May 25, 2006
Downes on Howell on Reflection
Stephen Downes comments on Catherine Howell's post on reflection: "Catherine Howell argues that the definition, in the IMS specification, is fuzzy, and insufficiently distinguished from an assertion. Quite right, but I would go further and ask what the point is of distinguishing between a reflection and an assertion at all, and even what the point is of placing these (via a form???) into an e-portfolio in the first place."
This echoes a discussion on the CETIS portfolio list, and in a SIG meeting in Liverpool recently on just how "structured" a portfolio really is. Reflection, to my way of thinking, is an attitude to communication as much as an intentional act, and it may be the case that a piece of writing is only really "reflective" in retrospect.
This has implications for systems in that, we may want to write and create more or less freely but later perhaps tag, group, or connect these artifacts to make a particular point. So choosing in a form "this is a reflection" prior to writing does seem a bit peculiar.
To be fair, the IMS EP spec includes reflection (ok, "reflexion") in its ontology as it can be a useful tag when overlaying metadata onto a collection (or conversation), rather than it necessarily being a useful distinction when engaged in the act of doing and creating.
The question then is, is it useful to do this after-the-fact classification? And does it have meaning for anyone other than the author?
You could argue that this post <reflects-on>Stephen's post</reflects-on> which itself <reflects-on>Catherine's post</reflects-on>. Does this need to be explicitly tagged? The answer, I think, depends on how you view the 'semantic web' approach in general.
Personally, I think portfolios are, and mostly will be, unstructured collections or networks of all kinds of media (and conversation fragments) intended to be interpreted primarily by other people. The structuring (and coordination) mostly takes place in the activity that results in this human-readable material being amassed and connected together in the first place. Sometimes we may even make a linking-artifact that tries to convey a particular message (e.g. a C.V. or an "about me" page, or a public profile) but again this is a media artifact and is primarily intended to be viewed, read or played by people.
I don't object to the notion of adding a semantic layer (either as tags, topic map, ontology, or even an IMS EP manifest) that makes machine-interpretation of this material easier, but I don't think it should be the critical concern. Portfolios, to me, are primarily for human-to-human communication.