Tools vs. practice

Jim McGee wrote a great piece on tools and practices. He argues that the ease of getting started with the first 5% is deceptive.

“We focus on the details of particular features and functions at the expense of ignoring the cognitive challenges of deep thought and collaborative work.”

I think it is important that we distinguish between “deep thought and collaborative work”. Tools for these two coincide only for roughly 5%.

It is useful, though, to recall that these 5% did have their merit. Almost 10 years ago, I wrote for my users (in German, here,) “What does my personal productivity have to do with ‘social’ software tools? … users use the same tools [for both].” (In particular, social bookmarks and annotations thrived because they were selfishly provided but were useful to others, and on the other hand, blogs and wikis often served also as notes to self.)

Also, I would like to defend the ease of getting started. Trying out a new tool is a risky investment of time. But there is no way of evaluating a think tool without personally trying it out. The closer it is to the mind, the more it can be compared to a tight garment which must perfectly fit.

For ‘deep thinking’, tools need to be more than just logistics (storage and transport) of thoughts. (In his recent blog post on Intelligence Augmentation, Clark Quinn tackled the tight integration of the ‘cognitive and computational architectures’ by carefully carving out some aspects of what the latter can do well — details, repetition, find correlations.)

Now for collaboration, the ‘tight garment’ cannot fit multiple persons simultaneously. A sad practice is often that the group settles on the least common denominator (email) because a stubborn member (the boss?) refuses to try out a better way.

Here it is useful to distinguish between passive acceptance/ refusal, and active choice of one’s pet function. And then apply Postel’s Principle: “Be conservative in what you send, be liberal in what you accept“. For example, messages without subject line or with the repeated, misleading subject line, are read but should still be avoided. The trick is then to try out what range of choices are tolerable for everyone to accept — not only which features optimally cater to individual expressiveness.

In practice, important examples of differing tastes are: How much should be sent across ‘push’ channels as opposed to ‘pull’ channels, or: where is more divergent information appropriate and where should convergent contributions prevail. Often, the intervention of a stout-hearted but gentle personality such as the ‘wiki gardener’ is needed. So, to answer Jim’s question “Where are good places to look?“: in gardening 🙂 His guess, software development, does not sound too promising to me — despite the GitHub workflow appears like magic — because tt requires more discipline than is usually accepted.

Posted in Knowledge management | Leave a comment

Polarization and Push or Pull

Great food for thought in @gsiemens and @realdlnorman’s exchange. So why can’t media be like a conversation? What makes the internet a “polarization factory” of ‘us vs. them’?

I think the problem is that is has turned too much towards ‘push’ media, and that these are prone to scale-free effects, with their potentially unlimited reach. Each statement uttered in this environment, is soon understood as advocating for a cause or for a political strand, because everyone seems to expect that their tweet may go viral, because that is how the platforms work. Then a differentiated statement can be ripped out of its context, and the ‘yes, but’ is distorted to just ‘no’.

By contrast, ‘pull’ media such as the comment section of a personal blog, resemble the conversation that George describes. The visitor on the blogger’s ‘front porch’ has bothered to come there, and sees the context and the background of a position.

(And yes, I will have to ‘push’ this URL out on Twitter, because RSS is almost dead. So there is no ‘them’ who I can blame for the “polarization factory”; it’s all of ‘us’ who operate it.)

Posted in Web | 3 Comments

Distinctively human?

1. In his latest presentation, Stephen Downes once more advocated personal learning. Previously I have often wondered if it is just a matter of style and taste whether one emphasizes an individual access to learning or not, i.e., whether one prefers learning in the same way a backpacker individual tourist learns about the world, or if one finds it easier and more effective to go with a guided group tour to get the proven optimized standard sight.

 Photo CC-BY by Flickr user supermoving

As a tourist, I never liked the group style, but for learning, the lazy way of letting stuff just wash over me, did have its charme. And it sounds even more promising now since the optimization is ever more perfected by personalization. And since it seems like a ‘science-based’ way to success.

2. There may be even a third type of learners: the machine ‘students’ who do their machine learning. Previously, I often wondered why machines needed to ‘learn’, at all: they don’t need to gradually accumulate their knowledge because they just do a database lookup; they don’t have to learn their rules because they receive them through programming; and learning for any kind of later independence seemed unimaginable.

Now I know that it is not rules, but patterns, what they learn: they learn from data that comes in large amounts and from data that is randomly distributed — much like lifelong human experiences. So our automated fellow learners are already far away from the group path, and they are rapidly approaching our human styles.

3. In an Australian conference last week, @SBuckShum talked about the Cognitive Automation and recommended that we

“Cultivate those qualities that are distinctively human”

So what will prove as ‘distinctively human’ ?

I think it is exactly the personal approach which cannot be standardized, that cannot be automated. It is the individuality and subjectivity that guarantees sufficient diversity for further evolution and to avoid collapsing into a ‘black hole’ of power law distributions, and that guarantees sufficient embodiment for staying grounded in reality.

By contrast, approaches that only accept objectivity, and strive for algorithms that promise an optimized, rational, assessable, uniform solution — are eligible for cognitive automation, sooner or later.

4. This also applies to the optimization of learning. Only standardized, easily assessable learning can be optimized, and in particular, content knowledge is the ideal example. If we compare learning to carrying a load upstairs to the attic storage, there are two distinct goals. One is that the load should be stored up there, and the other is to work out one’s muscles. If it is only necessary that the stuff is accumulated up there in our brain, a ‘lift’ would be a welcome optimization. But jobs that only rely on such content storage, are probably the first to be eliminated.

In a recent blog post, Stephen calls it an ‘either or’ dichotomy:

“are you providing content knowledge or, are you providing literacies?”

and I think it it really a shame how ignorant our universities are in dealing with this question.

Posted in Knowledge | Leave a comment

Seven habits of…

The seven habits of successful people? Well, yes, except that it is only one person (me), and ‘successful’ just means that I am highly content with the workflow that I am going to describe.

  1. I don’t open my email if I could not react, anyway.
  2. I prefer RSS over email news.
  3. I store an item whenever I have reflected on it.
  4. I don’t care much while capturing stuff,
  5. But I care a lot for grouping it later.
  6. Of course, I use my Thought think tool.
  7. I take an easy way to prune stuff.

Read more:

1. I don’t open my email if I could not react, anyway.

If I am preoccupied with a larger piece of work, or if I am not near my desk, a newly arrived email must wait. My curiosity about it is smaller than the reluctance to get interrupted. Enjoying an email means for me, to start thinking about a response rightaway, or to take the time to follow its scents.

2. I prefer RSS over email news.

My categorized RSS subscriptions come across to me as offerings rather than urges. They give me a sense of space, as if I was visiting the authors on their respective front porches, and I have already a certain idea of what awaits me there. Thus it is easier to notice the salient, unexpected, relevant news to pick.

Picking from RSS is more like ‘pull’ than ‘push’. It does not urge me to almost synchronously react — not even to decide if I delete an item or keep it or where to keep it: it rests on its web site without bothering me. Push vs. pull is a big issue today where lots of sites try to control our attention and patronize us. Joi Ito’s “Whiplash” has a whole chapter about pull over push.

Other kinds of notifications feel like a terribly linear assembly line. Once I get into such an unstructured linear list, I feel that it will mean a long time-span with almost no chance to interrupt and digress and still find my way back, and this is somehow daunting. ‘The stream’ is an ineffective construct just made to lure us into filter recommendations. It eats up power to connect items to their contexts rather than to relevant similar ones. Small bloggers are just ‘flushed’ down the stream. See also what Mike Caulfield writes about the stream.

3. I store an item whenever I have reflected on it.

Sometimes I am not sure if I have stored something in my ‘knowledge base’ of my harddisk. Then this little simple rule of thumb helps: If I have thought for some time about an item, I will store its url or a note. Even if it turned out to be rubbish (which I might note in brackets in the file name), I will vaguely remember it as somehow relevant and then I might try to find it again, so it is better to save a cue rather than search it in vain.

4. I don’t care much while capturing stuff.

Capturing an idea must be quick rather than optimized, to minimize the distraction from the task at hand. I don’t think about what would be the most appropriate medium for the note. I fire up Notepad and store the idea in a single separate file. I don’t think about the filename either, and I abbreviate ruthlessly. For my idea notes, the names are almost like a ‘private language’: typically, such a note is a one- or two-liner saying “xxxxxxxx has something to do with yyyyyyyy?”, and then the filename will be xxxxYyyy.txt which I will forget before I revisit my notes.

And I don’t care too much about the folder names. My filing is not optimized for directly searching, but for browsing in context and seeing what is there. So the folder names just need to be recognized within the context of the containing folder. Whenever I was looking in vain for something in one folder, I create a folder shortcut there, pointing me to the other folder next time. So over time, a dense network emerges, the traversing of which strengthens my awareness of the relationships.

And it is not important to find the optimized procedure having the shortest click path. Rather, a default procedure that is well-rehearsed will do to minimize distraction. So, my idea note files are saved to the desktop and dropped to a folder called ‘new’ — via a folder shortcut on the desktop. This minimally invasive note-taking guarantees that no tiny humble idea gets ignored and can be processed later — see next section.

5. But I care a lot for grouping it later.

In contrast to the careless capturing, I really cultivate my notes collection later. The crucial thing for me is not to classify them prematurely, but to be open for new similarities, connections, and patterns, and for new emerging categories becoming necessary. Once I put the notes into the disjoint categories of my private localhost WordPress, I have shuffled them around for so long that I know them well, and I rarely need cross-links.

6. Of course, I use my Thought think tool

Whenever I start to produce some output, I drop all pertinent stuff into my which helps me with the step between stuff collection and outline.

7. I take an easy way to prune stuff.

Sometimes I notice that a digression needs pruning, only after I have struggled quite a while with a difficult text, and I would not want to delete it entirely. Then I am glad about this well-rehearsed habit: I cut the text and paste it into a little text file called xxxxYyyy+1.txt — plus 1, for some distant next attempt that may never happen.

Posted in Personal Productivity | Leave a comment

Week 1 of new cMOOC

Finally I was accepted at the Federica.EU platform for the new cMOOC “Connectivism and Learning”. (Two earlier attempts failed even though I clicked on “I did not receive activation email”, and their support email address did not (yet?) work. But now the activation email was successfully received in my spam folder.) On the ‘plattaforma’, I did not find anything except Downes’s excellent videos and slides, but this should not be a problem in a cMOOC with blogs and all that. And the only remaining problem, the missing hashtag, can be added later on. So let’s start.


Stephen offered plenty of great questions for reflection, e.g. “Reflecting on your own learning, can you identify a process that you currently follow?” or “Do you think you could improve your learning if you had a better understanding of how you learn?” These questions remind me that I wanted to compile some notes about my workflows, hopefully soon.

Another question was about interactivity:

“Why do we need it ? […] second, we need to know that other people are in the learning experience with us.”

IMHO, this includes that we are watching our co-learners in their stage of recognizing, which is much more similar to the recognizing that we need to do ourselves, than the settled, packaged, resources presented by a traditional teacher. If the co-learners ‘teach’ us what they have just come to understand, they are ‘modeling and demonstrating’ their own understanding process.

And watching our co-humans, puts us automatically in a mode which is more open to immediate, contextual, multipoint, ‘all-at-once’ experience than consuming a slice of canned information. Frankly, I was long sceptical about the value of ‘social’ learning, and I suspected that it just helps those who are not willing to interact directly with the resources, and (of course) benefits those who repeat the stuff by explaining it to the weaker students. But in the meantime I have understood that there is more to it.

Posted in Learning | 3 Comments

Lighthearted Exercise

In time for New Year’s Eve, Magdalena Böttger has invented a hilarious-profound oracle game: You set goals for the following years, and throw the dice to determine your achievements.

I look forward to playing it, and in the meantime I noticed that her 20 interesting goals can also be used for something entirely different.

You can try to find connections between them. E.g., some are similar to others, or impact others in positive or negative ways, seem to include others, or be otherwise related. And thinking about their complex relationships can be too difficult if you only look at a linear list and most of the connections are only in your head. This might be a good opportunity to demonstrate the benefit of a think tool.

Select all the text in the list, drag and drop it into my tool — and start connecting and rearranging. (But don’t miss midnight…)

Posted in Misc | Tagged | Leave a comment

Magic of Zettelkasten

My think tool has now a new functionality when exporting to other note-taking systems: It can preserve hierarchical (tree) structures found in imported note collections (such as Mindmaps, or OPML including Scrivener exports), and it can even derive such a tree structure from my own network maps.

zettelkastenMy reason for the improvement was that there are systems which are not limiting their users to narrow-minded “box” thinking, but still use tree structures. One of them is iMappingTool which I have long admired for its aesthetically pleasing visual overview. Using my tool, you can now import your notes there and start imapping.

Another one is a system that has been invented long before software systems, and has always been surrounded by an aura of magic: The Zettelkasten (note box, card index) of the sociologist Niklas Luhmann. In a 1989 video, he shows us (minute 37:25, with English subtitles) how he worked with it, but it is only now that the University of Bielefeld analyzes his legacy and researches its magic.

What I understood is that, besides a dense network of cross references, the distinctive feature is the following: He allowed for arbitrary branching at every point in his hierarchical numbering scheme, in other words, a child note did not have to belong to the same topic category as its original ancestor notes. It was sufficient that its content was somehow connected to its immediate parent.

Why did this have such a powerful effect? Luhmann just recorded “which thoughts come to my mind and in what context”. So, the wider context enabled serendipitous inspirations. But it is very difficult to understand the effects of the small nuances of note-taking techniques if they are only discussed and not practically tried out. This is why my tool offers many import and export options — such that you can play with different systems and understand their differences.

Update 2017-06-14:

See the blue/ green cluster in the diagram: It turned the red/ yellow tree structure into a structure that was no longer purely hierarchical. zettelkasten-schemaFrom the orange ‘Foo 3.2’, a new idea “came to mind”: the green ‘Bar 1’, and from there a larger substructure emerged.

Posted in Knowledge management | Tagged | 3 Comments

New page: Recognizing

The new page is a remix of all my blog posts about McGilchrist’s (@divided_brain) and @Downes’ ideas on recognizing:

Posted in Uncategorized | 4 Comments

Sequential requirements

Does a neural network need domain knowledge? Stephen Downes shows how important this question is, so I’ll try my guess.

“if we require domain knowledge in order to learn, then we require memorization”

According to the traditional learning theories, we construct knowledge upon existing knowledge, for example by linking the new stuff into the known. When we see knowledge as representation, all the higher order concepts are nicely grown from lower level concepts. But at the very bottom there must be some special knowledge, sort of “seed” knowledge, to enable this recursive mechanism because otherweise we would run into an infinite regression.

Even if we use the less abstract idea of “recognition” we might misunderstand it and ask: how can we recognize what we do not already know? The trick is that recognition needs only a part of the features of a pattern to see the whole pattern.

The artificial neuronal networks described in the above post, apply such pattern recognition to learn solely from conversations. How far can this approach be extended? To the bottom, to eliminate the need for domain knowledge?

Let’s first look at human neuronal networks. The conversations begin long before the human can talk, when the baby has their first gaze “conversations” with their mother, at the age of just a few weeks. It is here that they recognize the world around them, long before they learn propositions about “he, she, it, they”, and even the “I” is learned only via the “thou” in these first bodily conversations.

(If it seems far-fetched to compare this kind of conversations with the sophisticated concepts of a given knowledge domain, consider how language covers a long scale of words: While it ends with isolated concepts coded in specialized, domain-specific terminology, it starts with basic ideas that suggest a very bodily context, and a full-senses/ all-at-once recognition, such as the deictic notions of “I”, “now”, and “here”, or other simple spatial or temporal descriptions which are gradually extended via synaesthetic metaphors, or with the modal words that extend from “wanting” to deontic use to epistemic use.)

Similarly, how do neuronal networks learn how to find out which response should be trusted? For the little human neuronetworks, the seed trust is given before they need to start reasoning, and it will enable them to add more trust criteria over time.

Artificial neuronal networks, by contrast, should always depend on some human who decides if their response is correct, or at least if their underlying algorithm is acceptable. So, there are limitations of what they can learn on their own. (This may be a consolation, but still it is unsettling how few humans might be controlling them.)

So, I think, while artificial networks need some prerequisite input, human neuronal networks use recognizing from the very beginning and require no indispensable prerequisites.


This sheds different light on the idea of prerequisites, and it also makes the idea of a “course” much different from the traditional temporal sequential arrangement. If I encounter prerequisites later, there is still time to cover them (like a rhizome is not necessarily pulled out from the top to the root but often in the reverse direction), as long as the course does not build the stuff of later weeks upon the stuff of previous weeks but uses the weeks’ skeleton only as a schedule of when people can simultaneously discuss a certain topic (i.e., if it is organized as a cMOOC rather than an xMOOC), or even lets the community “be” the curriculum.

This is the first, obvious meaning of

“how education may be linear, but learning certainly isn’t.”

(which Doug Belshaw spends “five minutes explaining”). Recognition explains the deeper mechanism of learning as not linear/ sequential (not via fixed isolated representations) but as laminar/ all-at-once (multiple connected features of a pattern). Thanks so much Stephen Downes for the idea of recognition.

The image above shows the logical dependencies of the chapters of an important textbook of my study (B. L. v. d. Waerden, Algebra I, 8th ed., 1971) which I encountered again as an illustration of linearity in the first German book on Hypertext by R. Kuhlen, 1991.

Posted in Knowledge | 2 Comments

Tangible Associations

After experimenting with other think tools, I understand better how the “magic” of my tool works: it makes associations tangible. It turns elusive mental relationships into “hands on” experience, and it compensates for the abstractness of some thought links, with a drawn line “at our fingertips”.


Many tools aim at a similar effect with their representation of our ideas, concepts, notes, text snippets and thought fragments: Each single idea is reified by one individual virtual index card on a cork board, or by a yellow rectangle that evokes the impression of a post-it which can be grasped and moved. One thought, one object.

The haptic impression caters to our worldly nature, and it mitigates the alienating virtual constructs that plague our everyday life as knowledge workers. And there are many occasions where people even insist upon the genuinely haptic tools such as their paper notebook, their fountain pen, or the brainstorming with paper post-its. I understand this preference even better since I can use a pen again on my Surface tablet, for handwriting and drawing, and to substitute both the mouse and (mostly) the keyboard.

The haptic representation of individual ideas is a useful thing, and Heiko’s iMapping (see my last post) does this better than any other tool I know. The items feel like “on your fingertips”.

But what about the links between the items? A complex network consists of nodes and relationships. Since I have encountered DeepaMehta many years ago, I have learned that these relationships can be recorded and manipulated in a similarly tangible way.

Now I tried this in an even more manifest way: Draw the lines with the pen. It is not as optimal as with the mouse (since I cannot use the third button and not even the Alt key), but it is excitingly “haptic”.

A similar tradeoff must be made when it comes to prioritizing nodes versus links. In Heiko’s tool, the notes feel a bit more tangible than my little circles which are 1 click and an eye saccade movement away from their detail. But in return, my links feel easier, and closer to the thoughts they are connecting, because the map is optimized for conserving space, since it is not intended as a long-term store (of items) but as an ad-hoc overview where you drop various text snippets to connect them. Because, you know, connectivism discovered that connections are more interesting than mere items.

Just try it out.

Posted in Personal Productivity | Tagged | Leave a comment