The moralizing gods saga has ended, or at least its first chapter. Two years ago, Nature published (Whitehouse et al 2019) a widely publicized paper that claimed to demonstrate that moralizing gods, who care more about relationships between humans than relationships between humans and gods, developed following major social complexification events, and not prior to them. This was a challenge to the so-called big gods hypothesis, which holds that gods who punish humans for bad social behaviour evolved first and allowed, or even caused, the formation of large, dense, complex social groups.

The paper drew serious critiques right away. Notably, Beheim et al. (2019, 2021) argued that the authors had wrongly assumed that there were no moralizing gods in times and places for which information was not available. The analysis in the original paper, it seems, had treated NA (not available) as FALSE (no gods), meaning that the result was an artifact of the structure of missingness in the data. I give a bit of a blow by blow of the debate as it unfolded in a series of posts here, here, here, and here.

Two years later, we have a resolution of the affair, although one that is in many ways unsatisfactory, and reinforces my belief that our academic publishing system is obsolete, more destructive than productive, and due for a major overhaul.

On July 7th 2021, Nature published the Beheim critique along with a retraction of the original moralizing gods paper. Thank the moralizing gods, though, the Beheim critique had been available as a pre-print since May 2019, and as far as most people are concerned, the debate had been settled at least a year ago after some back and forth between various configurations of Whitehouse et al. and Beheim et al.

So what do we learn here

This could have been a very brief, very productive conversation between colleagues over social media or through short blog posts, for example. Imagine that Whitehouse and colleagues tell us they think they found that large complex societies precede moralizing gods. They show us a table with their data. This could have been a tweet with an attached image. Looks pretty interesting at first glance.

In the discussion thread that follows, Beheim and a few others point out that the cells in the table where there are no moralizing gods are in fact mostly unknown, not false. Most everyone takes a second looks and agrees. Whitehouse and colleagues thank everyone for their input and keep working on their test with better methods and assumptions. This could have happened within an afternoon. In fact, most of the essential, helpful elements of the saga did unfold over just a few days, and with no help from journals.

Instead of this, we end up with a massively time consuming, professionally costly, and at times acrimonious battle of the Titans in Nature, over two years, mobilizing the efforts of who knows how many authors, reviewers, editors, comms people, etc. Not only that, but we have a bunch of colleagues who forevermore have a retracted tag from the biggest name in the business.

This doesn’t need to happen. All this fire, fury, stress, and damage, for basically no scholarly gain and no actual learning. Bizarre at best, and at worst, unconscionable.

The situation, which is far from unique, escalated and was so expensive because the stakes were so artificially inflated by our academic publishing, hiring, and promotion culture, and by the commercial journal industry that preys on its worst features.

How do we fix things?

This case cements for me, once and for all, just how much more powerful and preferable post-publication review is to pre-publication review. Despite the best good-faith efforts of most of the people involved in it, pre-publication review, in the end, amounts to little more than arbitrary gate-keeping. Because the number of reviewers is so limited, and because the pool from which they are drawn is so small, the outcome ends up being essentially random in terms of quality.

Post-publication review harnesses the full strength of the academic community and evaluates results on a much broader and more reliable basis than pre-publication review ever could. Narrowing the review to the “best experts” in a field only narrows it. Review needs to be broad, not narrow. Putting all your chips on Nature at the scholarly casino and trusting the pre-publication review roulette makes no sense for the authors or for the academic community as a whole, and this case shows it clearly.

Review also needs to be welcome, rather than feared. That means the salient consequence of review needs to be improvement rather than punishment. The moralizing gods saga escalated because some of the actors very reasonably feared the professional and personal unsupernatural punishment that can follow a helpful comment by a well-meaning colleague. We created the system that generates and sustains that fear, and we can change it if we want to.

Right now, the consequence of pre-publication review that looms large in people’s minds is rejection, and the feared consequence of post-publication review is retraction. What they have in common the prestigious journal as the source of professional worth.

In a rational academic world, we wouldn’t write huge papers, submit them for narrow review, and then live in fear of a career threatening retraction because of the much more powerful lens of broad review. We would work with each other, as we do when we’re writing the papers in the first place. Our collective knowledge and understanding would grow in the same way that our individual and small group papers grow.

People working together in a lab, or in a small research group, call each other over and say “hey, look what I just noticed”. Sometimes, everyone will say, “wow, that’s really interesting, we can build on that”. Sometimes, one of the people you call over will say “actually, if you recode these cells as unknown, the whole thing disappears”. Everyone has a laugh, everyone goes back to what they were doing and that’s the end of the drama.

We now have the tools to this this all together, no matter where we sit physically. It can all work like this. We can share our ongoing observations, results, insights and thoughts as they evolve, and help each other rather than punish each other. We don’t have to store them up for a high stakes gamble ultimately designed to protect commercial interests. Yes, it will have implications for funding and hiring, and yes, those changes need to happen too.


Beheim, B., Atkinson, Q.D., Bulbulia, J. et al. Treatment of missing data determined conclusions regarding moralizing gods. Nature 595, E29–E34 (2021).

Whitehouse, H., François, P., Savage, P.E. et al. RETRACTED ARTICLE: Complex societies precede moralizing gods throughout world history. Nature 568, 226–229 (2019).

5 thoughts on “Twilight of the moralizing gods: Lessons for a better academic world

  1. There’s a special place in “hell” for those that would use the word “compelexification” (not to mention “missingness”!”. For those that bemoan the lack of open access think about what those accessing that see bullshit academic words like “complexification”.

    Liked by 1 person

  2. Informal online discussion has many good features, but its weaknesses are the final synthesis and long-term archiving. Who has time to sort through a mess of rambling threads and pick out the good bits? Who can trust that those sites will still exist in 30 years when the next article on that subject is written? And they invite a gambit where someone asserts that something has been proven / refuted but does not point to where. So there is a lot of value in taking the discussion, reducing it to the key points, doing the work of documenting claims which seem obvious to you but not to everyone in the discussion, and publishing it in a format which will exist and be findable in a hundred years.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s