Content uploaded by Kentaro Toyama
Author content
All content in this area was uploaded by Kentaro Toyama on Dec 04, 2015
Content may be subject to copyright.
Content uploaded by Kentaro Toyama
Author content
All content in this area was uploaded by Kentaro Toyama on Sep 11, 2014
Content may be subject to copyright.
Where There’s a Will There’s a Way:
Mobile Media Sharing in Urban India
Thomas N. Smyth
†
thomas.smyth@gatech.edu
Satish Kumar
‡
satish_282003@rediffmail.com
Indrani Medhi
‡
indranim@microsoft.com
Kentaro Toyama
¶
kentaro_toyama@hotmail.com
†
School of Interactive Computing
Georgia Institute of Technology
Atlanta, GA 30332, USA
‡
Microsoft Research India
196/36 2
nd
Main
Sadashivanagar
Bangalore 560 080, India
¶
School of Information
University of California, Berkeley
Berkeley, CA 94720, USA
ABSTRACT
We present the results of a qualitative study of the sharing
and consumption of entertainment media on low-cost
mobile phones in urban India, a practice which has evolved
into a vibrant, informal socio-technical ecosystem. This
wide-ranging phenomenon includes end users, mobile
phone shops, and content distributors, and exhibits
remarkable ingenuity. Even more impressive is the number
of obstacles which have been surmounted in its
establishment, from the technical (interface complexity,
limited Internet access, viruses), to the broader
socioeconomic (cost, language, legality, institutional rules,
lack of privacy), all seemingly due to a strong desire to be
entertained.
Our findings carry two implications for projects in HCI
seeking to employ technology in service of social and
economic development. First, although great attention is
paid to the details of UI in many such projects, we find that
sufficient user motivation towards a goal turns UI barriers
into mere speed bumps. Second, we suggest that needs
assessments carry an inherent bias towards what outsiders
consider needs, and that identified “needs” may not be as
strongly felt as perceived.
Author Keywords
mobile phone, video, Bluetooth, sharing, social networking
ACM Classification Keywords
H.5.m [Information interfaces and presentation]:
Miscellaneous; K.4.m [Computers and society]:
Miscellaneous.
General Terms
Design, Human Factors
INTRODUCTION
A growing body of work dubbed ‘information and
communication technologies for development’, or ‘ICT4D’,
investigates the design of technological systems for
communities in developing countries, in the hopes that
information technology can play a role in improving lives
and alleviating poverty. The CHI community has shown a
steadily increasing interest in this topic. Projects in the area
often take the form of interventions, in which a novel
technology is introduced into a community with the goal of
improving its socio-economic situation. But as practitioners
in this field are intimately aware, the proportion of ICT4D
interventions which have meaningful, sustained impact is
very small.
One well-studied example is the telecenter
1
, in which a PC,
often connected to the Internet, is meant to serve as the hub
of socio-economic development activity, usually in poor
rural or urban areas [24]. Telecenter projects are usually
meant to provide social services such as remote healthcare,
distance education, or e-government. However, research
suggests that the vast majority of telecenters have little
measurable impact [19], and that many, if not most, close
operations within a few years. Failures are attributed to a
multitude of issues, some of which are well within the
bounds of consideration of HCI, including socio-cultural
barriers, lack of demand, mismanagement, economic
constraints, poor infrastructure, educational deficiencies of
potential users, and absence of sustained political support
[8,11]. These issues appear to plague projects well beyond
telecenters, including some of those with an HCI focus.
Due to the difficulty of crafting successful interventions,
developing regions are often described as environments
1
We use the example of telecenters here since there is a
relatively robust literature in cataloguing their successes
and failures. The same cannot yet be said for more recent
projects in HCI for development.
Permission to make digital or hard copies of all or part 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
bear this notice and the full citation on the first page. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior
specific permission and/or a fee.
CHI 2010, April 10–15, 2010, Atlanta, Georgia, USA.
Copyright 2010 ACM 978-1-60558-929-9/10/04....$10.00.
which are unsupportive of or unprepared for advanced
information technologies. Successful projects in ICT4D, in
fact, are generally portrayed as those that not only provide
new technology, but also intervene to overcome all of the
associated obstacles [4,14,18,23]. Rarely do researchers ask
whether these projects are meeting real demands.
This paper tells a different kind of story. We focus on the
sharing and consumption of entertainment media on low-
cost mobile phones, a practice which has evolved into a
vibrant socio-technical ecology in urban India. With the
availability of multimedia-capable phones for less than $50,
many users who have previously owned simple bottom-of-
the-line phones are upgrading. In turn, this expanding
infrastructure is supporting a phenomenon analogous to the
online file-sharing craze that developed on the Internet with
the advent of Napster. But in India, where Internet access is
not yet widespread, sharing is instead carried out via phone-
to-phone file transfer.
The system we describe here is complex. It involves end
users, who trade with each other; mobile phone shops, who
sell content to end users; and content suppliers, who sell
batches of content downloaded from the Internet to mobile
shops. This system has grown up through organic
processes, sanctioned or promoted neither by mobile phone
manufacturers nor service providers. This has occurred in
spite of the same myriad obstacles that interventionist
projects like telecenters have faced—from the technical
(interface complexity, limited Internet access, viruses), to
the broader socioeconomic (cost, language, institutional
rules, lack of privacy). The phenomenon also spans a range
of socioeconomic strata, and includes users with limited
incomes, education, and literacy skills.
The key difference between this video-exchange economy
and those of many ICT4D projects appears to be one of
motivation. Users themselves are strongly motivated by a
desire to be entertained. On the other hand, the so-called
user “needs” – of healthcare, education, etc. – addressed by
interventionist ICT4D projects appear not to be as urgently
felt, though they are undoubtedly important.
This raises several points of discussion. First, it suggests
that the motivation to adopt, if powerful enough, trumps
obstacles in the path of adoption—in short, where there’s a
will, there’s a way. Second, it suggests that developing
communities are more than capable of embracing advanced
information technologies if the appropriate motivation
exists. Finally, it suggests that we must reflect on the very
concept of ‘needs’ in the context of technology and
development: Are needs really more urgent than desires?
Who defines which is which? Do researchers exaggerate the
urgency of ‘needs’ due to their own biases and
preconceptions?
In this paper, we discuss each of these questions with
reference to the data we have collected on the
aforementioned mobile media sharing phenomenon. We
give special attention to the issue of usability, which
appeared to be of minimal concern to the users in our study.
BACKGROUND
Technological innovation and popular media go hand in
hand in India. In a 1993 book, Peter Manuel chronicles the
rise of the audio cassette in northern India, finding that the
low cost of cassette tapes had transformed the music
industry from a near-monopoly into a near-free-for-all, with
considerable impact on patterns of dissemination and
consumption [12]. In fact, variants of many of the practices
reported here have been observed in previous technological
generations. Digital audio and video carried on multimedia-
capable mobile phones exhibit many of the same
characteristics as cassette media (cheap, easy to copy) and
promises this time to affect both the music and film
industries.
Mobile Multimedia Sharing
With the worldwide explosion of mobile phones and
portable digital media players in recent years, it is not
surprising that the literature on mobile multimedia is a rich
one. Here we review selections that focus on media sharing.
Some of the earliest work in this area considered the
experience of live television on PDAs or custom devices
(e.g. [6,21]), but no wireless sharing was reported.
Kindberg has presented an in-depth study of camera-phone
use [9], but also did not find much evidence of file sharing,
instead finding that most sharing was done on-screen.
Several other studies have investigated the potential of
Bluetooth for communication with public displays.
Chevrest et al. [5] built a system allowing users to send and
receive files to and from a public photo display, while
Maunder et al.[13] proposed a similar system which
removed the need for special phone software. Neither of
these articles dealt with phone-to-phone transfers.
At least two reports examined peer-to-peer Bluetooth media
sharing in some detail. O’Hara et al. encountered the
phenomenon in her study of U.S. and U.K. mobile video
users [16], finding it to be common in high schools. Harper
et al. observed Bluetooth sharing in a study of ‘heavy
mobile phone users’ in the U.K. [7]. They dubbed the
practice ‘trafficking’, remarking on its peculiar social, legal,
and economic characteristics.
Our work is distinct from these studies for two reasons.
First, whereas other studies encountered sharing only in
passing, the sole focus of our study is media sharing, and
this focus permits a deeper examination of practices on the
ground and in the wild. Second, our work is based in India,
a developing country, which features more barriers to the
development of such an ecosystem.
Challenges and Motivations in ICT4D
A collection of ICT4D literature examines the all-too-
common phenomenon of project failure. For example,
Heeks has surveyed the success and failure of a host of e-
government initiatives—those which aim to computerize
basic government services in developing countries—and
painted a bleak picture in which 85% of projects were either
total or partial failures [8]. He attributed these failures to
“gaps” between project design and on-the-ground realities.
Meanwhile, Brand and Schwittay provided an in-depth look
at the failure of a single project—the LINCOS telecenter
project in Costa Rica [3]—and blamed it on the project’s
techno-centric focus.
But a seemingly greater number of ICT4D papers, short of
reporting failure, describe the challenges they have
encountered during their efforts. This style of report is so
popular that the narrative of ‘challenges being overcome’
seems to characterize much research in the field. For
example, Prasad et al. attributed the mixed success of their
video email technology to “surprising challenges,” most
notably the inability of illiterate users to “understand the
notion of asynchronicity” [18], and suggested that further
research is needed to overcome those challenges. Medhi et
al. examined the “broad challenges” facing their concept of
a job search system for low income domestic workers [14],
including a lack of formal organization and human
capacity. They went on to conclude that “computer-based
systems to solve developing-world problems often require
significant work above and beyond an implementation of
the technology.” Surana et al. described several challenges
in establishing and maintaining a long-distance WiFi
network between eye hospitals in rural India, the first of
which is listed as “limited local expertise” [23]. Kumar et
al. pointed to the “learning curve that is required” as an
explanation for the limited adoption of text-based mobile
applications by low-income users [10]. The challenges of
ICT4D research are the sole focus of an entire article by
Brewer et al. [4], which grouped them into three categories:
technical, cultural, and environmental.
These reports imply that interventionist research and
practice are challenging endeavours. We suggest, however,
that the question of the motivations and incentives of the
recipients of those efforts is often left out of these reports,
or considered only in passing as another challenge to be
overcome. We claim that a stronger focus on motivations
would enrich the field of ICT4D.
Some articles have ventured closer to this discussion by
pondering the seeming mismatch in the definition of
‘needs’ according to development experts on the one hand,
and the recipients of development on the other. In an
extensive review of literature on rural telecenter projects
[11], Kuriyan and Toyama echo the usual refrain in
reporting that telecenters are “difficult to sustain”, but go on
to remark that “what rural villagers want and what we
[ICT4D experts] think they need are frequently different.”
They find that in some kiosks which offer various
development-oriented content (government services,
agriculture, healthcare, etc.), such content accounted for
less than 10% of kiosk usage, with games, digital
photography, and desktop publishing drawing most of the
attention. Bailur’s ethnography of an Indian village radio
station offers a similar story [2]. Initial funding for the
project that created the station carried a strong social
development agenda. That funding was later discontinued
as the high demand for music and quiz shows edged out
development-oriented content.
In this paper, we take this discourse a step further, and
present a study of a successful and innovative technological
system that has emerged in urban India. We offer our
findings as compelling evidence that incentives and
motivations are of paramount importance to technology
adoption. Ironically, our report is cast in the familiar
narrative of challenges being overcome, but with one key
difference: the challenges in this case are overcome not by
researchers or development practitioners, but by the users
themselves.
METHODS
The work reported here comprises a set of informal
preliminary interviews, a set of formal, semi-structured
interviews, and several sessions of participant observation.
Data collection took place over a 9 week period in
Bangalore, India in the summer of 2009.
Bangalore, sometimes called India’s Silicon Valley, is a
thriving cosmopolitan city built on the IT boom in India in
the 1990s. But despite its wealth, Bangalore does not defy
Pareto’s power law of wealth distribution. Among its
residents, for example, are many low-income migrant
workers employed in the city’s service and construction
industries. Mobile phone adoption there is also strikingly
high even among low-income citizens, as in most of India’s
urban centres. An increasing number of phone users now
possess multimedia capable phones, a phenomenon which
gave rise to this study. It is thus an ideal environment for
investigations into emerging trends in technology use.
Preliminary Interviews
The study began with approximately 30 short preliminary
interviews, intended to build an initial understanding of the
practices under study, and to identify potential participants
for more thorough discussions.
Selection criteria. Interviews were conducted with passers-
by on the streets of Bangalore and with support staff
members at Microsoft Research India.
These were conducted mainly on the streets of the
Shivajinagar and Yeshwanthpur neighbourhoods of
Bangalore, both lower-income areas. Participants were
passers-by and proprietors of mobile shops, photo shops,
and Internet cafés. Two interviews were with members of
the support staff at Microsoft Research India. Interviews
typically lasted from 1-10 minutes and were conducted
according to a rough script. Participants were not
compensated. Data from these explorations were not
formally analyzed.
Formal Interviews
Our initial interviews led to 25 more formal, semi-
structured interviews. The guide for these interviews was
designed based on the information we gathered from our
preliminary exercise. Participants were drawn from three
groups: 16 users (those who consumed and traded media
content on their phones), 6 providers (those who earned
money from providing mobile content), and 3
miscellaneous (1 pirated DVD seller, 1 police officer, and 1
man who spuriously claimed to provide ‘anti-piracy
certification’ services for providers). Four participants were
female
2
, 21 were male, and the average age of participants
was 27. Interviews typically lasted about an hour, and were
audio recorded. Participants were offered Rs. 200 (about
$4)
3
as compensation. Both preliminary and formal
interviews were conducted with the aid of a translator when
the participant could not speak fluent English. This was the
case in 19 out of the 25 formal interviews, and a similar
proportion of the preliminary interviews. Alternate
languages used included Kannada, Tamil, and Hindi. All
questions were asked in English and responses were
translated back to English.
Selection criteria. We were interested in users who were
technology non-experts, and who might be typical
participants in a development program—those of lower
socio-economic status, limited household income, and
minimal access to advanced ICTs (beyond mobile phones).
Accordingly, participants in the ‘users’ group were selected
based on the condition that they did not own a computer.
They also had to possess a multimedia-capable mobile
phone. We also deliberately sought a minimum of
representation by women.
Most users were drawn from blue collar sectors, including 4
construction workers, 4 housekeeping staff, 1 driver, and 1
security guard. There were also 2 students, 1 fruit vendor, 1
salesman, and 1 homemaker. One participant was
unemployed. While we did not use a hard income cutoff,
the average monthly income of our participants came out to
Rs. 8,300 ($166), which is approximately equivalent to the
overall average for urban India [20]. Our sample thus
included some participants below the Indian average, as
well as some above it. However, none of our participants
could be considered ‘well-off’. All appeared to have
minimal disposable income, and many were the sole earner
for their household. It should also be noted that ‘average’ in
India is still poor compared to developed regions with
respect to purchasing power parity.
Sampling. Participants from earlier, informal interviews
who met our selection criteria were asked to participate in a
2
The small number of female participants was due to the
difficulty of finding females in our target demographic that
engaged in these practices. Female participation appears
higher among higher-income groups and college students.
3
All monetary figures in this paper are given in Indian
rupees with an approximate conversion to US dollars at Rs.
50 per dollar.
formal interview. These initial participants were then asked
to recommend others who might meet our criteria. This
‘snowballing’ continued until no new information was
being gleaned from interviews. We felt that snowball
sampling was especially appropriate in the case of this
study, due to the questionable legality of the practices being
studied. In many cases, it was clear that our interviews
would not have been granted without a personal
recommendation. However, we acknowledge that this
method may not lead to a representative sample.
Participant Observation
Two sessions of participant observation were carried out at
a Bangalore mobile shop. The first author sat behind the
counter and observed as customers interacted with the
employee in charge of mobile media sales. Each session
lasted about an hour. Brief field notes were recorded, and
subsequently expanded and incorporated into the analysis.
Analysis
Due to the relative novelty of this phenomenon, we avoided
a hypothesis-oriented approach to analysis, and instead
distilled themes of interest via inductive reasoning. This
strategy is characteristic of several well-known techniques
for qualitative analysis [22]. In an open coding exercise, all
transcripts and field notes were read, and interesting
excerpts or quotes were selected and assigned short codes
such as ‘innovative use’ and ‘privacy’. Following this
preliminary scan, the list of codes was reviewed several
times and similar codes consolidated. From this
consolidated list, overarching themes were further distilled,
and excerpts not matching predominant themes were
discarded.
At this point in the analysis, several of the authors noted
that many of the ‘themes’ that were uncovered could be cast
as obstacles or challenges which had been overcome by
users in constructing the sharing system, or motivations
which had driven them to do so. Themes were thus recast
as obstacles vs. motivations. It was further noted that the
only fundamental motivator in this story is entertainment,
as is described in the next section.
FINDINGS
Overview
Finding users with multimedia-capable phones was not
difficult. A vibrant, semi-formal second-hand market
supplies many such phones, often called ‘camera sets’.
Also, cheaper Chinese models, called ‘China sets’, were
available brand new within the same price range. One
mobile shop owner described how the China set was
making more capable phones accessible to a new class of
users:
… if you go around 5 to 6 years back only rich
[people] can have a cell phone with the touch screen
that may cost around Rs. 27000, Rs. 28000, Rs.
30000 [about $560]. But this year and last year the
same version like you can get around Rs. 5000 to
Rs. 6000 [about $110]. That’s a China made
exported from China…. Most of them who works as
a housekeeping, driver, they used to buy the same
type.
The Chinese models offer many of the same features as so-
called ‘branded’ handsets (albeit with dubious quality [17]).
This study focused on two of the most widely used such
features: multimedia viewing and Bluetooth file transfer.
By design, all the phones in our sample could both play and
capture videos, audio, and still images. Each of these
features was used heavily by many of our participants. Most
had a small collection of captured photos or videos, usually
of their family and friends. But non user-generated content
was more common. Music in MP3 format was perhaps the
most popular type of content. Several types of video content
were in evidence, with music videos (or “video songs”)
most popular, followed by short comedy clips or funny
advertisements as typically found on YouTube, and finally
short clips of popular film dialogs. Users often also
possessed several still images. Most popular were pictures
of Hindu gods, film stars, and exotic cars and motorcycles.
It was clear from our interviews that the ability to obtain
and consume such content was one of the main reasons our
participants had purchased more advanced mobile phones.
On being asked why he had gotten such a phone, one
participant answered:
All the effects is good, and the video is good, clarity
is good, GB can be expanded.
The last part of his comment refers to the MicroSD
expansion slot which was available on most of the phones
we encountered. Most such phones had been outfitted with
a 512MB or 1GB memory card, at additional cost to the
owner. This extra space is more or less required for users
interested in keeping multimedia content on their phones,
since internal phone memory usually exceed 64MB on most
models.
The second feature which emerged as highly important was
Bluetooth file transfer. We found that peer-to-peer sharing
of files in this manner is the main method by which users
obtain their content. People share variously with co-
workers, friends, and family. Though it was not a
prerequisite for participation in the study, fully all of our
participants were aware of this practice and regularly
engaged in it, save for one participant whose phone was not
Bluetooth-capable. All indications suggested that Bluetooth
sharing was widespread among users with capable phones.
It was also clear that multimedia content was an important
part of peoples’ daily routine, as opposed to an occasional
novelty. People reported listening to their music daily, and
watching videos once every two or three days. One
participant, a fruit seller, fired up his phone in the evenings
when business got slow:
In the night time the shop will be closed by 11:00
clock. By 9:30 I will start watching the music,
videos, and all because there will be no customers.
Users also refreshed their content through Bluetooth trading
on a regular basis, from once or twice a month to several
times a week:
Once in a week I will be changing the songs or
videos. I will be refreshing the videos because since
the MP3 is around 220 and videos clips around 21
that itself it takes time [to] watch. If I want within a
weeks time [I’ll] take new clip song.
As a further testament to the popularity of these practices,
an informal economy has grown up around mobile phone
multimedia. The most common commercial player in this
market is the ubiquitous “mobile shop”. Such shops are
found with almost as much frequency as basic provision
shops and restaurants in urban India. Their hitherto main
services have been mobile phone sales and service, prepaid
GSM SIM card sales, and sales of “top-ups”, or prepaid
mobile phone credits. In addition, many mobile shops now
also offer “downloading”, which is universally understood
to mean transferring multimedia content to mobile phones.
Prices for content vary, averaging about Rs. 150 ($3) for 1
GB of content, and Rs. 3 ($0.06) per song or video.
Obstacles
As noted, our thematic analysis revealed a diverse set of
obstacles which were surmounted, often in innovative
ways, in producing this system. We review each below.
Cost
Despite the reduced cost of multimedia-capable phones in
today’s market, buying such a phone still represented a
considerable sacrifice for most of our participants. Phones
often cost in excess of a month’s salary. As a result, some
participants were forced to save for long periods. Other
participants took loans to finance the purchase:
Yes it is one month cheque. … I will take the credit
to lead the family. Suppose if I get my salary and
from the salary if I buy the handset, the balance, if
there is a shortage to run [support] the family I will
take the credit from other family and run the family.
Purchase of second-hand phones was also common.
Interface Complexity
Bluetooth transfer is far more complex than making a
phone call or sending an SMS text message, with as many
as 19 steps required to send a file, and as many as 4
decisions points encountered in receiving one. Yet as stated,
all of our 16 participants in the ‘user’ category reported
sharing files via Bluetooth on a regular basis. This was
despite minimal experience otherwise—none of our
participants owned computers, most reported only
occasional access to a computer via a friend or an Internet
café, and few displayed a special penchant for technology.
Most participants had learned how to perform all these
steps from their friends, siblings, children, and co-workers.
Additionally, users had learned even more complicated
procedures for obtaining content from the Internet. One
user who had minimal previous computer experience had
learned from a colleague how to transfer videos from
YouTube to his phone via a complex process:
First I will go to YouTube. When I open the
YouTube, at the same time simultaneously I have to
open one more website .... I will copy and paste [the
URL of the YouTube video] into that zamzar.com
website. Then I will download it and I will put my
email ID and it will direct it to my email ID.
Zamzar.com is a service which converts YouTube videos to
mobile phone video format, and sends an email with a link
from which the converted file can be downloaded. Upon
downloading the file in this manner, the participant would
transfer the file to the phone using a MicroSD card reader,
which typically involves removing the phone’s battery
cover and battery to access the card. In one instance, the
participant reported downloading an entire popular movie,
in four 45-minute segments, one at a time.
Limited Bandwidth and Internet Access
Although Internet access is reasonably priced in India
where it is available, its use is negligible among low-
income groups. Even at mobile shops, the distribution hubs,
Internet access can be scarce. Several shops we talked to
had no connection, and another’s was a split line from a
neighbouring business and was slow and unreliable. The
dominant operating practice for the shops we interviewed
was thus to keep a large repository of content stored on a
local hard drive, and to periodically update the repository
from either the Internet (when access was available), by
copying content from customers, or from a paid supplier.
Suppliers appeared to be the ones with the most Internet
access and know-how. We interviewed one supplier, who
sold bulk collections of content:
I offer 15,000 MP3, 5,000 video songs, in any
converted formats, and applications of any Nokia
mobile, Samsung, any mobile category, applications,
games, plus whatever, themes etc. I will charge
about Rs. 2,500 [$50] for this.
He also provided a regular update service at Rs. 1,500 ($30)
per update, once the base package had been purchased:
Every month, new movies will be released, new
applications will be released, like that. I will
maintain relationships constantly with those
customers, and I will visit their shop [every] two to
three months.
This supplier reported obtaining almost all of his content
from the Internet, especially via BitTorrent Web sites.
Thus, the replication of content from supplier, to mobile
shop, and to hundreds or thousands of users via Bluetooth
sharing, all originates from a single Internet download. The
end result is an extreme savings on aggregate bandwidth
costs, as the cost of one single Internet download is
effectively distributed across so many recipients via
Bluetooth or flash memory cards.
Legality
Unauthorized duplication of copyrighted material is indeed
technically illegal in India. For proof of this, one need look
no further than the pirated DVD sellers which dot the
roadsides of Bangalore and other urban centers, keeping
their wares neatly arranged on top of a plastic tarpaulin,
ready to be bundled up and whisked swiftly down an alley
when the police show up. One DVD seller we interviewed
told of the weekly bribes he paid to the local police:
Starting I used to pay around weekly Rs. 300 [$6],
now I have become close friends to all the police
and all with crime people … Now I pay around Rs.
50 [$1].
It was quite clear that this seller knew what he was doing
was illegal. But curiously, the mobile shop content supplier
we interviewed insisted his business was entirely within the
law:
Q: So then are [DVD pirates] similar to you?
A: Actually no, they are only in that pirating
business.
Q: So how do you compare yourself to them?
A: Actually I am not against the laws.
Indeed, this attitude seems consistent with law enforcement
on the ground. One police officer we spoke to claimed that
the police didn’t bother with mobile shops unless they
received complaints from copyright holders, and to date she
knew of no such complaints. Only pornography, illegal to
sell in India, would draw their attention, and sure enough,
none of the mobile shops we spoke to admitted to selling it.
Still, digital transfers remain technically illegal, and one can
only imagine that as this segment of the entertainment
economy matures, it will come under greater scrutiny.
Time
As heavy users of Napster in its heyday likely know,
obtaining and maintaining a sizeable collection of
multimedia via informal channels can be a time consuming
exercise. Our participants gave evidence of several time
sinks in their practices. First, Bluetooth transfer itself can
be slow, especially when large files are being copied. In
tests we performed using typical phones, a 4.2 MB song
took 2:04 to transfer, and a 6.0 MB music video took 2:49.
Quick arithmetic suggests that transferring a large batch of
content can take upwards of an hour. This is especially
problematic when one works long hours or has a family to
care for—often the case for our participants.
One mobile shop owner described a strategy for
multitasking he had learned of from some construction
workers who frequented his shop. During their day on the
job, working side by side within Bluetooth range
(approximately 10 metres), they would initiate large
transfers and leave them running as they worked, with their
phones in their pockets:
I asked how did you send it via Bluetooth? I sold all
these things [to him]. So I told him like if he is
sending a 115 MB [file] through Bluetooth, how
long he has been spend a time across the phone?…
[He replied:] ‘Ah I have switch on my Bluetooth and
his Bluetooth. We will be working with masonry
works and put it into the pocket. After it gets
transferred we get a message. I will just save it and
keep it; that's a simple sir!’
Construction workers we interviewed confirmed this
practice.
Several other participants described swapping memory
chips as a faster way to transfer larger files, since copying
data from the chip to the phone’s internal memory is faster
than sending it over the air:
We will put the chip and we will copy the contents
to the phone. … In phone memory it will be free; we
will copy to the phone memory. … If it is less, like
one or two songs, we will share it via Bluetooth. ...
If the Bluetooth is around eight songs then it will
take one hour to transfer, and from the chip to phone
memory eight songs will be around five to six
minutes it will take to transfer.
Meanwhile, waiting for content to be downloaded at the
mobile shop can also be time consuming, so some
customers come with a prepared list of songs or videos,
written in the local script. They leave their memory chip
with the list, and come back to collect the full chip
sometime later.
Language
India’s constitution officially recognizes 22 languages [1],
and speakers of many of them are easily found in
Bangalore, whose thriving economy is an attraction for
migrant workers. Most of our participants spoke at least
three languages.
Each language also has its own body of popular culture.
Many participants had content from several different
languages on their phone. When asked to categorize the
content on their phone, participants almost always did so
first by language, rather than by genre or era.
One mobile shop employee told us that his customers
usually wanted content in their native language:
Here we have regional languages. Like as in your
country there is only one language, but here we have
each and every state we each languages. Bangalore
is a place where all sorts of people come here. … So
people come and they have various request, some
people ask for [many different languages].
To service this diverse clientele, the employee had built up
a large repository of content in 13 different languages, often
by copying content from the phones of customers who had
travelled from other states. He also seemed proud of his
expertise in suggesting attractive content to his diverse
clientele:
I need to attract customers. I need to understand
their choice. This person had a request for only 4
songs, … but I showed some [more] songs which
he’s interested in. These people who had migrated
from another place to Bangalore, … we need to
understand their minds.
Viruses
Nearly every participant mentioned fear of viruses as a
drawback of Bluetooth trading. However, it is unclear
whether the problems that participants described were
actually due to ‘viruses’ in the technical sense. System
errors or poor performance were likely more to blame. For
instance:
Once or twice [a virus] has come to my cell, then
immediately I used to switch off the cell, then on it,
then that would have gone. So whenever I listen to
the song and it used to get stuck and not go forward,
then I'll switch off and switch on, then I will listen to
the song.
But regardless of the cause, the fear and the effects are both
real, and the ecosystem has responded. Most mobile shops
offer a ‘virus cleaning service’. Participants we spoke to
reported paying between Rs. 30 ($0.60) and Rs. 100 ($2) to
have a ‘virus’ removed. We note that this does not appear to
be deception on the part of the shop operators or error on
the part of users; only that the term ‘virus’ seems to have
taken on an alternate colloquial meaning in this context.
Other reported strategies for avoiding ‘viruses’ included
refraining from swapping memory chips with other users,
only sharing with trusted phones, keeping the phone’s
Bluetooth function completely switched off except when
transferring files, scanning files on a PC before transferring
files to the phone, and refraining from Bluetooth transfer
entirely.
Institutional Rules
Several participants told of rules at their school or work
which forbade them to carry their mobile phones.
Housekeeping staff at one organization were asked to check
their phones with the security desk in the morning before
beginning work. Therefore, most of their trading took place
after work in the building’s basement, where the workers
changed into their street clothes before leaving the
premises.
Many schools in Bangalore had also banned mobile phones.
One student we spoke to described the measures students
took to hide their clandestine phones from inspections.
The best hiding places are the windows, the college
classroom windows. Just open it out, there's a small
space. We arrange all the mobiles, close the
windows, and sit. So if they're checking the bag, if
they check us also they don't get the mobiles.
Lack of Privacy
Several of our participants expressed concern that family
members would go through the content on their phone. This
usually came up when discussing pornographic videos, also
known as “blue-films”:
Q: So you've had [pornographic videos] but you've
deleted them, is that correct?
A: Yes, because my daughter, everybody used to
see. She is very clever, she will open the phone, oh
ba!
That participant carried two memory chips—one in his
phone, for normal use, and another in his wallet, where he
kept any objectionable content. Another participant had
created a folder with a diversionary name—
‘Presentations’—in which to store his explicit content. To
date, he said, the strategy had worked.
The Motivation: Entertainment
As mentioned earlier, interventionist ICT4D projects
struggle with the kinds of obstacles that this ecology has
overcome. If a single explanation could be offered, we
propose that the motivation of entertainment is far more
powerful than perceived “needs” of low-income
communities. That Indians have embraced multimedia
technology in service of entertainment needs is not
surprising, given the renowned Indian affinity for popular
film and music [12:xiii]. But as the popularity of YouTube
demonstrates, people everywhere, when given a powerful
tool, are as apt to use it for entertainment as for other
“productive” uses.
One participant, when asked why he had bought such a nice
phone despite his modest income, replied:
I just took this because I want some
entertainment…. Since the facilities is available in
this phone so I took it. In the normal phone we
won’t have any timepass but in this phone you will
have timepass.
“Timepass” is Indian slang for leisure activity which passes
the time in a non-productive way. Timepass and
entertainment are clearly related.
One participant singled out construction labourers as
mobile phone entertainment seekers:
From morning till evening they will be working hard
and in the evening for them [they will watch videos
on their phones] just for entertainment.
The overwhelming popularity of entertaining content on the
phones of our participants was also telling. As stated,
music, music videos, film dialogs, and comedy clips were
the norm.
Such content is also cheap. While the cost of handsets may
be relatively high, the cost of content for phones offers a
considerable savings. DVDs cost at least Rs. 50 ($1), and
the average reported cost of downloading a video using the
mobile phone network was almost Rs. 40 ($0.80), an order
of magnitude greater than purchasing the content from a
mobile shop. As a result, very few of our participants
reported ever having obtained, or even knowing how to
obtain, content via mobile Internet (GPRS), or multimedia
message (MMS), both of which deduct from one’s prepaid
balance.
I don't have that kind of technology or knowledge to
use it. If I use GPRS the amount will be cut so I
don't want it to get cut.
While we heard no reports of private individuals paying for
Bluetooth transfers amongst themselves, mobile shops
report that downloading has become an important part of
their business. This quote, from a supplier to several shops,
literally sums it up:
Yeah of course it's profitable. Actually, go for any
mobile shop, ... they will charge you for 1 GB Rs.
150-200 [$3-4].To fill up that 1 GB memory card
will be only 20-30 minutes. In one day they will get
customers like that around 15-20 customers daily.
Two hundred per head means how much?
Meanwhile, several participants stressed how appealing it
was to be able to get their entertainment while on the go.
Bangalore is beset with horrendous traffic problems, so
commutes can be quite long. Long bus rides to native
villages are also common. Prior to the advent of
multimedia-capable phones, most of our participants would
likely not have been able to afford a portable entertainment
device (other than a simple radio) in addition to their
landline or basic mobile. It is only the combination of
media player with phone that makes portable entertainment
affordable.
While the phone provides the convenience of mobility for
some participants, for others the phone is their only source
of multimedia entertainment. These include low-income
workers who don’t have a television, such as one
participant who lived with his mother and sister in a small
apartment:
By seven o'clock I'll finish off my food and we'll sit
round together, three of us, and I will tell my mother
that see Mom, this is the new songs I've got it today,
new video songs, and new MP3 songs.
Migrant labourers who live in makeshift accommodations
on construction sites also fall into this category. One
labourer said of his colleagues:
No they don’t have television—everything they have
in their native place—they are leaving [coming to
Bangalore] only with the mobile phones.
This scarcity brings about unique behaviours. One mobile
shop owner reported regular requests from labourers who
brought movies they had purchased on DVD and asked to
have the entire movie converted and transferred to a
memory chip to be watched on the phone.
Though we previously mentioned lack of privacy as a
barrier to be overcome, we also saw evidence that the
phone can provide privacy. Those same labourers, who live
in a communal situation, benefitted in this way. One said:
Televison like if you watch all of them in bigger
screen it disturbs for every person who is sleeping.
The phone allowed them to access the entertainment of
their choice without disturbing their bunkmates. Another
participant, whose living situation entailed a family of four
living in two small rooms, described taking his phone to a
private spot to watch videos when home life became too
hectic:
There is no place in the house since it is very small
house …. Outside the house there is a stone and I
will watch the video [there].
Frequent power cuts are another unfortunate feature of
Bangalore. Several participants reported turning to the
phone for entertainment when the power went out.
In summary, one can see why the desire to be entertained is
expressed so strongly through the multimedia phone.
Phones can provide entertainment anywhere, anytime, in
public or in private, irrespective of unreliable infrastructure,
and at little or no recurring cost. Stepping even further
back, one also appreciates just how strong the desire to be
entertained really is, having motivated the construction of
such an elaborate ecosystem in the midst of considerable
scarcity.
DISCUSSION
At this point, we return to the topics for discussion we
raised at the outset of this paper.
First is the notion, suggested by our findings, that
motivation appears to trump social, technical, and
environmental obstacles to adoption of a new technology –
particularly relevant to HCI is that complex UIs can be
mastered by undereducated users. Several previous studies
conclude that some systems are infeasible due to UI barriers
lesser than that for Bluetooth file transfer [10,18]. While
direct comparisons between projects in different contexts
should be made only with caution, it is easy to imagine
similar studies condemning Bluetooth file transfer as an
unsuitable technology due to interface complexity. Yet,
video exchange thrives in spite of the UI complexity.
The relative informality of the sharing system should also
be noted here. It is not as though the basic motivation of
entertainment was coaxed and cajoled toward Bluetooth
media sharing by powerful promotional forces. Quite the
opposite is true. During our 9 weeks in the field, we saw no
mention of Bluetooth media sharing whatsoever from any
official source, despite the constant barrage of
advertisements and promotional offers from mobile service
providers and other sources. While this is perhaps not
surprising, since nobody but small-scale mobile shops stand
to profit from the practice, it is further evidence that the
entertainment motivation alone was responsible for this
phenomenon.
Our second, related, point is that developing regions are
capable of embracing and building on sophisticated use of
ICTs. Perhaps the best evidence of this is the grassroots
innovation we observed within the system. Examples
include opportunistic use of hours spent working side-by-
side on construction jobs to complete large transfers;
swapping of memory chips and use of phone memory as a
temporary buffer to avoid long transfer times; and the
transfer of entire Bollywood films to mobile phones to be
watched in cramped sleeping quarters. Moreover, these
innovations were due to some of our lowest-status
participants: construction labourers and housekeeping staff.
Our study is also evidence that this adoption can occur on a
large scale. Although we do not yet have large-scale
quantitative data, anecdotal evidence from three other
metropolitan areas in India point to Bluetooth sharing being
a widespread phenomenon, spanning several social classes
and demographics. Though experience with Bluetooth
sharing was not a prerequisite for participation in our study,
fully all of our participants whose phones were capable of it
had engaged in the practice to some degree. Aside from the
findings reported here, we also draw on our collective
experiences in other parts of India, including urban Mumbai
and Delhi, and peri-urban areas of Tamil Nadu, in making
this claim.
It should also be noted that this system is not necessarily an
instance of 'leapfrogging', in which Indian users have
somehow skipped ahead of those in the developed
economies by watching and sharing videos on their mobile
phones. To the contrary, the majority of mobile media in
the developed world seems destined to be downloaded over
unlimited wireless data connections. Instead, the system we
have observed represents a unique response to the
socioeconomic realities of urban India.
The third point for discussion calls for a reflection on the
concept of 'needs' in the context of development. An
interesting thought experiment here would be to imagine
what kinds of need-oriented uses might have been
envisioned for this technology by development experts.
Language learning classes? Health information videos?
Community media initiatives? We can say with confidence
that each of these programs would indeed be beneficial to
many of our participants. But we ask: Would any of these
applications, noble as they are, have brought about the
system we have described? In the event that they didn't,
would the familiar cast of barriers have been used to
rationalize the disappointment?
This is certainly not to say that development experts are
never aware that the services they are offer may not be
considered highly desirable by the intended recipients. In
other words, we acknowledge that people from all walks of
life sometimes prioritize short-term gratification (‘desires’)
over long-term prudence (‘needs’). On the other hand, we
suggest that the perception of this reality as something to be
rectified itself carries certain value judgments which are not
necessarily universal. But the thorny philosophical
dichotomies of needs vs. desires, utility vs. freedom, and so
on, go far beyond the scope of this paper. Our point is
merely that the complexity of technology should not be
automatically blamed when adoption happens to fail.
A constructive corollary to this discussion is the notion that
development practitioners and researchers could consider
piggybacking on powerful motivations like entertainment in
service of their goals. This has been attempted before with
traditional forms of entertainment, such as theatre [15]. We
suggest that the mobile phone, and more specifically the
highly developed infrastructure we have described, offer the
potential for an electronic equivalent.
ACKNOWLEDGEMENTS
We thank Joy Joseph, Meera Laxman, and Sunandan
Chakraborty for assistance in carrying out this research, and
Bill Thies for his guidance and comments on earlier drafts.
We also thank the three anonymous reviewers for their
helpful feedback.
REFERENCES
1. Constitution of India. 2003.
2. Bailur, S. The complexities of community participation
in ICT for development projects: The case of "Our
Voices". Proceedings of the International Conference
on Social Implications of Computers in Developing
Countries, (2007).
3. Brand, P. and Schwittay, A. The Missing Piece:
Human-Driven Design and Research in ICT and
Development. Proceedings of the international
conference on information and communication
technologies and development, (2006), 2-10.
4. Brewer, E., Demmer, M., Ho, M., et al. The challenges
of technology research for developing regions.
Pervasive Computing 5, 2 (2006), 15-23.
5. Cheverst, K., Dix, A., Fitton, D., et al. Exploring
bluetooth based mobile phone interaction with the
hermes photo display. Proceedings of the international
conference on Human computer interaction with
mobile devices and services, (2005), 47-54.
6. Cui, Y., Chipchase, J., and Jung, Y. Personal TV: A
Qualitative Study of Mobile TV Users. Lecture Notes
in Computer Science 4471, (2007), 195.
7. Harper, R., Regan, T., Rouncefield, M., Rubens, S.,
and Al Mosawi, K. Trafficking: Design for the Viral
Exchange of Digital Content on Mobile Phones.
Proceedings of the international conference on Human
computer interaction with mobile devices and services,
(2007).
8. Heeks, R. Most egovernment-for-development projects
fail: how can risks be reduced? Institute for
Development Policy and Management, 2003.
9. Kindberg, T., Spasojevic, M., Fleck, R., and Sellen, A.
The ubiquitous camera: An in-depth study of camera
phone use. Pervasive Computing, (2005), 42-50.
10. Kumar, A., Rajput, N., Chakraborty, D., Agarwal,
S.K., and Nanavati, A.A. VOISERV: Creation and
Delivery of Converged Services through Voice for
Emerging Economies. Proceedings of the IEEE
International Symposium on a World of Wireless,
Mobile and Multimedia Networks, (2007), 1-8.
11. Kuriyan, R. and Toyama, K. Review of Research on
Rural PC Kiosks. 2007.
http://research.microsoft.com/en-
us/um/india/projects/ruralkiosks/Kiosks%20Research.d
oc.
12. Manuel, P. Cassette Culture: Popular Music and
Technology in North India. University Of Chicago
Press, 1993.
13. Maunder, A., Marsden, G., and Harper, R. Creating
and sharing multi-media packages using large situated
public displays and mobile phones. Proceedings of the
international conference on Human computer
interaction with mobile devices and services, (2007),
222-225.
14. Medhi, I., Menon, G., and Toyama, K. Challenges in
computerized job search for the developing world. CHI
'08 extended abstracts on Human factors in computing
systems, (2008), 2079― 2094.
15. Mlama, P.M. Culture and development: The popular
theatre approach in Africa. Nordiska Afrikainstitutet,
Uppsala, 1991.
16. O'Hara, K., Mitchell, A.S., and Vorbau, A. Consuming
video on mobile devices. Proceedings of the
conference on Human factors in computing systems,
(2007), 857-866.
17. Pan, T. Market review: Shanzhai phones in India.
2009. http://www.shanzai.com/index.php?
option=com_content&view=article&id=121.
18. Prasad, A., Medhi, I., Toyama, K., and Balakrishnan,
R. Exploring the feasibility of video mail for illiterate
users. Proceedings of the working conference on
Advanced visual interfaces, (2008), 103-110.
19. Sey, A. Public Access to ICTs: A Review of the
Literature. Center for Information & Society,
University of Washington, 2008.
20. Shukla, R. How India Earns, Spends, and Saves. The
Max New York Life Insurance Limited, 2007.
21. Södergård, C., ed. Mobile television-technology and
user experiences. In Report on the Mobile-TV project.
VTT Publications, Espoo, Finland, 2003.
22. Strauss, A.C. and Corbin, J. Basics of Qualitative
Research. Sage Publications, 1998.
23. Surana, S., Patra, R., Nedevschi, S., and Brewer, E.
Deploying a Rural Wireless Telemedicine System:
Experiences in Sustainability. Computer 41, 6 (2008),
48-56.
24. Toyama, K. and Keniston, K. Telecentre Debates.
Telecentre Magazine 1, 2008.
http://www.telecentremagazine.net/articles/article-
details.asp?articleid=72.
