My colleague, Gadi Ben-Yehuda, is deep into a discussion about using hashtags on Twitter and of strategies for broadcasting to certain audiences. His use of certain terms raised a few hackles, and he asked me to share my thoughts. Once I reviewed the conversation, I have to agree with the Tribe of the Raised Hackles.
In my view, Gadi is using the language and terms of control. Hashtags and other metadata, to me, fall in the same category as taxonomy. Once upon a time, we in the KM world appeared to have a singular faith that organizing content and messages within a common taxonomy would facilitate recovery and effective dissemination.
The problem with this approach is that it presumes you can establish relevance as the sender, and fails to consider the bias filters of the receiver (message decoding in Shannon Information Theory). In addition, the pre-categorizing of information for the receiver misses out on serendipity. When he reference one-to-many in his post, he is correct: but only for the first message. The value of SM for some is that the notion of an exchange, rather than a broadcast. This is not to say taxonomy is useless, it can be essential for interoperability across Agencies. The business reference architectures coming out of the OMB Line of Business research is an excellent example of establishing a common taxonomy across similar functions in disparate agencies. However, attempting to organize that which is designed as an anarchic form, strikes me as applying – inappropriately – the tools of control to what is a somewhat disintermediated forum. Not all attempts to organize are inappropriate, I should stipulate; but the notion that all messages should be categorized by the sender is – in my view.
Gadi speaks of adding structure to the medium, a medium that is fighting against existing structured channels for information flow. This is Shirky-esque; the notion that we are moving away from one-to-many broadcasts and towards a many-to-many construct. In one sense, Gadi is correct. Technically. When the Coast Guard uses the hashtag for #tsunami, they are speaking to people who have structured their feeds in such a way as to receive these efficiently.
However, this is only true for some. Others may note the preponderance of ‘tsunami’ in a word cloud, or search for the word if they hear snippet on the news. And while the USCG may have defined their audience as he notes, they cannot know who else may use the information.
The bottom line, at least for me (until Gadi responds): Attempts to structure what people need to know from your message rankle those who are angling for disintermediation in their communications and news. When he references one-to-many, Gadi is correct: but only for the first message. The value of SM is the notion of an exchange, an information commons, rather than a broadcast channel. Part of what hampers KM and information sharing (my soapbox) is the notion that we have to first structure what we are saying for the people we think we’re talking to. This leads to the enforcement of a document metaphor, rather than true sharing of granular information. Stop trying to answer a question you think your readers have, and just put it out there.
This is not to say organizing or tagging your content is always wrong – but it is wrong to claim you must always organize or tag your content. And this was Gadi’s original point. I think the salient question for those pondering this discusion are: Where do you spend your energies, and what design principles underlie your work?
One last thought: The driving notion when I was working (with a large team) on the DoD Information Sharing Strategy: We must be prepared to share information with unexpected partners, driven by unanticipated events. If that is the design principle, then pre-packaging, metadata, hashtags – all those actions that presume we know how our information will be used – become much less interesting.