Read, Think, Write

Criticizable Thoughts

#11, December 2, 2024

A thought only becomes criticizable for others when it is written down. Having an idea is cheap, writing it down less so: Deniability—“I was just thinking aloud!"—is gone, and a thought can be scrutinized.

Once you’ve written something down and handed it over to others, you are somewhat invested in it. Admitting that the idea was stupid comes at a cost. One has to give the idea up and admit failure. This is nothing bad per se, but not everybody is used to admitting failure. If you don’t like to admit that an idea of yours was half-baked or outright bad, better don’t pass it to others in written form.

The more you are invested in an idea, the harder it is to admit that it was a bad one. People who just talk about an idea, but never put it down in writing, don’t have to admit that it was bad, but rather can always claim that it was just misunderstood. As long as the idea is only uttered verbally, one can say that it wasn’t stated in full, and some crucial bits of it have been withheld.

Once written down, a certain completeness of an idea is expected. There are no longer excuses that something wasn’t put down completely, for there are higher expectations on the written word than on a casual conversation. One is expected to go over what he has written down before giving it to others, thereby noticing gaps to be filled in. A conversation can always remain incomplete, and it is totally acceptable to continue talking about an idea later on. Contrary to this, sending out a sloppy draft to others for serious consideration is rather rude.

Ideas can be discussed and further refined in brain-storming sessions and the like. However, as soon as further action has to be taken to follow up on an idea, it first needs to be written down to establish a common ground. This is especially important if other people than the originator of an idea have to act on it. Otherwise, those people can be blamed unjustly for “not getting the idea” once failure becomes obvious.

Somebody who’s not willing to carefully write an idea down doesn’t want to be criticized or held accountable for it. He probably is not convinced of his own idea. An idea should not be taken seriously as long as its originator refuses to write it down. The originator should also not be taken too seriously in that respect, for he prefers fantasizing to thinking, and rather remains dwelling in his cloud-castle than confronting criticism of the unwashed on the market-place of ideas. Lofty ideas and cloud-castles are both made out of thin air after all; neither of them can be grasped.

Digital Minimalism

#10, December 1, 2024

December is a good occasion for an information diet. Unlike with a new year resolution for digital minimalism in January, I’d like to start the new year cleansed from mental clutter. So I started my month of digital minimalism today.

I’m following the plan layed out in Cal Newport’s book Digital Minimalism; more precisely in the third chapter called The Digital Declutter. No matter if you call it an Information Diet, a Dopamine Detox, or a Digital Declutter, the idea is to wean yourself off of optional online activities for a thirty day period, after which you deliberately reconsider taking up some of those activities—or to stay abstinent from them for good.

The key to the process is that on the one hand you only have to restrict yourself for a very limited period of time, which makes it a lot easier. “I will be allowed to binge-watch Netflix again after that thirty day period, just not now.” sounds a lot more feasible than “I will never be allowed to binge-watch Netflix for the rest of my life.”

On the other hand, chances are that you no longer feel the urge to binge-watch Netflix after that thirty day period, because you got rid of that particular addiction. Yes: those products have been carefully engineered by multi-billion dollar companies so that you spend as much time on them as possible, which is the subject of the first chapter of Digital Minimalism called A Lopsided Arms Race.

Cal Newport emphasizes that you absolutely need to fill the void created by abstaining from optional online activities, which use up a lot of our leisure time. You need to pick up activities such as woodworking, exercising—and also something you can do without special equipment or a lot of preparation. In my case, it’s reading and writing to my daily dump. In December, I have a lot do do already, so I won’t be bored anyway. I just need to fill those idle minutes in between activities.

So what am I putting on hold for the rest of the year?

  1. Videos: I already kicked the YouTube watching habit earlier this year. Now I’d like to be more consequent. I deliberately deactivated my YouTube history, so that no recommendations are being shown when I land on the website. I’ll continue to record videos for the classes I’m teaching, which I review offline, but without a history, I won’t be pulled back into watching random videos. One exception is watching movies in a social setting. Showing video examples in classes is also OK, but I won’t research new examples for the time being. (Most students’ attention span is below twenty seconds anyway, so videos are not helpful to grab their attention any longer.)
  2. Social Media: I temporarely deactivated my LinkedIn account earlier this year, and I don’t feel the urge to reactivate it. I’ll spend some time on the German Debian Forum, for which I’ll write a couple of articles in December. I’ll also participate in the discussions surrounding my articles, which is kind of a duty for me as their author. But I won’t join any (political) discussions in the Smalltalk forum. Technical forums and StackOverflow are OK, as long as they support my professional goals.
  3. News: I kicked the news checking loop habit almost two years ago. However, I still follow some blogs, which often link to news stories. I refrain from clicking on those original sources, except if longer essays from non-news publications are linked to. However, I used to check those blogs multiple times a day. Checking those blogs only once a day didn’t work, so let’s go further down to zero checks per day, at least until the end of 2024.
  4. Books: I won’t buy any books this month, but rather read those I already have, which are a lot. My worst addiction is buying books. Putting them on a whish list, which I re-consider after a couple of weeks, saves a lot of money and shelf space, because most books don’t appeal to me a couple of weeks later. And I rather have those books on a whish list than on my bookshelf—and the money rather in my bank account than in Amazon’s. I also must not check the websites of technical book publishers, which I do a lot, even though they just publish a new book once or twice a week.
  5. Games: I stopped playing video games earlier this year, but I just make it explicit that I won’t take up that nasty habit again. Gaming makes it harder to focus afterwards, makes me fidgety, and leads to poor sleep. And I absolutely need some good sleep every night.
  6. Sports: No Formula 1 races, no Ski Alpine races—except in a social setting. I wanted to wean myself off of Formula 1 for ages, which just wastes my Sunday afternoons. The season will be over soon anyway, but I must not check any news about the sport in the meantime. For Ski Alpine races, I actually consider to gather some friends in my office space to have a good time together with a couple of beers. It’s a lot more enjoyable to watch races together, anyway.

I also set some rules for listening to music (not while working) and for online shopping, but those aren’t the biggest problems.

I’ll be reading and writing a lot more during that time. And hopefully, I’ll be more efficient and effective when doing my work, of which I have quite a lot to do in December.

Rowing for 10 kilometers every morning at five o’clock is a very beneficial habit. Since September 9th, I didn’t miss a single day. If I stay healthy, I’d like to stick to it through the entire winter, except when spending the night somehwere else, which is quite possible on Christmas.

I’ll also continue my Daily Frontend Grind, which is the activity following the rowing session (and the shower right afterwards). Learning TypeScript isn’t the most intellectually stimulating activity, but I just want to have that language available in my toolbox for the future.

I’m looking forward to start my new year with a refreshed and uncluttered mind, which I’ll plan to fill with mathmematics and maybe the Elixir/Phoenix tech stack—and with a lot of good literature.

Four Quadrants of Activities

#9, November 24, 2024

Our activities can be classified into four quadrants. The x-axis shows how we’re feeling during an activity, and the y-axis shows how we’re feeling after the activity. Both axes range from feeling bad—left/bottom—to feeling good—right/top. A z-axis showing how we feel before the activity is not considered for the sake of simplicity.

The four quadrants are enumerated from top-right counter-clockwise, which gives us the following categories:

Activities of quadrant I are pure pleasure. Those are playful activities we voluntarely do, like actual playing or having sex. However, the joy produced by those activities decrease quickly (under the law of diminishing returns), and we need to find other activities to do. It’s almost impossible to fill one’s day only with purely enjoyable activities, which explains why idle billionairs are so miserable.

Quadrant II encompasses effortful but rewarding activities. Exercising on a level that yields health benefits is not enjoyable, but makes us feel better afterwards. We need to put up with a certain amount of pain in order to reap the benefits of such activities. Focused learning or productive work belong to this quadrant, too. Ideally, we enter the state of flow, turning work into play (quadrant I). But the state of flow can only be entered after some initial strain and therefore is rather to be seen as the reward for a phase of hard work.

Nobody likes to do the activities of quadrant III. Unfortunately, a big part of our education and working life takes place in this quadrant. No matter how well we organize our lives, we have to spend some time doing activities of this kind. However, if we can reap some reward from the most tedious work, a quadrant III activity can be turned into a quadrant II activity.

First Insight: If we start with an activity that feels awkward, we still might turn it around to become a rewarding endeavor. If we’re happy with the result, we also judge the strenous process in a more positive light. Mindset and organisation are to a large extent techniques to turn quadrant III into quadrant II activities.

Quadrant IV contains activities that feel good while doing them, but make us feel bad afterwards. Drugs are the most obvious example. Spending time on social media, which arguably is a drug, feels the same. However, it’s the dose makes the poison: A dinner party with friends benefits from a couple of drinks, but once you surpass a certain threshold (let’s say: more than five to seven drinks), the evening might become a bit more enjoyable, but the next day less so. Quadrant IV activities always come with remorse.

For a lot of people, watching TV is a quadrant IV activity. After a long time of abstinence (now 17 years in my case), the activity itself doesn’t feel good anymore, but rather nauseating. The same is true for social media or following the news, after one has stopped using them for a couple of years. Video games probably were the worst offenders in my youth. They completely overexcite us, which not only makes it impossible to do any meaningful things for the rest of the day, but also cause bad sleep, ruining the next day.

Second Insight: Weaning yourself off of quadrant IV activities reduces their appeal.

Conclusion: Quadrant I activities are enjoyable, but not sufficient for a good life, which requires mostly quadrant II activities. With the right mindset, quadrant III activities can be turned into quadrant II activities. Quadrant IV activities are harmful, but loose their appeal when abstaining from them.

Mathematics, Language, and Programming

#8, November 21, 2024

I haven’t been writing a lot recently, but I read a couple of EWDs out of frustration. I’m seriously thinking about reading either A Discipline of Programming by Dijkstra himself, or The Science of Programming by David Gries, which is accompanied by a foreword written by Dijkstra. When I read something written by Dijkstra, I always agree with him—as far as I understand the subject he’s writing about.

One quote from EWD 498 always comes to my mind when programmers complain about the relative importance of languages in the school curriculum… and about math, of course:

Besides a mathematical inclination, an exceptionally good mastery of one’s native tongue is the most vital asset of a competent programmer.

I am a very poor mathematician, but unlike many other programmers, I am not proud of it. I have a mathematical inclination, but a very limited knowledge. However, I understand that mathematics can help us tremendously in software development, even if it’s just the simplest math. A couple of examples:

  1. I cringed when Kent Beck wrote about “All of the tricky code for rounding to three decimal digits […]” in the introduction to Test Driven Development: By Example. The code is not tricky, it just requires simple arithmetic. If you’d like to round a money amount x to, say, a granularity of 0.05, just multiply it by 1/0.05 (i.e. scaling it up), round it, and divide it by 1/1.05 (i.e. scaling it down): round(x*(1.0/0.05))/(1.0/0.05). It’s a literal one-liner. What bothers me is not the word tricky, because this one-liner can be considered an arithmetic trick, but the word “all”, which certainly is indicative of a significant amount of code having been written to solve this petty issue.
  2. I worked for a company that presents summarized textual content interspersed with quotes from the original text. Layouts used to be done manually, with quotes in the margins. For the web version, especially for the mobile application, the quotes had to be displayed in between paragraphs. The simplest solution was to just tuck them in between the topmost paragraphs, until there were no quotes left, leaving the gaps between paragraphs empty further down below. I figured out an easy solution by calculating the paragraph-quotes ratio (one quote after every n paragraphs), and then counting up a floating point variable while displaying the paragraphs, adding that ratio to the counting variable after each gap. Whenever the variable surpassed another whole number, a quote was due for display. Again: only addition, division, and rounding was required.
  3. For a time-tracking software, I had to figure out how to perform a proper holiday computation, which is a bit tricky when the workload of an employee changes during the year, especially when holidays were already taken in that year. I walked away from the computer, used a combination of the whiteboard and pen and paper to figure out a formula to compute the amount of holidays left. Thereby I realized that the formula could be simplified, eliminating a couple of parameters such a function would require, reducing the amount of data that needs to be loaded from the database, and thereby improving performance before writing a single line of code. Once more: addition, subtraction, multiplication, division, and a few simplifications on high-school math level solved the problem.

What’s most remarkable about these mathematical solutions is not that they are complicated in any way, but that nobody working on that codebase came up with them, even though the underlying issues caused significant pain and weren’t resolved for quite some time.

So having a mathematical inclination can be a far cry from being mathematically proficient. Somebody who is mathematically inclined just considers a mathematical solution to a programming problem a possibility that has to be investigated before resorting to trial-error programming done so often. In that sense, I am rather mathematically inclined.

English is not my native tongue, and I’m well aware of my limitations. This text probably contains a couple of mistakes, both gramatically and orthographically, and its style certainly could be improved by any high-school English teacher.

German is my native tongue, but even there a good high-school German teacher would be able to point out mistakes in my writing. However, I am probably more proficient in writing German than 95% or even 99% of the general population, because I read at least a dozen books written in German per year, and I also write almost every day. Therefore, I also tick the second of Dijkstra’s boxes: an exceptionally good mastery of one’s native tongue. Maybe I’m just the one-eyed man in the country of the blind, but one is compared to his peers, after all.

So I have the most vital assets of a competent programmer, now I just have to become that kind of programmer Dijkstra would consider competent. The question is: how to become that competent programmer I’d like to be?

Tackling The Science of Programming right away probably is a bad idea, because it introduces both new mathematical techniques and its application to programming problems. I’d rather learn about mathematical techniques before, and just then tackle their applications to programming.

Concrete Mathematics by Knuth et al. is certainly a good introduction for the kind of mathematics needed in computer science. With its roughly 450 pages, it’s considerably longer than The Science of Programming. It featuers exercises on six difficulty levels, of which I only consider doing the ones of the lower two or three levels: I’d like to become a mathematically inclined, competent programmer, not a mathematician, after all.

I always saw mathematics as an exercise in purification of the mind. It also purifies solutions for everyday programming problems, which then require far less testing than solutions built-up using the programming by trial and error paradigm that is so prevalent nowadays.

I must not forget to read good texts in my native tongue alongside. However, English is programming’s native tongue, so reading good English texts certainly is helpful, too. Morris Kline’s Mathematics in Western Culture, which is also recommended by Dijkstra in multiple EWDs, would certainly make for a good accompanying read.

Paul Graham on Intelligence and New Ideas

#7, November 16, 2024

In his essay Beyond Smart, Paul Graham argues that intelligence is a precondition but not sufficient for having new ideas. Summarized:

Einstein wasn’t just smart, but had important new ideas. Intelligence isn’t enough for high achievements. Discovering new ideas is more desirable than being smart. The idea that intelligence is the most important quality is an illusion fed to us since childhood. Having new ideas was historically less important than winning in conversation thanks to superior intelligence. What additional qualities are required for having new ideas? Unlike intelligence, many of those qualities can be cultivated. Once you accept that intelligence isn’t the most important quality, new possibilities arise. There’s no consensus about what matters if it’s not just intelligence. An obsessive interest in a topic is an important quality. Being independent-minded, which is related to intelligence, is another one. Working on your own projects and overcoming obstacles is a good technique. Other qualities are most prevalent in youth: enough sleep, good health, a lot of free time, little stress and few responsibilities. Writing ability allows you to discover new ideas by writing about them. The gap between intelligence and having new ideas is not a wasteland of unrealized potential, but a resource for new discoveries.

Again, the ability to write stands out as one of the qualities than can be cultivated. The other ingredients—health, sleep, the lack of responsibilities and stress—are life-style choices that require prioritizing. (Earning less to have more time is a hard tradeoff to make.)

The essay is the perfect text form to discover new ideas. I don’t know if the word has a latin origin—and since this is an essay, I don’t bother looking it up. But essayer is a French verb that means “trying (out)”, so an essay is a text in which the author tries out new ideas as he writes. Paul Graham again:

[Y]ou don’t think of the ideas first, and then merely write them down. There is a kind of thinking that one does by writing[.]

While summarizing the essay, particularly the paragraph about having an obsessive interest in a particular topic, I had a realization that is linked to the optimistic tone of the essay: What if I could turn my most important annoyances into a career? I’m clearly annoyed by the way so-called “knowledge workers” approach their work on a daily basis. This bothers me, because I’m constantly pulled into their ways of doing things, which prevents me from cultivating habits for undistracted, deep work that I value so much.

I suffer from this through my entire working life of more than 20 years. Why wasn’t I able to turn this annoyance into something meaningful? Changing the behaviour of other people is next to impossible, and working alone or only together with a few select like-minded people… well, I actually didn’t try that out yet. I actually had to stop in the middle of the sentence here to reconsider the issue.

If I would like to have good ideas once in a while, maybe I just should keep clear of people that don’t share my views on how to organize one’s work. Other annoyances to be avoided are business people that incessantly follow each trend of the year, e.g. adding AI to every product (and conversation) in 2024, or using a blockchain for anything in 2021. They might be good customers to a snake oil salesman, which is not the career path I’ve chosen.

I founded my own company so that I could be in control of my time and pick the people with whom I work together. The mode of working could become a selling-point: I’m not jumping on the AI bandwagon, nor will I spend my days in video conferences that achieve nothing; but I’m offering focused deep work hours for actual problem solving.

The way I worked on this essay looks antiquated: I copy the original essay in a text file, add some YAML-header for the meta data, then I convert the text using Pandoc (using the troff filter) to a nice but simple PDF, which I print out. I read the essay once and highlight the most important passages. On the second reading, I summarize each paragraph with a single sentence. When I’m done, I start a new essay by first summarizing the text I’ve just read.

This might not be the most efficient way to summarize the text, but it’s certainly the most effective way, because it lets me get into writing mode afterwards, while my mind is already charged with new ideas that already have been processed three times in different ways—reading, summarizing paragraphs by hand, summarizing the entire text at the keyboard. Most other people would rather summarize the text using some AI tool and then stare at the screen with no idea what to write about it. Very efficient, but totally ineffective.

This might be another quality for discovering new ideas: Replacing efficiency by effectiveness. Rather go deep the hard way than stay shallow the easy way. Real knowledge work requires depth, and depth requires effectiveness instead of efficiency.

Edsger W. Dijkstra: A Counterpoint

#6, November 15, 2024

In the short documentary Discipline in Thought, Edsger W. Dijkstra compares two different composition styles:

There are very different programming styles. I tend to see them as Mozart versus Beethoven. When Mozart started to write, the composition was finished. He wrote the manuscript in one go. In beautiful handwriting, too. Beethoven was a doubter and a struggler, who started writing before he finished the composition and then glued corrections onto the page. In one place he did this nine times. When they peeled them, the last version proved identical to the first one. That iterative method of programming is somehow a very Anglo-Saxon custom. British education is pervaded by it. People learn, when they write, not to try to get it right the first time. Just write what’s on your mind and then rewrite repeatedly to get the product you want. That’s partly why word processors are marketed so aggressively and partly why they have been so successful there. While it is one of the advantages of working with pen and paper that when you start a sentence, you should have it ready.

Dijkstra sees the need to have the sentence ready as an advantage. The late Dijkstra wrote his texts using pen and paper. A quick glance into the EWD Archive suggests that Dijkstra switched from a typewriter to his fountain pen (the Montblanc Meisterstück) as his main writing device in the early 1980s.

For Dijkstra, the process of writing is about bringing a finished thought to the page. He clearly is on the side of Mozart in this respect. Paul Graham seems to be on the Beethoven side, as he suggests in Putting Ideas into Words:

I’ll often spend 2 weeks on an essay and reread drafts 50 times.

Or in How to Write Usefully:

I’ve never tried to count how many times I proofread essays, but I’m sure there are sentences I’ve read 100 times before publishing them.

Paul Graham is an advocate for LISP and bottom-up programming.

How do you divide a program? The traditional approach is called top-down design: you say “the purpose of the program is to do these seven things, so I divide it into seven major subroutines.” […] Experienced Lisp programmers divide up their programs differently. As well as top-down design, they follow a principle which could be called bottom-up design—changing the language to suit the problem.

Dijkstra, on the other hand, developed an operating system for a computer that wasn’t even built yet. He had to get it right without receiving any feedback derived from running a program on actual hardware. LISP programmers, in contrast, use a REPL (read-evaluate-print loop) to get immediate feedback all the time.

Having been born in 1930, Dijkstra received most of his education during and after World War II. Back in those days, paper probably was rare and expensive. So you had to use it wisely. Even though paper became cheap later on, Dijkstra’s writing and thinking style was shaped under the idea that you should get it right once you put in on paper.

He supposedly designed his algorithm for finding the shortest path between nodes in a weighted graph (named after him as Dijkstra’s algorithm) during a coffee break:

One morning I was shopping in Amsterdam with my young fiancée, and tired, we sat down on the café terrace to drink a cup of coffee and I was just thinking about whether I could do this, and I then designed the algorithm for the shortest path. As I said, it was a twenty-minute invention.

Not having pencil and paper available was not an issue to him, quite to the contrary:

One of the reasons that it is so nice was that I designed it without pencil and paper. I learned later that one of the advantages of designing without pencil and paper is that you are almost forced to avoid all avoidable complexities.

Again, Dijkstra sees a constraint as something positive. He wrote and published the paper three years later. He also thought about the shortest path problem a long time before, but he didn’t bother writing about it until he solved the problem.

This stands in contrast to Paul Graham’s publication style (again from How to Write Usefully):

When I proofread an essay, there are usually passages that stick out in an annoying way, sometimes because they’re clumsily written, and sometimes because I’m not sure they’re true. The annoyance starts out unconscious, but after the tenth reading or so I’m saying “Ugh, that part” each time I hit it. They become like briars that catch your sleeve as you walk past. Usually I won’t publish an essay till they’re all gone—till I can read through the whole thing without the feeling of anything catching.

This approach can be described as debugging into correctness, whereas Dijkstra proofed correctness prior to writing and programming.

Having grown up with cheap paper and computers, my habits are more on Graham’s (or Beethoven’s) than on Dijkstra’s (or Mozart’s) side. It would be interesting to figure out if the two composers approached writing music differently: Did Beethoven work bottom-up by starting with a couple of nice tunes that he put together to great symphonies? And did Mozart start with a grand idea from which he worked down to the melodies? And most interestingly: How did Johann Sebastian Bach, whom I consider the most important composer of all time, approach his compositions?

Arguably, both Beethoven and Mozart—and both Graham and Dijkstra—created great works and influenced their fields of expertise. So there’s no “right” or “wrong” approach, at least not judging by the outcome. I like to listen to Mozart’s and to Beethoven’s music. And both Graham and Dijkstra wrote essays that I learned a lot from.

Mozart is considered a genius, a Wunderkind. In the field of computer science, Dijkstra is considered a genius, too. When one thinks of a genius, people like Einstein or Mozart come to mind immediately; Beethoven less so, even though many consider him a genius. And while Paul Graham’s work is highly influential, it is only so in the rather narrow field of IT startups and LISP programming.

Mozart did work hard on his compositions, and so did Dijkstra on his algorithms. But most of their work supposedly took place in their minds before they put it onto the paper. Maybe the genius just has a bigger working memory. The need to switch to paper comes up much later in the process. A genius knows how to organize his thoughts better without laying them out physically before them. And he’s obviously not afraid to loose them.

Not being a genius, I feel quite comfortable developing my thinking while writing. Thinking about a problem for a long time just creates loops in my mind rather than insights. Even though I bought a Montblanc Meisterstück a couple of years ago, I’ll never become a Dijkstra. However, writing with that pen lets me be especially careful. It’s a good contrast to my usual jotting down my thoughts. And the result looks rather nice, even with my awful handwriting.

So I won’t try to adopt Dijkstra’s or Mozart’s style; this simply won’t work out for me due to my rather limited mental capacity compared to those geniuses, and I simply was brought up with the iterative method due to the availability of cheap paper and computers early in my life. Even back in school, we wrote all essays twice: One draft and one clean copy, of which the latter was handed in.

So I’m feeling quite comfortable editing my texts multiple times before I submit them to my website. And I’m fully aware that probably nobody is going to read them. I write for myself, but I don’t like to read my name next to typos, grammatical mistakes, and poor reasoning. And I won’t be ashamed for publishing corrections, which I’ll make obvious.

For me, writing—and thinking—is more about the process, and less about the result.

Separating the Thought from the Mind

#5, November 13, 2024

Holding a thought in your mind and inspecting it at the same time is hard. Just thinking the thought already requires working memory. Thinking critically about it at the same time is probably too much mental strain for a non-trivial thought. The obvious solution is to write the thought down.

Writing not only frees up working memory, it also creates a distance between the thought and yourself. A thought held in your mind is still a part of yourself. A thought written down as a proper sentence is a separate object. It could have been written by you, or by any other person. (Re-reading one’s old texts often feels strange, because there was—hopefully—a further development of one’s thoughts in the meantime.)

As mentioned in the post about Putting Ideas into Words, improving a text requires reading it with the eyes of a pretended stranger. This is easier with written words than with thoughts held in your mind for the reasons stated above. Improving a text improves the phrased out thoughts. So it is more effective to write a rough thought down and further develop it as a series of written words than to improve a thought directly in the mind.

Thinking while walking is a habit held in high esteem by many thinkers. Taking a walk to think about an issue is also something I like to do. But I notice that I often tend to get stuck in a loop of thoughts rather than further developing it as I walk. Walking is beneficial in many ways, and I wouldn’t want to miss out on it. But to get more out of a walk than the exposure to fresh air and sun rays, I need to think more effectively while walking.

However, I don’t want to carry a notebook with me while walking. A better approach would be to write down my thoughts before the walk, quickly review what I’ve written before I leave the house, and then think intensely about what I’ve written during the walk. As I return home, I shall immediately write down my thoughts and insights gained from the walk, thereby further developing them.

Paul Graham: Putting Ideas into Words

#4, November 12, 2024

I wanted to follow up on yesterday’s idea that proper thinking requires writing down one’s thoughts. I found an essay written by Paul Graham called Putting Ideas into Words, which is exactly about this issue.

Quickly summarized: Writing about something reveals your incomplete understanding of it. It changes your ideas and uncovers misconceptions. The ultimate test is reading what you’ve just written with the eyes of a stranger. Hardly anybody thinks perfect thoughts that can be put into writing unchanged, at least not about non-trivial subjects. However, some ideas, especially within precisely defined domains, can be formed completely in one’s head; but usually those are rather paragraphs than entire texts. The need to explain an idea in writing puts unconcious knowledge to conscience. Writing might not be the best way to improve one’s understanding of every subject, but it always helps. Writing is the steeper hill than talking, because it lacks tone and voice, but allows for and requires more iterations. Thoughts that aren’t put under the test of writing are incomplete. Writing is not a guarantee for right thoughts, but a precondition.

In my first post, I mentioned that Socrates was known for not thinking very highly of writing. This conflict is now resolved: one could also improve one’s understanding of an idea by talking about it, but:

Less can go unsaid when you don’t have tone of voice to carry meaning. And you can focus in a way that would seem excessive in conversation. I’ll often spend 2 weeks on an essay and reread drafts 50 times. If you did that in conversation it would seem evidence of some kind of mental disorder.

We simply no longer argue on the city’s marketplace but rather on the Internet, which is often called a marketplace of ideas. One could argue that tone and voice could be conveyed in some limited ways through the written word, even without emojis. But the main point is that phrasing gives our thoughts a shape that allows them to be put to the test. Listening to an audio book or to a podcast might be more convenient, but re-reading the same passage over and over is certainly more convenient than re-playing it multiple times. The difference is not so much about the written versus the spoken word, but between a raw thought and a carefully crafted sentence.

As usual, Paul Graham hardly comes up with hard evidence in his essays. He rather appeals to the reader’s own experience and to common sense. This puts the burden of proof on the readers, who then have to watch out for evidence in their everyday experience. I certainly plan to do so in the future. Case in point: I didn’t know what to write about Graham’s essay before I summarized it, even though I read it multiple times prior to that. The thoughts only came to my mind when I reviewed my own writing on the page. One could argue that my thoughts are still incomplete, but now I can see evidence for the claims: writing about the essay shaped my understanding of it.

Less Input, Better Digestion

#3, November 11, 2024

In modern times, the ratio of received input to produced output has shifted in the direction of more input. Probably output increased also, but not in the same magnitude as input.

It began with the printing press. Soon, there was more material printed every day than could be read in a lifetime. Newspapers, radio, television, and the Internet overwhelmed us with input, so that nowadays there’s more information produced in every second that could be processed in a lifetime.

We have a similar issue with food, at least in the first world: There’s way more available than we could or should consume. There are more people dying from the diseases caused by excessive eating than due to malnutrition or starvation, at least in the first world.

So we decided to restrict our food intake both in quality and quantity. And it becomes increasingly clear that we also should restrict our intake of information. Quitting or at least restricting social media usage is one option. Not following the news is another strategy. And so is leaving away all kinds of digital entertainment from video platforms to gaming. I ditched the news for good almost three years ago, am no longer to be found on any social media site, didn’t play any games for half a year now, and very rarely watch videos. So far, this has at least been a time saver. Unfortunately, my ability to focus hasn’t increased considerably, at least not yet.

Most people have a tendency to fill every idle minute with digital input, and so do I. This is a bad habit, because there’s nothing useful to be gained from such short bursts of the usual low-quality, easy-to-digest content. It also has the tendency to turn idle minutes into wasted hours or evenings.

I often have this issue when teaching classes. Sometimes, there is just nothing to do for me, because all the students are working, and the instructions are clear enough for the students to cope without my help. So how can I fill this downtime? I could just stand or walk around and wait until there’s something to do for me. This is a good strategy during most classes, where issues usually pop up within minutes. But for some classes, especially during exams, there are long stretches with nothing to do, rarely interrupted by a single question.

So instead of reading email, checking something on the web, or starting something to work that I won’t bring far anyway, I tried to do just nothing. And doing nothing is hard in an otherwise busy environment on an otherwise busy day. Usually, thoughts pop up: I need to buy groceries after class; there’s a book I’d like to check out; there’s a blog post I started reading in the morning but couldn’t finish reading; etc.

So I wondered: Why don’t I just write all those thoughts down? And so I did. I created a private Git repository called “dump”, for which I create a single text file for every day, e.g. named “2024-11-11.txt” for today. And in the text file I write everything down that comes to my mind. I no longer use idle minutes to check something out, I use this downtime to write something down.

The best strategy to let go of a thought is to write it down somewhere. Everybody who writes grocery lists knows this: Once you wrote that bottle of milk down on your list, you no longer have to think about it—and you no longer will! This constant writing dump is a very good device to let go of thoughts. And if there’s an original thought once in a while, I could look it up in no time using a full-text search.

The idea is not to create sketches for essays, but just to fill the idle minutes with writing instead of reading: less input, more processing: better digesting. And writing is an activity to process one’s thoughts. Thoughts are unstructured fragments made up of words, but they are not proper sentences. Once you write them down, you have to really think through them. Writing adds more nuance. When you think “students don’t read properly nowadays”, then it’s just a random thought. Once you start writing it down, you wonder: “only these students here, or students in general?” and: “just nowadays, or how was it during my time as a students?”.

A written word has more weight than one solely existing in your mind. It reflects back from the page or screen and makes you think. Writing down a thought is a way to actually finish the thought. Thoughts come in fragments, writing comes in sentences.

This leads me to an interesting hypothesis:

Proper thinking requires writing down one’s thoughts.

One could argue that “writing” could be replaced by “speaking” in the hypothesis above, for it also requires phrasing: making sentences out of connected words. But the subject of this blog is writing, so I’ll stick to that hypothesis.

I already noticed how writing down my random thoughts—instead of filling idle minutes with more input—calms me down. There’s less noise in my mind, and more room for focus. Furthermore, writing 1000 words a day, which is my average so far, is a good exercise. (I write my daily dumps in German, but this blog in English; polyglotism is supposed to be a good strategy to postpone dementia. Which reminds me to pick up French and Russian again…)

Now I wonder if this frequent writing will improve the quality of my thinking. If not, it will at least improve my ability to write reasonably well.

The Experiment

#2, November 10, 2024

This blog is supposed to be an experiment. I’d like to practice three skills: reading, thinking, and writing. First, I must read some material worth thinking about. Then I must think about what I just read. And then I must write down my thoughts. I’d like to do this every day, even though it’s the busiest time of the year.

Building up my concentration is a project I can no longer postpone. I was able to build up some considerable discipline in many areas of my life, e.g. programming Rust every single day for more than half a year without missing one day, or getting up at 5 o’clock a.m. consistently and row for ten kilometers on my Concept2 ergometer.

However, when it comes to work, my discipline and concentration all of a sudden vanish. Maybe it’s the environment, maybe it’s the cognitive dissonance I feel when doing work for other people, or maybe it’s a lack of stamina when it comes to unpleasant things to do. (Rowing at 5 o’clock in the morning with a hangover after three hours of sleep is unpleasant, but in a different way.)

This blog is supposed to be a documentation about my efforts. I don’t know what I’ll write—and read, and think, for that matter—about, but most important is the consistent daily effort.

I just finished reading Cal Newport’s Deep Work, probably for the third time in the last couple of years. However, this time I marked a lot of passages. I did so because I’d like to summarize this book. But this time I’m not trying to write the usual neurotically comprehensive book summary about it. Instead, I’ll read a part of a chapter, think about it, and then combine the author’s message with my own thoughts about it. I’d like to write short essays.

This blog was created using Hugo. I first wanted to pick a minimalistic blog theme, but even the simplest ones seemed too bloated for my efforts. So I wrote my own minimalistic template, which doesn’t even support taxonomies. The only navigation feature is the anchor for each post next to its publication date, e.g. #2 referring to this post’s title.

If the page will get too long after 20 posts or so, I’d like to include pagination. But there will always be full blog posts displayed; no separate article pages being teased on a landing page. Pure simplicity. The only luxury is the font being used: TeX Gyre Schola, which weights ~200kb. It’s not the perfect font, but it has everything I want: serifs rendered nicely on a display, a nice italic style that can be combined with bold, and it looks rather bookish and conservative.

The Idea

#1, November 10, 2024

Paul Graham’s short and sober essay Writes and Write-Nots strongly resonated with me: Maybe writing is the next big thing, not despite the fact that it can easily be replicated using AI tools nowadays, but because of it: When everybody is delegating writing to AI tools, nobody will be capable of doing it on its own in the foreseeable future. And because clear thinking requires clear writing, as Paul Graham claims, only those who can write will be able to think in the future.

[W]riting is thinking. In fact there’s a kind of thinking that can only be done by writing.

This claim seems shaky, especially when one considers Socrates, who didn’t hold writing in high esteem. Paul Graham usually does not elaborate on his premises in his rather short essays. But let’s accept this premise for the moment.

If writing and thinking will become rare skills, those who are able to write and think will have huge benefits in the future. Those benefits will become clear within a generation, say, in 15-25 years, when those using AI tools for every task will establish themselves in the workforce.

So the next big thing isn’t AI (or Blockchain, or Big Data, or DevOps, or Cloud Computing), it’s thinking. Investing time in reading (to get new material for thinking), thinking, and writing is the ultimate long-term bet.