Who Do You Say That I Am? Part 3
The difficulties of going beyond self-identification (6 min)
In part 1 of this series we covered the fact that self-identification is unreliable by nature. We just can’t rely on polling data that simply asks people what they identify as. We cannot rely on people to tell the truth. This has many varied reasons. We covered several of them in part 2. While data on religious affiliation derived from self-identification is unreliable, it is understandable why organizations keep doing it.
Some of it is going to be intellectual laziness. Polling organizations, and others that rely on this methodology, often do it because it’s easier than having to come up with more reliable methods. Another reason is going to be the self-interest and strong incentives people have to allow an extremely elastic definition for certain identities. Religious ones in particular. An obvious example is the Christian identity itself, considering just how much Mormons, Unitarians, and so-called progressive “Christians” (Distrephists in reality) want to claim the label. These are just a few examples, among many, who would benefit from the social cachet that being labeled “Christian” would bring.
Accurately gauging religious identities is complicated by numerous significant obstacles, separate from the matter of defining heresy versus orthodox belief (which is explored further in part 4). Setting aside the indefensible lack of precision and the conscious ignoring of self-interest in feigned or deceptive self-identification, there remain significant problems that lead to higher costs and fewer available alternatives for any study. Some of these difficulties might even be intractable. Making it impossible to get accurate data, except in very limited ways. At least without some kind of development or clever solution. The problems are as follows:
1) Defining a religion/worldview
One of the very first problems is going to be the act of accurately defining a religion or worldview in the first place. This is not as easy as it might seem. That is because religions contain considerable amounts of teachings, assumptions, history, and philosophical views. Distilling a religion into a succinct definition is not very easy. It is also fraught with issues. One of the biggest is that trying to create a definition for a religion, based on only a handful of sentences (like to fit into a dictionary), will, at best, lead to an incomplete summary. A summary that can actually be misleading if it is taken as the standard for that religion.
These incomplete summaries might be useful to get an idea of a religion, but they will never give you the full picture. They are too broad for precise religious identification. Especially when such broad definitions are deliberately used to encompass groups that shouldn’t really qualify for inclusion. For example, dictionary.com gives this as the definition of Christianity:
The religion based on the life and teachings of Jesus Christ. Christians believe that Jesus Christ is the Messiah, sent by God. They believe that Jesus, by dying and rising from the dead, made up for the sin of Adam and thus redeemed the world, allowing all who believe in him to enter heaven. Christians rely on the Bible as the inspired word of God. (references to links removed)
This definition isn’t wrong, per se, but it is very incomplete. To the point that, if it were the standard, it would include groups that remain entirely outside the realm of true Christianity. Ironically, despite just how incomplete it is, it still contradicts beliefs that several supposed “Christians” claim to hold. Remember the Pew data? According to that, a full 12% of Christians say they don’t believe in the existence of heaven. Despite even having such an incomplete definition, it still ends up ruling out several people who get counted under this religious identification. Researchers include them anyway, which showcases just how misleading these studies can be.
Consequently, a dependable framework of criteria and definitions for various religions and worldviews will be considerably more extensive. That makes the problem much harder. Because it means you have to create a far more comprehensive set of doctrines, teachings, beliefs, or philosophical views. All of which need to be used to build well-defined criteria for all the different religions. This requires acknowledging the sheer scope of religions and worldviews, and organizing them with a greater number of detailed metrics and sophisticated algorithms, rather than relying on a brief question or definition. This rules out simple polls that do nothing more than provide a list of names and ask people to check a box for one they like.
2) Understanding a religion/worldview
Defining a religion is a problem all its own, one that leads into the next related reason. To accurately identify religious or worldview affiliation, you must understand those religions and worldviews at a level that goes beyond the superficial. Knowing only a religion’s dictionary summary means you aren’t qualified to perform accurate assessments of who belongs to that religion or what beliefs it is composed of. This severely limits accurate surveys of religious identification because it will require enough people who have enough expertise on the religions in question to provide the right set of criteria for religious identification.
This is an obvious problem found in a lot of mainstream media discussing religious affiliation. Often, the people writing these articles don’t have the slightest clue what a religion may or may not encompass. Worse still, they delude themselves into thinking they grasp the subject, but their overconfidence masks their lack of knowledge. This is an especially prevalent problem in the West regarding Christianity. It is extremely easy to find articles talking about Christian religious affiliation, but ones written by people whose understanding of Christianity is so distorted that it is obvious they don’t have any idea what they are talking about.
Understanding what Christianity is, what Islam is, what Buddhism, Hinduism, or Humanism are requires a level of research, time, and effort that cannot be bypassed. This is, of course, a serious obstacle to getting accurate data because it means that a level of expertise in many broad areas is required. Without this expertise, people rely on misleading heuristic definitions plagued with ignorant biases, preconceived notions, and outright falsehoods. That is one of the obvious problems behind the Pew research data. It’s obvious that many of the people who answer the questionnaire don’t have any idea what many of the religions listed actually are. Even when they identify themselves as part of it. The fact that atheists claim to believe in a god is a perfect example of such ignorance. This difficulty remains unavoidable, and polls or analyses offering merely a simplistic and incomplete summary of a religion or worldview cannot be trusted.
3) Working around self-interest
I have repeatedly mentioned the fact that self-identification is unreliable. But if we can’t ask people directly how they identify religiously, that means they have to be tested. This means you cannot ask only simple questions or rely on answers that are just checking boxes. Even if you had a perfectly acceptable definition of a religion, and provided these definitions, people cannot be trusted to answer honestly. Nothing guarantees people read these definitions, agree with them, nor answer truthfully. Self-interest, social incentives, and personal psychological incentives are too strong to guarantee reliable data.
This problem becomes even more difficult if you actually provide extensive, accurate criteria for all religions. Going beyond an incomplete, though accurate, summary definition of a religion means providing a considerable amount of text and reading for anyone being surveyed. You can’t accept people’s answers if in the end all they do is check boxes, with no validation. Even if you have accurate definitions and sufficiently comprehensive explanations of the religions or worldviews in question.
To reliably work around self-interest, people have to be put to the test and asked a rather large number of questions. This requires including mechanisms to catch people who answer in ignorance or who answer dishonestly. Implementing this will make any survey balloon into a very extensive test that has several deliberate guardrails built into it. Turning a questionnaire with a few dozen questions into an extensive test with hundreds of questions. The test would have to be designed specifically with the possibility of dishonesty or ignorance in mind and work to rule out false religious identification.
That’s just the problem as it relates to designing the survey or test itself. That doesn’t include dealing with the self-interest of people or organizations who have influence on the testing or design of the test. Without a doubt, there will be controversy in defining a religion, or how such criteria will affect the outcomes of any surveys or studies using such a rigorous methodology. That is because people have considerable incentive to change the test in order to generate outcomes that align with their own religion or worldview. A perfect example of that would be pseudo-Christian cults that need and want to be recognized as truly Christian (such as Mormons, Jehovah’s Witnesses, or Unitarians, as I mentioned prior). Another example would be so-called “non-believers” who have incentives to be viewed as “non-religious” (dealt with here).
Therefore, social pressure and individual or group self-interest are factors that have to be worked around to test accurately. Of all reasons, this is the one that is least recognized and the one most important to have solutions for. Because religious identification is not just a data point. It is a point of leverage and a weapon used to support desired narratives. Though it is one of the toughest problems to deal with.
4) Logistics
Considering the aforementioned problems leads to the idea that they could be effectively resolved by developing a comprehensive process. This process would utilize a meticulously crafted set of standards to ascertain religious identity. Something like an extensive exam that has many questions designed to test a person’s knowledge, beliefs on many topics, and lifestyle. You can rule out a good chunk of cheating or dishonesty by doing an interview with a trained interviewer. The interviewees could be required to provide documentation, or other evidence, of certain claims to catch people who would lie or exaggerate. Such as following up with a church, mosque, or synagogue (or equivalent) to ask about their attendance to validate claims they make about their religious life.
The issue with this is obvious. The complexity and cost of doing such a survey are much higher than doing a basic phone, internet, or mail survey. Collecting accurate data in this way would involve the logistics of having to design the test in the first place, having enough staff and time to perform interviews, the equipment for properly recorded interviews, and the time and effort taken to validate any provided evidence or documentation. Beyond the logistics of the survey itself, you would need to take legal precautions and deal with privacy issues. If you are collecting extensive personal information, documentation, or evidence to validate personal behavior and beliefs, or even recording interview results, you must do all of this in accordance with privacy laws.
Doing these things makes large-scale data collection significantly harder. The Pew data has thousands of survey results because they are relying on a relatively simple paper or electronic survey that uses 100+ questions with mostly multiple-choice answers. Doing interviews of nearly 37,000 subjects would be an enormous project costing a considerable amount of money. Realistically, you would have to reduce the sample size by a few orders of magnitude. Which would reduce the usefulness of the survey.
A multiple-choice survey or a more detailed written survey could be used. However, achieving comprehensiveness and covering a wide range of religious identifications might necessitate a survey with thousands of questions. It would need to test people in order to rule out self-identification and to mitigate dishonest or ignorant answers. While not every participant must answer every question, the survey must incorporate a process for directing individuals to relevant questions and enabling them to bypass inapplicable ones. An example of this would avoid questions that differentiate between Shia and Sunni Muslims for non-Muslim individuals.
There are a lot of logistical hurdles to overcome in order to get reliable religious identification. Testing religious identification and collecting reliable data is not as easy as giving out surveys that just ask people what they claim to be. It may be understandable why organizations don’t do this, but it isn’t excusable. Instead of accepting low quality, misleading polling data, we need to demand higher standards. Especially when people are trying to sell narratives based on this data.
In the next part, we cover major theological issues as they relate to religious identification and the nature of self-identification. Especially as it relates to Christianity. We need to stop pretending that the data we have is accurate in determining the religious makeup of our society. Any analysis that relies on uncritical self-identification is not to be trusted. It just isn’t based on trustworthy data. People simply cannot be trusted to self-identify accurately or honestly. To pretend otherwise shows ignorance or even manipulative behavior. Scripture even makes such things clear. The mainstream media is guilty of this kind of lazy, disingenuous “analysis” because they are propaganda outlets with a deliberate narrative they wish to push. We, on the other hand, need to see through the facade. It is not a trivial or academic matter.