Dissertations in the IsisCB…and what they say about our history
Recently, I have been analyzing the dissertation data in the IsisCB, looking for information that might help us understand the history of our discipline. The dissertation records have a lot of rich information in them, and that information can be put to use in understanding more about the institutional context of history of science. The first thing I did was to simply tally up all the dissertations by school. In this way we can see which schools were the most influential in graduating scholars in the discipline over the last forty years. (In the chart to the right, the top 50 schools are shown ranked by number of theses produced. Only some of the school names are immediately visible; to see the name corresponding to each bar on the graph, just hover your mouse pointer over the one you want to see.)
There are slightly over 6,000 dissertations in the IsisCB which can be used to explore the institutional development of history of science. Anyone familiar with the history of our discipline won’t be too surprised by the list of schools at the top, but the size and number of theses is impressive for a modestly-sized discipline like ours. One can go directly to the list of Theses in IsisCB Explore if you are interested, and you’ll be shown the whole list, including a number of dissertations that I’ve excluded from my analysis.
The graph raises interesting questions. I have not yet separated schools with specific history of science programs from other schools, but doing so would help us understand more about the reasons for specific strengths.
Even given their limitations, the data are revealing. Chronologically, they show how institutional output changed over forty years. In the table below, I have periodized the results by grouping the theses into five-year intervals. The values to the right of each school represent the number of theses in each period. (To keep it manageable, I’ve shown only the first fifteen schools in each column. You can see the whole table that shows all institutions with 5 or more theses in each five-year interval here.)
The patterns here seem to be relatively accurate for North America, indicating the ebb and flow of institutional strength over time as scholars come and go and institutional priorities change.
We can see this better if we explore the changes over time by institution. The graph below does this visually by focusing on what I consider the powerhouse institutions, those that produced an average of two dissertations per year during any five-year period:
This is a fascinating picture of how scholarship has emerged in the past forty years. You can see the major history of science institutions in the discipline at the bottom and how they are gradually supplemented by a more diverse cohort as new players enter the field, especially in the early 21st century.
To make that pattern more evident, the following graph to the right shows only those schools that have only broken through the 10-dissertation cutoff three or fewer times in the past 40 years. Notice how the number of new powerhouse schools shoots up in the 2000s and continues into the second decade of the century.
I’m not sure what to make of the decline in the last five-year interval. I’ve compared it with the DAI data, and it tracks well with that. According to the most recent NSF survey, humanities PhDs have not declined in real numbers during that time (https://www.nsf.gov/statistics/srvydoctorates), so the change seems to reflect something about the way that the field is developing relative to other humanities disciplines. I’ll certainly be looking into this as I move forward.
One should be careful not to read too much into those numbers. The CB data are inevitably skewed. There is no way that a single bibliographer can be entirely comprehensive or objective in his or her selection. The selection is affected by both the individual curator’s professional judgments about what constitutes a relevant work for inclusion and by the limitations of the resources available to find dissertations. The most serious gap that I find in this list is that it is almost entirely from the US. There are two schools in Canada that make the top 50 (Toronto and York) but you don’t even find Oxford and Cambridge until you get far down the list. European (and to a far lesser extent South American and Asian) institutions simply are missing. This is largely a gap in the resources available to find theses. Both John Neu and I have used Dissertation Abstracts extensively, and it has had a strong American focus over the years.
Looking at data visually like this can be very useful because it brings home the biases in the data. This visualization helps both me as curator and you as user to understand where the limits are and when you and I ought to start looking elsewhere to pick up items that are simply missing.