idiolect

And just a couple more comments which arrived by email, on the filing problem, and solutions

GP Says:

I opt for the ‘disorganised piles in boxes’ approach, and a brain which alway knows that ‘I’ve read something about that somewhere, now what was it and what did I do with it…’. It doesn’t work very well, but neither has it ever crashed…
If you come up with a good system let me know!

JS Says (in ps)

PS. My colleague, Dan, (who’s very into data visualisation) mentioned http://websom.hut.fi/websom/ – which probably doesn’t help too much, but is quite cool. Dan says “if your brother wants to talk about self organising maps, I’m more than willing”.

Just a few responses i had to my post about the filing problem. I’m still interested to hear from any one else who has good advice on information management systems

MD said:

You know you’re having a frustrated day when you reply quickly to emails about filing. Nonetheless:

All my bibliography stuff is handled by something called BibTeX which is part of a typesetting system called LaTeX. Maybe you’ve come across it? (It’s all very open source and free etc..)

You essentially have a text-based database that stores all your referebces, with as much or as little meta-info as you want. Then you write all your writing in LaTeX which is a mark-up language useful for mathsy documents (but I reckon dead useful in general) which can include citations. The LaTeX document looks at your BibTeX file and pulls out and formats the appropriate references.

LaTeX is generally great, especially when you want to combine two documents and you don’t want to re-number everything (sections, figures, equations etc). So my thesis will be a cut-and-paste job of my articles/reports, and I’ll not have to re-format everything. Plus it looks good.

LaTeX might represent an over the top learngin curve for non-maths stuff, I don’t know. Have no idea how to keep track of paper stuff.

RS said:

Personally I would just build a small access database. Each discrete item gets several properties: format, keywords, location etc. for hard copies you can then just have a numerical index linked to a database entry. For soft copies each item could be hyperlinked or embedded in the database. You would then find it easy to generate search queries based on keywords. You could easily flag sources that are referenced in individual papers you are working on and if laid out correctly you could then auto-generate a list of bibliographic references. Whilst I don’t know how you could exactly do it I think it should also be possible to link references within a document to the database entry. Whilst it would take a while to build I think it would be quite a powerful tool.

You could store notes separately and cross reference papers within the database directly. It would mean entering a bunch of data each time you read a paper but it would probably repay the time invested.

WJ, displaying an uncharacteristic lack of rigour, said:

Disorganisation is the key to innovative academic research 😉

I’m not a good example…. hard copies are rare…. pdfs are by
subject mainly (and with replications for when i’m writing a
particular paper)…. format is always of the type “Jennings (2004)
The Politics of Celebration.pdf” [i.e. name (date) short title for own
personal reference].

have a few articles stored by author. but only a few.

KG said:

M$ Access all the way. I’d love to know if there is a GPU licenced alternative with similar functionality, but last time I asked a free software geek he suggested that if you want a graphic interface you’re stuffed.

Access cos flexible, hyperlinkable, searchable, expandable and webable (if your really keen). I store references to all my info in one integrated database, (split according to source type: authored book, edited book, journal, web, other) including a section for notes on each. I store a list of things to read there too, and can search titles and authors in the whole lot in one go, but filter unread if you just want something you’ve actually seen. My bibliography creation is pretty clunky (read: doesn’t work) but will be sortable by the time I’ve got a thesis to add a bibliography to.

There’s also the potential bonus that if you’ve got a bunch of friendly academics or research team all using access on the same network you could actually use the same db and choose to just see your references or the whole project’s, including notes they’ve made on readings – if you trust x and he says paper y is crap, then thats one less to read!

Anyways, come round for a test drive, and you can have the shell of mine to play with if you like.

I am back in academia, and now i must decide how i’m going to sort out my papers. I’ve been avoiding the problem, but the rate at which i’m accumulating paper, and PDFs, means I’ve now got to settle on some method for keeping track of everything. Whatever the system is, it needs to facilitate the these things for me:

1. Keeping track of my paper copies of journal articles (which ones i have, where they are)

2. Keeping track of my PDF copies of journal articles (which ones i have, where they are)

3. Searching both paper and electronic copies of journal articles

4. Easily inserting (auto-formatted) journal references into manuscripts

5. Organising and searching of non-journal bits of paper (newspaper cuttings, unpublished manuscripts, miscellaneous text, etc).
[Or, at least, those are what I think my needs are]

I’m interested in how everyone else does this. I presume that all academics (and most informational professionals, for that matter) have to deal with the overwhelming amount of information it is possible to accumulate. I feel/hope there’s a good solution out there, but I’m struggling to reconcile all the different types of information i want to keep and use, and the different purposes for which I want to use them.

One solution I’ve seen is to keep all the papers in topic-organised folders, and a list of what you’ve got in something like Endnote. Problem with this is that I’m never very comfortable with topic-based organisation. I’ve never managed to specialise (read: focus) on one field, so there isn’t really the coherence in my papers that makes organising my topic natural. Papers speak to several topics, or don’t fit neatly into any.

My PhD supervisor uses biblioscape to keep a record of what he’s got on paper, and he organises the hard-copies by accession number – ie as he gets a paper he puts a number in the top corner and stores them all in numerical order. It’s a neat way of getting round issues with deciding how to store the papers. It means that recent papers are on top (so to speak), but it does mean you lose any other implicit organisation that might arise from, say, storing by first author or topic. Plus he’s dependent on his electronic index working (and by most reports biblioscape is a bit flakey. And the reference insertion add-on i never got to work perfectly too).

One other guy in the research group writes short pieces of text which reference each paper he’s got. It’s a good way or giving more coherence to papers he has read before filing them away. I don’t know how he keeps track of the actual paper (and PDFs as well).

Other people (including my co-author on the book) use a self-authored web-app to keep track of their notes, hyperlinks and documents. This has the advantage of being accessible anywhere, and – since it’s web based – there’s lots of cross-referencing and search functionality already ‘built in’. In fact, after the book, I was so taken with how useful wikis are, that i’ve started my own to keep my notes and hyperlinks – but it doesn’t seem ideal for organising my bits of paper. I’m happy storing my own short notes, memos and links on a combination of this blog and my personal wiki, but i’m still left with the paper problem (ie items 1 through 5 on my wish list).

This is the system I adopted during my PhD: A filing cabinet with all journal papers organised by author; Using Endnote to keep track of what is in the filing cabinet, with a keyword to indicate if it was in electronic form rather than paper form. PDFs stored all in the same folder on the harddisk with the name in form for [First Author’s Surname] + [Year of publication] (ie Brown04.pdf). This meant that i had to put paper copies back in the right place (which could be fiddly) and also download their details from Web of Science or Pubmed when i acquired them.

This all seemed to work okay, but it was quite high maintenance, and didn’t give me any good way of organising things that aren’t journal papers. So, at this point in time, i have ten years worth of random bits of paper i thought were interesting at the time, with no idea what is where or how they should be organised (they’re all in different piles by when i acquired them, and i acquired them usually on the impulse that the information felt important, but i didn’t know why – or i knew why but didn’t have anything to do with it). What do i do with all this paper information? And is there a way i can integrate the way I organise it with my academic paper collection?

I’ve started a new academic project as part of my post-doc, and am facing the need to do something with all the new journal articles i’m acquiring, as well as the feeling that i really should organise all the papers I acquired while writing the book, and papers from four years of research in boxes somewhere – and all under the knowledge-shadow that everything will interconnect with everything else, so there’s no good way to keep paper separate by project (and who would want to lose out on the serendipity of accidental cross-links anyway?).

Has anyone got any good advice?