Two can play this game... :)

So here's a little radar diagram that I've done up to represent some of the things we discussed this afternoon. Specifically, it relates to the idea I had about mapping congregations in terms of their expressed levels of diversity on the horizontal axis and tolerance of diversity on the y-axis. If we further break diversity and corresponding tolerance down into beliefs, behaviours and demographics categories, we could arrange measures for specific faith communities as such:

If we could swap perceived for actual the above schematic, then I think we’d have an interesting handle on what it means to be a tighter faith community (i.e. a relatively small surface area - Church B in this example). The problem with this of course is that doing so would require a list of all the beliefs individuals are supposed to subscribe to and getting a sense of existing drift for each community. Not intractable, but certaintly not feasible for the moment. So we may be stuck with perceived. Actually, now that I think about it, having all three would be ideal - actual, perceived, tolerance. As it stands, “behavioural diversity” is a little up in the air… do we mean life behaviours or worship/religious behaviours? 

Thoughts? Specifically, is surface area a decent measure of tightness? For example, what if group C is really high on diversity, but low on corresponding tolerance measures, and group D is low on diversity measures but really high on corresponding tolerance measures. Both would have the same area but be manifestly different groups. Clearly some distinguishing power is lost. Are both congregations effectively equally Tight but for different reasons? I don't think so, as one would be open to increasing diversity while the other is not. Does this mean the polygon created by the Tolerance axes is the Tightness measure and the other half really constitutes a Disparity measure between "ideal" and "actual" or "perceived"?

Rules as Genotype: Random Reflections

As much as I enjoyed the workshop last week on the topic of "rules as genotype," I've been increasingly disappointed that nothing was really accomplished by the participants there.  We had some amazing conversations, but we could never really get beyond various restatements of the metaphor.  In the end, I believe the only conclusions were either that the metaphor was still viable or that it had run its course.  No one, however, clearly articulated why they fell into one camp or the other and about the most useful comment was also the most obvious.  Namely, when you argue from metaphor, trying to map concepts from one domain to another, you're bound to be wrong in the details.

Aside from having too many different disciplines at the table to make much progress in two days, I think the workshop went off course primarily because no one wanted to get his or her hands dirty hammering out the messy details of vocabulary.  For instance, I don't recall anyone asking what a genotype was or what rules were or what aspects of a genotype rules were supposed to emulate.  Yet these are fundamental problems that have to be overcome if there's to be any hope of progress on this idea.

Thankfully, there a number of thinkers who are taking the idea of universal Darwinism seriously and who have done at least some of the basic groundwork.  I need to become far more familiar with this camp, but I believe Hodgson and Knudsen may have already answered this question in Darwin's Conjecture.  I don't want to try to summarize their entire argument here, or at least not now.  However, key to their argument is something they call a generative replicator.  For them, replicators require four conditions:

1. Causal implication:  The source must be causally involved in the production of the copy, at least in the sense that, without the source, the particular copy would not be created.
2. Similarity:  the replicated entity must be or contain a replicator.  The conditional generative mechanisms in the copy must be similar to those in the source.  Errors or mutations in these mechanisms must also be copied with some degree of fidelity.
3. Information transfer:  During its creation, the copy must obtain the conditional generative mechanisms that make the copy similar to its source from that same source.
4. Conditional generative mechanisms:  Generative replicators are material structures that embody construction mechanisms (or programs) that can be energized by input signals that contain information about a particular environment.  These mechanisms produce further instructions from a generative replicator to their related interactor that guide its development.  (pp. 122-123)

So, are rules a kind of generative replicator?  Before answering that, it's worth asking of genotypes are generative replicators.  According to Hodgson and Knudsen the answer is yes.  DNA cannot be replicated without the template, fulfilling the causal implication.  the replicated entity, the nascent strand of DNA, most certainly fulfills the similarity and information transfer requirements.  Moreover, the resulting daughter strands of DNA are capable of responding, through a variety of epigenetic mechanisms, to environmental inputs in order to guide development.  

If we can accept Hodgson and Knudsen's definitions, then it seems the question we should have been asking at the workshop was, really, are rules generative replicators?  Scanning the bibliography of Darwin's Conjecture, it looks as though Hodgson at least has already written a lot in this area that I should become more familiar with.  Unfortunately, the closest they come to discussing rules in the institutional sense is their exploration of judicial law as a major transition (Section 8.5 and assorted other places within the book).  For a variety of reasons I was unhappy with this particular discussion.  Perhaps the most glaring disagreement I had with them was their insistence that judicial law requires writing.  This simply flies in the face of historical fact, perhaps best exemplified in Icelandic law.  None of that, however, is particularly relevant to the concerns here.  

The authors argue that judicial law does constitute a generative replicator.  The causal implication requirement is a bit vague, but it seems to be met at some level either when laws are copied to new states (one form of replication) or when new laws are written which tend to rely on existing legislation.  It seems the argument here would be stronger if they included laws being passed from one generation to another, but I don't believe they consider this process in their formulation.  At any rate, these same processes, along with judicial review, which tends to clarify and refine laws, ensure that the similarity condition is met.  Undoubtedly, laws contain information so the copying of laws involves information transfer.  Conditional generation occurs because laws are typically invoked under specific circumstances and because they guide the development of the state institutions of which they are a part.

The hard work, then, seems to have already been done in answering the question of whether rules can be considered cultural genotypes.  It would be almost trivial to map Hodgson and Knudsen's argument onto Ostrom's formal conception of institutional rules.  For now I'd feel comfortable concluding that rules, like laws, could be considered conditional replicators.  A question that remains unanswered in my mind is what would be the cultural "phenotype" associated with rules.  In other words, what is the interactor that rules respond to?  An even bigger question, though, is so what?  If rules are generative replicators, then what are the implications for research or policy development?  In what ways might this formulation of rules provide new insights or help to organize existing knowledge?  These, too, were questions that didn't get addressed at the workshop and I think they are perhaps the most important questions of all.

Yet another concept map

It seems these are always hopelessly flawed.  Nevertheless, this map seems to capture most of what we've been talking about.  How useful it may be is another question entirely.

So, here I've partitioned the environment into different aspects, those that affect groups and those that affect individuals.  These, however, are not compartmentalized which is my way of attempting to show these are not always separate.  As I've been discussing, then, groups and individuals may have very different experiences within the environment and, as a result, may face different kinds of pressures.  The environment alone, however, is not the only source of influence/pressure/constraints/whatever.  Groups can also exert pressures on individuals and, of course, groups are affected by the traits of the individuals who comprise them.  All of these different influences work to produce outcomes.  Outcomes at the group level might be different from outcomes at the individual level.  Or they might not.  Perhaps I should have placed them in a box together the way I did for environmental influences?

I haven't quite unpacked this yet, but it may be in the outcomes that we find some traction with our conceptions of "alignment" between groups and individuals.

Does this do any useful conceptual work?  Again, I'm not entirely sure.  However, it does allow us to more methodically follow constraints and pressures (are these different?) to see how they might affect tightness or looseness.  For instance, let's imagine a harsh environment that offers little existential security for either groups or individuals.  This would limit the variation between groups operating within that environment and between individuals who are potential members of groups.   These should work together to create very tight groups.

Or what about an environment that is existentially secure for individuals but in which groups face existential insecurity by virtue, say, of fierce competition with other groups.  In this environment, groups should be far looser than in the previous scenario.  Nevertheless, they should be relatively tight in those particular traits that lead to differential success for the group.  For instance, let's suppose that successful groups within a given context are those with the greatest perceived entertainment value.  If groups can only survive by virtue of their entertainment value, then opinions about the relative importance of entertainment among group members should be tight even while opinions about other aspects of the group might be quite loose.  Clearly, I need to think of better examples but hopefully this gives an idea of what we're after.

One thing I do like about this map is that unlike some of my other recent formulations this provides space to consider the bottom up influence of individuals on the group and the top down influence of the group on individuals.  Aside from giving us a place to stick psychological variables like personal need for structure and social dominance orientation, it also gives us a way to consider historical trajectories.  For instance, if a tight group has survived several generations of individuals then it seems likely individuals raised within the group should have a higher need for structure or a higher social dominance orientation by virtue of their indoctrination over generations by the group.  Even if that group moves to a place where existential security at both the individual and group levels would lead us to predict greater looseness, it may take considerable time for the group to move in that direction due to to the bottom up influence of many tight individuals working in concert.

Okay, now I'm just rambling.  Enough!

Selection

Tonight in an email exchange with Ian, the subject of selection came up.  Or perhaps more accurately, the problem of defining selection pressure came up.  Over the past couple of entries, I've been excited about thinking of group tightness and looseness as responses to differential selection pressures acting at the group and individual levels.  There is something intuitive about the idea that just "seems right."  As is so often the case, though, translating that gut intuition into something meaningful and practical is proving more difficult.

One stumbling block in this is that the entire scheme relies on understanding selection pressures that might act on either the group or the individuals comprising the group.  But what exactly do we mean by this?  How can we operationalize it?  How can we measure it?  My tack so far has been to come up with as many examples as I can of different groups falling into different categories of tightness and looseness and to think about the selection pressures at work.  While this has been seductive, I've started uncovering examples where the whole scheme seems to break down.  These have been instructive, though, inasmuch as they haven't been able to completely refute the scheme; they've left some room for the overarching idea to still wiggle free.  It's in this ambiguous space that I've come to appreciate where more precision is required.

I think one place to begin is to define what we mean by selection.  Without a firm conception of selection, any talk of selection pressure become rather silly.  I'd like to have a more exhaustive inventory of accepted definitions, but it's already past my bedtime so for now I'll just share the definition given by Hodgson and Knudsen in Darwin's Conjecture:

Selection in a complex population system involves an anterior set of entities that is somehow being transformed into a posterior set, where all members of the posterior set are sufficiently similar to some members of the anterior set, and where the resulting frequencies of posterior entities are correlated positively and causally with their fitness in the environmental context.  The transformation from the anterior to the posterior set is caused by the entities' interaction within a particular environment.  (pp. 241-242)

Interestingly, they differentiate between two different kinds of selection:  subset selection and successor selection:

Subset selection is defined as selection through one cycle of environmental interaction and elimination of entities in a population, structured so that the environmental interaction causes elimination to be differential, and where survival outcomes are correlated positively and causally with fitness in that environment.


Successor selection is defined as selection through one cycle of replication, variation, and environmental interaction, which leads to differential replication, novel entities, and a changed distribution of population properties that correlates positively and causally with the fitness of entities in that environment.  (p. 242)

Fitness plays a role in all of these definitions, so it's probably a good idea to pin what they mean by that down, too.  This turns out to be a bit thornier than you might expect:

In biology, fitness is most usefully defined as the propensity of a genotype to produce offspring (DeJong 1994).  Survival of the fittest is no longer a tautology:  it is possibly false.  The fitness of a replicator is the propensity to increase its frequency (relative to other replicators).  In the social domain, this definition of fitness translates into the propensity of a social replicator (such as a habit or a routine) with a particular feature to produce copies and increase the frequency of similar replicators in the population.  The fitness of an interactor is the propensity of its replicators to increase their frequency. (p. 238)


Here I sense some slippage between these concepts.  Much, I'm sure, to Ian's delight Hodgson and Knudssen define selection in terms of elimination of entities due to their interactions with the environment.  However, their notion of fitness seems to be more in line with conceptions of  selection for entities.  Perhaps that's neither here nor there, but it is curious to note all the same.

So what to make of these definitions?  At this point, I'm not entirely sure.  More than anything, I just wanted to get them out there for us to begin playing with.  However, some preliminary thoughts are that if we define selection in a similar way, then selection and existential security are in many ways interchangeable.  Selection pressure, which oddly isn't defined at all by Hodgson and Knudsen, would seem to be a property of the environment which affects how much variation can be present within a given population interacting within a particular environmental context.  If this is so, then it would seem to be even more directly related to notions of tightness and looseness than I first thought.  At the very least, it's not hard to see how I could come to the idea that selection pressures are key to understanding these phenomena.

Partitioning selection to understand tight and loose groups

A couple of days have passed since my initial epiphany regarding the tight/loose distinction.  Having come down from that emotional high, and being rather regrettably sober for the moment, I thought it might be a good idea to revisit my initial intuitions and to frame them a bit more formally.  

I begin with the acknowledgment of individuals.  To facilitate discussion, I will talk about these most naturally as single organisms but, in principle, I think it could be any entity with a given set of properties existing in the world.  I further acknowledge that individuals can come together to form groups composed of multiple individuals.  For the time being, I'm not going to fret too much about how that group is defined or how it's bounded, though that's clearly something that needs to be considered.  I should also point out that I'm not concerned with the fact that individuals can be thought of as collectives of individual entities. That simply moves the level of analysis to differing levels of complexity without affecting the broad theoretical framework I hope to develop.

I propose that, once a particular group is constituted, it becomes an entity in the world in its own right.  If this is so, then a group, just like an individual, can become subject to selection pressure.  As an aside and in a nod to Ian's theoretical work, my intuition tells me this will hold true regardless of whether we think of selection for or selection against.  However, I haven't come close to thinking through any of that so I'll leave it for now.

This brings us to a crucial point.  Namely, individuals who are members of groups are themselves still subject to natural selection.  This means that natural selection can act at all levels of organization.  This, I think, is obvious but somehow it's escaped our attention so far as we've tried to think our way through what environmental factors might be favorable to tight or loose groups.  Furthermore, I think this may be absolutely key.  To see why, consider that there is no requirement that the selective pressures at both levels or organization be at all closely aligned.  I expect that groups can evoke emergent properties that make them subject to pressures not necessarily felt by individuals.  Whether the converse is true, whether individuals can be subject to selection pressures not felt by the group they belong to, is a bit thornier so I'll set that aside for now.  At the same time, even though selection pressures at different levels need not be aligned, it's not unreasonable to imagine that they often are.

This is all still a bit sketchier than I'd like it to be, but I think it sets up some interesting predictions.  For instance, individual variation in any particular trait should be inversely proportional to selective pressure working on that trait.  If there is strong selection for a suite of traits, then variation in those traits should be relatively low.  If selection is weak, then variation should be greater.  (And here I suppose the selection for vs. selection against changes the syntax a bit.  Whether it changes things more substantively is an open question.)

If all this holds true, then we can start asking interesting questions.  The simplest is, what happens when selection is weak at both the individual and group level?  The prediction in this case is that such a circumstance would allow high degrees of variation at the individual and group level.  Many different kinds of groups should be able to thrive in such an environment and the variation of individuals within each groups should also be high.  All groups would likely become loose groups, with high levels of variation.  

What if selection pressure at the group level is high, but selection at the individual level is low?  Under these conditions, we'd expect less variation in the kinds of groups that exist in an appropriate level of analysis, or perhaps more precisely, we'd expect groups existing within that context to exhibit less variation in those traits that are adaptive to the pressures being exerted.  On the other hand, the traits exhibited by individuals within each group could show considerable variation as long as the composition of individuals reliably evoked the group level adaptive traits needed for the group to survive.  Given that the group is under strong pressure in this scenario, we might expect there to be some core traits that are in some sense inviolable; without them, the group could not survive.  Again, groups facing this kind of differential selection pressure at the group and individual level should be loose.

Tight groups, then, should emerge any time that selection pressure at the individual and group levels are relatively high.  One thing I need to think more deeply about is what happens when selection pressures at the group level are weak relative to selection pressures at the individual level.  My sense right now is that this would not affect tightness so much as it would the variation in the kinds of groups that might emerge within the environment, but I'm not sure that really follows.  

However that may turn out, it seems to me that this basic framework, sketchy as it is, does at the very least accounts for many of our existing observations.  For instance, fundamentalism, which neatly fits our tight category, is associated with existential insecurity.  In other words, when individuals are living on the edge of existence, when they are experiencing strong selection pressures, they form tight groups. In the most extreme circumstances, the survival of the group and the survival of the individuals within the group are very closely aligned in that the traits that are individually adaptive within that context are likely very similar to group level traits that are similarly adaptive.  An interesting prediction from this that I don't think I've heard before is that because selection pressure at the group level is also high, you would expect to see a reduced variety of groups operating within that environment since relatively few kinds of groups will be adaptive.

In times of existential security, by contrast, selection pressures at the level of the individual are greatly reduced.  This should lead to a greater variety of groups and to a greater tolerance of diversity within those groups.  If these groups end up in competition with one another for some resource, then the variety of groups should diminish in response to the selection pressures, but tolerance within each group for diversity should remain relatively high as long as the diversity does not affect those group level traits that are adaptive.  This could account for why some of the loosest churches we've come across seem to have members who are middle class and uncharacteristically high levels of education.  This lines up neatly with some of the strict/lax literature, though it reaches the same predictions from a very different direction.  It also accounts for why many of those churches are struggling to survive.  Although in many cases they seem oblivious to the fact, they are in fierce competition, not only with other religious groups but, as David is so fond of saying lately, also with secular alternatives.  That they are failing to grow or even hold steady likely reflects that they have not found, despite (or perhaps because of) all their internal variation they have not hit upon the "trick" of surviving in the current social context.

Whew!  I could probably go on, but I doubt anyone has made it this far.  Hell, I don't even have the energy left to go back and proofread what I just wrote.  Nevertheless, I'm glad to have laid all this out and I welcome feedback.  I'm not quite there yet, but I think I'm getting closer and I think we're getting close to being able to make some novel predictions that we can test.  

We have been SO off course!

I'm in the middle of a reverie right now so this may all be bullshit.  I'd credit it to being drunk, but I've only had two glasses of chardonnay so I don't think I can get off that easily.  I do wish desperately, though, that I could have found a liquor store somewhere within walking distance of my hotel.

In any event, I had a revelation while eating pizza and drinking wine at a little joint here in Bloomington.  I started thinking about this tight loose distinction and when it might be adaptive to be a loose group.  As I've done so many times before, I started racking my brain about what sorts of environmental conditions might be conducive to looseness and I even came up with a couple of scenarios.  For instance, it occurred to me that it might be advantageous to be a loose group if you experienced a loss of membership through mass emigration.  This would allow you, at least in the short term, to reach out to new people to become members of your group.  However, this seemed at best to be a short term strategy.  I also wondered about colonization of a new habitat by multiple groups that, individually, were too small to recruit many resources so that it would be advantageous to join forces.

All of that may pan out, but it's ultimately focusing on the wrong parameters.  Completely!  I can't believe that I've been so stupid.

Yes, environmental conditions are important, but they're a distraction.  The real target we should have been pursuing all along is the conflict between group and individual level selection.  As we know, group level selection can only take place when between group selection is greater than individual level selection.  This is the story we're all familiar with.  What we've missed, however, is the relative strengths of these two levels of selection.  And therein lies the key!  If group level selection is only slightly greater than individual selection, that that should favor tight groups.  That's simply because individuals, though under less selective pressure than the group, are still under significant selective pressure.  This accounts for why fundamentalism and existential security are so tightly linked.  People living at the margins can do better when they belong to a group, but their security also depends very much on their individual adaptations.  This is not the case, however, when group level selection is very much greater than individual level selection.  Under those circumstances, what happens at the individual level becomes insignificant.  The individual can display all sorts of variation because the individual is under very little selection pressure.  This accounts for why existential security is so often associated with more "liberal" or loose churches.  People belonging to those congregations aren't under tremendous selection pressure because they're living quite comfortably as they are.

And that, in a nutshell, is it!  The rest is just commentary.  Or so it seems to me in my current state.

Biblical Behavior

So, I had hoped to have more to report from the workshop.  I may still post my notes when I finish transcribing them, but the fact is that absolutely nothing was resolved.  Given the diversity of disciplines, we probably needed at least a full week just to come to some sort of consensus regarding vocabulary.  That said, it was a good beginning of something, I'm just not sure what yet.

The sessions did, however, kindle some new ideas and I'm sure more will come to me as my thoughts on the meetings continue to gestate.  David presented our approach and preliminary results of the Biblical citation project and it was remarkably well received.  That was both gratifying and encouraging to say the least.

One question regarding the project that came up during one of the breaks was how or if we're measuring behaviors based on these Biblical citations.  That's something I've given some thought to, but aside from possibly getting some of Ian's surveys into the hands of parishioners in the churches we're studying I hadn't come up with much.

Today I had an idea as I was walking around in the Indiana sun.  Namely, why don't we have people read Biblical citations and play economic games?  After getting some basic demographic information and having participants answer some questions about their religiosity, we could have them play some kind of economic game to get a baseline.  Afterward, we could have them read some Biblical citations and respond to them in some way.  For instance, we could have them rate them for some quality such as submission or tell us how the citations make them feel.  Then we could have them play another round of the economic game or a different one to see if reading citations from different churches gets people to play differently.

I haven't formulated concrete questions this kind of methodology would address.  Basically, though, I want to be able to take a stab at answering the question of whether the Biblical citations different congregations choose are simply a reflection of the social dynamics within that congregation or whether there might be a causal relationship between reading the citations and behavior.

I'd welcome any thoughts y'all might have on this.  It seems like something we could pilot on Mechanical Turk and have ready to run in Evolution for Everyone this fall.  Hell, for all I know we might be able to run the whole thing on Mechanical Turk if we're clever about it.