Monday, February 18, 2013

Religion and Parasitic Memes

It's an obvious fact that organisms do not exist in isolation.  Every individual organism, no matter how simple or complex, must navigate a world of other organisms.  Often, these interactions are antagonistic such as in predator prey dynamics.  However, individual organisms also form positive relationships with other organisms in complex webs of mutualistic benefit.  Perhaps the most startling examples of this come from mammalian studies that have shown the majority of cells contained within a mammal's body are not mammalian cells at all.  Put more plainly, most of the cells in your body aren't human!

Most of the non-human cells taking up space inside the body you think of as you are not harmful.  Indeed, most of them are likely performing useful services that help you to survive and reproduce better.  As such, their biologic interests and yours are closely aligned.  This isn't always the case, though, as anyone who has ever suffered a bacterial infection can attest.  These relationships are unpleasant precisely because the biological interests of harmful organisms and our own are at odds with one another.  Staph doesn't need to worry about your welfare in order to survive and reproduce as long as there are plenty of other bodies in close proximity that it can infect when it's done exploiting yours.  These kinds of antagonistic interactions between organisms is what differentiates parastism from mutualism.  

It is within this context that the concept of memes was born.  A provocative implication of memetics was that cultural information, when considered as consisting of replicators in their own right, might behave similarly to parasites, using hosts' resources to replicate in direct conflict with their hosts' fitness interests.  This could explain why human behavior is often so maladaptive.  It's at once a clever and compelling idea.  Unfortunately, as best I can tell it's an idea that can only be explored within the context of maladaptive behavior.  It's not at all clear to me, for instance, how you could demonstrate selection at the level of memes in cases where memes are mutualistic since selection acting at the level of the individual has equal explanatory power.

Be that as it may, even if memetics can only be explored within the context of maladaptation then it is still a useful idea worth full consideration and there are seemingly endless examples of maladaptive human behavior to explore.  One of the favorite examples that memetics proponents, particularly those that make up the New Atheist movement, like to trot out is religion.  It's no secret that Dawkins, the one who coined the term "meme," is no friend of religion and his career has taken a polemical turn against religion because he believes it to be a parasitic meme, causing people to engage in all sorts of preposterous behaviors that are against their self interest.  He's hardly alone.

There are several problems with this line of reasoning, though.  First and most obviously such arguments generally treat religion as some singular, monolithic thing.  This gives great rhetorical freedom to march out any historical instance of religion behaving badly as an example in support of the hypothesis that religion is a parasitic meme.  However, it's an approach that doesn't give much, if any, empirical traction.  It ignores the fact that different religions are VERY different and even within a particular religious taxon, such as Christianity, there is such incredible variation that different denominations are better thought of as separate species, each of which is comprised of suites of traits that can be adaptive, maladaptive or entirely neutral.  It may be that Quakers display maladaptive traits, but using the excesses of Catholicism during the Crusades as evidence of this makes absolutely no sense, yet the broad brush arguments of many of those in the New Atheist movement seem to be just that undiscriminating.

Such arguments also tend to ignore context.  Generally speaking biological traits only become maladaptive within specific contexts.  Our evolved traits that cause us to find sweet and fatty foods desirable is entirely adaptive in environments where these energy sources are relatively hard to come by.  Only in modern times when these have become available in every corner store do these traits lead to severe obesity with all its associated problems that lead to lower fitness.  Branding Islam as maladaptive because some vanishingly small number of Muslims engage in suicide bombing is a bit like branding the heart as maladaptive because it sometimes leads to cardiac arrest.

This brings us to the last point I'd like to make.  Namely, is religion really maladaptive?  By any definition, this is probably true within certain contexts.  However, is it broadly true?  The answer seems to be no.  The most comprehensive treatment of this topic is Koenig, King and Carson's Handbook of Religion and Health, now in its second edition.  It is a hefty tome of meta analyses that reviews virtually every study that has been done related to religion and health.  I couldn't begin to summarize their findings here but the relevant take home message is that more religious people, on average, live longer and have more children than less religious people.

Now, survival and reproduction are the very sina qua non of biological fitness.  If religion actually can be thought of as being composed of memes then it would seem that most of the time religious memes' interests are closely aligned with the biological interests of their hosts and, as such, cannot be thought of as parasitic at all.  Indeed, if we look at fitness through a strictly biological lens, then the fact that atheists have fewer children and live shorter lives suggests that it is atheism, not theism, that is composed of parasitic memes and needs special explanation from an evolutionary perspective.  I somehow doubt that many within the New Atheist movement would be comfortable with that conclusion, but the best empirical evidence we have seems to point in that direction.


  1. Memes can be the heritable material of cultural symbionts that are mutualists, amensals or parasites. There is no "parasites-only" version of memetics - nor has there ever been. Of course it is true that memes can be easier to detect in cases where there are conflicts between memes and genes over adaptations. Thus the interest in memes that are genetically maladaptive. However, of course you can still detect mutualistic memes - just as you can detect mutualistic genes. You go on yourself to argue in your article that some religious memes probably qualify as being beneficial to their host's genes. Just so. Indeed, cultural mutualism is an old idea. It was promoted by Alexander and Wilson back in the 1970s. The idea that not *all* culture was adaptive to DNA-genes was what Cloak and Dawkins emphasized.

  2. I certainly didn't mean to imply that there was a parasites only version of memetics. Nevertheless, it does remain unclear to me what value a memetics view adds in those cases where memes' interests are be aligned with the host's genetic fitness.

    Be all of that as it may, this is very tangential to the point I was making in this entry. To reiterate, if we set aside any lingering doubts about memetics and grant that religions are comprised of memes, then we still cannot conclude based on the evidence that these memes are generally parasitic. In fact, the evidence supports the view that these memes are actually mutualistic, which rather flies in the face of the rhetoric coming out of the New Atheist camp.

  3. In mutualisms, the symbiote and its host rarely have completely aligned interests. The main idea is that both benefit. For instance strawberries and humans have a mutualistic relationship. However, few would say that strawberries and humans have completely aligned interests. Mutualisms involving cultural entities are like that - both parties benefit, but there are still optimising forces involved which are pulling in different directions.

    As for your theme, I have a post on the same topic, titled: "Religion is probably not maladaptive".

    1. I look forward to the day when such subtle differences can be sussed out. Unfortunately, that day seems rather far away. In particular, we seem to be lacking any methodology to clearly define the boundary of a meme. Granted, gene boundaries aren't always as clearly marked as we like to pretend. Nevertheless, we can with a bit of work meaningfully differentiate at some level of useful resolution where genes are located, when, where and why they're expressed, what makes them different from other genes, etc.

      Theoretically it seems entirely plausible to imagine subtle misalignments between a replicator's interests and those of its host even without knowing the the replicator's boundaries. As a practical matter, though, the inability to identify such boundaries for memes would seem to seriously limit empirical opportunities to demonstrate that this does, indeed, happen or to explore the implications for individuals when it does.

      Could you provide a URL or link to your post? I'd be interested in hearing your take, especially since you've thought far more carefully about these issues from an explicitly memetics perspective.

    2. The issue about where the boundaries of memes lie seems like a pointless hangup of critics of memetics to me. You can do frequency analysis of memes, just as you can of genes - and that's quite enough to be able to study them scientifically. Researchers don't obsess over delineating the precise boundaries of condom memes when studying things like their effects on fertility.

      Yes there are stop codons - but there are also overlapping genes and genetic recombination, which slices DNA genes up at any point. Memetics has its equivalent of stop codons - they are things like spaces and full-stops. It also has millions of memetic codes, not just the handful that are used by nucleic acids.

      Here's the URL of my article on the adaptive aspects of religion:

    3. When I have a chance to put my thoughts into a bit more order than I'm able to muster this morning I'll probably dedicate a post to why I think boundary issues are important, or at least when I think they're important. My intuitions on this stem from my work with analyses of sacred texts where it is not at all uncommon to have very small differences in, say, Biblical readings convey very different signals to an audience of believers. A given Biblical citation plus or minus a verse can make a significant difference.

      Then, too, there are the problems associated with how even the same bounded text are inflected through different communities. "John 3:16" connotes very different things for a liberal Methodist and a conservative Pentecostal. Indeed, one of the biggest problems I have with memetics as a discipline is that it hasn't, as far as I can tell, yet developed a theory of semiotics.

      Again, my thoughts are too scattered at the moment so I'll leave it there. I did, however, get a chance to read your piece on religion and it sounds that we're in broad agreement there. Thanks for sharing it with me!

  4. I tend to see things the other way around. Semiotics, though a pretty old discipline, failed to apply evolutionary theory to the topic of how signs changed. They missed out on something central. In the end, cultural Darwinism mostly came from biology-oriented theorists - who mostly ignored semiotics.