I’m blogging about the development of a new product in Mozilla, look here for my other posts in this series

In a presentation The Revolution Will Be Annotated, Dan Whaley begins with a very high-minded justification for annotation: that it is essential for our existence that we act wisely, and that we can achieve that through deliberation, and that annotation is a building block for open deliberation.

In response, let me digress wildly, and talk about elementary school.

It is common to cite large class sizes as a problem, and small class sizes as an opportunity to improve education. But there is debate about whether class size really matters; it certainly correlates with general privilege, but does it cause improvements? At the same time I’ve become much more familiar with the Montessori philosophy of education, and one of the surprising features is a relatively large ideal classroom size, in the 30s. And why? Dr. Montessori had positive theories about age mixture, community size, the culture of the classroom, and so forth – but I’ll add what I think is a Montessori-style reason it’s okay: it’s okay to have less teachers because learning isn’t caused by teachers. Learning is ultimately an internal process, an assimilation and construction of knowledge. Your environment can aid in that process, but the cause is still internal.

I share Dan’s enthusiasm about the importance of dialog to our collective wisdom. But I see dialog as supportive of personal growth, not of collective wisdom – our collective wisdom will increase as we individually grow.

Dan cites one problem with rationalism: we are good at constructing rational arguments to support what we already believed. The annotation remedy is to suppose that what we conveniently leave out of our arguments can be applied later by a diverse set of participants. Annotation makes it harder to make use of fallacies, harder to make use of limited narratives, because the annotator can add them back in.

I will cite another problem with rationalism: even a good rational argument is not good at convincing anyone of anything. A good rational argument is like teaching arithmetic by telling someone that 39301+9402=48703. Maybe even writing out the steps used to make that calculation. You can lay that in front of someone, you can lay a hundred similar examples in front of someone, and they will not learn arithmetic. If the person is disinterested they can just trust you – though then it hardly matters if you were right – but if they are interested I believe that the process of construction is necessary. You have to solve your own math problems to learn math.

Annotation is interesting because it gives another avenue for people to publish their beliefs and enter into dialog. But a global overlay of annotation is not a particularly appealing medium. It’s not a place to come to understanding, to practice the construction of ideas. And I think our collective wisdom depends on an incredible volume of discussion, not just on an increased quality, because you can’t get large scale individual growth without large scale discussion.

I see two features in the typical annotation system: one feature is the ability to talk about other things, with high fidelity. URLs have been a great start to being able to talk about things on the web, but they have some limits. The second feature is the ability to view these annotations implicitly. The second feature is the one I’ve seldom found interesting as a reader, and disagree with as a goal. Viewing annotations as a universal overlay of commentary asks too much of the annotations, provides too little to me as the reader, and I think is an attempt to pursue a kind of rational universal truth that I find little value in. It’s a sense that documents are there to teach us, and annotations make those documents even better teachers.

PageShot takes a different approach: it gives anyone the ability to talk about anything on the web, but each time you do that you create a new resource, your discussion lives at your document. You can write your commentary for a specific audience, and then give it to that audience, without having your intended audience confused with the original author’s intended audience. You can throw away what you have to say. The person who clicks on your link did it to see what you said, they aren’t some passerby. You can say implicitly through highlighting, here is what I thought was interesting. But PageShot applies no universality to what you say, it is a tool only of dialog. This makes PageShot more modest, but intentionally so.