The third chapter sets domestic relationships between humans and plants in the context of a broad range of plant-animal interactions, and of co-evolutionary relationships more generally. Such relationships are seen as completely natural phenomena, rather than as something mysterious. The fourth chapter looks at different aspects of domestication, considered as coevolution within an ecological framework. Human-plant relationship are classified according to which part of the plant is used - reproductive propagules, vegetative parts or asexual propagules - and according to intensity - into incidental, specialised and agricultural domestication. The ecological debate about the relationships between simplicity, complexity and diversity is touched on, along with the relationship between domestication and sedentism (including some basic foraging theory) and between domestication and different (r and K) selective regimes.
The fifth chapter is in many ways the core of the book. Here Rindos attempts to construct a completely mechanistic model for the origins of agriculture, building on the foundations laid down in the previous chapters. The argument is very difficult to follow (it took me as long to read this chapter as the first four chapters put together) and this is not because of the mathematics, which are actually quite simple. The main problem is that the distinctions between the premises, formal logic and conclusions of the model are not always clearly drawn, and that the material is repetitive. Despite some problems resulting from this, however, the chapter succeeds (as far as I can establish) in its main purpose, which is to construct a model for the origins of agriculture that requires neither intentional change in human behaviour nor population pressure as an external driving force. The major prediction of the model is that domestication is much older than agriculture.
The final chapter looks at the effects of agriculture on populations. In particular the development of the agroecology tends to increase instability and hence emigration rates and demographic fitness; the rapid spread of agricultural techniques is therefore to be expected. Some of the implications for the present and future are considered.
The Origins of Agriculture contains an truly extraordinary number of spelling and grammar mistakes; one suspects the copy-editor or typesetter must have borne some malice against the author! This is just annoying, however, and, apart from the comments on chapter five above, I can recommend this book wholeheartedly. In particular I am convinced of Rindos' central thesis, and feel that it must be taken into account (perhaps as a kind of 'neutral' model) by anyone theorising about the origins of agriculture.