navel-gazingpost-thesis analysis. Above is a plot of the number of primary sources (1908-1941) I cite by date of publication. (Published sources only, excluding newspaper articles -- of which there are a lot -- and government documents. Also, it's not just airpower stuff, though it mostly is.) I actually have no idea if it's a lot or not, and I'm sure there are some selection effects in there. But, although I've certainly not attempted any sort of statistical analysis (nor will I!), I think some features of the plot reflect real features of the airpower literature of period, at least as it relates to the bombing of civilians.
Firstly, there's a substantial increase in the number of sources in the 1930s, particularly from 1934 when there is a big peak. I argue in the thesis that this was only partly and indirectly due to the obvious reason (the arrival of Hitler in 1933). The more important reason was the World Disarmament Conference in Geneva, which ran between 1932 and 1934 (actually it went longer, but was dead in the water when Germany walked out). This roused airpower writers -- whether pro- or anti-disarmament -- to action, and gave them a reason to explain to the public the effects of bombing on cities. The slight rise from the late 1920s is also due to the conference, I think, or rather the optimistic Locarno-era preparations for it. The big peak in 1927 is a bit odd, though. Let's call that an outlier.
The other two noticeable peaks are in 1909 and 1938. The first was very early in the public's awareness of flight. That really started in 1908, but the possible defence implications came to the fore in 1909 -- the founding of the Aerial League of the British Empire, the first phantom airship panic, the publication of the first serious books on the topic. And of course the dreadnought panic -- it was a peak year for Anglo-German rivalry. The 1938 peak was the culmination of the building concern over the previous decade. What the plot doesn't show is that, unlike previous years, it was largely sceptical, based on evidence from the Spanish Civil War. The Sudeten crisis that September showed that the fear of the knock-out blow still had a strong grip on the public and the press. But afterwards there's a sharp decline in interest, which I maintain is real.
This is the same thing, but for secondary sources (i.e. published after 1941; again not just works on airpower). There's a superficial similarity, in that both plots slope upwards from left to right. But in this case that's much more likely to be an artifact, a function of the sources I've read and chosen to cite. Naturally I'm going to have a bias towards more recent sources, which build upon and extend earlier research. Earlier works will often lack the perspective that comes with distance, and they can be harder to find too, as libraries shuffle them to the stacks to offsite stores or dispose of them altogether.
Just as importantly, at least when it comes to policy studies, earlier secondary sources also lacked access to primary sources, despite being closer in time to them. That's something which does show up here. In the years after the Second World War, government documents were still confidential, and so it's mainly only the official histories which are of much use today, along with official document collections. (In some cases, in fact, they have not yet been superseded.) From the late 1950s, the (brand-new) 50-years rule meant that Edwardian-era documents began to become publicly available, and then First World War documents. If this had continued, it would have taken until the 1980s until historians had access to official sources for the 1930s! But luckily, in 1968 the fifty-year rule became a thirty-year rule, and by the mid-1970s the whole of the Second World War period was open for research. And that's exactly when the first detailed studies of British airpower policy, outside of the official histories, began to appear.
But I wonder what would have happened if the fifty-year rule had remained in place. Would airpower historians have been forced to look more widely for sources, instead of mining the (extremely rewarding) seams of government archives? Perhaps my own area would have been thoroughly worked over long ago?