The Draft Article 49 Opinion

After Jay Bybee’s departure, Jack Goldsmith, a distinguished University of Chicago law professor (now a Harvard law professor), took over the lea­dership of OLC. Goldsmith took several courageous stands against Administration hard-liners, stands for which he reportedly had to withstand the fury of David Addington, Cheney’s volcanic general counsel, regarded by many as the hardest of hard-liners.^[359] As early as December 2003, before the Abu Ghraib scandal and the leak of the Bybee Memo, Goldsmith advised the government not to rely on a March 2003 memo by John Yoo that had directly influenced the Defense Department’s working group on interrogation.^[360] And it was under Goldsmith’s leadership that OLC repudiated the Bybee Memo.

Nevertheless, Goldsmith too drafted a memorandum that exemplifies the kind of loophole legalism I object to in the other memoranda. (Let me emphasize, however, that Goldsmith’s draft was never given final approval, and that could indicate that Goldsmith thought better of it.) Written in March 2004, it concerned the question of whether detainees in Iraq could be tem­porarily sent out of the country for interrogation, despite plain language in Article 49 of the Fourth Geneva Convention stating:

Individual or mass forcible transfers, as well as deportations of protected persons from occupied territory to the territory of the Occupying Power or to that of any other country, occupied or not, are prohibited, regardless of their motive.⁸⁴

Goldsmith divided the memo into two sections, one on whether Article 49 would prevent US authorities from deporting illegal aliens in Iraq “pursuant to local immigration law,” and one on whether removing protected civilians from Iraq for interrogation violates Article 49.

In answer to the first question, Goldsmith contends that the drafters of Article 49 could not have meant to ban the removal of illegal aliens under an occupied state’s immigration law.

com/2005/09/silver-linings-or-strange-but-true.html>, which also provides a useful chronol­ogy and analysis. The March 14, 2003 Yoo memorandum has not been released or leaked. Levin’s letter mentions that twenty-four interrogation techniques are still approved; the implication is that the Yoo memorandum okayed techniques that OLC no longer approves.

⁸⁴ The Washington Post reports that Goldsmith had written an opinion five months earlier concluding that a ghost detainee named Rashul could not be removed from Iraq. By that time the CIA had already spirited Rashul away to Afghanistan, and after Goldsmith’s opinion they quickly returned him to Iraq. According to an intelligence source, “That case started the CIA yammering to Justice to get a better memo.” Dana Priest, Memo Lets CIA Take Detainees Out of Iraq, Wash. Post, October 24, 2004, A1, A21. However, Professor Goldsmith has informed me that this account is seriously defective: there was no previous memo on the topic, and he did not give in to any pressure. (Private e-mail communications, August 27 and 29, 2006.) The CIA’s deputy inspector general “told others she was offended that the CIA’s general counsel had worked to secure a secret Justice Department opinion in 2004 authorizing the agency’s creation of ‘ghost detainees’ - prisoners removed from Iraq for secret interrogations without notice to the International Committee of the Red Cross - because the Geneva Conventions prohibit such practices.” R. Jeffrey Smith, Fired Officer Believed CIA Lied to Congress, Wash. Post, May 14, 2006. Priest’s article states that even though the draft was never released, the CIA relied on it to remove a dozen Iraqis from the country. However, other sources assert that the dozen detainees were not Iraqis.

For that matter, Goldsmith never questions whether forcible removal by US forces of foreign captives taken in Iraq actually does accord with Iraqi immigration law. It doesn’t sound terribly likely, unless some conscientious American lawyer hastily rewrote Iraqi immigration law. Without the unar­ticulated premise that the US interest in Article 49 is nothing more than learning its implications for immigration enforcement, this portion of the memo has no point - unless, perhaps, “enforcement of immigration law” is the legal hook on which rendition of foreign insurgents hangs.

Goldsmith then turns to the question of whether Article 49 forbids sending Iraqi captives outside the country for interrogation, to which his answer is no. First he argues that “transfer” and “deportation” both imply permanent or at least long-term uprooting, not temporary removal for interrogation. To show this, he quotes authorities who indicate that uprooting and resettling people violates Article 49.^[363] However, none of his sources suggests that resettle­ments are the only forcible transfers or deportations that violate Article 49, and so this argument by itself amounts to very little.

To show that Article 49 permits temporary transfers, Goldsmith argues that reading Article 49 to forbid all forcible transfers is inconsistent with Article 24, which says that occupiers must facilitate the reception of youthful war orphans in a neutral state.⁸⁸ If Article 24 permits occupiers to evacuate war orphans, he reasons, then Article 49 cannot possibly mean to forbid all forcible transfers, such as sending Iraqi nationals to Afghanistan for interrogation.

Unsurprisingly, no commentator before Goldsmith ever noticed an “inconsistency” between the duty to evacuate war orphans and the obligation not to deport or forcibly transfer captives. No one would reasonably describe parents sending their child to safety as a “forcible transfer” or “deportation.” Nor, therefore, is it a forcible transfer or deportation when a child is moved out of harm’s way by responsible adults acting in loco parentis. The autho­rities acting in loco parentis, not the child, are the responsible decision­maker, so long as they are aiming at the child’s well-being. Goldsmith’s analogy between captives sent to be interrogated and children sent to safety boggles the mind - and that analogy is the sole basis of his argument that if Geneva doesn’t forbid the latter it doesn’t forbid the former. Like the Bybee Memo’s organ-failure definition of “severe pain,” this is legal formalism divorced from sense.

A second argument dispenses more senseless formalism. Goldsmith turns to two other Geneva articles, one protecting impressed laborers and the other protecting people detained for crimes. Among their protections, both articles prohibit such people from being sent abroad. According to Goldsmith, if Article 49 really meant to forbid any and all temporary removals out of state, these two articles would become redundant, and therefore “meaningless and inoperative.”^[364]

The short response is: no, they wouldn’t. The two articles say, in effect, that Article 49’s protection against forcible removal applies even to persons detained for a crime or lawfully impressed into labor. The articles ward off potential misreadings of Article 49 that find implied exceptions to it for impressed laborers or accused criminals. In that way, the two articles strengthen and clarify Article 49 - and unsurprisingly, that is precisely how the Red Cross’s official commentary to the Geneva Conventions explains the relationship among the three articles.^[365]

Goldsmith rejects the commentary’s explanation because Article 49 must not be read to make the other articles superfluous.^[366] Evidently, he believes that the anti-redundancy canon articulated in a 1933 US Supreme Court opinion trumps all other rules of treaty interpretation.

Finally, Goldsmith observes that a separate clause of Article 49 forbids occupying powers from deporting or transferring its own civilians into occupied territory. Presumably (he argues), that prohibition does not prevent the occupier from bringing civilian contractors or NGOs in for the short term. Hence, in this latter clause the words “transfer” and “deport” do not encompass short-term transfers and deportations. Thus, these words do not encompass short-term transfers of persons out of the country either, because “there is a strong presumption that the same words will bear the same meaning throughout the same treaty.”⁹³

Perhaps so, although the only legal authority Goldsmith cites for this “strong presumption” is a US Supreme Court dictum saying something different.⁹⁴ In opinions Goldsmith does not cite, the Court recognizes that in the interpretation of federal statutes, the same-words-same-meaning “presumption... is not rigid and readily yields” to good reasons for distin­guishing meanings in different contexts.⁹⁵ But even if there were a rigid same-words-same-meaning presumption, it hardly follows that words with the same meaning coincide in every respect. If a building code specifies safety requirements for “the cellar of a house” in one paragraph, obviously in that paragraph the word “house” refers only to houses with cellars. But it would be absurd to suppose that in other clauses of the code, dealing with other issues, the word “house” likewise refers only to houses with cellars.

customary international law. Restatement (Third) of the Foreign Relations Law of the United States, §325.

⁹³ TP, supra note 1, at 377.

⁹⁴ Air France v. Saks, 470 U.S. 392, 398 (1985). In the passage Goldsmith cites, the Court says that different words in a treaty presumptively refer to different things. That is the logical converse of Goldsmith’s principle, and neither implies the other. For good reason, then, Goldsmith cites this case with a “cf.” Presumably, if better authority existed, he would have cited it.

⁹⁵ General Dynamics Land Systems v. Cline, 540 U.S. 581, 595-98 (2004). For an even stronger statement to the same effect, see the unanimous opinion in Robinson v. Shell Oil Co., 519 U.S. 337, 343-44 (1997). meaning of the words ‘transfer’ and ‘deport’ is rather different from that in which they are used in the other paragraphs of Article 49.”^[368]

I describe these admittedly arcane details of Goldsmith’s memo because I have heard scholars who despise the Bybee Memo hold up Goldsmith’s as the gold standard of what a pro-Administration OLC memo ought to look like. It is no such thing. Like the Bybee Memo, it reaches a preordained conclusion by kabbalistic textual manipulations. The basic recipe in both memos is the same: lean heavily on “structural” canons of construction, take unrelated bits of law having to do with very different problems, read them side by side as though a legislator had intended to link them, and spin out “consequences,” “interpretations,” and “contradictions.” Where Bybee and Yoo interpret “severe” in the torture statute by looking at a Medicare statute, Goldsmith combines a treaty clause dealing with forcible transfer and a different clause dealing with war orphans to generate an imaginary con­tradiction. Neither memo writer asks the most basic interpretive question: What is the point of this law? To ask that question would have been fatal, because the object of both documents is to protect individuals in the clutches of their enemies, and here the captors - OLC’s “client” - wanted to unprotect them. Unmooring a law from its point leaves only the formal techniques of textual manipulation to interpret it.

At one point, however, Goldsmith pushes back against detainee abuse. In a final footnote at the end of his draft, Goldsmith warns that some removals of prisoners might indeed violate Article 49 and constitute war crimes.^[369] He also includes a reminder that a prisoner transferred out of Iraq for interrogation does not lose “protected person” benefits. These are important warnings, and they buttress reports of Goldsmith’s admirably anodyne role in resisting “the program” (as executive branch officials chillingly refer to their detention, interrogation, and rendition policies).

But then why not say specifically that those benefits include those of Article 31: “No physical or moral coercion shall be exercised against pro­tected persons, in particular to obtain information from them or from third parties”? Is it because a memo that explicitly said, “On the contrary, we believe he would ordinarily retain his Article 31 right against any form of coercive interrogation” would defeat the purpose of removing prisoners from Iraq? Why bury his vague warning in a footnote at the end of the memorandum? Why not quote Article 31 in the text, and point out that no form of coercive interrogation is permitted under Geneva IV?

It seems to me that the most charitable interpretation is that Goldsmith was working among hard-liners, and could subvert abusive interrogation only in a subtle and inconspicuous way. That may be the best an OLC lawyer could hope for. (Indeed, perhaps OLC never adopted his draft memo because even subtle and inconspicuous subversion was more than OLC’s clients could stomach.) But a huge potential for self-deception exists in this strategy. To bury a warning risks its dismissal. And to say, in effect, “You can forcibly remove detainees from Iraq for interrogation, but it’s up to you to make sure that the interrogation does not include coercion,” comes awfully close to Tom Lehrer’s Wernher von Braun (“ ‘Once the rockets are up, who cares where they come down? That’s not my department,’ says Wernher von Braun”).