The Piggydb Way: #2 Tags as First-Class Components

Most of you who happen to stumble upon are sure to know what you want, I imagine. You are looking for a tool (in the categories like wiki, outliner, personal database, etc) that helps you organize your information in a more effective way as an alternative to, for example, Evernote or Springpad. Then suddenly, I began to talk about the ‘reversal process’ instead of explaining the basic organizing functionality. You might be confused or think you didn’t need it because you knew what kind of information you were going to organize and how to organize it, and you just needed a tool to support it. Well, okay, but is that really the case? Are you really sure you know what you want?

When you are about to collect information and organize it, quite naturally, you know what kind of information you are going to organize, and anticipate the result. It may be information about your project, daily routines, a travel plan, or it may be a diary. You have already prepared the container in which you are going to just place pieces of information one by one. It would be quite straightforward, you think. However, once you begin it, something in you will be stimulated by what you are doing, and sooner or later, you will unconsciously step into the area of Knowledge Creation even with a most primitive container you use, such as a piece of paper.

Let’s suppose that you keep a record of things you have done in your daily life in a plain old paper notebook. You do so because you want to use these information later on when you need to know what you did and when you did it. And then, one day, after a while since it became a daily routine, you reflect on the record. In that reflection, you happen to find an interesting pattern in the log. You underline these places with a red pen and write an explanation of the concept common to them in your own way. This discovery is totally what you didn’t expect when you started this habit. I’m sure many of you have experienced something like this before if you have a note-taking habit. Let’s call this kind of practice ‘Weighting Information‘ (selecting a certain part of information and putting some meaning into it). So you acquired a new point of view as a result of weighting information.

Information weighting appears whenever you deal with information. For example, if you are reading this article on a web browser (I believe most of you are), it must have a bookmarking function. Bookmark allows you to save addresses of websites that you want to revisit later, and it can be regarded as a way of information weighting. You select a small bit of the vast sea of the Web and attach the concept of ‘bookmark’ to it.

As you may have noticed in the example of recording things done, there are two steps in information weighting. The first one is the discovery of an interesting commonality across random things and the second one is putting a label on it. In this process, the mechanism of the discovery is especially interesting and mysterious. One of the most interesting things about it is that you can’t expect the outcome in advance, in other words, you can’t plan what kind of thing you are going to discover.

The other day, I came across an interesting column on innovation titled “The Idea Idea” by Peter J. Denning ( Denning brought skepticism to the popular belief that an innovation is the result of adopting a good idea. He proposed a hypothesis that “practices rather than ideas are the main source of innovation” and “many ideas are therefore afterthoughts to explain innovations that have already happened“. The term ‘idea‘ here can be replaced by ‘concept‘ that we are discussing about. In the above example of information weighting, the concept discovery was not expected or planned but happened in the course of the practice of recording things done in daily life.

Stephen King, a world-renowned novelist, once wrote about his style of writing novels. He wrote in his memoir “On Writing” as follows: “I distrust plot for two reasons: first, because our lives are largely plotless, even when you add in all our reasonable precautions and careful planning; and second, because I believe plotting and the spontaneity of real creation aren’t compatible.” The ‘plot‘ mentioned in the quote would correspond to the ‘idea‘ in Denning’s column. As Denning said practices are the main source of innovations and ideas are afterthoughts to explain innovations that have already happened, King said stories pretty much make themselves in the course of writing and progress to theme. And King’s figurative expression for his creative process is really impressive: “stories are found things, like fossils in the ground“. Although I don’t know of many novelists, I imagine there are not a few novelists who take this type of approach. Haruki Murakami, who reportedly leads race for Nobel prize for literature as of writing this article (the winner turned out to be Mo Yan who happens to have the same family name as my wife), is one of them. This spontaneity would be essential for creative process and I think that human beings instinctively know it because, as King wrote, “stories are relics, part of an undiscovered preexisting world

After discovering a concept, you need to put a label on it so that it can be a part of your knowledge. And through this label, you can view the world around you in a fresh-new viewpoint. This kind of concept-oriented knowledge building is the main focus of the Knowledge Creation in Piggydb.

The most important feature of Piggydb in terms of concept-oriented knowledge building is Tag-Fragments. A tag-fragment, which is the kind of a knowledge fragment whose ‘as-a-tag’ attribute is enabled, can be used to represent a concept that you have discovered as in the above episode or selected as a theme in advance for your study or investigation. Originally, classic tags have played an assistant role in the Web 2.0 systems providing a lightweight way of organizing information. And, of course, you can use these tags to represent found concepts other than familiar categories. However, this means just grouping elements by concepts, which does not contain any structure for the concepts.

As you might know already, the labels of concepts themselves do not solely hold the value of your knowledge. They are just labels. The value resides in the context behind concepts, the context in which you have come to discover and build the concepts (the Internet is filled with fruitless communication triggered by responding to keywords without considering their contexts). Therefore, the important factor deciding the value of your concepts is how you structure this context information, and that’s why Piggydb introduces Tag-Fragment that supports two-layer structure allowing you to evolve found concepts into more rich and structured knowledge. While classic tags is no more than collective keywords for indexing, Piggydb’s tags can be treated as the same as the first-class information components (knowledge fragments) in a database, which means they can have their description and relationships to other components.

[To be continued]


The Piggydb Way: #1 Tag as Concept over Tag as Index

In this series of articles ‘The Piggydb Way’, I will try to explain why Piggydb is so unique and useful in terms of knowledge creation compared to other information management systems and how it will change the way you organize information. I will do my best by squeezing my limited English skills to convey the whole notion of what Piggydb is all about. Your feedback is always welcome.


I guess many of you regard or use Piggydb as an organizable notebook. But why do you organize your information in the first place? In most cases, its purpose would be to make it easy to find a piece of information you need later on.

There are countless applications and services that provide tools to do these things: folders, trees, tags, hyperlinks, etc. Compared to these applications, Piggydb has powerful tools: Fragment Relationships and Hierarchical Tags. But here in this series of articles, I’d like to explain that Piggydb aims not only to provide such tools for organizing information, but also to seek further value by providing a way to build your own ‘concept’ maps.

The most important feature of Piggydb is ‘tags’. However, you may not need them if you use Piggydb as an organizable notebook. Actually when I used it in the recent work to manage the project related information, I did not use tags at all except the system tags (#home, #bookmark, etc). Just connecting fragments was sufficient to manage the information (multi-parenting perfectly worked to organize the complex information). Heavy users may have realized that the roles of fragment relationships and tags are overlapped because both have similar functionalities in terms of ‘grouping’ fragments. So if you just need to organize information on your daily life or work, it would be sufficient to use only either of them.

Well then, why am I saying tags are especially important in Piggydb?

Tagging has become one of the primary ways people organize information since Web 2.0 services adopted it as their important features. Tagging was introduced as a more lightweight alternative to the existing classification systems such as hierarchies. Hierarchical classification systems manage categories by single-parenting tree structure and their vocabularies should be controlled carefully in order to work properly (controlled vocabulary). Although tagging was adopted as an effective alternative to hierarchies, there is a trade-off between these two systems. Tagging is certainly more flexible and easier to use in the way you can attach multiple labels you come up with to a piece of information while hierarchies force you to select a single category following the existing hierarchy. But what will happen when the amount and varieties of information increases?

Japanese economist Yukio Noguchi argued in his best-selling book published in early 1990s that any attempt to classify information to make it searchable is useless in the first place. He pointed out the problems in classification. One of them is the ‘Bat Problem’ which arises when classifying information and goods. Material things and information can have multiple attributes that are used respectively depending on the context (e.g. Bats have the properties of both birds and beasts). He also referred to Wittgenstein’s ‘Family Resemblance‘ principle, which states that “things which may be thought to be connected by one essential common feature may in fact be connected by a series of overlapping similarities, where no one feature is common to all (from Wikipedia)”. This idea also shows the limits of the traditional taxonomy (known now as ‘monothetic’).

The most difficult thing in classification is to maintain its consistency in the growth of the database. Not only will your database grow, but you yourself will also learn and then need to change your classification system. Noguchi explained this by introducing the ‘Theorem of the Ugly Duckling‘, which was proposed and proved mathematically by Japanese theoretical physicist Satoshi Watanabe. Here is the quote about this theorem from the book “Apoha: Buddhist Nominalism and Human Cognition“:

From the formal point of view there exists no such thing as a class of similar objects in the world, insofar as all predicates (of the same dimension) have the same importance. Conversely, if we acknowledge the empirical existence of classes of similar objects, it means that we are attaching non-uniform importance to various predicates, and that this weighting has an extralogical origin. When we employ a concept, we usually understand that there is a group of objects corresponding to this concept that any two members of the group resemble each other more than a member and a nonmember. Two sparrows are very much alike, while a sparrow and a rose are not alike. It is natural to translate the term “to resemble” as “to share many predicates in common.” But this interpretation can be shown to lead to a denial of the existence of a class of similar objects by the following theorem, which I have dubbed the theorem of the ugly duckling. The reader will soon understand the reason for referring to the story of Hans Christian Anderson, because this theorem, combined with the foregoing interpretation, would lead to the conclusion that an ugly duckling and a swan are just as similar to each other as are two swans. (Watanabe 1969, 376)

Watanabe thought that any classification is arbitrary and inevitably biased by our cultural background.

With these problems in mind, you should realize that hierarchical classification turns to be only a representation of a limited perspective and how fragile it is for ever-changing knowledge management. Even tagging, which is more flexible and solves some of the problems such as the Bat Problem, cannot escape from the problem of maintaining consistency. Rather, in tagging where users can freely choose words, classification can more easily lead to the lack of control than hierarchies, for example, the resulting vocabulary will be filled with homonyms (the same tags used with different meanings) and synonyms (multiple tags for the same concept). To begin with, tagging has been popularized in the context of Folksonomy, in which searchability or comprehensiveness is not the major purpose of classification, but it aims at promoting communication among users triggered by matching tags for finding more valuable information efficiently.

I have explained how fruitless it would be to classify information for searchability so far, but it might not even need those stiff theories. You have probably experienced a situation where you had been organizing your classification with great effort, but you ended up not using most of it, and what you really needed was just a simple keyword search, haven’t you?

Still, I would be certain that there are many people who insist that it works, at least, for them, and actually there are such cases, which I call ‘Stable Classification Schema Case‘. The demo site Table Tennis Videos is a good example for this. But I won’t discuss this case further because it is not directly related to Knowledge Creation I want to focus on here.

So finally, after a long digression, I’m returning to the main point: why am I saying, in Piggydb, which does not have any features related to Folksonomy, the most important feature is tags?

The theorem of the ugly duckling proved that any classification is arbitrary, and this arbitrariness makes it difficult to maintain the consistency of classification, therefore, classification is quite limited as a way to build an index for information. However, if we change our point of view a little bit, on the other side of arbitrariness, couldn’t it be said that classification is creative in terms of expressing one’s thoughts, ideas, or lessons? As you know from tag clouds, your tags represent the essence of your database or lessons learned in the course of growing your database. In a traditional way of information management, the essence comes before building the content of a database, like defining a classification hierarchy as a container of information. What will happen if you reverse this process? You don’t create tags as a mere index for information, but as a reflection of your growing database and make them the central part of your knowledge.

I’m going to explain this reversal process more in depth in this series of articles, but before diving into it, in order to distinguish between these two tag usages, let’s call them respectively ‘Tag as Index‘, where tags are created for searchability, and ‘Tag as Concept‘, where tags are created to represent the important concepts learned in the growth of your database.

Next: The Piggydb Way: #2 Tags as First-Class Components