Literature Reviews with Obsidian

When writing essays or introduction sections for your lab reports and finally your dissertation (or indeed for a PhD), you’re going to have to read a lot of papers, take notes, and make sure you have access to those notes in a usable form when you come to actually write. It’s useful to make the distinction between two phases: the reading / research phase, and the writing phase. There’s an odd and very pervasive tendency when we are in the reading phase to think we will remember all the key points from what we've read as well as our very important thoughts about them perfectly well when it comes to the writing phase. You won’t. Don’t believe yourself! Make notes! What I want to present in this little article is a really great free system for taking notes which is not only smooth and rapid on the input end (while reading), which means you’ll do it, but most importantly makes it really easy to extract the most important stuff on the output end (when you come to writing). The tool I’m going to recommend is ‘Obsidian’, but the basic ideas in this article (which I developed over the course of my PhD) can be applied to other software (such as the more well known ‘Evernote’).

            Once you've got the program up and running, in the ‘File explorer’ section you can make folders, and notes within folders just like most note taking programs, either with the buttons at the top or right clicking (e.g. right click on a folder to make a new note within that folder). When it comes to making notes on papers, I recommend just using one note for each article you read. For example, here is a snippet of my Obsidian ‘vault’ (they call it a vault because it sounds super cool). I have a few high level folders of areas of research which are very distinct for me, and notes on papers on that topic within those.

As you can see I’ve done, when you make a note for a new paper, make sure you put the year of the paper at the start – that makes it easy to sort all your papers by their year, which is a natural way to order them as we usually write literature reviews as a narrative in chronological order (you can easily sort earliest first or oldest first). For the title, after the year, you could use the paper title, or the authors (or both) – all you want is something which gives you the best chance of remembering which paper this is.

In terms of where to keep your actual pdf’s of papers I personally love Mendeley. It’s free and the best thing about it is when you drag a pdf in, as long as it’s from an official source, Mendeley automatically populates the title, year, abstract etc (and you can right click a paper to get a formatted reference too!). It also has a great search function making it really easy to find what you want. There’s not much else to say about Mendeley. I think you can use it to take notes on papers, but don’t. It sucks for that.

Taking notes (& using screenshots to speed this up)

            Back to Obsidian. Inside your note, I always start with a screenshot of the abstract of the paper + the title, and author info. Now when I say screenshot, I don’t mean of the whole screen. You should be able to take screenshots of just part of the screen (for me on windows it is shift + win + s) and I really recommend this as the easiest and quickest way to get content from paper pdf’s into Obsidian. At the very least you want to do this for images (e.g. tables and graphs of results). If you want to copy and paste text from pdf’s as editable text it usually gets jumbled, meaning you have to spend a lot of time fixing it which is annoying [1]. Either way, for really old pdf’s the text usually can’t be copied, so you’ll have to resort to screenshotting anyway. I also really like screenshotting text because it is (1) super quick, and (2) makes a clear visual distinction in my Obsidian note between the content I’ve taken from the paper and my own personal notes (written in text) [2]. See below for example:

[1] However, one trick for this if you’re on windows is to actually open the pdf in the Microsoft Edge browser – this must have some clever programming built in as usually text copied from a pdf opened in Edge comes out normal-looking.

[2] There's lots of options I won't go into for highlighting text in different ways to distinguish your notes from quotes from the paper, so there's other ways to go about this - my preference these days is for more speed and less work.

You can see at the top here I’ve made a little note to myself about this paper above the abstract. This won’t mean anything to you, but it makes sense to me and that’s all that matters 😊

Now, Obsidian notes have ‘preview’ and ‘edit’ modes. When you actually paste images into your Obsidian note in edit mode, it will actually look like this:

This is ‘edit’ mode. Click the little button in the top right, to the left of the ‘x’ to move between ‘edit’ and ‘preview’ mode. Now, you might find this a bit annoying, however, there’s a good reason for this. When you set up Obsidian you choose a place for all your notes to go (either locally on your computer or in a cloud folder like google drive, dropbox etc). They will be saved there in the most universal format in all of computing: as text files with ‘markdown’ formatting, and images. Barring some very unlikely change in computing culture, these are just about the most ‘future proof’ formats you can get, and so even if ‘Obsidian’ closes down or becomes obsolete, there are a gajillion other programs that use the same format. The problem with other purely cloud-based tools like ‘Evernote’ is that all your stuff is all saved on their servers in their special format and if they close down you are a bit screwed [2].

This is one of the reasons I recommend Obsidian specifically for literature reviews. This is the kind of knowledge you may want to have access to for your whole career (even more so for the literature reviews you will do on your PhD if you do one). I don’t think Obsidian is quite as good for lecture notes as it is for paper notes (for lecture notes I prefer ‘Outliner’ style programs that allow you to make nested bullet points and have more formatting options, like Dynalist or Workflowy), but if you have lecture notes you think you want to keep forever, it might be worth thinking about keeping them in Obsidian too.

When you start adding more stuff from the paper you want to be able to organise, and we can use ‘headings’ for this. Type # then a space then type your heading. For example you might want to separate content from the Introduction, Method, Results and Discussion sections of the paper you’re reading. For example, below you can see that in the ‘preview’ mode these become proper headings – if you go into settings and find the right option, you can even get the option to collapse them to navigate your note more easily. A single # makes the highest-level heading – if you want subheadings you can use ## and even ###. Just make sure for headings you put a space after the hashtag, otherwise you will make a tag, which we’re going to talk about next. Below you can see this in both edit mode and then preview mode.

[2] I wouldn’t bank on Evernote staying around forever, they are notoriously badly managed. They also have unreliable customer service in my experience so I wouldn’t trust them to provide a good ‘transfer’ service of your data if they did shut down. If you can’t tell, I used to use Evernote, but switched, for exactly this reason. I was slowly building up a bigger and bigger library of notes on important papers to me, and the prospect of losing them all became too much.

Tagging your notes up (& making connections)

Now, the next step in achieving the glory of your Obsidian vault is to start adding ‘tags’. For example, see the blue text below. These are some of the broad topics I want to keep track of within my research on human reasoning.

When you are writing a single dissertation, you’re going to be more focused on one topic than I am in my career now, but as you continue to read, important themes will start to develop (different threads in the topic you’re studying), and you can capture these by tagging. Obsidian makes this super easy so that it’s not a big effort. Just type # and a dialogue box will pop up – if you want to assign a previously used tag, start typing its name and it will appear in the drop-down box. If you want to make a new one just type it and press ‘Enter’.  Later we can use these tags to create an ‘overview’ note which will be a really important step from reading to writing.

Other than tagging, you can also make links in Obsidian between different notes. You do this by typing [[ and a similar drop-down menu will pop up, showing all your other notes. Making links between papers as such connections occur to you will also be really helpful to you down the line. For example I might be reading this paper and think oh this is similar to that other one I read! To make it as easy as possible to find whatever paper I want to link to, I tend to include year, title and authors in the titles of my notes – sometimes I remember a paper by any one or a combination of these – you want to minimise the amount of time it takes you to make these connections – if it’s cumbersome, you won’t bother.

Making a ‘summary’ / ‘review’ note (the secret sauce)

            Now, once you have completed your literature review (read all the relevant papers you know about and made your notes on each one) it’s time to start trying to synthesise. This means you need to move from individual notes on each paper to having a summary of all the papers you want in your write up in a single note, ideally in the same order you’re going to write about them. Essentially, we’re heading towards a ‘plan’ for your write up. If none of your current tags cover everything you want in your literature review, go through and make a new tag specifically for that write up e.g. ‘#Dissertation’. Be generous at this point i.e. include everything that might possibly make the cut – we will whittle them down later. Then, click on that tag in one of your notes (ctrl + click if you’re in edit mode) and Obsidian will automatically run a search for you. In the search menu, first use the ‘Change sort order’ button to go ‘A->Z’ to make sure the oldest papers are at the top (the natural order for a lit review: oldest first). Then click on the ‘Copy search results’ button, to the right of the ‘Change sort order’ button.

In the box that pops up, keep the default settings and press ‘Copy’ then ‘Done’. Now, make a new note (it's probably good to make a separate folder for these 'summary' notes) and paste your search results in. It will look something like this (I added the space between each paper):

So what we have here is a list of all the papers relevant to your current write up with handy links to your notes on those papers in chronological order (the order you are probably going to lay them out in your write up). But we’re not going to stop there. This is going to become your summary note – your ‘plan’ of your write up. You can basically use this to start mentally organising and developing your ‘narrative’ like this:

You can see here that by just writing little summaries below each paper I am starting to build what will be the start of my write up. In the process of doing this you might decide that some papers don’t work and decide to get rid of them. If you’re not sure about what a particular paper is you can easily click on it (if you’re in edit mode, ctrl + click) to go check out your notes on it, then press back (either the button in the top left, or alt + left arrow ) – this is why it’s so nice to have the links there. Of course when you go into your notes on the paper, if you’ve made links to other papers you’ll see those, and might decide to bring them into your literature review, so all your thoughts and connections made during the reading phase won’t be lost as you start to build towards writing.

So that’s basically it. Obsidian can do a lot more than this, but you can explore that as you go. Once you’ve got a working structure here the next phase is your full write up in a word processor, turning that informal summary into a polished piece of writing. What we’ve been working on here is the middle step. That leap from reading to writing is often quite a big one, and we don’t talk about it that much. These summary notes make a really nice bridge, easing the passage.

Just as a final note, if you start to use this system for more and more essays and lab reports, and especially if you make lots of links and tags as you go along, you will start to see something wonderful happening - you will be reminded of relevant papers from e.g. a previous essay when writing a lab report, maybe even a year or two later. This will make your life a lot easier and improve the quality of your submissions by helping you bring in the holy 'outside reading'. Without a system like this you would have to personally remember that there was a relevant paper on some module you did two years ago to the essay you’re writing now. Good luck with that! If you continue to maintain the same vault throughout your whole career (if you go into research) it can become an invaluable personal ‘external brain’, reliably holding all the connections between various different topics and areas of research that can be hard to hold in your mind.