A girl reads the Koran during the holy fasting month of Ramadan at a mosque in the village of Sanabis, Bahrain, on July 27, 2013. MOHAMMED AL-SHAIKH/AFP/Getty Images

In recent years, the United States and its “Western” allies have faced countless foreign policy choices involving the Islamic world, from engaging with Islamist governments in Egypt and Tunisia to negotiating with Iran to managing drone campaigns in at least three countries. While foreign policy decisions are shaped by many factors, public opinion is a major input. So how does the perceived Islamic character of actors influence foreign policy attitudes toward them?

Unfortunately, our existing understanding of these perceptions is limited. Research shows that religious differences are an important ingredient in foreign policy attitudes — recent survey experiments have shown that Western citizens were more willing to start a war against “Muslim” than “Christian” adversaries. But religious differences are often more complex.

Consider the key participants in the Syrian civil war: The Islamic State, Jabhat al-Nusra, the “moderate” Free Syrian Army, Kurdish rebel groups, Hezbollah and the Bashar al-Assad regime are all broadly “Muslim,” but their Islamic character is portrayed — by themselves as well as by Western media — quite differently. Do these differences shape foreign policy attitudes toward them? When are Western populations really fearful and mistrustful of Islamic political actors?

Our new study in Political Research Quarterly explores these dynamics. In an original survey experiment, we randomly assigned subjects different news stories about the ongoing Syrian conflict in which we manipulated the Islamic character of a fictitious yet realistic foreign actor — the “Free Syria Movement” (FSM) — seeking U.S. military assistance. Specifically, we examined whether giving the actor common Islamic language like “Allahu akbar,” policy goals such as sharia law, and/or labels including “Islamist” affected the respondents’ social affect, political attitudes and foreign policy preferences toward the group. Conducted in May 2015 via Amazon’s Mechanical Turk (MTurk) platform, the survey was completed by 1,095 respondents, with at least 120 in each of the eight conditions.

Here is what we learned.

1. Islamic cues do indeed matter.

Under normal circumstances, we found that respondents’ attitudes towards the FSM were relatively benign. Although they knew the group was Muslim, they tended to give neutral or mixed responses about its level of trustworthiness, compatibility with American values, emotional impact on them and potential role as an American regional ally. Likewise, respondents had mixed views about sending FSM the requested American military aid, although they leaned slightly against doing so overall.

In contrast, with the three cues incorporated, all of these responses shifted in a significantly negative direction. Respondents tended to see the group as untrustworthy, incompatible with their values and interests, a source of fear and a potential regional adversary. Their willingness to give it aid moved firmly toward opposition, dropping on average by more than seven percentage points. And other attitudes saw even larger negative shifts, with the average trust in the group dropping by 10 percentage points. Essentially, respondents did not inherently have hostile attitudes toward the Islamic actor, only when “cued” to do so.

2. Some cues matter more than others.

Yet we also found that some of the Islamic cues harmed attitudes toward the group far more than others. Of the three, insertion of “sharia law” as a policy goal had the most harmful impact, while use of the “Islamist” label did not yield any statistically significant negative effects on any of the outcomes. This is not wholly surprising. Although sharia can have many different meanings in the Muslim world — from inclusive welfare states to punitive morality codes — Western elites have characterized this concept solely in terms of violence and oppression. In the words of Newt Gingrich, sharia is “a mortal threat to the survival of freedom in the United States and in the rest of the world as we know it.” In fact, anti-sharia legislation had been proposed in 23 American states by 2011. This “sharia-phobia” is not unique: other broad Islamic political goals such as the pursuit of a caliphate have been received with similar apprehension in Western political discourse.

3. The influence of these cues depends on partisanship.

Finally, we found that the impact of the cues depends on party identification. With all three cues activated, for example, we see a 22 percentage point drop in trust in the group among Republicans, a 10 percentage point drop among independents and a 5 percentage point drop among Democrats. This also is not wholly unexpected. Republican political elites often describe national security threats in more explicitly Islamic terms — with a greater willingness to label terrorist groups as “Islamic” and invoke concepts such as sharia and the caliphate to characterize their goals. We interpret this mostly as Republican identifiers taking cues from their elites. Yet, as indicated above, independents and Democrats are not immune from these reactions either.

Western anxiety toward Muslim political actors is not just automatic. It is stoked by the rhetorical choices of Western elites who politicize Islamic terms and ideas and by the Islamic actors themselves who use these terms and are either ignorant or indifferent to this politicization. Contested Arabic-language terms such as sharia, khilafah (caliphate) and jihad may make the easiest targets.

This study suggests at least two promising areas of future research. First, we can examine the flip side of the coin: how adopting Christian language, policies and labels in the West influences foreign policy views in the Islamic world. This could help determine whether these processes mirror each other, in a Sisyphean cycle of religious politicization. Second, we could research whether and how these negative reactions to Islamic cues can be effectively countered. Does including brief translations and explanations of these cues that highlight their positive aspects, diverse meanings and/or Judeo-Christian equivalents ameliorate Western apprehension?

For now, we know that politicized Islamic cues such as sharia spark deeply negative Western perceptions and preferences toward their users. In the foreseeable future, Muslim actors seeking Western assistance or support would be wise to use them with great care.

Daniel Silverman is a PhD Candidate in the department of political science at Ohio State University.

Mujtaba Ali Isani is a research fellow at the Religion and Politics Excellence Cluster at the University of Münster in Germany.