Field of Science

Bad fonts decrease belief in God

So, do you believe in God? OK, well how about if I ask it like this: Do you believe in God?

Surprisingly, according to a new study by Will Gervais and Ara Norenzayan at the University of British Columbia in Canada,simply changing the font used to write the questions (to make them more difficult to read) can actually make you more likely to answer 'No'.

The explanation, they think, is that when the text is hard to read, we have to concentrate harder. We step our analytical brain up a gear, and quash our instinctive reactions.

In fact this effect, known as cognitive disfluency, has been shown in other studies to trigger analytic thinking strategies. 

There's good reason to think that triggering analytic thinking might reduce religious belief. Two recent studies (one last year and one last week)  have shown that people who are stronger in analytical thinking are also less likely to believe in god - even after controlling for basic intelligence.

But what's new in Gervais and Norenzayan's study is that they showed that you can manipulate this. They found that subtly encouraging people (well, Canadian students) to think critically also encourages disbelief.

For example, showing them a picture of Rodin's "The Thinker" (on the left) versus one of Discobolus, reduced reported belief in God by around one third.

Doing a word puzzle which featured words like "analyze", "reason", "ponder", and "think" had a similar, although somewhat smaller, effect (reducing belief in supernatural agents and also reported religiosity by around 20%).

And, in the font study, they found that simply using a greyed out, italic font was enough to have the same effect. Students who were made to read the questionnaire in a difficult font were around 20% less likely to report belief in God, the Devil, and angels.

So, has anyone else noticed that newspapers are a lot easier to read than they were at the beginning of last century? I wonder what effect that has had!

Will M. Gervais and Ara Norenzayan. Analytic Thinking Promotes Religious Disbelief. Science 2012 336, 493-496.

Creative Commons License This article by Tom Rees was first published on Epiphenom. It is licensed under Creative Commons.


  1. Sometimes the research you review here seems like a Poe. But I know it isn't.

  2. That's pretty fascinating. I had read how wearing a white lab coat could help increase your concentration, but having skepticism of religion triggered by such a physical cue is unexpected, at least to me.

  3. One thing in the article needs to be corrected. The University of British Columbia is in Vancouver, not Columbia.

    Other than that, a fascinating read and a very clever way of testing an idea.

  4. Peter - I meant to write Canada, not Columbia. Damned word blindness :) Fixed now!

  5. Wow, finally Comic Sans has a justification for its existence.

  6. Haha, at the ending. But I'm curious, were the students asked about the religious position before the test and after or during the test, for comparison? I mean, how exactly did the researchers know belief in the supernatural decreased? What two things were compared?

  7. It was a parallel-group study, meaning that half the subjects got the 'analytical thinking' prime, and the other half got the neutral prime. So, for example, half read the questions in a normal, plain font.

    Then they measure beliefs.

    So, strictly speaking we don't know that their beliefs went down after the analytical thinking prime, just that they were lower than the other group.

    But, assuming both groups were the same at the start, and the neutral prime had no effect, then that's the most likely explanation.

  8. If I were asked to fill in a survey with bad font, I'll be feeling ticked off that I'd be prime to answer negatively to any question.

    Were there other questions to control for this effect in the study?

    Similarly with the pictures if they are as pictured in your post. The second picture is more pleasant to the eyes.

    With the word puzzle, were the control group also given the puzzle but without those priming words? Or were they given no puzzles?


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS