A new book about deconverts from The Church of Jesus Christ of Latter-day Saints has successfully made its way into the broader conversation. Jeff Strong’s “Torn” has been relentlessly marketed, and many have presented it as either dangerous or groundbreaking or both. I am not convinced.
It is not the long-awaited key to Latter-day Saint disaffiliation. It is not some groundbreaking new approach. It is a sometimes useful, often sincere, frequently repetitive book about being nicer to people who struggle with faith. And it’s wrapped up in data that tells us far less than it intends.
That wrapping is important to consider. Strong has consistently advertised his book as “research-grounded.” He has spoken at length about the number of survey responses he has collected.
Like Jeff Strong, I am not a social scientist. I have also occasionally used surveys to attempt to support ideas I’m writing about. But when I do, I constantly hedge because my data isn’t well collected. I explain over and over which groups the data actually covers. I explain that the data is not representative. I explain that while the data may be an interesting window, the takeaways from it are inherently limited.
Strong has taken another approach, not only trumpeting his data, but building an entire worldview around it. And honestly, without the aura of “data-driven,” “Torn” would just be a familiar Latter-day Saint pastoral plea: listen more, judge less, be warmer, make our wards more welcoming. The kind of plea that Latter-day Saint leaders have been making for decades.
Stephen Cranney and Josh Coates handled the basic statistical problems in their own review of “Torn.” The basics are that Strong specifically took the survey to online groups of Latter-day Saints, particularly disaffiliated ones, and they shared it with those who shared their worldview. The survey was very long, so only those with some motivation to finish did. They compared his results to representative samples collected in the past, and unsurprisingly, Strong’s results are very different from the better-collected data.
If Strong wanted to collect data this way, he should have relied on qualitative research.
Methodologically, “Torn” is a nonstarter. If Strong wanted to collect data this way, he should have relied on qualitative research. A better understanding of the people who are attracted to religion-critical spaces online would be valuable. But terminally online Latter-day Saints are only a small segment of those you find in the pews on Sundays. They think and believe differently.
In Strong’s defense, his poorly drawn sample is very large. While Strong has agreed with these limitations when pressed, his book and its marketing rely on the assumption that the data is much more representative than it is.
But it’s not just the data that is a problem. Ralph Hancock has provided the philosophical critique. Hancock points out that Strong shifts from trying to listen and learn from those who have left the Church to not considering the possibility that they may be in the wrong. Strong presents deconverts as always perfectly innocent and wise. He tells us that sin or laziness cannot be the reason people leave the Church, because those who left did not report that as the reason in his survey.
That goes well beyond Strong’s sampling problems, to a kind of baked-in gullibility.
I spoke of a similar problem in an essay called “The Fiction of Self-Knowledge,” explaining how accepting individuals’ self-reports uncritically, as Strong has done, is bad social science. It’s not that these people are lying or attempting to trick us. It’s just that human beings do not fully and reliably understand their own motivations.
And in perhaps the best explanation of the book, Dan Ellsworth points out some of the underlying worldview embedded in Strong’s questions. The issue is not merely that Strong asked the wrong people, but that he is asking the wrong questions. Throughout his questions, “church culture” becomes the load-bearing explanation for nearly everything. Ellsworth’s point is that this is not a neutral analysis. It is a worldview. Strong’s framing repeatedly treats the central problem as a conflict between a protective, boundary-conscious church culture and a more open, nourishing, Christ-centered one. But that binary is doing more work than the data is. Once the options are framed as fear or love, exclusion or belonging, you’ve forced the respondent into an answer.
Strong also never attempts to compare Latter-day Saint deconversion rates to other religious organizations.
Strong also never attempts to compare Latter-day Saint deconversion rates to other religious organizations or explain why there are comparatively fewer deconverts than most other US religious bodies. Strong tries to tell us that there is something wrong with Latter-day Saint culture, but never tells us what culture he is comparing it to. It would be like criticizing baseball player Shohei Ohtani for getting out over half the time he goes to bat. Sure, that leaves a lot to criticize, but it would also be wildly off base.
So the case against “Torn” is fairly straightforward. The data is weak for multiple reasons. The conclusions are overdrawn. And the book’s main message seems to have been imposed on the data rather than taken from it.
But there is real good in the book.
Strong is right that Latter-day Saints should be kinder to those struggling with their faith. He is right that we shouldn’t automatically assume that those who struggle are lazy, shallow, wicked, or easily offended. He is right about the value of listening.
But the reason I’m confident he’s right is that none of this is new.
President Dieter F. Uchtdorf said in 2013 that we sometimes assume people leave because they have been “offended or lazy or sinful,” and then immediately added, “Actually, it is not that simple.” Elder M. Russell Ballard told Church educators in 2016 that students cannot be brushed off with “Don’t worry about it,” and urged teachers to know the Gospel Topics essays thoroughly enough to give thoughtful answers. President Russell M. Nelson taught in 2023 that a teenager who doubts his testimony does not need judgment, but the pure love of Jesus Christ reflected in our words and actions.
The Church’s resource on “Helping Others with Questions” tells members to respond with love, humility, kindness, and patience. It suggests avoiding dismissiveness and preserving relationships.
The parts of “Torn” that are good are basically a regurgitation of the best advice church leaders have been giving on this topic for over a decade.
Strong is right that Latter-day Saints should be kinder to those struggling with their faith.
Repetition isn’t always a flaw. Repeating and extending the voice of church leaders is a big part of what I strive to do myself. But “Torn” uses bad data and bad methodologies to smuggle in assumptions about how to do those things.
So yes, be kind, be warm, listen. Every question isn’t rebellion. Allow people to belong while they are learning, growing, and uncertain.
But it raises a more interesting question. If the data is poor, the reasoning so overextended, and the best counsel regurgitated, why has “Torn” so effectively captured the attention of the Latter-day Saint conversation?
The answer is that for many people, “Torn” is not being read as research or advice. It’s being read as a way to shift accountability for deconverts’ decisions away from themselves and onto a nebulous “culture.”
I remember as a student, when I wrote editorials, I had a terrible tendency to attribute problems to “they”—this thing I felt, but that didn’t actually correspond to any specific person or group. They were saying this. They were doing that. I was permitting myself to react to a phantom of my own making.
“Torn” reveals a community anxious about losing people. We want answers. We want ways forward. And while the advice is good and can continue to improve our spaces, making them better and more welcoming, for so many others, it gives them language to excuse choices they still feel defensive about because “they made me do it.”
“Torn” is frustrating. Its heart is clearly in the right place. But it seemed to have its conclusion locked and loaded before the start. And I think, unfortunately, because it externalizes control so thoroughly, it is likely to have the opposite effect from the one its author hopes for. I hope someone finds a better, more circumspect use for the copious data Strong collected.
In the end, what’s new in the book isn’t that good, and what’s good in the book isn’t that new. Read it if you want to know what everyone is talking about, but if you don’t, you’re not missing much of anything.
