Biologists recognise a number of common reasons why groups form - which include:
- Raising offspring;
- Safety in numbers;
- Hunting in packs;
- Gathering to mate;
- Kin selection -
Family groups are some of the most common reasons why groups form in nature.
Theres a well-established theory which covers this: kin selection. My main comment
here is to say that a symbiosis-aware version of kin selection theory is needed.
Groups can form not because their members are kin, but because their members'
symbiotic partners are kin. This idea should be taken to cover shared cultural
symbionts as well.
- Combined power - groups are often stronger and more capable than their individual members. They can do things which individuals cannot. For example, 10 ants may be able to transport a leaf which no single ant can lift. Groups are thus biology's multipliers. Groups can build bigger and better nests than individuals can. A pack of animals is can tackle larger prey than a single hunter. Similarly, a herd of animals can often mount a better joint defense against predators than a single animal can manage. Combining power does not require synergy - where the group is stronger than the sum of its component parts. Additivity (or worse) can still be enough. The term "power": is intended to refer to motor power, sensory power, compute power - or some combination of these.
- Sharing information - groups often form in order to share information between
their members. One classical way in which this hapens is the group members share their
genes with other group members by having sex with them. This happens on leks for example. This is part of the reason why groups consisting of members of the same species form so frequently. We clearly need a symbiosis-aware version of this idea. Hosts may form groups not so
they can have sex with each other, but so their parasites or symbionts can have sex.
Indeed, it isn't just sex - groups can form so that individuals can transmit their symbionts
to others who don't yet have them, or so individuals can gain symbionts from others.
Cultual symbionts also need to be included. For example, when Catholic groups form, one thing that happens is that the Catholic memes get to meet and mate with other Catholic memes. Some other things happen as well - which could be characterized as "collect the full set" and "get the latest upgrades".
Those with incomplete sets of Catholic memes can get new synergetic meme partners.
Those with out-of-date Catholic memes can upgrade to the latest versions. A good
generalized version of these ideas is to say that groups sometimes form in order to
share information between group members.
That's a case of sharing genetic information. However, there the information involved is shared across groups, not between individuals within the group.
Group selection is sometimes cited as a force driving group formation. However, the claims of group selection advocates seem to be covered by 1 and 2 above - at least if you use a post-1975 version of kin selection theory. The claim that group selection explains things that the 1960s versions of kin selection theory cannot does not seem to be worth very much.
Though not complete, I think the above principles explain most cases of group formation. The exceptions I am aware of are not far from this framework, but near to the edge of it.