The open-ended interpretation argument, however, is more interesting. It does not show that there are no propositions in these cases; but it does show that there is much more going on than a purely propositional account could allow. And one of the redeeming factors of Atran's account is that he recognizes this and does not (as many might be tempted to do) take this as a sign of their inferiority. There are, of course, a lot of perfectly meaningful statements that are clearly non-propositional (commands, for instance), and which could be called quasi-propositional; a theory of religion that doesn't recognize their importance, both in the field of religion and in other fields, will clearly be flawed. I would suggest that we regard Atran's account, i.e., of all counterintuitive religious beliefs as quasi-propositions, as based on an equivocation between religious beliefs and religious doctrines. The equivocation is very easy to make, and is quite common. I've attempted to rough out what would be involved in religious doctrine, and that was the point of the previous posts in this series. And while it might well be that the account I've given needs adjustment and perhaps even reworking, I would suggest that something like this not only fits the facts to which Atran appeals but also fits better with most of the conclusions he actually wants to draw than his own quasi-propositional account, which is vague, poorly supported, and not even clearly consistent.
Take for instance, this interesting passage from the conclusion of the work, discussing the pedagogical use of the solar system analogy in science:
Science aims to reduce the analogy to a factual description, where the terms of the analogy are finally specified, with no loose ends remaining and nothing left in the dark: atoms are scientifically like solar systems if and only if both can be ultimately derived from the same set of natural laws. Whereas science seeks to kill the metaphor, religion strives to keep it poetic and endlessly open to further evocation. In the case of religion, these metarepresentational ideas are never fully assimilated with factual and commonsensical beliefs. They are always held metarepresentationally: they are displayed, discussed, interpreted, and reinterpreted as doctrines, dogmas, sacred texts, or "norms" that further illustrate beliefs and bheaviors rather than describe beliefs and behaviors. The fact that religious beliefs do not lend themselves to any kind of clear and final comprehension allows their learning, teaching, exegesis, and circumstantial application to go on forever. (p. 277)
Now, if this is taken as about religious beliefs, it isn't a particularly reasonable thing to say; the parallel is thrown off, because the scientific analogy would be taken pedagogically (how it illustrates beliefs) and the religious case would be taken non-pedagogically (as belief itself). Actual religious beliefs do, in fact, lend themselves to clear and final comprehension and their "learning, teaching, exegesis, and circumstantial application" don't go on forever: they are quite definite, vary from person to person, are not reinterpreted (beliefs really aren't the sort of things that admit of reinterpretation). But let us take 'beliefs' here as a loose term for doctrines. Then much of the statement makes sense. As I've noted before, religious doctrines are pedagogical; thus the parallel works. Religious Core Symbols are, indeed, like the solar system analogy in how they function, but the difference Atran claims gives us the right idea about them: Core Symbols aren't ever fully assimilated to factual and commonsensical beliefs; they are "displayed, discussed, interpreted, and reinterpreted as doctrines, dogmas, sacred texts, or 'norms' that further illustrate beliefs and behaviors rather than describe beliefs and behaviors"; and it is true that they don't lend themselves to final comprehension and that this means that "their learning, teaching, exegesis, and circumstantial application" can go on indefinitely.
One can do something similar with other passages in Atran; and, except in a few cases which conflate belief and doctrine in a way that can't be saved, they make much more sense understood this way. Thinking in terms of religious doctrine rather than religious belief has the added advantage of overcoming a weakness of Atran's quasi-propositional account, namely, that Atran is far too quick to deny that religious beliefs admit of inductive and logical support. Any reading of Avicenna, Aquinas, Scotus, Maimonides, Teresa of Avila, Spinoza, Butler, or any number of others will show that the evidence is not so clearly on the side of Atran's conclusion. Looking at the matter from the perspective of religious teaching rather than belief avoids the dubious claim about verification, and the extremely dubious claims that it would require us to make about the history of theology and philosophy, e.g., that (say) Scotus was simply confused in thinking he was dealing with propositions. At the same time, it allows us to see more clearly how many of Atran's claims (which are often plausible and reasonably supported) can be true even given the existence of religious propositions about supernatural agents.
All of this is only a suggestion, but I think that if we look at the matter this way, we can both overcome a number of weaknesses in Atran's argument and accept the basic points he wants to make.
