updates, and opaque
applications. Legal and political frameworks have not even begun to keep pace
with these new realities. On the other side, as people who use fluid assemblages,
it is not yet clear how we should make sense out of personalization that is not
personal at all, or relate to machines that optimize their relations to us on the
basis of massive amounts of collected data and machine learning.
It is in this context that it seems we need practices of onto-cartography that are
geared toward following the tuning formations of fluid assemblages. Here Levi
Bryant’s concept of gravity is particularly helpful: gravity “refers to the way in
which the structural openness, movement, and becomings of one machine are
mediated by another machine” (Bryant 2014, 193). Onto-cartography involves
mapping the gravitational relations between machines. In the context of fluid
assemblages, we can see that there are certain entities that exert tremendous
gravitational force over others—what Bryant terms “bright objects” (Bryant
2014, 202). Examples could be the giants of the web that are almost impossible
for connected systems to avoid, such as Google, Amazon, and Facebook.
Government dataveillance programs can also be a significant factor. And then of
course we also need to map the relations among the various other entities
involved.
The purpose of such mapping is in order to understand and intervene in
worlds. Bryant calls this a practice of geophilosopy, with three dimensions: onto-
cartography—producing topographical maps of assemblages; deconstruction—
severing relations between machines that sustain particular patterns; and
terraformation—adding machines to existing worlds “to create new paths of
movement and becoming” (Bryant 2014, 274). The practice of deconstruction
aligns with the possibility that Tonkinwise identifies for design to be also about
designing away, not only creating the new (Tonkinwise 2014). The practice of
terraformation resonates perhaps most strongly though with design as an
intentional practice of bringing about change in the world.
So far the biggest overall gravitational force that influences the becoming and
operation of fluid assemblages is that of cybernetic capitalism. Indeed,
capitalism has always been the driving force that fuels industrial design, so in
this sense it is nothing new. However, what does seem to be decidedly new is the
culmination of the development of market feedback loops being embodied in
things themselves, and the ongoing and dynamic relations that they entail. It is
optimization of things based on market logic, users being constantly
interpellated as consumers and disciplined subjects.
In this new landscape of fluid assemblages, design as a primarily form-giving
practice is no longer adequate. To design, and to design with, fluid assemblages
is to design forms and processes of becoming; to assemble and be assembled; to
tune forms and formations.

Aesthetics of immanence
When we introduced the notion of fluid assemblages earlier in this book, we
argued that current design aesthetics seem to be oriented toward an ambition to
craft totalities. Indeed, there was a need for such an aesthetic orientation in light
of how industrial production transformed the way things were made, and how
the sense of a whole made and cared for by the craftsman had to be replaced
with something that would bring the various parts and machine-made
components together. In many ways, it is the very same orientation that has
made our current condition possible, that of providing strong and unitary
experiences of something that seems to us to be one thing, when the underlying
technology in fact is far from anything like a coherent and consistent whole. As
we have tried to show, however, the effectiveness of this approach hides other
issues that we also need aesthetics to address as we consider the social contract
that design is meant to establish; that is, we both literally and metaphorically can
trust the things we use. Indeed, the crafting of such a contract between design
and use can be seen as one of the most central implications of design being a
service discipline (Nelson and Stolterman 2012).
It has been remarked many times in interaction design discourse that the
temporality of computational things is central. Computational things have
behavior and a complexity unfolding over time that an aesthetics primarily
oriented toward visual form does not account for, and that we therefore need
aesthetics dealing also with acts, motion, process, etc. But following the
argument presented earlier, we need to push this point even further: the
difference here is not just one of time and temporality, it is a difference between
transcendence and immanence. Or in other words, whereas design expression is
currently built around an aesthetics of transcendence because this was the most
critical thing to achieve in industrialized mass production, fluid assemblages
instead call for an aesthetics of immanence.
While incorporating temporality certainly prepares for some of the issues in
how fluid assemblages can be designed to present themselves in and through
use, we also need to take care of how we actually are able to make sense of,
interpret and understand, these things as what they actually are, as assemblages.
To achieve this, we cannot rely on an aesthetic fundamentally oriented toward
making totalities. Indeed, just as industrial design needed new aesthetics to
address the challenges posed by industrial production, we believe we are going
to need new aesthetics to deal with fluid assemblages.
When it comes to aesthetics, therefore, it is not enough to address the fact that
these networked computational things change over time in the sense of changing
appearance, added functions, and so on—what we typically call “interaction”
and deal with in for instance interaction design. If such temporal aesthetics still
operate under the assumption that the meaningful whole conveyed in experience
needs to be a kind of “totality,” then we still fail to address what truly
characterize these things, namely that they are assemblages. Here, it is worth
recalling Jane Bennett’s introduction, and how “assemblages are living,
throbbing confederations that are able to function despite the persistent presence
of energies that confound them from within. … Each member and proto-member
of the assemblage has a certain vital force, but there is also an effectivity proper
to the grouping as such: an agency of the assemblage. And precisely because
each member-actant maintains an energetic pulse slightly ‘off’ from that of the
assemblage, an assemblage is never a stolid block but an open-ended collective,
a ‘non-totalizable sum.’” (Bennett 2010, 23–24)
This inherent instability, and in particular the complex agency of assemblages
that in no sense is subject to “user control,” is so central to how these things
work that to solve the issue of how fluid assemblages come to expression
through use—that is, how our aesthetics help us deal with these matters when
designing—by instead trying our best to hide them is in so many ways just the
same as trying to make mass-produced objects look like they were handmade. At
first this may seem comforting, but as the previous way of making never had to
deal with the levels of complexity that then come about, the limits of the
approach will eventually cause it to break down. We argue our current situation
is very similar: industrial aesthetics still work, but the issues it cannot deal with
are amassing.
And so, we arrive at the question: What would an aesthetic proper to these
new things then be like, and what would an aesthetics of immanence entail? This
question we cannot answer here, the reason being that it calls for something else
than what has been our current task at hand. This question about aesthetics is not
an analytical problem. This is a question for design: a question that calls for
extensive experimentation, for new ways of working, probably even for new
practices of design.

Concluding remarks
We cannot through analysis arrive at an answer to what the design and use of
fluid assemblages should be like. Nor can we through philosophical articulations
arrive at all the necessary tools we need to live well with these technologies. But
what we can do is to start to conceptually furnish a place where design can go to
work, and where use can be explored as a curious and critical activity.
Much like designing needs it spaces such as the studio and the workshop, it
needs its conceptual spaces, its tools to think with. But while designing in the
workshop makes use of the tools available, there is still a basic difference
between using tools and doing design. The situation here is quite similar. We
have aimed to present conceptual tools, along with introductions and ideas about
how to use them, but the basic distinction remains: also here there is a difference
between using conceptual tools and doing design.
In this book, we have tried to set up a workshop for a different kind of
conceptual work, along with some ideas about what materials, issues,
stakeholders, contexts, and concerns to work with in this space. We have also
argued that this workshop is not only meant for what we used to call “design,”
but importantly also for what we used to refer to as a matter of “use” (and
understanding use and its consequences). With fluid assemblages, this is no
longer a dichotomy but a spectrum of acts, agencies, intentions, and more. Thus,
above all, it is a “workshop” equipped for rethinking the ways things currently
are in light of what we hope they will become. In short, it is about creating a
place for changing things.

Notes
1 What is here translated as “making” (and in other translations as “producing”) is in the German original
hergestellen. Thus, one important aspect that is lost in translation is the intimate connection to the many
other words related to stellen that he is using, perhaps most notably Gestell (enframing).
2 This resonates with Hamid Ekbia and Bonnie Nardi’s notion of inverse instrumentality, describing a
sociotechnical system that requires humans to perform certain functions as component parts (Ekbia and
Nardi 2012).
3 Richard Coyne’s (2010) investigation of the role of media technologies in the tuning of place has
helpfully influenced the conception we develop here.
 References

Adorno, Theodor, and Max Horkheimer. 1997. Dialectic of Enlightenment. Translated by John Cumming.
London and New York: Verso.
Akrich, Madeleine. 1991. “The De-Scription of Technical Objects.” In Shaping Technology/Building
Society, edited by W Bijker, 205–24. Cambridge, MA: MIT Press.
Alaimo, Cristina, and Jannis Kallinikos. 2017. “Computing the Everyday: Social Media as Data Platforms.”
The Information Society 33 (4): 175–91. doi:10.1080/01972243.2017.1318327.
Appadurai, Arjun. 1996. Modernity at Large: Cultural Dimensions of Globalization. Minneapolis, MN, and
London: University of Minnesota Press.
Arnall, Timo. 2014. “Exploring ‘Immaterials’.” International Journal of Design 8 (2): 101–17.
Barad, Karen. 2007. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and
Meaning. Durham: Duke University Press.
Baudrillard, Jean. 2005. The System of Objects. Translated by James Benedict. London and New York:
Verso.
Bennett, Jane. 2010. Vibrant Matter: A Political Ecology of Things. Durham and London: Duke University
Press.
Benson-Allott, Caetlin. 2015. Remote Control. New York and London: Bloomsbury.
Bill, Max. 2010a. “Beauty from Function and as Function.” In Form, Function, Beauty – Gestalt, 32–41.
London: AA Publications.
Bill, Max. 2010b. “The Basis and Aim of Aesthetics in the Machine Age.” In Form, Function, Beauty –
Gestalt, 60–70. London: AA Publications.
Binder, Thomas, Giorgio De Michelis, Pelle Ehn, Giulio Jacucci, Per Linde, and Ina Wagner. 2011. Design
Things. Cambridge, MA: MIT Press.
Binder, Thomas, Eva Brandt, Pelle Ehn, and Joachim Halse. 2015. “Democratic Design Experiments:
Between Parliament and Laboratory.” CoDesign: International Journal of CoCreation in Design and the
Arts 11 (3–4): 152–65. doi:10.1080/15710882.2015.1081248.
Bogost, Ian. 2012. Alien Phenomenology, or, What It’s Like to Be a Thing. Minneapolis, MN, and London:
University of Minnesota Press.
Bolter, Jay David, and Richard Grusin. 2000. Remediation: Understanding New Media. Cambridge, MA:
MIT Press.
Borgmann, Albert. 1984. Technology and the Character of Contemporary Life. Chicago: University of
Chicago Press.
Bowker, Geof. 1993. “How to Be Universal: Some Cybernetic Strategies, 1943–70.” Social Studies of
Science 23 (1): 107–27.
Bowker, Geoffrey C., and Susan Leigh Star. 1999. Sorting Things Out: Classification and Its
Consequences. Cambridge, MA: MIT Press.
Boyd, Danah, and Kate Crawford. 2012. “Critical Questions for Big Data.” Information, Communication &
Society 15 (5): 662–79. doi:10.1080/1369118X.2012.678878.
Braidotti, Rosi. 2003. “Becoming Woman: Or Sexual Difference Revisited.” Theory, Culture & Society 20
(3): 43–64.
Braidotti, Rosi, and Timotheus Vermeulen. 2014. “Borrowed Energy.” August 12.
https://frieze.com/article/borrowed-energy/.
Brey, Philip. 2010. “Philosophy of Technology after the Empirical Turn.” Techné 14 (1): 36–48.
Brown, Bill. 2001. “Thing Theory.” Critical Inquiry 28 (1): 1–22.
Bryant, Levi R. 2014. Onto-Cartography: An Ontology of Machines and Media. Edinburgh: Edinburgh
University Press.
Bryant, Levi R., Nick Srnicek, and Graham Harman, eds. 2011. The Speculative Turn: Continental
Materialism and Realism. Melbourne, Australia: re.press.
Bucher, Taina. 2012. “Want to Be on the Top? Algorithmic Power and the Threat of Invisibility on
Facebook.” New Media & Society 14 (7): 1164–80. doi:10.1177/1461444812440159.
Burdon, Mark, and Mark Andrejevic. 2016. “Big Data in the Sensor Society.” In Big Data Is Not a
Monolith, edited by Cassidy R. Sugimoto, Hamid R. Ekbia, and Michael Mattioli, 61–75. Cambridge,
MA and London: MIT Press.
Bush, Vannevar. 1945. “As We May Think.” The Atlantic, July.
https://www.theatlantic.com/magazine/archive/1945/07/as-we-may-think/303881/.
Carr, David. 2013. “For ‘House of Cards,’ Using Big Data to Guarantee Its Popularity,” February.
http://www.nytimes.com/2013/02/25/business/media/for-house-ofcards-using-big-data-to-guarantee-its-
popularity.html.
Castells, Manuel. 2000. The Rise of the Network Society. Oxford: Wiley-Blackwell.
doi:10.1002/9781444319514
Ceruzzi, Paul E. 2003. A History of Modern Computing. Second edition. Cambridge, MA and London: MIT
Press.
Cetina, Karin Knorr, Theodore R. Schatzki, and Eike von Savigny. 2005. The Practice Turn in
Contemporary Theory. London and New York: Routledge.
Coleman, B. 2011. Hello Avatar: Rise of the Networked Generation. Cambridge, MA: MIT Press.
Coyne, Richard. 2010. The Tuning of Place: Sociable Spaces and Pervasive Digital Media. Cambridge,
MA: MIT Press.
Crease, Robert P. 2001. American Philosophy of Technology: The Empirical Turn. Edited by Hans
Achterhuis. Bloomington, IN: Indiana University Press.
DeLanda, Manuel. 2016. Assemblage Theory. Edinburgh: Edinburgh University Press.
Deleuze, Gilles. 1988. Bergsonism. New York: Zone.
Deleuze, Gilles, and Félix Guattari. 1987. A Thousand Plateaus: Capitalism and Schizophrenia. Translated
by Brian Massumi. Minneapolis, MN, and London: University of Minnesota Press.
Deleuze, Gilles, and Félix Guattari. 1994. What Is Philosophy? Translated by Hugh Tomlinson and Graham
Burchill. London and New York: Verso.
Deuze, Mark. 2014. Media Life. Cambridge, UK, and Malden, MA: John Wiley & Sons.
Dilnot, Clive. 2014. “Companion to DRS 2014: Design’s Big Debates.” Edited by Johan Redström, Erik
Stolterman, Anna Valtonen, and Heather Wiltse. Umeå: Umeå University.
DiSalvo, Carl. 2012. Adversarial Design. Cambridge, MA: MIT Press.
Dolphijn, Rick, and Iris Van Der Tuin, eds. 2012. New Materialism : Interviews & Cartographies. Ann
Arbor, MI: Open Humanities Press.
Dourish, Paul. 2001. Where the Action Is: The Foundations of Embodied Interaction. Cambridge, MA: MIT
Press.
Dubberly, Hugh. 2017. “Connecting Things.” In Encountering Things, edited by Leslie Atzmon and Prasad
Boradkar, 153–65. Broadening Design to Include Systems, Platforms, and Product-Service Ecologies.
London: Bloomsbury.
Duportail, Judith. 2017. “I Asked Tinder for My Data. It Sent Me 800 Pages of My Deepest, Darkest
Secrets.” The Guardian, September 26. https://www.theguardian.com/technology/2017/sep/26/tinder-
personal-data-dating-app-messages-hacked-sold.
Easterling, Keller. 2012. “An Internet of Things.” E-Flux Journal. http://www.e-flux.com/journal/an-
 internet-of-things/.
Ekbia, Hamid, and Bonnie A. Nardi. 2012. “Inverse Instrumentality: How Technologies Objectify Patients
and Players.” In Materiality and Organizing: Social Interaction in a Technological World, edited by Paul
M. Leonardi, Bonnie A. Nardi, Jannis Kallinikos, 157–76. Oxford: Oxford University Press.
Ekman, Ulrik, Jay David Bolter, Lily Diaz, Morten Sondergaard, and Maria Engberg, eds. 2015. Ubiquitous
Computing, Complexity and Culture. New York and London: Routledge.
Ellul, Jacques. 1964. The Technological Society. Translated by John Wilkinson. New York: Vintage Books.
Ensmenger, Nathan L. 2012. The Computer Boys Take Over. Cambridge, MA and London: MIT Press.
Feenberg, Andrew. 1999. Questioning Technology. London and New York: Routledge.
Feenberg, Andrew. 2002. Transforming Technology. New York: Oxford University Press.
Finn, Ed. 2017. What Algorithms Want. Cambridge, MA: MIT Press.
Floridi, Luciano. 2011. The Philosophy of Information. Oxford: Oxford University Press.
Forlizzi, Jodi. 2008. “The Product Ecology: Understanding Social Product Use and Supporting Design
Culture.” International Journal of Design 2 (1): 35.
Frenkel, Sheera. 2017. “For Russian ‘Trolls,’ Instagram’s Pictures Can Spread Wider Than Words.” The
New York Times, December 17. https://nyti.ms/2k8soRk.
Fry, Tony, Clive Dilnot, and Susan Stewart. 2015. Design and the Question of History. London and New
York: Bloomsbury Academic.
Galloway, Alexander R. 2004. Protocol. Cambridge, MA and London: MIT Press.
Galloway, Alexander R. 2012. The Interface Effect. Cambridge, UK, and Malden, MA: Polity.
Giaccardi, Elisa, Nazli Cila, Chris Speed, and Melissa Caldwell. 2016. “Thing Ethnography.” In
Proceedings of the 2016 ACM Conference on Designing Interactive Systems. 377–87. New York: ACM
Press. doi:10.1145/2901790.2901905.
Gibson, James J. 1979. The Ecological Approach to Visual Perception. Boston, MA: Houghton Mifflin.
Gillespie, Tarleton. 2010. “The Politics of ‘Platforms’.” New Media & Society 12 (3): 347–64.
doi:10.1177/1461444809342738.
Gillespie, Tarleton. 2014. “The Relevance of Algorithms.” In Media Technologies: Essays on
Communication, Materiality, and Society, edited by Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten
A. Foot, 167–193. Cambridge, MA and London: MIT Press.
Gitelman, Lisa. 2006. Always Already New: Media, History, and the Data of Culture. Cambridge, MA and
London: MIT Press.
Gitelman, Lisa, ed. 2013. “Raw Data” Is an Oxymoron. Cambridge, MA and London: MIT Press.
Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Anchor Books.
Golumbia, David. 2009. The Cultural Logic of Computation. Cambridge, MA and London: Harvard
University Press.
Goodin, Dan. 2015. “User Data Plundering by Android and iOS Apps Is a Rampant as You Suspected.” Ars
Technica, November 4. http://arstechnica.com/security/2015/11/user-data-plundering-by-android-and-ios-
apps-is-as-rampant-as-you-suspected/.
Greengard, Samuel. 2015. The Internet of Things. Cambridge, MA and London: MIT Press.
Gunn, Wendy. 2013. Design Anthropology: Theory and Practice. London and New York: Bloomsbury.
Hall, Stuart. 2012. “Encoding/Decoding.” In Media and Cultural Studies, edited by Meenarkshi Gigi
Durham and Douglas M. Kellner, 137–44. West Sussex: Wiley-Blackwell.
Harman, Graham. 2002. Tool-Being. Chicago and La Salle, IL: Open Court.
Harman, Graham. 2011. The Quadruple Object. Winchester and Washington, DC: Zero Books.
Harman, Graham. 2016. Immaterialism. Cambridge, UK, and Malden, MA: Polity.
Hayles, N. Katherine. 2008. How We Became Posthuman. Chicago and London: University of Chicago
Press.
Heidegger, Martin. 1971. “The Thing.” In Poetry, Language, Thought, translated by Albert Hofstadter, 163–
80. New York: HarperCollins.
Heidegger, Martin. 1977. “The Question Concerning Technology.” In The Question Concerning Technology
and Other Essays, translated by William Lovitt, 3–35. New York, London, Toronto, and Sydney:
HarperPerennial.
Heidegger, Martin. 1993. “Building Dwelling Thinking.” In Martin Heidegger Basic Writings, edited by
David Farrell Krell, translated by Albert Hofstadter, 243–55. London and New York: Routledge.
Heidegger, Martin. [1927] 2010. Being and Time. Translated by Joan Stambaugh and Dennis J. Schmidt.
Albany, NY: SUNY Press.
Heidegger, Martin. [1994] 2012. “The Thing.” In Bremen and Freiburg Lectures: Insight into That Which Is
and Basic Principles of Thinking, translated by Andrew J. Mitchell. Bloomington and Indianapolis, IN:
Indiana University Press.
Ihde, Don. 1990. Technology and the Lifeworld: From Garden to Earth. Bloomington, IN: Indiana
University Press.
Ihde, Don. 2008. “Introduction: Postphenomenological Research.” Human Studies 31 (1): 1–9.
doi:10.1007/s10746-007-9077-2.
Irani, Lilly, and M. Six Silberman. 2013. “Turkopticon: Interrupting Worker Invisibility in Amazon
Mechanical Turk.” In Proceedings of CHI 2013, 611–20.
Janlert, Lars-Erik, and Erik Stolterman. 2014. “Faceless Interaction—A Conceptual Examination of the
Notion of Interface: Past, Present and Future.” Human-Computer Interaction 30 (6): 507–39.
doi:10.1080/07370024.2014.944313.
Janlert, Lars-Erik, and Erik Stolterman. 2017. Things That Keep Us Busy:The Elements of Interaction.
Cambridge, MA: MIT Press.
Jenkins, Henry. 2006. Convergence Culture: Where Old and New Media Collide. New York: New York
University Press.
Jones, John Chris. 1992. Design Methods. Second edition. New York: Van Nostrand Reinhold.
Kallinikos, Jannis. 2007. The Consequences of Information. Cheltenham, UK, and Northampton, MA:
Edward Elgar Publishing.
Kellner, Douglas M., and Meenarkshi Gigi Durham. 2012. “Adventures in Media and Cultural Studies:
Introducing the KeyWorks.” In Media and Cultural Studies, edited by Meenarkshi Gigi Durham and
Douglas M. Kellner, 1–23. West Sussex: Wiley-Blackwell.
Kiran, Asle H. 2012. “Technological Presence: Actuality and Potentiality in Subject Constitution.” Human
Studies 35 (1): 77–93. doi:10.1007/s10746-011-9208-7.
Kravets, David. 2017. “Uber Said to Use ‘Sophisticated’ Software to Defraud Drivers, Passengers.” April 6.
https://arstechnica.com/tech-policy/2017/04/uber-said-to-use-sophisticated-software-to-defraud-drivers-
passengers/#p3.
Lanier, Jaron. 2010. You Are Not a Gadget. New York: Vintage.
Latour, Bruno. 1999. Pandora’s Hope. Cambridge, MA and London: Harvard University Press.
Latour, Bruno. 2004. “Why Has Critique Run Out of Steam? From Matters of Fact to Matters of Concern.”
Critical Inquiry 30 (2): 225–48. doi:10.1086/421123.
Latour, Bruno. 2005. “From Realpolitik to Dingpolitik.” In Making Things Public, edited by Bruno Latour
and Peter Weibel, 14–43. Cambridge, MA and London: MIT Press.
Latour, Bruno. 2007. Reassembling the Social. New York: Oxford University Press.
Lessig, Lawrence. 2006. Code: Version 2.0. New York: Basic Books.
Löwgren, Jonas, and Bo Reimer. 2013. Collaborative Media: Production, Consumption, and Design
Interventions. Cambridge, MA and London: MIT Press.
Maeda, John. 2000. Maeda@Media. London: Thames & Hudson.
Manovich, Lev. 2001. The Language of New Media. Cambridge, MA and London: MIT Press.
Marcuse, Herbert. 1964. One-Dimensional Man. Boston, MA: Beacon Press.
Markham, Annette N. 2013. “Undermining ‘Data’: A Critical Examination of a Core Term in Scientific
Inquiry.” First Monday 18 (10): 323–50. doi:10.1177/1077800405275056.
McLuhan, Marshall. 1964. Understanding Media: The Extensions of Man. London and New York:
Routledge.
Meehan, Eileen R. 2012. “Gendering the Commodity Audience.” In Media and Cultural Studies, edited by
Meenarkshi Gigi Durham and Douglas M. Kellner, 242–53. West Sussex: Wiley-Blackwell.
Meikle, Graham, and Sherman Young. 2012. Media Convergence. Houndmills, Basingstoke, Hampshire,
UK and New York: Palgrave Macmillan.
Meillassoux, Quentin. 2009. After Finitude: An Essay on the Necessity of Contingency. London and New
York: Continuum.
Menzel, Peter, and Charles C. Mann. 1994. Material World. San Francisco, CA: Sierra Club Books.
Miller, Daniel. 2013. Stuff. Cambridge, UK and Malden, MA: Polity.
Morton, Timothy. 2013. Hyperobjects: Philosophy and Ecology after the End of the World. Minneapolis,
MN and London: University of Minnesota Press.
Mumford, Lewis. 2010. Technics and Civilization. Chicago: University of Chicago Press.
Nelson, Harold G., and Erik Stolterman. 2012. The Design Way: Intentional Change in an Unpredictable
World. Second edition. Cambridge, MA and London: MIT Press.
O’Neil, Cathy. 2016. Weapons of Math Destruction : How Big Data Increases Inequality and Threatens
Democracy. Great Britain: Penguin Books.
Örstadius, Kristoffer, and Linus Larsson. 2015. “Mobilen Spårar Dig – Överallt.” Dn. August 31.
http://www.dn.se/ekonomi/mobilen-sparar-dig-overallt.
Pasquale, Frank. 2015. The Black Box Society. Cambridge, MA and London: Harvard University Press.
Peters, John Durham. 2012. Speaking into the Air. Chicago and London: University of Chicago Press.
Plantin, Jean-Christophe, Carl Lagoze, Paul N. Edwards, and Christian Sandvig. 2017. “Infrastructure
Studies Meet Platform Studies in the Age of Google and Facebook.” New Media & Society 20 (1): 293–
310. doi:10.1177/1461444816661553.
Puech, Michel. 2016. The Ethics of Ordinary Technology. New York and London: Routledge.
Raley, Rita. 2013. “Dataveillance and Countervailance.” In “Raw Data” Is an Oxymoron, edited by Lisa
Gitelman, 121–45. Cambridge, MA and London: MIT Press.
Redström, Johan. 2005. “On Technology as Material in Design.” Design Philosophy Papers 3 (2): 39–54.
doi:10.2752/144871305X13966254124275.
Redström, Johan. 2008. “RE:Definitions of Use.” Design Studies 29 (4): 410–23.
doi:10.1016/j.destud.2008.05.001.
Ries, Eric. 2011. The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create
Radically Successful Businesses. New York: Crown Business.
Shannon, Claude E., and Warren Weaver. 1959. The Mathematical Theory of Communication. Urbana, IL:
University of Illinois Press.
Shove, Elizabeth. 2007. The Design of Everyday Life. New York: Berg.
Simon, Phil. 2014. “Big Data Lessons from Netflix.” Wired. https://www.wired.com/insights/2014/03/big-
data-lessons-netflix/.
Simondon, Gilbert. [1958] 2017. On the Mode of Existence of Technical Objects. Translated by Cecile
Malaspina and John Rogove. Minneapolis, MN: Univocal.
Smythe, Dallas W. 2012. “On the Audience Commodity and Its Work.” In Media and Cultural Studies,
edited by Meenarkshi Gigi Durham and Douglas M. Kellner, 185–204. West Sussex: Wiley-Blackwell.
Spuybroek, Lars. 2016. The Sympathy of Things. London and New York: Bloomsbury.
Star, Susan Leigh. 1999. “The Ethnography of Infrastructure.” American Behavioral Scientist 43 (3): 377–
91. doi:10.1177/00027649921955326.
Star, Susan Leigh, and Karen Ruhleder. 1996. “Steps toward an Ecology of Infrastructure: Design and
Access for Large Information Spaces.” Information Systems Research 7 (1): 111–34.
Stark, Luke, and Kate Crawford. 2015. “The Conservatism of Emoji: Work, Affect, and Communication.”
Social Media+ Society 1 (2): 1–11. doi:10.1177/2056305115604853.
Stolterman, Erik, Heekyoung Jung, Will Ryan, and Martin A. Siegel. 2013. “Device Landscapes—A New
Challenge to Interaction Design and HCI Research.” Archives of Design Research 26 (2): 7–33.
Strong, David, and Eric Higgs. 2000. “Borgmann’s Philosophy of Technology.” In Technology and the
Good Life? edited by Eric Higgs, Andrew Light, David Strong, 19–37. Chicago and London: University
of Chicago Press.
Suchman, Lucy. 1994. “Do Categories Have Politics?” Computer Supported Cooperative Work CSCW 2:
177–90.
Suchman, Lucy. 2007. Human-Machine Reconfigurations. Cambridge, New York, Melbourne, Madrid,
Cape Town, Singapore, São Paulo: Cambridge University Press.
The Economist. 2017a. “Data Is Giving Rise to a New Economy.” The Economist, May 6.
https://www.economist.com/news/briefing/21721634-how-it-shaping-updata-giving-rise-new-economy.
The Economist. 2017b. “The World’s Most Valuable Resource Is No Longer Oil, but Data.” The Economist,
May 6. https://www.economist.com/news/leaders/21721656-data-economy-demands-new-approach-
antitrust-rules-worldsmost-valuable-resource.
Tonkinwise, Cameron. 2014. “Design Away.” In Design as Future-Making, edited by Susan Yelavich and
Barbara Adams, 198–213. London and New York: Bloomsbury.
Turkle, Sherry. 1997. Life on the Screen: Identity in the Age of the Internet. New York: Simon and Schuster.
Turner, Fred. 2010. From Counterculture to Cyberculture. Chicago and London: University of Chicago
Press.
Usborne, Simon. 2013. “Can Cycling App Strava Change the Way We Ride?” Independent, July 3.
http://www.independent.co.uk/life-style/gadgets-and-tech/features/can-cycling-app-strava-change-the-
way-we-ride-8685996.html.
Vaidhyanathan, Siva. 2012. The Googlization of Everything. Berkeley and Los Angeles, CA: University of
California Press. doi:10.1525/j.ctt1pn8d3.
van Dijck, Jose. 2013. The Culture of Connectivity. New York: Oxford University Press.
doi:10.1093/acprof:oso/9780199970773.001.0001.
Verbeek, Peter-Paul. 2005. What Things Do: Philosophical Reflections on Technology, Agency, and Design.
University Park, PA: Pennsylvania State University Press.
Verbeek, Peter-Paul. 2008. “Cyborg Intentionality: Rethinking the Phenomenology of Humantechnology
Relations.” Phenomenology and the Cognitive Sciences 7 (3): 387–95. doi:10.1007/s11097-008-9099-x.
Viney, William. 2014. Waste. London and New York: Bloomsbury.
Wakkary, Ron, Doenja Oogjes, Sabrina Hauser, Henry Lin, Cheng Cao, Leo Ma, and Tijs Duel. 2017.
“Morse Things.” In Proceedings of DIS 2017, 503–14. New York: ACM Press.
doi:10.1145/3064663.3064734.
White, Jeremy B. 2017. “These Are the Most Bizarre Russian Election-Meddling Posts That Have Just
Been Released.” Independent, November 2. http://www.independent.co.uk/news/world/americas/us-
politics/facebook-russia-ads-posts-hillary-clinton-most-bizarre-posts-election-2016-a8032436.html.
Wiener, Norbert. 1968. The Human Use of Human Beings. London: Sphere Books Ltd.
Wiltse, Heather. 2014. “Unpacking Digital Material Mediation.” Techné: Research in Philosophy and
Technology 18 (3). Philosophy Documentation Center: 154–82. doi:10.5840/techne201411322.
Wiltse, Heather. 2017. “Mediating (Infra)Structures: Technology, Media, Environment.” In
Postphenomenology and Media, edited by Yoni Van Den Eede, Stacy O’Neal Irwin, and Galit Wellner, 3–
25. Lanham, MD: Lexington Books.
Wiltse, Heather, Erik Stolterman, and Johan Redström. 2015. “Wicked Interactions: (On the Necessity of)
Reframing the ‘Computer’ in Philosophy and Design.” Techné: Research in Philosophy and Technology
19 (1): 26–49. doi:http://dx.doi.org/10.5840/techne201531926.
Winner, Langdon. 1986. The Whale and the Reactor: A Search for Limits in an Age of High Technology.
Chicago: University of Chicago Press.
Winograd, Terry, and Fernando Flores. 1986. Understanding Computers and Cognition. Boston, San
Francisco, New York, Toronto, Montreal, London, Munich, Paris, Madrid, Capetown, Sydney, Tokyo,
Singapore, Mexico City: Addison-Wesley.
Wu, Tim. 2016. The Attention Merchants. New York: Knopf.
Yelavich, Susan, and Barbara Adams, eds. 2014. Design as Future-Making. London: Bloomsbury.
 Index

abstraction here, here, here, here, here

active roles here–here
actor-networks here–here
actor-network-theory (ANT) here
distinctive aspects of here
methodological orientation and sensitivity here
ad hoc groupings here–here
advertisements here, here, here, here, here, here
advertisers here, here, here, here, here, here, here, here, here
aesthetics of abstraction here
aesthetics of immanence here–here
agencies here–here, here, here, here, here, here, here, here
agential realism here
“agile” development here, here
aging here
Alaimo, Cristina here
algorithmic arbitrage here
algorithmic culture here–here, here–here
aspects of here
algorithmic imagination here
algorithmic objectivity here
algorithms here
cultural consequences of here
pragmatic conception of here
alien phenomenology here
alterity relations here
ambivalence of technologies here
analog/digital converters here
analog technologies here
analytic distinctions here
analytic frames here, here, here
analytic playlist here–here
ANT. See actor-network-theory (ANT)
“app” genre here
Apple here, here, here
iAd program here
Apple Watch here
application programming interface here
Aristotle’s philosophy here
artifacts, ecological structure of here
artificial here, here, here
artificial intelligence here, here
assemblages here–here, here–here, here, here
agency of here
character here–here
characteristics of here
character of computers here
complex agency of here
design aesthetics here–here
design methodology here–here
emergent properties of here
ideas about here–here
notion of here, here
topographical maps of here
assembler here
assembly/assembling here–here, here–here, here
line here
automata, historical fascination with here
automatic action here–here

background relations here

Baudrillard, Jean here
Bennett, Jane here, here, here
Benson-Allott, Caetlin here
big data here–here
perspectives of here
black boxes here, here
Bogost, Ian here–here
Borgmann, Albert here–here, here, here
framework here
Bowker, Geoffrey here, here
bracketing here, here
branding here, here
bright objects here
Bryant, Levi here, here–here, here
business forms here

calculated publics, production of here

canonical view here
CD players here–here
"random" playback order in here–here
Ceruzzi, Paul here–here
chain of references here
change here, here, here, here, here, here, here
circulating references here, here, here
classic record here
CMC. See computer-mediated communication (CMC)
Coleman, B. here
collaborative media here
commodities
 capitalism here, here–here
characteristics of here–here
communication here–here
conditions of here
face-to-face here
mathematical model of here
media for here
practices of here
processes of here
situated practices of here
complex computational systems here
complex ecosystems here
complexity here, here–here, here–here, here, here, here, here, here, here, here, here, here, here, here, here
compositions and breach of contract here–here
computational components here–here
computational device here, here, here, here, here
computationalism here
computational technology here–here, here, here, here, here
computational things here, here, here, here, here, here, here, here, here, here, here–here, here, here, here
computer-mediated communication (CMC) here
computers here–here, here
assemblage character of here
programmability of here
conceptual tools here
connection mechanisms here
consumers here, here, here
consumption here–here
configuration of here–here
contemporary design here
contemporary media landscape here
contemporary music-playing systems here
contract, compositions and breach of here–here
correlationism here
craft methods here
cultural preparation here
cultural significance here
culture industry here
culture machine here
customers, demands of here, here
customization here, here, here, here–here, here, here, here
cyberculture here, here
cybernetic capitalism here, here
cybernetics here, here–here, here, here, here
significant legacy of here
cyberwarfare here–here
cyborg relations here
cycles of anticipation here

data
collection here–here, here
combinations of here
and information here–here, here–here
massive quantities of here
 points here
processing here–here
data-driven economy here
data-intensive processes here
dataveillance here, here
programs here
deconstruction here
DeLanda, Manuel here
Deleuze, Gilles here, here, here, here, here, here
democratic rationalization here
“democratization” of media participation here
design/designing here–here, here–here
approach, foundational for here–here
central implications of here
domains of here, here
intervention here
methodology here–here
studies, scholarship in here
theory here, here
types of here
worldview here
devices
connectivity and complexity of here
landscapes here
paradigm here–here, here
digital materials here–here, here
digital media here, here, here
digital music players here, here
digital networked things here, here, here
digital technology here, here, here. See also technology
disclosure, forms of here
discursive frameworks here
discursive practices here
distribution patterns here
Do Not Track signals here
duomining here
Duportail, Judith here
dynamic ad markets here
dynamic audio here

Easterling, Keller here

“ecological” mode of interacting here–here
ecosystems here, here, here, here, here, here, here–here, here
EDAA. see European Interactive Digital Advertising Alliance (EDAA)
effective computability here, here, here
eidetic reduction here
eidetic variation here–here
Ellul, Jacques here–here
embodiment relations here
engagement here–here, here–here, here–here
complex patterns of here
entanglement with practice here
equipment here, here, here, here–here
evaluation of relevance here
exteriority here, here, here, here
external storage device here, here
external vs. internal complexity here–here

Facebook here, here, here

advertising on here
marketing here
face-to-face communication here
feedback mechanisms here
Feenberg, Andrew here–here, here
Finn, Ed here–here, here, here–here
fixity here
flat ontology here
flexibility here, here–here
fluid assemblages here, here, here–here, here, here, here, here, here, here, here, here, here
becoming and operation of here
conception of here
context of here, here
description of here, here
designing here, here
example of here
function of here, here
outline of here
producers and consumers here
sense of here
shifting to here, here
specific instances of here
thingness of here
tuning formations of here
fluidity here, here, here
formation, processes of here
function here, here–here, here, here–here, here, here, here, here, here, here, here, here, here, here, here, here,
here–here, here, here, here, here, here, here, here, here, here–here, here
functional reality here

Galloway, Alexander here–here, here

general demographics here
geophilosopy here
Gibson, J. J. here, here
gig economy here
Gitelman, Lisa here
globalization here
God’s eye view here, here
Golumbia, David here
Google here, here
ads here
GPS here
gravity here
Guattari, Felix here, here, here

hacking here–here
handiness here
Harman, Graham here–here
HCI. See human–computer interaction (HCI)
Heidegger, Martin here, here, here, here–here, here–here, here–here, here
analysis here
heritage from here
ontology here–here
hermeneutic relations here
hippie entrepreneurs here
historical logic here
Hobbes, Thomas here
human activity here–here
domains of here
mediational structure in here
human agency here, here, here
human analysis here
human–computer interaction (HCI) here, here, here
human experience here, here, here, here, here, here
human perception here, here, here
human subjectivity here
human–technology relations here
human–world interface here
hypermediacy here
hyperobjects here

ICD. See international classification of diseases (ICD)

iCloud platform here
ICTs. see information and communication technologies (ICTs)
identity here, here, here, here, here, here, here, here
immanence here, here, here, here–here
immaterialism here–here
immediacy here
industrial capitalism here
industrial design here, here–here, here, here, here
branding here
mass consumption here–here
objective of here
product of here
streamlining here
industrial infrastructures here
industrial opportunities here
industrial production here, here–here, here–here
Industrial Revolution here
information
processes of here
types of here
information and communication technologies (ICTs) here
information-based processes here
infrastructures here–here
understanding of here
initiatives, motivation for here
intelligence here, here, here, here
intentionality, importance of here
interactions here, here
 design discourse here
interfaces here–here
characteristics of here
local here
software and here
structures of here
interiority here, here, here
international classification of diseases (ICD) here
Internet of Things (IoT) here, here, here
interpretation here–here, here, here, here, here, here, here
IoT. See Internet of Things (IoT)
iTunes here
data and experience here

Janlert, Lars-Erik here

Jones, John Chris here

Kallinikos, Jannis here

knowledge
logic here
production and certification of here
KOM here

Latour, Bruno here–here, here, here, here

“lean” development here, here
legal frameworks here–here
listening
equipment for here–here
to music here, here

machine here
animating agencies of here–here
concept of here
learning here
manually building here
Marcuse, Herbert here–here
market feedback loops, development of here
Marx, Karl here–here
mass consumption here, here
massive surveillance, programs of here
mass multiplication here
mass-produced object here, here
mass production, requirements of here
materialism here–here
ideas about here–here
materialist perspectives here
materials here, here, here, here
speculation here
McLuhan, Marshall here
means of production here, here
mechanical construction here–here
mechanisms of evolution here
media here–here
 for communication here
definition of here
dynamics of here
forms, convergence of here
landscape here
technologies here–here
mediations here–here, here
kinds of here
Meillassoux, Quentin here
meta-data here–here, here
military intelligence here
Millar, David here
Minimum Viable Product here
modern computing here–here
moments-based marketing here, here
Morton, Timothy here
Moses, Robert here
MP3 file format here
multiinstability here–here
multiintentionality here–here
multiple complex dynamics here
multistability here, here, here
concept of here
notion of here
Mumford, Lewis here, here
music here–here
distribution here
industry, business models in here
listening to here, here
playback, techniques for here
music players here
production and customization of here
short history of here–here
music-playing technologies here. See also technology
pre-digital examples of here
music-playing things here
assembling of here

“natural” physical properties here

Netflix, production of “House of Cards” here
net neutrality here
networking capability here, here–here

object-oriented perspectives here–here

object-oriented philosophy here
object-oriented programming here–here
objects here–here, here, here
concept of here
functional design of here
resembling the structure of here–here
"undermining" and "overmining" of here
ontocartography here, here
practices of here
ontology here, here, here, here, here, here, here
openness here, here, here, here
open playlist here
open-source software here

passive roles here–here

patterns of inclusion here
perceptual systems here
personalization of things here
phenomenological method here, here
foundations of here–here
phenomenology
of technics here
tradition of here
philosophical carpentry here
philosophical lab equipment here
philosophy here
“empirical turn” in here
philosophy-physics here–here
physical form here, here, here, here, here
changes in here
physical properties here, here, here
Piaget, Jean here
piracy here
Plantin, Jean-Christophe here
platforms here, here, here, here, here, here–here, here–here, here, here, here, here–here
Plato here
playlists here
pocket calculator here
political economy here–here, here
political frameworks here
politics here–here
postphenomenology here–here, here–here
development of here
praxis here
present-as-particular here–here
privacy policy here
processor here
product ecology here
production
configuration of here–here
forms of here
profit-extracting sieve here
programmed thing here–here
programming here
properties here
protocols here–here
connective possibilities of things here
definition of here
of networked computing here
of RFID here
proximal technosphere here
public relevance algorithms here
QOM here
“quantified self” movement here
quotations, extensive use of here–here

Radio Frequency Identification (RFID) here

Raley, Rita here
Rdio3 online application here–here
realization, process of here
“re-engage” listeners here
relations here–here
remediation here
remote activity here
representational practice here
RFID. See Radio Frequency Identification (RFID)
roles here–here
Ruhleder, Karen here
runtime assembly here–here
runtime customization, source of input for here
Ruskin, John here

scale here, here

scapes here
scripts here
seamless totality here, here
search engine here, here, here, here
segments here
self-expression here
“self-showing” of phenomena here
Shannon, Claude here
shape/form here
ship of Theseus here
Simondon, Gilbert here
Simon, Phil here, here
simplicity here–here, here, here
sleeping policeman here
smart home systems here
smartphones here, here, here–here, here, here, here, here
social contract here, here, here
social media here, here, here, here, here, here, here
social theory here, here, here
societal dimensions of things here
software
and interfaces here
start-ups here–here
updates here
solidity here, here, here
speculative realism here
speculative turn here
Spotify here, here
algorithmic culture and here–here
algorithms here–here
application here–here
 assemblage here
attractions of here
build experiences here
cookie policy here
data-fueled view here
description of here
dynamics in here
ecology here–here
functions of here
as infrastructural platform here
political economy here
politics of here
techniques here–here
workings of here
stability here, here, here, here–here
stages of life here
Star, Susan Leigh here, here–here, here
Stolterman, Erik here
streaming here, here, here–here, here
structure, issues of here–here
stuff here, here, here, here, here, here, here, here–here, here, here, here, here
subject here–here, here, here, here, here, here, here
Suchman, Lucy here–here
systemics here
system monitoring tools here
systems approach here

tape players here

technical apparatuses here
technical artifacts here, here
technical complexity here
technical evolution, internal structural rationality of here
technical novelty here
technical rationality here
technological advances here, here
technological determinism here, here, here
technological developments here, here, here–here, here
technological ecologies here–here
technological ecosystems, particular dynamics of here
technological progress here, here, here
technological things here–here, here
technological tools here, here, here, here, here
technology here
abstract initial version of here
ambivalence of here
complexity of here
development of here
domain of here
“empirical turn” in here
essence of here
instrumental conception of here
instrumental use of here
powerful role of here
 structural analysis of here
substantive theories of here
technospheres here–here, here
terraformation here
theoretical frameworks here
things/thingness/thinging here, here–here, here, here–here, here, here–here
actions and effects of here, here
aspects of here–here
assemblages here, here–here
characteristics of here, here, here
conventional view of here
designed here, here
development of here
digital networked here
ethnography here
experience and engagement here
fluid assemblages (see fluid assemblages)
forms and functions here
function of here
general orientations to here
“general” understanding of here
mechanical construction of here–here
objects/machines here–here
personalization of here
perspective of here–here
physical form here–here
relation to here
stuff here–here
technical artifacts here
understanding of here
value of here
thing theory here
time-sharing model here
tools
action and perception here–here, here
domain of here, here
influential philosophical analysis of here
for specific purposes here
technological here
totality here, here–here, here
seamless here
transcendence here, here
transdisciplinarity here
“trending” category here
tuning here
formations here–here
Turkopticon project here
turn to things here, here, here–here

unit operations here–here

“useful thing” here–here, here
user experiences here–here, here, here, here, here, here, here
Vaidhyanathan, Siva here
variations here
Verbeek, Peter-Paul here, here

“walled garden” approach here

Weaver, Warren here
web applications here
web-based music players here, here
wicked problems here
Wiener, Norbert here, here, here
WinAmp player here
wireless internet here
World Wide Web here, here, here

“Year in Music” analysis here–here

 BLOOMSBURY VISUAL ARTS
Bloomsbury Publishing Plc
50 Bedford Square, London, WC1B 3DP, UK
1385 Broadway, New York, NY 10018, USA
BLOOMSBURY, BLOOMSBURY VISUAL ARTS and the Diana logo are trademarks of
Bloomsbury Publishing Plc
First published in Great Britain 2019
Copyright: © Johan Redström and Heather Wiltse, 2019
Johan Redström and Heather Wiltse have asserted their right under the Copyright, Designs and
Patents Act, 1988, to be identified as Authors of this work.
Cover design: Louise Dugdale
Cover image © Mina De La O / Getty images
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying, recording, or any information
storage or retrieval system, without prior permission in writing from the publishers.
Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party
websites referred to or in this book. All internet addresses given in this book were correct at the
time of going to press. The author and publisher regret any inconvenience caused if addresses
have changed or sites have ceased to exist, but can accept no responsibility for any such
changes.
A catalogue record for this book is available from the British Library.
A catalogue record for this book is available from the Library of Congress.
ISBN: HB: 978-1-3500-0435-1
ePDF: 978-1-3500-0433-7
eBook: 978-1-3500-0434-4
Typeset by Integra Software Services Pvt. Ltd.
To find out more about our authors and books visit www.bloomsbury.com and sign up for our
newsletters.