Mini Disclaimer: This essay is a bit different from the other essays on this page. It was written as a formal academic rhetorical analysis and is intentionally more scholarly in register than the other work I post here. If you're new to rhetorical analysis, think of it as a close reading applied to political language. This examines not only what was said, but how specific word choices performed legal and political functions.

‍

“Genocide”: How the United States Government Weaponized the Word

In 1948, the Convention on the Prevention and Punishment of the Crime of Genocide, often referred to as the Genocide Convention, was ratified by the international community and unanimously adopted by the UN General Assembly, creating a binding legal obligation requiring all signatory states to prevent and punish genocide wherever and whenever it occurs. At the center of the Convention’s framework was a single word: “genocide”. Formal determination of that one word by any signatory power was designed to trigger legal obligation, humanitarian response, and collective international action. What the writers of the Convention had not considered was the enforceability of such action and that the word’s power would make it the most carefully managed word in modern diplomacy.

Language creates reality, and official language creates political reality. When a signatory state, such as the United States, publicly utters “genocide”, it is performing a legal act with binding consequences, and when it refuses to use the same word, that choice is equally performative (Austin, 1962). Securitization theory holds that declaring something a crisis is itself a political act addressed to an audience whose acceptance legitimizes extraordinary measures (Buzan et al., 1998). In more than one of the United States Government’s (USG) responses to genocide, the audience was domestic, and the extraordinary measure being legitimized was inaction. The cases of Rwanda in 1994 and Darfur in 2004 reveal that the USG understood this explicitly and exploited that single word for its own political advantage. Through these two cases, this essay argues that the international community's use of the label "genocide" functions not as a legal trigger for humanitarian action, but as a diplomatic instrument for managing national liability by either withholding or deploying the word based on political interest rather than moral or legal obligation, progressively undermining the framework the Convention was designed to create.

The Rwandan Genocide and Linguistic Avoidance

The evasive rhetorical strategy deployed by the USG in response to the 1994 Rwandan Genocide was precisely planned to absolve the US from legal obligation of involvement by creating a gap between public and private understanding of the situation. Declassified Defense Department documents reveal that US officials of the Clinton administration were explicitly told through internal communications not to use the term “genocide” when speaking to the public, as a “genocide finding could commit USG to actually 'do something'" (U.S. Department of Defense, 1994). Instead, US officials waited until after the killing had subsided in Rwanda to say “acts of genocide may have occurred”. This deliberate framing both acknowledged the situation sufficiently to satisfy public human rights pressure while simultaneously evading the legal obligation not only to punish but also to prevent the systemic destruction of a particular group of people imposed on the US and all signatory powers under the Convention.

An acknowledgement of an active genocide requires immediate action, but a post hoc phrase such as “acts of genocide may have occurred” incurs no such obligation (Heinze, 2007, p.366). Every word in this phrase was chosen with performative evasion in mind. Diminishing the genocide to “acts of” fragments a planned systemic destruction into isolated events. This framing removes the necessity for international interference by diminishing the gravity and magnitude of the violence. Next, US officials communicating that these acts “may have” taken place removes certainty from public understanding, a certainty that the USG held internally only weeks into the genocide (Power, 2002). Not only does this uncertain language plant doubt in response, but it frames the entire genocide as a passive event itself. Choosing to use the term “occurred” renders fortuitous an otherwise orchestrated event. But genocide doesn’t simply “occur”; it is actively committed by one group against another. This one phrase frames the situation as an unavoidable, passive occurrence, removing all political responsibility. The use of the term “genocide” creates a legal obligation to intervene under the Convention, but framing like this is engineered to prevent that obligation from ever forming. When this establishment of rhetorical evasion is allowed, it creates a template for future response to genocide.

Before their strategic acknowledgement in late May of 1994, the USG evaded the use of the word genocide completely. This strategy was deployed in real time during a press briefing held by State Department spokesperson Christine Shelley. Reuters correspondent Alan Elsner asked, “How many acts of genocide does it take to make genocide?” (PBS Frontline, 2004), a masterful question in itself. By adopting the State Department’s own language and questioning the logical basis for that particular use of words, Elsner unveils the absurdity of the phrasing. Placing “acts of genocide” next to “genocide” in a single statement reduces the terminology to its base framework, illogical. In response, Shelley gave a non-answer, stating, “Alan, that's just not a question that I'm in a position to answer.” (PBS Frontline, 2004). Here, we see rhetorical evasion used differently. The first part of Shelley’s response, “That’s just not a question…” delegitimizes the question and labels it not worth answering. She doesn’t offer follow-up or say she doesn’t know; she shuts the question down completely. The second part of the response, “I’m not in a position to answer”, allows for little to no pushback and signals to the press that she has no authority to answer the question. It’s not that she doesn’t know; it’s that she’s not allowed to tell them either way. This singular exchange is the clearest demonstration of US officials’ linguistic liability management.

This strategy came directly from the top, with Secretary of State Warren Christopher enforcing this decision institution-wide. Christopher told reporters during a June 1994 press conference that if there was “any particular magic in calling it genocide,” he had “no hesitancy in saying that” (Sciolino, 1994; Buchwald & Keith, 2019). However, the Convention did give the word genocide a particular magic, and Christopher knew this. It’s why he formally forbade all State Department officials from using the word to describe the Rwandan Genocide (Heinze, 2007, p.366). In late May of 1994, he sent a cable directive to US delegates in Rwanda stating: 

“Delegation is not authorized to agree to the characterization of any specific incident as genocide or to agree to any formulation that indicates that all killings in Rwanda are genocide” (Buchwald & Keith, 2019, p. 4).

His use of bureaucratic language turns the directive from a political decision into a procedural one. The use of the word “authorized” signals that this isn’t a denial or refusal of acknowledgment; it simply states that US delegates are not authorized to call it such. It is not a choice, but protocol, not to “agree to the characterization” of genocide. This phrasing reframes the term genocide as subjective rather than legally determined. Rhetorically, this positions the USG as declining to endorse an interpretation of the events, not as an evasion of truth. “Any specific incident” mimics the strategy deployed with “acts of” stated above. Again, the genocide is being turned from a systemic problem into individual acts. Since the legal definition of genocide agreed upon in the Convention requires both pattern and intent, this framing weakens any foundation for legal obligation. Christopher’s directive formally prohibits a speech act from occurring by manufacturing institutional silence. This directive alone turned an improvised shirking of legal responsibility into an institution-wide template of how to respond to global genocide. It gave every official spokesperson of the USG an excuse to evade answering the vital question: Was this a genocide?

Internal communication between US officials says, yes, it was a genocide. This linguistic evasion was only for the public, as US officials had privately acknowledged the violence in Rwanda for what it was. The Rwandan Genocide began on April 6, 1994, and within 16 days, US senior officials were privately using the word genocide in now declassified internal communications (Power, 2002). This proves that the uncertainty and inability to identify a genocide that was communicated publicly throughout April and May was an evasion of acknowledgment and the responsibility that would follow. The disparity between public and private understanding rested completely on that uncertainty. This rhetorical performance reconstructed public reality. Private communications between US officials are proof of the gap between the constructed framing and the reality it was meant to obscure.

The Rwandan Genocide ended in mid-July of 1994, and linguistic evasion by US officials was, at this point, ongoing. The political purpose, for the US to avoid any legal or moral obligation to interfere with the mass killing of Rwandan citizens, had been achieved. And yet, the rhetorical strategy lasted still, because the alternative would be to claim accountability for the evasion. The Rwanda case established that, despite global agreements, the word “genocide” could be strategically avoided and manipulated through language and rhetoric to avoid obligation. It set a precedent allowing the evasion of international responsibility. But it still leaves us with the question of what happens when a signatory state chooses to use the word after all?

Genocide in Darfur and the Neutralization of a Word

In September of 2004, Secretary of State Colin Powell used the term “genocide” before the Senate Foreign Relations Committee for the first time after months of avoiding questions regarding violence in Darfur, Sudan (Powell, 2004). Powell’s use of the word “genocide” marked the first time the US executive branch not only used the word to describe ongoing violence, but the first time a sovereign nation had formally accused another of an active genocide since the Convention in 1948 (United States Holocaust Memorial Museum (USHMM), 2026). After analyzing the lack of acknowledgement in the Rwanda case, this may seem like a stark improvement, but, like Rwanda, there is more to the why, when, and how the term “genocide” was used that leads to greater suspicion.

Powell’s remarks contained all the linguistic features of past humanitarian declarations; however, how something is staged is just as important as the words being said. It was September, only two months until the Bush administration was up for reelection. Not only was it an important election, but a coalition of three distinct domestic constituencies had chosen Darfur as a key issue. According to Political Science Quarterly, US human rights organizations had been publicly documenting events in Darfur, making it clear that the USG had to respond somehow. They could no longer remain ambiguous and stay in good standing with the public. The evangelical Christian viewed the events in Sudan through the lens of religious conflict, which consisted of an oppressed Christian South fighting back against an oppressive Islamic North. And finally, the third group of this coalition, the African American Congressional Caucus, framed Darfur as racial violence that demanded a response from the US as a global power (Heinze, 2007, p. 370). These constituencies sustained enough domestic pressure to push Congress to act. Six weeks before Powell’s remarks, Congress passed its own genocide resolution, since which Powell had been avoiding all questions alluding to genocide in Darfur. When asked by reporters, he claimed he was waiting for clear and absolute evidence to base his claim on, but the results of his team’s investigation had come back inconclusive (Hamilton, 2011). Along with this long-awaited and unhelpful evidence, something else had also arrived on his desk. The State Department’s legal advisor, William Taft IV, informed Powell that a formal and public determination of genocide would require no legal action from the USG, contrary to former belief (Hamilton, 2011). The Bush administration made its determination in response to domestic pressure while facing reelection, and their language aligned precisely with that goal. 

When Powell made his public remarks, he said the following:

"I concluded that genocide has been committed in Darfur and that the Government of Sudan and the Jingaweit bear responsibility" (Powell, 2004, para. 38).

And only moments later:

“However, no new action is dictated by this determination” (Powell, 2004, para. 47).

The gap between these two remarks isn’t hours or days; it’s mere moments. It’s a blatant example of the gap between the word’s intended power and what it was used for. Powell determines genocide, and immediately removes the need for action. By saying “no new”, he implies that despite the new determination, preexisting action has been sufficient. The one modifier, “new”, does enormous rhetorical work all by itself. It’s not that nothing must be done; it's that nothing additional needs to be done. It assumes that the current policy has been an adequate response. That second part of the sentence, “action is dictated”, is similar to the passive phrasing of “acts have occurred” in response to Rwanda. Human agency and ownership of the action are completely removed, allowing passive language to diffuse accountability. The passive voice makes it seem like a neutral legal determination rather than the political choice it was. The use of the word “dictated” matters here as well. It acknowledges that the Convention could require action theoretically, but in the case of Darfur, it simply does not. It’s a protective measure minimizing risk dressed up as a statement of fact. Powell isn’t saying that the USG chooses not to act, simply that the law does not require them to. This distinction protects the Bush administration from moral accountability while simultaneously abandoning the legal obligation that is supposed to accompany the use of the word genocide.

We know this method catered to those three domestic powers by looking at the situation and language surrounding the determination, but it was also confirmed through a post-hoc admission by former US representative to the UN, John Danforth, through an interview with the BBC in July of 2005, where he said: “The genocide label was something that was said for internal consumption within the United States” (Sudan, 2008, para. 14.3).

This internal context changes the core meaning of Powell’s announcement; the word genocide was deployed outward but aimed inward. The determination was the path of least domestic political resistance. Danforth’s use of the word “internal” confirms explicitly that the primary audience was not international. The word genocide exists in international law specifically to trigger an international response. When Powell’s legal counsel told him that using the word inversely to gain domestic traction would not demand legal obligation from the US to act on the international stage, he turned the international tool inward. His use of the word “consumption” tells us something else entirely. It implies that this determination wasn’t for diplomacy or law. Consumption is the language of markets and audiences, aimed at a consumer base. It frames the genocide determination as a product given to that consumer base (the domestic coalition) and frames the humanitarian declaration as a deliverable. The people of Darfur and international powers were never the audience.

The USHMM's 2019 institutional report on genocide determinations explicitly warned that “calling or not calling a situation a genocide cannot be a substitute for preventive and responsive action” (Buchwald & Keith, 2019, p. vi). The Convention was built on the assumption that the word “genocide” would hold power with all signatory states. To determine genocide was to trigger legal obligation, humanitarian aid, and collective action across the globe. Danforth’s admission that Powell’s determination was for internal consumption is exactly what USHMM’s report warns about. The label of genocide was used in place of any real action on the international stage. It was neutralized by the lack of new policy or action that followed. The USG’s response to Darfur weakened the legal power of the word completely. The word “genocide” had effectively been neutralized.

The international response to Powell’s determination intensified this damage. In January of 2005, the United Nations Commission of Inquiry on Darfur made its own determination: genocide had not taken place in Darfur, and the USG continued using the label of genocide anyway (Heinze, 2007, p. 374-376). The word had been deployed for domestic political gain, had been immediately removed from its legal power by Powell, and then formally contradicted by the primary international body investigating the situation. Still, the word persisted in official USG rhetoric. As a result, the term was entirely removed from its legal definition and its humanitarian purpose. It was no longer a legal tool, but a political one. It carried weight for domestic audiences while holding no obligation in terms of international law, carrying the weight of legal appearance, while the substance had been systematically removed.

Rwanda established that the word genocide could be suppressed to avoid obligation, and Darfur set the precedent for something much worse. The word could be spoken, contradicted, and manipulated, and the Convention’s writers had built no mechanism to prevent it. The USG had learned not to fear the word genocide, but to wield it.

An Obligation Both Avoided and Obliterated

Rwanda and Darfur unveil a systemic pattern in how the label of genocide is used to manage the speaker's domestic political position rather than protect the people the word was established to describe in 1948. In 1994, the Clinton administration refused to use the word genocide for fear that a genocide finding would commit the USG to take action. Ten years later, the Bush administration discovered what the Clinton administration had not: the Convention’s ruling had no reliable enforcement mechanism, and therefore could use the word freely without consequence. Both were different rhetorical methods towards liability management. In 1994, the word’s power was acknowledged privately and neutralized publicly; in 2004, it was exploited publicly and neutralized legally. Both strategies served the same goal: to protect the USG from obligation through language.

Saying something is an act in and of itself, and those utterances both describe the world and create it (Austin, 1962). The Darfur determination was a performative utterance that both deployed and eliminated the word’s consequences. The genocide label could now exist in the congressional record with no strings attached. Powell had not just failed to act, but he had set a public, documented, institutionally ratified precedent that the Convention’s international goal is unenforceable and therefore optional, and ultimately demonstrated to the international community the limits of the word (Buchwald & Keith, 2019).

Both instances showcase rhetorical choices made primarily to manage the speaking nation’s legal and political reputation. The passive avoidance in 1994 protected the US from legal obligation while preserving its credibility with human rights organizations and the domestic public. In 2004, the determination protected the Bush administration from domestic political pressure, while the "no new action" declaration protected it from international legal obligation. Both times, the primary audience was domestic. They weren’t speaking to other international signatory powers, or those in Sudan or Rwanda. Securitization moves are supposed to be addressed to an audience whose acceptance legitimizes extraordinary measures (Buzan et al., 1998). In both cases, language performed the function of action for domestic audiences by creating the appearance of response while ensuring none occurred. It was a purposeful and precise inversion of what securitization theory assumes and hopefully accomplishes.

The pattern identified is a product of the structural gap in the Convention itself. The word carries moral and legal weight in its naming, but has no automatic enforcement mechanism attached to it. These two case studies showcase the rhetorical range in which the US has managed the word genocide across a decade, through suppression and neutralization. Rwanda is a case study in performative evasion; Darfur is its mirror, a case study in performative retraction. The questions remaining are what this pattern means for the 1948 Genocide Convention’s integrity and usefulness as a tool of international law, and what becomes of international humanitarian communication when its most foundational linguistic tool has been demonstrated twice to be optional?

The 1948 Convention’s enforcement mechanism rests on the assumption that signatory states will use the word in good faith, with humanitarian relief in mind, and an intention to follow through. With Rwanda and Darfur, neither administration did, completely altering the framework surrounding this international law. That manipulation has not stopped. In January 2025, Secretary of State Blinken issued a second genocide determination regarding Darfur because the determination two decades earlier had not provided a resolution (USHMM, 2026). At the same time, genocide allegations are being actively adjudicated by the International Court of Justice regarding Gaza, while atrocities in Sudan, Myanmar, and beyond continue to test whether or not the international community will name what it sees and take action. The Genocide Convention’s power rests on a single word. The US didn’t just fail Rwanda and Darfur with that word; they illustrated to the world’s most powerful governments how to survive it and avoid its consequences.

Bibliography

Austin, J. L. (1962). How to do things with words. Oxford University Press.

Buchwald, T. F., & Keith, A. (2019). By any other name: How, when, and why the US government has made genocide determinations. United States Holocaust Memorial Museum, Simon-Skjodt Center for the Prevention of Genocide. https://www.ushmm.org/m/pdfs/Todd_Buchwald_Report_031819.pdf 

Buzan, B., Wæver, O., & de Wilde, J. (1998). Security: A new framework for analysis. Lynne Rienner Publishers.

Ferroggiaro, W. (Ed.). (2001). The U.S. and the genocide in Rwanda 1994: Evidence of inaction [Electronic briefing book]. National Security Archive. https://nsarchive2.gwu.edu/NSAEBB/NSAEBB53/ 

Hamilton, R. (2011, August 17). Inside Colin Powell's decision to declare genocide in Darfur. The Atlantic. https://www.theatlantic.com/international/archive/2011/08/inside-colin-powells-decision-to-declare-genocide-in-darfur/243560/ 

Heinze, E. A. (2007). The rhetoric of genocide in US foreign policy: Rwanda and Darfur compared. Political Science Quarterly, 122(3), 359–383.

PBS Frontline. (2004). Ghosts of Rwanda [Documentary transcript]. PBS. https://www.pbs.org/wgbh/pages/frontline/shows/rwanda/etc/script.html 

Penketh, A. (2005, July 2). US described Darfur as 'genocide' to please Christian right – Danforth. The Independent. https://sudantribune.com/article/11125 

Powell, C. L. (2004, September 9). The crisis in Darfur [Testimony before the Senate Foreign Relations Committee]. U.S. Department of State. https://2001-2009.state.gov/secretary/former/powell/remarks/36042.htm 

Power, S. (2002). "A problem from hell": America and the age of genocide. Basic Books. 

Reeves, E. (2004, December 15). Secretary of State Colin Powell's genocide determination: What it does, and doesn't, mean for Darfur. sudanreeves.org. https://sudanreeves.org/2004/12/15/secretary-of-state-colin-powells-genocide-determination-what-it-does-and-doesnt-mean-for-darfur-september-10-2004/ 

Sciolino, E. (1994, June 16). U.S. to supply 60 vehicles for UN troops in Rwanda. The New York Times. https://www.nytimes.com/1994/06/16/world/us-to-supply-60-vehicles-for-un-troops-in-rwanda.html 

Sudan, Permanent Representative to the United Nations. (2008, July 29). Identical letters dated 29 July 2008 from the Permanent Representative of the Sudan to the United Nations addressed to the Secretary-General and the President of the Security Council (UN Doc. S/2008/500). United Nations Security Council. https://undocs.org/S/2008/500 

United States Holocaust Memorial Museum. (2026, March 20). Darfur. Holocaust Encyclopedia. https://encyclopedia.ushmm.org/content/en/article/darfur 

U.S. Department of Defense. (1994, May). Discussion paper: Rwanda [Declassified document]. National Security Archive. https://nsarchive2.gwu.edu/NSAEBB/NSAEBB53/rw051294.pdf