Thanks Gregg for pointing me back at this one - a great paper on the importance of structural holes in networks. This is a formal representation of principles I and others have explored relating to parcellation in social spaces, and the need for (small) bridges and isthmuses to connect those parcellated spaces. Further proof and rationale from a different angle that (as is often so) Darwin was right about this.

It makes me wonder whether the principles the authors look at might be extended somewhat to consider temporal flux in more detail: the reason Darwin found this kind of hole-filled network interesting was that evolution happens more rapidly and branches sooner in small, parcellated spaces (like, famously, the Galapagos Islands). However, he noted that, very rarely and occasionally, these spaces need to connect with others in order for adaptations to flow into the broader ecology or, more often than not, be swamped by them. But maybe not entirely swamped: an interesting article in New Scientist recently suggests that early homo-sapiens coming out of Africa bred occasionally with existing hominids in Europe and Asia and, though only a small percentage of Neanderthal and Denisovan genes persist in modern populations, they definitely had an important impact, especially on immunity to disease. Lots of good lessons on social environment design from this.