I don't know why so many people make the point that hierarchical structures are how the brain works. Graphical structures are how the brain works. The hierarchical structure is how the physical world works (you can't put the same shirt in two closets), and we have an unfortunate habit of trying to force our brain to work the same way, but it's an impedance mismatch.
Unless I'm the only one that is frustrated that you never know where to find vegetable oil in the grocery store. Is it in the cooking aisle near the flour, or the condiments aisle near the vinegar? Seems like every store chooses differently since it doesn't make sense for them to put it in both places.
I agree with you that emacs org-mode is definitely not very graphical !
Regarding non-hierarchical information organization however, keep in mind that this article was only a very quick intro to the most basic possibilities of org-mode.
Org-mode actually provides a pretty powerful set of non-hierarchical structuring functionalities by using tags.
Within org-mode you can add arbitrary tags to your items, you can define custom sets of tags for auto-completion, you can get a filtered view based on a tag search across multiple documents and so on (see for example http://sachachua.com/blog/2008/01/tagging-in-org-plus-bonus-...).
In addition to tags, org-mod provides a second similar mechanism with properties (searchable key/value pairs annotations to your items).
So org-mode definitely does not restrict you to a hierarchical organization. It is a very cool tool box that, coupled with the programmability of emacs, offers a lot of flexibility to adapt to your own way to do things.
Yes I tend to agree. I still see people at work hunting for the correct folder to store email. Showing them how to use labels (in whatever client they are using) is often something of a revelation. Seeing the freedom you have just handed them can be quite surprising as well as a great way to get a free pint!
I'm a avid command-line and vim user and for a long time made use of text based todo and note taking tools. Whilst my naming schemes and structures always made perfectly logical sense, they never flowed when in use. After many years I concluded that I needed something more than just a river of text and started using tools like Trello and Todoist. Whilst my inner nerd feels like I have failed the command-line in some way, I'm really getting much more out of my tools than I have to put in.
Central todo-systems never worked for me. I use a combination of calendar entries, mails in my inbox, reminders, post-its, physical notes and text-files. I'm not a chaotic person, mind you, i just feel that forcing every task into the same shape (a line of text) tend to distort real priorities.
And if tags fail you, or you forgot adding a tag I found org-occur-in-agenda-files will almost always let me find what I am looking for fairly quickly.
I much prefer to work in a text based world. When I write I draft in vi and then have some scripts that have the formatting turned into something word can deal with for sharing.
So you have to share tables through excel that come from org mode? If so, how do you manage that?
Mind map is not really equivalent unless you restrict connections to rigid hierarchy, but ability to make more flexible connections is part of why you'd choose mind map software over an outliner.
FWIW, orgmode does export outlines to Freemind mindmapper (using ox-freemind.el which is part of the org-plus-contrib packaging at Emacs' elpa repository.
As some others have said, orgmode does get beyond hierarchical structure of outlines by allowing searches on tags, properties, dates, and other items, which then consolidates all relevant headings -- potentially from dozens or hundreds of different files -- in the Orgmode "Agenda". They are thus visually connected in the Agenda, which in addition to consolidating in one spot provides easy access for directly editing, jumping between the relevant headings, and more.
I use Orgmode, have used Freemind a little bit in the past, but find Orgmode a far more flexible tool; Orgmode has so many features that go beyond mere outlining or mind-mapping.
Freemind is primarily hierarchical but you can easily make non hierarchical links (hyper and visual). Also you can add multiple tags/attributes to nodes
Exactly. The org-mode -->> freemind conversion works fine, translating textual hierarchical outline to freemind's graphical representation. I don't think the current exporter has support for exporting non-hierarchical links (e.g., org-mode tags) to freemind, but that would be a nice feature.
Totally agree - being able to not only list things from top-to-bottom but more visually seems to work far better for me as well (though I suspect people's brains may work differently?)
My love for outlining like in Workflowy or org-mode but my need for more visual-based tools like Scapple or mind-mapping is what led me to scratch my own itch and create Mindscope. It's an attempt to be the best of both worlds. Would love feedback, of course... http://www.mindscopeapp.com
I'm a little put off by ipad-only: I want to be able to at least read my data from anywhere. I don't want it gone for a week when my iPad gets run over by a car. I'd also like some non-video content on the website: when I visited the page just now (on Android) I see two videos and an app store link. I'm not going to watch a video on the bus right now so that leaves me with no info at all.
I love the ideas behind this app. I haven't tried it since it's iPad only and the last update was in 2014, but looking at the video, it has so much promise. It's the first app that I've seen that looks like it has the potential to replace The Brain. Hopefully you can find a way to make some money off it and keep it going!
You're right about hierarchy in the world, but there's a deeper truth that you're missing that relates to hierarchical structures being more amenable to description:
The fact then that many complex systems have a nearly decomposable,
hierarchic structure is a major facilitating factor enabling us to
understand, describe, and even "see" such systems and their parts.
Or perhaps the proposition should be put the other way around. If
there are important systems in the world that are complex without
being hierarchic, they may to a considerable extent escape our
observation and understanding. Analysis of their behavior would
involve such detailed knowledge and calculation of the interactions
of their elementary parts that it would beyond our capacities of
memory or computation.
- Herbert A. Simon, "The Sciences of the Artificial", 2nd edition, page 218
Also, hierarchy in the physical world is not just due to one object being able to only be in one location at a time (or only one object being in a one location). It also arises due to the improved evolvability of hierarchic systems. Again, see Herb Simon's "The Sciences of the Artificial" for a delightful treatment of hierarchic systems and our cognitive relationship with them.
I guess my POV is that hierarchic systems are inherently limiting to us, for data that isn't truly hierarchical (as is usually true for data).
Using graphically-linked data, one could write a process that would generate a hierarchical structure that minimizes crosslinks, and that hierarchy would presumably remain stable right up until the point that a certain relationship is changed, at which point the entire generated hierarchy might be reshuffled. But if we manually assert such a hierarchy, then we have incentive to ignore relationship changes, or try to force them to work within the same hierarchical structure that doesn't serve it anymore. It inhibits our ability to process and understand.
We have more and more ability to visualize true graphical structures and I think it's underutilized in general.
Thanks for the recommendation; I will check it out.
The observation that hierarchical things are easier to describe is interesting. Maybe it has something to do with the adjointness of syntax and semantics [1], and that posets induced by hierarchical structures just naturally share more structure with our hierarchically-structured languages, so they are easier to glue together with Galois connections.
There are so many excellent org-mode export back-ends, it wouldn't surprise me if there was already one for graphviz[1], at least for arranging the header entries into a labeled branched graph (and with inter- and intra-document linking they aren't necessarily acyclical (trees)). And/or it shouldn't require too much effort to write one.
Orgmode supports export to the graphical mind-mapper, Freemind. Graphviz/dot is supported as a language you can use within org documents as part of org-babel functionality, but I don't think there's actually an exporter to graphviz (main use in babel would be to process the dot commands to generate graphviz diagram--e.g., a separate jpeg file--that will then be included as part of an exported org document).
As I understand it, Freemind is preferable to graphviz for export/translation of the org document itself, more oriented to combining text with visual mind-mapping than graphviz, which is focussed more on diagrams.
https://en.wikipedia.org/wiki/FreeMind
Language overloading strikes again... I meant graphical in terms of the data structure, not the UI. In FreeMind, can a child node have multiple parents?
Ahh, yes, in Freemind there is, strictly speaking, only one parent. I think they adopted this restriction because folding/collapsing of nodes is hard to define and/or hard to understand without this restriction, and folding is an important part of helping mind naturally grasp the big "picture".
However, within the visual Freemind diagram it is also possible to create arrows from one node to another that are not part of the main restricted hierarchy. I know some users have created these with the intent of creating multiple parents for a single node. So I would say that although Freemind doesn't support it very well (i.e., it's a hack, not natural like with graphviz), it does support showing visual relationships of single child with multiple parents. (And as far as export to Freemind from orgmode goes, I expect the export could be written to include arrows that are intended to show multiple parents of a single node, although I'm pretty sure it doesn't right now.)
Not my brain. Mine is spacially topographic, and ontologically N-way associative (and, fafaik, ontologies are hierarchical). But I can't think of any way to describe my brain as graphical.
Since this is a computer science (well, sort of..) board, I'm guessing he means 'graph like structures' in the sense that we're taught as first year undergrads -- nodes and edges.
hierarchical doesn't imply that every organizes their hierarchies the same. The vegetable counter example is moot.
The brain works in many different ways - I'm sure you're not wrong about the graphical structures being intuitive. But there's a reason why outlining is so ubiquitous in daily life and many professions.
My thinking seems to enjoy being structured as a directed mostly acyclic graph. That can represent causal links as well as containment, association, and so on. I've done brainstorming with GraphViz and found it quite wonderful. The syntax is also somehow inspiring and fun.
Maybe it's just me, but at my grocery store I've seen them stock the same item in multiple places because it makes sense both places. That said, they're a hippie health food place.
That is a solution, along the lines of xeroxing your insurance forms to store them in multiple folders in your filing cabinet. It has synch problems, though - if you have to update in one place, you might forget to update the copy. Or in the store, you might run out in one place and not the other, which causes waste of effort.
A step better is how libraries have card catalogs, but a step better than that is online grocery shopping - which definitely does not store their products in a hierarchical database.
It doesn't really have much to do with how brain works. Graphical way of arranging stuff is just something you are accustomed to. I find graphical ways of organizing information harder to use and much prefer hierarchical structures.
I don't know about Dominic System, but Memory Palace method is not about analyzing and/or grokking relations between items at all. It's simply a method that helps you remember stuff. Outlining is not about remembering stuff. It's about analyzing, breaking into component parts, understanding relations.
After outlining you will typically "remember" a subject better, but that comes about as a side-benefit of having a better understanding of the subject, not because of some method that simply helps brain recall things.
Unless I'm the only one that is frustrated that you never know where to find vegetable oil in the grocery store. Is it in the cooking aisle near the flour, or the condiments aisle near the vinegar? Seems like every store chooses differently since it doesn't make sense for them to put it in both places.