A post on James Malizewski's Grongardia blog has got me thinking about world building and the mixture of approaches that can be taken which I have called inspiration and perspiration. My definition of these is that inspiration is taking ideas and not ensuring that every little detail has been matched up and thought through or particularly not rejecting ideas as being out of place to readily. Perspiration is based on the painstaking development of the base concepts.
Of course pretty much any world building starts with a light bulb moment (the usual cartoon depiction of an idea), but after that there can be the divergent approaches of trying to construct a detailed, logically fleshed out setting with much perspiration or just detailing the immediately needed areas of the world and then improvising on the fly after that.
I suspect that the reason that the inspiration approach often produces more effective results is that it uses internalised knowledge of world building and existing settings that allows something that 'feels right' to be identified unconsciously. The other strength of the inspirational approach is that not elements in a world do make immediate sense or fit together in an obvious way allowing a more organic and realistic feeling with the setting.
Personally I've not been the biggest developer of my own worlds often preferring to detail an area in a published or historical setting such as Glorantha or the Wilderlands of High Fantasy, but even then I find that I can get things to fit together well, often from long experience of the setting. In Glorantha there is a much stronger sense of the world compared to the Wilderlands which has always had a fairly loose approach to anything like 'canon'.
I think my favourite source of inspiration is often listening to my players' comments as we are in a gaming session with that leading to me being able to create adventure hooks and seeds in the campaign and that can go on into other campaigns in that setting.