As well as definitive books, you need a carefully graded way of proceeding through them - plus a very good index. As David Palmer pointed out in his novel <i>Emergence</i>, one would have to devise a 'system to locate specifics in such huge collection. Useless otherwise; researcher could spend most of life looking for data instead of using.'<p>Other critical books should include philosophy and science: it took a <i>long</i> time to build a technological civilization, and the Scientific Method is the cornerstone of this. (And yet even now, much of the world is still mired superstition/religion.)<p>More crucial lessons from John Wyndham's <i>The Day of the Triffids</i>: 'The most valuable part of our flying start is knowledge. That's the short cut to save us starting where our ancestors did. We've got it all there in books if we take the trouble to find out about it...<p>'From my reading of history, the thing you have to have to use knowledge is leisure. Where <i>everybody</i> has to work hard just to get a living and there is no leisure to think, knowledge stagnates, and people with it. The thinking has to be done largely by people who are not directly productive - by people who appear to be living almost entirely on the work of others, but are, in fact, a long-term investment. Learning grew up in the cities and in great institutions - it was the labour of the countryside that supported them...<p>'A community of our present size cannot hope to do more than exist and decline. If we stay here as we are, just ten of us now, the end is, quite inevitably, a gradual and useless fade-out. If there are children we shall be able to spare only enough time from our labour to give them just a rudimentary education; on generation further, and we shall have savages or clods. To hold our own, to make any use of all the knowledge in the libraries we must have the teacher, the doctor, and the leader, and we must be able to support them while they help us.'