Guest Post: Self-Defence, Collateral Damage and Precautions in Attack

by Ian Henderson and Bryan Cavanagh

[Ian Henderson is a group captain in the Royal Australian Air Force and is currently posted as the Director Military Law Centre and Deputy-Director Asia-Pacific Centre for Military Law. Bryan Cavanagh is a squadron leader in the Royal Australian Air Force and is currently posted as a legal training officer at the Military Law Centre and Asia-Pacific Centre for Military Law. This note was written in their personal capacities and does not necessarily represent the views of the Australian Government or the Australian Department of Defence. This is the second in a four-part series and the first post can be found here along with a response here.]

This is the second in a series of four posts that address the relationship between self-defence and LOAC. Yesterday we looked at when self-defence does and does not apply during a period of armed conflict. Today we discuss whether it is lawful under the criminal law concept of self-defence to cause incidental injury (aka, collateral damage) and whether self-defence imposes requirements similar to the ‘precautions in attack’ under article 57of Additional Protocol I (API).

Killing and injuring people that are not the object of the attack

Under the Law of Armed Conflict (LOAC), the rules relating to collateral damage are fairly easy to state – but difficult to apply. LOAC permits expected incidental loss of civilian life and injury to civilians (collateral damage), provided that the collateral damage is not excessive in relation to the military advantage anticipated to be gained from the attack. What then is the law relating to causing collateral damage when acting in self-defence?

A response in self-defence must be reasonable – so the question becomes – when (if ever) is it reasonable to kill or injure people who are not attacking you when responding to those who are? We had some difficulty in tracking down authority on point — in fact we could not find any reported Australian case law that addressed the issue — but domestic law of the US does. Case law in the US has held that while acting in self-defence can excuse injury or even death to a bystander in certain circumstances (eg, when not acting carelessly), self-defence does not excuse knowingly or recklessly injuring or killing a bystander (see Henwood v People, 54 Colo 188 (1913) [8]). Our conclusion is that it is highly likely that reasonableness under the law of self-defence imposes a higher standard of care on a military member than LOAC to avoid causing any injury or death to civilians. In other words, where a person acts ‘lawfully’ in self-defence, self-defence would operate as a successful defence to a charge relating to injuring or killing the attacker but not to a charge relating to injuring or killing a bystander.

To remove some confusion around this topic, we spelt out what ‘proportionality’ means under LOAC and self-defence. Proportionality under LOAC is used as a reference to the collateral damage equation mentioned above. Proportionality under self-defence is about the degree of force used in response to a threat. Proportionality under self-defence does not directly address the issue of collateral damage.

Precautions in attack

Article 57 of API sets out a number of precautions applicable to those who plan, decide and execute attacks. For example, they must do everything feasible to verify that the objectives to be attacked are military objectives; take all feasible precautions in the choice of means and methods of attack with a view to avoiding, and in any event to minimizing collateral damage, and to cancel or suspend an attack if it becomes apparent that the objective is not a military one or that the attack may be expected to cause excessive collateral damage.

We concluded that it is unclear whether the requirement of reasonableness under the law of self-defence would extend to requiring a military member to take all of the precautions set out in article 57. Even if it could be said that these requirements are relevant to the reasonableness assessment, they are unlikely to be as effective as protecting the civilian population as the explicit requirements set out in the article. We view this as unsurprising, as the law of self-defence was not developed to specifically address these types of issues that are unique to military operations.

In tomorrow’s post, we will compare how LOAC and the law of self-defence deal with a number of discrete issues like use of prohibited weapons, obedience to lawful commands, and a ‘duty’ to retreat.

5 Responses

  1. Again, the international law issues are different from the domestic law issues regarding self-defense under domestic law. One would have to make a comparative study of the various domestic legal processes to even come close to providing “an answer.” I note that in our International Criminal Law casebook (Paust, Bassiouni, et. al, International Criminal Law (4 ed. 2013)–, Professor Bassiouni has, at least in the past, stated that there is a rule of “proportionality” that requires that one not take more lives than one is defending, etc. I do not believe that there are real world general patterns of practice and/or opinio juris to back up that claim. At its extreme, a soldier defending himself and 5 others could not kill 50 enemy soldiers advancing on their position, much less the 50 enemy plus 50 civilians that were human shields. But what is the domestic law re: is similar domestic circumstance — in some 200 states and nations??

  2. p.s. and this sort of problem comes up re: Rome Statute of the ICC< art. 31(1)(c)&(d) — because self-defense must be "in a manner proportionate…." and duress must be without an intent "to cause a greater harm than the one sought to be avoided." What will the ICC do with these??
    Professor Bassiouni helped to put together various drafts of various provisions at the Rome Conference.

  3. There are a couple of issues that you may wish to address in order to make your draft chapter stronger. First, you need to address the international human rights law governing the use of force for construing the Proportionality and SIrUS rules in GC-Prot.1. There is quite a bit of human rights case law addressing these rules. See Martin, et al. International Human Rights and Humanitarian Law 529-43(Cambridge Univ. Press 2006).

    Also, I am surprised that you did not address the right of inhabitants (which includes civilians) who spontaneously take up arms against invading enemy forces under GC3, art 4 (6), to be treated as POWs and, therefore, not to be prosecuted for killing the enemy under GC3 art. 99.

    Hope this is helpful.

    Francisco Forrest Martin

  4. And, hhmmm, what is the effect of the Europ. Conv. on H.R., art. 2(2)(a) (self-defense — “in defence of any person from unlawful violence”). Is there a human right to self-defense?

  5. Jordan,

    We agree that the international law rules and domestic self-defence rules are different. Indeed, that is the very point of our article. Having trained a soldier to understand and comply with the LOAC, we foresee potential problems if that soldier then relies on ‘self-defence’ as the relevant legal justification. And all the more so as it may not even be their own domestic law that is relevant, butt he law of self-defence is another jurisdiction.


    Re IHRL, that is almost exactly what my colleague Bryan Cavanagh said. We did not go there due to the ever constant curse of a word limit from our editors. As for levee en masse, again we had a word limit, but mainly we were primarily interested in looking at how soldiers are trained under LOAC and contrasting that with domestic self-defence law. However, both of your points are well noted and hopefully someone else might pursue the issues further in a longer document.


    We briefly flirted with discussing the HR of self-defence, but see above for the cursed word limit. In the end, we decided that we would look at two particular jurisdictions that had codified the law on self-defence (the Australian federal jurisdiction and the ICC). It will have to be for others to consider what relevance international or regional HRL might have in their own jurisdictions. At least for the Australian federal jurisdiction (unburned as we are with a charter or bill of HR) and the ICC, I expect any purported HR of self-defence is more of a policy issue than a substantive legal issue.

Trackbacks and Pingbacks

  1. There are no trackbacks or pingbacks associated with this post at this time.