A summary of the Chatham House International Law discussion group meeting held
on 1 November 2007.
The meeting was chaired by Elizabeth Wilmshurst. Participants included legal
practitioners, academics, NGOs, and government representatives.
Speaker: Michael N. Schmitt, Stockton Professor of International Law, United
States Naval War College
The event was sponsored by Cambridge University Press and marked the publication
of Perspectives on the ICRC Study on Customary International Humanitarian Law.
International humanitarian law (IHL) has long differentiated between the treatment of
combatants and civilians, affording certain protections to the latter. Customary and
treaty international law prohibit attacks against civilians, require that attacks which
might cause incidental loss of civilian life be proportional, and mandate the taking of
all feasible precautions to minimize such loss. Indeed, under the Rome Statue of the
International Criminal Court, it is a war crime to attack civilians.
Such special protection, however, is suspended if a civilian directly participates in the
hostilities. Professor Schmitt focused on three issues bearing on interpretation of this
customary rule. Codified in Article 51.3 of Additional Protocol I and Article 13.3 of
Additional Protocol II to the 1949 Geneva Conventions, the law allows, in particular,
civilians to be attacked “for such time as they take a direct part in hostilities.” 
Professor Schmitt is a member of an expert working group on the ‘direct participation
of civilians in hostilities’ (DPH). Convened in 2003 by the International Committee of
the Red Cross (ICRC) and the Asser Institute, the groups work will result, after a final
meeting in Geneva, in the publication of interpretative guidelines. However,
Professor Schmitt presented his own views on the subject, many of which have been
influenced by the working group process.
Legal clarification on civilian participation in hostilities has become necessary in light
of recent shifts in the nature of armed conflict. Traditional State-on-State conflicts
are increasingly characterized by asymmetrical warfare. In such hostilities, a
common tactic of the weaker party’s forces is withdrawal into populated areas so as
to deprive their stronger adversaries of the full advantage of superior firepower and
technology. This urbanisation of warfare expands the opportunities and incentives
for involvement of the civilian population. The shift of theatre has been accompanied
by the civilianization of many functions previously performed by military personnel.
Catering, logistics and security, for example, are commonly contracted out to private
companies such as Blackwater USA. Additionally, conflict is often characterized by
combat with non-traditional armed forces. For instance, in Iraq and Afghanistan,
American and British forces face a diverse and non-uniformed enemy which includes
remnants of the Iraqi military, insurgents, sectarian militia and transnational terrorists.
Question 1: Who is a civilian?
As only civilians enjoy the protections cited above, a preliminary issue is the
definition of the term “civilian”. For combatants (i.e. non-civilian), there is no need to
question the nature and currency of participation in the hostilities. They constitute
legitimate targets in respect of whom the obligation to take precautions to minimise
harm to the civilian population and the principle of proportionality are inapplicable.
Thus, the question of direct participation only arises once it is determined that an
individual involved in the conflict qualifies as a civilian. Professor Schmitt
emphasised that the civilian/combatant dichotomy in this context did not employ the
same test as the definition of combatants for the purposes of prisoner of war status
under Article 4 of Geneva Convention III.
With regard to international armed conflict (conflict between States), members of the
armed forces obviously qualify as combatants. For direct participation purposes, so
too do militia, volunteer corps and organized resistance groups that belong to a party
to the conflict (i.e. that are commanded and organized). Similarly, participants in a
levée en masse lose their civilian status and become legitimate targets. On the other
hand, autonomous groups, criminals, private contractors and government employees
(unless tasked by a party with combat functions) are civilians, and only lose civilian
protection for so long as they directly participate in the hostilities.
Categorization differs in non-international armed conflict (conflict within a State). The
State’s armed forces and rebel military forces (i.e. military forces that have gone into
rebellion) are combatants. This category includes non-military government forces
directly involved in the conflict (e.g. the police, intelligence services and the interior
ministry). On the other hand, whilst there is agreement that autonomous or
spontaneous fighters are civilians (who may be attacked if they directly participate),
disagreement exists as to the classification of other organised and commanded
Professor Schmitt advocated an approach which classes members of such forces as
combatants (in the DPH context) if the group in question was formed for the express
purpose of fighting (Option A). This test logically places rebel forces on a par with
government forces, who are automatically combatants by virtue of membership.
Under Option A, a member of an organised group which has both political and
fighting wings, such as Hamas or Hezbollah, is a combatant only if enlisted in the
latter. Professor Schmitt emphasised that, contrary to common misperceptions,
identification as a “fighter” is often possible; some wear uniforms, and intelligence
(e.g., phone intercepts) can reliably indicate membership in the fighting wing, etc.
Should doubt arise, the IHL presumption of civilian status would apply, such that the
individual could only be attacked if directly participating.
The alternative approach focuses on the function performed by individuals in
organised and commanded groups (Option B). By this approach, those who have
the specific task of fighting (or related functions such as military planning) qualify as
combatants (for DPH purposes); there is no need to perform a DPH analysis before
targeting them. On the other hand, if a function does not qualify because it merely
supports the group (e.g., cooking), it would be unlawful to target individuals
performing them unless they are directly participating. Indicia for this test might
include carrying weaponry, wearing a uniform and location.
Question 2: What is direct participation?
Because civilian protections under international humanitarian law are suspended
when a civilian directly participates in hostilities, the enquiry into the nature of direct
participation lies at the heart of the DPH analysis. Certain specified acts unarguably
do or do not constitute direct participation. Attacking the enemy, capturing
equipment, laying mines, sabotage and tactical intelligence collection on the
battlefield fall into the former category. Protection under international humanitarian
law, however, is not lost by civilians who care for troops or work in a factory (even
munitions factories). In a residual limbo lie many examples on which agreement is
more elusive. Examples include the civilian driver of an ammunition truck (a
quandary posed by Major General APV Rogers in Law of the Battlefield), political
leaders, aircraft maintainers, intelligence personnel, drone operators and computer
In order to resolve consistently the status of disputed functions, it is necessary to
adopt a test of general application. Various criteria have been proposed: proximity to
the battlefront, contribution to the fight, hostile intent, and extent of military command
and control. The ICRC official Commentary to Additional Protocol I states that direct
participation requires a direct causal relationship between activity and harm done and
that it includes acts, the nature and purpose of which are intended to cause actual
harm to the enemy.
In a Chicago Journal of International Law article, Professor Schmitt had earlier
proposed a test that looked at the criticality of the act to the direct application of
violence against the enemy. An act is direct participation if the civilian can foresee
that his action will harm or disadvantage the enemy in a relatively direct and
immediate way, although it is not necessary to foresee the actual harm inflicted.
Professor Schmitt demonstrated the application of his proposal to civilians involved in
intelligence and aircraft maintenance. He considered that those involved in the
development of strategic intelligence retain civilian protection, while individuals
performing operational or tactical level intelligence functions are direct participants.
Similarly, whilst flightline aircraft maintenance personnel are legitimate targets,
depot-level workers are properly classed as non-participating civilians.
Elements of this approach appear in the emerging consensus regarding the nature of
direct participation, although some critics have expressed concern about the difficulty
of ascertaining the intentions of civilian actors. That consensus centres on three
cumulative criteria. The first requires that the actions in question harm (or be likely to
harm) the enemy. Either enemy military operations must be affected (e.g. disrupting
logistics, gathering intelligence, clearing enemy mines and computer network
attacks) or civilians and civilian objects not under control of the attacker have to be
harmed. An unsettled issue in this regard is whether voluntarily shielding lawful
military objectives would qualify. Professor Schmitt’s own view is that although they
neither pose a risk to the attacker nor physically impede military operations, voluntary
human shields do inflict harm in the sense that they can absolutely prevent an attack
as a matter of law (through operation of the proportionality principle).
The second criterion is a direct link between the act and the harm to the enemy (or
civilian objects/civilians). In other words, is the act in question an integral part of a
concrete military operation, rather than one which merely supports or sustain the war
effort (such as revenue collection)?
Finally, an act of direct participation has to be designed to affect negatively the
enemy such that there is a belligerent nexus. As an example, even if harm inflicted
in a criminal bank robbery meets the first two tests, it does not equate to direct
Professor Schmitt applied these criteria to the case of defensive actions by civilians.
Civilians defending themselves against unlawful (even military) actions are not direct
participants. On the other hand, those who intentionally guard military objectives
against enemy action qualify (although individuals involved in such activities might
constitute combatants, thereby negating the need to do a DPH analysis). More
problematic is the case of civilians taking defensive actions when simply present at
military objective under lawful attack. Professor Schmitt opined that such actions
would be direct participation.
He then discussed the issue of how to resolve doubt as to whether an act amounts to
direct participation. Doubt regarding civilian or combatant status must, pursuant to
an accepted IHL principle, be resolved in favour of finding civilian status. In contrast,
the benefit of the doubt falls the other way when it is unclear whether there is direct
participation. He justified this approach on the bases that it provides clearer lines for
combatants (whose lives would be at risk if ambiguous conduct was interpreted in
favour of civilian status) and discourages the involvement of civilians in armed
conflict. Thus, it is an appropriate balance between military necessity and
humanitarian considerations (a balance that underlies most IHL norms).
The standard of proof for the test of direct participation is that of the reasonable
combatant in the circumstances. Professor Schmitt emphasised that it would not be
appropriate to transpose court room standards of proof, such as the criminal
standard of proof beyond reasonable doubt, to the decisions of combatants in the life
and death environment of the battlefield.
Question 3: What does “for such time” mean?
The final legal issue is the duration of participation, specifically, the meaning of “for
such time”. The approach espoused by many, and reflected in the ICRC
Commentary to Additional Protocol I, views immediate preparations for and return
from combat as participation; prior and later periods do not qualify. The
unsatisfactory consequence of this stance is that it creates a “revolving door,” by
which a civilian might, for instance, farm by day (and be protected from attack) but be
a guerrilla by night (i.e. a legitimate target). Professor Schmitt favoured an
alternative approach that regards participation as continuous until the civilian opts
out, either for an extended period or by an affirmative act. In addition to jamming the
revolving door, this approach places the risk of mistake in classification on the direct
participant (who after all chose to participate), rather than the attacker.
Professor Schmitt emphasized that the opting out requirement did not violate the IHL
presumption of civilian status. Opting out relates to the temporal component of
participation, not to the status of the individual. Indeed, the individual must be a
civilian before the issue of DPH even arises (see above).
Liability of Directly Participating Civilians
Professor Schmitt concluded by noting that direct participation is in and of itself not a
violation of IHL. However, the underlying conduct which constitutes the direct
participation (such as attacking civilians) might amount to either a war crime or a
violation of the domestic law of States with subject matter and personal jurisdiction.
The potential criminal liability of Blackwater security personnel involved in the
September 2007 killing of Iraqi civilians serves as an interesting case study of
domestic criminal jurisdiction.
With regard to prosecution in the Iraqi judicial system, Coalition Provisional Authority
(CPA) Order 17 gave US contractors immunity from the Iraqi legal process for “acts
performed by them pursuant to the terms and conditions of a contract or sub-contract
thereto”. When the CPA was dissolved in June 2004, Order 17 remained part of the
law of Iraq. Although the Iraqi government is moving towards rescission of the Order,
that had (at the time of his presentation) not yet been done.
Numerous US statutes extend US jurisdiction abroad for actions during hostilities.
The US War Crimes Act of 1996 provides for the domestic prosecution of grave
breaches of Geneva Conventions I-IV or certain specified violations of Common
Article 3 when committed by or against US nationals or service members. The
Federal Anti-Torture Statute of 1994 allows prosecution of US nationals or anyone
present in US who commits or attempts torture. No prosecutions of contractors have
taken place under either Act.
US Special Maritime and Territorial Jurisdiction (SMTJ) legislation specifies thirty
crimes, including murder, that are prosecutable if committed by or against US
citizens on overseas military bases or facilities. In 2007 David Passaro, a CIA
contractor, was sentenced to over eight years in prison for assaulting a detainee in
Afghanistan. However, because the September killings occurred outside US bases,
the SMTJ cannot be used to prosecute Blackwater staff.
The Military Extraterritorial Jurisdiction Act (MEJA) provides for the prosecution of
“persons employed by or accompanying the armed forces” who have committed
felonies. To date the only successful conviction under MEJA has been of a
contractor for possession of child pornography. Professor Schmitt noted that
prosecutions of Blackwater staff are unlikely, as MEJA covers contractors of the
Department of Defense (DoD) and other federal agencies so long as the contract
“relates to supporting the mission of the Department of Defense overseas”. The
relevant Blackwater security personnel were providing support to the State
Department, not the DoD.
In October 2007 the US House of Representative passed the MEJA Expansion and
Enforcement Act of 2007. Although too late for Blackwater prosecutions, this Act
extends the reach of MEJA to all government contractors overseas where armed
forces were conducting “contingency operations”. The Act has now been sent to the
Senate, which must approve it before it becomes law.
The Uniform Code of Military Justice (UCMJ) provides for the administration of
military justice. With regard to civilians, UCMJ jurisdiction was originally limited to
those “serving with or accompanying an armed force in the field” in time of war that
had been declared by Congress. As wars are no longer formally declared, the UCMJ
was amended on 1 January 2007 to further encompass those serving in “a
contingency operation”. Whilst no Executive Order implementing the legislation has
been issued to date, the Deputy Secretary of Defense, in a 25 September 2007
Memorandum, emphasised that “DoD contractor personnel (regardless of nationality)
accompanying U.S. armed forces in contingency operations are currently subject to
UCMJ action. Commanders have UCMJ authority to disarm apprehend and detain
DoD contractors suspected of having committed a felony offense in violation of the
[Rules for Use of Force], or outside the scope of their authorized mission, and to
conduct the basic UCMJ pretrial process and trial procedures currently applicable to
the courts-martial of military service members. Commanders also have available to
them contract and administrative remedies, and other remedies, including discipline
and other possible criminal prosecution.” Professor Schmitt noted that the
constitutionality of the extension of UCMJ jurisdiction over civilians (or at least some
of the civilians encompassed in the 2007 amendment) has been questioned. For
instance, some commentators argue that under the 1957 case of Ried v. Covert
civilians are entitled to due process standards which US military proceedings may not
meet. He noted that it will be interesting to see how the federal courts address this
Professor Schmitt concluded by noting that the United States has taken numerous
steps to monitor and control the use of force by contractors. For instance, US
Central Command (CENTCOM), which controls operations in Iraq and Afghanistan,
requires that civilians acknowledge its Rules on the Use of Force (RUF) in writing.
These rules provide that:
“Civilians Armed for Personal Protection: Are noncombatants. You may not engage
in combat actions with Coalition Forces, or defend Coalition military supplies or
facilities. You are armed for personal protection only.”
“You may use NECESSARY FORCE, up to and including deadly force, against
persons in SELF-DEFENSE, against a hostile act or demonstrated hostile intent
(threat of imminent use of force).”
Before being armed, civilians must undergo training in the Rules on the Use of Force,
Rules of Engagement (ROE) and the Law of Armed Conflict (IHL).
On 23 October 2007, the State Department issued a new State Department Policy
regarding security services. It requires a State Department representative to
accompany all contractors who are performing “Personal Protective Service”
movements. The policy further: requires additional cultural awareness training, as
well as training on the tactics and procedures of the Multi-national Force in Iraq
(MNF-I); restates the US Mission firearms policy, which limits the use of deadly force
to “aimed shots”, requires consideration of the presence of bystanders and mandates
the avoidance of civilian casualties; imposes technical requirements to monitor
operations; and develops procedures that will enhance State Department security
cooperation with MNF-I.
 Similar legal provisions are found in Common Article 3 of Geneva Conventions I-IV, Rule 6 of the
ICRC Study on Customary International Humanitarian Law and Article 8 of the Rome Statute.