The more obvious propaganda often escapes us purely because we’re so immersed in it. It took an Iraqi acquaintance to make me realize that, early in the Iraq war, The New York Times and other papers misused the word “insurgents” for people who attacked US troops. The term lent our side more legitimacy than it legally deserved. If Webster’s is to be believed, insurgents rise up against a recognized authority, and not against an occupying force that defied international law by invading.

Reuters, which prides itself on being the only true internationalist news organization, made a point of banning the word “terrorist” in reference to the September 11 attacks, with the argument that one man’s murderous extremist is another’s freedom fighter. The news agency aims to avoid emotive labels so that customers can come to their own conclusions based on facts. Reuters’ decision highlights what is, perhaps, an obvious point: The way conflict stories are written can substantially affect the public debate around those conflicts. Words matter.

Vocabulary twists apply to other types of violence, too. In Mexico, a “drug war”—an inherently debatable term itself—being waged between rival gangs and against authorities and the public has killed more than 47,000 people over six years. Officials usually avoid the phrase “drug cartels,” and instead refer to the syndicates as “organized crime.” The phrase doesn’t adequately convey the grisly methods of the drug gangs. One thinks of money laundering and numbers-running, not vicious groups that hang mutilated bodies from bridges and leave severed heads on streets.

Yet the media are beginning to consider their de facto role as propagandists who unwittingly help normalize violence. Last year, many of Mexico’s biggest media outlets signed a voluntary agreement to refrain from adopting the “language and terminology used by criminals” in order to avoid becoming “unwilling spokesmen” for the drug gangsters.

The pact left it to individual newsrooms to decide for themselves which words and phrases to shun. During a gathering earlier this year in Ciudad Juarez, the border town that has long been the epicenter of drug-related homicides, reporters debated the appropriate verb for “kidnap.” Until now, common usage was the passive and tame construction se levantó, or “lifted.”

“That implies no one was responsible,” one senior reporter argued. “We should use more direct language like secuestró—abducted.” The assembled journalists nodded, and then quickly requested anonymity so as to avoid reprisals.

Likewise, they discussed the prefix narco, which Mexicans place in front of anything relating to drug lords. It often has an allure for impoverished youths impressed by the glitzy lifestyle. Reporters at the meeting weighed the glamorous associations of terms like narco Polo (fancy dude who wears designer labels), narc-architectura (mansions), and narco zoos (kingpins have a predilection for exotic pets).

“Maybe we should just ban narco,” someone mused.

It’s a start. Yet just avoiding words is not enough. Conflict journalists need to be aware of words: where language comes from, what it means, who benefits by its use, and what it obscures. We especially need to consider these issues as we cover the heated rhetoric over Iran’s nuclear program. For example, many in the media confuse preemptive and preventive wars, although the two are quite different. A preventive war is initiated to destroy the potential threat of an attack by an enemy. This entails suspicion of an eventual assault, rather than one that is actually proven to be planned or imminent. By contrast, a preemptive war is launched in anticipation of immediate aggression, amid clear signs that the other side is going to attack.

The launch of conflict when no attack has occurred is a violation of international law, unless authorized by the UN Security Council. A preemptive strike is seen as justifiable, however, which is why the Bush administration strained to describe the 2003 invasion of Iraq as such. In the case of Iran, we lack indisputable evidence that an attack on Israel or anyone else is “imminent.” It is simply a fear—a well-founded one, perhaps, but nothing as solid as the proof of troop mobilization on the border when Israel struck against Egypt in the Six Day War in 1967.

We, as well as the public, must understand how and why language gets twisted by those who would market war. Those in favor of attacking Iran would like to sell any potential assault as a preemptive war. But unless it fits the criteria, journalists should remain wary.

Judith Matloff is a contributing editor to the Columbia Journalism Review. She is a veteran foreign correspondent, who teaches a course on conflict reporting at Columbia, and is the author of Fragments of a Forgotten War and Home Girl.