BITS Research Note 03.2
April 2003
ISSN 1434-7687


”Non-lethal” chemicals for law enforcement?

by Dr. Walter Krutzsch

  This note is also available as a PDF-File

Editorial Note:
From April 28 to May 9, 2003 the first ever Chemical Weapons Convention Review Conference will be held at The Hague. During the conference state parties to the Chemical Weapons Convention will review the progress made, since the convention, which bans chemical weapons, entered into force. They will review advances in science and technology, which impact the implementation of the treaty. And they will discuss a topic gaining more and more importance: The use of incapacitant agents and so called non-lethal-chemicals either in law enforcement or in warfare, and its implications on both the Convention and international humanitarian law. The latter is the topic of this research note.


1. Introduction

The Chemical Weapons Convention (CWC) was concluded to renounce forever all chemical weapons. Their prohibition must be strictly observed without interfering in the peaceful use of chemicals. This can lead into a dilemma when it comes to discriminate between the prohibited use of chemical weapons and the not prohibited use of certain chemicals by police forces against civil uproar. The provisions of the CWC, correctly interpreted, provide for discerning prohibited activities from not prohibited ones.

Some months ago, information appeared about a U.S. research and development program on non-lethal chemical weapons that violates the CWC. The ‘Sunshine-Project’ reported on such activities regarding anaesthetics and psychoactive substances, development of long-range military delivery devices for these chemicals, including an 81mm chemical mortar round. Their intended use is against "potentially hostile civilians", in anti-terrorism operations[ 1 ], counterinsurgency, and other military operations. The hostage drama in Moscow[ 2 ] was considered as one of the scenarios for which these new weapons could be needed. Speculations about the possible wartime use of those weapons stirred up discussions that raised doubts on the legal line of division between the use of chemical weapons on the one hand and law enforcement measures, including riot control purposes, on the other. The legal aspect of the issue is: Which chemical agents are ‘not prohibited’ under Article II, paragraph 9(d) of the Chemical Weapons Convention?

The statement of Ambassador von Wagner, former chairman of the Conference on Disarmament’s Ad Hoc Committee on Chemical Weapons, made when introducing the final draft of the CWC ten years ago, gained new relevance:

"4. Particularly riot control agents constitute a real problem. These irritants, physically disabling agents are used around the world in law enforcement and riot control, by police and other organs responsible for maintaining law and order. The same agents, however, would constitute an immediate risk and danger if they were allowed to develop into a new generation of non-lethal but nonetheless effective agents of warfare, causing insurmountable problems in trying to distinguish in the ensuing grey area between ‘real’ and ‘non-lethal’ chemical weapons as well as between ‘real’ and ‘non-lethal’ warfare units.

5. Only in the last week of negotiations a point near consensus has been reached on this important issue touching upon the very scope of the Convention. It was possible because a common view has emerged among delegations that the preparation and application of any method of warfare dependent upon the toxic properties of chemicals should be banned under the Convention..."[ 3 ]

2. Warnings against a legalisation of ‘non-lethal’ chemicals

Since the start of negotiations, as early as 1974, documented evidence exists that participants pleaded to confine the use of chemicals for civil purposes to irritants.

A Canadian paper from 1974[ 4 ] states in its paragraph 9 (page 3):

"In the case of harassing or irritating agents which are widely recognised as essential for civil riot control because of their quick reaction and short duration without injury, it is unlikely that governments would be prepared to ban their continued development, production and stockpiling. It might on the other hand be generally accepted that the development, production and stockpiling of incapacitating agents could be prohibited. This acceptance would stem from the unreliability and unpredictable effects of incapacitating agents, particularly the psychochemicals. It would seem unlikely that governments would wish to retain such agents for civil police use. In the event of there being a disposition to prohibit incapacitating agents but to allow irritating agents for civil use, an expert review committee could determine into which category fell those chemicals above the agreed threshold of effectiveness".

A paper submitted by the U.S. to the CWC negotiating body in 1977[ 5 ] stated likewise: "In addition to chemicals that kill or permanently disable, chemicals which have temporary, incapacitating effects are potential chemical warfare agents. For this reason, it is appropriate to consider their inclusion in a future arms control measure. The draft Conventions presented by the Socialist countries (CCD/361), Japan (CCD/420), and the United Kingdom (CCD/512), all appear to place restrictions upon incapacitants as well as on other agents. In addition, the 10-nation memorandum on CW (CCD/400) would seem to advocate the prohibition of incapacitants. One final conclusion of the U.S. paper is: "At present incapacitating agents do not appear to have become a major component of CW stockpiles. Their role could increase, however, if they were not covered in a CW agreement."

In addition to the states mentioned in the U.S. paper[ 6 ], different political and regional groups supported this view[ 7 ]. Towards the end of the negotiations, the negotiating committee of the CD warned against the serious dangers posed by non-lethal chemicals, if they were to develop into a new generation of effective agents of warfare[ 8 ]. Until the conclusion of the Convention there has been no position demanding that chemicals whose toxic effects transgress the boundaries of the definition of riot control agents should be eligible for law enforcement. Presently, strict observance of these boundaries has become even more important.

A paper issued in November 2002 by the Federation of American Scientists[ 9 ] characterises the present dangers posed by new non-lethal agents as follows:

"Biomedical sciences and the pharmaceutical industry are in the midst of a revolution in the science and technology of drug discovery that will significantly complicate the control of chemical and biological weapons (CBW). The Chemical Weapons Convention (CWC), the Biological and Toxin Weapons Convention (BWC) and the Geneva Protocol are thus challenged by these technological developments. Scientists contributing to this revolution need to understand the implications of their work, and arms controllers must recognise that there are profound changes underway that will affect the technical landscape of CBW control..." (page1)

"In fact, a categorical distinction between lethal and non-lethal agents is not scientifically feasible. Not only are certain individuals more susceptible to some agents, but synergy between two different non-lethal agents may make their combination highly lethal to everyone. Rational strategies to discover such synergistic pairs will soon be available. Thus, the development of multiple non-lethal agents may provide a lethal CW capability, in clear violation of the Convention. Even without synergism, stockpiles of non-lethal weapons and munitions would defeat a fundamental goal of the Convention, to exclude completely the possibility of the use of chemical weapons by preventing states from entering a war with a stockpile of CW whose use is proscribed, but which might nevertheless be employed under pressure of military necessity." (page 2)

3. Treaty provisions have to be interpreted correctly or will be negated

To prevent damage to the CWC, its legal authority must be strengthened in order to uphold the fundamental humanistic achievements of this legal instrument. Implementation of the CWC requires its interpretation. The rules of treaty interpretation serve to apply the agreed provisions. Correct interpretation is guided by the principle ‘Pacta sunt servanda’. The Vienna Convention on the Law of Treaties, concluded on 23rd of May 1969 (‘Vienna Convention’) codifies the norms of customary international law on treaties which is binding for states and international organisations. The general rule governing the interpretation of treaties is contained in Article 31 of the Vienna Convention. Neglecting that rule would mean negating the law. The following elements of this rule have a special bearing on the issues dealt with in this article:

Paragraph 1: A treaty shall be interpreted

  • in good faith
  • in accordance with the ordinary meaning to be given to the terms of the treaty
  • in their context and
  • in the light of its object and purpose.

Paragraph 3: There shall be taken into account, together with the context:...

(c) Any relevant rules of international law applicable in the relations between the parties.


4. The dividing line between activities prohibited under Article I and activities not prohibited

According to the concept on which Article II is based, any toxic chemical as defined in its paragraph 2 is a chemical weapon, "except where intended for purposes not prohibited under this Convention, as long as the types and quantities are consistent with such purposes…"

Paragraph 9 of Article II divides the exceptions into four categories of purposes ‘not prohibited under this Convention’. Each category stands by itself.[ 10 ]

Sub-paragraph 9 (d) exempts the use of toxic chemicals for the purpose of law enforcement including domestic riot control from regulation. Nevertheless, their use with the intent to cause harmful effects for humans moves them closer again to a prohibited CW employment. The interrelated provisions discussed in this chapter separate not prohibited action from prohibited use of chemical weapons.

  • ‘Law enforcement including domestic riot control’ presupposes a specific factual situation in which domestic law and order are violated or endangered. Use of force by police or other organs must be allowed within the scope of a state’s jurisdiction to re-establish law and order[ 11 ].
  • The consistency criterion of paragraph 1 of Article II stipulates, that types and quantities of chemicals have to be consistent with the legitimate purpose. Its impact on sub-paragraph 9(d) is significantly upgraded by paragraph 7 of Article II. A priory, it excludes the types of agents named in one the CWC lists from being used for riot control. The chemicals qualified as riot-control agents are characterised by their toxic effects. Such effects are defined in much broader scope than those related to irritants. Beside sensory irritation they also include physical disablement. Effects must manifest themselves rapidly after exposure and disappear within a short time after termination of use. These are the only type of chemical agents which are eligible under sub-paragraph 9(d)[ 12 ]. Further restrictions on these exemptions are: Use of riot-control agents in bombs, spray containers or artillery munitions, which would be inconsistent with the purpose. The same holds true for stored quantities in excess of the requirements for the allowed purpose. All those chemicals not in accordance with these purposes, together with their means of delivery, are prohibited chemical weapons. They have to be declared and destroyed.
  • The third provision in this context is paragraph 5 of Article 1 of the Convention. It prohibits any use of riot control agents as a method of warfare. It does not cover other categories of toxic chemicals since, according to paragraph 7 of Article II, they are not eligible for purposes of sub-paragraph 9(d). Paragraph 5 of Article I also relates to internal armed conflicts, including local, relatively small-scale outbreaks of violence. Therefore, lawful use of riot control agents has to be counterchecked to confirm that it does not constitute a method of warfare. The expression ‘method of warfare’ has been coined by the Geneva Conventions[ 13 ]. The humanitarian character of those norms requires their broad interpretation.
  • These provisions are further supported by Article III, obligating each State Party to declare chemicals it holds for riot control purposes.

5. Context

The provisions discussed under Chapter 4 are closely interrelated. The term ‘riot control’ in sub-paragraph 9 (d) is the same as that defined in paragraph 7 of Article II. This adds precision to the consistency criterion in Article II, para. 1 when applied to sub-paragraph 9 d). Since the same term ‘riot control’ is used in the 5th paragraph of Article I, any toxic chemical intended to be used for law enforcement including domestic riot control is prohibited as a method of warfare. Article III, paragraph 1 e) provides for the transparency required. This context gives conclusive proof, that sub-paragraph 9 d) cannot legitimise the use of any chemical agent except those defined in paragraph 7.


6. The ordinary meaning of terms

According to their ordinary meaning, ‘law enforcement’ is a term with a general connotation, ‘domestic riot control’, in contrast, with a special connotation. Consequently, when the question is raised which measures may be needed to enforce law and order and which measures would be adequate for domestic riot control, the answers will differ significantly:

A police officer on night patrol reprimanding individuals for disturbing sleep is a measure of ‘law enforcement’. The same applies for the dislocation and action of units of police and other security forces when riots are immanent. The difference lies in the character and degree of the enforcement measures – ten Euros fine on the one hand, water-canons, police-sticks and tear-gas on the other. In contrast, the term domestic riot control is unambiguous and requires enforcement measures only of one special character, implying, among others, riot control agents.

The two terms are interrelated, the specific term (domestic riot control) being contained in the general one (law enforcement). This relationship is correctly expressed by the word ‘including’. Law enforcement and riot control are not alternatives but parts of one category of purposes. This understanding is also supported by other official versions of the convention’s text:

  • French: des fins de maintien de l’ordre public, y compris de lutte antiémeute sur les plan intérieur
  • Spanish: mantenimiento ordon, incluida la represión interna de disturbos
  • Russian: pravoochanitel’nie celi, vkljucaja borbu s besporjakami v strane (law protection including fight against riots in the country).

Furthermore it should be mentioned that the treaty text reflects the sense of wording regularly used during the negotiations. Between 1982 and 1992 ‘irritants for law enforcement’ was used once[ 14 ] and the wording ‘irritants for law enforcement and riot control’ used in 14 editions of the Rolling Text.[ 15 ][ 16 ]

7. The context with the Geneva Protocol

The Preamble of the CWC refers to the Geneva Protocol in three paragraphs:

  • Paragraph 3 recalls that the UNGA condemned all actions contrary to the principles and objectives of the Geneva Protocol;
  • paragraph 4 recognizes that the CWC reaffirms principles and objectives of and obligations assumed under the Geneva Protocol;
  • paragraph 6 states the resolve that the CWC, by excluding completely the possibility of the use of CW through implementing its provisions, complements the obligations assumed under the Geneva Protocol.

These statements are part and parcel of the object and purpose of the CWC as such: To confirm the principles and objectives of the Geneva Protocol and to complement the obligations assumed accordingly. For this mutual relationship between two treaties Article 30, paragraph 2 of the Vienna Convention provides: "When a treaty specifies that it is subject to, or that it is not to be considered as incompatible with, an earlier or later treaty, the provisions of that other treaty prevail.".

As far as the scope of the prohibitions by the Geneva Protocol and customary international law are concerned, this paper can only refer to the legal literature commenting on the discussion of this subject, especially to volume III of the SIPRI study[ 17 ]. This study comes to the conclusion that the use of all toxic agents including non-lethal weapons, tear gas and other irritants in war and in other armed conflict shall fall within the scope of the prohibition of the Protocol as well as of the customary prohibition – according to the generally recognized rules of international law[ 18 ]. Important evidence for this interpretation is United Nations General Assembly Resolution 2603 A (XXIV) which affirmed the comprehensive character of the prohibition.

The SIPRI study examined the history of the emergence of those rules in the period following World War I up to the end of the 1960’s. New impulses, also in reaction to the use of irritants and herbicides in the Vietnam war, gave rise to a process that finally resulted in the conclusion of the Biological and Toxin Weapons Convention and the Chemical Weapons Convention. References in the Appendix prove that throughout the process there was constant awareness of the comprehensive scope of the CW prohibition:

  • the Report of the UN Secretary-General containing the unanimous report of the group of consultant experts[ 19 ];
  • the WHO Report on Health Aspects of Chemical and Biological Weapons;[ 20 ]
  • the Statement of President Nixon of November 1969, reaffirming renunciation of the first use of lethal chemical weapons and extending this renunciation to the first use of incapacitating chemicals[ 21 ];
  • President Nixon’s message of August 1970 initiating the ratification process of the Geneva Protocol.[ 22 ]
  • President Ford’s Executive Order 11850, of 4 August 1975 concerning renunciation of certain uses in war of chemical herbicides and riot control agents.[ 23 ]

Throughout the negotiations of the Chemical Weapons Convention, the achievement of a comprehensive CW ban with a comprehensive scope like the Geneva Protocol was the generally shared objective. From the reference documents contained in the Appendix to this research note it can be seen that the entire negotiation process of more than 20 years was a constant affirmation of this objective as reflected by state practice and opinio juris, in other words, by the belief that this practice is consistent with the existing law of this time. According to the principle of estoppel[ 24 ] such affirmative international commitments strengthen international law and protect it against opportunist interpretation. This process reached its climax in January 1989 at the Paris Conference where 149 representatives of the State Parties to the Geneva Protocol and other interested states (that means states not party to the Geneva Protocol) declared their determination "to prevent any recourse to CW by completely eliminating them"[ 25 ].

After closure of the CWC negotiations, the process of strengthening the Geneva Protocol continued through signature, ratification of and accession to the Convention. As stated in paragraph four of the Preamble, all States Parties (be they parties to the Geneva Protocol or not) are committed to abide by the principles and objectives of and obligations assumed under the Geneva Protocol. Therefore, those rules are applicable in the relations between the States Parties to the CWC as required in the Vienna Convention, Article 31, paragraph 3 (c).

From this follows: Any interpretation of Article II, paragraph 9 (d) of the Convention has to be in accordance with the comprehensive prohibition of use of chemical weapons stipulated by the Geneva Protocol. (Solely the use of irritants for civil purposes could be considered as not covered by the provisions of the Geneva Protocol.) This excludes interpretation of sub-paragraph 9 (d) that would exempt from prohibition any agent that is not a riot control agent as defined under Article II, paragraph 7.

8. Circumvention of the legal dividing line between prohibited and not prohibited activities

By false interpretation[ 26 ] non-lethal chemicals, while excluded by definition from use for law enforcement including riot control purposes, are made eligible for such purposes. ‘Law enforcement’ is considered to be a purpose in itself. The argument goes: Since agent X is not to be used for riot control purposes but for law enforcement purposes, it cannot be a riot control agent. Therefore, it does not fall under of Article II, paragraph 7. Consequently, the consistency criterion is not further specified and the relation with the undefined term ‘law enforcement’ deprives it of most of its relevance. Also the safeguard against misuse as a means of warfare in Article I, paragraph 5 is bypassed. As a surrogate for this provision, sub-paragraph 9 (c), despite having quite another function, is artificially constructed into a clause against use of chemical agents as means of warfare.

This false interpretation negates the rules of the Vienna Convention in toto: Pacta sunt servanda, good faith, object and purpose of the treaty, ordinary meaning of terms, context – here the interrelationship of sub-paragraph 9(d) with paragraph 7 and 1 (a) of Article II, paragraph 5 of Article I, and Article III,1 (e) as well as the relationship to the Geneva Protocol (Article 30, paragraph 2 and Article 31, paragraph 3 (c)).


9. Quintessence

The question whether a toxic chemical is ‘not prohibited’ under sub-paragraph 9 (d) will be answered by subsuming the intended purpose of its use, its chemical properties, types (including its weaponisation), and quantities under provisions of Article II:

  • Is it a toxic chemical as defined in paragraph 2 ? – if the answer is yes:
  • Is it intended to be used for purposes of law enforcement including domestic riot control as defined in sub-paragraph 9 (d) and is it not intended for use as a method of warfare? – if the answer is yes:
  • Are the types and quantities of the chemical consistent with the intended purposes, and more importantly, is the chemical not included in the CWC list and is it a riot control agent as defined in paragraph 7 and not prohibited by the Geneva Protocol?

If the final answer is ‘yes’, the result will be: the chemical is not a chemical weapon. However, it is prohibited as a method of warfare according to Article I, paragraph 5, and subject to declaration according to Article III, paragraph 1(e). If the answer is ‘no’, it is to be declared and destroyed under Articles I, III and IV of the CWC.

10. Further action

Consideration and action by the States Parties, the policy-making organs of the OPCW, especially at the CWC Review Conference, as well as public discussion on this subject is required:

The Review Conference should

  • reaffirm that the prohibitions of the CWC cover all chemicals regardless of their origin or method of production[ 27 ];
  • recall, that die CWC complements the obligations not to use such weapons, assumed under the Geneva Protocol of 1925 and
  • warn against any violation of these norms.


Dr. Walter Krutzsch was a member of the GDR delegation negotiating the Chemical Weapons Convention from 1985-1990; he served as a Senior Legal Officer of the Preparatory Committee for the OPCW from 1994-1998.





[1] See: Appendix, 57

[2] The opiate fentanyl, an anaesthetic, was used to free 700 hostages in a Moscow theatre; 128 hostages were killed and a great number injured by the gas.

[3] See Appendix, 51

[4] See Appendix, 14

[5] See Appendix, 19.

[6] See Appendix, 6, 11, 13, 18.

[7] See Appendix, 8, 9, 10, 14, 16, 20, 21, 24, 25, 27, 28, 30, 31.

[8] See Appendix, 51, 52, 53.

[9] Federation of American Scientists Working Group on Biological Weapons, Non-lethal biological and chemical weapons, Washington, November 2002

[10] This applies also for sub-paragraph 9 (c) that cannot influence the interpretation of sub-paragraph 9(d).

[11] Likewise, this wording is understood as implicitly tolerating a state’s practice of using toxic chemicals for capital punishment. This was considered necessary to avoid creating an obstacle to enter into the CWC without compromising the object and purpose and, at the same time, to avoid infringing upon the world-wide endeavours to abolish capital punishment. This intention would be counteracted by using it as an argument for a misinterpretation of sub-paragraph 9 d) in order to justify the would-be legalisation of non-lethal chemical weapons.

[12] See Krutzsch, Trapp, A Commentary to the Chemical Weapons Convention, Martinus Nijhoff Publishers, Dortrecht/Boston/London, 1994, p.36.

[13] The expression ‘method of warfare’ has been used, i.a. in the Protocol additional to the Geneva Conventions of 12 August 1949 relating to the protection of victims of international armed conflicts . (Protocol I ) 12.12.77, Part III, Methods and Means of Warfare and Section I, Methods and Means of Warfare. With regard to the obligation to respect the rules of warfare, Article 35 – Basic Rules – stipulates: ‘In any armed conflict, the right of the Parties to the conflict to choose methods and means of warfare is not unlimited.’ The applicability for non-international conflicts results from Article 43 – Armed forces: ‘The armed forces of a Party to the conflict consists of organised armed forces, groups and units which are under a command responsible to that Party for the conduct of its subordinates, even if that Party is represented by a government or an authority not recognised by an adverse Party.’ See also: Krutzsch, Trapp, ibid. p.18-19.

[14] Appendix 29

[15] Appendix 32, 33, 34, 35, 36, 37 ,38, 40, 41, 42, 43, 44, 45, 46

[16] Between 1982 and 1992, relevant text was used in at least 36 cases in the records of negotiation results (‘Rolling Text’) and official papers of participating delegations. In those cases the following terms were used synonymously, often two different versions in one paper in a number of cases:

law enforcement purposes (2), domestic law enforcement and riot-control purposes (2), law enforcement and riot-control purposes (15), domestic law enforcement and domestic riot-control purposes (14), domestic law enforcement or riot-control purposes, such as CS, CN and CR (1), law enforcement including domestic riot-control purposes (2)

[17] A comprehensive overview is given in "The Problem of Chemical and Biological Warfare", Volume III, CBW and the Law of War; SIPRI, Stockholm International Peace Research Institute. 1973 Almqvist & Wiksell.

[18] See also "The Problem of Chemical and Biological Warfare" ibid. p. 59 to 64.

[19] See Appendix, 1.

[20] See Appendix, 2.

[21] See Appendix, 3.

[22] See Appendix, 3.

[23] See Appendix, 15. However, of the four exceptions from the prohibition specified in the Executive Order only that concerning the use of riot control agents in prisoner-of -war camps may be in compliance with paragraph 9 (d) of the CWC. The others have been overruled by the CWC since paragraph 5 of Article I prohibits the use of riot control agents as a method of warfare.

[24] See "The Problem of Chemical and Biological Warfare" ibid. p. 64.

[25] See Appendix, 39.

[26] See: "Law Enforcement and the CWC, the CBW Conventions Bulletin, Quarterly Journal of the Harvard Sussex Program on CBW Armament and Arms Limitation, issue No.58, December 2002.

[27] See Malcolm Dando, Scientific and technological change and the future of the CWC: the problem of non-lethal weapons. disarmament forum (four . 2002), p. 42. United Nations Institute for Disarmament Research.



go to the next page for Appendix - Documents Related to "Non-lethal Che-micals and Law Enforcement including riot control"