Skip to content

Professor mines for historical memes

What makes a meme? This ques tion has sparked human curiosity for cen turies, but only recently have com puter sci en tists devel oped the tools to begin answering it. By scan ning enor mous data sets of dig i tized infor ma tion, researchers can now iden tify rep e ti tion in the global con ver sa tion, uncov ering the fea tures of media ranging from viral videos to out ra geously pop ular blog posts.

But the tools at researchers’ dis posal aren’t lim ited to inves ti ga tions of our present media con sump tion. In the 19th cen tury, before modern copy right laws were in place, article reprinting was wide spread and socially accept able. “It was kind of the wild west of pub lishing,” said Ryan Cordell, an assis tant pro fessor of Eng lish who is a member of the NULab for Texts, Maps, and Net works. He is col lab o rating with com puter and infor ma tion sci ences assis tant pro fessor David Smith to study the texts of that era. “One reason I’m inter ested in that period is that in many ways it’s anal o gous to the present day.”

With local, state, and fed eral ini tia tives to dig i tize his tor ical texts—including classic novels and small town news pa pers alike—the work of detecting the most pop ular cul tural ele ments of an era can be extended well into the past.

Cordell was recently named one of 20 mem bers of the inau gural class of Andrew W. Mellow Fel low ship Scholars in Crit ical Bib li og raphy, a pro gram housed at the Uni ver sity of Virginia’s renowned Rare Book School. The pro gram aims to pro vide young scholars with hands-​​on instruc tion in inter preting the mate rial forms of tex tual arti facts, from medieval man u scripts and early Amer ican hand-​​press books to born-​​digital materials.

“Part of why I’m excited about this fel low ship is a lot of my work tends to take place in these dig ital archives, these large scale repos i to ries we have of his tor ical texts,” Cordell said, noting that his research bridges a gap between tra di tional and new media investigations.

Once the news pa pers, mag a zines, pam phlets, and novels are trans lated into zeros and ones, Cordell and Smith develop algo rithms to comb through the entire data base. “It breaks all of that text down into little chunks—five-word chunks, 10-​​word chunks—and looks for repeated strings of texts,” said Cordell. If the algo rithm finds these chunks repeated sev eral times, the researchers can be fairly con fi dent the text was reprinted, and thus of cul tural sig nif i cance at the time.

So far, one of the most sur prising find ings has been evi dence of wide reprinting of “quasi-​​truthful anec dotes,” as Cordell put it. These were arti facts ranging from a poten tially fic tional letter from a wife to her dying hus band or an alcoholic’s story of reli gious sal va tion and reform. “What we’re finding in this study is that seems to have been an incred ibly pop ular genre in the 19th cen tury,” said Cordell.

Whether it’s an anec dote, a poem, or a pres i den tial speech, finding the rep e ti tions in printed media can lend insight into the people and values of the day. “My next incli na tion as a scholar of the period,” said Cordell, “is to ask why. What’s going on? What do these texts tell us about the people who wrote them and read them and wanted to pass them on? That part of the research is really just beginning.”

The next step, he said, is to figure out what they tell us about culture.

By Angela Herring

More Stories

Photo of the Capitol Building at night

High stakes for politics, SCOTUS in 2018

Photo of the crashed truck that was used in the October 31st attack in Manhattan.

Weaponizing Language: How the meaning of “allahu akbar” has been distorted

Northeastern logo

Why I love studying Spanish