Sunday, June 29, 2008

Data First or Structure First?

For an interesting entry on this question, see

Data First vs. Structure First

Without further comment.

Tuesday, June 24, 2008

The Hundred Most Used Words in English

The hundred most used words in English, according to The Reading Teachers Book of Lists , Third Edition, are the following:
1. the
2. of
3. and
4. a
5. to
6. in
7. is
8. you
9. that
10. it
11. he
12. was
13. for
14. on
15. are
16. as
17. with
18. his
19. they
20. I

21. at
22. be
23. this
24. have
25. from
26. or
27. one
28. had
29. by
30. word
31. but
32. not
33. what
34. all
35. were
36. we
37. when
38. your
39. can
40. said

41. there
42. use
43. an
44. each
45. which
46. she
47. do
48. how
49. their
50. if
51. will
52. up
53. other
54. about
55. out
56. many
57. then
58. them
59. these
60. so

61. some
62. her
63. would
64. make
65. like
66. him
67. into
68. time
69. has
70. look
71. two
72. more
73. write
74. go
75. see
76. number
77. no
78. way
79. could
80. people

81. my
82. than
83. first
84. water
85. been
86. call
87. who
88. oil
89. its
90. now
91. find
92. long
93. down
94. day
95. did
96. get
97. come
98. made
99. may
100. part

I pulled the list from the following Website:

While these words are all fairly short, it stands to reason that applying a method to shorten them further will result in an improvement in typing speed.

The Website says that the first three hundred words "make up about sixty-five per cent of all written material in English." It also lists the second one hundred, third, fourth ... up to 10th. But shortening the first three hundred words would be the most effective way to increase typing speed.

One easy way to implement this would be to take out the vowels of all the words longer than three characters and add the following substitutions to AutoHotkey:


I will try to learn these abbreviations over the next few days and see whether this works. If it does, I will do the same for the next one hundred.

Monday, June 23, 2008

Auto-Complete Applications - and what is Wrong with them

Last week, I downloaded another brand of "auto-completion" software. I tried others before, but none really appealed to me. Given that I find AutoHotKey's auto-completion feature very useful, I would expect that the kind of software that specializes in completing every word would be even more useful than software that only completes or corrects a limited number of words. But my experience does not support this hypothesis. I believe I now understand why this is so. Let me explain why I think that auto-completion software can never live up to its promise.

AutoHotKey gives you the ability to substitute easily "Best wishes, Your Name" for "bw," or almost anything else for any abbreviation or word you chose. This feature is called "auto-completion." AutoHotKey also allows you to enter a list of words you often misspell or mistype together with the correct spelling, and it will silently correct the wrong spelling. This feature is called "auto-correction. I use both extensively. Using "bw" for "Best wishes," Return, My Name" and "/bm" for inserting an HTML bookmark, etc. has become completely automatic. I don't think about having to enter these abbreviations for word, phrases, and formulae at all. I concentrate on what I want to write, not how to make it show up on screen. The same holds, of course for auto-correction. I have a tendency to type "waht." Auto-correction transforms this into "what." I don't have to worry about that mistake. This feature makes life (or thinking) easier.

In some ways this is just like touch typing or writing words by hand. If you want to write any kind of text, you concentrate on what you want to write not how you write it. When you learned to touch type or form written letters that was different. You had to concentrate on the how, but you exercised for so long that you did not have to think about the how any longer, that is, you exercised until pressing a certain key for a certain letter became automatic. The same holds for many other activities. Riding a bicycle or swimming is very much the same. You usually don't think about how you do it, but just do it. In fact, if you were to try and think about what you are doing, this would actually interfere with the actual doing of it. (This is not to say that you should never try to think about how you do something, i.e. how you actually move your arms in doing the breast stroke. Indeed, if you want to improve you must do so -- but only until you have fixed the problem and until the new habit has become automatic.)

The kind of automation that AutHotKey's auto-completion and auto-correction allows also depends on habituation. You have to concentrate on using "/bm" for a bookmark until it is automatic or has become second nature. But no longer than that. I realize how habitual this use of short cuts has become when AutoHotKey has stopped working for some reason, and "^n" does not start Notetab or "^#d" does not open the Document folder. It takes a second or so to recover from the amazement.

Now, the kind of auto-correction software I was talking about at the beginning, tries to guess every word you type. This is very difficult, as many words start out the same way. Thus "exer" could indicate that you want type "exercise", or "exertion" or "exercises". Since the program cannot know which word you will actually type, it will give you a choice of all three or more, and then make you decide by selecting the correct word, and pressing "enter" or some other key to make it appear on the screen. The program I downloaded has become very good at "guessing" what I want to write during the last few days, but it does not increase my typing speed. In fact, it seems to show it down. The reason is, I think, that I have to think not just about what I want to write, but also about how. I cannot just concentrate on putting words on the screen, but must also concentrate on the little window that suggests words in order to see whether the one I want has appeared already. This is distracting. It breaks my concentration on what I want to say and thus interferes with my thinking, which ultimately slows me down more than typing the few characters that I save in this way.

You might ask why the selection of the word cannot become "automatic" in the same way as typing and substituting a shortcut for a word or a phrase. It cannot, because it takes a decision to select the words. In riding the bike, you decide on which direction to take, or how fast you want to ride, not how you do it. Similarly, in writing I want to concentrate on what and how I say something, not on how I type it. This is why I have given up on universal auto-completion. It complicates things.

Sunday, June 22, 2008

ConnectedText and NoteTab Outlines

It is a mark of a good application that it does not lock you into proprietary data. Your information should always be available to other applications, preferably also in the form of plain text. ConnectedText allows you to do this in at least two ways: (i) you can export all of your topics into separate text files, and (ii) it can save a backup of every topic in a backup folder under the project.
If you would like to back up your project in a different application, you might have to consolidate your backup files in order to import it. I have written an AutoHotkey script that does this:

FileSelectFolder, WhichFolder, C:\Users\*\Documents\ConnectedText\Projects, 0, Select Project to Backup ; Ask

the user to pick a folder.
InPutBox, myFile, File, Select the File name for the Output File
FileSelectFolder, myOutputDir, Choose directory to write to
myDir = %whichFolder%
outputFile = %myOutputDir%\%myFile%

Loop, %myDir%\*.cbk
FileRead, aFileCentents, %A_LoopFileFullPath%
FileAppend, `n@@`n%aFileContents%, %outputFile%

You need to change the starting path to a directory of your own.

You can import the resulting file into a NoteTab Outline File by first opening it in Word, another Text Editor, or Word Processor that allows you to save the file in Windows format, not in UTF *. (If you click "Save as", you will be presented with the option "Windows (Default)".) You also need to insert the following line at the beginning:

= V4 Outline MultiLine TabWidth=30

also replace "##CT##" with "H="

and replace "[[" with "[" and ""]]"" with "]"

When you save the file, use the extension "otl".

The result is a NoteTab Outline file with working hyperlinks.

If you have more than 5400 topics, you are out of luck, however. NoteTab Outlines have a limit of 5400 outline items.

Friday, June 20, 2008

On Michael Ende's "Zettelkasten" - a Pet Peeve

This is a post about the usage of the word "Zettelkasten" in German. So it is of limited interest to those who do not speak German. "Zettelkasten" is the equivalent of "card index," "index card file," "slip box," etc. Literally translated it is a box of small pieces of papers (of uniform size). No one would be tempted to employ these terms to a collection of loose pages, arranged in folders or files. But in German, there is a tendency to do just that. So, people have a tendency to speak of Hegel's Zettelkasten, even though his collection of had nothing to do with "Kästen" or "Zettel."

Frederieke Mayröcker's so-called "Zettelkasten", on the other hand, seems to be nothing more than a messy room full of papers. Even though this does seem to include shoe boxes and smaller pieces of paper, but it's not the kind of thing you would think of when you talk about a "Zettelkasten." A card file does suggest at least a semblance of order. While a Zettelkasten may indeed involve "Unordnung" or disorder, it is one with a "nichtbeliege Struktur" (non-arbitrary structure).

But the misuse of this term that bothers me most is that of a book containing the posthumous notes of Michael Ende (of the Never-ending Story fame). Endes was a thoughtful author children's stories that are also of some interest to adults who have remained children at heart.

Michael Endes Zettelkasten. Skizzen und Notizen (Stuttgart: Weitbrecht Verlag, 1994) has nothing to do with a card index either. It is just a collection of manuscript material: "In seinem Archiv hat Michael Ende über viele Jahrzehnte seiner schriftstellerischen Arbeit hinweg stapelweise beschriebene Zettel, Blätter, aber auch ausgefalleneres Schreibmaterial wie zum Beispiel Theaterkarten und Rechnungen angesammelt. Auf ihnen sind, teils handschriftlich, teils mit Schreibmaschine geschrieben, irgendwelche Romanfänge, einfälle jeglicher Art, einzelne Szenen aus aus Theaterstücken wie auch zahlreiche Entwürfe ... auch Gedanken und Überlegungen ..., die ihn irgendwann einmal beschäftigt oder bewegt haben: selbst längere Abhandlungen über die unterschiedlichsten Themen ...

Some of this is interesting, but it fails to constitute a Zettelkasten in any interesting way. To call it by this name is arbitrary and not very useful either.

Wednesday, June 18, 2008

Otlet's Mundaneum

There is an interesting article about Paul Otlet in the New York Times of June 18, 2008. It includes pictures and a short movie clip.

Otlet's Mundaneum may be be considered as an early attempt to achieve, with index cards and a device called "electric telesope," what the World Wide Web has achieved.

Also interesting in this context is the following entry on edge-nodged cards.

Without further comment.

Bachelard on Classification and "Reasoning Cabinets"

I picked up and read yesterday Gaston Bachelard's The Poetics of Space. It had been on my list of books to read for along time. I had seen many references to it as an interesting book about space in the physical and imaginative sense. Alas, I found it rather disappointing. In fact, it read as if the author had collected literary passages by various writers, psychologists, and philosophers about space on index cards, classified and arranged these cards in roughly ten heaps, called "chapters:" "The House. From Cellar to Garret. The Significance of the Hut;" "House and Universe," "Drawers, chests and Wardrobes," "Nests," "Shells," "Corners," "Miniature," "Intimate Immensity," "The Dialectics of Outside and Inside," and "The Phenomenology of Roundness." It is not that these passages are uninteresting, it's just that they do not add up to anything all that interesting. If Bachelard is right, then this may have to do with the fact that "the poetic image is essentially variational, and not as ... the concept, constitutive" (xix). But I am just as doubtful about his confident claims about "the poetic image" and "the concept," as I am about his so-called "phenomenological" method.

At one point, he comments on the way he ends a chapter, saying: "No doubt it is very rash on the part of a writer to accumulate, in the final pages of a chapter, disconnected ideas, images that only live in a single detail, and convictions, however, sincere, which last only for an instant." And then asks "what else can be done by a phenomenologist who wants to brave teeming imagination ...?" (147). I don't know. But then I don't know about "images that live in a single detail" either. I only know that it isn't just the final pages of the chapter on "corners" that is an accumulation or disconnected ideas and images.

Still, there are interesting ideas, juxtaposed in interesting ways. The one I found most interesting is found in the most interesting heap or stack, the one on "drawers, chests, and wardrobes." It concerns Bergson's prejudice against classificatory concepts. According to Bachelard, Bergson thought that a "philosophy of concepts" is inadequate. "Concepts are drawers in which knowledge may be classified; they are also ready-made garments which do away with the individuality of knowledge that has been experienced." Concepts represent "lifeless thinking" because they represent "classified thinking" (75). Bergson wrote: "Memory ... is not the faculty for classifying recollections in a drawer, or writing them down in a register. Neither register nor drawer exists ..." "Faced with any new object, reason asks ... 'in which of its earlier categories the new object belongs? In which ready-to-open drawer shall we put it? With which ready-made-garment shall we invest it?' Because, of course, a ready-made garment suffices to clothe a poor rationalist ..." (75). Drawers, "cerebrial" or otherwise are the wrong image.

To his credit, Bachelard finds that Bergson's "'drawer' metaphor remains a crude instrument for polemical discussion" (76). It invariably kills imaginative thinking just because it is a metaphor or a "false image," not having "the direct virtue of an image formed in spoken revery" (77), whatever that means.

Bachelard, turning to the next index card, then introduces the story told in Henri Bosco's Monsieur Carre-Benoit, in which, "it is not the intelligence that is a filing cabinet, the filing cabinet is an intelligence" (77). Carre-Benoit has real affection only for his solid oak filing cabinet. "It replaced everything, memory as well as intelligence. In this well-fitted cube there was not an iota of hazyness or shiftiness, Once you had put something in it ... you could find it again ... Forty-eight drawers! Enough to hold an entire well-classified world of positive knowledge. M. Carre-Benoit attributed a sort of magic power to these drawers concerning which he said that they were 'the foundation of the human mind" (77)

Shades of Luhmann, one might think ... but, no, there is no foundation of the human mind in Luhmann -- and positivism is even more of a "no, no." What we need is infinite internal branching ability! Without it, no serendipitous discoveries. But, if Bachelard is right, Bosco "has succeeded in embodying the dull administrative spirit" (77). And wasn't Luhmann an admistrator first? Doesn't this critique of "Schubladenken" concern die "Zettelkastenmethode" at least indirectly? Perhaps, but we must return to Bachelard ...

... who continues the Bosco story and reports that the maid abused "the foundation of the human mind" to store mustard, peas and lentils in it: the "reasoning cabinet had become a larder," which leads him to the sagacious but trite conclusion that "many erudite minds ... lay in provisions." End of story. Or shoud I say index card?

So, now you know why Bachelard collected all the passages about spaces. I just wish he had thought more about their contents and not simply used them for "variation."

Disclaimer: I don't know whether Bachelard actually used cards or "fiches." But the book reads like it.

Tuesday, June 17, 2008

Notebook: 2003 - 2005

Late in 2002 I discovered Wikis. This was perhaps a little late, but they really caught on in a big way only during the early 2000s. The Wiki Way was only published in 2001, and I was never really interested in the collaboration aspect of wikis. Rather, I liked the idea that I could have a desktop wiki for my notes. Nor did I like the idea of it being on a server (and possibly exposing my notes to the whole wide world). Wikit makes hyperlinking easy, and it does not require a server. So I first began to use it. Soon after (in May) I discovered an application called Notebook by Will Duquette, called Notebook. It must still have been version 0.9, as version 1 was released only in June of 2003. It offered everything that Wikit could do and then some.

"Notebook is a personal notebook application." It was ideal for note-taking. Even though the search function was anemic, not really allowing for Boolean searches, and even though it could not print (at least not without invoking some other applications), I loved it. But it was programmable. In fact, I taught myself the rudiments of Tcl/Tk to do some programming of my own. Notebook is available for Windows, Unix, and the Mac. the "Starkit" version is the easiest to use. It became a most interesting project for me, and it seemed for a while the solution to "the problem with index cards."

Notebook stores its pages in a flat database without any page hierarchy. The database is actually one large textfile that is loaded into memory when the program runs. This works very well. The program is fast and very reliable. And when the time came to move things from Notebook to ConnectedText, it was easy to transform Notebook's markup to that of ConnectedText with a capable text editor.

There is a very good description of Notebook in the article What is a Wiki? by Nathan Matthias. (While you are at it, you might also look at Caffeinate Your Hypertext, which includes an interesting discussion of the relation of hypertext and outlines.)

Notebook is freeware, and it is still being developed, albeit slowly. I am fond of it, but I have moved on to a much more capable application, and it now resides -- you guessed it -- on the flash drive that also contains other "old applications." What made me move to ConnectedText was not just the much more capable search function, but also the active and concerned development team. It reminds me of the early days of Ecco. And I hope that it won't be bought out ... though I am sure I would even then use it for a long time, as there is nothing that even compares - at least in my book. See also ConnectedText 3.

Agenda - an influential program I never used seriously

There is one program that greatly influenced the way I think about personal information management and note-taking but that I never really used, and that is Agenda. It clearly also influenced Ecco and InfoHandler. Ecco's folders and views were clearly an attempt to implement some features of Agenda's automatic categories. And InfoHandler's "unique concept of categorizing information" has a similar precedent. Its master/slave categories replicate the hierarchy of categories at least to some extent. InfoHandler can even import Agenda files, and thus provides an obvious path from Agenda to a Windows application.

The Projects View:

A Note:

Another application that learned a lot from Agenda is Zoot. In fact, Jim Fallows, who wrote an article on Agenda in the Atlantic that is, at least in part, responsible for the continuing popularity of Agenda, also wrote an article about Zoot. (See the file section of the Zoot Forum. But, since Zoot is another application I never seriously employed, I will not say any more about it here. In any case, one might say that Agenda has influenced every serious contender in the note-taking application business since the early 1990s. While it is said the application "flopped," the latest release (2.0b) of 1992 is still being used by some.

Agenda is a freeform database program that is characterized by "items," "categories," and "views," where items are headings with freeform notes attached to them. Categories are used to classify different items into groups, and views are grids that allow you to see the items in different contexts. Category membership formed the basis for retrieving information. Since these categories could be arranged hierarchically as parents and children, any item that was assigned to a child category also became a member of the parent category. The notion of a "smart folder" was born. Thus "a 'When" category was included automatically so that if a date was embedded within the text, it was interpreted as a date and an assignment was made. For example, the item 'See Wendy on Tuesday 3pm' was automatically assigned to the following Tuesday at 3pm. If a category 'Wendy' had also been created then an assignment could also have been made as well." (See Lotus Agenda.)

The notion of dynamically assigning categories that allow one to sort the information automatically as it were is very appealing. Some people claim that the employment of categories as a fundamental construct for organizing and retrieving information is more powerful than and inherently superior to a hypertext-based model of information management, like that of a personal wiki. But I would argue that this is misleading, because there seems to me not the slightest contradiction or tension between organizing information by direct hypertextual links AND by way of categories (arranged hierarchically or non-hierarchically). It is possible to do both at the same time, and all of the better Wiki-application allow you to do both.

In ConnectedText, for instance, the basic structure of "items" (called "topics"), "categories" (called "categories"), and "views" (called "smart topics" and "category" topics) is preserved. While it does not automatically assign "See Wendy on Tuesday 3pm" to a date topic, there is nothing that prevents one from writing a script that would do this. In any case, if there is a smart topic that collects items containing "Wendy," it would do this even without a script.

In any case, one might say that Agenda formed an interesting step on the way toward my own solution of "the card file problem," even though all I ever did with it was to "play with it" - and, yes, it does have a place on the USB drive with the other "old applications."

Monday, June 16, 2008

InfoHandler from 1999 to 2003

Because I began to realize during the late 1990s that a hierarchical organisation is not the most effective means for keeping one's notes, I began looking for an alternative. I found it in InfoHandler.

InfoHandler presents itself as the solution to "The Card File Problem," asking you to "[t]hink of a card file box. In this box you have several thousand[s of] cards with evaluations of articles in periodicals. Every card with a header line - author, title, and reference - and below that extensive notes. All cards are ordered accurately by subject. If you are looking for something for a given subject, you only need to find an appropriate index and view all cards belonging to this topic. -- So far so good - but now you are looking for something that was published for a certain subject in a certain magazine. You scan the few hundred cards for the subject and select those with the wanted periodical. Because you are uncertain about the subject[,] you search other subjects as well. Now you see you got far to[o] many cards, you couldn’t examine them all. You guess that really you are only interested in certain products together with this subject. Unfortunately, after deciding to order by subject, you couldn’t have ordered the cards by products too. So now you begin to search the collected cards for the wanted product. Don't forget to return all the cards that you have removed to the right places in the box when you have completed your task. -- Libraries ... solve this dilemma by having two card indexes with duplicated cards, one ordered [by] authors, the other [by] subjects. -- Another problem [arises] if, [in] preparing a new card, you find that several of the categories seem to fit and you cannot decide where to put the card - unfortunately, a card cannot be simultaneously at several places."

The solution: the developers of InfoHandler "conceived [of] the unique concept of categorizing information that [they] implemented with InfoHandler. This is much more than just attaching 'keywords' to the info. Most important is the centralized management of the Categories and their organization in groups. Categories are not just 'words' or 'character strings' but 'objects' with their own properties and functionality. Shortly, with the help of these grouped Categories , a kind of multi-dimensional view on the data is accomplished, as opposed to the one-dimensional view of the common card indexes or with the hierarchical organization that is offered by most of the available tools for personal information management (PIMs)."

This approach sounded good to me; and it is a good solution to the problems that card files present. InfoHandler also provides the means of connecting or linking cards directly. Although its method was cumbersome, it was no more so than that of any of the other programs I knew. It was the discovery of wiki technology that opened my ways to much better ways of doing this.

But it was not just wiki technology that made me abandon InfoHandler. Its way of handling rich text and copying of Html was not consistent. something that looked good in one version of the program, looked entirely different in the next. Sometimes, cards seemed to have no information (though I found out later - after having deleted the cards - that the font had changed for no reason apparent to me to the color silver. In other words, it was there -- it just could not be seen. It also was not "simple," but it's busy look (see screenshot) began to grate on me more and more. And the development of the program took a turn that "turned off" many people, including me. So, I did not renew my license. Version 8 was the last I bought. It is now at version 10, sort of.

In spite of these criticisms, I must say that InfoHandler is a very capable program, with a powerful search feature that allows complex (Boolean) queries on your data, using the categories you have assigned.

To say it again, InfoHandler is a good solution of "the card file problem," if only because it is based on a powerful database application. But it is not the best solution of this problem because it depends too rigidly on categories and filtering. For the longest time, there was not even the ability to search and list the results of such a search. More, even redundant, capabilities of searching, listing, filtering, and jumping from one card to another would, in my humble opinion, have improved the product more than moving it to the Microsoft.NET Framework.

On the other hand, it is much more capable and still promising than the otherwise very interesting Freeware Zettelkasten application by Luedecke. For me, it was an interesting step on the way to ConnecteText. It is no longer a live option.

Ecco from 1993 to 2000

In a Journal entry of Tuesday, August 10, 1993, I find: "I also bought a copy of ECCO, and familiarized myself with it over the weekend. Imported the addresses and the bibliography file from InfoSelect. Though I had gotten quite used to Ascend, the program has some things that I really did not like. It is not very flexible. You really cannot use it very well for long-range planning. Ecco is more flexible, but perhaps it is too flexible. It does not give me the kind of structure that I really need. However, as an outliner and a research tool, it is really great. I'll try it out for scheduling over the next few weeks."

Well, I used it for several years, at the beginning parallel to InfoSelect, but increasingly as the only application for note-taking. It implemented outlining well. In fact, it may be the best outlining application ever developed.

But, not all my data were in Ecco. I also used paper Notebooks from this time forward, indexing them in Ecco. Ultimately, however, I gave up on it for two reasons: (i) I realized that the outlining metaphor was not the best for keeping a database of information or notes, and (ii) Arabesque, the company that had published Ecco, was bought up by NetManage about a year after it was released and I had bought it. While the original owners were very responsive to their users, NetManage seemed to have no interest in them whatsoever. Soon, the product became "abandon ware." For more on its features and its history, see Ecco Pro

Even though I have installed the last version of Ecco on every computer I ever owned, I did not become one of the die hard accolites of the program. The main reason has to do with my realization that the outlining metaphor is inherently limited when it comes to note-taking. If Luhmann was right about anything, then he was right in his claim that a hierarchical structuring of one's notes gets in the way of future intellectual growth, which has to do, at least to some extent, with extending our concepts "as in spinning a thread," twisting "fibre on fibre. And the strength of the thread does not reside in the fact that some one fibre runs through its whole length, but in the overlapping of many fibres" (Wittgenstein). A network of notes serves me better than an outline.

Put differently, my original fear was groundless. Ecco was not too flexible. It turned out that it was not flexible enough, though I sometimes wonder what would have become of Ecco (and my note-taking habits), if the original developers would have continued developing it.

I have transferred all the notes I ever had in Ecco to Connectedtext. I used it occasionally in outlining papers and projects, but the outline function in ConnectedText has made this largely superfluous. So, I don't know whether it will find a place on the hard drive of my next computer or will become another "toy" on a flash drive.

InfoSelect for Windows in 1992

InfoSelect, which I bought late in 1992 as a replacement for the combination of "Organize!" and "Scraps," was a big improvement over both. You could use it like a big shoe box, into which you threw everything you thought worth keeping -- and that is how I used it. The good thing was that it allowed (and allows) you easily to find everything again. It allowed Boolean searches, using AND and OR. In fact, it even had something called "neural search," which allowed you to enter multiple words and would then search the "stack" or the active database for entries that contained "as many of these words as possible."

It had some problems that seemed to have more to do with the limitations of Windows. Stacks had what I perceived as a severe limitation of size. Umlauts were always a problem. But on the whole it incorporated "the power of the simple tool." It was amazing to see the little windows neatly arranging themselves after a completed search.

I still have a copy of this program on a Compact Flash chip, and I often think I should use it more, but it has become just a toy. I did buy the versions that followed, up to number 4, but I increasingly felt that InfoSelect had lost their way, trying to be everything to everyone. I also found the implementation of the outline for managing information klutzy - almost a betrayal of the original vision of Tornado notes (its predecessor) and InfoSelect for Windows 1. While the search feature improved significantly, the price of updating to get features that for the most part I did not really want proved to be too much. So, version 4 was the last I bought (in 1997).

I sometimes wish that they would come out with InfoSelect Standard or InfoSelect Classic that does away with most of the functions developed later (at a reduced price), concentrates on easy data entry and search and re-implements the old interface. (You can, of course, have the little Windows in the new version as well, but it isn't as convenient.) I am sure that such a version would not make me abandon ConnectedText, but it might complement it, as I still like the "shoe box" way of organizing, if "organizing" is the right way for this approach. In any case, it would only be the first step.

Sunday, June 15, 2008

Keeping Notes around 1991

Another program that I used to keep notes together with Organize! in 1991 and 1992 was a quirky database program named Scraps, written by Raymond Lowe.

Viewing all scraps:

Looking at one scrap:

Scraps billed itself as a "freeform data base" that "relieves you of the straight-jacket imposed by most data base programs but, at the same time, .. takes away a lot of the structure and safeguards with which you may be familiar." The usage tips are still somewhat relevant:

"Keep retrieval in mind: The real importance of the stored information is not the storing but the retrieving of it. That being the case it is worth always considering how you might want to look something up when you enter it.
Use standard words: Sprinkle your scraps liberally with standard keywords, pick those words yourself so they make sense to you. For example you might choose to put the word "PHONE" in every scrap that contains a telephone number, and "ADDRESS" in ones that holds an address. This ensures that you will not fail to find something because you used a term in the scrap different to that which you are using to search for it.
Throw in keywords: When several different words can be used to describe a key concept put all of them into the scrap. In a scrap about cheap airline tickets type "AIR PLANE AIRPLANE CHEAP ECONOMY TICKETS FARES" across the bottom of the scrap. That way you can be sure that you'll find the scrap when you need it, regardless of how you ask for it.
Use project keywords: If you have a whole bunch of scraps revolving around a single task or project try putting a standard code word in each. Every scrap about the West District Housing Project might have "WDHP" or "WESTHOUSING" in it. That way you can quickly find all the those scraps by searching for this key word. By using an invented word or acronym you can be sure that no other scrap will accidentally happen to include the word."

The most important function was a fairly powerful search capability, in which the Boolean operators of AND and OR could be used. It worked best with many small items. Even though each "scrap" could be as long as memory allowed, memory was scarce in those days. More importantly, however, breaking down large documents into chunks has virtues of its own, and I still prefer smaller notes over long ones: between 250 and 800 words seem to work best for me.

Scraps could also do reminders and alarms, but I primarily used it for keeping scraps or snippets of information.

Being Organized in 1990

I used for some time around 1990 a quirky, but useful program, called "Organize!" It still works, sort of:

Its author was Kwek Sing Cher. The manual was written by Jim Murray. Here is a quote from the Introduction, which, I think, remains as true as ever:

"The Power of the Simple Tool: There are a dozen or so Personal Information Managers (PIMs) on the market. Many are quite expensive. All are useful, to some degree or another. Why should you choose this one? -- Consider the typewriter and the pencil. The typewriter, at first glance, seems the more powerful. It's certainly more complex. For some uses (business correspondence, novels) it is indispensable. It is also cumbersome, expensive, and demanding to operate. -- The pencil, on the other hand, is a simple tool. It works on almost any solid surface. It's convenient; you can carry a dozen of them in your pocket. It's flexible, useful for everything from shopping lists to artwork. Chances are, you have already used it many more times than you will ever use a typewriter. -- In point of fact, the pencil -- like many simple tools -- is the one that's really powerful."

I have always liked simple tools. But simplicity is a moving target.

Organize! worked just like a calendar, in which you made notes that could then be searched by means of keywords like the following:

1.DO: (include the colon)
3.APP*: (notice the difference the asterisk makes)
4.02 APP:
5.02 APP*:
6.WEEK #

What you needed then was: An IBM AT, IBM PS/2, or compatible; EGA or VGA/SuperVGA monitor. The minimum recommended was a 12Mhz 286 PC, and a mouse, for easier use.

The software needed was: MS/PC-DOS 3.0 or later, and Windows 3.0 or later.

The computers of today allow for a different kind of simplicity, one that I could not have dreamt of in 1990.

Saturday, June 14, 2008

Wittgenstein "on" Note-taking

According to Aristotle, definition is a two-step process. To define any kind of thing, you must first establish what this thing or phenomenon has in common with other things, that is, you have to determine to which class of things it belongs. Secondly, you must determine what makes this thing different from others in the same class. This is also known as definition by genus and difference. Thus, he famously claimed that "man" belongs to the class of living things or "animals", but is different from other animals in being rational. This also has to do with the "essence" or nature of thing.

Ludwig Wittgestein suggested that this way of defining or classifying things does not always work . In fact, it fails in some very interesting cases. Thus language is a rather complex phenomenon, or perhaps a complex of phenomena that are rather disparate from one another and cannot be properly understood, by simply relying on the traditional Aristotelian notion of definition. "Instead of producing something common to all that we call language, I am saying that these phenomena have no one thing in common which makes us use the same word for all, - but that they are related to one another in many different ways. And it is because of this relationship, or these relationships, that we call them all 'language'".

He further suggested that "language" was similar to "games" in this regard. There are many different kinds of games, ranging from board-games, card-games, ball-games, games people play, to electronic games, which he claimed have nothing in common, except similarities and relationships. In fact, what we see is "a complicated network of similarities overlapping and criss-crossing," not one property they all share. Therefore, he could not conceive of a "better expression to characterize these similarities than 'family resemblances'; for the various resemblances between members of a family: build, features, color of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way." Furthermore: "'games' form a family. And for instance the kinds of number form a family in the same way." Often, when we give a name to things, we do so in order to point out that it has "an indirect relationship to other things we call the same name." In this way, we extend our concepts "as in spinning a thread we twist fibre on fibre. And the strength of the thread does not reside in the fact that some one fibre runs through its whole length, but in the overlapping of many fibres". (See Ludwig Wittgenstein, Philosophical Investigations (65, 66, and 67).

Leaving open the question whether this is true of language, it certainly describes the way he investigated the philosophical phenomena that interested him.

I also think that it is highly suggestive for the methodology of note-taking. To try and force one's notes into a rigid hierarchy of categories and definition too early can be counter productive. It will stand in the way of seeing indirect relationships to other things. We should develop the concepts and theories based on the notes "as in spinning a thread we twist fibre on fibre." Furthermore, the strength of the thread" will not not necessarily "reside in the fact that some one fibre runs through its whole length, but in the overlapping of many fibres". This is one of the reason why I like the wiki approach to note-taking as exemplified to my mind best in Connectedtext.

I still think that in the end, that is in publication (which is the end of all ends, isn't it?) "a complicated network of similarities overlapping and criss-crossing" is not enough. You need to present a thesis or two and defend them, as Wittgenstein ultimately did too.

Sunday, June 8, 2008

Pile of Index Cards

An interesting site on how to use index cards:

Pile of Index Cards

see also: Pictures

Without further comment.

ConnectedText: Scrivener for Windows?

There is an interesting program available for OS X on the Mac. It's called "Scrivener," which is described as a "word processor for writers." Many people have asked for a Windows version of this application, but, as the developer has made clear, a Windows version of this excellent program will not be forthcoming because OS X is internally so different from Windows that it would require programming it from scratch.

Naturally, people have searched for windows programs that come close. The ones mentioned most often are:

Page Four
Liquid Story Binder

I tried all three, but none of them really appeals to me. There is, however, an application that I have used for a long time and that I have recently realized can be used very much like Scrivener. This is ConnectedText. Let me explain.

Scrivener is a program that is intended to integrate the process of research, outlining, storyboarding and writing. ConnectedText is much better as a repository for research, has an outliner that is almost as capable as Scrivener's, and it allows you to write distraction free, in the same way as Scrivener. While it does not have a "storyboard," it does have a Navigator that can fulfill a similar function. It also allows you to write in full screen mode. In addition, it has something that Scrivener lacks, namely the ability quickly to link entries by using wiki links.

All you have to do is to save a desktop of the sort that I have illustrated in the following screen shots. This desktop for writing has two panes: one for writing, and one for navigating, containing views for the Outline, the Navigator and the Table of Contents (for the topic on which you are working). This is the screen with the outliner and a topic in edit mode:

This is the screen with the navigator and the topic in view mode:

Scrivener is thought to be the word processor for the "non-linear" writer, i.e. those who do not compose longer pieces by starting at the "beginning" and ending at the "end," but those who compose different parts of the piece and then later worry about where they fit. These are also writers who do not make a clear break between a "research" and a "writing phase" in their work, but do both in tandem. ConnectedText approaches researching and writing in the same way. This should be equally appealing to those who have to write academic texts as to those that write non-fiction.

Neither application is meant to prepare the final copy of the text. Where Scrivener uses "restructured text," ConnectedText uses a wiki markup that is close to that of MediaWiki's and can easily be transformed into html. There is a rudimentary macro on the Website that allows transforming the markup into rtf as well.

One area where ConnectedText really shines is version control. Every revision of every topic or note you write is kept. You can, accordingly, measure the "progress" you have made in developing your ideas, paragraphs, or papers at any time.

I know no better alternative to Scrivener on windows than ConnectedText.

Wednesday, June 4, 2008

Thinkmap Visual Thesaurus

I find a thesaurus indispensable in writing. My favorite is

Thinkmap Visual Thesaurus

I don't use the online version, but bought the desktop application.


There is an interesting Web application that allows you to create a mind map from indented text outlines and download it as a jpg file that you can then use in an application that allows you to embed pictures, like ConnectedText, for instance.

It makes attractive mind maps in the style of Visual Thesaurus.

Mind map

Just don't use long (or too many) headings. I think it is useful for visualizing information.