1 Jamie Shea, at the NATO press briefing on 26 March 1999, website: www.nato.int.

2 Air Commodore David Wilbey, at the NATO press briefing on 1 April 1999.

3 Air Chief Marshal SirJohns, Richard, “Air power in a new era”, Royal Society of Arts Journal, 3–4 of 1999, p. 99Google Scholar.

4 As when a bomb fell 600 metres short of the target, at Aleksinac. There was also the incident on 12 April when missiles were fired at a railway bridge near Leskovac, just as a train from Belgrade to Thessaloniki was crossing it.

5 Indeed, the preamble to the Declaration of St. Petersburg of 1868 proclaimed that the only legitimate object of war was to weaken the military forces of the enemy and that, for this purpose, it was sufficient to disable the greatest possible number of men.

6 For the British Empire forces, 908, 371 military battle deaths and 2,090,212 military wounded out of a mobilized military strength of 8,900,000. See the table in the British Army Review, August 1996, p. 79.

7 The Independent, 4 November 1999.

8 Op. cit. (note 6), p. 80.

9 Ibid. According to Erik Durschmied, The Hinge factor, Hodder & Stoughton, 1999, Iraqi military losses were over 100,000 whereas Coalition forces suffered a total of 192 killed, of whom 35 died from “friendly fire” and two were killed dismantling a bomb. “In military terms, such a ratio is called the Zero Factor.”

10 “NATO won the Kosovo conflict without a single life lost in combat operations on its own side (…) for just two aircraft lost”, Nick Cook, “War of extremes”, jane's Defence Weekly, 7 July 1999. Two Gurkha engineers, Lt. Gareth Evans and Sgt. Balaram Rai, were killed dealing with unexploded ordnance during the operation.

11 According to the Swiss Federal Office for Civilian Protection, the ratio of the First World War was 200 military: 1 civilian; in the Second World War nearly 1:1 and in the Vietnam War, 1 military: 20 civilians. See Bouvier, M. Sassoli/A., How Does Law Protect in War?, ICRC, Geneva, 1999, p. 145Google Scholar.

12 “While such errors were inevitable, observers contend, widespread public faith civilearly on in the campaign in the ability of modern surveillance systems and smart weaponry to restrict collateral damage heightened reaction against the bombing campaign when civilian casualties did occur and led to intense pressure at times on politicians to call a halt to the attacks … If politicians insist on minimum — and perhaps even zero — attrition rates among their military personnel in future air campaigns, a range of new weapons, up to and including UCAVs, will need to be introduced as soon as the technology allows, The alternative for the military is to brace civilearlyians - politicians, media and public alike - for the realities of attrition in parallel with an increased investment in equipment and tactics that support manned close air support and battlefield air interdiction missions at low as well as high altitudes.” Cook, op. cit. (note 10).

13 Johns, op. cit. (note 3), p. 96.

14 Allen, Charles (ed.), Thunder and Lightening, HMSO, 1991, p. 74, 80Google Scholar.

15 Ibid. p. 81/2.

16 Ibid., p. 137: “Aircrew usually had the comfort of being remote from the destruction caused by their bombs or rockets.”

17 Ibid., pp. 113–114: “That really brought it home (…) we were actually killing people “

18 According to Wing Commander Greg Bagwell, “Precision weapons”, Royal Air Force Air Power Review, Spring 1999, p. 7, the Al Firdos bunker in Baghdad was assessed as a communication node but, unknown to allied intelligence, it was also being used as a civilian air raid shelter. The Chinese Embassy in Belgrade was hit by mistake on 7 May 1999 owing to the use of faulty intelligence at the planning stage which went unnoticed through subsequent checks. See the report in The Independent, 11 May 1999.

19 Where thermal, acoustic or radar “blueprints” of enemy military equipment are programmed into a weapon's memory. Op. cit. (note 18), p. 7.

20 Apart from the new US joint direct attack munition (1DAM) which can be deployed from altitudes above cloud cover. See Joint Statement on the Kosovo After Action Review, Department of Defense, 14 October 1999.

21 Op. cit. (note 14), p. 81: “… in reality you cannot have a surgical war. There is no such thing. For a bomb to go down an elevator shaft is just luck. It probably has a CEP - a computer error probability - of maybe 30–40ft. That can make the difference between hitting the place you're aiming for and the one next door which is full of civilians.” Incidentally, other authors define CEP as “circular error probable”, i.e. the radius of a circle centred on the target within which 50% of all bombs dropped will land. Bagwell, op. cit. (note 18), p. 14. Bagwell, p. 8, also comments that “the explosive force of a 5001b warhead does not constitute keyhole surgery!”

22 Ibid., p. 13, quotes $9,000 for a simple GPS guided bomb and $500,000 for a Block IlI Tomahawk missile,

23 Ibid., p. 3.

24 Defined by Bagwell, op. cit. (note 18), p. 14, as those requiring no human input during the final phase of flight.

25 At the NATO press briefing on 18 April 1999, an account was given of how a pilot launched an attack against a radar and, noticing that the site was close to a church, pulled his weapon off the target so that it exploded harmlessly in a wood.

26 Op. cit. (note 10).

27 Robertson, Lord, Kosovo: An Account of the Crisis, Ministry of Defence, London, 1999, p. 13Google Scholar.

28 At 1800 hours BST on 17 April 1999.

29 NATO press briefing, 15 April 1999.

30 Or possibly Terzick Most Zrze.

31 The Independent, 17 April 1999.

32 To be found on www.fas.org/man/dod/docs99/s99O419b.htm.

33 At the time of the Kosovo conflict, for example, the USA and France were not party to Protocol I but Germany and the UK were.

34 Protocol I, Art. 57, para. 2(a)(i).

35 The United Kingdom made a declaration of understanding to this effect when ratifying Protocol I. See IRRC, No. 322, March 1998, pp. 186 ff. - The Conventional Weapons Convention of 1981 contains similar language in the Amended Mines Protocol of 1 May 1996, Art. 3, para. 10.

36 Protocol I, Arts 50, para. 1, and 52, para. 3. With regard to Art. 50, para. 1, the United Kingdom, on ratification of the Protocol, made a statement of understanding that the rule applied “only in cases of substantial doubt still remaining after the assessment (…) has been made, and not as overriding a commander's duty to protect the safety of troops under his command or to preserve his military situation, in conformity with other provisions of the Protocol.” Ibid.

37 This passage is adapted from the author's text in the Model Manual on the Law of Armed Conflict, ICRC, Geneva, 1999Google Scholar.

38 The United Kingdom, on ratification of Protocol I, declared that “Military commanders and others responsible for planning, deciding upon, or executing attacks necessarily have to reach decisions on the basis of their assessment of the information from all sources which is reasonably available to them at the relevant time.” Loc. cit. (note 35).

39 Bagwell, op. cit. (note 18), p. 9, thinks that this should be dealt with in rules of engagement: “… some targets could require such guaranteed accuracy that the aircraft/platform might be placed at increased risk to enemy defences. In this case the ROE will need to make the desired identification criteria extremely clear so that crews, who are effectively judge and jury, will be able to take the right course of action.”

40 Rogers, A. P. V., “Conduct of combat and risks run by the civilian population”, Military Law & Law of War Review, 1982, p. 310Google Scholar.

41 British Defence Doctrine (JWP 0–01) issued by the British Minster of Defence in 1996.

42 Protocol I, Art. 58.

43 See Rogers, A. P. V., Law on the Battlefield, Manchester University Press, 1996, p. 79Google Scholar.

44 Statute of the International Criminal Court, Art. 8, para. 2(b)(iv).

45 Lee, Roy S. (ed), The International Criminal Court, Kluwer Law International, 1999, p. 111Google Scholar.