Bob Nicholson – Remixing Digital Archives: The Victorian Meme Machine, IHR Digital History Seminar, 10 November 2015
Live notes, so an incomplete, partial record of what actually happened.
Tags: dhist
My asides in {}
Stream/Deck: http://ihrdighist.blogs.sas.ac.uk/2015/10/28/tuesday-10-november-2015-bob-nicholson-the-victorian-meme-machine/
Project came out of a pitch to the British Library Labs http://labs.bl.uk/
Idea to mine for jokes, create memes out of them, share them via @victorianhumour in Twitter. But the question perhaps never asked satisfactorily was why?
Themes of the talk
- what are the problems with Victorian jokes
- archive remixing (to make them work for us...)
- creative remixing
- The Victorian Meme Machine
Ideas underpinning the project - how we do research, how we organise our archives - are interoperable.
We are not amused has haunted the Victorians. But the Victorians loved jokes. Huge value placed on wit and punning in construction of sociability. Cheerfulness a Christian duty. Jokes sit in a liminal space between oral culture and visual culture, often captured by accident. Aim to build a collection for the world to use.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>.@DigiVictorian demolishing the lazy stereotype of the humourless Victorians. Exhibit A: The Punster's Pocketbook (1826) #dhist
— IHR Digital History (@IHRDigHist) November 10, 2015
Pocket books super important. The Wallet of Wit.
Puniana: thoughts wise and other-wise https://archive.org/details/puniana00rowlgoog
Just because stuff was thrown away doesn't mean they weren't important in the moment...
What are we going to do with these then?
Well, look at how these jokes for circulated. And how they kept going. Moving between private situations, newspapers, and books, and - presumably - various other ephemera that does not survive.
Jokes give us access to language and culture. Attitudes and ideas. And of course they have a historian significance.
How to go about working jokes through digital archives?
Lots of good work in this area, but often focuses on the canon. Exception Jack Bratton and Ann Featherstone, Victorian Clown (CUP). But all a tip of the iceberg, because there are millions of these things.
Archival structures give us access to archives in particular ways. 'Show me all the jokes' isn't built into the construction of the archive. Are there other structures?
Digital archives have destabilised these structures, to a point... Keyword search subverts the structure by which media was organised for contemporary readers.
Newspapers are extraordinarily varied texts. Heteroglossia. They are different texts. Texts that move between texts, independent things united by the newspaper {as technology}. And if we organised each thematic chunk for itself, we'd use different metadata for each.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>''You're constantly fighting against the weight of an archive when you're [keyword] searching' - well said @DigiVictorian ! #dhist #musetech
— Mia Ridge (@mia_out) November 10, 2015
Taking jokes out and bringing them together. There is a de-contextualisation here. But I'm very aware of this. And believe we need to get them out of text to deal with the noise.
Goal to have a high quality annotated archive of a million jokes. How we are going to do this is a slow burner.
What are the challenges:
- well for this one at least, we don't need to do much digitising
- instead, how do we get them out of archives? Access. Copyright... feels like work is limited by permissible rather than possible, even when permission is granted. Philosophical problem here. Data formats
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>#dhist @DigiVictorian used to think his research was limited by what's possible, now feels it's limited by what's permissible. #copyright
— Mia Ridge (@mia_out) November 10, 2015
- software that recognises jokes within other texts. Because jokes had structure
- integrating them into a new database
- meta-data, OCR, annotation. Modern ABBYY Finereader actually does quite a good job (though because jokes often at the bottom of pages, they can end up on a bend which makes the OCR go wonky)
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>The same text through different OCR tools (the non-free one is on the right) #dhist @DigiVictorian pic.twitter.com/0XIs1YVQd5
— Mia Ridge (@mia_out) November 10, 2015
How do you motivate people to add information? Crowdsourcing. Idea working with is a form of gamefication. Creative building of jokes that captures metadata: 'reciprocal impact'. Using daftness to build a rich research archive.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>#DHist @DigiVictorian 's hope is that crowdsourcing participants will be motivated by a tool that dynamically constructs cartoons of jokes
— Mia Ridge (@mia_out) November 10, 2015
So, two forms of remixing: archival remixing and creative remixing.
If we love the things we research, we should find ways of getting them out there, that place them in creative contexts.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>More questions than answers from @DigiVictorian at #dhist; but what is really important is the ambition - some fantastic challenges!
— Martin Hewitt (@ProfMHewitt) November 10, 2015
Q: Crowdsourcing mechanics. Would people spend more time on the cartoon than than the metadata? How would you manage that tension? A: Tension is there. Trying to capture energy of online expression.
Q: What is your 5-year hypothesis about what you are going to find? Where does this get you to? A: don't know... Spent years working on Victorian jokes. On the level of what I'd like to do with them: language, gender, trends, reprinting.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>#DHist Lots of work in tracking the spread and re-use of text and images at the moment - is that tech-led and/or a historiographical moment?
— Mia Ridge (@mia_out) November 10, 2015
Q: is this a subset? (because of their filler nature). A: yes, lots of genres of joke that could make the archival. What is the line between jokes, short stories, anecdotes?
Having to think about how a machine would read a joke makes you think about the structure to jokes in interesting ways.
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>I like that. How would a machine read a joke? as a way to think like a digital humanist #dhist
— Adam_Crymble (@Adam_Crymble) November 10, 2015
<script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>#dhist what do ‘serious’ scholars think of all this remixing jokes stuff? does this ruffle victorianist feathers?
— Adam_Crymble (@Adam_Crymble) November 10, 2015
Q: What do you feel are the relative benefits of crowdsourced meta vs a unified glossary of keywords / tags (created by archivist rather than users)? A: we want both. Create ways of filtering by types of data when the database built.
Q: You mentioned archival remixing and creative remixing. You are also remixing the reading experience (or perhaps, if Victorians read newspapers in the patchy way I do, faithfully recreating it).
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Exceptions: embeds to and from external sources, and direct quotations from speakers