Pros and cons of diversity in genetic and memetic evolution

Diversity has become a liberal buzzword. But is diversity always a good thing?

Diversity is good in a biological or cultural ecosystem because it implies resiliency. Some of the resiliency is due to redundancy and some to complexity. If one pathway fails, another can be a backup. There are many sources of food and each species or culture has many functions

Diversity is also good for getting out of a local fitness maximum which is not global, in a very specific way.

But the resiliency/inertia of diversity can also mean being stuck in a rut (aka local but not global fitness maximum), due to inertia. In order to get from the rut to a better place, one needs not diversity within a group but unity of purpose. A small population with a single mutation (or epigenetic change) in the right direction is more likely to overcome the  entropic or (negative) fitness "barrier" than a large population with a diversity of mutations. Many diverse groups (but internally genetically or memetically similar) have a better chance for at least one of them to find the right mutational direction for getting over an entropic or a (negative) fitness "barrier". Diversity helps to provide many possible mutations, but they must not interact, because they will most likely effectively cancel each other out.

Diversity not only reduces focus of direction (which is important for finding the uphill direction, or in some cases the mountainpass direction in fitness space), but reduces the ability of a group to improve its internal as opposed to external fitness, something we can call nesting. When a new direction is focused on, an incipient culture (even more so than an incipient species) is able to provide an internal environment where that new meme or gene, and all the ones networked to it, are selected for, while the external environment of the mother culture or species most likely selects against these genes or memes. Diversity would be harmless for either directional focus or nesting, if the genes or memes that are diverse within the population are all part of the same network and are acting in tandem even as they are different than the unmutated ones in the mother species or culture. But this is extremely unlikely, as gene or meme networks impose stringent constraints on what changes are selected for or against. Independent changes of many genes or memes are almost certainly going to be selected against, internally. Changes in one gene or meme have a better chance of being internally selected for. And if that one gene of meme can adaptively control other genes or memes (non-mutationally), than it is possible for many genes or memes to change, but that is not an independent change in all of of them, just in one.


The most successful sequence for getting out of a rut, finding a better fitness maximum and then being able to be resilent and responsive from there, is:
1. form many groups, each having a single mutation within the group, but different than all other groups.
2. Once the group gets over the fitness or entropic barrier, reintroduce memetic or genetic diversity into the group, perhaps through reduced isolation and interaction with other groups.


There are many examples of groups that were united in one meme and able to get out of a cultural rut into a new culture:
The Amish, the Bruderhoff, the Hutterites, the Hare Krishnas. Whether these groups are resilient or not remains to be seen and depends on how much diversity they introduce.

There are also examples of groups that have not made it out of cultural ruts because they are paralyzed by  premature diversity.