The Schematicity of Religious Thought

Human thought is schema-driven and reconstructive. By that I mean that, in most cases, we do not store information in memory in its entirety, but instead store its "gist" (in the form of a schema). When we later need to retrieve that information, we reconstruct it based on its particular schema, along with certain general schemas (e.g., relevant ontological intuitions) and background knowledge. Human conceptual and communicative systems have evolved (culturally and/or biologically) to take advantage of this. For instance, when attempting to communicate novel concepts to others, we naturally ground those concepts in pre-existing schemas (for example, through analogy, or by placing the new concept in a larger conceptual system). This allows people to use their background knowledge to produce inferences about the new concept, and thereby reduces the amount of information that we need to convey about it. Because religious concepts must utilize the same cognitive mechanisms that nonreligious concepts do, we would suspect religions have developed certain common conceptual features that are also designed to take advantage of the schema-driven nature of thought. In this post, I'm going to discuss a few of those features, along with the ordinary cognitive mechanisms that make them useful.

One way of thinking about schemas is as variables with relationships between them. Schemas will tend to be stored with default values for the variables, which serve as expectations For example, our restaurant schema contains a variable for the time when the server will bring the check, and the default value for that variable will be sometime between two other variables, which stand for the time we finish eating and the time we leave. Research has shown that how new information fits with our schemas will determine how well it is remembered, and how easy it is to comprehend and transmit. Information that is relevant to our existing schemas, i.e. information that fits into one of the variable slots, will be more easily remembered. There are two types of schema-relevant information. The first, schema-consistent information, is information that fits with our expectations. When the server brings us our check between our finishing and leaving, this is a schema-consistent instance. In these cases, we tend not to store particulars about the situation, but instead store the instance with the schemas default values. This allows us to reason about the situation using the other default values in the schema, but it also creates situations in which our memories are distorted. When reconstructing the situation from memory, later, we may include default values from the schema that were not present in the actual instance.

The second type of information that is easily remembered is schema-inconsistent information. This type involves values on the same dimensions as the schema's variables, but that differ significantly from the default values. For example, if our server brought us the check at the beginning of the meal, the information would fit into a slot in our schema (the time that the check is brought) but would be very different from the default value. When we encounter this type of information, it demands our attention, and we tend to store it in detail. Thus we will remember particulars about the instance, and are less likely to experience memory distortions. A third type of information, which is schema-irrelevant, does not line up with any of the slots in our schema. This type of information tends to be ignored and forgotten.

Given what we know about memory, we can make some predictsion about the content of religious concepts. As Atran and Norenzayan¹ note, memorability may be one of the most important factors in determining which religious ideas are successfully transmitted over time, and which are not. Thus, religious concepts that survive over time will likely exhibit features that make them easy to remember. These concepts will therefore tend to be grounded in existing schemas (e.g., God concepts will be similar in many ways to ordinary agent concepts), but because they must be remembered and transmitted across generations, they will also tend to differ from existing schemas in schema-relevant ways. In other words, religious concepts will tend to have many schema-consistent properties, along with some schema-inconsistent properties, and few schema-irrelevant properties. In the research conducted so far, this is exactly what cognitive scientists have found.

First there are the schema-consistent aspects of religious concepts. As I mentioned in a previous post, it's important to distinguish betwee "theological" concepts and everyday religious concepts (usually just called "religious"). In theological contexts, concepts often bear little similarity to nonsupernatural concepts. In these contexts, God (or equivalent agents) may be omnipotent, omnipresent, omniscent, and so on, wheras natural agents are limited in power, presence, and knowledge. In religious contexts, however, God is often much more similar to natural agents. As I mentioned in the previous post, in everyday use, people tend to treat God as a spatio-temporal being much like natural agents. Barrett² uses the example of Hindu theologians who understand that Shiva is omniscient, and thus aware of their every thought, but who, in their everyday practice of their religion, still feel that they must pray to him to make him aware of their thoughts.

The two examples of similarities between religious concepts and nonreligious concepts illustrate one of the most striking types of similarity: the athropomorphization of God concepts. People seem to have a strong tendency to attribute anthropomorphic attributes to God in religious contexts that are inconsistent with their theological representations of God. To study this tendency, Barrett and Keil³ conducted a series of experiments designed to show how natural agent schemas influenced our everyday representations of God. First, they had answer a questionaire about God designed to get a picture of their theological representation of God. The authors assumed that the questionaire would cause participants to reflect on their beliefs, and thus produce theologically-correct answers. The answers to the questions were highly consistent across participants, and painted a picture of God that is very different from natural agents (see the table below).



Table 1 from Barrett & Keil (1996). Answers to the questionaire, representing people's theologically-correct representations of God. Click for a larger view.

In the first experiment, after completing the questionaire, participants were asked to read stories, some of which were about God, and some of which were nonreligious. Following a short delay, participants were then given statements and asked to indicate whether they had been in the stories they had read. This type of recognition memory task has been shown to tap into participants' existing schemas, which can distort their representations of the passages. In this case, some of the statements in the recognition task were new statements (i.e., they had not been in the original passages) that implied that God had anthropomoprhic properties inconsistent with the theological representations. Thus, if participants' religious (rather than theological) concepts of God contained anthropomorphic properties, they should mistakenly recognize these anthropomorphic statements has having been in the original passages. Consistent with this, participants' correct recognition rate for the passages about God was about 40%, which was significantly worse than their correct recognition rate for control passages about nonreligious concepts (80%). This large difference in accuracy was due in large part to participants mistakenly remembering the anthropomorphic statements as having been in the original passages.

In a subsequent study, participants were given passages about God and asked to paraphrase them. Their pharaphrased versions of the passages regularly attributed anthropomorphic properties to God that were not contained in the originals, such as limiting God to a single time and place, and thus making God unable to attend multiple events at once, or representing God as needing to be present to know about certain events. These two studies show that while people may be able to produce theologically-correct representations of God through reflection, during ordinary tasks such as reading comprehension, they will tend represent God as more similar to non-supernatural agent schemas.

Even everyday religious concepts are not completely like natural agent concepts, though. In fact, research has shown that people tend to believe that "counterintuitive" concepts, particularly "counterintuitive" agent concepts, are likely to be religious concepts⁴. Religious concepts will not, however, tend to be counterintuitive in arbitrary ways. Instead, they will tend to exhibit the same dimensions (or variables) that ordinary agent concepts do, but have different values on some of those dimensions. In other words, they will differ in ways that are relevant, but schema-inconsistent. Cognitive psychologists studying religious cognition have termed this type of difference "minimally counterintuitive." Studies have shown that minimally counterintuitive religious concepts, while not as easy to remember as intuitive ones, are much easier to remember than extremely counterintuitive ones (i.e., schema-irrelevant concepts)⁵.
Further research has shown that the schema-inconsistencies in religious concepts that are best remembered are those at the "domain-level" (e.g., person, animal, artifact), rather than at the "basic-level" (e.g., dogs, chairs). In paricular, only domain-level inconsistencies that allow for inferences based on the schemas are well-remembered⁶.

Finally, the degree to which our on-line representations of religious concepts include schema-inconsistent properties will depend on several factors. For instance, people who ordinarily use more schematic representations (e.g., images) to represent God will tend to attribute more anthropomporphic properties to him⁷. More interestingly, it appears that cognitive load can influence the degree of "counterintuitiveness," or schema-inconsistency in our online representations. One benefit of schemas is that they allow us to reason about things without having to process a lot of information about the particular situation. Instead, we can simply use the default values in our schema to make inferences during a task. This is particularly important when tasks involve a high cognitive load. In these cases, we have fewer cognitive resources to process information about the particular situation, and schemas allow us to reason in these situations without a significant decrease in accuracy. Consistent with this, research has shown that the higher the cognitive load of a task, the less counterintuitive, and more schema-consistent, our religious concepts will tend to be⁸. For example, while performing tasks that use up a significant amount of our cognitive resources, our God concepts will tend to be more anthropomorphic than in tasks with a lower cognitive load.

The take-home message of all of this is that our religious concepts tend to exhibit the features that we would predict based on what we know about schematic memory. They tend to be similar to our ordinary schemas from the same domains (e.g., God concepts are similar to other agent concepts), but "minimally counterintuitive" in domain-level, schema-relevant ways. The degree to which they resemble or differ from nonreligious concepts is largely dependent upon the context in which they are representd. In contexts in which we can use a lot of our cognitive resources to reason about religious concepts, they may be quite different from nonreligious concepts. However, when performing ordinary cognitive tasks that use up resources, they will tend to be similar to nonreligious concepts. All of these features make religious concepts more memorable, and easy to understand. This in turn makes them more likely to be transmitted and survive over time. In this way, religious systems, like most cognitive systems, have "evolved" to fit the cogntive mechanisms that they must utilize.


¹ Atran, S., & Norenzayan, A. (In Press). Religion’s evolutionary landscape: Counterintuition, commitment, compassion, communion. Behavioral and Brain Sciences.
² Barrett, J.L. (2000). Exploring the natural foundations of religion. Trends in Cognitive Science, 4(1), 29-34.
³ Barrett, J.L. and Keil, F.C. (1996) Anthopomorphism and God concepts: conceptualizing a non-natural entity. Cognitive Psychology, 3, 219–247.
⁴ Pyysiäinen, I., Lindeman, M., & Honkela, T. (2003). Counterintuitiveness as the hallmark of religiosity.Religion, 33, 341-355.
⁵ Atran & Norenzayan (In Press).
⁶ Boyer, P., & Ramble, C. (2001). Cognitive templates for religious concepts: cross-cultural evidence for recall of counter-intuitive representations. Cognitive Science, 25, 535–564.
^{7 4} Barrett, J.L., & VanOrman, B. (1996). The effects of the use of images in worship on God concepts. Journal of Psychology and Christianity, 15(1), 38–45.
⁸ Barrett, J.L. (In Press). Theological correctness: cognitive constraint and the study of religion. Method and Theory in the Study of Religion.