Sunday, May 30, 2010

Stephen Jay Gould on Science and Religion

According to Stephen Jay Gould, science and religion have separate domains of teaching authority, or “non-overlapping magisteria.” If so, then it is not evident that any major revealed religion has ever confined itself to its proper magisterium. But that does not mean that Gould is wrong.

  Jesus and Mo (image linked to site) 

Not long before his untimely death in 2002, Stephen Jay Gould advanced what he described as “a blessedly simple and entirely conventional resolution” to “the supposed conflict between science and religion.”1 “Supposed,” because the said conflict “exists only in people’s minds and social practices, not in the logic or proper utility of these entirely different, and equally vital, subjects.” By nature, according to Gould, science and religion do not and cannot conflict, because their respective concerns are entirely distinct:
Science tries to document the factual character of the natural world, and to develop theories that coordinate and explain these facts. Religion, on the other hand, operates in the equally important, but utterly different, realm of human purposes, meanings, and values—subjects that the factual domain of science might illuminate, but can never resolve (4).
Gould adopted the Catholic ecclesiastical term “magisterium” (from the Latin magister, “teacher”) to describe these spheres of concern. “A magisterium . . . is a domain where one form of teaching holds the appropriate tools for meaningful discourse and resolution” (5). Science and religion, on Gould’s view, are “non-overlapping magisteria”—a phrase that, perhaps for reasons of euphony, he abbreviated to “NOMA” (surely “NOM” would have been more accurate). The magisterium of science comprises “the empirical realm: what the universe is made of (fact) and why does it work this way (theory),” while that of religion “extends over questions of ultimate meaning and moral value” (6). No question belongs within both domains: hence their “non-overlapping” character.

Gould’s thesis has not been well received, as far as I know, among defenders of science. Members of this audience have argued that the thesis rests on a highly questionable dichotomy of fact and value; that it grants religion a dubious and undeserved authority in questions of meaning and value; and that it does not describe any state of affairs that has ever actually existed between science and religion.2 In what follows, I will mostly be concerned with the last of these three criticisms, though I will have a bit to say about the second one toward the end.

Gould is well aware that many people have invoked and continue to invoke religious sources to make claims about the character of the natural world, as well as scientific sources for claims about meaning and value in human life. He notes that “NOMA does challenge certain particular (and popular) versions of religious belief . . . For example, if your particular form of religion demands a belief that the earth can only be about ten thousand years old (because you choose to read Genesis as a literal text, whatever such a claim might mean), then you stand in violation of NOMA” (93). The thesis of non-overlapping magisteria concerns the domains within which religion and science have their respective authorities, not the scope of actual claims that human beings make in the name of the one or the other.

What are commonly adduced as examples of the conflict between religion and science, such as the persistent conflict over the teaching of evolution in American public schools, are, according to Gould, typically political clashes between one group representing the interests of a specific religious group and an opposing group representing not merely scientific but very often opposing religious interests as well. In fact, Gould takes the battle over creationism in American public schools to illustrate rather than to counter his thesis:
Modern creationism, alas, has provoked a real battle, thus supporting NOMA with a positive example of the principle that all apparent struggles between science and religion really arise from violations of NOMA, when a small group allied to one magisterium tries to impose its irrelevant and illegitimate will upon the other’s domain. Such genuine historical battles, therefore, do not pit science against religion, and can only represent a power play by zealots formally allied to one side, and trying to impose their idiosyncratic and decidedly minority views upon the magisterium of the other side (125–6).
Similarly, social Darwinism—the real object of William Jennings Bryan’s opposition to the teaching of evolution in public schools—was an ethical and political view based on an illegitimate inference from how nature works to how human beings should conduct themselves (162–3, 165–6).

As I said before, Gould’s thesis has not been well received among defenders of science. Thus Massimo Pigliucci offers the following as one of “several intrinsic reason why NOMA does not hold water”:
It is not true that (most) religions do not make claims about the natural world. Besides the tens of millions of people who believe the Earth is 6,000 years old, the Bible was never meant as a book of metaphors. It is read that way by enlightened Christians today precisely because of the long battle between science and religion, with the latter constantly on the losing side. (“Gould’s Non-Overlapping Magisteria, A Review”)
Richard Dawkins, in a review of Gould’s book, writes in a similar vein with reference to certain doctrines of the Catholic Church:
The Virgin Birth, the bodily Assumption of the Blessed Virgin Mary, the Resurrection of Jesus, the survival of our own souls after death: these are all claims of a clearly scientific nature. Either Jesus had a corporeal father or he didn’t. This is not a question of “values” or “morals”; it is a question of sober fact. We may not have the evidence to answer it, but it is a scientific question, nevertheless. You may be sure that, if any evidence supporting the claim were discovered, the Vatican would not be reticent in promoting it. (“When Religion Steps on Science’s Turf”)
And Daniel Dennett, responding to Gould’s thesis, says in an interview:
There are no factual assertions that religion can reasonably claim as its own, off limits to science. Many who readily grant this have not considered its implications. It means, for instance, that there are no factual assertions about the origin of the universe or its future trajectory, or about historical events (floods, the parting of seas, burning bushes, etc.), about the goal or purpose of life, or about the existence of an afterlife and so on, that are off limits to science. After all, assertions about the purpose or function of organs, the lack of purpose or function of, say, pebbles or galaxies, and assertions about the physical impossibility of psychokinesis, clairvoyance, poltergeists, trance channeling, etc. are all within the purview of science; so are the parallel assertions that strike closer to the traditionally exempt dogmas of long-established religions. You can’t consistently accept that expert scientific testimony can convict a charlatan of faking miracle cures and then deny that the same testimony counts just as conclusively—“beyond a reasonable doubt”—against any factual claims of violations of physical law to be found in the Bible or other religious texts or traditions. (“Daniel Dennett’s Darwinian Mind: An Interview with a ‘Dangerous’ Man”)
The common argument here seems to be, in briefest form, that because the actual beliefs of most religions include matters of natural fact, most religions intrude upon the magisterium of science; therefore science and religion do not have non-overlapping magisteria.

It is not difficult to see how Gould could parry this objection. His thesis, as he says in a passage already quoted, does not concern the actual practices of the world’s religions but rather “the logic or proper utility of these . . . subjects” (3). It concerns the respective domains in which religion and science hold the means of legitimately answering questions, not the domains in which people do in fact invoke religion or science to answer questions.

Such a reply, however, merely displaces the force of the criticism to another point in Gould’s conception. How plausible can a conception of the “logic and proper utility” of religion be if it implies that the greater part of the world’s religions overstep the proper bounds of religion itself? As Pigliucci suggests in the passage quoted earlier, it is only because religion, as represented by its more enlightened adherents, has been beating a retreat in the face of scientific advance for the past 400 years or so that Gould’s conception of its proper sphere has even an appearance of being workable.

Now there is one part of Gould’s thesis on which none of these arguments cast any doubt, namely that religion has no teaching authority with regard to questions of natural fact. Pigliucci, Dawkins, Dennett, and Gould are of one mind on this point. The trouble for the thesis of NOMA is that, at least historically, most of the adherents and most of the authorities of the world’s religions have received and propounded teachings about such questions, and proponents of science have continually encountered resistance from religious quarters whenever their findings came into conflict with those teachings. One may miss this point if one confines one’s attention to questions of natural-scientific theory. In such matters, one may take it for granted that, apart from the views of those whom Gould rightly disparages as a minority of zealots, religion, for the most part, got out of that line of work long ago. But, as noted by Dawkins and Dennett, the questions on which religion has had to retreat include highly specific questions of human history, such as questions about the life of Jesus or, to return to the topic of one of my previous posts (“Three Kinds of Religious Beliefs”), the supposed exodus of the Israelites from Egypt and the supposed reception of the Torah by Moses at Mount Sinai.

To be sure, Gould is able to cite pronouncements by Popes Pius XII (Humani Generis, 1950) and John Paul (“Message to the Pontifical Academy of Sciences,” 1996) as illustrations of NOMA (76–82). There are plenty of Protestant denominations (I think the majority, though I don’t know) that have officially accepted the findings of modern science, evolutionary biology in particular. The liberal denominations of Judaism—Reform, Conservative, and Reconstructionist—have certainly been receptive to the findings of modern science, even, indeed perhaps especially, in questions of ancient history and the origins of Judaism’s sacred texts (see my post “What Beliefs Are Jewish Beliefs?”). On the other hand, Modern Orthodox Judaism, as I understand it, professes to accept the findings of science, but whether its rabbinate actually does so in practice is another matter (see my post “Dishonesty in Hertz’s Torah Commentary”).

In his interview, Dennett cites an unnamed or unknown “wag” who said that Gould’s thesis “amounts to rendering unto Caesar that which is Caesar’s and unto God that which Caesar says God can have.” In other words, NOMA leaves science in charge of questions of how the world works, and leaves religion some subset of the leftover questions that science cannot answer. This subset, over time, has been getting smaller and smaller. As noted earlier, Gould’s critics have attacked his attempt to identify it with the set of questions concerning meaning and value in human life. “Philosophy,” says Pigliucci, “using the tools of logic and informed by the discoveries of science, seems to me a much better candidate for that magisterium.” I am inclined to agree; but I suspect that this is merely a point on which Gould’s thesis was underdeveloped, not a point on which it was flatly wrong.

Jesus and “Mo,” in the cartoon shown above, may be closer to the mark, or at least less susceptible of refutation, when they identify the proper sphere of religious claims as “the supernatural.” In an earlier post (“The Natural versus the Supernatural”), I identified the supernatural as a putative order superior to that of nature. Setting aside for the present the questions of what beliefs a human being might hold regarding such an order, how and why he or she might do so, and by what observances he or she might enter into relation with such an order, I am, at any rate, content to say that religion has its magisterium, which is the supernatural. The natural is the magisterium of science: a religion may contain beliefs regarding natural matters, but it has no authority with regard to them.

The sum of my consideration of Gould’s arguments and those of his critics is that the thesis of NOMA remains defensible, but implies, at least in historical terms, a radically revisionary and restrictive conception of religious authority. The “proper” sphere of religious claims is much, much smaller than what any of the Abarahamic faiths has historically claimed for itself, as it excludes most of what is narrated in scripture. Every religion is entitled to its own version of history, of course, but it cannot claim specific authority for versions of events that are unsupported by the available evidence, much less ones that are contrary to such evidence. On such a view, revelation can extend no further than a supposed supernatural realm. I can live with that, but I don’t know what proportion of the world’s religious believers can do so.


1Stephen Jay Gould, Rocks of Ages: Science and Religion in the Fullness of Life (New York: Ballantine Books, 1999), 3. Gould presents his thesis more briefly in his essay “Non-Overlapping Magisteria,” Natural History 106 (1997).

2Such criticisms may be found in Richard Dawkins, “When Religion Steps on Science’s Turf,” Free Inquiry 18.2 (1998); Massimo Pigliucci, “Gould’s Non-Overlapping Magisteria: A Review” (source unknown); and Chris Floyd, “Daniel Dennett’s Darwinian Mind: An Interview with a ‘Dangerous’ Man,” Search Magazine, May/June 2000.

Previous entry: Tom Tomorrow on the BP Oil Disaster

Next entry: More on Gould on Science and Religion


  1. All of this of course does not answer what exactly religion is supposed to do in its magisterium. Does religion really tell us anything about purposes, meanings, and values? As far as I can tell all Judaism really does is tell us what to do. Maybe that's under the category of purposes?

    Also does Gould discuss the historical origins of religion or is that unimportant to him? Surely when religions were started they were not intended to have a "separate magisterium" (Mainly because there was no science)

  2. “All of this of course does not answer what exactly religion is supposed to do in its magisterium.” Indeed not. I described that point in Gould’s talk about meaning and value as “underdeveloped,” but I admit that my own suggestion about the “supernatural” is not much more illuminating. I share the aversion of Pigliucci, Dawkins, et al., to the idea that religion has authority over questions of moral value, but I think that religion does contribute something to ethics and to the general business of taking an attitude toward life and the world that philosophy cannot supply. Unfortunately, though, I have not worked out exactly what that “something” is.

    “As far as I can tell all Judaism really does is tell us what to do.” Really? Is there not a long tradition of rabbinical writing about what the observances mean?

    “Also does Gould discuss the historical origins of religion or is that unimportant to him?” He does not, though he discusses some historical episodes of religion and science trespassing on each other’s turf.

    “Surely when religions were started they were not intended to have a ‘separate magisterium’ (mainly because there was no science).” Indeed. Gould writes at one point: “The magisterium of science is a latecomer in human history. Faute de mieux, theology once occupied this realm of factual inquiry as well” (Rocks of Ages, 64). Gould seems to say here that there was at one time no magisterium of science. But the idea that the boundaries of the magisteria are historically determined leads to incoherence, as it makes them fall wherever actual scientific and religious practices in fact place them. On that view, the magisteria will overlap, contrary to Gould’s thesis. To be consistent, it seems to me, Gould must maintain that the magisteria are timeless, and allow only that religion has intruded upon the magisterium of science through most of its history (because science as we know it was not yet there) and has only comparatively recently been driven out of it. This, of course, underscores my point that Gould’s thesis assumes a conception of religious authority that is much more restrictive than anything possible in pre-modern tiems.

  3. >Is there not a long tradition of rabbinical writing about what the observances mean?

    Yes but after 1000 years the general consensus is that nobody is quite certain

  4. A blatant characterization of all religion under singular categories won't do.

    Islam (since its inception) and catholicism (at least in its modern form) have completely accepted what the scientific worldview reveals. Protestants are not as progressive.

    One of the reasons I find some religious believers oppose evolution is because they don't have a clue what it is. Evolution equates to them "we come from monkeys", which faulty as it is, is all they know.

    The more enlightened christians and jews are able to reconcile evolution with Genesis is because they already know the entire Genesis order of creation is wrong (Sun made before plants, etc), they already know about the contradicting 2 creation accounts, they know that books like Joshua and Esther are complete historical fabrications, and so on. Christianity in the last few centuries created the doctrine of "inspiration" to accomodate errors in the Bible.

    As for muslims, problems have never really occured with science mainly because Mohammed was a very pro-science person. He told his followers to always seek knowledge, even if it is unto China. The Quran and hadith literature make no statements about cosmology, or matters we would not suited to be discovered by methodological naturalism. And besides Nasir Al Din Al Tusi came up with evolution centuries before Darwin, he was one of the highest religious clerics of the time.

  5. A blatant characterization of all religion under singular categories won't do.

    I'm not sure what you mean by a "blatant characterization" or by "singular categories," but if you are going to use the word "religion" in a serious way, I don't think that you can just shrug off the question of what religion is and what its proper business is, especially not if, like Gould, you want to answer both religious types who think that they have authority to give directives to science and defenders of science who think that religion is just a corruption of human reason that needs to shut up and go away. Gould attributes to religion and science respectively a pair of tasks or essential concerns calculated to keep the two out of each other's business. I don't think that his attempt is successful, but you seem to think that such a task is not even worth attempting. Why?

    Islam (since its inception) and catholicism (at least in its modern form) have completely accepted what the scientific worldview reveals. Protestants are not as progressive.

    That is not my impression of history, much less of what is going on in Islam these days. I don't think that Galileo would concur with your judgment about the acceptance of scientific findings by the Catholic church. Further, professing acceptance of scientific findings is one thing, while actual acceptance of them is another, as can be seen in the record of Orthodox Judaism toward biblical archaeology and criticism. Remember that "accepting what the scientific worldview reveals" means not just accepting general conclusions about cosmology, biology, etc., but also verdicts about history, including the history of a religion's sacred texts and the events recounted in those texts. To say that Islam not been receptive to attempts to extend this kind of scientific investigation to the Qu'ran would be an understatement.

    And besides Nasir Al Din Al Tusi came up with evolution centuries before Darwin, he was one of the highest religious clerics of the time.

    My first thought on reading this was that Darwin never pretended to originate the idea of evolution, but only—simultaneously with Alfred Russel Wallace—the theory of natural selection, which explained how evolution occurs. Still, the Wikipedia article on Tusi provides quotations that confirm that he had the idea of speciation through the differential survival value of traits 600 years before Darwin and Wallace.

    I think it is well known that, in the high middle ages, Muslims were the custodians of all higher learning in the West, while Christians and Jews were either learning from them or remaining in ignorance of science and philosophy. But that state of affairs did not persist into the modern age, and things are on a very different footing now.

  6. >As for muslims, problems have never really occured with science mainly because Mohammed was a very pro-science person

  7. Thanks for the link, Shilton. There is some very interesting reading there, even for someone like me who cannot read the Hebrew passages. (I can use Google Translate, though, if I need to.)

    The title of the post speaks of "Jewish, Islamic and Classical Perspectives On the Conflict Between Authority and Truth." I would emend that description. Each side invokes an authority, and each side claims truth. The conflict, it seems to me, is really between claims to two different sources of authority: religious tradition, whether scriptural or clerical, on the one hand and empirical evidence and scientific reasoning on the other. The difference in principle, in Gould's terms, is that the religious side in these cases is trying to operate outside of the sphere in which such sources actually have authority. It recognizes no epistemological difference in principle between a question about the possible length of a human pregnancy and, say, a question of religious observance: it answers both by citing scriptural or clerical sources. From the point of view of anyone outside of the particular religious tradition in question, this is a positively insane way of answering questions of natural fact. We can sympathize with people who answered all questions this way a thousand years ago, but it is more difficult to sympathize with those who do it today, like the Muslim clerics cited in the article.

  8. "...but I don’t know what proportion of the world’s religious believers can do so."
    Not enough of them.

  9. Science and religion can work together to create wonders for all that is involved.

  10. I think as elegant as all the arguments made here are, they miss the point. Maybe you can call it a category mistake. ( Misapplying a time worn psychological dynamic which I will call for brevity: denial plus. Denial plus = Transcendence.

    To wit. The issue is side stepped by Gould. The issue is the offering of immortality, an immortality ideology (Ryle, as defined, used by Pigliucci) . Without that religion is empty. All religions offer some form of immortality.

    And the claim made that human reason is impotent to deal with this (false, see Becker, Lamont)

    The following is from the frontispiece by Stephen Gould, his anthology to

    Dinosaurs in the Haystack

    For my only brother Peter (1944 - 1994)

    My dearest and constant companion
    May we someday, somewhere and somehow live together in that two-family house of our lifelong dreams.