Network Analysis and Simulations: Six Degrees of Francis Bacon

These visualisations, developed by Brian Ball, Tyler Gore, Marta Weber, and Peter West at Northeastern University London, present the findings of research carried out as part of the FY25 TIER 1 funded project New Digital Methods for Understanding the Impacts of Early Women Writers on Science and Philosophy.

This project is a London-Boston, collaborative project between Northeastern University London, the PolyGraphs project, the NULab, and the Women Writers Project (the project PIs are Sarah Connell, Julia Flanders, Brian Ball, and Peter West). The project examines and highlights the impacts of early women scientists and natural philosophers on the development of these disciplines during a formative period of the Enlightenment. The project takes Margaret Cavendish’s (1623–1673) engagements with the Royal Society as a case study in the relation between early women’s writing and early institutional science.

As part of this project, the team in London employed network analysis and computer network simulations to evaluate the status of women in early modern intellectual networks. Specifically, the team focused on the relationship between Cavendish and the founding members of the Royal Society. The primary database for this research was the Six Degrees of Francis Bacon Project project. Six Degrees of Francis Bacon is a digital reconstruction of a network consisting of 13,309 ‘nodes’ (i.e., historical figures) and the ‘edges’ that connect them to one another. In turn, the information about these nodes and the edges between them is drawn from the Oxford Dictionary of National Biography.

The visualisations below represent two kinds of findings of this networks research. First, findings pertaining to the status of women in the Six Degrees of Francis Bacon network. Second, findings of computer network simulations carried out in order to evaluate the impact of the structure of intellectual networks on the transfer of knowledge.

Gender Analysis

Drawing inspiration from the late Helen de Cruz who suggested that something like the ‘Bechdel test’, usually applied to films, might be applied to philosophy and history of philosophy research, this visualisation (Figure One) shows how many women in the Six Degrees network only share edges with men. There are a total of 12,068 men in the network and 966 women. When men are removed from the network, 358 women are left as isolated nodes with no direct edges.

Further analysis of the status of women in the Six Degrees network revealed that 88.8% of women in the network are in a relationship of what Sangiacomo and Beers call ‘familiar strangers’. To be in a relation of ‘familiar stranger’ is for two individuals to be connected to a third, but not connected to one another. So, if A and B both know C, but do not know one another, then A and B are ‘familiar strangers’. The findings below (Figure Two) thus further emphasize the degree to which women in the network are more likely to be connected to men than other women. The most well connected women in the network (perhaps unsurprisingly) are monarchs and women of very high social standing (Figure Three).

This visualisation (Figure Four) focuses specifically on Cavendish’s connections within the Six Degrees network. The image presents an ‘ego network’, i.e., a series of nodes and edges, from within the wider Six Degrees network, which are all connected to Cavendish. It is worth observing that, in line with the wider findings of women’s connections in the network, the majority (16 out of 68 nodes) of Cavendish’s connections are with men.

Simulating Knowledge Transfer

The second set of visualisations (Figures Five through Fourteen) represent the findings of computer network simulations that focus on Cavendish’s connections to the founding members of the Royal Society: William Brouncker (c.1620–1684), Robert Boyle (1627–1691), Robert Hooke (1635–1703), Henry Oldenburg (c.1618–1677), William Petty (1623–1687), John Wilkins (1614–1672), Christopher Wren (1632–1723), and Prince Rupert of the Rhine (1619–1682). Note that Prince Rupert does not appear in the first visualisation as he does not share any edges with other founding members in the Six Degrees network.

In four steps, these visualisations depict the findings of simulations that were carried out on three variations of this sub-network extracted from the wider Six Degrees network. The first network contains simply the founding members of the Royal Society listed above. The second contains the founding members plus Cavendish; in this variation, Cavendish retains her historically accurate connections (as per the Six Degrees data) to those founding members (so if she is not connected to a founding member in the wider Six Degrees network, then she is not connected to them here). The third variation is also the founding members plus Cavendish, but here Cavendish is connected, i.e., shares an edge with every other member. Roughly speaking, the aim was to speculatively assess how the transfer of knowledge within the Royal Society network may have been different if (a) Cavendish was also a founding member, and (b) if Cavendish was an especially well-connected founding member.

Roughly speaking, the simulations we employed work on the assumption that there is a ‘truth’ that each member of the network has a certain degree of belief in. For instance, let’s assume that each individual is working out whether they believe in the truth of the proposition a certain coin is weighted towards heads. Each step in the simulation is an opportunity (in this scenario) for each individual to flip the coin and re-assess their level of belief in that claim. What’s important is that individuals are able to consult with others they are connected to—and e.g., share their own beliefs. What our findings indicate is that better connected individuals (‘nodes’ in the network with more ‘edges’) on average reach the truth in fewer steps than those who are less well-connected.