How to physically organize academic papers

Question

I'm looking for an efficient way to organize academic papers on my bookshelf. I'm a CS PhD student and continue to amass a large number of papers that I periodically need to refer to. The PDFs of these are very easy to deal with, but the physical prints end up in small piles throughout my office (each project/article/paper gets a pile). This is getting out of hand and not very efficient in many ways.

I really like reading hardcopies instead of PDFs as there's only so long I can stare at a screen. This must be a common problem among PhD students, right?

My plan is to place papers into labeled manilla file folders, then horizontally stack these folders on my bookshelf. Is this the way to go? Has some academic found a clever way to handle this better?

Answer 1

As suggested in comments, use a filing cabinet, ideally with hanging file folders that require no punch holes. But don't mix physical location and semantic organisation. This would be highly inflexible.

Do this instead: Note down a running number or date on each hardcopy. This is the call number of your hardcopy.^* Now write down a range of numbers or dates on each folder, and file the hardcopies accordingly. For example, hardcopy number 143 goes into file 130-150.

To find and retrieve your hardcopies, use a catalogue, either on paper (unwieldy) or digital (better). The latter can be in the form of some database, BibTeX file or reference manager like Zotero. If necessary, just repurpose one of the lesser-used fields to store the call number (e.g. "note" or part of the "abstract" field).

In your catalogue, you can assign multiple tags and projects to each hardcopy, no matter where it's physically stored. This removes the need for duplicate hardcopies.

In case you want to temporarily reshelve some papers, say, for a project or because you take them to a conference, just assign a secondary "call-number" that describes the current location, like "widget conference binder" or "brief case". This isn't strictly necessary if, every few weeks, you put straying hardcopies back to where they belong, which is easy enough, since they now have an address written on them.

_{* You could also use more elaborate call-numbers based on some classification, but I prefer to keep things simple. A running number or date also maintains the only advantage of the pile "system": Quick retrieval according to "It must be here somewhere, I've only used it a week ago".}

Answer 2

I used to use a piling system - it works great until you have to move office, and the piles are all in new places. Also, putting things on top of each other makes it awkward to pull out those further down the heap. Filing cabinets were invented for a reason, and work really well ... again, until you have to move. Which may be quite often if you are still a postgrad.

It depends a bit on how many papers you have.

I'm not a great fan of arch files/binders as they are awkward to handle when reading and - especially ring binders - are very inefficient in terms of shelf space taken up compared to how much paper they have inside. But, for 20-30cm worth of A4 I'd stick to binders and stand them on the shelf.

For more than that, you could look at either "archive boxes"/"box files" (which can stand) or the traditional cardboard "document wallets". Both can be labelled on the side, so you can read the label even while they're in the pile.

Of course, it also helps if your reference management system can record which folder a particular article is in (and also write this on the paper itself, so you can easily put it back in the right place).

Answer 3

I've found it useful to use large ringed binders for papers. One binder per project/subject is usually sufficient, though you might sometimes have to use two. This works well if you can separate your papers into groups based on subject area or project, but you might need to double-up papers in some cases (i.e., have more than one printed copy of a paper). Within each binder I separate the papers using coloured tabbed separators, with papers included in chronological order (i.e., reading earliest to latest). I add a first page in the folder that lists the full citations of all the papers in chronological order. Reading papers in this order is nice to give you an overview of the development of the literature over time.

This type of organisation is nice if you are working on a project or trying to refresh your knowledge of the literature pertaining to an old project. You can easily grab a single binder and take it down to the cafe to read it while you have lunch/coffeee. I am like you --- I much prefer to read the printed copy than stare at a screen all day. There is a bit of initial time required to organise the binders and cover-sheets, or update new cover-sheets occasionally, but I find it is generally worthwhile.

In practice I actually use a hybrid system that consists partly of these binders and partly of the "piling system" you described. I have quite a few left-over papers that are in piles, which didn't fit into particular projects or which were printed out of curiosity about a subject, without attendant literature. Ideally I would take the time to be a bit more systematic, but business sometimes gets the best of us!

Answer 4

(I've since been reminded of another option which I don't think will work particularly well in this case, hence this separate answer, but ultimately it's for the OP to decide such, especially if they don't have as many paper papers as us oldies.)

In the past I've had good results from making up mini-compendia of related articles using semi-permanent binding techniques such as wire or comb binding: a wire-bound A4 booklet of papers is much more pleasant to browse and read than stuff in a ring binder with heavy rigid covers flapping about. The down side is that it's more effort to put together, and of course you need access to the specialised punch/binding press.

While I generally feel that wire-binding makes for a nicer volume, in the present context I'd note that combs have spines that can be labelled, and also with some care it's possible use the machine to partially unbind a comb-bound volume so as to add newer papers.