So you think I need to design an experiment to just say there’s no existing experiment? Okay.
This sounds actually more like how I’m using Obsidian. I was, honestly, put off by the whole Zettelkasten idea because it doesn’t seem like a proper fit for what I’m intending to do with notes. (Please don’t argue with me about that, Zettelkasten proponents. 
)
I took some inspiration on what MacSparky explained his general system looks like, which has evolved over a long and diversified working life (and his private life, too, obviously).
I use Obsidian to capture meeting notes, to capture quick how-tos when I solve a specific problem (stuff like “How do you create the proper launch deamons for wiki.js to autostart?”, “Which partly non-obvious traps do you have to avoid when shutting down our company’s servers?”), capture thoughts and logs regarding ongoing projects, and all these kinds of things.
I tried doing this in DevonThink, but the friction of using DevonThink to just quickly capture stuff is higher and while I like the clean interface of DT 3, Obsidian with a proper custom theme and some custom templates is super fast and somehow more enjoyable.
I have templates for meeting notes, telephone notes, daily notes, weekly reviews, general thoughts…
Whereas before I was inconsistent in taking a note for any external, potentially somehow relevant phone call, I now naturally open a new note before calling someone/when I’m called and at my desk.
Whereas I couldn’t be bothered before to do daily notes because it only got in my way, I now put in some random thoughts each day about what I did, what could be improved etc.
The power of Obsidian, to me personally, really comes down to being a very efficient note taking app that doesn’t look me in and is relatively enjoyable to use. It made me rethink my (simplified) structure and it’s future proof, because we’re only talking about a folder of text files and some very few pictures as attachments. If Obsidian goes away, I can just as well continue with DevonThink or any other app, even though I would have to recreate some of the workflows, especially my template usage.
That being said, Electron still sucks and the more I use the app, the more I want to rely on complete keyboard navigation with does not work as well as with native editors. Compare that to a native code editor like Nova from Panic and try not to cry…
But since I’m not being locked into the system, I don’t worry too much about that. Either Obsidian gets some further polish and optimizations for macOS, or I’ll see where Craft is standing in a year or so. It seems like Obsidian has really put some pressure on other note taking apps to become more open and compatible…
6 Likes
I don’t believe any of us were attempting to start an argument. I for one appreciate your emphasis on going beyond mere subjective experience and intuition in assessing the value of applications and systems. That is good and necessary.
That said, and with due respect, nothing I read or “argued suggested that science and logic have no place in productivity.” Behavioral and other branches of the sciences have studied productivity extensively.
What I said is that we must be careful to avoid scientific reductionism as science cannot measure everything and should not make pronouncements outside its epistemological sphere. As I stated, there are aspects of productivity and systems that can and should be measured by science. There are others that science cannot measure or measure in any meaningful way. Science can measure what is and can predict what likely will be, but can say nothing about “what ought to be” and little about the quality of anything that is not physical.
Social scientists may assert the impact of xyz policy outcome on the poor, e.g, x number of people are less poor due to policy z but that rest on the non-scientific presupposition that outcome x, fewer people are poor, is a good thing and therefore is a preferred outcome. That assessment is not scientific, its ethical and political.
Science cannot measure the quality of emergent ideas as ideas are not physical and therefore must be judged by non-scientific standards–which in turn are derived from a host of epistemological, teleological, ethical and other presuppositions.
Would it be possible to measure systems by counting the number of new connections or generated ideas? Of course. But counting numbers and measuring correlations say nothing of the quality of those ideas and it would be devilishly difficult to ascertain with any statistical certainty whether the idea emerged as a result of the system, was already subconsciously present, or emerged as a result of being measured or assessed (think quantum physics)
As to your statement that “Intuitively, it might work for research in social sciences, where arguments are atomic and abundant so a linked system helps to find new angles of attacks. On the other hand, I just don’t see how it’s going to be relevant for STEM, where ideas are highly regulated by logic. [emphasis added] If I can’t link to subjects logically in my head, no note linking is going to magically create new “emergent” connections,” I doubt you intended this but your post seems to imply that logic is the sole province of the hard sciences. I’m sure that is not what you intended to imply because if so, then this entire thread falls outside the sphere where logic resides. After reading my comments you might conclude that that is indeed true! 
Regarding Zettle, my take is that you are correct in that any such system is a worthy subject for studying its impact on productivity but such a study will necessarily be limited in scope and value because the quality, ultimate origin, and the value of any connections and emergent ideas cannot be adequately measured by science. Science, while extremely valuable, does not encompass all truth. In fact, assessing most of human experience falls outside the domain of science, which is why we have all the other branches of knowledge.
They inform each other but they are not each other.
Which brings me to the conclusion that each person must assess the impact of any PKM system for him or herself based on his or her assessment of what is good, beautiful, and true in the system for his or her work and person. A scientific study cannot answer that question, it can only measure or count. As I said previously, the mechanics of a system can be tested for certain outcomes, e.g., retrieval speed, capacity, etc., it cannot measure the effectiveness at generating creative, good, ethical, imaginative, original or other ideas as those are not subject to the scientific method.
Thus far, I am enjoying Obsidian and I have in fact noticed connections through the graph that I have not seen before. Could another system have produced the same result? Probably but what I know is that using Obsidian did.
I will not take the time to get into it here but I found a connection between Steve Jobs’s statement that “we should do what we love” compared to “we should love what we do.” I believe the second statement to be truer because most people don’t have the luxury of doing only what they love, in part because we have ethical responsibilities well beyond ourselves. This is a fascinating distinction that I had not seen before I started using the Zettle method in Obsidian. I doubt that science can measure and determine if this was Zettle, serendipity, or could have just as easily emerged from another system.
Anyway, we are all grateful for science. But science is limited and must be limited to its sphere of competence.
Thanks for starting a fascinating conversation!
7 Likes
I’ll give you one in an academic situation. Sheep produce wool in varying colors. In my breed, the Black Welsh Mountain, we have had in the North American population pop up some variants, a true breeding brown or moorit color we call chocolate and a white. The genetics of sheep color also vary among breeds. I’ve been researching sheep color papers and finind out quite a bit of relatively recent research done on the genetics and variants of common wool colors done by using DNA analysis.
I am also very interested in textiles. I spin my own yarn, weave, knit and do naalbinding. I also research archeological evidence of old textiles because of my interest, Whether it’s old weave patterns or how to replicate yarn used in existing garments found in archeological digs I find that fascinating.
As I’ve been slowly entering my population of scientific papers into Zotero. Annotating them and then saving the highlights and my own notes or extracting my existing annotations and putting them into Obsidian I have been linking terms and things I found interesting in each field. Many of the papers already had notes in DEVONThink or in files on my hard drive.
Obsidian pointed me to some mentions of moorit from an old 1959 scientific paper on medieval wool types as I was linking my notes from a paper written in 2018. Following up on that led me to several more (some the subject of my rant on the cost of papers because all I’ve seen so far are the abstracts) where gas chromatography was used to determine breed of sheep in textiles recovered from archeological digs. Well not breed exactly, breed is a modern construct, but at least the types of sheep.
Now I might have made the connection without Obsidian but who knows when. I’m now using that data to make some changes to some research that I’m proposing we do in our breed of sheep this coming year.
I have had those papers and notes for at least 10 years. I’ve made many notes on them and others. I am making new connections all the time now. It’s almost an exponential growth as I get more and more of my historical notes into my Obsidian vault. For me at least, the zettl concepts and the system that Obsidian uses to implement them has resulted in many new emergent ideas.
7 Likes
So what sort of evidence do you want.? Can you show any similar uses of using statistical data to determine the effectiveness of software?
1 Like
Can you give an example of an idea emerging from the network of notes, which wouldn’t have been possible without Zettl? Are they often in world building or plots?
Absolutely 
The question, though, is flawed. You’re looking for things that are enabled by Zetteling; nothing is enabled by that in itself, in the same way that you can still write a novel by a clay tablet and a stylus, but using a computer is a matter of clarity, speed and comfort, and makes the process magnitudes better.
This being put out of the way: plot lines and characters arcs, especially in complex stories, feed on each other and influence each other all the time. The simplest way I can put it is that Zetteling allows you to develop those things very simply in parallel by maintaining a fluid structure and seeing the order of scenes, the themes, emerge freely as things get into place. You have an idea for a cool scene or even a punchline: you capture it in the system without wondering where it goes, as would be with a more traditional system. You link it to character A, with a note that maybe it could go to character B. Weeks later, as you are maybe struggling on some plot point, you can easily rediscover this snippet that was relevant but did not yet fit anywhere. Now you know. Or maybe it will never get used; that’s fine either. But the core thing, at least for me, is that the order emerges. The nature of all creative work is that, by definition, there is no preestablished order. You discover it as you go. The order is done when the work is done. Zetteling embraces this as a process.
5 Likes
I’d love to see some fo your templates as I build out my own system. I too am starting to develop templates but seeing others could bootstrap my system and make it better faster.
@Christoph I too would love to see your templates. I have created two but would gain inspiration from your examples.
That is a fantastic concrete example!
1 Like
There’s the black swan.
1 Like
Love a Popper reference, or maybe you mean that Natalie Portman film
2 Likes
I agree that it’s very hard to measure this precisely in science. I didn’t think of some absolute conclusions that can be drawn from such studies. What I had in mind was some some study that at least aims to be objective and scientific. So far all recommendations are based solely on subjective experience, which is valuable but not sufficient for adopting systems of this complexity. I could try a calendar app or task manager based entirely on someone’s words. I’d be reluctant to convert all my notes into a complete new system if you know what I mean.
I’m definitely not implying logic is exclusive for hard sciences. Instead, hard sciences are much more rigid in the sense that it has to be logical. I’m struggling to think of any scenario where a long link of related notes can brew new logic conclusions. Because usually the link has to be put in after the concepts of the notes are already understood, at which point the concepts are already linked in the mind.
Thanks for the example!
I agree that my original wording was flawed. I should be asking what is made easier by zettl, instead of what is only possible with zettl.
1 Like
Well said. Just a couple of additional short thoughts.
- If we wait for every app to be scientifically assessed before trying and adopting we’ll be waiting a while!
- As to “ Because usually the link has to be put in after the concepts of the notes are already understood, at which point the concepts are already linked in the mind”, that is true insofar as creating the manual links via [[ ]] is concerned. The “magic”, if you will, happens when links are created and then one notices a conceptual link that was not manually created. In my simple example above of the Steve Jobs quote, I had not created nor noticed a conceptual link between two completely different articles/books related to work and calling until I’d put the information in Obsidian. When I brought up the graph I noticed two ideas that were not yet linked and immediately saw the “opposing” ideas of “do what you love” versus “love what you do.” I had not seen nor linked this prior to the articles and quotes being put in Obsidian. The base information and arising concepts are related to a chapter in a book I’m writing. The chapter has to do with our universal and particular calls and their intersection. Should I have noticed the connection? Perhaps but I didn’t until I used the graph feature. In other words, the Obsidian tool facilitated my thinking. I think that is all we are suggesting. I don’t think I’ll wait for all the apps to be scientifically verified as effective and efficient. I think I’ll just continue using Obsidian.
1 Like
My request was that you should provide us with an example of a case where you have used statistical evidence or logical analysis to make a hard decision about a superior method of approach to do something equivalent. For example, show us why you made a firm decision to use Things instead of OmniFocus because statistical evidence or logical analysis proved conclusively to you that Things was superior to OmniFocus.
When we have this example from you, we can understand what effort we should provide back to you.
In the meantime, we have …
Oh. No! You are not allowed to do this! You should get the statistical evidence or logical analysis first.
Serendipity. Simply serendipity.
I would wish that all of my successful studies in science or engineering could be predicted in advance.
What is made easier is the ability to review a collection of notes to discover connections that you did not recognize existed at the time that you took the note in isolation.
No. I don’t know what you mean. If you are willing to try Things instead of OmniFocus (or vice-versa or pick Fantastical or Notion or …) simply based on someone’s word, then you should be willing to try the ZK method simply based on our word as well.
Put a different way, no one has asked you to convert all of your notes.
–
JJW
1 Like
Incidentally, I recently presented at a symposium on data and serendipity. The line of research is focused on how the ways in which we capture and represent data can limit its serendipitous potential. (It’s not directly relevant here, but the recorded talk is on YouTube if anyone’s interested.)
A thesis of my work is that the more we can granulate information, the more easily it may be (re)combined. (This is not proven out yet, but the theory is pretty reasonable!) This recombination facilitates serendipitous discovery, which supports the line of thinking about Zk you’ve put forward in this thread.
In other news, I’m also very interested in the science of personal productivity and knowledge management. I have many thoughts on the matter—too many to write up write now!—but I’ll point out that the field of design science is dedicated to the study of “technology artifacts,” which includes both Zettelkasten methods and the apps that support it. If folks are interested in these questions, I’d recommend looking into design science.
In general, I think it’d be bad (design) science to study Zettelkasten as a system outright. It makes more sense to break up what Zk as a system is: mechanisms for reading, taking atomic notes, linking notes together, and so on. Each of these individual components are (a) part of other systems and (b) more easily studied on their own than wrapped up in the confounds of one another.
Last, I would be surprised if, once you’ve broken down a system like Zk as I have above, there isn’t research out there examining it already. The fields of knowledge management, human-computer interaction, information systems, and so on are quite old—you’d be surprised at what’s already been done!
3 Likes
I loved this description of your workflow. But, how did Obsidian point descriptions of moorit wool ? Were you doing a search for moorit wool ? Could DevonThink find the same connections it it were indexing your Obsidian files ? What was the key workflow in Obsidian that provided this insight vs taking notes and links in DevonThink ? Or was it the act of linking notes that lead you to this insight ?
unlinked mentions
I am indexing my Obsidian vault in DT right now. I suppose I could have found the connection in DT but the thing is I’d have to look for it specifically by crafting a search. Obsidian showed me possible links before I even thought to search for it that way.
1 Like
My templates so far are really not too sophisticated and not too exhaustively developed.
But I’ll try to put together an “example vault” next weekend. Happy if it maybe provides some useful ideas!
5 Likes
Not exactly. Maybe lighten up?
1 Like