The Problem with the Primacy of Primary Sources


By Andie Alexander
Andie Alexander earned her B.A. in Religious Studies and History in 2012. She is completing her M.A. in Religious Studies at CU Boulder. Andie also works as the online Curator for the Culture on the Edge blog.

Over the past few weeks I have heard repeated talk of primary sources vs. secondary sources, privileging the former over the latter in every case. The argument that was made in these instances is premised on highlighting the legitimacy and groundedness of the primary sources, as if they focus on something “real,” (such as religion on the ground, or “real world” examples). These are then juxtaposed with the secondary sources, which are seen as subsequent discourses on primary sources, mostly concerned with meta-discourses — regarded mainly as “thinking about thinking” or “talking about talking” — which are given that secondary ranking because they aren’t talking about something “real” or answering the reader’s question (in fact, often they leave them with more questions than answers, whereas primary sources are all about answers). In some cases, these secondary sources are dismissed with a “So what?” I say “dismissed” not with a negative connotation, but with a sense of such indirect relation to the “real world religion” that the “meta-discourses” are seen as not worth the same time and engagement as their juxtaposed primary sources.

What strikes me as interesting about this distinction between primary and secondary sources is that the use of that very language largely relies on some essentialized notions of religion, as if it exists authentically, apart from subsequent claims about it.

For, to me, sources and claims about sources are much the same (they are all instances of human discourse), so to see so-called primary sources as anything more than yet another discourse taking place, is to miss the mark. Primary sources certainly do not represent or depict the “real” any more than the secondary sources. For, inasmuch as we presume them to be human artifacts, primary sources have an argument and methodological approach that is necessarily rooted in certain perspectives and biases.

Consider archaeologists working at a new dig site. How do they begin to determine what is an artifact worth studying and preserving and what is just simply dirt or unimportant material? There is an active decision in the process of setting one broken shard of clay apart from another — that itself is meaning-making. There’s no telling how many pieces of broken pottery one might run across in this process. But what is set apart now becomes a relic. How old does it have to be? Does it have to have a certain shape? These questions are just the beginnings of the active critical work necessary to studying ancient artifacts (as opposed to dirt).

So just because primary sources work with case specific data they are not consequently mirrors of “reality” or religion on the ground. Yet many seem to assume that primary sources are more rooted in reality, a view that necessarily relies on the notion that religion is a “real” thing and is inherently interesting in and of itself. One can certainly find religious practices interesting (or should I instead opt for the more active verb “make”?), and no doubt they are of great importance to their practitioners. But relying on the assumption of importance qua importance seems like shaky ground on which to do productive scholarship (for it means our work is based on obviousness, suggesting that such don’t really have any particular skills other than an ability to recognize what others claim as important).

If we take seriously the argument that religion is not essentially anything — something you’ll hear in a Dept. of Religious Studies ad nauseam — , and therefore is malleable socio-political rhetoric, then why do we fall back on these essentializing discourses to distinguish the sources we study from those that we produce? For, as Merinda Simmons aptly argues,

But that fact is too often forgotten, it seems, even now in this academy wherein talk of “critical theory” proliferates but wherein its implications are curiously absent. Scholars thus do themselves… no favors by… professing their love for it, and calling that progressive academic work. What we are left with, in that case, are dueling essentialisms in the service of respective passions.

While Simmons was referencing the AAR’s 2016 theme of Revolutionary Love, I think the argument applies quite well here. There is always much talk of applying “critical theory” but I rarely see evidence of it when it comes to how we do our own work. Instead, there is a brief acknowledgment of the necessity of so-called critical approaches in scholarship before moving on to studying the “real” stuff.

This way we can have our cake and eat it too — and thereby prevent our critical theories from being a little too critical.

I was recently jokingly chastised for not finding my area of work inherently interesting. Having made the comment that my topic was no more interesting than anything else, I was asked why I was even in the field of religious studies. But the point of my comment was not that I found my work uninteresting — quite the opposite, rather. What I meant to highlight in that remark is that no area of study is inherently or self-evidently interesting. For example, I didn’t choose my thesis project because it is self-evidently interesting, but I instead found that it was a great way in which to explore identity politics, issues of globalization, and the practical effects of classifications — all those meta-discourses that are seen by some as secondary to our “real” work. But wait: if not for those second-order discourses, I would have no way to even begin talking about my research project, much less identify something to study and then develop and carry a project through to fruition. For, arguably, it is these second-order discourses that allow any work to be done. Given this position we’d claim that there is no way to study (much less represent) any sort of data without employing some sort of theoretical approach to that topic. Description is therefore theory — it’s embedded in assumptions. Without them, there would be nothing to describe, much less analyze. So whether one is engaged in rigorous critical analysis or what one might portray as disinterested descriptive work then one is equally engaged in, and producing, secondary sources.

So you can understand why seeing primary sources as somehow different from secondary sources is, to me, a bit puzzling. Case specific data does not equal “real” religion or actual experiences — it is just one of many ways in which those specific data sets can be identified and then discussed.

What’s more, if we recognize that we cannot have the data without the theory, then we might ask what this dichotomy of primary and secondary sources is doing? The way I see it, it allows us to feel like were talking about something real, studying real issues as opposed to acknowledging how unequivocally rooted we ourselves are in these theoretical discourses. However, if we take a step back and instead of essentializing these primary sources, acknowledge that they, too, are just as much discourse on discourse, then we can shift the conversation more to the implications of these discourses rather than their origins, ground, or sources. Rather than teaching data as the product of disinterested description only to then change the game by teaching the complexities of how we only later talk about religion, why not instead start from the latter position and examine how our assumptions make things into religion (or not)? For if — as so many claim — they do not agree with essentialized notions of religion, then why do we still continue to rely on them and teach them in our classes? While it is certainly simpler, I go back to what I said in the start of this piece: relying on the assumption of importance qua importance, seems like shaky ground on which to do productive scholarship (and I’ll add pedagogy here as well.)