» HOWTO: Organize Humans Hierarchically
In discussing Dunbar’s Number and urban tribes, I’m looking for fundamental scales of human organization. Ryley pointed me to two quite similar hierarchies:
| Colloquial term | Dunbar term | Dunbar size | military term | military size |
|---|---|---|---|---|
| family | support clique | 3 – 5 | team | 4 |
| extended family | sympathy group | 12 – 20 | squad | 9 – 10 |
| band | band | 30 – 50 | platoon | 16 – 44 |
| tribe | clan | 150 | company | 62 – 190 |
| ? | megaband | 500 | battalion | 300 – 1000 |
| subculture | tribe | 1500 | brigade | 3000 – 5000 |
The military grouping is often said to be consistent throughout history, although with support units it’s hard to be sure apples are compared to apples. I think the numbers have been tending downward in the last 100 years due to area-effect weapons and mobilization (you can control less troops when they’re moving around really fast and you don’t want to keep them too close together). This guy claims that you need 6 people to maintain sentries, which might have something to do with why teams are almost always deployed at least in pairs (eg: one of my favourite video games, Full Spectrum Warrior).
“Support cliques” and “sympathy groups” are so named by Dunbar because of specific ways of measuring them: your support clique is your first line of support in really tough times and your sympathy group is the people who’d be devastated if you died. This table makes it clear that the phrase “urban tribes” was coined by Ethan Watters without much background reading. Watters is talking about post-kinship, post-economic families and needed a snappy name.
Anthropologists are convinced that these group sizes are based on human cognitive limits. I’m not sure I’m ready to give up my dedication to 8ish as a common human grouping (for example, as the maximum size for an ad-hoc meeting) given that 7±2 is the most famous human cognitive limit.
I don’t think I’ve ever been to a useful meeting of more than four or five people. That was in video games though where there are only four or five disciplines. Meetings tend to be organized around team leads and their support personnel / area experts. Most actual production teams are sized in the clique/group range, the effective ones closer to clique.
All the larger meetings were upper management fapping about, panicking.
The most effective meeting I’ve ever been in was a three-person story conference on the set of my short: the Screenwriter, the Director, and the Producer. The Director actually kicked everyone else out of the room (grip, actors, hair & makeup) so the brainstorming session wouldn’t pollute / be-polluted-by them. It went like this:
D: “We can’t shoot scene W, there’s no time.”
S: “We could try X, Y, or Z.”
D: “I can film X!”
P: “We can afford X!”
Done!
Jack
9 Jul 09 at 11:26 am
Academic institutions use these size classes. I wonder if that’s intentional.
Work group: 3-5
Seminar Group: 12-20
Class: 30-50
I don’t know if tribes have an equivalent.
Cohort: 500
Department: 1500-2000
A lot of universities go larger than this with classes and department size, but one of the big criticisms you hear about big universities is that they’re “impersonal” and “you’re just a number”.
Kyla
9 Jul 09 at 9:33 pm
Cohort sizes for UVic CS when I was there were in the tribal (~150) range, certainly after the weeding courses.
Jack
9 Jul 09 at 11:17 pm
Well it doesn’t count if they just happen to be that size, the theory says that these sizes are a direct result of human cognitive abilities. In the military, everyone has three units reporting to them (eg: this is the case for absolutely everyone in Generation Kill).
A professor can only care about 30-50 students in a lecture. Each professor teaches 3 classes per year. When departments get to 12-20 professors they split into subspecialties*. So we should max out around 2000 students per department.
* Somewhere in The Structure of Scientific Revolutions Kuhn gives a number for global specialists in each area of research.
Jared
9 Jul 09 at 11:37 pm