I cannot really provide an objective "review" of Daniel C. Dennett's Breaking the Spell: Religion as Natural Phenomenon (Penguin, 2006) due to certain peculiar prejudices which the reader probably does not share. But I can recommend the book whole-heartedly. Read this book if you have any interest at all in contemporary arguments about the status of religion in society.
By "contemporary" I mean after September 11, 2001. Just as American isolationism effectively came to an end on December 7, 1941, tolerance for religious extremism, at least among American intellectuals, is now effectively over. Books like Sam Harris' The End of Faith: Religion, Terror, and the Future of Reason (2004), Richard Dawkins' The God Delusion (2006) and Christopher Hitchens' God is Not Great: How Religion Poisons Everything (2007) dominate the best-seller lists. Wired Magazine coined the term "New Atheism" to describe this intellectual trend.
When one reads Dawkins' book (The God Delusion) and he so often quotes Dennett, one gets the feeling one would be better off reading this book instead. Dennett is a philosopher by trade (a professor at Tufts) in the Anglo-American mode which worships empiricism and natural science and strives to bring clarity where needed by asking pertinent questions. I have to admit that in this case, a bit of clarity is what fits the problem. While the questions he asks seem to have obvious answers ("can religion be studied as a natural phenomenon?" Yes. "Should it?" Of course) he does a good job of marshaling the objections and dealing with them.
Take the case of Stephen Jay Gould's "Non-Overlapping Magisteria" ("NOMA") approach, first articulated in a 1997 essay by that name for "Natural History" magazine, and later in his book Rocks of Ages (1999). NOMA contends that conflicts between science and religion can be resolved by restricting both to their appropriate domains or "magisteria": science would be limited to the empirical realm, to theory grounded in observation, while religion would deal with questions of ultimate meaning and moral values. It seems all the "New Atheists" are united in their rejection of this approach, for the simple and obvious reason that religion declines to be limited to its proper "magisteria".
Religion is a vast and complicated subject, even when you limit yourself to just a slice of it, as Dennett does when he defines religion as "social systems whose participants avow belief in a supernatural agent or agents whose approval is to be sought." (More on this limitation in a moment). There are, however, certain themes in the study of the evolution of religion which can be identified and are useful in bridging and relating more specific studies and controversies, such as the faux-controversy between evolution and creationism, the separation of church and state in law and politics, and dealing with the corruption of religious enthusiasm in violent extremism. Dennett does a great job with this in Part II of Breaking the Spell ("The Evolution of Religion"). There's a lot here which is very thought-provoking and could even be read and enjoyed by some people outside his target audience of non-believers ("brights" as he mischievously calls them).
Ultimately, however, I find this book irksome. Dennett ignores liberal Christianity, and the long history of mysticism and reformation which has struggled to break Christianity out of superstition, i.e. what Dennett defines as religion itself, i.e. "social systems whose participants avow belief in a supernatural agent or agents whose approval is to be sought." Dennett mentions Paul Tillich a few times, he doesn't mention Rudolf Bultmann (or "de-mythologizing") at all, and seems largely ignorant of the kind of theology taught all our elite liberal seminaries.
Perhaps this is fair because the influence of liberals in American religion is declining. Still, if religious extremism in general, and conservative Christianity in particular, are seen as social and political problems, would you not want to enlist the support of people who are trying to reform the religion? Apparently not. Dennett takes at face value the criticism of Stark and Bainbridge (sociologists who have studied the spread of religion) that God as "essence" (identified with Tillich's notion of God as the "ground of Being") will never have the persuasive power of an anthropomorphic personal God. This may well be true. It is certainly the case that the vast majority of people who call themselves "religious" fit Dennett's definition, rather than what might be taught in liberal seminaries in America. In sociology at least, I suppose what most people believe is the object of study, and not "the truth" of what they believe.
Dennett seems to understand that liberal Protestantism exists (he refers to Paul Tillich, for example) but he either excludes it from his study, by definition, or accepts the (reactionary, conservative Christian) critique that Tillich's God (as the "Ground of All Being") is an unpersuasive belief in a divine essence rather than a person which no religious person can possibly take seriously. The argument there is that a divine essence cannot provide what people seek from a personal God. (There is an age-old answer to that: your personal God is just an idol. An idol is not God, but merely a representation of God.) Dennett's definition of God ("a supernatural agent or agents whose approval is to be sought") runs counter to 2000 years of theology. Paul and Luther taught that God's approval cannot be sought, it can only be freely received (we are justified by faith through grace). I realize that this is not widely understood. Many so-called Protestants contend, for example, that you have to say you believe in Jesus to be saved, relying on Romans 10:9 ("if you confess with your lips that Jesus is Lord and believe in your heart that God raised him from the dead, you will be saved.") So they emphasize praising God and shouting "Jesus is Lord!". I don't think God is impressed with bumper stickers. I think you only have "faith", a positive relationship with God, when you give up constantly seeking God's approval and just accept God's love.
This error manifests itself almost daily in the politics of religion, for example, when a preacher says a hurricane is God's punishment for allowing homosexuals to live. When he says that, he is not expressing a religious insight, but rather a political position: he wants to persuade people to take homosexuals outside the city gates and stone them to death (or failing at that, making their lives as miserable as possible). Granted, this is a subtle piece of theology, understood only by the elite, the saints and the mystics. What theologians call "works righteousness" is so thoroughly embedded in popular thinking, what Dennett might call religious "memes", that it very much resembles the parasitic viruses that Dennett likes to talk about.
The legalistic, "works" preacher (the one who says God smote us with a natural disaster to punish us for our disobedience) has as his object purifying the city or nation, making people conform to a standard of behavior and a sense of identity (which incidentally enhances the power and prestige of the preacher's class in society). If the legalistic preacher has to increase emotional suffering (anxiety, fear, despair, loneliness) to accomplish these political objectives, than so be it. In this way, religion has been hijacked and its purpose thwarted to serve a political purpose: social cohesion and conformity.
Dennett's definition fails to exclude this purely political behavior, and indeed, a lot of the New Atheists seem to think that religion is nothing more than a means of social control. That's fine for polemics and rhetoric, but not for science. This is a serious methodological flaw if you ask me. Following Dennett's definition and the methodology of some of the people whose work he describes, we are studying not "religion" as a natural phenomenon but popular errors of religion, analogous to deadly genetic diseases. That's a problem for someone like Dennett or Dawkins whose bag of intellectual tools includes frequent analogies and references to replicators and their error-correcting mechanisms.
The flaw becomes even more obvious to me when Dennett addresses religion today and the consolation and moral support it provides. Once again he feels like he has accomplished something if he disproves superstition (e.g. that prayer will cure cancer) but downplays the emotional and spiritual consolations which religion provides in the face of death. He feels like he has accomplished something when he says that atheists can have moral notions (right and wrong) and many so-called religious people do bad things, and thus misses the reinforcement and discipline which religion provides (or somehow failed to provide, in many cases). For example, an alcoholic who has come to an AA meeting doesn't need to be told what is "good" and "bad". The alcoholic who has taken that step knows drinking is "bad" and recovery is "good". The epistemological origin of this is somewhat beside the point. What the alcoholic needs is something to keep him or her going during those times when they are not at a meeting or talking with their sponsor on the phone. That something is traditionally provided by a "Higher Power". If you don't believe in such a thing, you are encouraged to "fake it 'till you make it". And that seems to work, too. It may just be a psychological trick, designed to empower that part of the alcoholic's mind or psyche which wants to recover, over the part which just wants to give up. It may be, in other words, that the "Higher Power" is the alcoholic himself or herself, exerting self-control (in spite of the AA theology that the alcoholic is "powerless" before alcohol). It is a trick that works (what Dennett calls a "Good Trick") and it works not just to perpetuate the AA "meme", but also to help people quit drinking.
Now I would be remiss if I were to give the impression that Dennett is not aware of my objections here. His observations about "belief in belief" (page 200 and what follows) delve into precisely these issues. And I would completely agree that science is useful to expose religion's "Good Tricks" (the features that either perpetuate religion for its own sake or because they are beneficial). Still, I object to the definition of religion as superstition. Dennett may not feel comfortable with distinguishing true religion from false, lacking some sociological or empirical study to back him up, but I would contend that equating superstition with religion is like equating bacteria with "germs": it ignores the existence of the useful parasites.
I would define religion as cultural artifacts (stories, music, art) and social practices (assembly, ritual, prayer) which have as their object the intent the alleviation of emotional suffering (anxiety, fear, despair, loneliness). The "supernatural agent" isn't essential to the purpose. I can illustrate this with a question: if God knows what you want, why pray? Because the prayer is for our sake, not God's. The point is not to appease God and seek God's approval by prayer and praise, but to set our minds right and our hearts at ease. This doesn't fit the agenda of the New Atheists in their battle with Creationists and other evil extremists, and so I wouldn't expect Dennett to frame the question this way.
Nonetheless, I'm not willing to let Creationists define what religion or Christianity is, no matter how many polls show that most people understand God their way. The same great masses who believe God is "a supernatural agent" whose "approval must be sought" also don't have a clue what a Higgs boson is (nor do I, to be honest). If we study physics, then, should we take polls, and ignore what the so-called experts say? (It would certainly save a lot of time and money wasted on elite projects like particle accelerators.) By ignoring us, Dennett would have us liberal Christians just give up on correcting the legalistic and exposing the un-Christian political agendas of fascist faux-Christians. This I am not willing to do.