Ryan Burge, a sociologist who garners attention as a pundit on US religion, recently published an essay on Substack about a poll on spirituality. The poll and his spin on it relate to things I’ve written here, such as this and this, and I write to make a few comments.
First about the study. Pollsters asked 3600 people to react to thirteen keywords, the most interesting of which turned out to be love, values, peace, and transcendence. Their key question was how many respondents associated each keyword with religion and/or spirituality. (They could choose one, both, or neither.) The pollsters also asked people how religious they considered themselves to be, then the same question about how spiritual they felt.
Why Do We Care and What Do We Expect to See?
Before we get to summarizing results, let’s step back to consider two background questions. First, suppose we don’t care all that much who wins the thirteen horse races for popularity among religion, not-religion, spirituality, and not-spirituality. Rather, suppose we want to know how to strengthen aspects of our culture that promote values of peace and social justice—and so by extension we valorize some forms of religion/spirituality over others. More pointedly, suppose we want as many horses as possible from all the races—whether they self-profess to be religious, spiritual, or “none”—to build on shared values and make common cause against militarism and toward economic justice. We may care to some extent which keywords are resonant and which horses edge out others by a nose or big margins. But our core interest is whether they are running in the same direction. Hold that thought, I'm coming back to it. Spoiler alert, I'll be worrying that this focus gets lost in Burge’s world of spin.
Second, consider what we might “normally” expect to hear about these races—so that actual results could reinforce or unsettle what we thought we knew. Here we bump up against pundit wisdom on spirituality—non-trivially informed by Burge’s work as a public intellectual, although in this case he is largely floating along in a mainstream of common sense and, to his credit, he underlines how badly our working definitions of spirituality are muddled.
It is fair to note three generalizations that often float amid such muddled ingredients.
First is a widespread assumption that religious people (especially white Christians) trend heavily conservative, while liberals skew “secular” (another muddled term) and answer “none” if pollsters ask about their religious proclivities.
Second, one battle line or blurry zone between the resulting camps—religious conservatives and non-religious liberals—centers on people who self-describe as “spiritual but not religious” or say things to pollsters that get them lumped there.
Third, with or without due caution about muddled definitions, the difference “spiritual” is assumed to make, as distinguished from religion, often signals some drift toward “none-ness” and/or the liberal side of the two-party conception. Moreover, there is fairly settled wisdom that a drift toward none-ness is more pronounced among younger people, linked to discontent with the alliance between Republicans and the Christian Right.
These ideas range from true and important to half-true and pernicious. But suppose we expect our poll to reinforce them—and conversely that if we enter expecting this, we will be predisposed to interpret data through this lens.
What Burge Mainly Saw
Burge loves to write about nones and tease out complexities. Much of his cachet is based on trumpeting the rise of nones and quantifying trends in various subcategories—for example, exploring whether Trumpism is capturing a growing fraction of nones or semi-nones to complicate the main trend. I’ll say more about how this relates to two other muddled categories, "evangelical" and "Christian nationalist," in another post.
Accordingly, when we turn to today's data—the thirteen horse races created by asking people if they associate words like “love” with religion and spirituality—Burge zeroes in on how respondents were more likely to pick spirituality.
It isn’t wrong to notice that. Although results vary widely, overall the spiritual votes came in around 10-15% higher. In a typical case, 41% associated purpose with religion but 46% associated it with spirituality. We can’t tell exactly how many of the 46% voted with the 41%, but can gather that spirituality is somewhat more popular and seems to be an umbrella category, with religiosity as a subset.
One of the bigger gaps emerged in people's reports on how much they consider themselves religious and/or spiritual. Since they could pick “not at all,” “slightly,” “moderately”, or “very” for each, there are sixteen possible combinations. Virtually no one combined high religious interest with low spiritual interest, but 9% said they were moderately or very spiritual but not at all religious. Factoring shades of “slightly” into this mix pushes the numbers higher.
Here we have data about a spiritual but not religious cohort. Likely it overlaps with “nones,” although the poll doesn’t address this directly. Let’s not forget the muddle: literally the same person might appear in different polling data as a none, a secularist, spiritual but not religious, or just a plain Christian or Jew who doesn’t go to church or synagogue very often. Meanwhile, spirituality can refer to conventional religious practices like prayers to transcendent gods “out there,” and/or sensed inner-worldly or immanent presences, and/or simple ethical sensibilities (perhaps with belief in neither transcendent nor immanent gods). That's a full spectrum from orthodox piety to atheism. Still, this study does capture useful data about the word associations of some people near the heart of these ambiguities.
We could dive into the weeds, but for now let’s focus on Burge’s takeaways. He notes, as we already did, that “lots of people [are] not religious but somewhat spiritual, but the opposite is much less prevalent.” And he says, “people tend to associate many words with spirituality, but fewer terms describe religion…only a few terms [are] clearly associated with religion.”
Such a summary can dovetail with common wisdom about religion signaling conservatism and nones rising. By the end of Burge’s piece, he says something related. This comes after he inquires how people's religious and/or spiritual interests impacted their reported political activity and civic engagement. Barely over 10% strongly agreed that religion or spirituality made any difference, while 70-75% claimed little to no impact. However, spirituality did track somewhat higher, in line with its overall lead. This prompts Burge to say two things. First, he notes that these findings do not reinforce a stereotype of spirituality being individualistic; rather they suggest that religion and spirituality do similar cultural/political work. (Fair enough, although I might put more stress on the 10% vs. 75% part). Then, in his closing sentence, he lands a conclusion with fishier smelling overreach: “it’s not religion that drives political engagement, it’s spirituality.”
(By the way, many white evangelicals are trying furiously to rebrand as moderate/apolitical in light of their terrible reputations, while liberal churches are busy seeking least common denominator formulas while ostentatiously saying “we’re not like those politicized evangelicals.” Yet all these church folks do have political and civic entanglements, however explicit these are and whatever weight they carry. Some respondents’ ideas about such entanglements are probably mixed up with their churches' talking points enough to affect the precision of these numbers.)
To wrap up this section, note how we can fit Burge’s comments within pundit common sense about religion signifying conservatism while none-ness signals liberalism and growth. (Clearly many independent variables are in play, but my point is about how interpretive streams flow.) If so, supposing that we circle back to my question about values that one might mobilize toward social justice, a default extrapolation might be that the spiritual and/or none would be preferred horses to ride, with religion somewhat getting in the way while waning in importance.
What Stood Out to Me
Many readers likely skimmed Burge, took away something like this message, and moved on. But he got a rise out of me with his summary formula: “only a few terms [are] clearly associated with religion.”
When I looked into it further, what stood out were not the divergences—in many cases fairly small—that Burge stressed, but how much the religious and spiritual overlapped. Consider the sixteen combinations we noted. Yes, they included a “spiritual but not religious” cohort with no “religious but not spiritual” counterpart. However, the key trend was high spiritual interest and high religious interest tracking together. By far the most popular combination was moderate religion with moderate spirituality. That was 23% of the sample; another 9% claimed to be very spiritual and very religious; 8% more picked “moderately” for one and “very” for the other. Disinterest in religion and spirituality also trended together, with 12% saying “not at all” for both and another 12% choosing “slightly” for both. Nancy Ammerman also discovered this, as I discussed in my earlier essay, linked here again. Although Burge flags this trend, it's not where his spotlight falls.
Turning to the thirteen keywords, the big picture suggests neither two poles, nor nearly as many either/or choices as one might gather from skimming Burge. It doesn't show the spiritually-interested on one side and religiously-interested on the other, but simply a stronger overall tendency to associate terms with the spiritual.
The most significant (paradigmatic?) cases included two of the three terms with the highest combined votes. In the “love” race, the spiritual horse edged out the religious one 55% to 47%, but then turning to “values,” the religious horse won 51% to 47%. (I’ll discuss the third shortly.) In NBA playoff terms, this is like a 1-1 tie to kick off a high-stakes series. Meanwhile the terms with the lowest overall numbers—transcendence and transformation—scored low on both spirituality and religion, within the overall pattern of spiritual as more inclusive.
(Speaking of transcendence, I might be inclined to throw out the results, somewhat like I would treat answers to an exam question with ambiguities that confused my students. Although intriguing, it seems fishy that only 12% associated religion with transcendence. One might guess that even people skeptical of religious/spiritual talk about transcendence might make an association, even if they went on to disavow believing it. Think about how one might associate the eucharist with Catholics even if one was anti-Catholic—but only if one had a clear fix on what “eucharist” means.)
Re-reading Burge’s summary phrase—the one where few keywords are “clearly associated with religion”—one grasps that this is shorthand for a more precise idea: religion only won three races, bucking a pattern in which spirituality won eight. (That's 11 of 13; “security” and “morality” tied.) Here we could veer into the weeds, so If you wish to stay out, just remember that second place horses ran nearly 90% as fast as the winners. It’s not that they “weren’t clearly associated with racing.”
Still there is interest in some of the divergences. Let’s start with terms where religion led. It won (35% to 29%) for belonging. As noted already, it scored 51% in an important “values” category where spirituality only scored 47%. It also polled far higher for “structure,” with 32% compared to spirituality’s 22%. Here, despite a stereotype of individualistic spirituality not being reinforced, a stereotype of disorganized New Agers versus rigid Christians stayed on track. But there is a complication: spirituality trounced religion for clarity (32% to 25%) so one wonders why structure and clarity diverged so much.
Turning to cases where spiritual numbers were higher, the most important case concerned “peace,” the term with by far the highest composite score. Peace and spirituality went together 68% of the time—easily the top number in the poll and well above the 53% who found religion peaceful. However, should we say peace is “not clearly associated with religion” when 53% was religion's highest score among all thirteen terms, and the third highest overall? Also, if we circle back to my opening concerns, sometimes it is a mixed blessing to score higher on peace in a world where evangelicals are saying “all lives matter” and protesters are in the street chanting “no justice, no peace” (that is, without justice there can be no peace). We need to bring this down to cases—although, of course, some such cases touch on the common sense that religions cause and/or worsen wars.
(If you think this latter idea is settled wisdom, I urge you to read this book. It argues powerfully that modern secular states within a capitalist mode of production, not religious difference, have been the overwhelming drivers of war—and that states deploy the idea of keeping violent religion under control as a major ideological justification for their militarism. This leads toward discursive and legal strategies that privatize and sideline those parts of religion that oppose the wars. How much of the gap in our poll reflects this set of ideas?)
Lower down the popularity list, spirituality also outdistanced religion for “wonder” (the largest proportional gap, 17% vs. 27%), “inspiration” (in fourth place for popularity, with spirituality ahead 50% to 43%), and a few others. You can dig down into the weeds with graphs in Burge’s report.
My own takeaways were as follows. First, amid the thirteen concepts, peace, love, and values were the most popular, and all three were largely both religious and spiritual, albeit with the complexities we noted around peace. Second, this tracks with a trend of spiritual and religious overlapping more often than not—although if one came in expecting a tie, then religion would seem to be underperforming, while by the same token if one expected huge zero-sum conflicts, this would largely be disconfirmed. Third, spirituality seems somewhat more inclusive and/or valorized—perhaps by 10 or 15%—although somewhat less structured and/or conducive to belonging as compared with religion, and less prone to steer into conflict for better and worse.
Two Complications to Distort These Numbers
Now let’s factor in two huge complications that I underlined in my earlier articles. First, it is a common rhetorical and theological habit of self-identified pious evangelicals to say things in the ballpark of “I’m not religious, I’m spiritual.” In fact, this sometimes signals seriousness of commitment. Adding more complexity, this is an outgrowth from historic Protestant polemics against Catholicism, some probably warranted and many certainly unwarranted.
Meanwhile, by extension, many self-identified nones may be exhibiting unacknowledged “secular Protestant” behaviors/assumptions when they say similar things. This is especially true if they were raised Protestant and are now in conscious rebellion. Carrying forward secularized Protestant values can work for better and for worse, with practical impacts big and small. Sometimes this ramifies in ways that are important for peace and justice work. In any case, bearing this in mind clearly matters if our concern is to understand and win debates over cultural common sense. Equally clearly, this entails large limitations and margins of error if we hope to draw wide conclusions from polls.
I suppose that if pollsters could quantify these two phenomena—self-described serious Protestants claiming not to be religious alongside self-described nones covertly shaped by secularized Protestant habits of mind—it would narrow the gaps between spiritual and religious in today’s findings. It is hard to be sure about it, and I doubt that quantitive polls are a promising way to clarify it, but it’s fascinating to think about.
Tying it Together
Let’s return to my opening questions. How does this matter? Some people, including myself, do not care much whether slightly more people associate, say, “love” with spirituality (55%) compared to religion (47%). This is already true if we posit that the data is straightforward, and it becomes emphatically true if we factor in ambiguous definitions and the rising prevalence of “garbage in garbage out” polls with small samples and low response rates.
(I have no particular reason to doubt this poll. It seems useful and Burge is good at testing various polls against each other. It’s just that I’ve learned to distrust religion polls as my default stance, especially for spinning big conclusions out of smallish numerical differences. And I note that the most cited of all sociologists in this field over the past two generations has recently weighed in—at book length with extensive receipts—with similar distrust.)
What is it that we do care about? For me, above all, it is understanding and strengthening prospects for cultural work toward less militarism and greater economic justice. Here, just when one might want this essay to end, we risk veering into a whole other essay to unpack why I say this. Briefly, my premise is twofold First, historically and still today, it has often been necessary to have strong religious/spiritual buy-in to create successful social movements. That’s not without a need for non-religious and even anti-religious parts of these movements, and it’s not without a known bug: certain religious groups can mess up certain coalitions and/or annoy people by claiming more credit than they deserve. Here again we need to bring our analysis to cases. Still my point stands: commendable kinds of religious buy-in have typically been a necessary but not sufficient piece of such puzzles.
Second, what “we” think we know about the decline or marginality of leftward fractions of US Christianity is a prime example of what I earlier dogged as a “pernicious half-truth.” The last thing we need is to make the supposed death of the religious/spiritual left into a self-fulfilling prophecy. About half the time such prophecies are more like right-wing talking points—something to push back against rather than echo uncritically. If you click around on this site, numerous essays unpack aspects of this problem.
So suppose our take-away for today is that the religious and the spiritual largely overlap. Indeed, let’s go further—suppose we understand them collapsing into each other as a “spiritual-religious,” which is probably the best way to think about it most of the time. Then, recall that this poll's strongest findings were approximate convergences between the spiritual and religious on love and values, plus a strong showing for religion in the peace category (in this case with a gap which in some activist contexts might ramify as reflecting better on religion.)
What follows? Maybe it’s a gestalt switch: a lot less default sorting into two polarized camps (conservative religious vs. liberal nones, anti- vs. pro social justice) and a much more fluid conception, both of the spiritual/religious and of dividing lines between the religious and secular.
Being religious, spiritual, or none are all contested categories, with conflicts over values inside each camp. There are many conservative nones, a large minority of left-liberal religious people, and plenty of debates all around. But all three of these groups have a great deal they might agree about or build upon. It’s not that their commonalities always support peace and justice agendas; it's that shared values can help draw a subset of people toward such agendas. Certainly our pundit common wisdom should not forget—as often as it does—that left-trending religious people are relevant, simply based on such people complicating two-party oversimplifications or being numerical minorities in certain (often muddled) categories.
Burge probably agrees with a fair amount of what I’ve written. Still, his spin on spirituality, floating as he does near the mainstream of cultural common sense and zeroing in as he does on smallish differences in the horse races instead of directions in which the horses are running, might easily lead us away from the gestalt switch we need.