The Boundaries of Belief - Sam Harris

According to a recent Pew survey, 21 percent of atheists in the United States believe in “God or a universal spirit,” and 8 percent are “absolutely certain” that such a Being exists. One wonders if they were also “absolutely certain” they understood the meaning of the term “atheist.” Claiming to be an atheist who believes in God is like claiming to be a happily married bachelor. Rarely does one discover nonsense in such a pristine state. Still this hasn’t stopped many people from concluding that there is a schism in the atheist community.

The inclusion of a “universal spirit” might have muddied things for some of these putative atheists, but this would not account for the 6 percent of them who rejected such a spirit in favor of a “personal God.” Granted, it is not clear what the phrase “personal God” might mean to men and women who have wandered so far from the plain meaning of words, but we can only assume that they believe in a God of the sort that 71% of Americans worship: a deity who can hear earnest and blameless prayers—as for the remission of childhood cancer—and fail to answer them, while granting those of far lesser gravity nearly every day (I rely upon the reader to insert here the most mortifying expression of religious awe ever uttered at the Grammy Awards).

Open the newspaper tomorrow morning, or any morning thereafter, and reflect upon the fact that half of your neighbors (51%) are “absolutely certain” that a “personal God” presides over all this casual destruction. The incongruity and moral carelessness of such certainty is reason enough to keep atheists (the real ones) awake at the ramparts until a proper war of ideas can be finally waged and won.

The Pew survey produced a few more anomalies: 3 percent of “atheists” are “absolutely certain” that a personal God exists and believe that the Bible is His “literal” Word; 4 percent attend religious services weekly; 5 percent pray daily; 2 percent receive answers to their prayers “at least once a week,” have witnessed “a divine healing,” and draw their morality straight from scripture. It may well be that some atheists, lacking the requisite fear of hell, find it amusing to maliciously waste a pollster’s time. I think, rather, that these figures are simply what it sounds like to ram against the error bars in this particular survey.

Pew’s sample of 35,556 Americans included 515 respondents who identified themselves as “atheists” (1.6 percent). The margin of error for this subgroup appears to be around 5 percent – which clearly makes a hash of many of the above findings. Among 35,556 people, Pew seems to have found 40 especially confused God-fearing men and women who think they are “atheists.” Their mutterings do not offer any special insight into the nature of belief.

In search of such insight, we recently conducted a much more detailed poll of atheists and devout Christians through my website. Our sample of respondents was almost the inverse of Pew's: we had 36,781 surveys completed (some respondents completed more than one), mostly by atheists. Rather than accept each persons self-description as an “atheist” or “Christian” at face value, however, we filtered our results by each person’s response to the following two statements:

Please indicate your degree of belief in the God of the Bible.

1. Disbelieve strongly

2. Disbelieve somewhat

3. Don't Know

4. Believe somewhat

5. Believe strongly


Please indicate your degree of belief that the Bible is the word of God.

1. Disbelieve strongly

2. Disbelieve somewhat

3. Don't Know

4. Believe somewhat

5. Believe strongly

We then focused on those who responded with a 1 or a 5 to both of these statements. The primary purpose of this poll was not opinion research, in fact. Rather, we were designing stimuli for an experiment that we are now running on atheists and Christians using functional magnetic resonance imaging (fMRI). The goal of survey was to produce stimuli of two categories – factual and religious – which would behave appropriately once we put members of each group inside our MRI scanner. We needed factual statements that both atheists and Christians would accept with the same order of confidence and religious statements that would divide them more or less diametrically.

In addition to vetting our experimental stimuli, however, we took the opportunity to solicit the opinions of believers and nonbelievers on many psychological and social topics that are not strictly relevant to our neuroimaging work. Many of these results are now available for viewing on my website.

Sunday, July 6, 2008