I may regret this...
A book came across my desk yesterday, containing the following paragraph:
Mating consortships are also sometimes seen in chimpanzees and are particularly common in bonobo. In fact, a male and female bonobo may spend several weeks primarilly in each other's company. During this time, they mate often, even when the female isn't in estrus. But these relationships of longer duration aren't typical of chimpanzee [...] males and females.Esentials of physical anthropology / Robert Jurmain, Lynn Kilgore [and] Wendy Trevathan. 6th. ed. Belmont, CA : Thomson Wadsworth, 2006. ISBN 0495030619.
I only loked at it coz of the cute chimps on the cover. Anyway - that paragraph jumped out at me because of my recent discussions with Stuart. Basically, it suggested something that seems to me to drive a coach and horses through Knight's theories.
If consortship/pair bonding type relationships are imminent within our nearest evolutionary cousins, then we could well infer that such a practise was in existence within human evolution at some point - bonobo style relationships. If then we further consider that all it would take would be selective pressures (and benefits) to extend this kind of relationship, then we can begin to see modern human sexual relations emerging through a simple evolutionary process quite easilly.
The lack of oestrus Knight notes could in fact just be a consequence of bonobo sex - if oestrus stops being important we may simply have stopped selecting for it.
That is, I'd suggest, there is no need for a sex strikle theory to account for observable features.
Stuart asked for a rival theory - there's one. Hominids were engaging in consortships, those who had more solid consortships which lasted longer produce more succesful offspring, passing on the tendency for consortship and long term relationships we see in human history.