An Exciting Note-Taking Adventure!

Redundant I realize – how could note-taking not be exciting?

Thus we embark on yet another adventure in the wild-and-woolly underside of historical research. This time we’re hunting in holes and corners for that wily beast, the Note! So pay attention, take good notes, and maybe you’ll learn something along the way.

As I’ve discovered recently, any note-taking system worth its salt needs to fulfill several functions: store information, summarize information, search information, sort information, and site information (just checking to see if you were paying attention). And, since even Ancient Greek philosophers complained about having too much to read, we humans have felt the need to condense that overflowing basket of information down into a more digestible volume – i.e. to take notes. No surprise, then, that a whole variety of systems have been developed over the millennia to meet this need; Ann Blair’s Too Much To Know is a recent fascinating account of the history of note-taking up through the early modern period. A solution popular with 17C academics, and a supplement for the older commonplace book, was to take notes on little slips of paper – a small piece of paper was just large enough to record one piece of information on it, and it could easily be stored (on hooks, in drawers…) and quickly shuffled into a new order, even sorted several ways (first by a keyword, then by date…). Wikipedia credits our favorite Swedish botanist Linnaeus with the invention of the modern variation of these slips of paper, the notecard. By 1900 if not before, scholars everywhere were reading about the utility of notecards, particularly when you add keywords to the margins for even faster sorting. Turn-of-the-century contemporaries were also seeing one useful implementation of them in library card catalogs. So popular was this notecard idea that a few companies even introduced technical advances, befitting a mechanical age, such as an ingenious 1896 edge-notched version of a notecard that allowed one to quickly separate the cards with topic X from all the rest. The cards had prepunched holes along their edges, each hole representing a particular category. You (or cheap labor you hire) notched out those holes for the categories relevant to each card. Then, with a long needle you skewer through the hole of the topic you want to select. Since the cards you are looking for don’t have the ‘top’ of the hole for that category, they fall to the table while the rest dangle from the needle:

An improved 1961 version of the edge notched note card system.
This post has an interesting discussion of edge-notched notecard technology.

In short, the practice of notecards, and before that the idea of using little strips of paper organized on hooks and in little drawers, has been around for centuries.

Yet I don’t recall ever being taught anything about notecards when I was a kid. Admittedly, my earliest note-taking memories aren’t that memorable – probably because I didn’t take good notes. My recollections only date back to high school debate, when we had to carry around a briefcase full of notecards during our debate tournaments. Onto the front of the each 3″x5″ notecard we pasted a cut-out quote from a photocopied book, or else handwritten notes. I don’t even recall what we included beyond those quotes: I don’t remember including any summaries, but I’m sure we had to write a keyword or two somewhere on the margins. I do remember that those who had their act together would develop sheets of prepared arguments that they could just read off of, instead of rifling around in a notecard file for just the right card (time limits being a key part of a debate round). To be a master debater required a long notecard file; I can remember the deflated feeling seeing your opponent wheel into the debate arena with a handcart stacked with several boxes full of notecards. But I have no recollection of any broader or theoretical discussion of the notecard method, or of applying it to any other endeavor. The only other memory of notes in high school was when my mother, a nursing faculty and the proud recipient of EndNote (the year was 1988), asked me to teach her the software so she could keep a bibliography. Since I had no idea of the point of a bibliography at that stage, I wasn’t much help. Come to think of it, high school wasn’t much help at all when it came to note-taking.

Thus the utility of the notecard system managed to pass me by, if I had actually been taught it at all. I abandoned debate in college, and instead muddled my way through school without any particular note-taking system worthy of the name. I don’t recall much about my study process, other than the fact that I copied big chunks of several books on the library photocopies so I could have my own copy. Not much to learn there either.

Grad school – there you need to take good notes! But there too, nobody that I was around, including the profs, talked much about their note-taking system, just that you were supposed to have one that “worked for you.” Some friends used notecards, others used other systems. You could (and I did) take notes in the margins of books, but such annotations are stuck within the covers, not doing much good. With notecards still on the periphery of my consciousness, I did what many of my current students do as a default: read through Source A from the beginning, taking notes on paper as you go. In other words:

Grad School notes, circa 1994ish

OK as far as it goes, as long as you want to analyze the notes by the chronological order in which the letters were written – and have a filing cabinet full of notes. If not, you need to do a lot of work to flip back and forth between many pages (1000s of letters, folks), which of course is undoubtedly what prompted people to cut them up into slips and paste them onto cards in the first place. Dissatisfied with my early note-taking in grad school, I did make one significant improvement: I transitioned from paper to computer, sort of. This was the go-go 90s, and now almost everybody could afford a personal computer – even though the Internet still consisted primarily of Gopher and various BBSs. So, I figured, I’ll be “efficient” and take notes on the computer in MS Word! I’m so smart. So I did, and the results tended to look like this:

An example of Word notes, circa 1995ish

Pretty spiffy, huh? Each document with its citation info, quoting and even paraphrasing the letters rather than just summarizing or transcribing them (summary = short overview of whole, paraphrase = author’s content summarized point-by-point, transcript = verbatim copy). Look how I even created boxes around the citation info! Very useful no doubt. But more often than not, I would get lost in the weeds with so many sources. Even worse, a hybrid system naturally evolved from the primordial ooze that was my mind: I just couldn’t keep myself in the digital world, scribbling additional notes by hand in open white space. Now that’s what I call note-taking.

This was the ramshackle system I used during my research stint in Europe. Months of research in the archives only added to the mess. The result when I returned? Oodles of Word documents (one document per source, and then I aggregated – read copy-and-pasted – as much as I could in separate topical Word documents), a bunch of old paper notes (some of which I still haven’t digitized), as well as some new paper notes from the archives (mostly for tables and other formats difficult to replicate in Word), as well as both paper and Excel tables with siege information (more on that in the next post). As I tried to put them all together, I realized that I couldn’t deal with what I had – too much info in too many formats organized too many different ways. But I still didn’t consider notecards seriously. Once again, I convinced myself, technology must be the answer! So I started reading up on how I could manipulate all these Word documents – it was impossible to sort entries like the above (paragraphs) by more than one keyword, and then only if you duplicated the document and changed the leading word in each paragraph for each keyword. One document for the original source sorted by page number or letter date; another Word doc sorted by level of war; another sorted by side of author… Conveniently for me – one might even say prerequisitely – my wife was at this time going back to school for a second bachelor’s degree, this time in computer programming (lite). So I dragooned her into frequent trips to Borders bookstore to look through their MS Word VBA manuals in an attempt to make Word do all sorts of complicated contortions that, I eventually realized, were impossible to ask of it. But surely technology is the answer!

My failure at VBA programming initiated what eventually became a multi-year excursus into the nature of relational databases. My ever-suffering wife probably regrets that she recommended I look at databases, which sounded like just the ticket for what I wanted – she was probably thinking that it would be easier for her than learning Word VBA! Luckily for me, she was right. Even more fortunately, she was also there to disabuse me of my initial, ignorant notions of what good database design should look like, introducing me to crazy concepts like normalization (do you take one form or two?) and entity-relationship diagrams. I recall a time when we would actually amuse ourselves by describing how some ad on TV would be structured if it were a database. Fun times.

But they weren’t all good times. Through a lot of sweat and trial-and-error (this was before you could simply look online and find all sorts of info on databases and Access) I managed to develop a tricked-out note-taking database to surpass all others. To fill it in I scanned and OCRed (ABBYY FineReader) thousands of pages of published sources. Like a proud father, I even presented my database at the 2000 Western Society for French History – silly me, proselytizing ER diagrams to the historical masses. I won’t belabor the result too much, other than to link to my website where I described it in gory detail c. 2000, and to provide a snapshot of its current permutation:

Notes database, 2012 version

It’s not quite Vannevar Bush’s Memex, but it was light years ahead of what my colleagues appeared to be using back in 2000. And it still is, if my recent recce in the British Library is any indication. In retrospect, with a greater appreciation of note-taking under my belt, its main advantages are four-fold:

1. I really focused on speeding up data entry – with lots of combo boxes, automatically-filled fields, as well as key shortcuts to quickly jump to commonly-used fields. Of course entering the transcript, notes, and keywords will continue to require the ol’ noggin’. Computers are still dumber than a six-year-old.
2. More importantly than data entry speed, I created fields for all three types of ‘notes’: plenty of room for the transcript (which probably shouldn’t be considered a ‘note’), a field for a summary, and then another field for paraphrase. Plus a separate field for my comments and thoughts. All segregated so you never need to worry about confusing a quote for your summary.
3. But my two biggest ‘breakthroughs’ relate to keywords, which I use extensively. In particular I dealt with the thorny issue of how to tie specific keywords to shorter portions of text. One letter could discuss five different topics after all, so I made a one-to-many relationship between the top bibliographic info on a source (the ‘one’) and the bottom subform/table with the transcript (the ‘many’). So I can split a single document into as many parts as I want, tagging each chunk of text with the appropriate keywords, resulting in many possible keywords for a single letter. You want to avoid reading having to read through a two-page letter and waste time trying to figure out (perhaps for the fourth time) which section deals with which of your keywords.
4. The keywords themselves are actually one of my proudest achievements: after using EndNote for awhile in early grad school, it quickly became obvious that one keyword field was not enough, at least for me. I like to consider lots of categories, and since I have more than a dozen different types of keywords, you can’t expect to remember to tag each record with each of those in just a single field, much less remember what all the options are for each keyword. My Access database, on the other hand, however, allows for that: a dedicated keyword field for the level of war referred to, another for the theater, for the side discussed, for the specific combat under discussion… And, just as important, is my implementation of a Library of Congress-style Topic keyword – detailed enough to identify nuanced topics. Identifying that a letter is about the author’s “View of the French” is great, but keywording a letter “View of the French-Don’t like” is even better because you can easily compare it with those who do like the French, or those who are ambivalent. It goes without saying that the real power of a relational database comes from the fact that you can query any field, or any combination of fields, so the more info you put into those short keyword fields, the more analytical power you have available – you don’t have to reread the letter to remember the answer to the keyword question of what their view of the French actually was, the compound keyword tells you. I was only copying the format of the Library of Congress, but it seems to be a rare interpretation of the use of keywords. In fact, with the Web we seem to have digressed back to dumbed-down tags that eliminate all sorts of possibilities for analysis.

Store, summarize, sort, search and cite. All organized so as to minimize having to re-read the original every time it appears in your searches. You need summaries for speedy perusal. You need paraphrase for comprehension. You need transcription for full-text searching and going back to the original as new questions arise. You need detailed keywords to quickly and precisely sort and search. You need multiple keywords to ask interesting questions. Whatever note-taking system you use, it should be able to do all these things.

So as you can probably tell, I hate the “just choose a system that works for you” line. Within certain parameters it is reasonable, but it is certainly not the advice you should give to a newbie like a grad student – give them some structure at the start and then they modify it over time. More understandable for those already in the middle of a big project, but why was it considered normal to start a project without a good note-taking system even five years ago? Sure, you can obviously get by and even prosper without such a system, but your work could be so much better if you set up the proper system, or, if you prefer, you could spend less time on your research and more on something else. Most note-taking systems I’ve seen don’t meet the 5 S criteria, and serious, sustained research should demand doing all 5 criteria as quickly and efficiently as possible. Only particular types of note-taking systems meet these important functions: notecards are the best paper option, but digital notes alone (like in Word documents) can’t match the feature set of a well-designed notecard system. Dedicated databases are clearly the digital winner – you don’t necessarily need MS Access or File Maker Pro or Scribe, or any relational database to do all these things. But you should be able to do them easily and quickly. To get a bit heretical here, if you truly believe in the “one (small) thought-one note” rule of notecards, Excel is actually a surprisingly-good off-the-shelf choice. It’s main disadvantage is that it’s not set up for textual display (and historians are generally numerophobic and thus unfamiliar with the program), but if your thought or note is short, it can easily fit in a wide column. The key benefit is that you can sort by multiple columns, in whatever order you want, and resort just as quickly. Unorthodox I realize, but I don’t think we stress enough to our students how important short keyword categories are for how we analyze (and organize) history.

The way I present it in my methods course:

Different types of notes (keyword, summary, paraphrase, transcript) are needed for different functions, so you should ideally have them all. If one doesn’t have the time to do all of these (and I did take longer to graduate than many History Ph.D.s, for a variety of reasons), you obviously have to prioritize. Detailed note-taking usually seems to get shorted. It’s quite understandable that an undergrad can do well without using a robust note-taking system. On the other hand, it’s an embarrassment that history grad students don’t have dedicated note-taking software given to them on their first day – a small amount of money could develop a whole series of modular databases that would make all History Ph.D.s far more productive than they are, and far more rigorous to boot. (George Mason’s Center for History and New Media developed a simplified version of a notecard database called Scribe. It’s still available, but of late they’ve been obsessed with Zotero and Omeka, which don’t do the same thing but are more sexy and certainly more fund-worthy. So Scribe has fallen by the wayside, which is too bad. As I’m finally starting to appreciate, downloading bib info doesn’t actually save that much time compared to searching and researching and keywording and analyzing your notes.)

Regardless of whether we have an awesome or a slipshod note-taking system, we shouldn’t hide behind the fiction that every note-taking system is as good as any other, because it fits our personal style, as if our research method is unchangeable, or as if we have an inherent right to not have our note-taking system offended. At the least, we should actively search out tweaks to make our system a little bit better whenever possible. Yet I have met surprisingly few historians who seem willing, much less eager, to discuss how we research, and I find that bizarre. The honest ones may admit that they don’t have much of a system, which is at least a start. But we should be doing more.

We certainly need even more discussion now that the allure of the digital deceives many into thinking that anything on the computer is inherently better than the same content on notecards. Megabytes of plain text are great for looking up specific text strings, doing word frequencies and the like, but if you don’t go through and keyword all those early modern primary sources, you still won’t be able to find all the references to logistics unless you do a thousand separate searches for all of the words that could denote supply. And how you keyword can make a big difference. But maybe note-taking is rarely discussed because we historians tend to be more anecdotal, not feeling the need to fully exhaust a source base before drawing our conclusions? The historians that discuss their databases are usually the ones who use relational databases for specific computational tasks – recording and (statistically) analyzing structured information in wills, censuses, and so on. Or maybe we historians just don’t care enough about our subject to commit the time and effort?

I recognize that note-taking is one of those things that people, particularly scholars, can get quite defensive about – hence the “use whatever system works for you.” A phrase uttered, we should note, without bothering to specify what exactly a note-taking system needs to do (how it “works”), and what benchmark we should use to compare systems. But it’s shocking how rarely we talk about something so fundamental to our craft. Back in the early modern period some scholars recommended that you keep your research methods a secret so as to heighten the illusion of your vast wisdom, as well as your value as a tutor or advisor. But that shouldn’t be the case today – surely this is one area where we can learn a lot from each other, and not worry about our content ideas getting scooped. Secrecy wasn’t even the rule back in the 1600s, since just as many scholars sold books to college students eager to discover effective techniques for learning a subject. We still have the equivalent today, such as The Craft of History, Doing History and the like. Yet even Barzun’s The Modern Researcher spends surprisingly little time discussing why notecards are such a good system – maybe because he grew up in an era where everybody understood how useful notecards were, being used in many institutions from libraries to hospitals to mail-order catalog companies. I assign my methods students to read the Chicago Manual’s discussion of notecards, but it’s only when I literally show them what you can do with notecards and can’t do with notepaper (or a Word doc) that they finally get it – you mean you can lay out your cards on the table and look at them all, organize them, then gather them up in their groupings and plug them into your paper one after the other? To be fair to them, and more evidence of our general ignorance of note-taking, I didn’t even appreciate all those nuances until post-grad school. Which means we need to be discussing this more openly.

There is, however, some hope. Blogs are much more willing to discuss such matters, even if sporadically. Notably they are written (like most blogs) by historians almost a generation younger than me, usually graduate students and postdocs. Often in the formative stages of their professional development, they are naturally dealing more immediately with how to organize a large project for the first time; they are also challenged by the massive increase in historical information now available online and via more lenient photographing policies of many libraries and archives. Born “digital natives,” they are far more at home with Web 2.0 and its ilk, they focus much more on organizing photographs, and they seem to use the omnipresent tag metaphor, often in rather simplistic ways. (And Macs seem to rule the roost FWIW.) The most recent Perspectives (October 2012) has a forum with several articles on digital technologies and history, including one on DevonThink, a well-spoken of Mac note-taker. The author’s blog posts on her use of DevonThink as a postdoc are the kind of thing we should be talking about more broadly. Unfortunately such discussions can be short circuited (or at least ghettoized) when most historians (I would guess) still use PCs, while many of the bloggers discuss Mac-only programs. Even more reason for a broader discussion of the general principles of what should be included in an efficient note-taking system.

All that being said, my Access database system has its disadvantages, which is why I’m currently in the process of switching to an even more powerful program, a personal wiki called ConnectedText. As with my relational database adventure, this too requires me to create a new text database from scratch, but it’s turning out to be surprisingly simple thus far, having gone through all the hard conceptual work and data cleaning switching from Word to Access.

But I’ll save the details and pause here, since I don’t want you to get too excited. But don’t worry, there’s more to come! Part II will talk about the simultaneous battle I waged against information overload: siege data! Part III – yes there will be at least three posts on this topic of note-taking – will discuss my conversion to ConnectedText, a slightly more texty type of program that takes advantage of links in surprising (to me at least) ways.

You can’t hardly wait, can you? But wait you must.

While you’re waiting, tell us what type of note-taking system you use. What are its advantages and disadvantages?