Here Winchester recounts the life of Joseph Needham, author of the authoritative series of books on the history of science in China. His press minions were sufficiently active last year that I somehow recalled the title of this book while looking for something else.
Overall it is quite well written, if a tad too salacious, and a tendency towards a shallow engagement with the academic side of things. (I object to his overly salacious treatment of Needham's eroticism.) More background on the political organisation of the Middle and Celestial Kingdoms in antiquity would have been most welcome, as would be a discussion on how China related to the region; technologically speaking, what came out of their entanglement with the Mongols, Indians and Vietnamese? The extended section on how Buddhism got introduced is the sort of thing there should have been more of, but even a thorough journalistic biography of Needham himself is probably beyond a book of this length.
The Needham Question, as to why China's progress stalled for so long, receives a cursory treatment and is largely dismissed along "you can't prove a negative" lines. I struggle with this attitude, as it implies that historicism can never really isolate the causes and effects of events, a charge that Popper levelled against Marxism. Also I fail to see why a similar question can't be asked of Egypt, India and Arabia, with their early innovations in mathematics and engineering. Perhaps the question cannot be resolved in some absolute way, but the kinds of discussion it generates are fascinating. For example, one line is that Western thought allowed the natural (phenomenological) to be decoupled from the supernatural (noumelogical), whereas the Chinese approach required holistic explanations. Roughly, that perhaps science proper requires modularity, some means of delimiting the claimed scope of purported laws of nature.
Winchester gives the impression that the ancient Chinese were excellent and creative engineers, but somewhat less interested in building abstract models of scientific phenomena; he (but perhaps not Needham) says nothing about Chinese parallels with the great strides taken by Newtown, Leibniz et al in developing the differential calculus in the 17th century. I really would have liked to understand what sort of logic the Chinese employed.
I was irritated to find that the bibliography is enormous; the non-specialist reader would have been better served by a much shorter list of entry points into this expansive topic.