I'd like to map and track ideas that were debated in several little magazines of the 1920s to present a visualisation of where those "conversations" took place and how they changed with each contribution. It would be great to learn how others would approach this. Thanks!
Best way to map & track debates through several little 1920s magazines?
(7 posts) (4 voices)-
Posted 5 years ago Permalink
-
Replying to @Rosvita TEXTWinder's post:
The approach would depend on a few factors. First, do you need to analyze the data or has this already been done? If not, what format are these magazines in? How many pages? Have they been scanned and are they available as plain text? What's the unit of measurement you're interested in tracking? A term? How many interlocutors? Is the visualization to be interactive or a static infographic? What audiences will use the visualization and what knowledge work tasks will they want to preform?
Posted 5 years ago Permalink -
Replying to @Jim Ridolfo's post:
Thanks Jim, very useful questions. I am in the process of analyzing the data. The magazines are in hardcopy print format, and due to copyright restrictions, not going to be digitized any time soon. I have images that I have made personally, but I would not be permitted to post them. Indeed, I am attempting to work around the copyright restriction by "visualizing" my critical analysis to show how magazines transferred ideas about modernity.
Posted 5 years ago Permalink -
I wonder if you could add a few more details about what you're working with and what you want to do. I would suggestion the following, though how relevant these suggestions are will be a function of the particular materials you're working with.
If you were working with some large amount of plaintext data (borrowed from somewhere like The Modernist Journals Project), "text mining" techniques (like topic modeling) might suggest themselves (and one might want to use a tool like MALLET.
But it sounds like you're using images just as an aid to your research, perhaps (am I right?) not even OCR'ing them. In which case you won't be be working with any plaintext data, but will be generating some sort of data yourself by hand (rather than algorithmically) and are looking for a way to visualize/understand/interpret the results.
In which case, you might consider network analysis; this post, which uses a piece of software called Gephi, might serve as a model.
The real task then will be figuring out how to go from the magazines you're interested in to data. You might simply record: author/issue data/tags and then go from there.
I hope those links are at least a little helpful; and I'd love to hear more about what you're working on.
(Hat tips to both Jeff Drouin and Ted Underwood, linked to above.)
Posted 5 years ago Permalink -
Thank you, Chris! Yes, a very useful tweet from the Magazine Modernisms (Jeff Drouin, perhaps?)compelled me to click through to http://magmods.wordpress.com/2012/02/17/network-analysis-text-mining-and-emergence-in-the-september-1918-little-review/#comments . I had read the blog entry before, but sometimes a nudge can confirm that one is heading in the right direction. I'd also come across the Ted Underwood article, but while topic modeling has been on my horizon for almost a decade, as I understand it, the word level of analysis seems, to me, almost too detailed, while the modeling aspect remains intriguing. The fact that I am now returning again and again to the same places is, however, heartening.
Basically, the issue for almost anyone looking at modernism through post-1922 magazines is that it is difficult to work around copyright restrictions. As you will know if you check sites like the MJP, the little magazines after 1922 are not usually accessible online; if you are an "alt-ac" then gaining access to libraries for hardcopy research can be complicated... Where one is allowed to take photos of the magazines (thanks to Emory University, I have jpegs of the HOUND & HORN pages I currently need) it's only under agreement that those images not be published. Requesting copyright from the estates of those involved in publishing little magazines is a difficult process with little reward. This is a well-known tale of copyright clearly hindering research...
Response: I am interested in modeling the NETWORKS that were established through little magazines of the late 1920s. As I understand Katherine Hayles idea of emergence, through Jeff Drouin's blog (as a “a particular kind of complexity that arises not from the individual elements of a system, but only from their interaction (15)” there is more to be understood about the construction and definition of modernism if we map its emergence through the conversations, the "interactions," as they took place in the dialogues between the little magazines.
Mapping the networks/conversations, rather than digitising the magazines, would mean that copyright would not be an issue; the dynamic of emergence could be researched further; and ideally, collaboration between others working on little magazines (and facing the same restrictions) would be fostered. From working on Latin American little magazines, I know where this conversation played out in some of those contemporaneous magazines. I'm sure others working on, say, transition or THE HOUND & HORN know where those conversations took place. Ideally, as copyright permission was received, then one could link through to digitised pages, but for now, bibliographic reference to and critical analysis of those inter-publication debates would have to suffice. A digital project sans digital texts, for now.
In essence, my question here was: What would be the best way to begin that mapping?
I'm currently working my way through Jeff Drouin's report on their work on the Little Review issue and beginning to consider what kind of information a dataset, in my proposed project, would entail. Any further insights, suggestions and ideas would be greatly appreciated. That is, all except the now well-worn "tip" of working pre-1922!
Thanks! Rosvita
Posted 5 years ago Permalink -
Replying to @Rosvita TEXTWinder's post:
Hi Rosvita,
Which networks are you referring to? Magazine publishers? Authors? Distributors? Places? Ideas within the text? Do you have this information already, or were you hoping to derive it from the text?
One thing that's confusing me about your question is that you seem to be conflating digitizing these magazines (i.e., scanning them and running text-recognition software on them) with publishing them. The former is generally permissible for your own research; the latter can get you into trouble with copyright. (Here's an example of someone who did the former with a copyrighted work.) What Chris is asking, I think, is whether you have run text-recognition software on your page images. If you have done this, you have a dataset to work with and can begin to use tools to produce visualizations. If you haven't, you'll need to either digitize your sources or create networks by hand.
I wrote a post on creating a basic network visualization that you may find helpful. Perhaps this can give you some idea of how to get started.
All best,
MiriamPosted 5 years ago Permalink -
Replying to @Miriam Posner's post:
Hi Miriam, and thanks for taking the time to respond; every insight helps.
The questions being posed are really stimulating, and merit more than just a quick response, but in short, there are "conversations" that take place in the little magazines as contributors and editors respond to ideas they have picked up from other publications. They often comment on those ideas, adding their own considerations, and then they are picked up elsewhere. I have tracked conversations from a publication in Madrid, over T. S. Eliot's Criterion, to a small Cuban publication in the late 1920s as it moved from the main section of a magazine to a translation and to a mention, almost an aside. If it got picked up again, I'm not yet aware of it, but someone else might be.
Along the way, of course, there are slippages in meaning, new interpretations, re-inventions and a whole host of often incremental readings that vary from context to context but that clearly result from the dynamics between little magazines. Often, for this reason, reading from magazine to magazine can be more useful than reading cover to cover. Recuperating those cross-publication readings is what I am trying to achieve.
Doing a thematic tracking is possible, of say, the debates around humanism through a number of contemporary little magazines. Or it would be possible to re-create the network magazines quite obviously constructed the fairly ubiquitous "Periodical Reviews" that seem to have been de rigeur. Adding location/publication information to either of those mappings would not be difficult and would reveal where certain ideas played out with greater or lesser intensity. For instance, a mapping that reveals who took up the humanism debates, where, why, when, would be useful: for instance, did they ring "truer" for Argentines than for Peninsular writers? Or did they wane first, then return in greater strength because of factor X, Y or Z? Yes, I can write about this in an article, but there seems to be an opportunity for imaginative collaborative work here that quickly goes beyond the bounds of traditional publication.
I wasn't aware that I was conflating digitising with publishing. As I indicated, I have digital images of magazines that I can use, but know that I can not publish them. What may have been confusing is that I stated my hope that a map I create now may ONE DAY have links to digitised OCR'd pages. What I can do now is track the conversations. Once I figure out how!
I have digital images, but they are not OCR'd. As an "indie," those are not resources I have available to me, and I'm not entirely sure that's where my project is aimed. In any case, I will need to create the datasets by hand. But as I go through the images, to analyze and write about them, I have a growing list of tracked debates. What might be useful at this stage would possibly be to examine a variety of datasets to see how people have organized their data in relation to maps that have resulted from them?
Does this go some way to clarifying what it is I am trying to do?
Again, all consideration is deeply appreciated.
RosvitaPosted 5 years ago Permalink
Reply
You must log in to post.
