Selected Authors from the Archive of Neutral Intentionality
Which questions to ask of the data? Which answers to communicate? How to communicate? Which answers are left out?
These are the problems we decided to grapple with during the Data as Collections Hackathon event in NUIG. Deciding to take a critical perspective on the archive and on the discipline of digital humanities allowed a type of 'performative process' approach the Hackathon. In this overview I will first address the formal elements of our project. In part two I will look at the project within my own disciplinary languages including some performative archival histories within the discipline of art, and in part three I will attempt to take a critical overview of the event in its entirety with the hope it might inform my own future research in transdisciplinarity within academic practices. While this is a text heavy overview, a more creative output of the work can be found by clicking here
Group formation was key to the event and somewhat by chance the group that I found myself in formed organically, stemming from brief and quite diverse conversations of interest. My group seemed to be the only one out of five which followed the organisers stated intent to consist of one coder (Lucas Azevedo), one digital humanites researcher (Patrick Egan, Ethnomusicologist and Web Developer) and one artist (myself). The importance of methodologies of group formation within an inter/transdisciplinary setting will be further discussed in part 3. Noting that other groups had formed which were either coder heavy, or artist/archivist heavy I feel this is the fundamental basis upon which the organisers reflections on efficacy may rest.
During one of the introductory presentations Andrea Fitzpatrick (director of the Cimera Art and Science Programme) facilitated a short discussion on some of the perceptions around the role of the artist within a collaborative setting. This discussion allowed an early assault on some of the perceptions of what collaboration with artists might look like, including artist instrumentalisation and the perception of the role of art as a representative outcome as opposed to a process of engagement, problematisation and synthesis. I felt this conversation was most useful to both the artists and those participating who might not be clear on the use of creative thinking in collaborative practice, and also allowed me optimism that the organisers were open to creative outputs which were unexpected or even subversive. As I will discuss in part three of this event analysis, the use of synthesis within a creative approach to any collaborative undertaking cannot be emphasized enough. The ability to synthesize multiple ideas within a group setting requires non linear thinking, strong negotiation skills and the ability to leave the ego behind and ones own disciplinary boundary gates open. A strong understanding of the importance of process as opposed to outcome becomes the difference between walking away having learned from and with each other, as opposed to imparting ones own knowledge as a way of self assertion or affirmation.
Patrick Egan particpated in the conversation about the role of the artist, as did Lucas Azevedo. (Patrick has created a very useful and specific blog post on the ins and outs of the event here). I had previously discussed a critical frame in addressing the hackathon with Patrick which I feel helped frame the final angle we were going to take on our project. Lucas Azevedo described his coding work to me during group formation, which looks towards the machine learning subject, with a particular interest in lie detection in text. I was interested in his work instantly and found conversation to flow freely with both Patrick and Lucas during the 'getting to know phase' of the workshop, leading us to form a group organically and without too much deliberation. Patrick seemed interested in pursuing his idea of working with three disparate databases, including spotify. I initially did not fully understand his interest but felt it facilitated the idea of the data we were goignt o look at as arbitrary, which in turn allowed me to form ideas of critical framing of the work in relation to both archives and digtial humanites practices.
Once we had formally decided on our group we were given a board and post it notes and we began the process of brainstorming, idea sharing and getting to know each others style of information dissemination about their own discipline capabilites. Patrick tells us that
"The original idea which led to our group formation was to interrogate on multiple levels what happens when you skew the truth in a collection/archive and what role data harvesting within such an archive can play in such a skewing. We decided to look for ways to find out how the archivist, transcriber or coder make decisions on the data which changes how the truth is projected or produced into the results. We decided to use an archive of grey literature, namely the Abbey Theatre meeting minutes from 1904-1939.
My own interest in grey literature stems from Boltlanski and Chiapellos The New Spirit of Capitalism along with my own textual, performative and artifact works on instituent practices. What may seem initially as benign, or as Lucas pointed out pretty dull minute texts of little interest or excitement, I was also able to make a correlation between grey literature and the vast reams of data which may or may not hold importance but to which the public are not normally subject. Grey literature is perhaps the predecessor to big data in that both can remain unseen while having societal and political effect. The subtle difference of the materiality of of grey literature and the non materiality of big data is something I wished to explore within any artistic outputs generated by our collective work.
The group embarked on a close reading of a random set of minute book pages. Two names that seemed to be common within the archive were Lady Gregory and WB Yeats. Patrick and Lucas began to use Python code and found 114 references to Lady Gregory, and over 500 references to WB Yeats. To understand this difference they performed some more data analysis with Natural Language Toolkit (NLTK) software from Stanford software and found that WB Yeats numbers were different if you use different algorithms for natural language (in context). This differential in output sparked my interest in my previous assumptions that data or numbers are fixed. The work Patrick and Lucas produced highlighted to me something I hadn't properly considered, that the problems coders must address and solve in some ways correlate to the problems artists must ask of themselves namely: am I asking the right question, am I happy with this set of results, is it possible there are different results from the same questions?
Patrick then used an alternate method, namely "Simple Concordance Program" to get the frequencies of other surnames, and keywords such as 'raise', 'proposals', 'difficulty' and 'decisions'. He found many problems with these words as they had been written by the minute taker. For Keywords in Context (KWIC) many of the entries were ambiguous, for example for the word 'proposal,' they may have been referring to proposals that happened in a previous meeting, so this is a duplication. Patrick found there were also other issues with words, for example "raise his salary, raised strength, raise the floor" arose as mistakes in the code. ( I would like to thank Patrick for his written summations on this work!)
Witnessing this type of data work was illuminating. Noting the coding language, the use of software, the seemingly arbitrary or nuanced questions we were asking of the digital archive and also addressing Lucas' concerns of our having no set reasoning why we were doing what we were doing, along with no perceived outcome. As an artist I found this very amusing as it is this indeterminacy from which much artwork and collaborative projects stem, and it is an early determinancy which can see such workshops fail in the desired outcome of something 'different'. Of course if you try to determine something from the start you follow a linear trajectory to reach a pre-determined goal. Without this pre-determinancy we were able to work in a rhizome-like way, weaving in and out of ideas, concepts and plans of action. This allowed us to make the whole process fun, playful and at times ridiculous and ultimately allowed us to come up with a multi layered response from our findings as well as address systemic questions of data, archive and art.
Ultimately after some deliberation we decided to run with the data we discovered on the concordance of Lady Gregory and Yeats. These concordance tests showed us a number of things, including the gender disparity we based our final outcome on. Different methods of coding the text gave different results, and also informed me of the concept of 'dirty data'. During one of the coding Lucas realised the empty lines in-between the lines of text were being counted as data. He was questioning why some texts were coming back as much longer or as having much more data than others and realised the machine was simply 'reading between the lines'. I understood it as the code reading the 'negative spaces'.
The code also was limited in some ways depending on how the text was originally justifed. Lucas spent alot of time 'cleaning' the data he was producing. He then made a graph with both sets of data from the two different concordance methods and ultimately we had four graphs to work with, along with the excell sheet Patrick was working on charting the whole process. My input was to try and find a way of synthesising all this data, and to try to find some interesting correlations between all the disciplines that we could work with to facilitate a concrete outcome from our seemingly arbitrary findings.
We continually referenced back to our post it notes . I suggested perhaps we create a new type of archive, one which included actual texts, projections of data, and placed it within a relevant time and space, considerations within the discipline of sculpture. For some of this process I felt it useful to leave the two working on their computers and occasionally input suggestions or thoughts. In some instances I did feel I wasnt fully understanding the processes they were using as I was the only one out of the three with no coding expereince. In hindsight I feel this was indeed helpful with our group dynamic. Three coders may well have seen us just reproducing work which coders normally do, which in my view would have been undesirable with regards to the nature of the event.
I also suggested we leave the room we had been working in for the duration. This is also a suggestion I would pose to the organisers. Conference rooms are specific spaces, a space which many of the participants would be familiar with in their working environment as teachers, researchers etc. This familiarity perhaps can impinge the 'outside of the box' type of response. I will elaborate on this in section 3.
I felt that as we were working within the biggest repository on the campus, namely the library, we might do a site visit and have a look at some of the books. I had my own mini projector with me and told the others I could do very basic projection mapping and perhaps we could combine the materiality of knowledge via books alongside the immateriality of data and digital archives in a way which might illuminate the problems and findings of our days work coding the archive of the Abbey Theatre minutes. We decided to see could we borrow the same amount of books as bars in our two graphs (14) and make the titles of these books relevant to the many stages and interests we had encountered in the workshop. I also noticed their was an exhibition in the foyer of the library of an archive, and thought it might be interesting to 'hack' our findings into this archive by way of an installation of the books and a projection of dirty data, representative in a moving video mapped onto the books of 'white noise'. The projection would map the graph visual onto the books, we would then video the installation, our archive, within and archive, and then archive the video onto a digital platform such as website.
In some ways I felt like the whole process was performative due to the fluid way we were working with ideas. I felt we were participating in our own performative archive, and that the outcome should be performed to the other groups, as opposed to a powerpoint presentation or any other traditional method. Patrick began our presentation in a formal way, describing the work undertaken in detail. Lucan described his coding output and I was then able to highlight the synthesis of our ideas which occurred when we decided to create ouptut which was multi layered and creative.
We presented the work last and received some positive feedback that our methodology or approach to the workshop had been inventive and creative. Although I did not present a 'performative lecture' I feel this might have been a great way to illuminate the work we had performed, however I am aware that this may be perceived as me sticking within the strict confines of my own field, and the field of archival performance which has some precedence in art and its histories.
"With digital archives and data collections drawn from humanities research as your foundation, you will collaborate in small groups of researchers and practitioners over two days to explore and create. You will walk away with a community of support, and an idea of the possibilities of using collections as data.
Over this free two-day hands-on hackathon-style event, you will explore what people from diverse backgrounds can create when they work together.You will work in groups of 3 people. Each group will consist of a humanities researcher, a developer / engineer, and an artist / designer. Over the course of the two days, you will work on a collection of your choice to produce an output that draws on the team’s range of interests, ideas and skills." - Collections as Data Event Information
As with any participatory or collaborative undertaking having a rooted understanding in ones own relevant disciplinary histories can only be beneficial. The archive within art histories as a form of practice has taken many forms, from simple techniques of archival representation to huge participatory performative archival productions. In The Public Art of Performative Archiving by Panos Kouros we are told "Performative archiving is construed here as a range of archival art practices which produce reflection for empowered public spheres in the context of today’s generalised archive culture." and "The archive has been discussed in the context of contemporary art practices as an impulse which encompasses a variety of different approaches to collection, documentation, classification, in combination with a wide range of media (photography, video, installations, actions)."
As stated in a previous blog post I was also familiar with Walid Raads work with the Atlas Group. In the late 1990s, Raad created a fictional foundation called The Atlas Group in order to better organize and more subtly contextualize his growing output of works documenting the Lebanese civil wars. Raad presented all the outputs of the archive he invented digitally. Raad used his fictional archive as a means to an end, as a method to question power.
Raads works seek to raise questions on the neutrality of history and how it is presented to us, and also interrogate the notion of time and how historical narratives can both be fact, fiction, familiar and unfamiliar. Raads work also looks at the archive as institution and performs institutional critique by creating multi layered fictional narratives. I was particularly interested in this type of approach for the Hackathon as felt it may help make the shift of the imaginary, that outcomes do not have to be immediately 'user friendly' or 'commerical' in order to achieve the shared goal of learning from others and working