Melbourne Journal of International Law
[Diplomatic assurances raise a number of legal as well as communicative questions. Whom do they assure and what do they assure? How are they structured and what does this structure imply about the authority of human rights? This article explores the effects of diplomatic assurances on the international law of human rights. First, it demonstrates that a diplomatic assurance can constitute an agreement to disagree on the precise content of international human rights law, and that diplomatic assurances must be regarded as treaties under international law. There is an uneasy relationship between diplomatic assurances and multilateral human rights instruments and the law of treaties is unable to reduce ambiguity in this context. Second, this article will explain the structures of which diplomatic assurances form part and will show how such assurances deny the articulation of human rights violations on multiple levels: by courts, by the captive, by the monitoring diplomat, and by the norm itself. Finally, it is argued that the silence thus produced is not accidental, but must be seen as an integral part of the productivity of human rights law in a system of nation states.]
Certain states resort to diplomatic assurances to facilitate and legitimise the removal of non-nationals to third states with dubious human rights records. The removals that these diplomatic assurances facilitate typically involve two countries where human rights are conceived of in different ways and are institutionalised to different degrees. Traditionally, assurances were mainly seen as a migration tool. Over the past few years, however, they have entered the domain of high politics due to their practical importance in the programme of ‘extraordinary renditions’ in the ongoing ‘war on terror’. There are now reports of cases in which such assurances have provided a prelude to the interrogation or incapacitation of perceived enemies in less protective jurisdictions.
Diplomatic assurances can be given different readings. One optimistic interpretation would be that such assurances offer a transmission belt between nation-statist and territorialised systems of human rights protection and are but another step in an ongoing development and specification of universal human rights obligations. It follows from this reading that where they bring about effective monitoring arrangements (for example, through the diplomatic agents of the sending state), they should perhaps be welcomed by the human rights community. However, although it is possible to interpret diplomatic assurances in this way, this optimistic reading has no apparent support in the human rights community.
Another reading would interpret diplomatic assurances as a formal aberration. On this reading, receiving states are already obliged to respect the human rights of a rendered person in accordance with treaties and customary international law. Diplomatic assurances appear as a mere iteration of pre-existing obligations to protect human rights, adding no further substance. The two competing promises involved in such a reading — the universal and multilateral obligation of human rights treaty law versus the bilateral obligation emerging from the diplomatic assurances — seem to cancel each other out and reveal the impossibility of any functioning cross-cultural transmission belt. If the multilateral promise of third-party protection is so frail that it needs backing by a bilateral one, will the latter not share the constitutional defects of the former? Or, if the bilateral promise ‘strikes a balance’ between the different degrees of human rights protection in the two states, will it not demand too little from one, while demanding too much from another? Against this backdrop, it is tempting to reject the emergent practice of diplomatic assurances for the sake of saving the conception of human rights.
There is yet another conceivable reading, which would interpret diplomatic assurances to be indicative of a broader shift in our relation to human rights. This reading would rest on two premises. First, it would look into the communicative processes unfolding around diplomatic assurances. Who do they assure, and what do they assure? How are they structured, and what does this structure imply about the authority of human rights? Second, it would take the legal formalism of assurances seriously, and argue that assurances make sense beyond the doctrinal logic of human rights. However, this is an interpretation that the human rights community might not wish to endorse, as such a reading would use diplomatic assurances to illustrate the ongoing disenchantment with human rights.
In this text, I shall argue that the combination of diplomatic assurances and renditions represents a contemporary form of ban. Their main function is to perform a double move — to assure the rendering community of its own faithfulness to human rights law, while the captive is subjected to human rights law in its withdrawal.
This argument will be developed in three main stages. Parts II–IV will explore the effects of diplomatic assurances on the international law of human rights. I will demonstrate that a specific diplomatic assurance constitutes an agreement to disagree on the precise content of the torture prohibition in international human rights law; that diplomatic assurances must be regarded as treaties under international law; and that there is an uneasy relationship between diplomatic assurances and multilateral human rights instruments — a relationship in which treaty law is unable to reduce ambiguity. In Part V of the article, I will try to explain the structures of which diplomatic assurances form a part, and to show how the assurances deny the articulation of human rights violations on multiple levels: by courts, by the captive, by the monitoring diplomat, and by the norm itself. Finally, I will argue that the silence produced by the assurances is not accidental. Rather, I will argue, it must be seen as an integral part of the productivity of human rights law in a system of nation states.
There can be little doubt that renditions play an important role in the practice of the United States. The US has rendered persons to its own territory as well as to other states since the presidency of Ronald Reagan, with a marked increase since 2001. Other states have contributed to this practice by permitting the removal of persons at the request of the US, or by receiving such persons. In a number of such constellations, diplomatic assurances played a decisive causal role in bringing about renditions. Hence, diplomatic assurances are used frequently enough, and are sufficiently associated with high politics, to warrant an exploration of their effects on human rights law.
To date, a number of authors and organisations have pointed out that diplomatic assurances are insufficient to avert human rights violations in the context of renditions and removals. Yet it seems that the practice of states is not markedly affected by this critique. If we were to imagine that assurances of the type analysed below were to be given for a period lasting another decade or so, and were to involve more and more states, it cannot be ruled out that they might affect the content of human rights obligations under international law. Hence, it seems that a mere reiteration of their incompatibility with human rights obligations by critics will be insufficient in the longer term.
The practice of states in giving and receiving diplomatic assurances could affect the content of human rights as part of both customary law and treaty law. In this article I will limit myself primarily to analysing the effect of diplomatic assurances on international treaty law. In order to facilitate this analysis, I will use the case of Ahmed Agiza and Muhammed El Zary, two Egyptian citizens rendered from Sweden to Egypt at the behest of the US Government.
Agiza and El Zary, both Egyptian nationals, were excluded from refugee status in Sweden on the basis of links to terrorist organisations alleged by the Swedish security police. Subsequently, they were deported from a Swedish airport to Egypt in December 2001 on a US Government plane. Before deciding to execute the removal decision, the Swedish Government obtained diplomatic assurances from its Egyptian counterpart. These assurances took the form of an exchange of aide-mémoires, which have partially been made public by the Swedish government.
In the relevant document, dated 12 December 2001, the Swedish Government formulated its position as follows:
It is the understanding of the Government of the Kingdom of Sweden that the above-mentioned persons [Agiza and El Zary] will be awarded a fair trial in the Arab Republic of Egypt. It is further the understanding of the Government of the Kingdom of Sweden that these persons will not be subjected to inhuman treatment or punishment of any kind by any authority of the Arab Republic of Egypt and further that they will not be sentenced to death or if such a sentence has been imposed that it will not be executed by any competent authority of the Arab Republic of Egypt.
The Egyptian Government responded as follows:
With reference to your aide memoire [sic] dated 12 December 2001, concerning repatriation of the following Egyptian citizens … We, herewith, assert our full understanding to all items of this memoire, concerning the way of treatment upon repatriate [sic] from your government, with full respect to their personal and human rights. This will be done according to what the Egyptian constitution, and law stipulates.
In an address to the Swedish Parliament on 26 January 2005, the responsible Swedish Minister explained the role of the assurances as a precondition for removal:
The Egyptian Constitution prohibits torture, but we are well aware of the methods employed in prisons and detention centres in Egypt. The assessment made was that there was a risk that [Agiza and El Zary] would be exposed to torture. On this issue, our position and our legislation is clear. No one shall be expelled to a country where he or she risks being exposed to torture or other inhuman treatment.
The Government had to address both the risk of terrorist action and the risk of the men being mistreated. Thus the question was whether there was a way to remove these men to their country of origin, where they could be put before a court, by obtaining guarantees that they would be treated well. This was the path chosen by the Government. The Government requested and obtained guarantees, by which Egypt assured that the men would not be exposed to torture or other inhuman treatment, and that they would be given a fair trial. Swedish representatives would be following future trials and regularly visit the men in prison. With this guarantee, the Government took the decision to expel the two men to Egypt immediately, and that the State Security Police would take care of their expulsion.
The effectiveness of the Swedish-Egyptian agreement has been brought into question in light of the treatment to which Agiza and El Zary have been subjected. Under the rudimentary monitoring mechanism in the agreement, the Swedish Ambassador to Egypt visited the rendered Egyptians on a number of occasions. Initial official accounts of the Swedish Government reported no instances of torture. However, when Agiza’s case was brought to the attention of the Committee against Torture, the torture allegations transpired and the Committee found that Sweden had violated its obligations under arts 3 and 22 of the Convention against Torture by removing Agiza.
In the speech quoted above, the Minister addressed the effectiveness of the Egyptian assurances as follows:
Has Egypt lived up to the guarantees it gave Sweden? At present, we do not know. In the period of time since the expulsion took place, there has emerged, on different occasions, serious information to the effect that the men would have been subject to treatment contrary to the guarantees. Egypt has denied this. Therefore, Sweden has demanded an independent international investigation.
At the time of writing, the difference of views persists between the countries as to the necessity and appropriateness of an independent international investigation. Perhaps unsurprisingly, Sweden has not made use of the dispute settlement mechanism available under art 30(1) of the CAT, which would have paved the way for arbitration as well as conditional reference to the International Court of Justice.
In order to understand whether Egypt did in fact honour the guarantees it gave Sweden it is necessary to examine both aide-mémoires and to determine exactly the standards to which they required Egypt to adhere. This will illuminate the effect that diplomatic assurances have on human rights treaties. The Swedish aide-mémoire does not refer to human rights in general. Instead it refers specifically to three dimensions of treatment: fair trial; freedom from inhuman and degrading treatment ‘of any kind’ (which, a fortiori, includes torture); and the non-imposition of a death sentence in the future as well as the non-execution of any previously imposed death sentence.
In the best of cases, the rationale behind this selectivity might be a specific risk assessment on the part of the Swedish Government, which simply saw it as unnecessary to use the broader language of human rights. To wit, all three scenarios could raise issues under Sweden’s obligations under the Convention for the Protection of Human Rights and Fundamental Freedoms.
The Egyptian response combines excess and restraint in a single move. First, it appears to confirm the content of the Swedish aide-mémoire: ‘We, herewith, assert our full understanding to all items of [the Swedish] mémoire’. However, it adds that treatment upon repatriation will be ‘with full respect to [Agiza’s and El Zary’s] personal and human rights’, which exceeds the text of the Swedish aide-mémoire. Then, it ends by confirming that ‘[t]his will be done according to what the Egyptian constitution, and law stipulates’. This last sentence is identically structured as reservations of a type against which Nordic states routinely protest. In this case, however, the Swedish Government apparently had no objections to such language. This is significant because the practice of objecting to ‘domestic law reservations’ by Sweden and other Western countries is well established and widely known. A departure from such practice attains a meaning of its own, and must be reasonably interpreted by Egypt as a tacit endorsement of the reference to domestic norms as a limit to the international obligation relevant for the aide-mémoires.
The Egyptian aide-mémoire thus constitutes a significant reduction of the scope of commitment suggested by the Swedish aide-mémoire. Where the Swedish text demands freedom from inhuman or degrading treatment ‘of any kind’, the Egyptian response specifies this freedom as constituting ‘personal and human rights’ but qualifies the respect for those rights with a reference to the stipulations of the Constitution of the Arab Republic of Egypt and to Egyptian laws. Obviously, the quoted sentence would reduce the assurance to whatever the Egyptian Constitution and Egyptian law stipulates.
What, then, does Egyptian law stipulate on this matter? In its dialogue with the Committee against Torture, the Government of Egypt has argued that art 1 of the CAT forms part of its internal law. However, this notwithstanding, the Committee against Torture has repeatedly recommended that Egypt bring its domestic definition of torture in line with that of art 1 of the CAT. An examination of the relevant norms of the Egyptian Constitution and its penal law confirms that there is a considerable gap between that legislation and the ambit of art 1 of the CAT. Article 42 of the Egyptian Constitution provides that ‘[a]ny person arrested, detained or whose freedom is restricted shall be treated in a manner concomitant with the preservation of his dignity. No physical or moral harm is to be inflicted upon him.’
Egypt’s Penal Code recognises torture as a criminal offence, yet the definition of the crime of torture falls short of the definition in art 1 of the CAT. First, under art 126 of the Penal Code, torture is limited to physical abuse. Second, the quoted provision presupposes that the victim is ‘an accused’. Third, the Penal Code’s provision requires the intent to coerce a confession. Mental forms of mistreatment are excluded, as are cases in which the torture is committed against a person who is not to be regarded as ‘an accused’, or for purposes other than producing a confession.
Even if the argument that the CAT is part of the internal law of Egypt is taken at face value, it obviously does not inform the State’s practice. Formally, it is open to serious doubt whether the reference to ‘what the Egyptian constitution, and law stipulates’ in Egypt’s aide-mémoire indeed includes the CAT, as the norms of the CAT hardly constitute a stipulation of Egyptian law. The better view seems to be that this reference reduces the ambit of the torture definition to a scope falling short of art 1 of the CAT. Consequently, the parties’ exchange of aide-mémoires defers to an overly narrow definition of torture.
One might argue that this cannot really do any harm, as multilateral obligations supersede those of Memoranda of Understanding (‘MOUs’) such as these aide-mémoires. After all, Sweden and Egypt are bound by the International Covenant on Civil and Political Rights and the CAT. However, this conclusion would be hasty. If the exchange of aide-mémoires constitutes a binding instrument of international law, it might have legal consequences independently of the stipulations in the ICCPR and the CAT. Therefore, the question of its binding nature must first be resolved. If the exchange is indeed legally binding, the question of its relation to relevant human rights treaties would arise.
Having determined that the question of whether aide-mémoires will impact on states’ human rights obligations under the ICCPR and the CAT will only arise if the mémoires are legally binding, it is now necessary to ascertain the aide-mémoires’ legal status. While a traditional reading of MOUs would suggest that they should be seen as a legally binding instrument, Aust has suggested that this might not be so in all cases. In his view, legally binding MOUs must be distinguished from MOUs not possessing that quality. Without necessarily endorsing that view, I shall test the exchange of aide-mémoires between Sweden and Egypt against the criteria suggested by Aust. As enunciated by art 2(1)(a) of the Vienna Convention on the Law of Treaties, a treaty is a document ‘governed by international law’, which has been interpreted by the International Law Commission to hinge on the parties’ ‘intention to create obligations under international law’. Thus it must be asked whether the aide-mémoires were exchanged between Sweden and Egypt with the ‘intention to create obligations under international law’. There are two possibilities.
One line of argument would answer this in the negative and conclude that the exchange merely created ‘political obligations’ outside the realm of international law. The language used in the documents could be taken to support this conclusion. It could be further argued that the parties did not intend to create new obligations but rather intended to confirm the relevance of pre-existing obligations to the case.
A more convincing line of argument would affirm that Sweden and Egypt did intend to create obligations under international law. First, from the Swedish perspective, the guarantees were a sine qua non for the legality of removal. No expulsion would have taken place if the guarantees had not been obtained. This fact emerges with all necessary clarity from the Minister’s speech in Parliament quoted above. It is further affirmed by the Swedish representations to the Committee against Torture, which stress the significance of the guarantees for the legality of removal under art 3 of the CAT. It would hardly be conceivable that such reliance be placed on a non-binding, ‘political’ instrument, from which Egypt would be free to release itself at any time without breaching any bilateral legal obligation.
Second, certain elements of the guarantees exceed pre-existing Egyptian obligations under international treaty law. Those parts of the guarantee which put in place a monitoring mechanism and give a Swedish representative access to the captives in detention, as well as the right to observe trials, do not replicate obligations which are part and parcel of human rights law at present. Without a legally binding commitment, this part of the guarantee would have been quite meaningless. Consider removal to a state where torture usually takes place, in defiance of that state’s international obligations. Either the guarantee is legally binding and may, therefore, alter the risk assessment undertaken by a removing state, or it is not binding, and will not affect the risk assessment, in which case removal would constitute a violation of pertinent human rights norms. Therefore, the exchange of aide-mémoires must be seen as a binding instrument of international law, falling within the ambit of the VCLT. By the same token, it must be presumed that states generally intend to create binding obligations when giving and receiving such diplomatic assurances.
As demonstrated above, the exchange of aide-mémoires between Sweden and Egypt constitutes a treaty under international law, and the subject matter that it regulates overlaps significantly with that of multilateral human rights treaties such as the ICCPR and the CAT. These instruments both refer to the prohibition on torture or inhuman treatment, as does the exchange of aide-mémoires. In light of this fact, two relationships must be considered here: first, an inter se relationship between two states exchanging diplomatic assurances, and second, each of these states’ relationship erga omnes partes to the multilateral instruments. Had it not been for the third party interest that is implied in human rights treaties, both could have coexisted without interference. However, as the prohibition of torture is so central to the diplomatic assurances, so is the interplay of obligations both inter se and erga omnes partes.
I have shown above that a conflict of norms looms in this particular area. In particular, removal to Egypt and interrogation by Egyptian authorities in line with the conception of torture set out in Egyptian laws would be incompatible with obligations under art 7 of the ICCPR and art 3 of the CAT, compared with art 1 of the CAT. Were the exchange of aide-mémoires to be interpreted in strict conformity with Egyptian laws, we would be faced with a real conflict. To be sure, the conflict between these norms is horizontal, as the conflicting norms both form part of international law.
How does the norm on torture in the exchange of aide-mémoires relate to the norms on torture in multilateral treaties? Can it alter the content of the said treaties? Is there a risk that the giving and receiving of guarantees will decrease the protective scope of human rights treaty law as it existed before the exchange of aide-mémoires? Or is there any rule insulating human rights law against such alteration, which would call into question the validity of the overlapping regulation in the exchange of aide-mémoires? Inevitably, damage will be done, either to the diplomatic assurances, or to multilateral treaties protecting human rights. Or, one may add, to the coherence of international law.
Methodologically, an interesting pattern emerges in the pursuit of these questions. The outcome of treaty law arguments depends entirely on a prior decision as to whether a specific diplomatic assurance constitutes an agreement to legalise certain forms of torture inter se, or an agreement to enhance protection from torture by setting up a monitoring mechanism. Treaty law cannot answer this prior question. At most, it offers a language to express a preconceived answer.
Zuleeg has emphasised the decisionist moment in treaty conflicts. Where the parties to separate treaties differ, a conflict of treaties without clauses on the relations to other treaties is to be resolved neither by the rule lex prior derogate legi posteriori, nor by the rule lex posterior derogat legi priori, nor by a combination of principles. The solution is found in the principle of political decision rendered by the State facing contradicting treaty obligations. This would leave us with the uneasy conclusion that Egypt and Sweden each possess the prerogative to decide the content of their obligations under the exchange of aide-mémoires.
Let us verify that treaty law offers no better answer. Stating the obvious, the group of states parties to the earlier multilateral treaties and to the successive diplomatic assurances is not identical (there are at present 156 parties to the ICCPR, and 141 parties to the CAT, while the aide-mémoire merely binds two states). Therewith, we may leave the issue of an amendment aside, and focus on the possibility of a modification. Neither the ICCPR nor the CAT expressly prohibits a modification. Therefore, the pertinent rule is art 41(1)(b) of the VCLT, which provides that two or more contracting parties to a multilateral treaty may conclude an agreement to modify the treaty as between themselves alone if:
the modification in question is not prohibited by the treaty and:
(i) does not affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations;
(ii) does not relate to a provision, derogation from which is incompatible with the effective execution of the object and purpose of the treaty as a whole.
Based on a strictly formalist approach, the rendition of a person under diplomatic assurances does not affect the enjoyment by other parties to the ICCPR and the CAT of their rights, nor the performance of their obligations.
To invoke the concept of erga omnes partes obligations at this first stage in order to assert the invalidity of the treaty would deprive the lex specialis at art 41(1)(b)(ii) of its independent operation. Excluding such a conflated interpretation might allow the diplomatic assurance to survive the bar set out in art 41(b)(i). However, any modifications of provisions on torture appear to fall neatly under the criteria set out in art 41(1)(b)(ii). Derogating from the torture prohibition and the definition linked to it would certainly be incompatible with the object and purpose of both the ICCPR and the CAT. Yet this only makes sense if we assume that such modifications actually operate to diminish the legal protection provided for by the multilateral treaties. What if Sweden and Egypt argued that the objective of the diplomatic assurances was not to define torture, but to agree on a bilateral monitoring mechanism? In that case, the protection against torture could have been enhanced rather than diminished, and a modification would be perfectly permissible.
This confusion can be escalated further. Let us consider that the diplomatic assurances both put the torture definition at risk and offer an enhanced monitoring mechanism. Can the overall effect of this couplet of measures be regarded as a derogation from the torture prohibition? After all, Klabbers has shown that an approach identifying the object and purpose of a treaty with single provisions is unwarranted. If we are to describe the object and purpose of the ICCPR and the CAT as augmenting the protection of the individual against certain human rights violations, we are left with the difficult question of how to judge a modification that offers better monitoring at the price of diminished scope ratione personae.
Whatever position one chooses to adopt regarding art 41(1) of the VCLT, it is obvious that both States have failed to follow the notification procedure owed to other states parties under that article. Other states have not been given the opportunity to consider the intended modification by Egypt and Sweden, and to take appropriate steps (as the VCLT is silent on what these may be and what consequences they might have for the inter se relationship). A logical conclusion would be that the modification inter se cannot attain validity in the absence of notification. But what if that absence merely reflects that Sweden and Egypt did not perceive the diplomatic assurances as treaties modifying the ICCPR and the CAT? And what if other states fail to protest against that view?
This brings us to yet another treaty law argument, assuming that diplomatic assurances are regarded as ‘relating to the same subject-matter’ as the ICCPR and the CAT. Formally, art 30(4) of the VCLT would suggest that the latter treaty — the exchange of aide-mémoires — prevails in the relation between the parties bound by it, while the obligations of other parties to the ICCPR and the CAT would remain unaffected. This would open up a ‘flexible geometry’ of human rights protection, where two parties or smaller groups of states could act together to downgrade protection without the relative degree of transparency imposed by the provisions governing amendment or modification. Obviously, this operates to the detriment of the universality of human rights. As a safeguard, doctrinal writers have described human rights norms as being exempted from the applicability of the lex posterior rule.
Would it help to introduce the concept of jus cogens and argue that art 53 of the VCLT makes the exchange of aide-mémoires void, because it violates the prohibition of torture, which is held by a number of authorities to be peremptory? Not really. The objective of the aide-mémoires was not to legalise torture bilaterally. This leaves us with the question of how torture is defined in jus cogens: logically, this could coextend with, or diverge from, the definition provided by art 1 of the CAT (which, in turn, is narrower than the torture concept used by the European Court of Human Rights and the Human Rights Committee). It is precisely this indeterminacy which makes it difficult to argue a divergence between the jus cogens definition of torture and the Egyptian conception of torture, which we have earlier shown to be ambiguous as well. Faced with this tension, art 53 of the VCLT, with its contemplation of clear-cut conflict, seems strangely out of place. Furthermore, the dispute is not necessarily about the precise wording of a legal definition alone — it is as much about whether conduct qualifying as torture took place.
At most, there is a form of damage control mechanism insulating the interpretation of a multilateral treaty from bilateral state practice, which would stop the proliferation of any restrictive interpretation emerging from the exchange of aide-mémoires. Article 31(3)(b) of the VCLT requires the interpreter to take into account, together with the context, ‘subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’. Where subsequent practice is not shared by the ‘parties’ in their entirety, but only by some, this practice shall not, in principle, affect interpretation under art 31 of the VCLT. However, this would not necessarily lead to the conclusion that the interpretations by Sweden and Egypt respectively violate international law. On the contrary, it goes to show that diplomatic assurances create legal ambiguity by contract. The missing reference to human rights law in the Swedish aide-mémoire quoted above and the relativisation of human rights law with a reference to Egyptian legislation can be taken at face value. Neither of the States felt the need to refer to a unifying concept, nor to solicit a unifying interpretation through a monitoring body. Rather, the reference points to the Egyptian legislation as the ultimate source of authority. Neither Egypt nor Sweden felt a need to clarify the relationship between the exchange of aide-mémoires and multilateral human rights treaties — the absence of a notification under art 41(2) of the VCLT is the most explicit indicator of this fact.
This is a disturbing sign of cultural relativism in international human rights law. By accepting guarantees without specifying further the treatment that is to be regarded as prohibited — as compared to the general formulations of human rights instruments — and without setting up a formal mechanism for determining differences of interpretation, Sweden is practically deferring to Egypt’s reading of human rights norms. This strikes against the authority of human rights and subsumes it under that of the state. Legally, both States may argue that the diplomatic assurance merely reiterates relevant domestic and international norms relating to torture, and that any alteration of their international obligations would merely provide for better monitoring.
Whether we like it or not, extraordinary renditions reflect the way in which rendering states interpret and perform their human rights obligations. Ironically, the language of rendition tells us so. Apart from ‘surrendering’ a person to a foreign power, the term ‘rendition’ may also signify ‘interpretation’ or ‘performance’. Important legitimation tasks are performed through the human rights dimension of diplomatic assurances, although in a quite different sense to that which the human rights community’s critique would suggest. Ultimately, human rights are assured not to the captive, but, in a circular move, to the rendering community. This circular assurance can only be properly heard and believed by the rendering community if it is undisturbed by claims that human rights violations nevertheless take place.
Thus, the assurance requires subsequent silence. This silence is ensured by a structure that precludes the suffering of the rendered from being articulated as a human rights violation. First, it creates the immunity needed to enable secret services to network across national boundaries and to build relationships of exchange. Such immunity disallows the articulation of human rights violations within the courts. Second, a ‘double secret’ restricts the captive’s capability to articulate his sufferings as a human rights violation, even when he is visited by diplomats of the rendering state. Third, a ‘double bind’ restrains the monitoring diplomat and ties her or him into the silence produced by the assurance. Fourth, as has emerged earlier, the norm’s capacity to ‘speak for itself’ is hampered by the production of legal ambiguity through the diplomatic assurance. Each of these will now be explained in turn.
Diplomatic assurances regulate the activities of three bureaucratic communities in a single move. By assuring the judicial community (courts, NGOs and legal representatives) that the captive is treated in accordance with certain legal standards, the security community may be given access to that captive at a preferred location. Through the assurance, the judicial community is invited to hand the case over to the security community, and a severely limited monitoring option is offered to the diplomatic community. Compared to a territorial interrogation, the judiciary is unable to effectively monitor the treatment of the captive after removal. In lieu of exercising control, it is asked to have faith in the assurances of a foreign country ex ante. Analogously, international monitoring will remain structurally challenged; even if the Committee against Torture finds Sweden to be in breach of art 3 of the CAT, it will be near impossible to act upon that finding, given that the captive is no longer within Swedish jurisdiction. In all, this shifts power from the judiciary to diplomacy and security networking. The assurance creates the immunities needed for the cooperation of a transnational security community, to which the courts effectively render control. Whatever human rights considerations might inform the work of secret services in international cooperation, they will not be subject to public scrutiny.
Diplomatic assurances revolve around a double secret. Within each, a source of terror is concealed. One is the secret that the state seeks to extract from the captive, whilst the other is that of the interrogating state; it conceals that which makes the captive speak. The first is the source of the state’s fear, the second is the source of the captive’s fear. The second is negatively circumscribed by the assurance: whatever the captive fears, it cannot be revealed due to security imperatives. At any rate, it is not to be articulated as a human rights violation.
This circumscription of the second source due to security imperatives is, again, a double one, applying both to the receiving state and the captive. The state shrouds the source in the veil of silence; making its interrogation methods known might give away benefits to its enemies. But, tragically, the captive will also conceal the source of his fear from his potential saviour — the diplomat who visits him. Maher Arar, the Canadian citizen whom the US rendered to Syria, recalls the visits of the Canadian consul after two weeks of detention, interrogation and torture:
I was taken from my cell and my beard was shaved. I was taken to another building, and there was the colonel in the hallway with some other men and they all seemed very nervous and agitated.
I did not know what was happening and they would not tell me. … I was told not to tell anything about the beating, then I was taken into a room for a ten-minute meeting with the consul. The colonel was there, and three other Syrian officials including an interpreter. I cried a lot at that meeting. I could not say anything about the torture. I thought if I did, I would not get any more visits, or I might be beaten again.
The consular visits were my lifeline, but I also found them very frustrating. There were seven consular visits, and one visit from members of Parliament. After the visits I would bang my head and my fist on the wall in frustration. I needed the visits, but I could not say anything there.
The reason Arar kept silent was simply that he feared for his own security; he would be punished either by further torture, or by the cancellation of further visits. Further, if the visits stopped, might he not be killed without anyone taking notice, his tormentors exempted? Visits by diplomats are simply unsuitable to effectively detect or avert violent interrogations of rendered captives. Arar did not meet the consul alone, and indeed, the 1963 Vienna Convention on Consular Relations does not bestow a right to meet a diplomatic representative without the presence of officials from the receiving state. His account gives a graphic reflection of his inability to speak. Or, to be precise, Arar cannot name the human rights violation inflicted upon him. This ensures again that whatever it is that has happened and that will happen is not to be articulated as a human rights violation. The receiving state and the captive share the same bind of security imperatives. The captive becomes an accomplice in concealing torture.
Requests for renditions are directed through embassies — a process that assigns a bundle of communicative functions to diplomats, and burdens them with multiple loyalties. One loyalty is owed to the captive, which suggests faithfulness to the rights of the individual. Another is owed to the sending state. This loyalty relation is complicated, as the interests of different bureaucracies involved in renditions are not necessarily the same. Obviously, a secret service and a consular protection unit may have different priorities. In addition, there is a functional imperative to maintain good relations with the receiving state and its authorities. The diplomat is as much a traitor as a saviour in this regard.
The case of Maher Arar provides a graphic illustration. Two weeks after Arar had been rendered from the US to Syria for interrogation purposes, Arar was visited in detention by the Canadian Ambassador to Syria. While the Ambassador saw it as his task to secure Arar’s release, he also asked Syria to relay any evidence it had gained against him. That is, the Ambassador acted as an emissary of human rights protection whilst passing on information derived from the violation of those rights. In this case, ties to the receiving country and its state apparatus apparently trumped the protection owed to a fellow citizen. When confronted with conflicting information about the length of Arar’s detention, the Ambassador chose to believe the head of the Syrian intelligence service rather than Arar. There is a fatal logic to this choice. As the Canadian security service had taken an interest in the outcome of Arar’s interrogation, the ambassador will assume that this renders Arar a potential enemy. And why should the enemy tell the truth? From this, it follows that the security service of a receiving country is a better source of information than the captive.
The diplomat thus becomes the guardian of a double secret. First, there is the secret of that which the state fears. The diplomat is the potential transmitter of security information leading to the capture and rendition of the captive. The diplomat might also be used to pass on information obtained from the captive in interrogations by the receiving country. Second, there is the secret of that which the captive fears. The diplomat is the potential transmitter of the fact that the captive has been tortured. To acknowledge torture, however, would mean that the diplomat acknowledges a failure. Obviously, shifting the responsibility from the court to the diplomat was unreasonable. The trust vested in the diplomat’s monitoring would then have been misplaced.
Earlier parts of this article have established that the Swedish–Egyptian diplomatic assurances create ambiguity as to the content of human rights law as well as to their own validity. From a structural perspective, this ambiguity can be perceived as being quite productive. After all, who is assured by a diplomatic assurance? Each of the two States assure themselves, their institutions and their populations, rather than each other, or an undefined omnium. Each State assures itself that it is neither violating human rights nor altering international human rights law.
Diplomatic assurances constitute a double promise. One is the universal promise of human rights enshrined in multilateral human rights treaties such as the ICCPR and the CAT. The second is a particularised form of human rights promise given through the diplomatic assurance itself. The logic of universal human rights would suggest that the latter promise is infected by the breach of the former.
Why should we rely on a diplomatic assurance when it was made necessary by the breach of multilateral human rights obligations in the first place? The logical answer is that a diplomatic assurance promises something other than a human rights treaty, which saves it from being tainted by the dismal human rights track record of the receiving country. A multilateral human rights treaty such as the ICCPR promises human rights within a self-contained setting: the triad of state, territory and population. By contrast, a diplomatic assurance promises that the rendition of the subjected person is without prejudice to the human rights promise of the multilateral treaties. Viewed in this light, it is understandable that the norms of human rights law as well as the norms of the diplomatic assurance remain silent when we test them in their application to actual renditions.
It is frequently argued that terrorism will prevail as soon as human rights are compromised in the struggle against it. The ambiguity of such statements should give us pause; apart from its common sense, it can also be understood as an imperative to protect the concept of human rights from being tainted by the violence unleashed in the ‘war on terror’. With regard to diplomatic assurances, the same ambiguity is at work. Rather than assuring the captive of protection against harm, it may be that human rights as such are merely assuring themselves. Thereby, the rendering community may feel that the conception of human rights remains unharmed by the struggle that unfolds itself during the rendition process.
For good reasons, the debate on extraordinary renditions and diplomatic assurances has focused on the ‘outsourcing of torture’, that is, the question of the captives’ treatment in the receiving country under international human rights law. However,
according to US intelligence officials, captured terrorists were rendered over to coalition partners for interrogation, not so much for their coercive techniques as for the cultural affinities that enable them to reach out and induce, or goad, the captives into talking.
Taking this argument at face value (and leaving aside the reduction of judicial monitoring for a moment), it would imply that rendition is about moving the captives to an ‘appropriate’ cultural context, where a ‘proper’ interrogation may take place. It would further confirm that the coalition’s security services lack the capability to conduct such interrogations in a satisfactory manner, for whatever reasons. Alternatively, an essentialist edge can be added, pushing the limits of the argument: multicultural societies have failed when captives need to be rendered to an ‘originary’ setting in Egypt or Syria.
At the core of the renditions and assurances accompanying them is a simple exchange between what we could call ‘cultures of human rights’. As truth is contingent on culture, the rendering state removes the captive to her or his ‘true’ culture. In return, the state representing that culture assures the rendering culture that whatever it does, it will not torture or otherwise violate the captive’s human rights. Both interrogation scenarios rest on cultural relativism and render human rights to contingency. But the corporal dimension is disturbing. Is the body of the captive tortured so as to reveal where it ‘truly’ belongs, the community of which it ‘really’ forms a part?
In this exchange, human rights are repoliticised. This signifies more than the loss of the judiciary equating to the gain of a transnational security community. There is a Schmittian concern too. Schmitt cast the essence of the political as resting on the distinction between friend and enemy. When placing the treatment of a rendered captive beyond judicial control, states are actually no longer assuring the judiciary of a certain conduct in the sense that rights-respecting conduct will be ‘likely’. Rather, they assure their public that whatever is done to the captive in the future, it does not constitute a violation of human rights. Seen as such, diplomatic assurances posit a legal definition. In doing so, they cast the captive as an absolute enemy in the Schmittian sense. Herein, we encounter the move from inimicus to hostis; from a private to a political enemy.
The combination of diplomatic assurance and rendition effectively abandons its subject in Nancy’s sense: ‘To abandon’, Nancy writes, ‘is to remit, entrust, or turn over to such a sovereign power, and to remit, entrust, or turn over to its ban, that is, to its proclaiming, to its convening, and to its sentencing.’ Or, as Mills explains Agamben’s appropriation of the term: ‘To be abandoned means to be subjected to the unremitting force of the law while the law simultaneously withdraws from its subject.’ The absence of any distinguishable content in the analysis of treaty law is a product of this ongoing double move, of the law being ‘valid precisely insofar as it commands nothing and has become unrealizable’. Diplomatic assurances perform the movement of withdrawal through their self-referentiality: human rights are assured not to the captive, but to the rendering community. To quote Nancy again, the ‘[a]bandoned being finds itself deserted to the degree that it finds itself remitted, entrusted, or thrown to this law’.
There should be no misunderstanding: renditions and assurances are no regression to a ‘state of nature’, nor a reaching back to a prelegal stage. The secrecy of renditions notwithstanding, human rights law has perhaps never been operated so vigorously and transparently. After all, the sovereign is back at the negotiating table, drawing up specific treaties defining the human rights of a single rendered person. Where states are seen to be dealing with human rights bilaterally and individually, largely undisturbed by the critique of advocates, NGOs and monitoring bodies, it must be apparent that there has been no emancipation of human rights from state authority. Their occult element is rendered transparent, and thereby loses its persuasive power. If anything, we are assured that human rights are a matter of positive international law. Forced to observe this, I will conclude that the authority of human rights is reduced to the authority of the state, and can be construed as its derivative. Yet it is precisely this vigour of the lawmaker and the black-letter of the aide-mémoires which voids the law of content, silences human rights, and therewith condemns the captive to abandonment.
[*] Associate Professor of International Law, Faculty of Law, Lund University, Sweden. I would like to thank Karen Engle, Jan Klabbers, Karen Knop, Ulf Linderfalk and Anne Orford, as well as the participants of the Second Melbourne Legal Theory Workshop on the Cultures of Human Rights for their helpful comments on earlier versions of this text. I am much indebted to Linda Engvall for her research assistance. All errors remain my own.
 Academic writing on the subject is still sparse. See, eg, Julia Hall, ‘The Treatment of Asylum Seekers and Migrants in the Context of the Global “War on Terror”’ (2004) American Society of International Law: Proceedings of the 98th Annual Meeting 258. Hall states that ‘diplomatic assurances and post-return monitoring are inadequate safeguards against torture’: at 260. For an earlier and generally critical view of the United States’ policy on non-refoulement, see William Aceves and Paul Hoffman, ‘Using Immigration Law to Protect Human Rights: A Critique of Recent Legislative Proposals’ (2002) 23 Michigan Journal of International Law 733. See also Valerie Neal, ‘Slings and Arrows of Outrageous Fortune: The Deportation of “Aggravated Felons”’ (2003) 36 Vanderbilt Journal of Transnational Law 1619, arguing that aggravated felons may be returned under US law to their country of origin after diplomatic assurances and that before someone can be returned, there must be a perceived absence of risk of contravening art 3 of the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature 10 December 1984, 1465 UNTS 85 (entered into force 26 June 1987) (‘CAT’).
 Neither ‘rendition’ nor ‘extraordinary rendition’ is a legal term of art. Extraordinary rendition has been defined as the ‘transfer of an individual, with the involvement of the US or its agents, to a foreign state in circumstances that make it more likely than not that the individual will be subjected to torture or cruel, inhuman, or degrading treatment’: Association of the Bar of the City of New York and Center for Human Rights and Global Justice, Torture by Proxy: International and Domestic Law Applicable to ‘Extraordinary Renditions’ (2004) 4. The report offers a well documented introduction to the topic, mainly from a US perspective. For an overview of the US policy context and the ensuing debate on renditions, see Edward Greer, ‘“We Don’t Torture People in America”: Coercive Interrogation in the Global Village’ (2004) 26 New Political Science 371.
 According to Michael Scheuer, the former CIA employee heading the unit responsible for tracking Osama bin Laden, President Clinton and his successor President Bush preferred to have captives rendered to third countries, because rendering them to the US would have turned them into prisoners of war and augmented their legitimacy. Scheuer states that when the CIA raised the question of human rights protection (in the target countries of rendition) with the Clinton Administration, the latter was not interested in details about the treatment of captives: Thomas Kleine-Brockhoff, ‘Die CIA hat das Recht, jedes Gesetz zu brechen’, Die Zeit (Berlin, Germany), 29 December 2005, 10.
 Typically, the human rights community remains sceptical of diplomatic assurances and calls for ‘safeguards’ in an attempt at damage control. Human Rights Watch took a rather restrictive position in 2005, advising governments to prohibit reliance on diplomatic assurances ‘in any case where there is a risk of torture or prohibited ill-treatment upon return directly related to a person’s particular circumstances’: Human Rights Watch, ‘Still at Risk: Diplomatic Assurances No Safeguard against Torture’ (2005) 17(4D) Human Rights Watch 1, 80. A similar position was taken by the Commissioner for Human Rights of the Council of Europe in Report by Mr Alvaro Gil-Robles, Commissioner for Human Rights, on His Visit to Sweden, 21–23 April 2004  13 CommDH:
The weakness inherent in the practice of diplomatic assurances lies in the fact that where there is a need for such assurances, there is clearly an acknowledged risk of torture and ill-treatment. Due to the absolute nature of the prohibition of torture or inhuman or degrading treatment, formal assurances cannot suffice where a risk nonetheless remains. As the UN Special Rapporteur on Torture has noted, such assurances must be unequivocal and a system to monitor such assurances must be in place. When assessing the reliability of diplomatic assurances, an essential criteria [sic] must be that the receiving state does not practice or condone torture or ill-treatment, and that it exercises effective control over the acts of non-state agents. In all other circumstances it is highly questionable whether assurances can be regarded as providing indisputable safeguards against torture and ill-treatment: at .
Let me try to contribute to this debate by offering three simple essential elements which could make rendition less risky (and simultaneously less attractive): 1) Regular monitoring meetings of rendered persons with independent legal and medical experts at neutral premises; 2) An obligation to publish expert reports; and 3) An obligation to return the rendered person to the rendering country in instances where visits reveal indications of human rights violations.
 It is worth noting that the Committee against Torture has regularly requested further information or has voiced concern about the use of diplomatic assurances: Committee against Torture, Conclusions and Recommendations of the Committee against Torture: United Kingdom of Great Britain and Northern Ireland, Crown Dependencies and Overseas Territories, UN Doc CAT/C/CR/33/3 (25 November 2004) [4d] (raising concern regarding the use of diplomatic assurances by the United Kingdom), [5i] (recommending the provision of information to the Committee regarding the number of extraditions that have occurred after 11 September 2001, the content of such assurances and the monitoring measures adopted in these cases); Committee against Torture, Conclusions and Recommendations of the Committee against Torture: Canada, UN Doc CAT/C/CR/34/CAN (7 July 2005) [4b] (expressing concern regarding Canada’s role in the expulsion of Maher Arar), [5e] (recommending the provision of information regarding: the use of diplomatic assurances to the Committee, especially the number of extraditions that have occurred post 11 September 2001; the minimum requirements for these assurances; the measures currently used to facilitate monitoring and the legal enforceability of these assurances); Committee against Torture, Conclusions and Recommendations of the Committee against Torture: Switzerland, UN Doc CAT/C/CR/34/CHE (21 June 2005) [5j] (recommending provision of information to the Committee of any complaints regarding the use of diplomatic assurances); Committee against Torture, Conclusions and Recommendations of the Committee against Torture: Germany, UN Doc CAT/C/CR/32/7 (11 June 2004) [D5(e)] (recommending that Germany provide information regarding the use of diplomatic assurances); Committee against Torture, Summary Record of the First Part (Public) of the 607th Meeting: New Zealand, UN Doc CAT/C/SR.607 (18 May 2004) ,  (New Zealand stating that it does not use diplomatic assurances).
 Former CIA Director George Tenet has described extraordinary renditions as one of the ‘principal strategies employed against the threat of … terrorism’: Kareem Fahim, ‘The Invisible Men: Canadian Inquiry May Reveal CIA Secrets on Outsourcing Torture’ (2004) 49(13) The Village Voice 37. In Tenet’s 37-page memorandum to the 9-11 Commission, the phenomenon of renditions was mentioned on five occasions in a manner confirming their strategic importance: George Tenet, Written Statement for the Record of the Director of Central Intelligence before the National Commission on Terrorist Attacks upon the United States (2004) <http://www.9-11commission.gov/hearings/hearing8/tenet_statement.pdf> at 22 May 2006. Michael Scheuer has confirmed his involvement in approximately 40 renditions between 1995 and 1999: see Kleine-Brockhoff, above n 3.
 For an excellent overview of the program’s history, see Jane Mayer, ‘Outsourcing Torture: The Secret History of America’s “Extraordinary Rendition” Program’, The New Yorker (New York, US), 14 February 2005, available at <http://www.newyorker.com/printables/fact/050214fa_fact6> at 22 May 2006. See also Stephen Grey, ‘United States: Trade in Torture’, Le Monde diplomatique (English Edition) (Paris, France), April 2005, available at <http://mondediplo.com/2005/04/04usatorture> at 22 May 2006.
 See, eg, the call for a moratorium on the use of diplomatic assurances by the state officials in the report of the Association of the Bar of the City of New York and Center for Human Rights and Global Justice: above n 2, 8.
 Certainly, there are other interesting cases, yet those of Agiza and El Zary offer a unique degree of accessibility due to the media attention and to the monitoring of the cases by the Committee against Torture. The Durmo case offers parallels, however there is no information on any diplomatic assurances exchanged in the context of rendition: Eslam Durmo v Bosnia and Herzegovina and the Federation of Bosnia and Herzegovina (Admissibility and Merits), Human Rights Chamber for Bosnia and Herzegovina, Case No CH/02/9842 (10 January 2003). At the behest of the US and Egypt, Bosnia-Herzegovina rendered its citizen Eslam Durmo to Egypt on 6 October 2001. The Federation of Bosnia-Herzegovina was held to be responsible for violating, inter alia, art 3 of the European Convention for the Protection of Human Rights and Fundamental Freedoms, opened for signature 4 November 1950, 213 UNTS 221 (entered into force 3 September 1953) as well as art 3(1) of the European Convention for the Protection of Human Rights and Fundamental Freedoms Additional Protocol No 4, opened for signature 16 September 1963 (entered into force 2 May 1968).
 Agiza v Sweden, Committee against Torture, Communication No 233/2003, UN Doc CAT/C/34/D/233/2003 (20 May 2005) [2.4]–[2.5].
 Kingdom of Sweden, Aide-Mémoire to the Arab Republic of Egypt (12 December 2001) (copy on file with author); Arab Republic of Egypt, Aide-Mémoire to the Kingdom of Sweden (12 December 2001) (copy on file with author).
 Kingdom of Sweden, above n 11.
 Arab Republic of Egypt, above n 11. The omission in the quote contains five lines of classified text.
 Barbro Holmberg, Swedish Minister for Migration, ‘Anförande vid riksdagens särskilda debatt om regeringens agerande och omständigheterna kring avvisningen av Ahmed Agiza och Mohammed El Zari’ (Speech delivered to the Swedish Parliament, Stockholm, Sweden, 26 January 2005) (author’s own translation).
 Agiza v Sweden, Committee against Torture, Communication No 233/2003, UN Doc CAT/C/34/D/233/2003 (20 May 2005) [13.8]–[13.15].
 Holmberg, above n 14 (author’s own translation).
 Above n 9, arts 3, 6. Article 3 pertains to the prohibition of torture or inhuman or degrading treatment or punishment; art 6 to the right to a fair trial. See also Protocol No 6 to the Convention for the Protection of Human Rights and Fundamental Freedoms concerning the Abolition of the Death Penalty, opened for signature 28 April 1983, ETS 114, art 1 (entered into force 1 March 1985). Article 1 of the Protocol pertains to the abolition of the death penalty. According to the case law of the European Court of Human Rights, all three articles may unfold to apply extraterritorially, that is, states’ freedom to remove persons to third states where such risks might materialise is limited by the said articles. For an overview of case law, see Gregor Noll, Negotiating Asylum: The EU Acquis, Extraterritorial Protection and the Common Market of Deflection (2000) 454–8.
 Arab Republic of Egypt, above n 11.
 In a statement made on behalf of the Nordic Countries to the UN General Assembly Sixth Committee, Sweden has claimed that there is ‘persistent State practice from the Nordic countries and an increasing number of other States making use of such objections in relation to invalid reservations to human right [sic] treaties’: ‘Elements of Nordic Practice 2003: The Nordic Countries in Co-Ordination’ (2005) 74 Nordic Journal of International Law 131, 149.
 Kingdom of Sweden, above n 11.
 Arab Republic of Egypt, above n 11.
 Constitution of the Arab Republic of Egypt (1970) (‘Egyptian Constitution’).
 Report of the Committee against Torture, UN GAOR, 44th sess, Supp 46, UN Doc A/44/46 (8 June 1989) . Article 1 of the CAT is as follows:
For the purposes of this Convention, the term ‘torture’ means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in or incidental to lawful sanctions.
 Report of the Committee against Torture: 29th Session (11–12 November 2002) and 30th Session (28 April – 16 May 2003), UN GAOR, 58th sess, Supp 44, UN Doc A/58/44 (2003) [42b]. Since the consideration of Egypt’s first periodic report in 1989, the Committee against Torture has repeatedly made this recommendation.
 Egyptian Constitution art 42. The remaining sentences in art 42 read:
He may not be detained or imprisoned except in places defined by laws organising prisons. If a confession is provided to have been made by a person under any of the aforementioned forms of duress or coercion, it shall be considered invalid or futile.
 See Human Rights Watch, Egypt’s Torture Epidemic (2004) <http://hrw.org/english/docs/2004/02/25/egypt7658_txt.htm> at 22 May 2006.
 International Covenant on Civil and Political Rights, opened for signature 16 December 1966, 999 UNTS 171 (entered into force 23 March 1976) (‘ICCPR’).
 See, eg, the introductory chapter in Jan Klabbers, The Concept of Treaty in International Law (1996) 1–14.
 Anthony Aust, Modern Treaty Law and Practice (2000) 20–2, 41–4. I shall test my argument against Aust’s demands, given that the question of a non-binding MOU never arises under Klabbers’ argument.
 Opened for signature 23 May 1969, 1155 UNTS 331 (entered into force 27 January 1980) (‘VCLT’).
 Sir Arthur Watts, The International Law Commission 1949–1998: Volume Two: The Treaties Part II (1999) 623.
 The term ‘understanding’ is employed on several occasions. The documents also use the verb ‘will’ rather than ‘shall’. If we follow Aust, both choices — ‘understanding’ and ‘will’ — indicate that the agreement was perhaps limited to the political sphere only: Aust, above n 31, 27. However, Aust acknowledges that it is ultimately the intention of the party that governs the nature of an MOU or an exchange of notes: at 46.
 Above n 14.
 Agiza v Sweden, Committee against Torture, Communication No 233/2003, UN Doc CAT/C/34/D/233/2003 (20 May 2005) [4.22].
 Once it has entered into force, the Optional Protocol to the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature 4 February 2003, 42 ILM 26 (2003) (not yet in force) will put into place a system of regular visits to places of detention to strengthen the protection of detained persons against torture or inhuman or degrading treatment and punishment. One could interpret the Swedish–Egyptian agreement to monitor visits by the Swedish Ambassador as a precursor to such a future system. What makes such bilateral monitoring stand out is its focus on individual cases and its ability to add diplomatic ‘punch’ by packaging human rights issues with other bilateral relations. What makes it vulnerable is precisely the flip side of the coin: the barter trade in which unrelated bilateral matters such as trade policy may be pursued at the expense of human rights protections. As explained elsewhere in this text, the diplomat is a problematic protector, given that she or he is untrained in detecting signs of torture and acts within multiple loyalties.
 Above n 32, art 2(1)(a).
 In a recent case concerning the extradition of UK citizen Babar Ahmad from the UK to the US, a UK court at first instance considered that an unsigned assurance submitted to it by the US Embassy would bind the UK Government: The Government of the United States of America v Babar Ahmad  (Unreported, Bow Street Magistrates’ Court, Judge Workman, 17 May 2005). By way of contrast to the guarantee in the cases of Agiza and El Zary, this assurance was not part of an exchange of notes, but a unilateral declaration.
 In limiting the problem to a conflict of norms, rather than a conflict of treaties, I follow Pauwelyn’s approach: Joost Pauwelyn, Conflict of Norms in Public International Law: How WTO Law Relates to Other Rules of International Law (2003) 6–10.
 I neither need nor want to answer the question of what the correct interpretation of the definition of torture should be according to the exchanged aide-mémoires.
 Pauwelyn, above n 41, 11. In addition, both the ICCPR and the aide-mémoires address the fairness of trial, creating another overlap. A similar point could be raised with regard to monitoring, where the exchange of aide-mémoires, the ICCPR and the CAT provide different institutional solutions. These overlaps potentially host conflicts as well, but the issue of torture suffices to substantiate my argument.
 Manfred Zuleeg, ‘Vertragskonkurrenz im Völkerrecht. Teil I: Verträge zwischen souveränen Staaten’, German Yearbook of International Law (1977) 246, 267–8.
 See Office of the High Commissioner for Human Rights, Status of Ratifications of the Principal International Human Rights Treaties (2006) <http://www.ohchr.org/english/bodies/docs/RatificationStatus.pdf> at 22 May 2006.
 The pertinent provisions on torture in both instruments are cast as non-derogable norms, which gives a technical indication of their close relationship to the ‘object and purpose’ of the treaties: see ICCPR, above 29, art 7; CAT, above n 1, art 3.
 See Aust, above n 31, 222.
 Jan Klabbers, ‘Some Problems Regarding the Object and Purpose of Treaties’ (1991) 8 Finnish Yearbook of International Law 138, 151–2.
 VCLT, above n 32, art 41(2).
 See ibid art 30(5), stating that art 30(4) of the VCLT is without prejudice to art 41 of the VCLT.
 Ibid art 30(4) reads:
When the parties to the later treaty do not include all the parties to the earlier one:
(a) As between States parties to both treaties the same rule applies as in paragraph 3;
(b) As between a State party to both treaties and a State party to only one of the treaties, the treaty to which both States are parties governs their mutual rights and obligations.
Paragraph 3 contains a lex posterior rule.
 The term is taken from the context of European Union integration and describes a system whereby different levels of obligation and integration coexist within one and the same integrationist framework.
 For an overview of arguments supporting exemption made by successive Special Rapporteurs Lauterpacht and Fitzmaurice as well as other doctrinal writers, see Jan Mus, ‘Conflicts between Treaties in International Law’ (1998) 45 Netherlands International Law Review 208, 222–4.
 VCLT, above n 32, art 53 reads:
A treaty is void if, at the time of its conclusion, it conflicts with a peremptory norm of general international law. For the purposes of the present Convention, a peremptory norm of general international law is a norm accepted and recognized by the international community of States as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character.
 In the case of Furundžija, the Trial Chamber of the International Criminal Tribunal for the Former Yugoslavia stated that the absolute prohibition of torture decreed by international law had the character of a jus cogens norm, imposing obligations erga omnes on states: Prosecutor v Furundžija, Case No IT-95-17/1-T (10 December 1998) –. For a current take at the inventory of jus cogens norms with further references, see Andreas Paulus, ‘Jus Cogens in a Time of Hegemony and Fragmentation — An Attempt at Re-Appraisal’ 74 Nordic Journal of International Law 297, 305–8.
 If art 1 of the CAT were to narrow down the jus cogens definition of torture, the former norm would be void, which would effectively do away with the CAT. The malleability of the torture concept produces an acute tension with the need to define precise limits to jus cogens.
 Above n 32.
 This cultural relativism seems to be well received with CIA operatives. As former CIA Director Michael Scheuer commented: ‘Human rights is a very flexible concept. It kind of depends on how hypocritical you want to be on a particular day’: Grey, above n 7.
 Egypt apparently does not believe that the behaviour of its officials as reported by Agiza to the Swedish Ambassador constitutes torture or inhuman or degrading treatment, whereas the Committee against Torture necessarily must have regarded that behaviour as constitutive of torture (otherwise, it might not have determined that Sweden was in breach of art 3 of the CAT when removing Mr Agiza to Egypt). Then again, the Committee against Torture does not possess the power of a court, and can hardly overrule Egypt’s interpretation in any formal way.
 The dissenting opinion in the Agiza case before the Committee against Torture pinpoints the importance of choosing an appropriate point in time for judging whether or not a state has complied with art 3 of the CAT: Agiza v Sweden, Committee against Torture, Communication No 233/2003, UN Doc CAT/C/34/D/233/2003 (20 May 2005) 38.
 The case of Ramzy currently before the European Court of Human Rights sheds some light on the way in which the transnational security community produces rendition cases by denying courts the information to convict. Algerian citizen Mohammed Ramzy was acquitted of terrorism charges by the Dutch courts. The Court did not allow certain intelligence reports to be used as evidence by the prosecution, and access had accordingly been denied to the defence, as officials of the Dutch intelligence service wished to maintain the secrecy of the reports. Ramzy’s asylum claim was turned down and his removal was prepared in cooperation with Algerian authorities. See Registrar, European Court of Human Rights, ‘Application Lodged with the Court: Ramzy v The Netherlands’ (Press Release, 20 October 2005) <http://www.echr.coe.int/Eng/Press/2005/Oct/ApplicationlodgedRamzyvNetherlands.htm> at 22 May 2006. On Dutch–Algerian cooperation in preparation for Ramzy’s removal and readmission, see Ramzy v The Netherlands, Application no 25424/05 (Unreported, European Court of Human Rights, 15 July 2005) [14.78] (copy on file with author).
 In the Agiza case, it became clear that the setup of diplomatic assurances also invites the rendering state to protect the secret of that which makes the captive speak. The Swedish ambassador visiting Agiza during his detention in Egypt related information on mistreatment and threats in one of his reports. This information was initially withheld from the Committee against Torture: Agiza v Sweden, Committee against Torture, Communication No 233/2003, UN Doc CAT/C/34/D/233/2003 (20 May 2005) [12.34]. The Committee against Torture concluded that this constituted a breach of Sweden’s obligations under art 22 of the CAT.
 The term ‘torture’ is used in the sense of art 1 of the CAT here.
 Maher Arar, ‘Statement’ in Rachel Meeropol (ed), America’s Disappeared: Secret Imprisonment, Detainees, and the ‘War on Terror’ (2005) 68–9.
 Vienna Convention on Consular Relations, opened for signature 24 April 1963, 596 UNTS 261 (entered into force 19 March 1967).
 Ibid art 36(1)(c), which reads:
1. With a view to facilitating the exercise of consular functions relating to nationals of the sending State:
(c) consular officers shall have the right to visit a national of the sending State who is in prison, custody or detention, to converse and correspond with him and to arrange for his legal representation. They shall also have the right to visit any national of the sending State who is in prison, custody or detention in their district in pursuance of a judgment. Nevertheless, consular officers shall refrain from taking action on behalf of a national who is in prison, custody or detention if he expressly opposes such action.
 For full documentation of the case, see Arar Commission, Maher Arar Commission of Inquiry (2004) <http://www.ararcommission.ca/eng/index.htm> at 22 May 2006. For the report of the Factual Inquiry, see Stephen Toope, Commission of Inquiry into the Actions of Canadian Officials in Relation to Maher Arar, Report of Professor Stephen J Toope, Fact Finder (14 October 2005), available at <http://www.ararcommission.ca/eng/17.htm> at 22 May 2006.
 CBC News, Ambassador Believed Syrian General over Arar (2005) <http://www.cbc.ca/story/canada/national/2005/06/14/arar-pillarella050614.html> at 22 May 2006.
 CBC News reported that:
Pillarella [Canadian Ambassador to Syria at the relevant time] has been accused of playing conflicting roles as Canada’s ambassador to Syria. He was trying to win Arar’s release while at the same time asking Syria to pass over any evidence it gained against him. Among documents that have been made public as a result of the inquiry was a memo Pillarella wrote to his superiors in Ottawa. He said a Syrian contact had told him Arar was being interrogated and had confessed to links with terrorist organizations, alluding to groups based in Pakistan. Pillarella said the Syrians ‘promised to pass on to me any information they may gather on Arar’s implication in terrorist activities’: ibid.
 In the Arar inquiry it was found that the Canadian authorities received information imparted by Arar during interrogations in Syria: Commission of Inquiry into the Actions by Canadian Officials in Relation to Maher Arar, Summary of Information Received at In Camera Hearings (20 December 2004) , , available at <http://www.ararcommission.ca/eng/16.htm> at 22 May 2006.
 See, eg, the last Foreign Policy Declaration from the office of Anna Lindh, Swedish Minister of Foreign Affairs, who was killed on 11 September 2003, as quoted by her successor:
International terrorism is a threat to our rights. Intolerance based on a black-and-white view of the world is its foundation, financing through criminality and illegal money transactions its precondition, ruthlessness and terror its tools. No goal, no cause, no struggle can justify terror aiming to kill and maim innocent civilians. Sweden will always be in the vanguard when it comes to fighting terrorism. But if we let the fight against terrorism take priority over human rights, we will be the losers and the terrorists will triumph. Human rights are at the heart of the tolerant lifestyle we aspire to and they oppose.
Laila Freivalds, Minister for Foreign Affairs, Sweden, ‘Security under the Rule of Law’ (Speech delivered at the Swedish Foreign Ministry, Stockholm, Sweden, 1 December 2004).
 Evidence of access to different interrogation methods, including torture, and of the capacity to make persons disappear, played a role in framing US rendition policies. Former CIA agent Robert Baer recalled from his experiences until the mid-1990s that Egypt was suitable for disappearances, Jordan for serious interrogations, and Syria the preferred location for torture. See, eg, Steven Macpherson Watt, ‘Torture, “Stress and Duress” and Rendition as Counter-Terrorism Tools’ in Rachel Meeropol (ed), America’s Disappeared: Secret Imprisonment, Detainees, and the ‘War on Terror’ (2005) 82.
 Martin Rudner, ‘Hunters and Gatherers: The Intelligence Coalition against Islamic Terrorism’ (2004) 17 International Journal of Intelligence and Counter Intelligence 193, 220.
 However, it appears that US interrogators may have been present at extraterritorial locations during the interrogations in the Arar case: Arar, above n 64, 65–7. This would not exactly maximise ‘cultural affinities’ in the interrogation situation. According to Watt, US agents were also personally interrogating three UK nationals and one UK resident with refugee status, detained in Gambia in 2002. At least one was threatened to be handed over to Gambian agents, who would beat and rape him: Watt, above n 72, 84–5.
 Carl Schmitt, Der Begriff des Politischen: Text von 1932 mit einem Vorwort und drei Corollarien (1963) 29.
 ‘Feind ist nur eine wenigstens eventuell, d.h. der realen Möglichkeit nach kämpfende Gesamtheit von Menschen, die einer ebensolchen Gesamtheit gegenübersteht. Feind ist nur der öffentliche Feind, weil alles, was auf eine solche Gesamtheit von Menschen, insbesondere auf ein ganzes Volk Bezug hat, dadurch öffentlich wird. Feind ist hostis, nicht inimicus im weiteren Sinne’ (‘An enemy only exists when there is a potentially — ie when there is the real possibility of a — fighting collectivity of people confronting another such collectivity. The enemy is only a public enemy, because everything that relates to such a collectivity of people, especially an entire people, becomes public. The enemy is hostis, not inimucus, in the broader sense’): ibid 29 (emphasis in original) (author’s own translation).
 Jean-Luc Nancy, ‘Abandoned Being’ (Brian Holmes trans, 1993 ed) in Werner Hamacher and David Wellbery (eds), The Birth to Presence (1993) 44 (emphasis in original). For an introduction to the significance of Nancy’s concept of abandonment, see Benjamin Pryor, ‘Law in Abandon: Jean-Luc Nancy and the Critical Study of Law’ (2004) 15 Law and Critique 259.
 Catherine Mills, ‘Agamben’s Messianic Politics: Biopolitics, Abandonment and Happy Life’ (2004) 5 Contretemps 42, 45.
 Giorgio Agamben, Potentialities: Collected Essays in Philosophy (1999) 172.
 Nancy, above n 77, 44.
 Pryor describes the ‘historical a priori’ as ‘a site of abandonment’: see Pryor, above n 77, 266.