«The Social, the material, and the ontology of non-material technological objects»

Abstract

Recent calls for more attention to the impact of the material in organizing processes tend to neglect the non-material nature of many of the technological objects that populate our world. We address this shortcoming by setting out a theory of non-material technological objects, a central feature of which is the distinction between such objects and what we call their bearers. We then use this theory to elaborate various key features of computer files, to generalise Faulkner and Runde's recent theory of the identity of technological objects, and in a critical evaluation of the emerging Sociomateriality perspective.

Full text

The social, the material, and the ontology of non-material technological objects
Philip Faulkner
Clare College
University of Cambridge
pbf1000@cam.ac.uk
Jochen Runde
Judge Business School and Girton College
University of Cambridge
j.runde@jbs.cam.ac.uk
Abstract: Recent calls for more attention to the impact of the material in organizing processes tend to
neglect the non-material nature of many of the technological objects that populate our world. We
address this shortcoming by setting out a theory of non-material technological objects, a central
feature of which is the distinction between such objects and what we call their bearers. We then use
this theory to elaborate various key features of computer files, to generalise Faulkner and Runde’s
recent theory of the identity of technological objects, and in a critical evaluation of the emerging
Sociomateriality perspective.
Acknowledgments: we are grateful to Nathan Crilly, Martin Kilduff, Wanda Orlikowski and
participants at the Judge Us seminar, Cambridge Judge Business School, June 2010, for helpful
comments on earlier drafts of this paper.

1
1. INTRODUCTION
There have been numerous calls in the recent literature for a more sophisticated treatment of
materiality in organizational life, especially as represented by the technological objects that people
interact with all the time (Kallinikos 2002, Kallinikos, Hasselbladh and Marton 2010; Latour 2005;
Leonardi 2010a; Leonardi and Barley 2008, forthcoming; Orlikowski 2007, 2010; Orlikowski and
Scott 2008; Pinch 2008, 2010; Volkoff, Strong and Elmes 2007). A recurring theme in these
interventions is the need for a conception of how the social and the material come together that can
accommodate, not only interactions between people and material things, but also the apparently often
mutually constitutive relationships between much of the social and the material world.
We are sympathetic to these ideas and the work that is emerging in response to them
(Leonardi 2009, Svahn 2009; Svahn, Henfridsson and Yoo, 2009). However if we are to pay more
attention to the technological objects implicated in organizing processes, then it is necessary to look
beyond the material, and to consider also the non-material, that is, technological objects that have no
intrinsic physical being. The things we have in mind here include ICT-related objects such as
computer programs, web pages, digitised images and so on, as well as such things as product designs,
mathematical algorithms, DNA sequences and literary and musical compositions. There is no need to
labor how thoroughly our lives are inundated with such objects, and the extent to which both what we
do and how many of the things we employ in our doings, depend on and are shaped by them. Readers
will need no reminding of the enormous organizational consequences of the proliferation of non-
material ICT-related objects in particular, which have deeply affected things ranging from the
temporal organization of work (Jones et al. 2010), to modes of innovation (von Hippel and von Krogh
2003) and the types of information produced and used in organizations (Leonardi 2007, 2010b) to
mention just three.
One of the pieces still to be put in place if organizational analysis is to give due weight to
technological objects is a systematic account of the ontology of such objects, namely a theory of their
mode(s) of being, properties, identity, and so forth. Given how much has been written on technology
in organizational theory, quite a bit of which is ontological in orientation (Mutch 2009; Orlikowski
1992, 2000; Pentland and Feldman 2007), the claim that there is still a lacuna on the nature of

2
technological objects in their own right might come as a surprise. The reason that the existing
organizational literature is generally of little help in this regard, is that it tends to focus on the
networks, routines, institutions and so on that connect things — on “the space between things”, as
Leonardi (2010a) so perspicaciously puts it — rather than on the things connected. Our aim in this
paper is precisely to shift attention to things, and to non-material technological things in particular,
and in this way contribute to enlarging the small stream of work that is beginning to emerge on this
topic in organizational theory (Kallinikos 2002; Kallinikos, Aaltonen and Marton 2010; Leonardi
2010a; Volkoff, Strong and Elmes 2007).
We begin in Section 2 by setting out the conditions for something to be a non-material
technological object and introduce the distinction between non-material technological objects and
what we will call their “bearers”. Section 3 moves on to the subject of bitstrings, an immensely
important category of non-material technological object, and the ontology of computer files in
particular. The remainder of the paper then draws out some implications of the foregoing account by
relating it to some recent contributions on the ontology of technology in organization theory. Section
4 focuses on the identity of non-material technological objects, building on the theory of the
“technical identity” of technological objects recently proposed by Faulkner and Runde (2009), and
broadens the discussion to how non-material technological objects “slot” into the social world.
Section 5 explores various implications for the emerging “Sociomateriality” perspective (Orlikowski
2007, 2010; Orlikowski and Scott 2008), perhaps the most prominent of recent approaches urging
organizational theorists to engage more closely with technology per se. Finally, Section 6 concludes.
2. MATERIAL AND NON-MATERIAL TECHNOLOGICAL OBJECTS
Any ontological exercise has to begin somewhere, at some level, or with some element, of reality.
We will start with the general category of objects, under the assumption — one that we trust will be
shared by most readers — that objects are one of the basic kinds of existent found in reality, alongside
other basic kinds of existent such as events and properties. The defining feature of objects from our
point of view is that they endure and, save for those that are so basic so as not to be composed of
constituent parts in any meaningful sense, that they are structured.i

3
In saying that objects endure, or exist through time, we mean that they are things that are fully
present at each and every point in time at which they exist. Objects can therefore be said to be
“continuants”, in contrast to events or “occurrents” that take place and whose different parts occur at
different points in time. The length of time an object typically endures, what we will call its lifespan,
depends on the nature of the object under consideration. Thus while an organism such as a housefly
might have a lifespan of no more than a few weeks, an artefact such as a hammer or skyscraper might
endure for decades or even centuries.
The second defining feature of objects is that they are “structured”, that is, composed of a
number of distinct parts that are organized or arranged in some way. Thus a motorcar, for example,
comprises an engine, steering wheel, foot pedals and numerous other components, that are arranged in
a particular way. We will have more to say about the structure of objects in due course, but for the
moment we restrict ourselves to two points. The first is that the constituent parts of any object are
themselves objects in their own right. Thus an engine, for example, is itself an enduring entity
consisting of various suitably arranged parts. The second point is that the structure of an object, its
constituent parts and their organization, need not be fixed over time. Thus while the structure of a
highly engineered artefact such as a motorcar may be relatively stable (at least as long as it receives
regular maintenance), the structure of a living organism generally changes throughout its lifetime.ii
Within this general conception it is possible to distinguish various different kinds of object,
and for our purposes there are two main distinctions that need to be drawn. The first is to distinguish
the subset of objects that are technological objects. Here we proceed on the basis that technological
objects, usually but not always artefacts, are, roughly, objects to which members of some community
of human beings have assigned one or more uses in pursuit of their practical interests.iii We will come
back to this idea in more detail when we consider Faulkner and Runde’s notion of technical identity
below. The second distinction is between material and non-material technological objects. Material
technological objects are things such as bicycles, bridges and bathtubs that have a physical mode of
being, namely that they necessarily possess spatial attributes such as location, shape, volume, mass
and so on. Non-material technological objects are ones that do not have a physical mode of being and
therefore lack attributes of this sort.iv Examples of this second kind of object, which are sometimes

4
called information goods (Shaprio and Varian 1999), include product designs, sales reports,
mathematical algorithms and, perhaps most ubiquitous of all, the computer files that are the subject of
the following section, as well as things such as literary and musical compositions.
To explore some of the basic issues surrounding the idea of non-material technological
objects we will initially focus on a relatively simple example of this kind of object, namely a cooking
recipe such as Jamie Oliver’s recipe for barbecued Thai chicken legs (Oliver 2010). A recipe, we
contend, is a non-material technological object par excellence: a “syntactic” entity, similar to
newspaper articles, instruction manuals, employment contracts, musical scores and so on, in
consisting of a set of well-formed expressions written in an appropriate language, and where “well-
formed” means that these expressions adhere to the syntactical and semantic rules of that language
(Lando et al 2008).
It is easy to see that a recipe satisfies our three criteria for non-material technological
objecthood: (1) that it is an object, (2) that it is a technological object and (3) that it has a non-material
mode of being. On the first criterion, a recipe is an object in virtue of being both a continuant and
structured. With respect to the former property, a recipe is a continuant in virtue of it enduring over
time rather than being something that takes place in time. With respect to the second property, recall
that an entity is structured if it is composed of a number of distinct parts organized in some way. In
the case of material objects, structure refers to their physical components, their spatial arrangement,
interactions and so forth. In the case of non-material objects structure again refers to their constituent
parts, their arrangement and interactions, but here these are not physical attributes of the object.
Returning to our recipe, the component parts are words and their arrangement is a logical property
insofar as the words are arranged so as to conform to the rules of the language in which it is couched
and to convey the intended meaning of the author. This point applies to syntactic entities generally:
the component parts of such objects are symbols and their organization refers to their logical
arrangement so as to conform to the rules of the language in which the object is expressed.
Given that our recipe is an object, we now need to confirm that it meets our second criterion
of being a technological object. We argue that it does so in virtue of many human communities
assigning a use to such objects in pursuit of their practical interests, in this case to impart information

5
about how to prepare certain kinds of dishes. Finally, on our third criterion, our recipe is a non-
material technological object because it has no intrinsic physical being. That is to say, as it is no more
than a set of expressions, a list of ingredients and instructions for combining ingredients in the present
case, it has none of the spatial attributes of material objects listed above.
An immediate consequence of the last point is that it is important not to conflate syntactic
objects with the material objects, the recipe book, newspaper cut-out or computer printout for
example, on which they may be inscribed. That is to say, however much syntactic objects may
depend on material objects to be stored, accessed, transferred and communicated, they are always
distinct from any and all material things in virtue of their non-material mode of being. This
distinction, between a non-material object and the objects on which it is inscribed, applies quite
generally and is a key feature of our conception of non-material technological objects, making it
possible to recognize the particular properties associated with their non-material mode of being. We
will call the objects on which a non-material object is inscribed the “bearers” of that non-material
object.v In many cases the bearers of non-material objects are material objects, such as books and
magazines, computer screens or printouts, CD-ROMs and so on. Yet there may be non-material
bearers of non-material objects too. This possibility arises once we allow for bitstrings, to which we
now turn.
3. BITSTRINGS, INFORMATION AND THE ONTOLOGY OF COMPUTER FILES
The emergence and increasingly rapid take-up of new information and communication technologies
over the last fifty years vividly illustrates the role that non-material technological objects, particularly
computer files, now play in organizational life and their capacity to transform previously existing
institutional arrangements and practices. There is considerable value therefore in thinking through
some basic issues in the ontology of computing, first to say something about the nature of computer
files from the perspective developed in the preceding section, and second to unpack some further
aspects of our general conception of non-material technological objects. We begin with an important
and very specific class of non-material object called bitstrings. By a bitstring we mean a syntactic
entity made up of bits, the 0s and 1s employed in a binary numbering system. The importance of

6
bitstrings lies in that almost all silicon-based von Neumann computers, such as traditional transistor-
based PCs, store and manipulate information encoded in binary form. Bitstrings, comprising bits
usually arranged into larger eight-bit elements known as bytes, therefore lie at the heart of the
information processing carried out by computers.
The most common form of bitstring is the computer file, a discrete body of related pieces of
information encoded so as to be accessible to the central processing unit of an appropriate computer.
By discrete we mean that the information contained in a computer file is separable or isolatable from
other information. An individual computer file is therefore a distinct entity in its own right, with its
own characteristics such as filename, size and so on. Computer files are usually stored on, and
accessed via, non-volatile forms of computer media including devices such as hard disk drives, CD-
ROMs and flash memory cards, which retain the stored information even when not powered. Using
our earlier terminology these latter objects are the material bearers of computer files, things that while
closely associated with, and essential to the use of, computer files, are distinct from those files.
Computer files are classified in a variety of ways, with perhaps the most common approach
being to distinguish them according to either the type of information they contain or the format in
which that information is encoded. As regards the former, information of any kind can be stored in a
computer file, with the most basic distinction being between data files containing information that
capture things like text, images, audio recordings and so on, and program files that access these data
files. A file “format” refers to the particular way in which the information contained in the file is
encoded in binary form, with well-known file formats including JPEG, GIF, PDF and MP3. The two
ways of classifying files are closely related, since the format used to encode a particular piece of
information depends in large part on what kind of information it is. Yet the correspondence is not
one-to-one, since information of a given kind can often be encoded in a variety of different formats
(e.g. a still image might be encoded as a JPEG, GIF or BMP file), while the same file format can be
used to encode different types of data (e.g. both still images and animation can be encoded in GIF
format).
Perhaps the defining characteristic of the modern computer is programmability, the capacity
to accept and implement different sets of instructions, and in particular the stored-program

7
architecture associated with von Neumann computers. Computer programs, files in which the
bitstring represents a set of instructions which when executed on an appropriate computer enables a
particular set of computations to be carried out on a set of data, are thus one of the key types of
computer file.
vi Programs are typically divided into two broad categories, systems programs and
application programs, where the former refer to programs such as operating systems and utilities that
enable a computer to function, while the latter refer to programs such as spreadsheet packages,
database systems and games that allow particular (sets of) tasks to be achieved using a computer.
Our brief review of bitstrings and computer files makes it possible to extend our general
account of non-material objects by highlighting the possibility that non-material objects may
themselves be bearers of other non-material objects. In the preceding section we concentrated on
material bearers of non-material objects, such as the hard disk drives and CD-ROMS on which
computer files are commonly inscribed. But consider what happens when a non-material object such
as an old family recipe is digitized, namely where a bitstring representation of that object is created in
the form of a computer file generated, for instance, when a grandson types that recipe into an MS
Word document. The resulting .docx file is itself a non-material object, but is at the same time a
bearer of the original non-material object, the recipe. The same goes for the products of digitisation
more generally. Whatever the original non-material object – an image, sound or piece of text – the
resulting bitstring is a non-material bearer of that object.
Just as a non-material object is distinct from any and all of its material bearers, the same goes
for the non-material bearers of a non-material object. Thus a non-material object like our recipe may
be borne in a variety of different bitstrings corresponding to the different file formats, such as HTML,
PDF or TXT, which may be used to encode an object comprising text. Yet the recipe is distinct from
any and all of these bearers, since while the sequence of words that comprise it will be the same in
each case, the bitstring bearers will differ, having different structures, properties etc.
Further properties of bitstrings
Three further properties of bitstrings warrant consideration before we proceed, namely non-rivalry in
use, infinite expansibility and recombinability. Our aim in discussing these properties is again

8
primarily ontological, to clarify the meaning of each in light of the account set out above and thereby
say more about the nature of bitstrings and computer files in particular. We also briefly highlight
some of the main issues and phenomena that follow from these properties.vii
An object displays non-rivalry in use when its use by one person does not affect its
simultaneous use by others.viii The classic example is a coastal lighthouse designed to aid ships
navigate a hazardous coastline, where its use by one vessel does not affect its use by other vessels.
The relevance to bitstrings is that they too are necessarily non-rival in use: one person using an
application such as Microsoft Excel, for example, in no way restricts the number of other users who
can do so at the same time. But note that this property applies only to the program itself, understood
as a bitstring, and not to the individual material objects, such as DVDs and hard disk drives, on which
that bitstring resides. The same applies to all bitstrings. While bitstrings generally are non-rival in
use, their particular material bearers are more likely to exhibit rivalry in use.
This last point becomes clearer once it is recognised that many of the things ordinarily
regarded as non-rival are in fact only so up to a certain number of users, beyond which “congestion”
occurs and they exhibit rivalry in use. Thus the lighthouse is only non-rival so long as there are not
already so many vessels in the surrounding waters that the arrival of one more would interfere with its
use by the others. This cannot occur in the case of bitstrings because congestion is an inherently
physical phenomenon arising from there being a limit to the number of users who can simultaneously
access, or locate themselves around, an object without interfering with the access of others. Due to
their non-physical mode of being, bitstrings do not become congested regardless of the number of
users. They therefore exhibit a “pure” form of non-rivalry, a property they maintain no matter how
many people concurrently use them.
In the case of material objects like the lighthouse the significance of non-rivalry revolves
around difficulties associated with the use of private markets to organize provision, problems that
arise from the fact that, once it exists, the cost of providing a non-rival material object to an additional
user is zero.ix With this in mind one further point regarding bitstrings deserves mention, that use of an
existing bitstring by an additional person also requires that person to have access to a material bearer
of that bitstring. Thus unlike non-rival material objects, which can necessarily be used by others at no

9
additional cost, it is only if additional bearers of a bitstring are easily accessible that the cost of
providing an existing bitstring to an additional user will be low. This observation leads us to our
second property, infinite expansibility.
The degree of expansibility of a non-material object refers to the ease with which additional
material bearers of that thing can be made available to potential users. Infinite expansibility denotes
the limit case and refers to the property of a non-material object whereby the number of accessible
material bearers can be made arbitrarily large arbitrarily quickly at no costx. Bitstrings exhibit a
particularly high degree of expansibility, and to see why it is useful to consider the factors that
influence the expansibility of non-material objects in general. The key point here is that the
expansibility of an object depends not on the nature of that object alone, but also on the technologies
associated with the production, distribution and use of its material bearers. Consider a syntactic entity
such as an early Renaissance text. Prior to the emergence of movable type-based printing processes in
the 15th century the expansibility of such objects was extremely limited, with each new material
bearer having to be created by hand, a labour intensive and time-consuming process with a high
likelihood of transcription errors. With the development of printing additional bearers could be
produced much more quickly and with fewer inaccuracies, dramatically increasing the expansibility of
such syntactic entities despite the objects themselves being unchanged. Even then, access to these
additional bearers was limited by the constraints associated with physical modes of distribution, a
point we return to below. In short, expansibility depends in part on the technologies associated with
the material bearers of non-material objects and the degree of expansibility of an object is therefore
not fixed.
Now consider computer files and other types of bitstring, and in particular the impact the
growth in downloading – the practice of transferring information to a local computer system from a
remote system via the internet – has had on the expansibility of these objects. Computer files
nowadays exhibit a very high degree of expansibility in virtue of the fact that once created,
downloading enables users to produce material bearers of files on their own media, almost
instantaneously, at negligible cost and with very low likelihood of error.xi Notice that an important
aspect of downloading in this regard is that, in contrast to printing which affects expansibility only

10
through the ease of production of new bearers, downloading also eliminates any additional need to
distribute these bearers since production takes place on users’ own equipment. Three developments in
computing and internet use stand out as especially significant here, with all three being particularly
prominent in the last twenty years; (1) the enormous growth in the number of people with access to
the internet; (2) the rapid increases in data transfer speeds; and (3) continued falls in the cost of data
storage which mean that most internet users already own mass storage devices with spare capacity as
part of their existing computer equipment. The combined effect of these developments has been a
massive increase in the expansibility of bitstrings over the last two decades, perhaps even to the point
that they can now usefully be considered infinitely expansible.
In the case where a computer file is itself the bearer of another non-material object, such as an
MP3 encoding of an audio recording or a PDF version of a textbook, the above noted developments
have also greatly increased the expansibility of the non-material objects being borne. And it is of
course this aspect of the high degree of expansibility of computer files that lies at the heart of debates
about the impact of the internet, particularly file-sharing technologies, on the markets for such things
as recorded music and video, computer programs and literature, raising a host of issues including the
viability of legal (e.g. intellectual property right regimes) and technological (e.g. forms of copy
protection) restrictions on expansibility, the growth of open source software initiatives and so on.
The final property we will discuss, recombinability, refers to the ease with which an object
may be combined with others in order to generate new kinds of objects. By “combined” here we
mean more than merely the grouping together of objects in order to produce a collection of otherwise
unaltered and distinct things. Rather what we have in mind is the merging or blending together of
different objects, or parts of those objects, to produce something new, something that is more than the
sum of its constituent parts. The basic idea is nicely illustrated by a material object such as Meccano,
the model construction system comprising various pieces – strips, girders, plates, wheels, axles, gears,
nuts and bolts – that allows children to build simple mechanical devices such as bridges, cars,
locomotives and the like. As objects that can be put together in numerous different ways to produce a
variety of new entities, pieces of Meccano exhibit a high degree of recombinability.

11
With the exception of things, like Meccano, designed with recombinability in mind, material
technological objects typically exhibit a much smaller degree of recombinability than non-material
technological objects. Syntactic entities, for example, are often highly recombinant and it is this
property that underpins the widespread practice of reusing existing syntactic entities, or parts of those
entities, in a variety of different settings. A mundane example of this is the reuse of pieces of text, for
instance using passages originally written for one set of lecture notes in a second, subsequent, set of
notes.xii More significantly, designs for new products invariably incorporate elements taken from
existing products (e.g. think of how the spinning disc of the phonograph turntable was reused in the
design of the compact disc player or how the QWERTY keyboard layout came to be used in modern
computer keyboards). Reuse, and therefore the property of recombinability, is thus a prerequisite for
the evolutionary nature of technological change (Arthur 2009, Basalla 1988, Ziman 2000).
Reuse is important in the context of bitstrings too. One notable example of this is “code
reuse”, whereby an existing program or part thereof is reused in another program (Haefliger, von
Krogh and Spaeth, 2008; Mohagheghi and Conradi, 2007; Sherif, Appan and Lin, 2006). Forming
one of the fundamental principles of efficient programming, code reuse ranges from practices such as
the ad hoc copying of code from an existing program to a new one, to the systematic integration of
reuse into programming methodologies, as reflected in the use of objects such as software libraries
that enable programmers to easily reuse code associated with common operations such as converting
data between different formats and incorporating standard applications such as calendars, drop-down
country lists, spell-checkers and so on, into larger applications. A topical example of code reuse is the
mashup phenomenon, namely the practice of combining data, functionality or other content, either via
application programming interfaces (APIs) or other sources, into a single, often web-based, integrated
user experience (Benatallah, Casati, Daniel and Yu, 2008; Butler 2006; Kulathuramaiyer 2007).
Examples include mashups of Flikr images over a Google map, or of freely available weather data
from geospatially-indexed temperature feeds with the Google Maps interface. While mashups are not
a new idea, the rapidly increasing number of APIs and data feeds is providing ever-greater scope for
users with the necessary programming skills to rapidly develop novel web applications.

12
4. LOCATING NON-MATERIAL TECHNOLOGICAL OBECTS IN THE SOCIAL WORLD
The conception of technological objects outlined above provides a basis for distinguishing between
material and non-material technological objects, and so provides a starting point for studying the
particular properties of the latter. To this end, we now consider some of the implications that follow
from adopting the perspective described. We will focus on two areas of organizational theory in
which non-material technological objects are currently under-theorised, and where we believe
incorporating an account such as ours might be of value. The first relates to the identity of
technological objects, which is the subject of the present section. The second relates to the emerging
Sociomateriality perspective, which we consider in the next section.
We have said very little about the identity of technological objects so far, save that they are
that subset of objects to which people have assigned one or more uses in pursuit of their practical
interests. This leaves unaddressed the important questions of what makes any object an instance of a
particular kind of technological object and how technological objects come to occupy such generally
stable and readily identifiable “slots” in the social world. One of the few places in which these
questions has been directly addressed in the organizational literature is a recent paper by Faulkner and
Runde (2009), who provide a theory of what they call the “technical identity” of technological
objects. As Faulkner and Runde focus purely on what we have called material technological objects,
our aim in this section is to generalise their theory so as to include non-material technological objects.
Faulkner and Runde’s argument proceeds in two stages. The first develops a version of what
in the philosophy of technology is sometimes called the “dual-nature” conception of technology
(Kroes, 2010; Kroes & Meijers, 2006; Meijers, 2000). Central to this view is that the identity of an
object flows from what Faulkner and Runde term its “function” and its “form”. Function refers to the
use to which members of the relevant community put an object, and here Faulkner and Runde build
on the philosopher John Searle’s (1995, 1999, 2001) ideas about “agentive” functions that humans
impose on objects and other kinds of entities in pursuit of their practical interests. The key idea is that
such functions, for example of a watch to tell the time or a camera to capture still images, are
collectively assigned to objects by members of social groups. Rather than being intrinsic to an object,
therefore, the function assigned to it (and so its technical identity) is necessarily community-relative.

13
With regard to the form of a technological object, Faulkner and Runde argue that in order for
the function assigned to an object to be sustained, that object must generally possess the physical
characteristics and capabilities required to perform that function.xii i As objects of a particular
technical identity often vary considerably in the details of their physical characteristics, Faulkner and
Runde adopt a “family resemblance” view of technological objects, recognising that many objects
have definite common physical features and capabilities on the basis of which they can be grouped as
tokens of the same type, even when there may be no single set of physical features shared by them all.
According to the dual nature conception, the function and form of an object are together
constitutive of that thing being a particular kind of technological device within some community or, in
other words, of it possessing a particular technical identity in that community. The second stage of
Faulkner and Runde’s argument locates this conception within a broad structuration theory framework
(Giddens 1984; Orlikowski 1992; Sewell 1992), represented in their case by the transformational
model of social activity (TMSA) associated with Archer (1995), Bhaskar (1979), and Lawson (1997,
2003). This is not the place to go into the details of structuration theory or the TMSA, but the main
insight on which both are based is that human activities and social structure are recursively organized.
Human activities are things that people do, whether these doings issue from conscious, perhaps
sustained, reflection or are largely automatic and routinized. Social structure includes things such as
social rules, relations, positions and the like, continuants that shape those activities in both enabling
and constraining ways. “Recursively organized” refers to social structure being at once drawn on in
human activities and at the same time reproduced (and potentially transformed) as a largely
unintended consequence of those activities.xiv
The link between the dual nature conception and the TMSA is provided by the idea that
collectively assigned functions are a type of social rule. Faulkner and Runde define social rules as
generalized procedures expressible by suitable transformations of the formula “if X in situation C, do
Y”, where these rules are sustained in virtue of being accepted by, and implicated in the activities of,
the members of a social group. The collective assignment of an agentive function to a certain type of
object is no more than one such procedure, expressible as an injunction of the form “objects with
such-and-such physical characteristics are for this purpose in such-and-such a situation”. Of course,

14
Faulkner and Runde recognise that the members of the group who assign that function do not
generally think explicitly about the various physical components and capabilities of that object or that
such-and-such a function is assigned to it. Typically people just see a watch or a camera and know
what they do or are for, i.e. they have either internalised the relevant rules as tacit knowledge or else
developed capacities or dispositions to act in accordance with those rules – what Searle (1995: 127-
147) calls “Background causation” or Bourdieu (1990) calls “habitus”.
As Faulkner and Runde portray it, so long as the people in some community use and / or
reference any technological object in the appropriate manner, that is, in accordance with the function
collectively assigned to it, they contribute to sustaining the function associated with that object, and
so also its technical identity. Collective assignments of function and the technical identities they
support are thus elements of social structure on their account — no less than social positions, social
relations and social rules more generally — the continued reproduction of which accounts for the
relative stability of the contents of our artefactual worlds. But that is not to say that things can’t ever
change. New technical identities may emerge, be this by accident or design, most commonly where
there are significant changes in the form of the objects to which a particular function has been
assigned (e.g. in the move from propeller to jet engines in the aerospace industry), but possibly also
with significant changes in the function assigned to objects of a particular form (e.g. in the
transformation of the turntable from a playback device into a musical instrument in the DJ/turntablist
community). And there is always the possibility of the emergence of technological objects that might
not have existed before, and where it may take a while before the function assigned to them become a
matter of general policy and their technical identities become established more widely.
Non-material technological objects
One of the shortcomings of Faulkner and Runde’s theory is that it is restricted to material
technological objects. This shortcoming can however be overcome in a relatively straightforward
manner, by substituting the notion of an object’s form with the more general notion of an object’s
structure. Recall that objects are entities that are structured in the sense of being composed of
constituent parts that are organized in some way. For material objects such as bridges, bolts and

15
bathtubs, structure is equivalent to what Faulkner and Runde term physical form, namely an object’s
material components, their spatial arrangement, and how these elements interact. For non-material
objects, structure again refers to the constituent parts of an object, their arrangement and interactions,
but in this case these are not physical attributes. Recall that in the case of a syntactic entity such as a
recipe, for example, the component parts are words and their arrangement is a logical one – requiring
conformity with the rules of the language in which the object is expressed – rather than a spatial one.
With this amendment the generalised theory of the technical identity of an object can be stated as
follows: an object possesses a particular technical identity within the community in which it is used
and / or appropriately referenced if:
(1) that object has assigned to it the function associated with that technical identity; and
(2) the structure of that object is such that it is generally able to perform that function.
Faulkner and Runde’s original theory thus becomes a special case of a more general theory in which,
so far as material objects are concerned, structure is equivalent to physical form. However, the more
general notion of an object’s structure makes it possible to use it to account for the technical identity
of non-material objects, material objects, and indeed hybrids of the two that have both material and
non-material components. Thus a syntactic entity is a recipe within the community in which it is used
and / or appropriately referenced if:
(1) that entity has assigned to it the function of informing people how to create a certain kind of
dish; and
(2) the syntactical and semantic structure of that entity is such that it is generally able to inform
its users how to create that dish.
Note that we do not deny that a recipe may have all kinds of functions besides informing users how to
create a dish, e.g. to entertain through the quality of the prose in which it is written, or to promote
cultural awareness or a particular brand of ingredient. But functions of this second sort are not what
distinguish cooking recipes from other syntactic objects. It is analytic to what a recipe is that it
provides information about how to produce a certain kind of dish.
Exactly the same idea applies to bitstrings and their technical identity within some community.
Consider a word-processor application. According to our generalised account set out above, a

16
bitstring is a wordprocessor within the community in which it is used and / or appropriately
referenced if:
(1) that bitstring has assigned to it the function of enabling people to create and edit documents
on a computer; and
(2) the structure of the bitstring is such that when it is executed on a computer (of the appropriate
type) it can generally be used to create and edit documents.
A word-processing programme such as MS Word thus derives its identity, not only from the structure
of the bitstring comprising it, but also from the function assigned to that bitstring: in this case, to
facilitate the production and manipulation of documents on a computer.
Just as in the case of material objects and their physical form, there is generally no single
structure that a non-material object must possess in order to perform a given function. Thus, in the
case of a recipe, the sets of instructions that part-constitute different recipes can vary significantly in
their details. One question this raises is on what basis, in the face of such diversity, people are able to
group objects as tokens of the same type. We maintain that the “family resemblance” idea mentioned
above applies here too, that such objects often share enough in the way of distinct functions, structural
features and capabilities to be grouped as tokens of the same type, even if there is no single set of
features common to them all. Thus while GarageBand, Cubase Studio 5, Logic Pro and their many
competitors all vary somewhat in their architectures, system requirements, ease of use, and so on, they
share sufficient commonalities to be grouped as digital audio workstations (and then, save for their
meeting conditions corresponding to (1) and (2) in the examples given above, even when there is no
single feature shared by them all).
Bitstrings are typically designed to achieve very specific ends. But there are also analogues of
the possibility raised by Faulkner and Runde, of technological objects of some pre-existing material
form being used differently from what their designers or originators may have intended. A good
example here is the repurposing of Web 2.0 open-source weblog software in the development of a
low-cost digital image library reported by Uzwyshyn (2008). Examples of this kind raise the further
question of whether the repurposing concerned might alter the technical identity of the item
concerned. Of course this is an empirical matter that will vary from case to case. But it seems

17
eminently possible that while repurposing will not always be sufficiently radical to affect the technical
identity of the object concerned, it might be so in some cases. For example, it is quite conceivable
that the weblog software mentioned above might develop a new technical identity as digital archiving
software in its new role, especially if it comes to be marketed as such.
While we have taken a realist view of non-material technological objects once they have
come into existence — that is, that they and their identities (once established) exist independently of
their observation by any single observer — this is of course not to deny that they are deeply social
things. There are at least three ways in which this is so. In the first place, many non-material
technological objects are trivially social in the sense of being the product of the efforts of more than
one person (and then sometimes spread over more than one generation). Second, non-material
technological objects are social in the sense of being human artefacts that presuppose pre-existing and
well-formed languages. Third and most interestingly, they are social to the extent that, as in the case
of material technological objects, their identity is sustained through human practices in the
communities in which it has currency, where every time that a member of that community uses and /
or references the object concerned in a way that is consistent with the function assigned to it,
contributes to reinforcing that identity. Note that this relation is always two-way: the human practices
that sustain the identity of a particular kind of technological object are in part sustained by objects of
that identity being implicated in those practices. Thus while the various practices associated with
wordprocessing help sustain the identity of wordprocessing programs, the existence of
wordprocessing programs at the same time helps to sustain the practices of wordprocessing. In these
cases, in other words, there are aspects of the social world that are mutually dependent.
Finally, in taking a realist view of the identity of non-material technological objects we mean
only that their structure and the function assigned to them is independent of any individual observer
(save of course for the very special case in which there is only one person doing the assigning and that
same person is also the individual observer). How that object is interpreted, that is, whether it is
considered good, efficient, indispensible, unsafe or whatever else is a different matter entirely —
which of course is true also of material technological objects, as is so well illustrated in the social
construction of technology literature (Bijker 1995; Pinch and Bijker 1987). In short, we do not

18
suggest that everyone in a community that assigns a function to some object will have identical
attitudes to that object or even always recognise it as an instance of a class of objects that have a
particular technical identity. Neither do we deny the influence of political, economic and other forms
of power in the historical processes in which the structure and identity of technological objects are
forged. But we do claim that discussions of different attitudes towards / interpretations of a particular
kind of technological object presuppose that that object already has assigned to it the function that is
constitutive of that kind of technological object. For otherwise studies such as Wiebe Bijker’s (1995)
investigation of the different early forms of the bicycle, or Pinch and Trocco’s (2002) study of the
emergence of the analogue synthesizer, would not have got off the ground. Exactly the same goes for
non-material technological objects.
5. SOCIOMATERIALITY
In the foregoing section we began to shift attention from technological objects as objects, towards
their identity and positioning in society more widely. Indeed, we see the analysis of the last section as
one way to bring non-material technological objects into a structuration theory framework, by
rendering collective assignments of function and the technical identities they underpin as part of
social structure. Our aim in this section is to explore these ideas a little further, by considering how
our account aligns with the “Sociomateriality” perspective advocated by Wanda Orlikowski (2007,
2008 in collaboration with Susan Scott, 2010), perhaps the most prominent advocate within the
organizational and IS research community of paying more attention to the stuff of technology rather
than focusing exclusively on peoples’ interactions with technology. We thus come full circle at this
point, returning to the theme we began with in this paper.
Orlikowsksi’s work on the theoretical foundations of Sociomateriality draws heavily on the
writings of Barad (2003, 2007), Latour (2005) and Suchman (2007) and is closely linked with fellow
travellers such as Bijker (1995), Pickering (1995) and Knorr Cetina (1995) (see Jones (2010) for a
review). The concept of Sociomateriality is characterised by a fundamental ontological commitment
to the “constitutive entanglement of the social and material in everyday organizational life”
(Orlikowski 2007: 1435). The image that emerges is of a highly relational social world in which

19
“entities (whether humans or technologies) have no inherent properties, but acquire form, attributes,
and capabilities through their interpenetration” (Orlikowski and Scott 2008: 455), and where humans
and technology have “so thoroughly saturated each other that previously taken-for-granted boundaries
are dissolved” (Orlikowski and Scott 2008: 455). If humans and technologies (not to mention
organizations) have no inherent properties, or indeed boundaries or meanings, this raises the all-
important question of where to draw the line between them, be this for analytical or even simple
practical purposes. Orlikowski (2010) invokes the performative metaphysics of Karen Barad (2003)
in answer to this question, according to which the lines humans draw between things are always local
resolutions, the product of “agential cuts”, performed and (temporarily) stabilised through human
practices.
We are supportive of Orlikowski’s efforts to foreground the material in organizational
research, and it is therefore pleasing to be able to point to some important commonalities between her
position and our own. On the view that we have been advocating, the identity of technological
objects, both material and non-material, is not intrinsic to those objects but something that involves an
intertwinement with the social (through the recursively reproduced functions collectively assigned to
them). We would therefore agree that there is a sense in which human cognition and intentionality
contribute to constituting the artefactual world, in and through human practices, making the objects
that populate it what they are in the communities in which their identities have currency. Indeed, as
far as material technological objects are concerned, and as our allegiance to the dual nature theory of
technological objects might already have suggested, we would be happy to describe their identities as
‘socio-material’. Further, it seems to us self evident that technology has deep organizational
consequences (witness how news provision, publishing and the delivery of music has changed in the
wake of developments in ICT) and, potentially, at least in a derived way and over long periods, even
for the physical constitution of the human body (e.g. through advances in medicine and food
technology, changes in working conditions, and so).
There are however some important differences between Orlikowski’s account and our own.
Many of these relate to the ontological claims that “people and things exist only in relation to each
other” (Orlikowski and Scott 2008: 455) and what we will call the interpenetration thesis that “entities

20
(whether humans or technologies) have no inherent properties” but become what they are “through
their interpenetration” (Orlikowski and Scott 2008: 455). In our view, both claims are problematic as
general statements about socio-technical reality. Nevertheless, we will argue that there are domains in
which they hold, and that, suitably qualified, they add nuance to the Sociomateriality perspective.
To assess the claim that people and things exist only in relation to each other, it is useful to
distinguish between internal and external relations (Lawson 2003: 227-228). Any two entities are
internally related if they are constituted in part by the relationship in which they stand to each other.
Examples include husband and wife, computer-programmer and computer, and nut and bolt (human to
human, human to non-human, and non-human to non-human internal relations respectively). Any two
entities that are related in some other way, but are not constituted in part by this relationship, are
externally related. Examples here include two students taking the same class, an iPhone and the
student who uses it, the drive shaft and the steering wheel of the car. Orlikowski does not distinguish
between internal and external relations in her writings on Sociomateriality, and, given her purposes, it
is understandable that she should focus on classes of people and classes of things that are internally
related. But not every relation is an internal relation, and it is therefore necessary to specify the
particular relata involved when making claims about internal relations. While videogame players and
videogames presuppose each other, professional footballers and videogames probably don’t, even
though many professional footballers are avid videogame players.
The interpenetration thesis is a significantly stronger claim than people and things existing
only in relation to each other, and some qualification is necessary here too. Interpenetration, as
Orlikowski describes it, implies a merging of bodies ontologically (sometimes, it seems, to the extent
of melting into or “saturating” each other) whereas two things being internally related or mutually
constitutive does not. While computer programmers and computers are internally related, this does
not mean that they morph into one thing or even merely meld in some way, except perhaps
metaphorically such as when an obsessive computer programmer is described as “disappearing into”
or “being glued to” his or her computer. Of course we accept there may be cases in which people and
technological objects have combined to some extent, and where it is perhaps legitimate to speak of
their fusing as “composite bodies” (e.g. where people have dental fillings, pacemakers or artificial hip

21
replacements). Further, there may be other ways in which interpenetration might occur, something we
will return to below. However, we would maintain that people and technological objects, while no
doubt often internally related and sometimes interpenetrating, are nevertheless generally distinct and
different things, with their own intrinsic properties (e.g. while people are conscious beings,
technological objects generally are not, while people have biological bodily functions, most
technological objects do not, and so on). And if so, this opens up the possibility that where we draw
the line between things, while surely still ‘agential cuts’ performed in human practices, may actually
reflect these differences – real differences that exist independently of any individual human observer
– rather than coming down to mere convention, local resolutions and so on.
Orlikowski’s observation that “there is no social that is not also material, and no material that
is not also social” (Orlikowski 2007: 1437) is perhaps also a reflection of the interpenetration thesis.
Again, this is a claim that needs qualification. While there do seem to be many things that are both
social and material, not least all material technological objects, it seems to us that there are many
material things whose nature and existence does not seem to presuppose human involvement (e.g.
mineral deposits that are exploited in the construction of human artefacts and which there is very good
reason to believe had already been in existence long before the emergence of homo sapiens). More to
the point in the present context, we would also claim that there exist things that are social but not
material, e.g. human ideas, social rules (including assignments of function), relations and positions,
and, importantly, the non-material technological objects that are the subject of this paper (Leonardi
2010a). While all of these things certainly presuppose the existence of a material world, that alone
does not make them material.
All this brings us to what for present purposes is the most important difference between
Orlikowski’s account and our own, namely that the theory of Sociomateriality, at least as it has been
developed so far, does not appear to have an explicit place for non-material technological objects.xv
And if organizational life depends on and is governed by such objects as much as it appears, then we
would urge that the ontology of Sociomateriality be expanded to include them as distinct entities in
their own right. Of course to the extent that non-material technological objects presuppose material
bearers and the wider social and physical infrastructures on which these bearers and their operation

22
depend, it might be possible to get quite far even while keeping intangible objects in the background
— e.g. restricting focus to material objects implicated in the use of non-material technological objects
such as computer software (Orlikowski and Iacono 2001). This may be sufficient for many purposes.
However, there are as we have seen also many situations in which the specific properties of non-
material technological objects are decisive in their impact, and in which the specific mode of being of
such objects needs to be taken into account.
An immediate and interesting positive consequence for the Sociomateriality literature of
focusing on non-material technological objects is that it might lend some degree of support to the
interpenetration thesis. Although we see little evidence of interpenetration of humans and nonhumans
that have a material form (at least as yet) as we have said, it can be argued that many material IT-
related devices such as hard-disk drives and CD-ROMs are penetrated by bitstrings from all kinds of
sources, be these systems or application software or data files of all descriptions. Of course, this is
not the same as saying that the bitstrings become one with their bearers. Rather the notion of
penetration here refers, somewhat metaphorically, to the idea that once an object such as a CD-ROM
has a bitstring inscribed onto it, there is a sense in which the resulting disc may be regarded as a
composite entity, comprising both the material CD-ROM and the bitstring that has “penetrated” it.
Similarly there is a sense in which non-material objects are themselves penetrated by other non-
material objects, such as when a bitstring representation of an existing sales invoice is created or an
audio recording is encoded as an MP3 file. In an analogous way, it may sometimes be right to say
that humans themselves are penetrated by non-material objects (items such as memorised procedures,
pin numbers, passwords and the like, which are simple enough to retain their fidelity when passing
into the human mind). Indeed, being penetrated in these ways seems to be a prerequisite for what it
means to be a bearer of a non-material object in the sense we introduced above.
In our view, then, it is possible to talk of non-material objects penetrating material and other
non-material technological objects, as well as, in some cases, humans too. Interpenetration requires
something more, however, since this implies two-way penetration between objects (i.e. that it is not
only that A penetrates B, but that A penetrates B and B penetrates A). Perhaps the most compelling
examples of this kind of reciprocal penetration occur in relation to computer networks, where each

23
individual computer connected to a network may be regarded as both penetrating, and being
penetrated by, other computers on the same network. By connecting different networks the internet
can then be viewed as promoting the interpenetration of computer systems on a global scale, the scope
of which grows ever wider as internet access becomes possible on ever more kinds of device (mobile
phones, media players, games consoles, etc.). Nor is the interpenetration that the internet facilitates
limited to material technological objects; understood as a system of inter-linked hypertext documents,
the world wide web displays the interpenetration of a vast number of syntactic entities.
6. CONCLUSION
Trevor Pinch (2010: 87) closes a recent paper with the thought that it is time for technology studies to
make technological objects − nonhumans in his terminology − reappear. Orlikowski’s writings on
Sociomateriality reflect a similar concern, as does the insistence in actor network theory that
nonhuman “actants” be seen as actors symmetric with their human counterparts (Latour 2005). But
the technological objects that come up in this work tend to be of the material variety that have a
tangible physical form, such as the cobbled shoulders, fences and roads mentioned by Pinch, and with
little attention to what distinguishes non-material technological objects from their material
counterparts. While we agree on the need to bring the material to the fore in the study of organizing
processes, we have argued that many of the technological objects that inundate our lives are in fact
non-material entities, and that, in view of their many special qualities, they deserve close theoretical
attention in their own right.
Our aim in writing this paper was accordingly to outline a theory of non-material
technological objects, to provide an account of their mode of being, their properties and how these
differ from those of material technological objects. One of the things that became clear early on in
our investigation — and this is something reflected in many discussions of computer files — is how
easy it is to conflate non-material objects with the media on which they are stored and through which
they are accessed and communicated. Since many of the properties of non-material technological
objects that have led to their having such an enormous impact in human organization are intimately
connected with their non-material mode of being and changes in the way that they are borne, it is

24
important to avoid this conflation. It is to this end that we introduced the distinction between non-
material objects and their bearers.
While non-material technological objects of one sort or another have no doubt been in
existence since the emergence of mankind, the advent of computers and the internet has led to a step
change in the levels of production, accessibility and general traffic in such objects. None of this
would have been possible without the emergence of the bitstring, now perhaps the most ubiquitous
non-material technological object of all, which displays the properties of non-rivalry in consumption
and, given the availability of suitable bearers and low transmission costs, expansibility, to almost
perfect degrees. A key aspect of this development is that it became possible to create bitstring bearers
of other non-material technological objects, which greatly contributed to the expansibility of the latter
by facilitating the production and transfer of their bearers on ever more efficient systems.
The second half of the paper considered some implications of our theory and involved a shift
in focus from non-material technological objects per se to wider questions of their identity and their
positioning in the social world. We first proposed a generalisation of Faulkner and Runde (2009),
extending their notion of the technical identity of an object to encompass both material and non-
material technological objects alike. We then highlighted some of the ways in which our framework
informs ongoing debates within organizational theory, by comparing our theory to the
Sociomateriality perspective, perhaps the leading approach in contemporary organizational theory
working on moving technological objects higher up the theoretical and empirical agenda. We
concluded that, while there are differences between the Sociomateriality perspective and our own,
non-material technological objects are probably one respect in which our account provides support for
the interpenetration thesis associated with the Sociomateriality perspective.
As is typical of ontological investigation this has been primarily a ground clearing exercise,
an attempt to articulate a coherent and comprehensive framework for thinking about technological
objects with the objective of promoting conceptual clarity in both theoretical and empirical work that
touches on technology in organizing processes. But we believe that our account also raises various
more specific ideas for future refinement, and we close by mentioning three of these. The first has to
do with the distinction between material and non-material technological objects, and the particular

25
properties exhibited by non-material technological objects. We have highlighted three of these
properties in relation to bitstrings – non-rivalry in use, infinite expansibility and recombinability – but
their full implications, as well as those of others such as non-degradability, have yet to be fully traced
out. Amongst the questions that arise here are the significance of the fact that computer users
typically employ an assemblage of technological objects comprising both rival (e.g. CPUs and hard-
disk drives) and non-rival (e.g. computer programs) elements, and the extent to which the non-rivalry
of non-material objects is rendered irrelevant in the absence of a high degree of expansibility.
The second idea relates to the notion of hybrid technological objects, devices that are
complexes of both material and non-material components. We have had little to say directly about the
ontology of such objects, their particular characteristics, different varieties and so on, yet such is their
pervasiveness, most notably in relation to ICT, that a more detailed account of the nature of these
objects would be of considerable value. Finally, there is work to be done looking at the relationship
between the various material and non-material bearers, larger system infrastructures and
complementary goods that the use of non-material objects presupposes. Amongst the questions that
arise here are the likely effects of the emergence of new kinds of material or non-material bearers and
the conditions under which such developments imply radical changes to existing practices and
organizational arrangements. In short there remains much to be done, both at the level of the broader
social ontology of non-material technological objects that we have been pursuing ourselves, as well
on these (and other) more substantive questions.

26
ENDNOTES
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
i Since technological objects are usually highly structured we will ignore these very basic,
unstructured, objects in what follows.
ii This last point raises well-known philosophical questions concerned with how the identity of a thing
that undergoes change is nevertheless maintained over time (e.g. Theseus’ paradox). For a detailed
account of these issues see Lowe (2002), chapters 2-4.
iii There are also examples of other species imposing functions on objects, but which we leave to one
side here (Aunger 2010a, 2010b).
iv Leonardi (2010) has something similar in mind where he proposes that something can be touched
(tactile experience) as a criterion of the tangibility or material nature of what we call material objects.
Intangible or non-material objects are then ones that cannot be touched. While we agree with the
spirit of this idea, it does however suffer from the problem that there are material technological
objects that cannot be the objects of human tactile experience, e.g. the recently developed
nanodragster, the size of which is 1/50,000 of the width of a human hair (Vives, Kang, Kelly and Tour
2009). !
v The term “copies” is often used in this context, for example when a DVD on which a Hollywood
film is inscribed is described as being a copy of that film. The same applies to paperback and
hardback copies of a novel, to copies of a live audio recording on CD, and so on. We avoid using the
term copies, however, because it is potentially confusing insofar as it risks collapsing, or at least
obscuring, the distinction between a non-material object and the bearer of that object. The DVD on
which a film is inscribed is not literally a copy of that film, just as the flash drive on which an MP3
version of a song resides is not literally a copy of that song.
vi The term computer program is also used to refer to a particular process of computation resulting
from the execution of a set of instructions on a computer at a particular point in time. Since we are
concerned with objects rather than processes, in what follows we will use the term program to refer
exclusively to sets of instructions (for a more detailed discussion of the ontology of computer
programs see Eden 2007, Eden and Turner 2007, Lando et al 2008 and Suber 1988).

27
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
vii See Quah (2003) and Rayna (2008) for further discussion of these and other properties of bitstrings.
viii The notion is closely linked with the study of public goods in economics (Samuelson 1954). Public
goods are goods that are both non-rival in use and “non-excludable”, where the latter refers to the
property of a good in virtue of which it is not possible to exclude anyone from consuming it once it
has been provided.
ix See chapter 14 of Gravelle and Rees (2004) for a useful exposition of the standard “market failure”
analysis.
x Unlike non-rivalry, use of the term infinite expansibility is far from settled and it is worth
highlighting two points about our own usage of the term. First, unlike authors such as David (1993,
2004) who use the terms non-rival and infinitely expansible interchangeably, we follow Quah (2003)
in regarding the two ideas as closely related but distinct. Second, Quah’s definition of infinite
expansibility as the property of an object whereby “its quantity can be made arbitrarily large
arbitrarily quickly at no cost” is problematic from the point of view of our own account. The
difficulty lies in the notion of the “quantity” of a thing when used in the context of non-material
objects. According to our account, a non-material object either exists or it doesn’t and in this sense its
quantity is trivially either exactly one or zero. Beyond this, however, the non-physical mode of being
of these objects means that it makes little sense to talk literally of the number of distinct instances of a
non-material object. What Quah has in mind when he refers to the quantity of a non-material object,
in our view, is the number of material bearers of that thing.
xi In this context it is helpful to distinguish downloading from the closely related practice of media
streaming. While both involve the transfer of computer files over the internet from a remote system
to a local one, streaming does not usually result in those files being stored on the local system and so
does not involve the creation of additional material bearers. Nevertheless, the growth in streaming
has contributed to the increasing ease with which non-material objects can be accessed.
xii Although we will not develop the point further here, it is also worth noting how the recombinability
of syntactic entities such as text is increased when that entity is borne by a bitstring and so able to be

28
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
manipulated digitally. To take a simple example, consider how easy it is to recombine existing pieces
of text using the cut and paste function in modern word processing software.
xiii Faulkner and Runde adopt the qualifier “generally” to allow for the cases in which objects with a
particular identity are not able to perform the function for some reason, perhaps because they are
malfunctioning or because the circumstances are not appropriate to their use. We typically do not
stop regarding what is normally a kettle as a kettle when it has lost its capacity to boil water because
of a defective heating element or a power outage.
xiv While we regard the TMSA as an instance of structuration theory, proponents of the TMSA tend to
differ somewhat from Giddens on the ontological status afforded to social structure. Whereas the
former group tends to view social structure as distinct from human activity, with structure being seen
as something that generally pre-exists human activity and which is then sustained and perhaps
transformed by it, Giddens seems to regard social structure as a generalised feature of activity,
existing only in its momentary instantiations in human practices and as memory traces (Giddens 1984,
17, 23-28).
xv It is true that Orlikowski invokes the example of the Google algorithm, which would be a paradigm
non-material technological object on our account, and that her use of the word material has been
interpreted more liberally than we do here simply to mean “stuff” that could include objects of a non-
tangible nature (Leonardi 2010). However, and while we accept that the term “material” has different
connotations, it seems that this is stretching the meaning of the term too far in a context in which the
non-material mode of being of certain technological objects has so many distinct and highly important
organizational and economic consequences. Further, it seems to us that Orlikowski’s commitment to
what we have called the interpenetration thesis is of a piece with what we see as her reluctance to
make sharp ontological distinctions between material and non-material technological objects.

29
REFERENCES
Archer, M. 1995. Realist Social Theory: The Morphogenetic Approach. Cambridge University Press,
Cambridge, UK.
Arthur, W. B. 2009. The Nature of Technology: What It Is and How It Evolves. Allen Lane,
Middlesex, UK.
Aunger, R. 2010a. What’s special about human technology? Cambridge J. Econom. 34(1) 115-123.
Aunger, R. 2010b. Types of Technology. Tech. Forecasting Soc. Change. 77(5) 762-782.
Barad, K. 2003. Posthumanist Performativity: Toward an Understanding of How Matter Comes to
Matter. Signs: J. Women Culture Soc. 28(3) 801-831.
Barad, K. 2007. Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter
and Meaning. Duke University Press, Durham, NC.
Basalla, G. 1988. The Evolution of Technology. Cambridge University Press, Cambridge, UK.
Benatallah, B., F. Casati, F. Daniel, J. Yu. 2008. Understanding Mashup Development. Internet
Comput. 12(5) 44-52.
Bijker, W. 1995. Of Bicycles, Bakelites, and Bulbs: Towards a Theory of Sociotechnical Change. MIT
Press, Cambridge, MA.
Bhaskar, R. 1979. The Possibility of Naturalism. Harvester Press, Hemel Hempstead, UK.
Bhaskar, R. 1989. Reclaiming Reality: Critical Introduction to Contemporary Philosophy. Verso,
London, UK.
Bourdieu, P. 1990. The Logic of Practice. Polity Press, Cambridge, UK.
Butler, D. 2006. Mashups mix data into global service. Nature. 7072 (5 January 2006) 6-7.
Collier, A. 1994. Critical Realism: An Introduction to Roy Bhaskar’s Philosophy. Verso, London,
UK.
Eden, A. 2007. Three Paradigms of Computer Science. Minds and Machines. 17(2) 135-167.
Eden, A., R. Turner. 2007. Problems In The Ontology Of Computer Programs. Appl. Ontology. 2(1)
13-36.
Faulkner, P., J. Runde. 2009. On the identity of technological objects and user innovations in
function. Acad. Management Rev. 34(3) 442-462.

30
Giddens, A. 1984. The Constitution of Society. Polity Press, Cambridge, UK.
Gravelle, H., R. Rees. 2004. Microeconomics, 3rd ed. Prentice Hall, Harlow, UK.
Haefliger, S., G. von Krogh, S. Spaeth. 2008. Code Reuse in Open Source Software. Management Sci.
54(1) 180-193.
Jones, M. 2010. Sense and sociomateriality: advancing the debate on the relationship between
technology and organizations. Working paper, University of Cambridge, UK.
Jones, M., K. Munir, W. Orlikowski, J. Runde, L. Nikolychuk. 2010. In the shadow of dead trees:
temporal changes in the news industry. Working Paper, University of Cambridge, UK.
Kallinikos, J. 2002. Reopening the black box of technology artifacts and human agency. Paper
presented at the twenty-third International Conference on Information Systems.
Kallinikos, J., A. Aaltonen, A. Marton. 2010. A theory of digital objects. First Monday. 15(6) 1-17.
Kallinikos, J., H. Hasselbladh, A. Marton. 2010. Governing social practice: technology versus
institutions. Working Paper, London School of Economics, UK.
Knorr Cetina, K. 1997. Sociality with objects: Social relations in postsocial knowledge societies.
Theory, Culture Soc. 14(4) 1-30.
Kroes, P. 2010. Engineering and the dual nature of technical artefacts. Cambridge J. Econom. 34(1)
51–62.
Kroes, P., A. Meijers. 2006. The dual nature of technical artefacts. Stud. Hist. Philos. Sci. 37(1) 1-4.
Kulathuramaiyer, N. 2007. Mashups: Emerging Application Development Paradigm for a Digital
Journal. J. Universal Comput. Sci. 13(4) 531-542.
Lando, P., A. Lapujade, G. Kassel, F. Fürst. 2008. An Ontological Investigation in the Field of
Computer Programs. J. Filipe, B. Shishkov, M. Helfert, L. Maciaszek, eds. Software and Data
Technologies. Springer, Heidelberg, Germany, 371-386.
Latour, B. 2005. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford
University Press, Oxford, UK.
Lawson, T. 1997. Economics and Reality. Routledge, London, UK.
Lawson, T. 2003. Reorienting Economics. Routledge, London, UK.

31
Leonardi, P. 2007. Activating the Informational Capabilities on Information Technology for
Organizational Change. Organ. Sci. 18(5) 813-831.
Leonardi, P. M. 2009. Why Do People Reject New Technologies and Stymie Organizational Changes
of Which They Are in Favor? Exploring Misalignments Between Social Interactions and
Materiality. Human Communication Research. 35(3) 407–441.
Leonardi, P.M. 2010a. Digital materiality? How artefacts without matter, matter. First Monday. 15(6).
Leonardi, P.M. 2010b. From road to lab to math: the co-evolution of technological, regulatory, and
organizational innovations for automotive crash testing. Soc. Stud. Sci. 40(2) 243-274.
Leonardi, P.M., S.R. Barley. 2008. Materiality and change: Challenges to building better theory about
technology and organizing. Inform. Organ. 18(3) 159-176.
Leonardi, P.M., S.R. Barley. 2010. What’s Under Construction Here? Social Action, Materiality, and
Power in Constructivist Studies of Technology and Organizing. Acad. Management. Ann. 4 1-
51.
Lowe, E. J. 2002. A Survey of Metaphysics. Oxford University Press, Oxford, UK.
Meijers, A. 2000. The relational ontology of technical artifacts. P. Kroes, A. Meijers, eds. The
empirical turn in the philosophy of technology. Elsevier Science, Oxford, UK, 81-96.
Mohagheghi, P., R. Conradi. 2007. Quality, productivity and economic benefits of software reuse: a
review of industrial studies. Empirical Software Engrg. 12(5) 471-516.
Mutch, A. 2010. Technology, Organization, and Structure – A Morphogenetic Approach. Organ. Sci.
21(2) 507-520.
Oliver, J. 2010. Barbecued Thai chicken legs. http://www.jamieoliver.com/recipes/chicken-
recipes/barbecued-thai-chicken-legs
Orlikowski, W. 1992. The Duality of Technology: Rethinking the Concept of Technology in
Organizations. Organ. Sci. 3(3) 398-427.
Orlikowski, W. 2000. Using technology and constituting structures: A practice lens for studying
technology in organizations. Organ. Sci. 11(4) 404-428.
Orlikowski, W. 2007. Sociomaterial Practices: Exploring Technology at Work, Organ. Stud. 28(9)
1435-1448.

32
Orlikowski, W. 2010. The sociomateriality of organizational life: considering technology in
management research. Cambridge J. Econom. 34(1) 125-141.
Orlikowski, W., C.S. Iacono. 2001. Desperately seeking the “IT” in IT research – a call to theorizing
the IT artefact. Inform. Systems Res. 12(2) 121-134.
Orlikowski, W., S. Scott. 2008. Sociomateriality: challenging the separation of technology, work and
organization. Acad. Management. Ann. 2 433-474.
Pentland, B.T., M.S. Feldman. 2007. Narrative networks: Patterns of technology and organization.
Organ. Sci. 18(5) 781-795.
Pickering, A. 1993. The Mangle of Practice: Agency and Emergence in the Sociology of Science.
Amer. J. Sociology. 99(3) 559-589.
Pinch, T. 2008. Technology and institutions: living in a material world. Theory and Soc. 37(5) 461-
483.
Pinch, T. 2010. On making infrastructure visible: putting the non-humans to rights. Cambridge J.
Econom. 34(1) 77-89.
Pinch, T., W. E. Bijker. 1987. The social construction of facts and artifacts: Or how the sociology of
science and the sociology of technology might benefit each other. W.E. Bijker, T.P. Hughes, T.
Pinch, eds. The Social Construction of Technological Systems. MIT Press, Cambridge, MA, 17-
50.
Pinch, T., F. Trocco. 2002. Analog Days: The Invention and Impact of the Moog Synthesizer. Harvard
University Press, Cambridge, MA.
Quah, D. 2003. Digital goods and the new economy. Jones, D., ed. New Economy Handbook.
Academic Press, San Diego, CA, 289-321.
Rayna, T. 2008. Understanding the Challenges of the Digital Economy: The Nature of Digital Goods.
Communications and Strategies. 71(3) 13-36.
Samuelson, P. 1954. The Pure Theory of Public Expenditure. Rev. Econom. Statist. 36(4) 387-389.
Searle, J.R. 1995. The Construction of Social Reality. Allen Lane The Penguin Press, Middlesex, UK.
Searle, J.R. 1999. Mind, Language and Society. Weidenfeld & Nicolson, London, UK.
Searle, J.R. 2001. Rationality in Action. MIT Press, Cambridge, MA.

33
Sewell, W.H. 1992. A theory of structure: duality, agency and transformation. Amer. J. Sociology.
98(1) 1-29.
Shapiro, C., H. Varian. 1999. Information Rules. Harvard Business School Press, Boston, MA.
Sherif, K., R. Appan, Z. Lin. 2007. Resources and incentives for the adoption of systematic software
reuse. Internat. J. Inform. Management. 26(1): 70-80.
Suber, P. 1988. What is Software? J. Speculative Philos. 2 89-119.
Suchman, L.A. 2007. Human-machine reconfigurations: Plans and situated actions. Cambridge
University Press, Cambridge, UK.
Svahn, F., O. Henfridsson, Y. Yoo. 2009. A Threesome Dance of Agency: Mangling the
Sociomateriality of Technological Regimes in Digital Innovation. International Conference on
Information Systems, Phoenix, USA.
Svahn, F. 2009. The Sociomateriality of Competing Technological Regimes in Digital Innovation. J.
Molka-Danielsen, ed. Selected Papers of the 32nd Iris Seminar. Tapir Academic Press,
Trondheim, Norway.
Uzvyshyn, R. 2008. Repurposing Open Source Software for Agile Digital Image Library
Development: The University of West Florida Libraries Model. D-Lib Magazine. 14(9/10).
Vives, G., J.-H. Kang, K. F. Kelly, J. M. Tour. 2009. Molecular Machinery: Synthesis of a
“Nanodragster”. Organic Letters. 11(24) 5602–5605.
Volkoff, O., D. M. Strong, M. B. Elmes. 2007. Technological Embeddedness and Organizational
Change. Organ. Sci. 18(5) 832-848.
Von Hippel, E., G. von Krogh. 2003. Open Source Software and the “Private-Collective” Innovation
Model: Issues for Organization Science. Organ. Sci. 14(2) 209-223.
Ziman, J. 2000. Technological Innovation as an Evolutionary Process. Cambridge University Press,
Cambridge, UK.