Semantic cartography to support conceptual development through blogs

Blogs recording the academic process

For a while now I have been using Warwick Blogs both for developing academic research and for recording my activities as an E-learning Advisor. It works well for both of these purposes. It is most effective in how it allows me to very quickly record an event. Such events may be meetings, lectures, seminars, readings (of a book chapter), or even just sessions of thinking. Theses are recorded sequentially as they occurr, within the datetime ordered structure of the blog. And I can easily jump back to a point in time to see the recording of events from that time.

Real-time engagements

We could say that these recordings are recordings of real-time engagements. Such real-time engagements are always engagements with more than one distinct but related concept. For example, in philosophy I may have a seminar that deals with modularity of mind and with artistic confinement, as two explicitly stated and applied concepts. The resulting blog entry should record how these two distinct concepts have operated together, or at least how I moved from one concept to another. These concepts are then the explicit subjects of the meeting and the resulting blog entry. Our expectation has been that the recording of real-time engagements would also result in the recording, as implicit concepts, of other aspects present in the engagement. For example, I may wish to identify that I used a specific skill. That skill and my use of it becomes a concept when I start to define it's nature, scope, and my capabilities. Other such implicit concepts may concern the intentions, agendas, perspectives, and subsequent and consequent trajectories of the engagement.

In this way, real-time engagements always consist of complex assemblages, formed from both explicit and implicit concepts, and reflectively recorded as involving both explicit and implicit concepts. These may be represented in a more or less effective way in a blog entry about the engagement.

One of the reasons for developing the Warwick Blogs system was to support the development of these concepts, both explicit and implicit, over an extended period of time. We know that concepts only get defined, skills only get developed, when they are applied to each other, in relation to each other, in real-time engagements. This is how the research process actually works. To support this development we have two mechanisms that seek to link together real-time engagements and the concepts that appear within them: categorisation and follow-up entries.

Categories as developing concepts

The most obvious mechanism is the category. By placing a series of entries in a single category the blogger indicates that they develop the concept represented by, or held within, that category. For example, there may be a category called PDP, and the concept that it represents is the quite nebulous concept of Personal Development Process.

It is true that, as entries are created in a category, over time they tend to define and develop the concept behind that category. But these categories and the concepts that they represent are broad. PDP is a big and potentially all-encompassing idea. To achieve a greater degree of granularity, one could define many different categories to represent all of the concepts that will be developed in a blog. This could ,ake for a very complex top-level view of the blog. We would also lose an important aspect of categories, a role that they perform at the top level view of the blog, in that they provide a means for dividing up the various uses of a single blog (social, academic, etc). We could use sub-categories to deal with this, but again categorisation gets quite complex.

There is, however, a more serious limitation with the method of categorisation. We know that blog entries usually report on real-time engagements. These engagements always include more than one concept, implicitly or explicitly. In the case of PDP, the engagements are likeley to be explicitly related to a subject-specific concept, and only implicitly related to PDP issues. If categorisation were to be our mechanism for identifyng the concepts dealt with by an entry, the entry would have to exist in more than one category. We could allow a single entry to appear in more than one category, but that would be both complex and unintuitive. Consider this scenario. I go to a seminar on "the transcendental deduction". The seminar deals quite intensively with Kant, but even so I notice during the seminar that I am using my notes in a different and more effective manner in response to some PDP work that I have been doing. I get home and just about have time to write a single entry to record the engagement. This entry mostly deals with Kant, but with a brief mention of the use of my new skill. Do I then classify the entry under both PDP and Research? Would my PDP category get cluttered up with academic entries that have some hard to find mention of PDP? But if I didn't in some way identify that the entry contains something of relevance to PDP, the important event of an advancement in my PDP agenda would just be lost within the complex body of entries within my Research category.

Thus categorisation is not sufficient.

Follow-up entries extracting and refining concepts

The second mechanism that we have for developing concepts in blogs is the 'follow-up' entry. This is designed to deal with the limitations of categorisation. It works as follows: I write an entry recording an engagement, I publish the entry, at a later date I re-read the entry and have some further thoughts (or possibly I have a related engagement), I use the 'write follow-up' link on the entry to write a new entry linked back to the first entry.

We had hoped that bloggers would use this to explicitly develop a concept that had been only implicitly recorded in an entry about a more complex engagement. For example, I may have a seminar that involves, amongst others, an implicit PDP related concept. I would intuitively, and due to time-constraints, record that in an entry that deals with all of the concepts involved in the engagement. In order to isolate an develop the implicit PDP concept, I would write a follow-up entry that extracts that concept out of the complexity of the engagement, defining it and reflecting upon it with greater concentration.

Note that this only works one way, I cannot currently take a set of concepts dealt with in their own blog entries and then write an entry that deals with all of them together. So a key process, synthesis or summarization is missing. However, the 'follow-up' mechanism still does present a powerful tool for working with concepts.

In reality, it is probably the case that very few people use the follow-up system because they are just too busy recording those complex real-time engagements. They hardly ever get a chance to abstract out the single concept and work on it in isolation. We may be able to encourage the use of a pedagogy that gets students to work with concepts in this way, but I suspect that even then it will not become the intuitive mode of operation.

We can therefore conclude that the 'follow-up' facility is very useful, but for the concept development that we are after, it probably requires just too much extra work to be done on a consistent and widespread basis.

Semantic mapping of real-time engagements

To find a more intuitive solution, I have concentrated on what has to be the only realistic strategy: model intuitive and commonly used workflows for concept development, and use the technology to replicate them, but with some key enhancements that give the new tools an advantage over traditional methods. I'm sure there are several ways in which researchers and students develop concepts, reflecting the diversity of methodologies between and within disciplines. Although I think we should not overemphasise this. But to start somewhere, I will start with that with which I am familiar: philosophy. So how does it work in philosophy?

Firstly, what is a concept? Without delving too deeply into the relationship between mind, language, technique and material, we could simply say that it is some repetitively appearing notion that makes a definite and reproducible difference to the way in which an aspect of the world is conceieved. For example, the concept of 'human-rights'. When this is refered to, whether I agree or disagree with the proposition within which it is deployed, I will always concieve of a human as something that may or may not lay claim to a universal right. A concept doesn't prove anything, but it does raise a set of questions, it does posit the existnce of some thing or some attribute. And most importantly, it alters the outcome of an argument or a decision.

Surprisingly, when an individual concept is introduced to a student or first encountered by a researcher, it is rarely done so with much of a definition. A concept new to the student or reader tends to be applied immediately. We may provide some indication of the key concepts that need to be observed, possibly with a rudimentary definition, but that is all. The concept is quickly deployed to make a key difference to an argument or a decision. In the context of that argument or decision, in relation to the other more familiar concepts used, and in relation to the outcome of the argument or decision, the student gets a feel for the concept. Not necessarily a complete understanding, but at least a sense of it. This sense is fundamentally a cartographic sense, a feeling for the layout of arguments and discourses with the relative and repetitive positioning of concepts forming the peaks and contours. Intuitively the student develops a semantic map of the meaning of the concepts.

As the concept is then re-encountered in different contexts, for different arguments and decisions, that sense becomes more definite, or at least greater. The students gains a stronger orientation and sense of the landscape in which they are operating. In fact the concept itself, as it is put into action repeatedly, changes and gains definition. It rises from the background cartography, with features extending outwards to meet with other concept. The activity of definition itself, textual definition, usually only occurrs at this later stage. They know where the mountains are and now set out to conquer them. It is primarily concerned with setting out the scope of the concept, stating what work it can do, what difference it can make to arguments. We could say that this activity of definition, scoping and tracking the use of concepts, is one of the most important, if not the most important, academic activity.

If we now return to a consideration of how the blog system could be used in academic activity, we can see that it fits with this analysis of the workflow and entities involved in philosophy. For example, we want first year undergraduates to encounter the key concepts of PDP as applicable to their own activities and decision making processes. We may give them a primer that identifies the useful concepts, a rough map, with some sense of their scope, importance and suggested directions. But to actualy make sense of those concepts, the student needs to start applying them. And as has been made clear, the use of those concepts, the development of that understanding, happens within real-time engagements. The same is true of philosophical concepts. So we need to give them the ability to identify when they have used the concepts, directly within the context of their use: the blog recordings of the real-time engagements. We need to make it obvious when more than one of the concepts is used in relation to another. We need to allow them to track and visualise their use of the concepts over time, as they develop. And finally, we need them to focus the application of the concepts into the development of definition. The actual functionality that is required to help with this is as follows:

1. Priming of the development with a set of key concepts to be recorded in the blog. This could be more or less self-directed. The process of identifying key concepts and adding them to the list could be the responsibility of the student/researcher. A tutor could provide the list in the module blog, module web pages, with the students creating their own instances using 'blog this'. Or they could just be suggested in lectures and seminars. Or the student could be prompted to record these key concepts and register them in their blog. We could even pre-load a blog with an entry for each key concept. Would there be a separate category for 'Key Concepts and Definitions'?
2. Use of the concepts during the recording of real-time engagements in the blog. A simple means for identifying words or blocks of text that refer to a concept. This should trackback the use of the concept to the entry in which it was defined/introduced. More than one concept can therefore be used and linked to in a single entry.
3. When reading the entry it should be clear which concepts have been used and where they have been used within the text.
4. A search facilities that returns a list of instances in which the key concept has been used, or where a set of key concepts have been used together.
5. The student extends the entry with which they define the key concept as they use the concept within their blog.
6. Relationship specifiers that identify the relationships between key concepts as they appear in entries.
7. Visualisation tools that demonstrate the relationships between key concepts and the blog entries that they appear in – concept maps of the semantic cartography.

Robert O'Toole : 07 Nov 2004 17:17 |  Tags: E-Learning Warwick University |  Comments (4) |  Report a problem