The ins and outs of home networking: The case for useful and usable domestic networking.

Full text

Page 1

8
The Ins and Outs of Home Networking:
The Case for Useful and Usable Domestic
Networking
REBECCA E. GRINTER, W. KEITH EDWARDS, MARSHINI CHETTY,
ERIKA S. POOLE, JA-YOUNG SUNG, and JEONGHWA YANG,
Georgia Institute of Technology
and
ANDY CRABTREE, PETER TOLMIE, TOM RODDEN, CHRIS GREENHALGH,
and STEVE BENFORD
University of Nottingham
Householders are increasingly adopting home networking as a solution to the demands created by
the presence of multiple computers, devices, and the desire to access the Internet. However, current
network solutions are derived from the world of work (and initially the military) and provide poor
support for the needs of the home. We present the key findings to emerge from empirical studies
of home networks in the UK and US. The studies reveal two key kinds of work that effective
home networking relies upon: one, the technical work of setting up and maintaining the home
network,andtheother,thecollaborativeandsociallyorganizedworkofthehomewhichthenetwork
is embedded in and supports. The two are thoroughly intertwined and rely upon one another
for their realization, yet neither is adequately supported by current networking technologies and
applications. Explication of the “work to make the home network work” opens up the design space
forthecontinuedintegrationofthehomenetworkindomesticlifeandelaborationoffuturesupport.
Key issues for development include the development of networking facilities that do not require
advancednetworkingknowledge,thatareflexibleandsupportthelocalsocialorderofthehomeand
the evolution of its routines, and which ultimately make the home network visible and accountable
to household members.
This research was supported by the National Science Foundation CNS-#0626281 and the EPSRC,
through the Equator project.
R. E. Grinter and W. K. Edwards were resident at the University of Nottingham when this research
was conducted.
Authors’ addresses: R. E. Grinter, W. K. Edwards, M. Chetty, E. S. Poole, J.-Y. Sung, and J.
Yang, School of Interactive Computing, College of Computing, Georgia Institute of Technology,
Atlanta, Georgia 30308; email: {beki, keith, marshini, erika, jsung, jeonghwa}@cc.gatech.edu. A.
Crabtree, P. Tolmie, T. Rodden, C. Greenhalgh, and S. Benford, School of Computer Science and IT,
University of Nottingham, Jubilee Campus, Wollaton Road, Nottingham, United Kingdom; email:
{axc. pdt. tar. cmg. sdb}@cs.nott.ac.uk.
Permission to make digital or hard copies of part or all of this work for personal or classroom use
is granted without fee provided that copies are not made or distributed for profit or commercial
advantage and that copies show this notice on the first page or initial screen of a display along
with the full citation. Copyrights for components of this work owned by others than ACM must be
honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers,
to redistribute to lists, or to use any component of this work in other works requires prior specific
permission and/or a fee. Permissions may be requested from Publications Dept., ACM, Inc., 2 Penn
Plaza, Suite 701, New York, NY 10121-0701 USA, fax +1 (212) 869-0481, or permissions@acm.org.
C ?2009 ACM 1073-0516/2009/06-ART8 $10.00
DOI 10.1145/1534903.1534905 http://doi.acm.org/10.1145/1534903.1534905
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 2

8:2
•
R. E. Grinter et al.
Categories and Subject Descriptors: H5.m [Information Interfaces and Presentation]: User
Interfaces—Miscellaneous
General Terms: Human Factors
Additional Key Words and Phrases: Home networking, human computer interaction
ACM Reference Format:
Grinter, R. E., Edwards, W. K., Chetty, M., Poole, E. S., Sung, J.-Y., Yang, J., Crabtree, A., Tolmie, P.,
Rodden,T.,Greenhalgh,C.,andBenford,S.2009.Theinsandoutsof homenetworking: Thecasefor
useful and usable domestic networking. ACM Trans. Comput-Hum. Interact. 16, 2, Article 8 (June
2009), 28 pages. DOI = 10.1145/1534903.1534905 http://doi.acm.org/10.1145/1534903.1534905
1. INTRODUCTION
In the last decade, Human-Computer Interaction (HCI) research has moved
out of the office and into the home. While much is different between the two
domains, the increased presence of a home network, based on the principles of
networking designed for the military-industrial complex, suggests an impor-
tant relationship. Simply put, the home is an increasingly networked entity,
comprising a multitude of connected devices and services distributed through-
outthehome.Thistrendtowardsthehomenetworkisbeingdrivenbyanumber
of needs: sharing a single computer in a single location is an increasingly un-
realistic proposition for household members; other devices in the home have to
share access to the Internet; devices also and increasingly need to communicate
with each other as well. Consequently the networked home has rapidly become
a part of domestic computing.
The purpose of this article is to present the results of a series of research
studies that explore the work householders engage in to embed computing in
domestic life and thus “make the home network work.” These studies were
conducted in two countries, the United Kingdom (UK) and the United States
(US). Each focused on the work required to integrate the network into everyday
life in the home [Chetty et al. 2007; Grinter et al. 2005; Tolmie et al. 2007].
Prior to presenting our studies, we review research related to four key features
of technology uptake and use in the home: adoption of domestic technology,
communication, computing, and networking. We then describe our methods,
and the households that participated in our research. Our reflection on these
studies is presented in two sections each with a distinct focus. First we describe
the work it takes to set up a network and the devices on it and then, second,
we consider the practical work required for ongoing maintenance, which we
characterizeasdigitalhousekeeping.Finally,indiscussion,weturntoourmain
point which is that the work of domestic networking relies a) on technical work
to construct the network that is poorly supported by design at this point in time
and b) on collaborative work that embeds the home network in everyday life
and opens up new possibilities for design. We conclude that our studies suggest
the need for HCI involvement in a more radical reconsideration of the nature
of the networked domestic infrastructure.
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 3

The Case for Useful and Usable Domestic Networking
•
8:3
2. RELATED WORK: STUDIES OF THE DOMESTIC
Over the last ten years HCI research has built a significant corpus of knowledge
aboutdomestictechnology.InthissectionwereviewkeyliteraturefromtheHCI
community, as well as work that predates HCI but speaks to questions of tech-
nology in a domestic context. We caveat this section by noting two things. First,
much research has focused on industrialized nations—therefore it is important
to state that this research (our own included) makes certain assumptions may
not follow in other parts of the world where domestic technologies are begin-
ning to take hold, such as, for example, the so-called Global South. Second,
given that our empirical work took place in the UK and the US, we also also
focused on related work that gave us context for understanding the history of
domestic technology adoption in these two nations. Nevertheless, while we rec-
ognize that the ways in which technology is woven into everyday life exhibits
distinct cultural differences, the current limitations of network technology and
the collaborative character of the work that use necessarily relies on speak to
broad concerns. After all, whether we live in New York or Yemen we still have
to create and maintain the home network and incorporate it into the cultural
milieu. Creating and maintaining the local home network and incorporating
the network into the immediate social environment are issues that cut across
a great many cultural distinctions and divides.
2.1 Adopting Domestic Technologies
Although a relatively new focus for HCI, the study of the history of the adoption
of domestic technologies is an established concern. For example, at the turn of
the 20thcentury domestic scientists focused on studying a variety of technical
systems and their role in supporting “housework.” In the 1950s Lillian Gilbreth
applied scientific management techniques, and time and motion studies in par-
ticular,toredesignbothdomesticpracticeandthephysicallayoutofthehometo
make housework more efficient [Gilbreth et al. 1954]. The results of Gilbreth’s
“domestic engineering” influenced the built structure of the home, shaping the
“magic triangle” kitchen layout that became widespread throughout the West-
ern world. Indeed, the 1950s were a boom time for domestic engineering as
white goods and a host of novel labor-saving devices flooded the market. Their
uptake was accompanied by a growing understanding of the systems that com-
prise the home: new indoor plumbing facilities and how they supported the
movement of water into, and waste out of, the home [Leavitt 2002].
Theinterrelationbetweentheseemergingtechnologiesandhomelifebecame
a subject of some interest. In the 1940s and 1950s, in both Canada and Sweden
women-led groups focused on consumption and its relationship to homemak-
ing emerged to tackle debates about what consituted enough work [Parr and
Ekberg 1996]. In the 1960’s feminist scholars took up the question of gender
and domestic technologies [Ravetz 1965]. More recently, Cowan’s seminal study
of housework, for example, focuses on the gap between the promise of domestic
technologies and the reality of housework [Cowan 1983]. She notes that while
domestic technologies such as washing machines were marketed to women as
labor-saving devices, their adoption did not save as much time as promised.
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 4

8:4
•
R. E. Grinter et al.
Rather, it triggered a change in expectations about how often people would
change their clothes. So although washing was no longer a fully manual chore,
the amount to be washed increased significantly. More generally, Cowan argued
for a broader analysis of what it means for technology to be adopted and that
there is a particular need for us to explore the patterns of action and interac-
tion that actually surround a system. This, in turn, may enable us to establish
whether initial visions of labor saving (or any other promissory note) inherent
within a technology actually exist in practice, or whether work shifts from one
type of activity to another.
Surveys of labor in the home continue to suggest that women bear the chief
responsibilityforhouseworktaskssuchascookingandcleaning.However,other
studies also point to different kinds of home labor (such as Do It Yourself), and
include activities such as the assembly and maintenance of home electronics
including audio-visual (AV) systems [Gelber 1997]. In our work we took a broad
view of what constituted digital housekeeping, focusing on the work being done.
Whatever the actual status of domestic work with respect to its gendered
character, it is clear that the study of domestic technologies has a long history of
looking at the relationship between infrastructure (including the built environ-
ment), the technology within it, and the work involved in using it. A dominant
analytic perspective has emerged that focuses on the relationship between gen-
der and technology, particularly on understanding who is doing the work and
accounting for the role of technology in terms of the broader social order. The
research reported here suspends a concern with gender in the study of technol-
ogy [Chetty et al. 2007; Grinter et al. 2005; Tolmie et al. 2007] and focuses, for
reasons of design, on the practicalities of technology use in the home. We are
less concerned at this point in time with whom as we are with what, though we
take seriously the advice that we should look at the actual patterns of action
and interaction or ‘work’ that surround technology. In this respect we suggest
that our research may offer scholars of domestic life concrete insights into the
nature of domestic work, how technology supports or hinders that work, and
what the relationship between technology and housework encompasses.
2.2 Domestic Communications
Ever since the widespread adoption of the home telephone people have asked
questions about the impact of communication technologies on domestic life. In
the early 1900s, for example, executives of the Bell System (which was grow-
ing in dominance in the US telephone industry) worried about the recreational
uses of the telephone. Indeed they went as far as to actively discourage the
use of the telephone for social calls through advertising. Nonetheless it was the
sociability of the phone, and its use in the domestic context in particular, that
drove uptake and use and ultimately shaped the adoption patterns associated
with this technology in the US at least [Fischer 1997]. Studies of communi-
cation technologies in the home have been and are perennially accompanied
by concerns with its effects on people, children and teenagers included. With
respect to the home telephone researchers have suggested that the technology
“depersonalizes” communication and has been responsible (at least in part) for
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 5

The Case for Useful and Usable Domestic Networking
•
8:5
the increased sexual liberation of American teenagers [Lynd and Lynd 1929].
Similarconcerns,andmoralpanics,havepermeatedanalysesofsuccessivecom-
munications technologies, including radio, television, mobile phones, and com-
puters [Millwood Hargave and Livingstone 2006]. Discussion of the impacts of
communications technologies on the inhabitants of the home is now a standard
part of social commentaries on what it means to communicate from within the
home and, in turn, what it means to use digital technologies for that purpose.
The telephone still plays a central role in domestic communications
[Anderson et al. 2002; Anderson et al. 1999; Palen and Salzman 2002; Palen
et al. 2000] and despite the intentions of those early telephone executives, it re-
mainsadeviceinusetopromoteandreinforcefamilialandsocialties[Anderson
et al. 2002; Anderson et al. 1999]. However, the last decade has seen the whole-
sale arrival of the mobile phone and computer in the home, both of which have
come to be used within the context of domestic communication. Examination
of Short Messaging Service (SMS)—a text-based communications system orig-
inally available on Groupe Sp´ eciale Mobile (GSM) networks but subsequently
replicated on other wireless systems—provides some insight into the roles that
domestic communications fulfil. SMS research has often focused on teenagers
across Europe and Asia because they were among the earliest adopters of this
technology for communications purposes [Grinter and Eldridge 2001; Harper
et al. 2005; Ito et al. 2005; Ling 2000; Taylor and Harper 2002]. These stud-
ies report a variety of findings that uncover how SMS fits into the everyday
circumstances that teenagers find themselves in: being able to transcend the
physical limitations of circumstance, for example, and talk to friends at a dis-
tance; working around the schedule constraints imposed by the family; and
using SMS to coordinate activities in real time as opposed to having to arrange
events and meetings in advance, etc.
Just as mobile phones have been rapidly appropriated into domestic com-
munication, so too have computers. In addition to email, bulletin boards,
chatrooms, and Instant Messaging have all found a place within the home
[Grinter and Palen 2002; Livingstone 2002; Turow and Kavanaugh 2003;
Wellman and Haythornthwaite 2002]. Indeed in a comparative study that
sought to answer the question of what drives people to use computing, Kraut
et al. [1999] found that householders tended to be drawn to communication ac-
tivities over information activities. Recent surveys in the UK similarly revealed
that over 75% of computer use in the home revolves around communication,
with 99% of survey respondents saying they use it to read and send email; 56%
for instant messaging; 26% for chatrooms, and 13% for Internet telephone [OII
2009]. For over a century the landline telephone has been one of the primary
technologies of communication for householders. However, in the last 20 years
other technologies, mostly noticeably the mobile phone and the computer, have
augmentedit.Further,someofthesenewdigitaltechnologiesandmeansofcom-
munication have been built on the telephone’s infrastructure. Today modems,
Digital Subscriber Line (DSL) routers, and WiFi piggyback on a global com-
munications infrastructure and are rapidly populating the home to make new
forms of communication possible. For example, research in Ubiquitous Comput-
ing has explored a range of ambient as well as direct communications devices,
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 6

8:6
•
R. E. Grinter et al.
such as the InterLiving Project’s InkPad system that allows family members
to draw on a computer surface and by so doing share notes between houses
[Lindquist et al. 2007] or the Digital Family Portrait that supports the sharing
ofactivitymovement[RowanandMynatt2005],orPlantDisplaythatvisualizes
how much time people spend communicating [Kuribayashi and Wakita 2006].
2.3 Domestic Computing
Mirroring previous interests surrounding other technologies, the growing pop-
ularity of personal computing throughout the 1980s led to the emergence of
domestic computing as a distinct research focus. One of the earliest studies of
domestic computer use highlighted the role of computers in telecommuting—
using a modem in the home to connect to the corporate network. In the mid-
1980s telecommuting was in its infancy and being positioned as a new mode
of working. An initial survey of 282 homes [Vitalari et al. 1985] highlighted
the makeup of these early adopters of computers revealing (somewhat unsur-
prisingly) that 96% of telecommuters were male with a higher than average
education level, that 63% of respondents reported being in a technical profes-
sion, and that the computer was used for purposes of work. More significantly,
they noted that time spent on the computer was a trade-off against other ac-
tivities that take place in the home, and that having a computer at home was
a significant commitment, requiring technical knowledge to set up, run, and
maintain. Ten years later Venkatesh [1996] saw different uses of the computer
in the home. The rapid adoption of Internet technologies—in particular email
and the web—had changed what was possible to accomplish with a computer at
home. Domestic computer use was becoming increasingly diverse (a trend that
continues today, and has been well captured by HCI research). Studies showed
that in addition to telecommuting, recreational uses were also emerging which
focused on using the resources of the Internet in support of home leisure activi-
ties [Kraut et al. 1996; Lally 2002; Venkatesh 1996]. In the UK, a 2005 national
survey reported that 77% of respondents reported using the Internet to plan
and make travel arrangements, 54% for downloading music, 48% for playing
games, and 18% for managing their photo collections. 50% used it for online
shopping, 45% for banking, and 30% for paying bills [OII 2009]. Outside of this,
40% used it for accessing central and local government services. Creative use
such as developing web pages and blogs was reported as being engaged in by
20% of the survey population and surfing the net, particularly in relation to lo-
cal concerns (such as weather, traffic, and local news information) was reported
by almost all respondents.
While communication remains at the forefront of computer use in the home,
domestic computing has clearly burgeoned over the last decade. This has been
accompanied by a growing interest in domestic computing by the HCI commu-
nity to explore the possibility of designing applications for the home [Mateas et
al. 1996; O’Brien and Rodden 1997] or understand the full spectrum of domestic
activity [Taylor et al. 2008; Wyche and Grinter 2009]. One major line of domes-
tic research has focused on understanding the routines of the home, including
those implicated in technology use. This research has sought to explicate the
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 7

The Case for Useful and Usable Domestic Networking
•
8:7
ways in which householders collaboratively organize and conduct all aspects of
domestic life. Through detailed attention to the nature of household routines,
this line of research has sought to understand the barriers to technology adop-
tion in the home [O’Brien, et al. 1999], provide insight into what it means to
design technologies that can be incorporated into everyday life in the home
[Tolmie, et al. 2002], and highlight opportunities for richer domestic technol-
ogy design that resonates with the demands of the home [Crabtree and Rodden
2004; Rodden et al. 2004; Taylor and Swan 2005]. Another line of research
has focused on the use of novel methods that suspend a concern with rational-
ity, functionality and utility to explore domestic values, often in playful and
provocative ways [Gaver et al. 1999; Gaver and Martin 2000; Hutchinson, et al.
2003; Lindquist et al. 2007]. Given the centrality of communication within the
home, a number of systems have also been built and deployed in homes to ex-
plore new types of technologies to support communication. Some systems have
explored the possibility to extend explicit communications—for example, pro-
viding new mechanisms for holding conversations [Hindus et al. 2001]. Another
approach has focused on facilitating communications by raising awareness of
the whereabouts of family members [Brown et al. 2007]. Instead of being de-
signedforcommunicationdirectly,thisclassofsystemshelpshouseholderstake
advantage of potential opportunities to communicate as needed.
Behind much of this research is the presence or assumed presence of a net-
work connection, both to connect the home to the digital world beyond the front
door and to connect to an increasing array of devices and services distributed
around the home. Network access is increasingly central to the computer’s do-
mestic utility, so much so that a great many households have expanded the net-
work from a simple Internet connection for a single computer, to an intrahouse-
hold network that shares the network connection across multiple machines and
devices (such as Personal Video Recorders and dedicated game consoles), and
supports services within the home (such as networked music players and home
file servers). And it is the desire of householders to have networked applications
and services that is driving commercial broadband and wireless service growth.
And yet, although the infrastructure underpinning intrahome networks holds
muchpromise,researchalreadyshowsthattherealizationofthehomenetwork
does not come without complications.
2.4 Domestic Networking
Given existing interest in both communication and computing use in the home,
the explosive growth of the network during the early 1990s raised questions
about how the spread of networking beyond a technologically sophisticated mi-
nority would impact the household. As early as 1996 research was showing
that domestic networking was proving difficult. Franzke and McClard [1996]
and Kiesler et al. [2000] reported how difficult users found creating even the
simplest network case: connecting one computer to the Internet. Their findings
stressed how participants needed technical knowledge to diagnose and deal
with networked technologies, and that they turned to friends and family for
help, a finding also observed in more recent research [Poole et al. 2009]. Over a
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 8

8:8
•
R. E. Grinter et al.
decade since, two things have changed. First, the field of Ubiquitous Computing
has emerged with strong visions and an aggressive research agenda to build
smart homes and then leverage that intelligence to provide new classes of ap-
plications to support domestic life [Intille 2002; Kidd et al. 1999]. Second, an
increasing number of homes have adopted more complex networks that connect
large numbers of devices inside the home together [Horrigan and Rainie 2002].
However, these two developments point to a serious tension or gap [Shehan and
Edwards 2007]: the promise of future applications rests on the ability of house-
holders to manage the home network, something that our collective research
shows has not become easier since the first reports of connecting computers to
the Internet.
We are not the first to comment on the complexity of technology demanded by
the Ubiquitous Computing agenda [Harper 2003]. In a seminal study of the Or-
ange Smart Home, Randall [2003] identified that even in use, the smart home
creates a paradoxical situation for its residents. He learned that for the resi-
dents, albeit temporary ones in a research setting, the control system designed
to provide householders with increased ability to manipulate the home was cre-
ating sufficient confusion as to leave them feeling out of control. Others have
focused on the work needed to understand what it means for householders to
be able to participate in or completely own the experience of making their home
smart. Edwards and Grinter [2001] observe that for most people, and unlike
most laboratory-built smart homes, making the home smarter means adding
technologiestoanexistingstructure,ratherthancommissioningabuildertode-
sign a house that is smart from the ground up. This theme was further taken up
by Rodden and Benford [2003] who applied Brand’s [1994] architectural frame-
work to analyse the complex relationship that potentially exists between the
structureofthehomeandthetechnologiesassociatedwithdomesticUbiquitous
Computing and then further extended to including networked technologies by
Chetty et al. [2007]. Beckmann et al. [2004] focused on the complexity inher-
ent in sensor networks designed to help computers determine the presence and
activity of users through an array of devices that can detect movement, temper-
ature, and so forth. They found, perhaps not surprisingly, that people struggled
to install these types of networks and questioned whether it was appropriate
to gather certain types of data in their homes at all.
The issue of network complexity, and how householders might come to man-
age it for themselves—thereby embedding computing in, and adapting it to, the
ongoing circumstances of domestic life—is the principal concern of this paper.
We begin by noting that, despite the difficulties, households are taking up a
networked domestic life. Accordingly we seek to examine current practices sur-
rounding the networked home and, from those practices, learn about possible
solutions to issues of complexity. Our research shows not only the practices
and routines that have emerged around home networking, but also the ways
in which home networking remains nontrivial for even the most qualified of
people—even those with advanced degrees in Computer Science. The research
sheds light on the Ubiquitous Computing agenda and the real-world character
of “intelligence” in the networked home, revealing what householders are seek-
ing to do with their home networks and how they are making them “fit into”
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 9

The Case for Useful and Usable Domestic Networking
•
8:9
the infrastructures—technical, physical, and social—that inhabit the domestic
setting.
3. METHODS AND PARTICIPANTS
In this section we describe the methods we used to collect and analyse data,
and the participants from whom we collected that data. Our studies took two
somewhat different forms. In the United Kingdom (UK) our research took an
ethnomethodologically informed ethnographic approach. In the United States
(US) we employed a qualitative approach using several different techniques to
draw out the experiences of our participants. We describe each of these in turn.
3.1 Methods and Participants in the United Kingdom
The study in the UK focused upon tracking over time the efforts of three differ-
enthouseholdstoinstallandmaintainhomenetworks.ThestudyranfromMay
to December 2006 and involved a mixture of direct observations and “catch-up
conversations” designed to offer participants an opportunity to report on their
ongoing experiences with the network between observations. Both the ethno-
graphic capture of data and subsequent analysis were conducted from an eth-
nomethodological perspective [Suchman 1987], which is to say that we sought
to describe in fine detail the social organization of the home network as it was
given in the methodical ways that members encountered and managed the net-
work in the course of their day-to-day actions and interactions. Participants in
the study included:
—Household A, consisting of two adults, one male, one female, 44 and 30 years
old respectively, both computing professionals, living in a large two-bedroom
apartment.
—Household B, a family consisting of 2 adults, one male, one female, 38 and
36 years old respectively, and 3 children, one 9 (a girl), one 7 (a boy), and one
15 months (another girl), living in a semi-detached house; one of the adults
is a computing professional, all other members of the household have very
limited technical experience.
—Household C, a family consisting of 2 adults, one male, one female, both 43
yearsold,and2children,one12(agirl)andone9(aboy),alsolivinginasemi-
detached house; once again one of the adults is a computing professional but
alloftheothersinthehouseholdhavenospecializedexperienceoftechnology.
While at least one member of each household was involved in computing
in some sense, none of the households involved in the study could really be
described as having advanced technology set-ups. Instead, as became quickly
evident to us, each of the computing professionals involved was very reluctant
togettooheavilyinvolvedincomputingactivitiesathome.Theprincipalreason
cited for this was that working with computers already consumed a significant
partoftheirday.Havingsomeonetechnicalinthehousedoesnot,itwouldseem,
by any means result in rapid technology adoption. Rather, it transpired that
any technical undertaking in the home was and is accountable to a whole range
of other everyday household concerns. The building of a home network was
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 10

8:10
•
R. E. Grinter et al.
not driven by technical interest then, but was instead motivated by household
members’ concern to develop a solution to burgeoning technological complexity:
multiple computers, multiple devices, and multiple demands being placed on
them by various household members warranted and drove the construction of
home networks for the participants in our study. With broadband connections
amounting to over 70% of all Internet connections in the UK and a rapid uptake
ofwirelesstechnologies,itishardtomaintainthenotionofhomenetworkbuild-
ing being about experimentation done by “geeks.” Instead people are installing
home networks because it makes sense for them to do so in the face of comput-
ing technology that is increasingly distributed throughout the home and used
by a variety of equally distributed different household members. Construction
of a home network is a members’ solution to the problem of distribution.
Our study of network construction and maintenance was conducted through
monthly site visits, which were complemented by regular catch-up conversa-
tions over the telephone. The site visits were a critical part of being able to
witness and understand the reasoning involved in a range of situated activ-
ities associated with building and maintaining the networks. However, as is
apparent in the following sections, it transpired that network setup and main-
tenance rapidly becomes an ongoing and routine feature of the broader pattern
of household activities. We therefore kept in regular contact with the house-
holds in order to capture some of what that ongoing work involved as part of
the participants’ day-to-day experience. The site visits were conducted through
direct ethnographic or participant observation [Crabtree 2003]. This entailed
shadowing participants as they went about their activities in order to produce
a fine-grained or “thick description” [Ryle 1971] of the actions and interactions
involved in setting up and maintaining their home networks. The aim of the
approach was and is to uncover and explicate the various ordinary, in situ, and
frequently tacit competences and collaborative activities through which every-
day courses of action (such as network maintenance) are accomplished and
organized. Data capture here is comprehensive as there is no prior presump-
tion as to what might or might not be significant. Instead the ethnographer
aims to become party to the gamut of lively action and reasoning applied to sit-
uated circumstances as they arise. In this way matters such as network setup
and maintenance are seen as a practical accomplishments done by social ac-
tors performing their activities in the face of a host of local contingencies that
inhabit their work [Garfinkel 1967].
The record of practice that emerges is similarly subject to ethnomethodolog-
ical analysis where the focus is upon what can be learned about recurrent pat-
terns of action and reasoning through the inspection of particular “instances”
[Sacks 1984] in which the members of particular settings engage with and
display their orientation to the ongoing and organized work of those settings
[Button 1992]. Thus, while findings may be articulated through specific field-
work vignettes, it should not be understood that what is being said is only of
relevancetotheparticularsofeachobservedinstance.Insteadweareinterested
in broad characteristics of practical action and practical reasoning that make
homes, households, housework, network maintenance, etc., recognizable as the
organized accomplishments of members—that is, as accomplishments that you
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 11

The Case for Useful and Usable Domestic Networking
•
8:11
or I as well as those studied might recognize as organized accomplishments too.
Thus, as people engage with concerns such as where to put technology, where
to plug things in, how to organize furniture, what to tell children about using
things, and so on, we are interested in both how the particular arrangements
and characteristics shape how technology is “made at home” in some particular
setting, and how the social organization of practical action and practical rea-
soning is of broader relevance to our understanding of home networks and the
potential for their continued development.
3.2 Methods and Participants in the United States
In the US we conducted two studies that followed the same protocol. The pro-
tocol consisted of two steps. First, we asked potential participants to fill out
an inventory of their technology. The inventory was organized into three parts.
Part one focused on technologies, in particular infrastructure technologies such
as home control, security, cable, satellite systems, and network type (e.g., WiFi,
Ethernet). Part two asked the participants to locate technologies in each room
of their house. Part three focused on those technologies that do not tend to
be associated with a specific room, but rather with a particular householder.
The goal of this inventory was to get a sense of the technologies that we were
likely to encounter during the interview and to customize our interview for that
particular household.
The second step of our home protocol consisted of a home visit, which was
scheduled for a time when all members of the household would be present and
available. This was limited to all “typically occupant” householders—we did
not attempt to schedule times when sons and daughters would be home from
university or military service, for example. The home visit itself was broken
into three distinct parts. First, we asked our participants each to independently
sketch their current audio-visual (AV) and computer networks, and then draw
what they thought they would ideally like (for more details see Poole et al.
[2007]). The current network sketching exercise allowed us to understand what
individualparticipantsthoughttheirnetworkcomprised(andfromthiswewere
able to identify differences among householders). We asked about their ideal to
see what people aspired to, particularly where it differed from visions that we
mightholdinHCIorUbicompresearch.Italsoturnedoutthatthesketchesalso
served as a useful tool for “warming up” the participants: being forced to think
explicitly about their network helped them to reconnect to sometimes “invisi-
ble” infrastructure technologies [Star 1999; Tolmie, et al. 2002]. After sketching
we asked the participants to take us on a guided tour of their home, to visit
the locations where they had home networking or AV equipment. We included
AV networks in this study for two reasons. First, AV represents a predecessor
network in the home, meaning that such networks represent, for most people,
an earlier instantiation of a complex constellation of interconnected devices. We
anticipatedthatAVnetworksmighthavesomeinfluenceinhowpeoplehandled
their computer networks. Second, AV and computer networks are increasingly
converging with devices requiring the services of both (i.e., an MP3 player that
gets content from the computer but plays it out through the speakers of the
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 12

8:12
•
R. E. Grinter et al.
stereo). At each location we visited in the home, we asked the householders to
describe what was going on, and prompted them to talk about their network.
Again, we found proximity to some of the devices triggered memories about
victories and disasters associated with networking. When the home tour was
complete, we returned with the participants to the living room to finish asking
questions.
We recruited participants in two metropolitan areas of the United States:
San Francisco and Atlanta. In San Francisco our sample consisted of 8 house-
holds that were made up of two people (a man and woman) with dual incomes
and no children. Our sample in this city intentionally focused on early-adopter
home network users, with complex network needs and configurations, in order
to reveal both possible futures for home networking, and to understand how rel-
atively expert users approach the challenges of networking in the home. Thus,
these couples all had at least one householder with some formal or practical
knowledge of networking, which took the form of an advanced degree in Com-
puter Science, or many years of systems administration experience. In Atlanta,
our sample consisted of 11 households, with a total of 28 individual partici-
pants. We sampled to broaden the types of household we visited beyond simply
early adopters. Six of the households we visited included parents and children.
Despite broadening our sample, all of our participants had higher than average
household incomes (pointing to the costs of home networking, and of the reality
of a “smart home”). More details about the participants and the methods are
availableinourearlierreportsofthesestudies[Chettyetal.2007;Grinteretal.
2005].
4. RESULTS: CONSIDERING THE NETWORK FROM WITHIN AND WITHOUT
In this section we present the results of our collective research organized into
two sections that broadly reflect the work of “digital housekeeping” [Tolmie
et al. 2007]. First we wish to reflect on the work required to introduce new
devices and services into the home and to make them fit into the network and
the household. Second, we talk about the ongoing work required to keep the
devices and services working once they have been configured. In both sections,
we wish to stress two interrelated types of work, the work to understand and
work with the network as a technical artifact, and simultaneously the need to
manage it as a social artifact.
4.1 Setting Up Technology in the Home
A common theme that emerges across our research in both the UK and US is
the complexities that households face in setting up their technology at home
and making it part of their home network. This complexity manifested itself
in positioning the technology, maintaining the wider order of the home, and
planning for change. We describe each of these in turn.
Many, if not all, of the households we visited discussed issues of locating tech-
nology inthehome—thatis, wheretheyinstalledthe devicesthatmadeuptheir
home networks and why they chose particular places. In these conversations,
a number of properties that home networks must fulfill came to light. Some
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 13

The Case for Useful and Usable Domestic Networking
•
8:13
stemmed from the physical and infrastructural properties of the house itself,
while others spoke to the domestic order of the household, family members’
patterns of action and interaction with each other, and quotidian “logics” that
organize the home and the work within it. The physical properties of the house
required that householders reason about the network in a variety of ways. The
presence of wireless networks, marketed in part to work around the constraints
of the physical house, did not always mean that our participants did not have to
think about their home. For example, participants described learning about the
physical properties of their homes, such as the thickness of the walls, through
their experiences of locating their wireless base station(s) within their home.
Householders told us about the reach of their wireless networks, the strength of
the signal, and the places where parts of their home hidden behind their walls
(load-bearing walls being attributed as being particularly thick) blocked or re-
duced their network connectivity. Wireless networks—both those belonging to
the household, and those belonging to neighbors—appeared at various places
in the physical environment of our participants’ homes. Some families talked
about not only where they could and could not connect to their own network,
but also where they saw someone else’s signal. Indeed, we saw evidence in the
form of repositioned furniture to capitalize on free network access in parts of
the home that their own household wireless networks did not cover.
The need for power was a significant constraint on device location. Devices
had to be situated by the wall jack, or power had to be “moved” to where house-
holders wanted the technology. In older houses (where power outlets were less
common), literally fulfilling the need for power required the development of
complex schemes including plugging multiple extension cords into each other
(forming a chain) in order to cope with a situation where there was a jack with
at most two outlets serving between six and ten devices competing for power.
In some cases infrequently used devices might be disconnected, but many com-
ponents (e.g., the television, the router) occupied such a central role that they
had to be constantly plugged in. Participants recognised that the overloading
of wall jacks with chains of extension cords was hardly the safest solution. As
we discuss below, the presence of young children can make this type of arrange-
ment unacceptable. Another problem we found with this solution was that the
circuitry could not meet the power demands of the devices, leading to other dif-
ficulties. In one case, it took the householders several months of living with a
problem before they figured out that when they turned on certain combinations
of devices, their circuit would fail to provide enough power (brownout), which
in turn caused their router to lose its IP address, thus disabling the network.
After they finally determined the problem with their router, this put the house-
hold into the position of having to redesign their network to accommodate their
electrical wiring.
Despitetheobviousneedtoconnectdevicestoapowersupply,extensioncords
also served another purpose, to move devices so that they were positioned in the
placesthathouseholderswanted.Inthesecases,welearnedaboutlocationsthat
were grounded in our shared sense of where certain activities should happen.
Discussionsabouttheshapeoftheroom,thepossibilitiesofarrangingfurniture,
and most importantly expectations about what types of activities took place
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 14

8:14
•
R. E. Grinter et al.
where (which orientation, with access to what lighting, because the furniture
would not fit any other way, because it was important to have access to other
parts of the home, etc.) led our householders to arrange their rooms and in so
doing configure their activities in particular ways. They then worked hard to
ensure that devices could be connected into that space in ways that fitted into
that usage plan.
Another type of consideration that emerged in our households focused on
children, who often surfaced a set of logics concerned with physical safety of
both children and household devices. Households confronted the challenges
posed by children by putting devices in places where they could not be reached,
and making sure that wires were installed in such a way that they would not be
tripped over. Decoration also mattered in the sense that tidiness and appropri-
ate visibility/invisibility came up for our householders. We found cases of DSL
modems being hidden under couches, because householders did not like to see
the blinking lights in their living rooms. We saw “nests” of wires behind the
backs of televisions and other large devices in order to hide them from view.
In one instance, having abandoned an attempt to hide these wires, one family
had decided instead to decorate them, as if to acknowledge their presence and
attempt to integrate them more aesthetically into the home. In another case
potentially unsightly nests of cables were placed on a windowsill where preex-
isting clutter would render them relatively invisible, stressing that invisibility
for householders is often not about literal absence of perceptual availability but
rather a matter of making features wholly commonplace [Tolmie et al. 2002].
These ecological placements—which are rooted in the household’s desire to
have their network reflect their household order—turned on not just the power
needs of devices but also data requirements. Wireless networks provided some
flexibility in this regard, but we also encountered other types of networks in use
in order to allow data to reach a device. Both phoneline and powerline bridges,
technologies that use the existing home phone and electrical wiring infrastruc-
ture, respectively, to pass data traffic, were present in some of the homes we
visited. Participants explained that these technologies allowed the home net-
work to evolve in ways that supported how they used their physical space.
Relatedtotheworkofpositioningparticularnetworkelements,wealsoheard
about the work of making changes while retaining wider order during instal-
lation. Installing a single device, or making more complex changes and/or en-
hancementstothenetwork,welearned,wastypicallyacomplexactivityinvolv-
ingdisruptiontothehousehold.Inmostbutnotallhouseholds,wesawthatthis
activity tended to fall to one individual—typically a person with some type of
formal knowledge (either acquired through formal education in Computer Sci-
ence and related disciplines and/or through professional experience of holding
jobs that entailed some level of computing competence) took up the responsibil-
ity associated with making changes to the network. The fact that this pattern
repeated itself across the homes in our collective studies immediately suggests
one challenge for Ubiquitous Computing and related disciplines, the necessity
of not requiring such specialized knowledge in order to set up, maintain, and
evolve home networks. This is particularly important if Ubiquitous Computing
is to reach out to broader sections of the population.
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.

Page 15

The Case for Useful and Usable Domestic Networking
•
8:15
The person responsible for these changes often described undertaking two
kinds of related activities. First, the person thought about the addition of new
technology, or the reconfiguration of the network, in its technical terms. By this,
we mean that part of this job was to consider the entire network, and its topol-
ogy,tomakesurethattheundertakenchangenotonlyprovidedthedesiredgoal
but also did not break existing services. For many of our householders this goal
turnedouttobesurprisinglyhard.Justknowingwhatservicesthenetworkwas
providing was sometimes more complicated than it might appear. For example,
households with teenagers—who reported making their own changes and mod-
ifications to the network—sometimes did not have a unified knowledge of what
was on their network or the services it provided (and where it delivered those
applications).Inothercases,weheardaboutseldom-usedservices,easilyforgot-
ten when making changes and discovered only after the change had been made.
This set of considerations was closely coupled to a broader, out of network,
set of plans that also needed to be made. Additions and changes were often
described as being very disruptive. The physical mess associated with the tech-
nologywasoftenungainly,clutteringupspacearoundthenetwork.Butthiswas
not the only problem that these householders considered: they also described
thinking about the network as a collection of services embedded in the broader
routines of the home. Changes to the network frequently meant disruption to
those routines. Additions of print servers might mean that printers would be
offline; recabling an audio-visual network to include a Personal Video Recorder
or iPod would mean that others could not watch television or use the receiver.
For all these reasons, making changes to the network was frequently de-
scribedintermsthatemphasizednotonlythemodifications,butalsothemeans
by which order within the network and the home would be preserved. These
changes were also sometimes framed in terms that highlighted the anxiety
and apprehension about the complexities of maintaining order while making
change. In fact, one family who had been trying to resolve a problem on their
network described how it had been on their to-do list, in one form or another
(referring to the different solutions that they had tried) for over 3 months, and
because of that they saw any new change as one of great risk. One solution
that some home administrators used was to minimize the disruption caused by
change by waiting for an appropriate time, such as when other householders
were not present. In a more ambitious case, when a household had decided to
install an Ethernet-based network, they waited until their home was under-
going renovations to make the change, leveraging a more intense change to
the physical structure of the home as an opportunity to make this more minor
modification.
In addition to making decisions about the location of devices, and thinking
about how to make changes while maintaining order, other planning activities
also took place. This planning work had two foci: first, the technical work of
making the network work, and second, making the system fit into the domestic
order of the household. The technical network focus manifested itself in at least
two ways. First, we were amazed, particularly in households that had elaborate
home networks (multiple machines and subnets, wired and wireless, powerline
and phoneline, connections to one or more corporate network and so forth), by
ACM Transactions on Computer-Human Interaction, Vol. 16, No. 2, Article 8, Publication date: June 2009.