The following graph, in various forms, is making the rounds:

[Image removed.]

The suggestion of the graph (and usually mentary by those who share it) is that Sec. Hillary Clinton lost to President-elect Donald Trump because Democrats didn’t turn out to vote for her like they did for President Obama.

The idea is that Hillary Clinton was a historically unpopular candidate. This is true. Second only to Donald Trump, she was the least liked candidate of all time, at least since anyone has been keeping track. Her career, though long and plished, has been plagued by scandal, much of which surfaced in the final weeks of her campaign. It makes sense that maybe Clinton just didn’t get enough Obama voters to show up at the polls.

I’m unsure the source of the data. It may pletely accurate, but even if so it is misleading. As Carl Bialik wrote last week for FiveThirtyEight, “On average, turnout was unchanged in states that voted for Trump, while it fell by an average of 2.3 percentage points in states that voted for Clinton. Relatedly, turnout was higher petitive states — most of which Trump won.”

So turnout was depressed for Clinton, but apparently only in those states that she won. Low turnout, then, can’t explain why she lost the states she didn’t win. And, in fact, this doesn’t even capture the phenomenon accurately, since she is on track to win some states by a greater margin than Obama did in 2012. Thus, depressed turnout in the states she won might mean fewer Republican-leaning voters there and not that she failed to turn out her Democratic-leaning base.

When we look at the states Trump won, and swing-states in particular, at least two things seem necessary: 1) Trump won many voters who had previously voted for Obama. 2) Trump actually did bring some voters to the polls who had not voted in recent elections, as he claimed he would do.

That many voters who voted to reelect Barack Obama, our first African American president, would four years later vote for a man who questioned Obama’s citizenship and whose campaign was plagued by association with alt-right, white identity politics suggests that issues of race were not important to these voters. Of course, that doesn’t mean that race wasn’t an important factor for other voters, nor that these voters don’t actually care about racism in general, but only that this narrative, just like any other, has limits and flaws.

Digging into voter exit poll data from this highly contentious election has been insightful to me for this very reason. Like anything in life, it paints a much plex and layered picture than what many people would like to admit. Even as one of the big stories of this election is the failure of polls to correctly indicate the winner beforehand, the exit poll data we have has the opposite plementary effect of the many criticisms of the pre-election polls: plicates our assumptions and points to the limits of the stories we like tell ourselves. Data has limits, but so do narratives.

My concern over the last week has been to point to these limits not for the sake of “gotcha” punditry but rather humility. Any knowledge worthy of the name ought to be panied by humility. At least since Plato, philosophers have repeated the dictum that the more we learn, the more we discover we do not know.

All of this amounts to a challenge to rationalism. By rationalism, I do not simply mean reason. Without reason we couldn’t know anything. But rationalism is the belief that human reason can explain everything. If that were true, there could not be anything beyond our ability to know prehend. But from a Christian point of view, this is an essential tenant of the faith.

In perhaps the most terrifying passage of the Old Testament, at the end of the book of Job the Lord speaks out of an ominous whirlwind to answer Job’s questions … by asking Job a ton of questions he can’t answer. Job struggled with the problem of evil: Why do bad things happen to good people? (In particular, why did bad things happen to Job?) He and his friends had a long debate over the nature of God and justice, only to end with no resolution, until God shows up.

On the one hand, God honors Job’s request to appear before him and argue his case. On the other hand, God reminds Job how little he really knows, saying,

Who is this who darkens counsel

By words without knowledge? (Job 38:2)

A lot of pollsters are probably feeling like Job at the moment:

What shall I answer You?

I lay my hand over my mouth. (Job 40:4)

I’m less optimistic about the pundits, however. For many, this is a time to lay blame on others. For others, it’s time to gloat. In the midst of it all, the voices of real people who once again took the time to participate in our democratic process can get drowned out.

Some fear a Trump presidency over worries of how some of his most radical supporters might be emboldened. (Trump himself actually spoke out against this sort of thing on 60 Minutes last night, even if downplaying his own knowledge of it.) Others celebrate Trump’s victory, feeling that finally someone has listened to them. In reality, both are justified in their beliefs to some degree. But we run the danger of losing all of that and more amidst the “words without knowledge” of overly mentary.

All this is not to say that anyone who shared the graph at the start of this post is some self-serving huckster looking for Facebook “likes” or even that they are therefore rationalists. I almost shared it myself, in fact. It is interesting, and I’m thankful that someone shared it with me. Rather, my point is only to highlight that while turnout is another piece of the puzzle, it also turns out to be more than it appears. Correcting our assumptions about the existence of unexplainable aspects of reality can help us maintain our humility and safeguard against making hasty conclusions, mistakenly presuming that all of reality can fit into our heads, even as we admirably seek to know all that we can.

For those of us who are religious, at least, practicing that humility is something that ought to be considered essential to our faith.