Pile of Index Cards

“Might depend on what you need to file. I have a colleague that tried to file his emails chronologically and could never find anything. He finally started using the technique I use, which was to set up personal folders by subject names using the groups and topics I interact with most. …”

“I’m kind of like him. I can place things on a calendar in my mind. I have an idea when I did/wrote/filed something, but not necessarily why and where. I used to try to sort my photos via subject–that lasted a month. I have thousands upon thousands of photos, both in film and on hard drives, and trying to sort them all by subject or whatnot was just too much for me, so I resorted to a chronological system. Everything is filed by date, and even has a file name, based upon date. I’m able to find just about any photo in mere minutes now, and usually seconds. …”

I found an interesting discussion above in a thread at D*I*Y Planner. They discuss about sort file by subject or chronological order. Failure and success. What makes difference? Following is my analysis.

In the case of Star’s friend, the matter is “email” which is mainly send by “someone else”, brute-forcely, no matter what he want. Even if he try to manage file like email by chronological order, it will be difficult because the information is “less impressive”, and as a result, is forgettable.

In the case of jonglass, on the other hand, the matter is “photo” which is mainly taken by “himself”. We can imagine when and what situation we take photo. We take photo as a record or of something that moves our heart. Even though the information is come from outside, the motivation to take photo is comes from completely “inside”. They are “more impressive”, and therefore remains in mind.

Return to PoIC. Among Four Cards, strength of impression can be expressed as follows,

Discovery cards are something like “snapshot” of our mind/thought. We write them because they are impressive. The ideas are loop, loop, loop in a brain, and then come to this world as index cards. If the Discovery cards are dominant in a system, “sort-by-chronological-order” works good, because, like jonglass’s case, most of information comes from “inside”, as a result, “more impressive”. Each of us have own time-axis. Index is directly connected to our own history. In addition, our own idea right here and right now is includes own past ideas.

Now the difference between two cases above is clear. Star is talking about most right-hand-side, and jonglass is talking about most left-hand-side of the equation, respectively. Note that needs for sort-by-subject has completely opposit sign of inequality in the equation.

Jonglass says about this skill as “I suspect it’s a brain thing…” in the title of his comments. In my case, I learned the word “chronological order” in twofold : from Noguchi Yukio’s book “Cho-Seirihou” and Joan Cusack’s movie “High Fidelity“. I’m quite not sure this skill is special. Because people in flickr first manage their pictures in chronological order, and later organize “set”. In the movie “High Fidelity”, a guy stack all of his record collection (huge) in chronological order, and later organize “compilation album” for his girl friend. They don’t have index because simply they don’t need it. What is different from PoIC’s chronological order and Task Force? This is a matter of a strength of impression and whether one aware time-axis inside oneself or not.

“A farmer went out to plant some seed. As he scattered it across his field, some seed fell on a footpath, where it was stepped on, and the birds came and ate it.

Other seed fell on shallow soil with underlying rock. This seed began to grow, but soon it withered and died for lack of moisture. Other seed fell among thorns that shot up and choked out the tender blades.

Still other seed fell on fertile soil. This seed grew and produced a crop one hundred times as much as had been planted.”

“Bicycle.” It’s more appropriate word to express what is PoIC. One must pedal by oneself. All power to move emerges from inside, and the system just help and accelerate it.

On 2 channel, the largest discussion board in Japan, a guy told that efficiency of bipedal locomotion of human is the worst, however, if together with “bicycle” it turn out to be the best among all species. This is the exactly same what Steve Jobs expressed about what computer should be. A computer getting much faster and complex vehicle than the beginning though.

Unfortunately, my job doesn’t allow me to be all-time analog person. When I have to use car (computer), I use car without hesitation. The right thing should be used in the right place.

Then why I ride a bicycle (index cards)? It may be because I can feel my own breath, movement of muscle, sweat on face, breath of air, drifting landscape, etc. There is no digital way to replace analog PoIC. If a car driver ask for me “Why are you torture yourself? A car is much easier and faster than bike! Are you insane?” Then I may smile and answer back, “Just for fun.”

When someone says “PoIC works great for me!”, then I could share experience with him/her. All I can say for him/her is just “Yeah. That’s it!”. Sometime it’s difficult to express everything in word because experience is more “chemical reaction in a brain” thing.

If you carefully read the manual, you will notice that design of PoIC’s system/method is extremely simple. Then you succeed to capture PoIC’s philosophy. I removed everything that causes air resistance and friction during long ride as possible. Perhaps, it looks something “different” from ordinary bicycle. Write, stack, reuse : In fact, there are no reasons to stop PoIC.

How long we are going? As long as we want to go.

PoIC’s principle is extremely simple: Use four card to express inside/outside of us, stack them chronologically and, reuse them later. That’s all.

I write PoIC manual as I use it as discovery database. Because that the only thing I can do. However, the principle may not limit usage of PoIC for user. So some may use as GTD processing system, and some may use as reference system.

PoIC is generic because of it’s simplicity. It fits for anybody, but there is no “average” person. So user must modify and customize PoIC as it fit for own life, with keeping the principle. For example, there are four possible cards, Record, Discovery, GTD, and Reference. When I optimize PoIC suit for me, the Discovery Card become dominant.

What I have done in PoIC manual is only half. You have to build up the other half by yourself. You may ask how to build. The answer is already in your hand. Use PoIC. If PoIC doesn’t fit for you, you must ask yourself why? How to solve it? How to make it better? Write your idea on Discovery Card. In fact, PoIC is developed by using PoIC like that way.

This proposition will be a good exercise especially for PoIC beginner. When you got your own PoIC, congratulations! Then you can build up your productive/creative system/method on the solid basis of your own PoIC.

Recently, I read Paul Graham’s “Hackers and Painters“, and am very impressed. Programming, painting, PoIC, they are essentially same: making own world.

ScanSnap is state-of-the-art document scanner from PFU/Fujitsu.

To tell the truth, I had been negative with using scanner to scan index cards. My index cards are more than 3,000 now. Add tags of contents for all the cards from now? No way. Make digital database? I have analog database already. Why I have to duplicate it? Even I scan and handle index cards on digital, it is physically impossible to spread it on desktop like this picture. Handling index cards in “analog” is much fun.

All time I see someone who scan index cards, I ask him/her “what do you do with the flood of index cards?” in my mind. It just make double information in the system. As a result, complexity of system is also doubled. That is why number of scanned index cards in Cardlog is so limited.

Recently, however, I have been interested in another use of scanner : Statistical analysis. That is, scan PoIC tag and date/time stamp of all index cards in the dock. Then I will see a sequence of letters like,

with timestamp, where R: Record, D: Discovery, G: GTD, C: Cite (= Reference). I can analyse not only a number of index cards, but also distribution of four kinds per month or per day, transition of the distribution, or when discovery comes (such as morning or evening?). I think I can see some kind of pattern. And it will be a new knowledge.

Such statistical analysis makes sense with greater number of sample. Perhaps 3,000 index cards are not yet enough for it though. A scanner might be good for this kind of routine, troublesome job. It’s something like Perl in Bioinformatics. According to Jazzmasterson’s recent experiment, it is possible to scan 5×3 index cards by ScanSnap.