THE RUSI JOURNAL

kILLER DRONES
THE mORAL UPS AND DOWNS
DAvID WHETHAm

The use of drones has increased exponentially in recent years, causing a mounting wave
of concern amongst the media and public about the implications of using unmanned
systems often misunderstood in their nature above all in terms of accountability,
legitimacy and fairness. David Whetham explores the many facets of this question,
delving into the often-overlooked nuances of the use of remote-controlled systems and
its practical as well as moral implications.

rones are now ubiquitous in

the battle space. The US Army
alone has more than 7,500
unmanned aerial systems up from
seventy-six in 2002 at a total cost of
$5.23 billion.1 Although the numbers
from the UK perspective are not quite
as extraordinary, the British military
still deploys a range of unmanned aerial
surveillance platforms ranging from the
Black Hornet mini-helicopter providing
enhanced situational awareness for
troops on the ground through to the
Reaper (a medium-altitude, longendurance remotely piloted air system)
at the other end of the spectrum.2
First deployed by the RAF in
Afghanistan in October 2007, the
capabilities provided by Reapers in
terms of surveillance, reconnaissance
and, since May 2008, firepower (a
combination of laser-guided bombs and
Hellfire air-to-ground missiles) have
been used since then in conjunction
with other platforms and capabilities.
The Reapers are launched from airfields
in Afghanistan by crews deployed in
theatre, but flown using secure satellite
communications by pilots sitting in
Creech Air Force Base in the US, and now
also from RAF Waddington in the UK.3
As former Assistant Chief of the Air Staff
Baz North has explained, Our experience
of operating RPAS has confirmed that

they have unique capabilities that

complement those of traditional combat
and ISR [intelligence, surveillance and
reconnaissance] platforms; maximum
effect is achieved by employing them
in a mixed grouping.4 It is no surprise,
therefore, that the RAFs current modest
contingent of five Reapers is set to double
in the near future.5 This technology is
certainly not restricted to just the US
and its allies. Indeed, what is very clear
is that drone numbers around the world
are increasing, and it is difficult not to
conclude that they will continue to be an
integral part of future military operations.
The debate as to whether this
increasing reliance on drones is a
positive or a negative development
is extremely polarised. As with most
complex situations, however, there is
a great deal of confusion as to what
exactly is being argued about, making it
difficult to reach a considered view. For
example, while some appear to be very
critical of the very existence of drones, it
is possible that what they actually object
to is the use of drones by the US to carry
out targeted killings on the territory of
other states.6 This is, in fact, not a dronespecific concern, as such actions could be
carried out through less remote means,
such as helicopter gunships or even
special forces on the ground. As such,
the ethical issues surrounding the use of

RUSI JOURNAL JUNE/JULY 2013 VOL. 158 NO. 3 pp. 2232

RUSI 158_3 TEXT.indd 22

targeted killing deserve to be discussed

more broadly and at greater length than
can be accommodated here.7 Others
object to the concept of killer machines,
a perceived unfairness in the imbalance
of risk inherent to the use of drones,
or what they see as the indiscriminate
nature of drones due to the excessive
civilian casualties associated with them.
This article will discuss such issues in
order to offer a balanced appraisal of
some of the more profound ethical
implications of recent developments in
drone technology.
The commonly used term drone
can, of course, cover a huge array
of unmanned aerial technologies,
ranging from micro air vehicles (MAVs)
measured in inches through to the huge
Global Hawk surveillance aircraft with
a wingspan of over 130 feet (almost 40
metres). Drones come in many guises
and can be used for a huge array of tasks,
and are considered particularly attractive
for carrying out what are known as dull,
dirty or dangerous jobs. For example,
different platforms can be used around
the clock to monitor conservation and
anti-poaching campaigns, detect forest
fires using sophisticated heat-detection
equipment, or conduct search-and-rescue
missions in treacherous conditions using
advanced all-weather optical sensors and
infrared capabilities. These alternative
DOI: 10.1080/03071847.2013.807582

14/06/2013 10:30:15

Protesters burn an effigy of a US military drone during a rally outside the US Embassy in Manila, Philippines, April 2012. Courtesy of AP Photo/Bullit Marquez.

uses notwithstanding, this article

focuses specifically on the ethical issues
surrounding the use of military drones to
kill people during armed conflicts.

Autonomous Killing

Responding to public perceptions

and, indeed, fears that drones are
cold, calculating, emotionless machines
dispatched to eliminate any and all
identified threats, the UK Ministry of
Defence (MoD) has chosen to re-brand
its Unmanned Aerial Vehicles (UAVs)
with the much less threatening term
Remotely Piloted Air System (RPAS).
The latter is far more accurate given its
emphasis on the person in control of
the system, rectifying the notion that
the platform operates entirely without
human input. While drone technology is
sophisticated enough to permit a wide
range of automated functions, from takeoff and landing through to loitering over
a particular geographical area for many
hours at a time, there is still a person
supervising its activity. As such, the

aircraft does not make life and death

decisions autonomously. According to
Noel Sharkey, a professor of artificial
intelligence and robotics, despite the
many and various developments in this
field in the past twenty years, no system
yet exists that can discriminate between
a combatant and a non-combatant in a
combat zone.8 As such, the UK position
remains very clear: A fully autonomous
system would have to be capable of
making the qualitative assessments
currently required by the Law of Armed
Conflict, until this is possible, the human
must remain within the decision-making
process.9
Clearly, autonomous killing is not,
as yet, an issue in practice. However, it
is worth noting Sharkeys concern that
research into the ability of artificialintelligence systems to select targets,
for example, is being enthusiastically
pursued in many parts of the world,
even if the systems themselves are not
currently being equipped with weaponry.
It is not inconceivable, therefore, that if,

or more likely when, the military situation

or political landscape changes at some
point in the future, politicians seduced
by promises of casualty-free conflicts
are very likely to give the order to arm
such systems.10 In this vein, Human
Rights Watch called for a pre-emptive
ban on all fully autonomous weapons in
November 2012 and the organisers of the
Campaign to Stop Killer Robots a global
effort launched in April 2013 argue
that advances in robotic technology
mean it is only a matter of time before
fully autonomous, human-out-of-theloop systems capable of identifying
and firing on targets on their own are
developed.11 This will indeed raise some
profound ethical challenges, not least
over the question of responsibility and
accountability should the systems prove
inaccurate in their targeting.

Killing at a Distance

Of
course,
stand-off
weapons
themselves are nothing new and there is
a spectrum in terms of such capabilities:

23
RUSI 158_3 TEXT.indd 23

14/06/2013 10:30:17

kILLER DRONES

a rock thrown a few yards might be

at one end, while a Tomahawk cruise
missile might be at the other. A soldier
with a sniper rifle, able to watch a
target through a scope from a distance
of several kilometres before pulling the
trigger, is also clearly on the stand-off
spectrum, as is a fast-jet pilot delivering
close air support to those on the ground.
From this perspective, therefore, the
drones degree of remoteness is much
greater when considering that the
operator may sit many thousands of miles
away from where the drone is actually
flying its mission. However, while the
degree of stand-off is obviously greater,
this does not necessarily imply that there
is any moral difference between the
soldier looking down the rifle scope, the
pilot using an aircrafts sensor display to
target a laser-guided bomb and the drone
operator looking at targets on a screen.
One of the obvious physical
differences is that however safe the
sniper is, or however high above the
front line the fast-jet pilot is, both are
still in, over or at least near the theatre
of operations, sharing certain risks with
those on the ground, including the
targets. These are risks that the drone
operator is simply not concerned with.
Yet the question remains as to why this
should matter at all. If the effect of the
combatants action is going to be the
same whether they are pulling the trigger,
flying a plane or piloting a drone, it would
appear perverse to insist on putting ones
own personnel in harms way, and to
claim it as somehow ethically preferable,
when the same effect could be achieved
without doing so. Bradley J Strawser, an
assistant professor in the defence analysis
department at the US Naval Postgraduate
School and a research associate with
Oxford Universitys Institute for Ethics,
Law and Armed Conflict, suggests there
might actually be a moral imperative at
work that compels the West to pursue
and deploy this type of technological
development to minimise unnecessary
risk to personnel.12 Indeed, it is wrong
to command someone to take on
unnecessary potentially lethal risks in an
effort to carry out a just action for some
good.13 As a programme manager at
Honeywell puts it: every time a T-Hawk
goes down it means a human didnt.14

Surely, this should be a good thing, based

on a clear common-sense, if not ethical,
requirement to reduce risks to ones own
personnel where it is possible to do so.
This appears to be a strong argument
in favour of using more drones rather
than fewer. However, there are also other
concerns related to this distance between
the pilot and the target primarily that
of moral disconnection. Some military
training is designed to allow individuals
to achieve emotional distance from
their enemies and thus enable them to
overcome an innate reluctance to kill.15
One of the tools traditionally employed
in this regard is dehumanisation the
promotion of a sense of otherness in
the group that is deemed to be a threat.
Those who fall into such a group can then
be perceived as non-entities, expendable
or undeserving, making the act of killing
them easier to carry out.16 Given Prince
Harrys recent comparison of his work as
a co-pilot gunner in an Apache gunship
to a video game,17 it is easy to see how
some people would wonder about the
moral effect of killing when the targets
are just pixels on a screen and there is
no need to look anyone in the eye a
clear, perhaps even extreme, example of
dehumanisation.
It also raises the question as to how
much easier this would be if the person
pulling the trigger were 8,000 km away
to begin with. The English philosopher A C
Grayling notes that an RAF bomber crew
in the Second World War could unleash
its bombs from 20,000 feet and knowingly
kill hundreds or even thousands of
women and children. If, however, one
gave the same bomber crew a knife and
told them to slit the throats of the family
in the room next door, they would not be
able to do it.18 Indeed, physical separation
appears to ease the suspension of moral
concerns, even to the extent of making
terrible things possible. Interviews
with contemporary military pilots with
combat experience show that they tend
to agree that not only are decisions
to kill [from the air] rarely perceived
as emotionally charged, the death of
friendly, yet physically distant combatants
is emotionally dulled.19
It would appear, then, that such a
lack of physical contact associated with
remote killing contributes greatly to the

alienation of each side from the other.

Whether this remoteness is achieved
through the planting of a roadside
improvised explosive device (IED) for
which people bid money via the Internet
to win the right to detonate the bomb
and watch the results live on a website
or a drone operator wasting tiny avatars
on a computer screen, it is thus difficult
to ensure the essential mutual respect
that combatant equality the bedrock
of the law of armed conflict requires. If
a faceless enemy can be so debased, it is
easy to see how lines can become blurred
and civilians can come to be seen as just
another target.
However, physical distance does
not always guarantee a corresponding
emotional separation. An American
Second World War veteran recalled
the moment at which he opened his
aircrafts bomb-bay doors on his first
mission over Europe: He felt terrible
resistance, nausea, sickness, headaches,
despair. He couldnt do it, but his crew
chief screamed at him, Now! Now! If he
didnt, the mission would be a failure and
it would be his fault. He finally pushed
the button. Then he vomited.20 Of course,
this refers to an act that was part of a
total war in which large-scale, devastating
attacks on civilian populations were
commonplace, rather than the type
of highly accurate, precision-targeting
policies of wars of choice in the
contemporary age, in which the pilot
can even be on a different continent.
Nevertheless, the moral comfort that
distance can provide might not extend
as far as many would think. Just because
the target is viewed through a screen
rather than a rifle or bombsight does not
mean that taking life has no effect on
the person pulling the trigger. Perhaps
surprisingly, just as fast-jet pilots can
suffer from post-traumatic stress disorder
(PTSD), there are also cases among drone
pilots who have never personally even
been in (or over) the theatre of military
operations.21
One wonders if operating a drone for
offensive missions in the contemporary
operating environment might actually
be closer to the experience of military
snipers.22 Snipers, too, are separated
by distance, but can also be intimately
aware of their target, much like the drone

RUSI JOURNAL JUNE/JULY 2013

RUSI 158_3 TEXT.indd 24

14/06/2013 10:30:17

DAvID WHETHAm

operator who might be tracking a target

for hours using real-time video feeds
before the decision is taken to strike. A
sniper deployed in Iraq recalls:23
Theoretically, sniping was supposed to
be a matter of clinical, dispassionate
killing. Even when we were in Iraq,
killing Iraqis, it was target one, target
two. Target ones on the left; target
twos on the right. OK, scan target one.
Target ones down. Scan target two. Fire.
Target twos down. Thats it. Theyre just
targets; you try to convince yourself of
that. [However,] imagining a man purely
as a target was not easy when you had
to aim specifically at him and fire and
then watch him fall over, screaming and
arching his back in agony.

Drone footage can be just as graphic.

This raises questions about whether
the geographic dislocation is actually
providing moral dislocation at all, or at
least any dislocation greater than that
experienced by a sniper.24 An RAF RPAS
operator noted that: 25
[A] day, two days, maybe a month can
be spent watching and then when it
happens, we dont leave once weve
dropped the weapons, we stick around
for battle damage assessment and
because of the way of the Muslim faith
they like to bury their dead before
the sun goes down generally you will
see the funeral procession, the women
and children arrive, it can be very
emotional and thats one of the things
theyre looking for to make sure that
that doesnt turn into a problem in ten,
fifteen years.

Such considerations feed into a wider

debate regarding the long-term
emotional effects of combat on drone
operators and in particular whether
the geographical and emotional distance
entailed will prove a positive or negative
factor. Dave Grossman, in his powerful
book On Killing: The Psychological Cost
of Learning to Kill in War and Society,
notes that those who do not dehumanise
their opponents are most likely to be
able to deal with the aftermath of war
and go on to lead happy and productive
lives.26 If it is true that the innate (or

at least the socially programmed)

inhibitions against killing do not need to
be overcome in the same way for drone
operators as they do, for example, for
infantry who need to be prepared to
close with and then kill the enemy, then
the psychological repercussions of war
might be significantly reduced. It will be
interesting to see if incidences of PTSD
among drone operators in the longer
term turn out to be significantly different
both quantitatively and qualitatively
to those among combatants who are
physically present in (or over) the battle
space.
Putting aside this question
of
emotional
and
psychological
consequences associated with remote
combat, it remains the case that the
absence from the physical situation
has some obvious implications for the
effectiveness of the drone operator. The
fact that the operator is not directly at
risk when on a mission provides him or
her with some advantages over a person
physically in harms way. Another drone
operator refers to this as the ability
to step back and have a bit more of
a Hamlet moment as it were you
can hopefully double-check what
youre doing is correct.27 This
detachment provides time and space for
decision-making that is, quite literally,
a world away from the experiences of
soldiers on the ground or even of pilots
in hostile airspace. The ability to remain
cool, calm and detached arguably
allows better decisions to be taken in the
heat of battle.
There are undoubtedly issues of
situational awareness linked to the
physical limitations of the information
feeds, such as the need to ensure
sufficient bandwidth for receiving
all of the required sensor data, and
potential time delays (of about one to
two seconds) caused by the distance
the signals need to travel, but these are
all technical issues that are constantly
being minimised or ameliorated. This
author has been told by current RPAS
pilots who have also flown combat
missions in fast jets that, in practice,
problems
surrounding
situational
awareness generally have more to do
with the quality and chosen field of view
of the sensors being employed rather

than whether those sensors are on a

manned jet or a drone. Just as manned
aircraft can increase their situational
awareness by calling on visual back-up
from other aircraft or other assets in
theatre, a drone operator can also refer
to images provided by other drones and
ISR assets to build up a more detailed
and accurate picture of the situation. Of
course, this must be balanced against the
possible negative effects of drone pilots
switching between different platforms
or coming on shift halfway through an
operation, and therefore not having the
continuity of focus required to build up
an understanding of the story playing
out on the ground. However, such issues
should be avoidable and one might
expect that as technology continues to
improve, the quality of the decisions
being made should also therefore
increase.
There may be other advantages to
being physically remote: a corresponding
emotional distance between the operator
and events on the ground may well
have a positive influence on the
behaviour of the former towards the
target. As Strawser notes, Once fear
for their own safety is not a pressing
concern, one would assume the operator
would be more capable, not less, of
behaving justly.28 By contrast, a 2006
report by the US militarys Mental Health
Advisory Team (MHAT) avers a strong
correlation between anger among armed
forces personnel and the mistreatment
of non-combatants. It also suggests that
soldiers and marines who were members
of units that had suffered casualties
were more likely to treat civilians in
negative ways.29 Furthermore, Paolo
Tripodi argues that the ability to remain
slightly detached from ones immediate
situation is one of the attributes of
a good commander.30 As such, the
enforced detachment of the drone
operator might well be a positive
thing.
The distance between the drone
and the operator also raises the question
of whether the operator needs to be a
combatant at all or whether civilians
could undertake such roles. In doing so,
however, they would relinquish their
immunity from attack, meaning that
they may be directly targeted and, if

25
RUSI 158_3 TEXT.indd 25

14/06/2013 10:30:17

kILLER DRONES

captured, they may not be recognised

as combatants or receive the protection
afforded by prisoner-of-war status.31
If the civilian operator is not based in
the country in which the actual military
operation is taking place, this might seem
to be a small or acceptable risk to take.
However, as Wing Commander Allison
Mardell notes in her contribution to
an RAF Directorate of Defence Studies
report on UAVs, allowing civilians to
operate drones during armed conflict
may have significant implications for
them if their activity amounts to a
direct part in hostilities.32 As such,
attention should be paid to the legal
questions surrounding the employment
of civilian drone operators; indeed,
it must be remembered that civilians
employing lethal force that is not in
direct self-defence may be charged with
murder.
Continuing in this vein, however,
the removal of physical risk raises the
question of whether drone operators
are really at war at all and whether
even those in uniform are actually
fellow combatants. In turn, this highlights
some interesting dilemmas over the
nature of direct and indirect participation
in conflict that the geographic separation
from the battlefield might appear to
cloud a little. If a man knowingly helps
his brother to plant an IED by driving
him and the device in his car, he might
be a civilian, but he is facilitating
the planting of the device and can
therefore be considered to be directly
participating in hostilities at that time.
As such, he may be legitimately targeted
while carrying out this role. Consider,
then, a husband on the morning school
run who drops his wife at work at an
air base in Nevada where she will be
piloting a drone employed in offensive
operations. The question must then be
raised whether he, too, becomes a direct
participant in hostilities while acting as
chauffeur.
This has implications for the way
in which the contemporary operating
environment is conceptualised and
the changing moral landscape this
generates. Some of these challenges
were acknowledged by the same RPAS
operator cited above who noted that
his hour-long commute was important

for adjusting from home life to combat

mentality:33
[I]t gives you time to adjust you do
have to be in the right frame of mind
to do this. You are carrying munitions
and weapons that can end lives so its
not the kind of thing where you can
enter the box worried about the next
mortgage payment or worried about
how the kids are going to get to school
tomorrow.

Clearly, expecting people to go into work

as if it were simply a day in the office,
be directly involved in events that kill
people, and then go home again at the
end of the shift to play games with the
children may be introducing a cognitive
dissonance that is difficult to maintain
without significant, and perhaps longlasting, effects. Such issues will require
in-depth consideration as drones
become increasingly prevalent in defence
doctrine.

Accountability

By highlighting the potentially huge

distances between what is happening
on the ground and the person operating
it, the question of accountability comes
into sharp relief. The person targeted by
a drone will probably remain completely
unaware of the weapon system and,
in turn, of its operator until his or her
world, quite literally, comes crashing
down. Similarly, the marine unit on the
ground has no idea which individual
is controlling the drone in the skies
above when it calls in fire support. A
key consideration, therefore, is what
happens if something goes wrong, and
how individual responsibility can be
determined.
While it is, of course, always a
question of how the technology is
employed in practice, formally, there
are strict guidelines in place regarding
decisions to deploy munitions, as in the
current military theatre of Afghanistan,
for example. Unless acting in self-defence
(in which case no further authority is
required), any use of lethal force by ISAF
members has to be duly authorised by
a correctly empowered engagement
authority. In practice, this is normally
the commander on the ground who,

having gone through and satisfied both

the target-eligibility procedures and
collateral-damage assessments, passes
instructions through a qualified joint
terminal attack controller (JTAC) to the
operator of the platform required to
deliver the munition, be that an artillery
battery or a drone.34 Such command-andcontrol procedures lead some to argue
that the ground commander owns the
bomb: they make the decision and so
should also ultimately be accountable
for the effects of that decision. This
would appear to remove responsibility
from the drone pilot. However, there is
also the clear requirement that everyone
involved in prosecuting a target, from
planning through to pulling the trigger,
has the obligation to comply with the
Law of Armed Conflict and to reject
any order that is deemed unlawful. A
weapon-release authorisation does not
turn anyone into a moral automaton and
the drone operator can still decide not to
prosecute a target. As such, responsibility
is spread across everyone involved.35
Culpability, in turn, is closely linked to the
honesty rather than just the accuracy
of the situational assessment made
by the person carrying out the action or
giving the order. Therefore, anyone in a
position to influence a targeting decision
who fails to pass on pertinent information
bears the moral responsibility for any
resultant mistake be they on the ground
with eyes on, someone at the back of
the Operations Room with a view of a
different camera feed, or the drone pilot
him- or herself. While this attribution
of responsibility and culpability is not
unique to the use of drones, it does
emphasise the importance of trust in
military operations between the different
parties involved in decision-making.
Another consideration with regards
to accountability is that the very nature
of this new technology actually offers a
much higher degree of oversight than in
virtually any other area of military activity.
Many of this authors conversations with
military personnel recently returned
from operations in Afghanistan highlight
the complete lack of effective scrutiny
in theatre; no journalists are foolhardy
enough to be embedded on long-range
patrols so there is no media presence and
such teams can be out of contact with

RUSI JOURNAL JUNE/JULY 2013

RUSI 158_3 TEXT.indd 26

14/06/2013 10:30:18

DAvID WHETHAm

their own bases for extended periods. It

is easy to see how what happens out on
the open plains of Afghanistan might well
remain known only by those involved as
a result.36 However, the drone operators
experience could not be further removed
from this: every movement of the
joystick, every frame of camera footage
and therefore every decision or,
indeed, hesitation or omission made
by the operator is recorded and can be
pored over at great length following
any incident.37 There is nowhere to
hide from a bad decision if the military
wishes to use such information to hold
the drone operator to account (and
recent experience suggests that if they
do not, there is at least a chance that the
information may well end up in the public
domain through Wikileaks or some other
forum, enforcing accountability anyway).
Such scrutiny would appear to
place profound restraint on any triggerhappy behaviour; however, it goes
further than this. The 2006 MHAT report
mentioned above recounted that 45
per cent of US soldiers and 60 per cent
of marines surveyed stated that they
would not report a fellow unit member
for killing an innocent non-combatant.
These figures rose to 57 and 70 per cent
respectively for those claiming they
would not report a fellow unit member
for unnecessarily destroying private
property.38 (It should be noted, however,
that this research was undertaken in
very particular circumstances in Iraq and
that a great deal of work has since been
done to correct such attitudes.) Adequate
oversight allied with an attendant fear
of being caught is one of the ways
that such attitudes and behaviours can
be adapted.39 While the prospect of
using drones to spy on both friendly
and hostile forces is not necessarily
palatable, the fact that the whereabouts
of such assets is generally unknown,
combined with their ability to observe
events on the ground in intimate detail,
means that there is at least a chance of
misdemeanours being captured on film
or even being watched live. Knowing
or even thinking that there is a small
chance of someone watching remotely
might help improve both behaviour and
accountability among those other than
the drone operator as well.40

Precision (and its Paradox)

It is a necessary, if obvious, caveat that

just because technology is capable of
being used in a particular way, it does
not mean that it actually will be. In this
sense, drones are exactly the same as any
other military technological development
throughout history: each is potentially
open to abuse or can be put to use in
pursuit of illegitimate ends. However,
in terms of precision, many of the
advantages of current drone technology,
when used appropriately, appear obvious.
Rather than saturating a target with
multiple sorties and ordinance in order
to have confidence in its destruction, the
same military outcome can be achieved
with fewer and smaller weapons
due to their improved accuracy. The
resultant reduction in foreseeable but
unintended and unwanted civilian death
and destruction collateral damage is
therefore significant.41 As such, given the
removal of risk to drone pilots, discussed
above, it is clear that Such weapons can
therefore help to preserve life from both
perspectives (and are more financially
efficient into the bargain).42
Of course, sometimes things do
go wrong. For example, discrimination
decisions
and
proportionality
assessments for actions taken with any
weapon system might not be conducted
with the diligence required, while
any decision is only as good as the
information or assumptions that inform
it. Indeed, history shows that mistakes in
war do happen: consider the destruction
of the Chinese Embassy in Belgrade in
1999 or the bombing of a wedding party
in Wech Baghtu, Afghanistan, in 2008
both of which were entirely accidental.43
One report reviewing the use of drones in
Pakistans remote tribal areas concluded
that 32 per cent of those killed in drone
attacks since 2004 were civilians.44 While
the numbers are debated (for example,
the Brookings Institution puts the number
of civilians killed at between one in seven
and one in ten45), any civilian deaths
caused by such accidents are obviously
highly regrettable and the fact that any
occur at all leads some to the idea that
they were intentional. They argue that if
the weapons are as accurate as is widely
asserted then, surely, whatever those
weapons hit must be the intended target.

This highlights the paradox of precision:

that is, the more accurate the weapons
employed, the more attention upon any
misses or mistakes.46 The importance of
managing expectations surrounding the
Wests ability to carry out attacks with
unfailing pinpoint accuracy is essential,
as is ensuring that policy-makers
understand that the potential accuracy of
the technology does not mean that there
will not be civilian casualties. Warfare
has never been clinical in this sense and
is unlikely to suddenly become so given
the persistence of the fog of war and
friction encompassing those factors
that cannot be controlled and that
therefore have the potential to derail
plans. The situation is exacerbated when
an opponent deliberately seeks to draw
the foul, as it is known in basketball, by
positioning themselves in such a way that
any military response will either violate,
or be perceived as violating, the rules in
bello by creating substantial numbers
of civilian casualties.47 The decision to
abort an attack with Storm Shadow
cruise missiles during Operation Unified
Protector in Libya in 2011 rather than risk
killing civilians allegedly being used as
human shields illustrates this dilemma
only too well.48 Of course, as with so
many other issues, this is not a problem
unique to drones, but rather an issue for
all military activity in the contemporary
operating environment.

Asymmetric Implications and

Perceptions of Legitimacy

The potential mismatch between the

technological capabilities of opponents
may also create a particular ethical
challenge.
The
asymmetric-threat
concept normally focuses attention on
those hostile agents who seek to turn
the tables on the sophisticated military
machines they oppose by playing by
different rules because they cannot
hope to match them technologically.
Their tactics might include hit-and-runstyle guerrilla attacks that prevent the
preponderance of military power being
brought to bear through to the use of
IEDs against dismounted troops or their
vehicles. It is important to recognise,
however, that the deployment of
drones reverses this asymmetry: indeed,
any military activity carried out by a

27
RUSI 158_3 TEXT.indd 27

14/06/2013 10:30:18

kILLER DRONES

technologically sophisticated protagonist

against a less sophisticated one is
inherently asymmetrical.
This is, in fact, merely the
recognition that one normally seeks
to exploit an opponents weakness
rather than attacking their strength.
Admittedly, however, conducting a war
from thousands of metres in the air
without the pilot even being in the same
hemisphere as the weapon might be seen
to be taking asymmetry to the extreme.
Peter Singer of the Brookings Institution
notes Arthur C Clarkes observation that
[a]ny sufficiently advanced technology is
indistinguishable from magic.49
The work of the German political
scientist Herfried Mnkler might be
invoked in response to the question
whether it is even ethical to use such
magic against those who cannot employ
it themselves. He suggests that: The
pilot of a fighter-bomber or the crew
of a man-of-war from which Tomahawk
rockets are launched are beyond the
reach of the enemys weapons. War
has lost all features of the classical duel
situation here and has approached, to
put it cynically, certain forms of pest
control.50 This conceptualisation of
modern warfare evocatively captured
here and even more pertinent to drones
becomes of real concern when it is
considered in relation to the principles
of jus in bello. As Singer notes, war can
become like playing God from afar, just
with unmanned weapons substituted
for thunderbolts.51 Disturbing though
these images might be, the question
is, however, fundamentally misguided
if it is motivated by a concern to create
a level playing field: indeed, the
principle of proportionality is not about
being fair, it is about not using more
force than is necessary to achieve the
required ends.52 As such, it would be
morally perverse to refrain from using
any weapon unavailable to the opponent
thereby eschewing precision munitions
and causing additional and unnecessary
death and destruction, or putting
armed forces personnel in harms way
unnecessarily out of some misguided
notion of fair play.
Nonetheless,
domestic
and
international public perceptions of
morality will continue to shape the way

that drones contribute to present and

future military operations:53
Although the decisions of democratically
elected governments can claim a natural
legitimacy, they prefer to be working
with public opinion and are apt to
become nervous if they have failed
to persuade international opinion.
The legitimacy of a military operation
is a subjective attribute, relating to
questions of legality and morality as well
as security. Because it is subjective this
is an area where the inability to develop
persuasive narratives about the whys
and wherefores of controversial policy
will make itself felt. An operations
legitimacy will be hard to obtain and
sustain if it is not in accord with the
prevailing political culture, and in the
West, that means with liberal values.

The discussion above has drawn out

some of the ways in which drones can be
used in accordance with such values. It is
interesting that for all of the arguments
made against the use of drones by
the US along and across the Pakistani
border with Afghanistan, there was very
little angst over the use of drones in
conjunction with other tools to provide
protection for the civilian population
in Libya in 2011. The general public
may have been unaware of or simply
uninterested in the presence of drones
there, but perhaps the lack of concern
had more to do with a general sense that,
at least when the civilian population was
in direct need of protection from its own
government, this was a worthy cause
and whatever appropriate tools were
available should be employed. The fact
that no Western personnel lost their
lives during Operation Unified Protector
was an added bonus. However, the next
time that a pilot participating in a just war
is lost in action while carrying out a task
that could have been undertaken just as
effectively by a remotely piloted drone
may well prompt a more fundamental
shift in public perceptions of what is
acceptable or legitimate.

The Threshold to the Use of

Military Force

Precision-strike from stand-off weapon

systems such as drones offer policy-

makers an option for direct action that

would simply not be possible using
more conventional tools in the military
toolbox. Experience suggests that
public support for military action will
fall away as the number of casualties
both in terms of national military
personnel and civilians mounts.
If such numbers can be minimised
through the use of precision weapons,
delivered safely from afar, the argument
follows that public resolve can
therefore be maintained an essential
consideration when democracies want
to use military force, particularly where
national interests are not obviously at
stake. In 1999, the dark observation
was made that NATO was willing to
bomb to stop a massacre, but only
as long as pilots were not put at risk
by flying below 15,000 feet: the life
of one NATO soldier is worth 20,000
Kosovars.54 This classic argument for
the utility of, and preference for, air
power has been taken to the next
level by removing the risk of harm to
the pilot as well, adding resonance to
Michael Ignatieffs prescient question:
are we on the eve of a new age of
drive-by wars, in which American
power can strike anywhere, in near
certainty that neither its civilians nor
its soldiers will ever be put at risk?55 It
must also be considered whether such
capabilities increase the likelihood of
war, potentially transforming it from an
option of last resort into a first or at
least an early response to certain types
of crisis instead.
Of course, in some instances
this lowering of the costs of action
may well be considered a good thing:
perhaps, for example, the West could
be willing to prevent a future genocide
such as in Rwanda in 1994 using such
technology.56 Generally, however, the
worry may be that any development
that makes the decision to go to war
easier must be profoundly worrying,
even if it is difficult to see how this
objection can be tested empirically.
If the argument that drones make
the decision to use force easier than
it would otherwise be appears to be
intuitively plausible, Strawser counters
that it still ultimately fails as a valid
objection to their use because there is

RUSI JOURNAL JUNE/JULY 2013

RUSI 158_3 TEXT.indd 28

14/06/2013 10:30:18

DAvID WHETHAm

still a moral imperative to use drones

or other stand-off weapons when they
are available and equally effective,
presumably even if this means that
they may be used more frequently.57 It
is indeed wrong to command someone
to take on unnecessary potentially
lethal risks in an effort to carry out a
just action for some good.58 However,
a policy preference for the limitation
or, indeed, the elimination of risk
in combat might ultimately result in
a strategic failure, suggesting that
taking some risk might be necessary to
achieve ones political aims. Perhaps,
for example, those on the losing side
might not accept defeat if they feel
as though they have never even met
their opponent in battle. In this vein,
it is possible to draw a lesson from
the experience of early months of
the most recent conflict in Iraq: the
Sunni Triangle, the future hotbed
of rebellion[,] wasnt occupied until
weeks after Baghdad fell in 2003, and
local would-be insurgents instead got
the signal that they had never been
defeated.59
Similarly,
new
generations
of stand-off weapons seem to
demonstrate an ability to kill but little
or no willingness to die for the Wests
causes.60 Technological advantages
might actually signal to opponents
a fundamental lack of resolve to see
an issue through to the end. Michael
Ignatieff argues that the willingness
to take mortal risk is what makes
military
deterrence
believable.61
The use of drones over Beirut, for
example, had the result of spurring
mass identity politics as an antidote
to the technology discrepancy, and
gave the impression that they [Israelis
and American] dont want to fight us
like real men, but are afraid to fight.
So we just have to kill a few of their
soldiers to defeat them.62 Furthermore,
it should be noted that it was not the
bombing from a safe distance of 15,000
feet that led to success in Kosovo, but
the eventual threat of substantial
numbers of boots on the ground
combined with real political pressure
from the international community
a demonstration of both credible
means and genuine resolve. If the use

of drones does indeed make it easier

to resort to war because of the
apparent low cost of action, the result
may actually be a crisis that cannot be
finished without paying heavier costs.

Unintended Consequences?

Of course, this risk-free approach to

warfare might also have implications
for ones own civilian population,
which might be seen as a legitimate
target as a result. If it is impossible to
strike a states military assets due to
its predominant reliance on drones
or other military tools that leave its
armed forces all but invulnerable,
those wishing to inflict damage on
the state may well broaden their list
to encompass civilian targets, should
the stakes be high enough. This calls
into question whether the breaking of
international law becomes justifiable
in the event of a supreme emergency,
in which an effective response in
alignment with the principles of jus in
bello is simply impossible.
The term supreme emergency
was coined by Winston Churchill to
depict the terrible situation in which
Britain found itself in early in 1941,
faced with imminent invasion and with
only RAF Bomber Command available
as an effective offensive weapon. There
was little doubt regarding the terrible
cost of defeat to Germany and the
existential crisis was seen to justify
using those bombers against the only
target that could be struck given the
limits of technology at the time, which
included not only military facilities
but also the civilian inhabitants
of German cities.63 This supreme
emergency passed as other theatres
of war opened up, new allies joined
the struggle, technology improved,
allowing more accurate targeting, and,
most importantly, the threat of the
imminent invasion of Britain abated. As
the emergency passed, the continued
bombing of German civilians became
increasingly difficult to justify on
moral grounds.64 The justification of
targeting civilians in this way is clearly
difficult, but it is easy to see how such
arguments might be convincing to those
who feel powerless to respond to the
overwhelming technological superiority

of a faceless enemy whom it appears

otherwise impossible to counter.

Conclusion

In many ways, drones can be seen as

the continued evolution of manned
airpower in the twenty-first century
and, as such, the type of ethical issues
associated with them are often similar
to those debated regarding their
manned counterparts, differing only in
degree rather than posing radically new
questions. As with any other military
technology, it is open to abuse and
misuse, but it can also be employed
responsibly within established legal
and ethical frameworks. Drones are
currently deployed in combination
with, rather than as a straightforward
replacement for, many other types of
military asset in theatre. It would be
churlish not to acknowledge that, when
used appropriately, drone technology
offers a whole range of military
advantages and that many of these
are to be genuinely welcomed from an
ethical perspective.
However, there are a number of
challenges that potentially come to a
head if drones are transformed from
being one of many tools into the tool of
military and political choice. The ability
to conduct stand-off operations in a way
that minimises risks to non-combatants
and eliminates them entirely for national
military personnel might well lower the
political threshold to employing military
force, making the occurrence of war
more frequent. If this means a greater
willingness to intervene and use force
to protect those that cannot protect
themselves then it may well be a good
thing, but it might also make it easier
to become involved in situations that
prudence would otherwise rule out.
At the same time, it might also make
those conflicts more difficult to resolve
due to a lack of will either perceived
or real to put ones own people in
harms way when required. Meanwhile,
involvement in other peoples affairs
while relying overwhelmingly on military
tools that reduce or eliminate risk to
ones own combatants, ironically, might
also increase the risk to ones own civilian
population. Some may see robots as an
answer to the suicide bomber,65 but they

29
RUSI 158_3 TEXT.indd 29

14/06/2013 10:30:18

kILLER DRONES

should also consider that some suicide

bombers may themselves be motivated
by a sense of helplessness in the face of
that remote approach to war in the first
place. Therefore, even if in some contexts
they can indeed offer some real practical
and ethical advantages over other
military tools, drones may not always be
the right means to match the political
ends.

David Whetham is a Senior Lecturer in

the Defence Studies Department of
Kings College London. His publications
include Ethics, Law and Military
Operations (Palgrave, 2010) and Just
Wars and Moral Victories: Surprise,
Deception and the Normative
Framework of European War in the
Later Middle Ages (Brill, 2009).

The article presented here is an updated

and substantially revised version of David
Whetham, Remote Killing and Drive-By
Wars, in David W Lovell and Igor
Primoratz (eds), Protecting Civilians During
Armed Conflict: Theoretical and Practical
Issues During Violent Conflict (Farnham:
Ashgate, 2012). Any opinions expressed
here should not be taken to represent any
official institution or position.

15 See Dave Grossman, On Killing: The

Psychological Cost of Learning to Kill
in War and Society (Boston, MA: Little
Brown, 1995).

Notes
1

Geoff Fein and Grace Jean, Budget

Stalemate: US Military Seeks Improved
UAS Capabilities Despite Financial
Squeeze, IHS Janes International
Defence Review, April 2013.

Nick Hopkins, British Military has 500

Drones, Guardian, 6 May 2013.

BBC News, Armed Drones Operated

from RAF Base in UK, Says MoD,
27 April 2013.

Assistant Chief of the Air Staff, RAFs

Reaper logs 10,000 Hours over
Afghanistan, 19 May 2010. <www.
gov.uk/government/news/rafs-reaperlogs-10-000-hours-over-afghanistan>,
accessed 9 May 2013.

Royal Air Force, Reaper, <http://www.

raf.mod.uk/equipment/reaper.cfm>,
accessed 9 May 2013.

CNN, Anti-Drone Protests Take Off in

Britain, 28 April 2013.

For an exploration of some of the moral

issues specifically relating to this, see
David Whetham, Drones and Targeted
Killing: Angels or Assassins?, in Bradley
Jay Strawser (ed.), Killing by Remote
Control: The Ethics of an Unmanned
Military (Oxford: Oxford University Press,
2013). See also Michael L Gross, Moral
Dilemmas of Modern War: Torture,
Assassination, and Blackmail in an Age
of Asymmetric Conflict (Cambridge:
Cambridge University Press, 2010),
particularly Chapter 5.

Noel Sharkey, The Life Scientific,

BBC Radio 4, 29 January 2013. For a
more optimistic view regarding future
developments, see Ronald C Arkin, The
Case for Ethical Autonomy in Unmanned
Systems, Journal of Military Ethics (Vol.
9, No. 4, 2010), pp. 33241.
Allison Mardell, Unmanned Aerial
Vehicles: The Legal Perspective, in Owen
Barnes (ed.), Air Power: UAVs: The Wider
Context (Northolt: Ministry of Defence,
2009), p. 82.

10 Sharkey, The Life Scientific. For

discussions on some of the ethical
implications of fully autonomous
weapon systems, see Special
Issue: Ethics and Emerging Military
Technologies, Journal of Military Ethics
(Vol. 9, No. 4, 2010).
11 See Human Rights Watch and
International Human Rights Clinic, Losing
Humanity: The Case against Killer Robots
(New York, NY: Human Rights Watch,
2012), <www.hrw.org/sites/default/
files/reports/arms1112ForUpload_0_0.
pdf>, accessed 9 May 2013. See also
Stuart Hughes, Campaigners Call for
International Ban on Killer Robots,
BBC News, 23 April 2013.
12 Bradley Jay Strawser, Moral Predators:
The Duty to Employ Uninhabited Aerial
Vehicles, Journal of Military Ethics (Vol.
9, No. 4, 2010), pp. 34268.
13 Ibid., p. 344.
14 Andy Simms, A Magnificent Man and
His Flying Machines, Soldier Magazine
(October 2010), p. 37.

16 Susan Opotow, Moral Exclusion and

Injustice: An Introduction, Journal of
Social Issues (Vol. 46. No. 1, Spring
1990); see also Edward Tick, War and
the Soul: Healing Our Nations Veterans
from Post-Traumatic Stress Disorder
(Wheaton, IL: Quest Books, 2005), p. 82.
17 Jon Boone, Taliban Retaliate after Prince
Harry Compares Fighting to a Video
Game, Guardian, 22 January 2013.
18 For a fascinating and balanced discussion
of these issues, see A C Grayling, Among
the Dead Cities: Is the Targeting of
Civilians in War Ever Justified? (London:
Bloomsbury, 2007).
19 Nick Tucker-Lowe, Does the Advent of
Uninhabited Systems Fundamentally
Affect the Ethical Landscape of
Contemporary Conflict?, Defence
Research Paper (unpublished Defence
Studies MA dissertation, Kings College,
London), July 2010, p. 11.
20 Tick, War and the Soul, p. 91.
21 For example, see Jane Mayer, The Risks
of a Remote-Controlled War, National
Public Radio, 21 October 2009. This
is also discussed in P W Singer, Wired
for War: The Robotics Revolution and
Conflict in the 21st Century (New York,
NY: Penguin, 2009).

RUSI JOURNAL JUNE/JULY 2013

RUSI 158_3 TEXT.indd 30

14/06/2013 10:30:18

DAvID WHETHAm

22 Dave Cummins, an RAF Reaper operator

based at Creech Air Force Base in
Nevada, made precisely this observation
on BBC Radio 4, World at One, 4
January 2013.

35 This debate is currently taking place

at the highest levels of ISAF military
leadership. The UK tends to support the
latter view of spreading responsibility
across all involved.

23 Jeff Sparrow, Killing: Misadventures in

Violence (Carlton, Victoria: Melbourne
University Press, 2009).

36 For a disturbing account of what can

happen when effective oversight is
lacking, see Mark Boal, The Kill Team,
Rolling Stone, 27 March 2011, <http://
www.rollingstone.com/kill-team>,
accessed 12 May 2013.

24 This emphasises the idea introduced

above that, in many ways, remote killing
is in reality a spectrum rather than sharp
differentiation. There is a connection
between the Lancaster bomber crews in
the Second World War and the English
and Welsh archers at Agincourt in 1415,
for example.
25 Cummins, on BBC Radio 4, World at
One.
26 Grossman, On Killing.
27 Individual referred to only as Fire Control
Officer, BBC Radio 4, World at One.
28 Strawser, Moral Predators, p. 353.
29 Mental Health Advisory Team (MHAT)
IV, Operation Iraqi Freedom 05-07: Final
Report, 17 November 2006, pp. 3839,
<http://www.armymedicine.army.
mil/reports/mhat/mhat_iv/MHAT_IV_
Report_17NOV06.pdf>, accessed 9 May
2013.
30 Paolo Tripodi, Understanding Atrocities:
What Commanders Can Do to Prevent
Them, in David Whetham (ed.),
Ethics, Law and Military Operations
(Basingstoke: Palgrave, 2010),
pp. 18586.
31 Mardell, Unmanned Aerial Vehicles,
p. 79.
32 Ibid.
33 Cummins, on BBC Radio 4, World at
One.
34 This is also known as the Forward Air
Controller in the UK but the US term
JTAC (Joint Tactical Air Controller or Joint
Terminal Attack Controller) has become
commonly used for the suitably qualified
ground-based person who carries out
this role.

37 In this sense, the drone pilots

experience is very much like that of the
fast-jet pilots.
38 MHAT IV, Operation Iraqi Freedom 0507, p. 37.
39 See Tripodi, Understanding Atrocities,
pp. 17388, and Peter Wall, The Ethical
and Legal Challenges of Operational
Command, in Whetham (ed.), Ethics,
Law and Military Operations, pp. 221
29.
40 While a non-military issue, one can see
such thinking at play in the March 2013
suggestion by the League Against Cruel
Sports that they may begin using drones
to monitor (and presumably deter) any
potentially illegal hunting activities in the
UK. See BBC News, Drones May be Used
to Target Illegal Hunting, 16 March
2013.
41 Singer notes that during the Second
World War, it took an average of 108
planes to successfully prosecute each
target. In Afghanistan sixty years later,
each aircraft could be expected to
successfully engage more than four
different targets per sortie. See Singer,
Wired for War, p. 100. While accurate,
the Hellfire missile one of the
weapons used by the Reaper may not
be the most discriminate of weapons
simply due to the size of its warhead
(originally designed for targeting tanks
and bunkers). Just as with any decision
to deploy force in a combat zone, this
means that appropriate precautions
have to be taken before the use of these
weapons can be authorised, but it is
also the case that smaller and smaller
warheads will likely be utilised in the
future as certainty regarding accuracy
improves.

42 David Whetham, Ethics, Law and

Conflict, in Whetham (ed.), Ethics, Law
and Military Operations, p. 20.
43 BBC News, Embassy Strike a Mistake,
8 May 1999; Reuters, Karzai Says Air
Strike Kills 40 in Afghanistan,
5 November 2008.
44 Dean Nelson, One in Three Killed by US
Drones in Pakistan is a Civilian, Report
Claims, Daily Telegraph, 4 March 2010.
45 Douglas Murray, Drones Save Lives,
Wall Street Journal, 7 March 2013.
46 See Whetham, Ethics, Law and Conflict,
p. 21.
47 See Michael Skerker, Just War Criteria
and the New Face of War: Human
Shields, Manufactured Martyrs, and
Little Boys with Stones, Journal of
Military Ethics (Vol. 3, No. 1, 2004),
p. 28. That lessons are being learnt might
be indicated by the apparent reduction
in civilian deaths associated with
drones. For example, see Ken Dilanian,
CIA Drones May Be Avoiding Pakistan
Civilians, Los Angeles Times, 22 February
2011.
48 Ian Dury, Mission Aborted on Orders of
SAS: RAF Attack is Halted after Troops
Spot Human Shields, MailOnline, 12
May 2011, <http://www.dailymail.co.uk/
news/article-1368626/Libya-RAF-abortattack-SAS-spot-Gaddafi-using-humanshields.html>, accessed 12 May 2013.
49 P W Singer, The Ethics of Killer
Applications: Why is it So Hard to Talk
About Morality when it Comes to New
Military Technology?, Journal of Military
Ethics (Vol. 9, No. 4, 2010), p. 310.
50 Herfried Mnkler, Die neuen Kriege
(Hamburg: Rowohlt, 2003), quoted and
translated in Uwe Steinhoff, Torture:
The Case for Dirty Harry and against
Alan Dershowitz, Journal of Applied
Philosophy (Vol. 23, No. 3, August 2006),
pp. 33753.
51 Singer, Wired for War, p. 324.
52 See David Whetham, The Just War
Tradition: A Pragmatic Compromise, in
Whetham (ed.), Ethics, Law and Military
Operations, p. 21.

31
RUSI 158_3 TEXT.indd 31

14/06/2013 10:30:19

kILLER DRONES

53 Lawrence Freedman, The Transformation

of Strategic Affairs (London: IISS Adelphi
Paper Series, 2006), p. 80.

57 Strawser, Moral Predators, p. 358.

54 Singer, Wired for War, p. 324.

59 Singer, Wired for War, p. 308.

55 Michael Ignatieff, To Fight but Not to

Die, World Today, February 2000, p. 21.

60 Whetham, Ethics, Law and Conflict,

p. 22.

56 See Zach Beauchamp and Julian

Savulescu, Robot Guardians:
Teleoperated Combat Vehicles in
Humanitarian Military Intervention,
Strawser, Killing by Remote Control,
pp. 10625.

61 Ignatieff, To Fight but Not to Die, p. 23.

63 See Michael Walzer, Just and Unjust

Wars: A Moral Argument with Historical
Illustrations (New York, NY: Basic Books,
1977), pp. 25168.

58 Ibid., p. 344.

64 See Grayling, Among the Dead Cities.

65 US Navy researcher Bart Everett, quoted
in Singer, Wired for War, p. 62.

62 Rami Khouri, an Arab moderate, quoted

in Singer, Wired for War, pp. 30809.

Need to become more strategic in your thinking?

Want the benets of up-skilling?
MSc Full-time and part-time study in Defence Acquisition Management
If you are involved in any aspect of
defence acquisition then this CIPS,
CILT and CMI accredited course is
relevant to you. Whether you are
involved in establishing or managing
capability requirements; engaged
in procurement, commercial or
contracting activities; or developing

and implementing sustainable

support solutions, this course will
provide the underpinning tools and
resources to develop the critical
analytical and thinking skills required
for the eective management of
defence acquisition.

Tuition fees: 15,950

For further details, please contact
Dr Julieanna Powell-Turner
Course Director
T +44 (0) 1793 785755
E: j.powell-turner@craneld.ac.uk
www.defenceacquisition.co.uk
Crown Copyright images from www.defenceimages.co.uk
Cranfield University 116DS1213

RUSI JOURNAL JUNE/JULY 2013

RUSI 158_3 TEXT.indd 32

14/06/2013 10:30:19