Steven Johnson's new book The Invention of Air is on the face of it about Joseph Priestley's discovery of oxygen, and the intellectual milieu which he lived and worked. Some parts of this story were very familiar to me, others less so. I for one was unaware of the extent and importance of his contacts with America's founding fathers, and his influence on them. One of the chapters describes Priestley's role in the clash of views between Thomas Jefferson and John Adams during the latter's presidency, after he had fled Britain and settled in rural Pennsylvania. However, this is not a straightforward biography, nor quite a history of ideas and society, of the kind exemplified by Jenny Uglow's The Lunar Men or Richard Holmes's The Age of Wonder. It would be a challenge to come up with a succinct elevator pitch. Not that this is a criticism. I thought it was terrific.

There are three other themes in the book. The first is that discovery is a social process, and the institutional arrangements supporting scientists shape the kind of discovery that takes place. There are reasons great discoveries occur when they do. (This reminded me of one of the messages of Malcolm Gladwell's Outliers, reviewed recently on this blog.) Johnson describes the contexts in which Priestley worked, the different constraints and the importance in particular the information networks in each case – the London coffee house, the patronage of a politically ambitious nobleman, the support of the Midlands industrialists who belonged to the Lunar Society and commissioned work useful for their businesses, and the much slower correspondence which characterised his years in the United States.

The second is that science, politics and faith or morality should not be regarded as separate realms. One of the features of Priestley's life Johnson admires is that he was active in each. He's scathing about modern politicians who know nothing and care less about science. And he rightly points out that there are many issues in modern politics where it's impossible to maintain the separation that has developed between these realms – think only of stem cell research and biotech. Or the recent political uproar in the UK about the Government's sacking of a respected scientist, having turned down the recommendation of his expert panel on the classification of cannabis, and objecting to his continuing to voice his own opinions about drugs. The border between technical or scientific advice, political choice and moral opinions is fraught, but we are not well served by pretending they can be kept separate.

The third theme is the importance of inter-disciplinary approaches. Johnson credits Priestley, in a correspondence with Benjamin Franklin, with being the first person to understand that the whole Earth is a system, of which the absorption of carbon dioxide and exhalation of oxygen by plants is just one part. He goes on to argue that a systems approach is essential and requires more inter-disciplinary science in future. Presumably, it's always been needed but is bound to be in tension with the growing specialization of science. I was struck by Johnson's aside in this context that radicals in Priestley's day were those who believed so ardently in progress that they thought institutions and society needed to be entirely reshaped to realize the promise of discovery – whereas self-appointed 'radicals' today are likely to be opposed to science and the very idea of progress (p238-239).

There are plenty of other fascinating thoughts, almost asides, in the book. For example: “This is a recurring theme of human history: major advances in civilization are almost invariably triggered by dramatic increases in the flow of energy through society.” Thus the north of England became the workshop of the world in the 19th century because that was where the coal was. I'm not sure this is true as otherwise surely Saudi Arabia would be the modern industrial dynamo? But it's an intriguing thought.

An interesting sidelight comes from a lunch at which I heard Steven Johnson talk about the software he uses to keep tabs on all the research he does – the things he reads, the papers, cuttings and notes stored on his computer. It's called DevonThink and appears to bring order to filing and to use machine intelligence to find connections between different pieces of stored material. As Johnson said, sometimes you wonder who's having the idea. Having heard him talk about it, I think the footprint of the software is visible. There are all kinds of thought-provoking connections in The Invention of Air. But on reflection, I suspect Johnson would have made them for himself. I enjoyed other books by him – Emergence and The Ghost Map. This new one is, more than they are, a loose network of ideas but still an intelligent, entertaining read.