One of my biggest issues with my masters notes is knowing what is processed and what isn’t, and then when my permanent topic notes were last synthesised. I’ve tried various methods but ultimately, I’ve felt overwhelmed with the number of notes I have, and the lack of certainty of what I’ve reviewed.

It turns out the solution was rather accidental, but now I have system where the tool itself tells me the status of the knowledge: tools as status.

Here is a summary of how it works below:

Read more if that interests you :)

This time last year I was obsessed with knowledge acquisition, or rather connection acquisition. I wanted all the links for everything, and I didn’t care what that did to the usefulness of my notes. Logseq was amazing for this and just let me link and connect and dig down into these connections thanks to its block references and embeds, but I rarely reviewed the knowledge. To be clear, Logseq wasn’t preventing me reviewing knowledge, it just wasn’t my priority.

In September, I realised this wasn’t the best way of working. My needs had changed. I no longer needed to acquire everything, I needed to process and review and synthesise what I was reading.

The attempts I did make at synthesis in Logseq were fine, but I wanted to use the whole page, add in images, routinely add metadata etc. I wanted something more visual to help my chaotic brain. Capacities has turned out to be the tool I’ve been missing. I carried on using Logseq for my literature notes and initial summaries, which I’d then transfer into Capacities as referenced, summarised information.

This was me giving Capacities and Logseq relative tasks. I then realised I’ve always preferred to collect thoughts in Apple Notes and links in Raindrop even though Logseq has a mobile app that works perfectly well and could be used for Capture too.

It became obvious to me that I’d accidentally assigned the tools I use to a specific status of knowledge in my PKM system.

I thought about that for a bit and realised it was actually the perfect system for me. This split has come about because of how the tools work right now, as opposed to a structure I have forced, which is why I think it works so well, and why I enjoy it so much.

With this system, I know exactly how much processing a source or a thought has had dependant on where it is in my system.

Step 1: Capture

Words are captured in Apple Notes, links are captured in Raindrop. I sort captured content into its various homes about once a week, I do it slowly and whenever I want to, because I really enjoy the process. Whilst I work towards ‘inbox zero’ in these apps, unprocessed information doesn’t worry me as the system is so simple. There are only two places to look for anything I haven’t sorted: Apple Notes or Raindrop.

One small change I’ve made to bolster this system is putting random thoughts I’d like to action in Apple Notes even if the brainwave comes to me whilst writing in another app. I can just switch to Apple Notes and write the thought there. This means I know this thought will move along my information pipeline. I don’t need to develop or support a review system outside of what I’m discussing here, which is more efficient and easier to maintain.

Step 2: First Process

I regularly save academic articles to my PKM system. I find them in many places, save the link to Raindrop, and then save to Zotero so that they can automatically collect reference information for me. There are many reference managers, but Zotero has an integration with Logseq which made it the easy choice.

I’ve written about my workflow here, but essentially, I read and take preliminary notes in Logseq with its PDF viewer and work with the source until I am able to confidently summarise it, and my response to it.

My source pages look like this in the end:

With this system, I know that whatever information is on a Zotero page is summarised and referenced. This means I can confidently move that information onto its next home.

Step 3: Capacities

Once I have my summarised and referenced literature notes, I can start synthesising that information with other knowledge, working on its component parts e.g., extracting information about people, events, themes, definitions etc to build into my web of knowledge.

Capacities is built to do just that, and its amazing text editor, linking capabilities and data structure makes developing the knowledge really fun.

To start, I literally copy over the summary from Logseq to Capacities into a page in my media content type. This is because it’s a summary from a literature note, aka from a piece of media. This is an important distinction to make because Capacities isn’t like Logseq or Obsidian where you have a blank page, and you can do whatever you like with it. Every piece of content belongs to an ‘object’ which is a noun: an event, a topic, a definition etc. Those objects live in databases which you build and assign properties to. Each piece of content within those databases is then assigned those properties. It’s a far more structured approach to knowledge management, but one that works very well.

I can then create links to other pieces of content from other objects, so I can break down information from its unified object (the notes from a piece of media) into its component parts (topics, academics, countries, definitions) in order to bolster what information I already have on those matters, with the newly referenced and summarised information. This means my knowledge is being constantly developed.

Once the literature notes are in Capacities and linked to other entities, that literature source will be one of many backlinks to that entity e.g., ‘Structural Realism after the Cold War’ is linked to neorealism, and my neorealism topic (below) has several backlinks. I need to know when I last reviewed these backlinks in order to know up to date that entity (neorealism) is.

To do this, I just have a date property that I fill in once all backlinks have been reviewed and written up into the page I’m in. Dependant on what type of page it is (e.g., topic, event, person), there are different properties to fill in. Capacities’ data structure means the field is always there to fill out, so I don’t forget to note it down.

However, it’s important to note that I do not force myself into reviewing everything regularly, I am just doing so when it makes sense. However, I feel the structure is in place for me to confidently review and synthesise my knowledge whenever I need it.

The system overall

This system works for me for two reasons.

1. It lets me use the best bits of all the apps I like

I recently wrote a little bit about why I don’t like the one app solution mentality, like Notion says they offer. I can get higher functionality by using multiple apps, I just need to make sure I’m using them efficiently, which I believe I am here.

Thanks to its block-based outliner structure and amazing PDF integrations, Logseq is excellent at helping me dig down into a source, and then I can use Capacities to organise the component parts of my knowledge more effectively in a way that works for my brain.

2. I am less affected by shiny new tool syndrome

The Capacities mobile app has just launched for Believers and it’s great, works amazingly and it is good to have my notes on the go. But instead of trying to make it a capture tool, I am reassured that my process and system work, and therefore this app is just given the job of me accessing notes on the go.

Other tools interest me, but the urgency of making that tool my next new lifeline has gone, because I finally feel my note-taking needs are met with the system.

I really enjoy this system. I’m not sure if something so well defined would work with all knowledge work, but for my masters it’s great :)

What is your PKM stack? How do your different tools fit together into your system? Let me know!