Paul Revere and the Sociologists

Multiple people have sent me links to Prof. Kieran Healy’s satire “Using Metadata to Find Paul Revere,” which is also available at Slate. With all the talk about the U.S. government collecting metadata on our electronic communication, this case study of social networking in pre-Revolutionary Boston is getting extra attention. So, before more links show up in my inbox, I figured I should comment.

This paper has its roots in David Hackett Fischer’s fine 1994 book Paul Revere’s Ride. In an appendix Fischer and his grad students laid out seven lists of Whig activists in pre-Revolutionary Boston, such as the anti-Stamp Act Loyall Nine, the North End Caucus, and the town’s Committee of Correspondence. Their analysis was very simple: Dr. Joseph Warren appeared on more lists than anyone else, and Paul Revere appeared the second most often. Therefore, Fischer wrote, it made sense for those two men to be at the nexus of Patriot information-gathering and dissemination in Boston in 1775.

Furthermore, Fischer argued that Revere’s networking habits also extended outside town. He supposedly knew the militia officers of the country towns, so he knew whom to rouse on the night of 18-19 April. In contrast, William Dawes didn’t wake up anyone on his parallel ride to Lexington, and thus the south side of the “Battle Road” responded later than the north side.

I’m no longer convinced that personal habits are the whole explanation. Dawes was also a networker. He kept the records of the Boston militia regiment before the war and was secretary of the private Ancient & Honorable Artillery Company afterwards. His wife had family in Roxbury. Most sources speak of Dawes as a gregarious man. I think he didn’t wake up people because he didn’t think that was his mission.

Malcolm Gladwell picked up on the Revere case study in The Turning Point. Fans of that 2000 bestseller spread the word further, as in this essay at Search Engine Watch.

As Healy notes, he’s not the first social scientist to work with Fischer’s data. Shin-Kap Han wrote a similar, more serious paper (P.D.F. download). Both scholars apply far more sophisticated techniques for analyzing social networks than Fischer’s head count, but they get pretty much the same results: Warren and Revere had links everywhere.

However, any analysis is only as good as its data. And some of the info Fischer used isn’t as reliable as one would hope. Han recognized this when he used only five groupings:

The other two—the list of men known to have participated in the Boston Tea Party and the list of Whig leaders on a Tory enemies list—are not used, for those were not membership groups.

In fact, the “enemies list,” transcribed here, is the committee of men that Boston’s town meeting appointed in late 1774 to enforce the Continental Congress’s Association, or boycott on goods from Britain. The nasty gossip written beside those names, possibly by printer John Mein, make it a more interesting document, but the list of men itself is a significant group selected by the Patriots themselves.

The Tea Party list, on the other hand, is indeed problematic. It comes from Francis Drake’s Tea Leaves, published in 1884. Drake was quite credulous in accepting local and family traditions about which men participated in the Tea Party, so he listed David Kinnison (a minor living well outside Boston), Thomas Machin (still a soldier in His Majesty’s 23rd Regiment in New Jersey), and Dr. Thomas Young (seen by eyewitnesses at the Old South Meeting-House during the tea destruction).

Furthermore, the Long Room Club is another retrospective list. The first book to mention this group seems to be Samuel A. Drake’s Old Landmarks and Historic Personages of Boston in 1873, and he offered no source for his list of members. Fischer lists those same members in a footnote without citing Drake or any other source. But there seems to be no period confirmation of a political club that met above Edes & Gill’s print shop.

Finally, Fischer chose to include the members of the St. Andrew’s Lodge of Freemasons, where both Warren and Revere were leaders. That’s the only nonpolitical group on the list. There’s no doubt its members were involved in Whig activities, but including that group in the analysis and no similar groups (the St. John’s Lodge? the Boston bar? the Brattle Street Meeting? the state artillery regiment?) tilts the scale toward Warren and Revere.

In sum, I think Fischer’s analysis is probably accurate: Warren and Revere were among the most active political networkers in pre-Revolutionary Boston. Their contacts were important. But treating those seven lists as valid sociological data is a mistake, and a waste of good methodologies.