More Templetonian spin: Scheitle

Over the past few years, I’ve learned to read Templeton-funded sociology articles very closely. This has come in handy in several recent cases, including a new paper by Christopher Scheitle, “U.S. College Students’ Perception of Religion and Science: Conflict, Collaboration, or Independence? A Research Note.”* The paper suffers from an incorrect framing of the question, a misleading presentation of the data, and – not surprising given the first two – unsupported and incorrect conclusions.**

I’m going to quote the full introduction as it illustrates so many of the problems with the current foundation-flush sociology of religion:

The public and scholars alike have long been interested in the relationship between religion and science. The primary question has centered on whether these two institutions are waged in a conflict over their respective claims to truth and sociopolitical authority, or are they independent from or even in collaboration with each other (Evans and Evans 2008)? A popular strategy among social scientists to evaluate this question has been to assess whether scientists are less religious than nonscientists (Ecklund 2008, 2010; Ecklund and Scheitle 2007; Gross and Simmons 2009; Larson and Whitman 1999; Leuba 1916, 1934; Stark 1963). The assumption is that, because they are the most knowledgeable about scientific matters, scientists will be most likely to demonstrate some conflict with religion if such a conflict exists (Wuthnow 1989:143). If scientists are less religious that nonscientists, then the inference has been that there is an inherent conflict between scientific knowledge and religious belief.

As Ecklund and Park (2009:280) point out, such inferences do not tell us much about whether individuals actually view the relationship between religion and science as one of conflict. Such perceptions may be more important than any association between scientific knowledge and religious belief, as it is individuals’ opinions about the relationship that will play a significant role in public debates. Regardless of individuals’ personal religiosity or scientific knowledge, how they approach the relationship between religion and science could have important consequences in schoolrooms, courthouses, and legislatures. Presented here is research on how undergraduates, some of whom will serve as leaders within those forums, perceive the relationship between religion and science. Utilizing longitudinal data from the Spirituality in Higher Education Project (SHEP), a nationally representative survey of undergraduates, I examine the association between students’ religiosity and field of study and their view of the religion and science relationship.

The conflict narrative of understanding religion and science has often been the driving force in scholarly and popular discussions (Evans and Evans 2008; Russell 1997:7–18). The assumption is that religion and science each make claims about reality or truth and, because their respective claims often differ, they must be in conflict with each other. This conflict has both personal consequences as individuals are forced to choose one version of the truth (Russell 1997:7–18), as well as social and political consequences, of which the trial of Galileo, the Scopes Monkey Trial, or the more recent Dover School Board Intelligent Design Trial (Slack 2008) are often offered as examples.

While the conflict framework often receives the most attention, others have claimed that religion and science are not in conflict because they address fundamentally different types of truth. Quoting Cardinal Baronius, Galileo argued for this independence perspective when he said that “the intention of the Holy Ghost is to teach us how one goes to heaven, not how heaven goes” (Barbour 1997:14).1 One of the most famous and eloquent explanations of this independence position was provided by Stephen Jay Gould in his writing on “nonoverlapping magisteria” (Gould 1998). Gould argues that the lack of conflict between science and religion arises from a lack of overlap between their respective domains of professional expertise—science in the empirical constitution of the universe, and religion in the search for proper ethical values and the spiritual meaning of our lives (1998:271).

This introduction is a giant, hulking strawman that confuses several distinct questions - epistemic, institutional, and individual.

First, there is the question of an epistemic conflict. Accommodationists use a variety of means to evade this fundamental issue, and one is to attempt to reframe the question in psychological or sociological terms: if people can be good scientists and religious, if they don’t recognize a conflict, or if religious people or organizations can coexist with science, then the epistemic conflict doesn’t exist. Another twist comes when they then claim, as here, that the noncompatibilist argument the reverse of this: if it can be shown that scientists at some place or time are more likely to be irreligious (or that the majority are), that people recognize this conflict, or that it plays out in the form of religious people or organizations interfering with science, this is evidence of the epistemic conflict. (Scheitle here uses “assumption” and “inference” liberally to imply that this is the gist of the incompatibilist case.)

But this is a misrepresentation. The epistemic conflict exists. That’s what we’re saying, and there’s no question about it. We’ve spelled out the nature of this epistemic incompatibility many times, and its substance continues to be ignored or evaded by accommodationists at every turn. The evidence for the conflict is the nature of science and faith themselves; its existence is entirely independent of the level of its recognition, including among scientists. It wouldn’t make sense to argue that scientific acceptance of the conflict is the evidence for it, as this varies in line with a number of social factors.

Sociology and psychology can never answer the question of whether an epistemic incompatibility exists. This would be fundamentally fallacious: an argument from popularity or authority. However, based on what science and religion are and the existence of this fundamental incompatibility, we would expect there to be observable effects at the individual and institutional levels. Because this epistemic conflict exists in reality, the prediction is that greater scientific understanding and knowledge will result in decreased religiosity and a greater likelihood of recognizing the conflict (of course falling on the side of science). This is similar to the acceptance of anything in science with the weight of the evidence behind it: heliocentrism, the germ theory of disease, AGW,…

Of course, there are differences across social contexts in terms of contravening social factors and opposing campaigns, levels of ignorance, and so on that will affect levels of recognition amongst the general population and scientists. But in general, because the epistemic conflict is real, those with more scientific knowledge and understanding would be expected to be more likely to recognize the incompatibility, reflected in both reduced religious belief and explicit recognition, and the level of increase should advance along with the advance of knowledge (at both the individual and the social level). As far as I know, this is supported by research about scientists, including that which Scheitle cites in his paper.

In terms of institutional conflict in the US, some conflict is obviously in evidence, both historically and in the contemporary era, as Scheitle recognizes. Its causes and variation over time and space can be investigated, but the fact is that there is conflict between religion and science (and science education) in the US. You don’t need a survey of college students’ perceptions to see this.

So, contrary to his framing in the introduction (which, oddly enough for a sociology paper, does not present the research as responding to a sociological question or sociological literature), the research reported in the article says essentially nothing about the existence of the epistemic (or institutional) conflict between science and religion. The only “conflict narrative” or “conflict framework” relevant to a survey-based sociology article concerns people’s perceptions of a conflict: a claim or findings from previous research showing that people in the sciences are more likely to recognize the incompatibility of science and religion, or research on these perceptions in general.

So the questions Scheitle’s article does respond to that are at all relevant are: Are freshman science students in the US more likely to be more secular in general and to hold a conflict-science perspective? By their junior year, are they more or less likely to do so? Do college students in STEM or other politically relevant fields tend to move in a more secular direction, and particularly does the percentage of those with more religious-dominating views threatening to secular governance or education decrease?

I’ll leave aside the less relevant aspects of the research (including criticisms about the combining of the two “nonconflict” views and the ignoring of the apparent significant correlation with SAT scores) and discuss the data Scheitle presents concerning these questions. As he doesn’t present it in a form that allows us to see the important patterns, I’ve had to work with it a bit. [I have the tables, but don’t know how to reproduce them here. :/]

The data show, unsurprisingly, that students – including STEM students - holding the conflict-science view in the US are not a majority. Also clear is that, as predicted from the fact of epistemic incompatibility (again, the evidence of conflict is not sociological, but, duh, epistemic), STEM students, even at this early stage, are the most likely to hold a conflict-science perspective, and the least likely to hold conflict-religion views. As Scheitle acknowledges:

Overall, natural science students are relatively low in holding a conflict perspective, but they are among the highest in holding a pro-science conflict perspective. A little over 20 percent of natural science students “side” with science. Only engineering and mathematics students report a higher allegiance to science in a perceived conflict with religion. (179-80)

So what changes between freshman and junior years? Scheitle:

…27.4 percent of students who held a pro-religion conflict perspective in their freshman year still held this perspective in their junior year. A little over 70 percent of these students now said that they view the religion and science relationship as one of independence or collaboration. Looking at the pro-science conflict column, we see that students holding this stance in their freshman year are more stable in their view than the pro-religion students, as 53.2 percent did not change their opinion between the two surveys. However, of the 46.8 percent that did have a change of opinion, 45.9 percent moved to the independence or collaboration perspective and only 0.9 percent moved to the pro-religion side. Students with the most stable opinion are those who held an independence or collaboration perspective in their freshman year, as 87.0 percent of these students held the same opinion in their junior year. Those who moved away from this opinion were fairly evenly split between the pro-religion and pro-science groups, with 5.2 percent and 7.8 percent moving to these groups, respectively.

To summarize these changes, we can say that few students move from viewing the relationship between religion and science as one of independence or collaboration to viewing it as one of conflict. The more common change is from a conflict perspective to an independence or collaboration one. College and/or aging seems to temper the views of those who held a conflict perspective. It is also worth noting that pro-science conflict views tend to be more entrenched than pro-religion conflict views. (182)

STATISTICS ABUSE! The discussion has now collapsed the conflict-science and conflict-religion perspectives, which is telling and confounding. But this business of talking about the percentages of the percentages that changed is extremely obfuscating.*** What we want to see is the net change overall across groups and across disciplines. Scheitle does not provide this, but it can be constructed from the data presented in Tables 2 and 6.

If we ignore the trivial percentages switching from conflict-science to conflict-religion and vice versa, we see conflict-science and I/C percentages remaining quite stable in STEM fields. Overall, we’re looking at a net loss from conflict-science of a handful of percentage points (4.9%) in the natural sciences and a mere 1.3% in M&E. In the social sciences, we see a tiny increase (1.5%) in the conflict-science view. (A similar increase, of 2.5%, can be observed in the Arts & Humanities.) Now, I doubt these small net changes have any real significance in one direction or another. If a series of surveys showed such small percentages of decline over an extended period, it would be different, but what this suggests is stability over time.

The key lines to look at in Table 6 in evaluating Scheitle’s claim that “few students move from viewing the relationship between religion and science as one of independence or collaboration to viewing it as one of conflict. The more common change is from a conflict perspective to an independence or collaboration one. College and/or aging seems to temper the views of those who held a conflict perspective” are: “started pro-science and switched to I/C” and “started I/C and switched to pro-science.” Again, the picture is one of stability, with small net losses of conflict-science to I/C in STEM fields (-4.6% in the natural sciences, -1.2% in M&E). (There are gains of 1.8% from I/C in the social sciences and 2.5% in Arts & Humanities.) Thus, this “tempering” notion is incorrect with regard to conflict-science views, at least among students in sociologically relevant fields. (I haven’t the time or energy to look at every field, but you can simply compare these “switched to” lines to get a sense of the broader patterns and variation.)

Things are slightly different when we look at movement between conflict-religion and I/C perceptions. In the natural sciences (again ignoring the tiny percentage that switched from conflict-science), the percentage that begins conflict-religion is 9.5%, but by junior year is down to a miniscule 1.7%. The conflict-religion view loses a net 6.4% to I/C in the natural sciences and 3% to I/C in M&E (of course, it was only 7.5% in M&E to begin with), and 8.4% to I/C in the social sciences and 12.1% to I/C in the Arts & Humanities. In education, conflict-religion loses a whopping 22% to I/C (compared to a 1.8% net loss for conflict-science). The drops in STEM and even the social sciences could be simple variation, but the double-digit losses in A&H and education are substantial.

In his conclusion, Scheitle argues:

The predominant narrative surrounding the religion and science relationship has been driven by the assumption that these institutions are engaged in an unavoidable conflict resulting from their contradictory claims to truth (Evans and Evans 2008). However, the analysis conducted above found that most undergraduates, regardless of their area of study or even their religiosity, do not hold a conflict perspective. Furthermore, many more students move away from a conflict perspective to an independence/collaboration perspective than vice versa. This finding might be especially surprising since many people, especially religious families, assume that higher education has a secularizing influence on students (Smith and Snell 2009:248), which might be expected to increase perceptions of a conflict. Despite its seeming predominance, the conflict model of understanding religion and science issues does not seem to have much support within the undergraduate population. Ecklund and Park (2009) made a similar conclusion in their analysis of the views of academic scientists.

Still, some of the patterns seen in the analysis above might be disconcerting for those looking to move beyond the public battles for power between religion and science. The finding that scientists and engineers are among the most likely to have a pro-science conflict perspective could mean that some of the most influential voices in these public debates might be more likely to fuel the debates than attenuate them. Similarly, future educators are among the most likely to hold a pro-religion conflict perspective. Given that classrooms and school boards have been one of the central forums for the struggle over religion and science, this does not bode well for a reduction of those struggles. (185)

But this is ridiculous. First, a survey of college students can tell us essentially nothing about epistemic or current institutional conflicts. The first isn’t a sociological question, and the second involves a completely different sort of analysis. The idea of a “predominant conflict narrative” that comprises all of these and can be investigated via surveys is silly. Second, who is arguing that the majority of US college students in any field currently recognize the incompatibility of religion and science? This is what he appears to be arguing against, but I don’t know that anyone has made this claim or suggested that it would be evidence of anything.

The claim that follows “Furthermore,…” is misleading and basically wrong. As we’ve seen, when the two conflict views are separated, Scheitle’s own evidence supports a general view of secularization. While conflict-science views hold steady, conflict-religion views – in the most relevant fields – decline. This broadly supports the secularization argument, as secularization is consistent with any movement away from the conflict-religion view (including, probably, that from collaboration to independence, though since Scheitle collapses these we don’t have that data).

As should be clear from what I’ve argued above, “Despite its seeming predominance, the conflict model of understanding religion and science issues does not seem to have much support within the undergraduate population” is silly and wrong. If “the conflict model” means epistemic conflict, a study of college students’ views has no bearing on its truth (though their views and changes therein generally conform to incompatibilist expectations). If it means institutional conflict, this study doesn’t speak to it. If it means individual perceptions of conflict, it depends on what you mean by “support.” More STEM students perceive a conflict, and this percentage holds pretty steady from freshman to junior years. At the other end, it looks like some larger percentages of people with conflict-religion views have moved in recent years in a secular direction. The percentages going directly from the conflict-religion to conflict-science are (as expected) miniscule, but the percentages moving in a secular direction are rather larger, and in some cases (e.g., education) substantial. It’s a direction. Will percentages keep shifting in the conflict-science direction or will they be “stuck” in independence or collaboration? I think the former is more likely, but depends on a number of sociological factors. Again, this is sociological data. The fact of an epistemic conflict is independent of it.

With regard to “public battles,” how I loathe this insinuation that pro-science incompatibilist views are damaging. There is a real worry in the US about secular education and politics and the practice of science, and this concerns religious interference in these institutions. The views compatible with secular politics and education are science-conflict and independence. Collaboration is a bit iffy, depending on how it plays out in action. The view that’s obviously the clear threat to education, politics, and science free of religious interference is conflict-religion. It’s quite strange that Scheitle points to the higher percentage of STEM students with conflict-science views as being in any way unfortunate in that they “might be more likely to fuel the debates than attenuate them.” Tough darts, Scheitle. The conflict-science perspective is not only correct but no threat to the constitutionally-mandated separation of church and state or to secularism in general. (To the extent that those promoting it reduce the level of religious belief, it is a great boon to this cause.) If you want to argue against secularism in the sense of no religious interference in these spheres, be my guest.

Scheitle’s conclusions about education students are even more bizarre. Again, the threat to secular education in general and good science education in particular is people, especially teachers and educational administrators, acting in accordance with the conflict-religion view. He suggests that “future educators are among the most likely to hold a pro-religion conflict perspective. Given that classrooms and school boards have been one of the central forums for the struggle over religion and science, this does not bode well for a reduction of those struggles.” But as his research demonstrates, between freshman and junior years, education students show a 22% drop in the percentage of people holding religion-conflict views, reducing the percentage who hold this view to single digits. This is very strong evidence of secularization, and bodes quite well, in fact.

*Journal for the Scientific Study of Religion (2011) 50(1):175–186.

**It’s interesting that this is a bit of research in the sociology of religion that might not appear as “pro-religion” to anyone unfamiliar with the debates in which it intervenes.

*** To offer an extreme example, let’s say they surveyed 100 people, and two had science-conflict views while 98 had no-conflict views. Then they surveyed them a second time a few years later and one from each group had changed her mind. They could say that 50% of the first group and less than 1% of the second had changed their minds. Would this support Scheitle’s interpretation? Sigh.