There's no question that, prior to last week, President Donald Trump held fairly anti-interventionist views on foreign policy. Breaking dramatically with Bush-era Republican orthodoxy, Trump ran on a message of "America First" and of avoiding spending blood and treasure on adventures overseas. And yet, with last week's Tomahawk missile barrage against the Syrian government in response to a chemical attack on civilians, this administration took a step in the exact opposite direction from Trump's earlier rhetoric.

If Trump was elected by voters who supported non-intervention, wouldn't breaking with that position then cost him support? Online backlash from the "alt-right" certainly led to speculation that the move in Syria would potentially cost Trump his "base." But this all assumes that Trump's voters held anti-interventionist views and therefore gravitated toward him, rather than the other way around: gravitating toward Trump for one reason or another and adopting a more complete, weakly-held slate of Trump views only after the fact.

And indeed, polling suggests that last week's actions in Syria will neither strongly help nor strongly hurt Trump's position, despite plaudits from Hillary Clinton and skeptical tweets from Laura Ingraham. Trump's job approval has remained flat for the last week. But shouldn't Trump's voters be angry with him for doing a 180 on intervention in Syria?

Political science has long tried to tackle a fundamental question of voter behavior: Do voters choose politicians because those politicians hold views that they like, or do voters choose policy positions because they politicians they like say those positions are correct? Do voters have a set of preferences and go searching for a candidate that best fits those preferences, or does it go the other way, where voters pick a candidate because of something emotional or embedded in their identity, and that leader shapes voter views? It's not hard to assume that voters do not have deeply considered views on each and every policy issue before them, but instead perhaps have one or two strongly held views, and then allow their favored political leaders to fill in the gaps on the rest of the issues.

Perhaps no modern political issue tests this question quite like that of Russia and Syria. Indeed, looking at polling data from 2013 on the question of whether or not to engage in military activity in Syria is like looking at polling conducted in an alternate dimension on another planet, with views on these issues having been completely reversed in a very short time span.

Let's start with the question of Syria. In 2013, when Barack Obama was commander-in-chief, the GOP shed their affinity for overseas engagement, with more Democrats than Republicans saying that they would support launching missile strikes against the Syrian government if chemical weapons were found to have been used. More broadly, Republicans in the Obama era leaned away from supporting global intervention, with more Republicans than Democrats holding the view in 2016 that the US should deal with its own problems rather than try to help the rest of the world.

Fast forward to today, and while Democrats' views toward the value of action in Syria is effectively unchanged at 37 percent support, Republicans have dramatically shifted toward supporting getting involved, leaping from 22 percent supporting such action under Obama to 86 percent supporting it today.

There's also the wildly swinging pendulum of views toward Russia. In that 2013 Washington Post poll, Democrats were more confident than Republicans that Russia could be a partner in finding a diplomatic solution to getting chemical weapons out of Assad's hands.

But things move quickly: just between the late summer of 2016 and December of the same year, the percentage of Trump voters who reported viewing Russia as more of a foe than a friend fell from 67 percent to 56 percent. (Meanwhile, post-election, Clinton voters overwhelmingly reported viewing Russia as unfriendly or an enemy, with 77 percent saying as much in YouGov's February polling.) On the question of Putin himself, the GOP – which had formerly been more negative about Putin than the Democrats – moved to nearly neutral by the end of last year. Gallup found something similar, with views of Putin being very negative across parties in 2015 but with his favorability tripling among Republicans by 2017.

Fast moving views are not likely to be strongly held views. Instead, they're much more likely to be about people mirroring back the signals they see coming from the leaders they support. People can resolve dissonance by shifting their own view on issues that aren't top of mind. So while Trump's actions may look radically different than what his Twitter feed of the last few years might suggest, don't assume this is going to mean his voters will abandon him. It's much more likely that they'll change their minds about the policy rather than the person.

Kristen Soltis Anderson is a columnist for The Washington Examiner and author of "The Selfie Vote."