On Monday 11th of May we are introduced to the working methods, goals andsignificance of the Digital Humanities Hackathon and digital humanities as an emerging field of study. We divide into groups, and start familiarizing ourselves to the project and materials. Our group is led by historians of technology and industrialization Petri Paju from University of Turku and Mats Fridlund from Aalto University, and the material we work with are the three digitized Finnish technical journals Teknikern, Tekniska Föreningens i Finland Förhandlingar och Suomen Teollisuuslehti from 1880-1910 in DIGI – Finnish National Library’s Digital Collections. We are eager to begin work, and below is a summary of our daily progress.
Monday, 11th of May
We are familiarizing ourselves with the material by close reading and talking to our group leader historian of technology Petri Paju about the history of the industrial breakthrough in Finland. Through the discussions we are generating possible research topics, the light bulb soon emerges as a possibly enlightening case study. Various quantitative digital methods are emerging: how to locate occurrences of “light bulb” glödlampa (Finnish and Swedish terms and related words, such as “arc-lamp” and “electricity”)
We discuss the possible problems related to OCR – eg. garbled text in the wrong order will inevitably pose a problem for word frequency, so will advertisements with images (these will not necessarily be recognized). A preliminary word list of light bulb and related words for electric light are generated manually. We decide to look for co-occurring words near mentionings of light-bulb words in the articles: place names (foreign and domestic), personal names, value words. we all work together at our table and our digital literate computer scientists and computational linguists generate a coding script/algorithm to extract occurrences of words from our journal material.
Tuesday, 12th of May
Our second group leader historian of industrialization Mats Fridlund from Aalto University joins the group. Today the work continues on code/script as well historical background reading on the adoption of electricity in Finland.
The possible research topics are narrowed down and geographical data becomes our focus. We are interested in whether place name references co-occur with “light bulb” in any significant way. We debate the best way to look for place names. Compiling a list manually? We decide to search automatically for words beginning with a capital letter in the vicinity (+/-25 words) of “light bulb”, and manually pick out place names (as opposed to personal names etc.) We present our work and get feedback from other groups. Our first practical preliminary results on word frequencies are obtained. Graphs based on frequency over time show that the occurrence of “light bulb” and the old lighting technology “arch lamp” are roughly in correlation. Graphs=Science!
Wednesday, 13th of May
Today our group is joined by historian Jessica Parland-von Essen that help us out with advice and questions. We are collecting pictures of electric lamps from the advertisements in the journals and these pictures are used to illustrate our poster that will present a preliminary version of the project and that is to be used after Friday’s final presentations. Our poster is now ready to be printed, and it contains the outlines of the project, some cool pictures and a visualization.
Now we move on to tackle the issue of formulating our larger research question; brainstorming! Everybody suggests a research question and we select the best ones, coming up with 4 relevant questions. All of us have a slightly different approach to the research question, depending on our own areas of expertise; we attempt to find the questions that can best be answered with the material and the methods we have. we continue to edit the word lists, coming up with a list of words which we believe to carry the most interesting meanings, to be searched from the contexts of the ‘light bulb’-words.
In the afternoon we present some preliminary results of our work to the other groups, and get valuable feedback on our research questions. Our leader’s alsostress the importance of the ‘three-year-plan’: we want our research plan to be extensive enough to be a three-year project and not just five intense days!
Thursday, 14th of May
We continue working with the word lists, and our computational linguist Antti teaches us about the SVD Singular Value Decomposition method that we are applying on them -> the humanists learn something digital!
Friday’s presentation is approaching, so we focus on writing our research plan and preparing material for the presentation, while work on code/script also continues. The first map of the place names mentioned in the contexts of light bulbs is finished -> excitement! Our ethnologist Anna finnish her close reading of one of the earliest Finnish articles mentioning light-bulbs, to get another qualitative point of view on the material complementing the quantitative distant reading using keywords and SVD. The day ends with us presenting our work to the other groups once again and everything works awesome!
Friday, 15th of May
The morning of the final day of the hackathon is spent on finishing our presentation. After lunch there is an invited guest speaker and then the group have their presentations. After we suceeded in crashing the presentation systems we give our presentation Digital history re-reading of Finnish industrialization: Digital distant & cultural close reading of technology transfer, diffusion & use in Finland, 1880-1910 which is awesome. Our group leader Mats Fridlund walks around taking pictures and live tweeting. All goes well now and we get good questions from the commentator. After all the presentations wine is served and outside guests come up to us where we are standing next to our poster and asks us about our project work. It was exciting to introduce our project to the audience, and we were all quite happy with what we accomplished!