In response to my recent mindmap of Andrew McAfee’s conference keynote (one of a number of mindmaps I’ve done), I got this comment:

Does the diagram work as a useful way of encapsulating the talk for someone who was there? Because, speaking as someone who wasn‘t, I find it almost entirely content-free. Just kind of a collection of buzz-phrases in thought bubbles, more or less randomly connected.

I‘m not trying to criticise his talk – which obviously I didn‘t hear – or his points – which I still have no idea about – but the diagram as a method of conveying information is a total failure to this sample size of one. Possibly more useful as a refresher mechanism for people who got the talk in its original form?

Do mindmaps work for readers?   Well, I have to admit one reason I mindmap is completely personal.   I do it to help me process the presentation. Depending on the speaker, I can thoughtfully reprocess the information, or sometimes just take down interesting comments, but there are several benefits: In figuring out the ways to link, I’m capturing the conceptual structure of the talk (really, they’re concept maps), and I’m also occupying my personal bandwidth enough to allow me to focus on the talk without my mind chasing down one path and missing something.   Er, mostly…

Then, for a second category, those who actually heard the talk, they might be worthwhile reflection and re-processing.   I’d welcome anyone weighing in on that. I don’t have access to someone else’s example to see whether it would work for me.

Then, there are the potential viewers, like the commenter, for whom it’s either possible or not to process any coherent idea out of the presentation.   I looked back at the diagram for McAfee’s keynote, and I can see that I was cryptic, missing some keywords in communicating. This was for two reasons: one, he was quick, and it was hard to get it down before he’d moved on.   Two, he was eloquent, and because he was quick I couldn’t find time to paraphrase.   And there’s a more pragmatic reason; I try to constrain the size of the mindmap, and I’m always rearranging to get it to fit on one page.   That effort may keep me more terse than is optimal for unsupported processing.

I will take issue with “more or less randomly connected”, however.   The connections are quite specific.   In all the talks I’ve done this for, there have been several core points that are elaborated, in various ways by talk, but each talk tends to be composed of a replicated structure.   The connections capture that structure.   For instance, McAfee repeatedly took a theme, used an example to highlight it, then derived a takehome point and some corollaries.   There would be ways to more eloquently convey that structure (e.g. labeled links, color coding), but the structure isn’t always laid out beforehand (it’s emergent), and is moving fast enough that I couldn’t do it on the fly.

I could post-process more, but in the most recent two cases I wanted to get it up quickly: when I tweeted I was making the mindmap, others said they were eager to see it, so I hung on for some minutes after the keynotes to get it up quickly.   McAfee himself tweeted “dang, that was FAST – nice work!”

I did put the arrow in the background to guide the order in which the discussion came, as well, but apparently it is too telegraphic for the non-attendee. It happens I know the commenter well, and he’s a very smart guy, so if he’s having trouble, that’s definitely an argument that the raw mindmap alone is not communicative, at least not without perhaps some post-processing to make the points clear.

Really valuable to get the feedback, and worthwhile to reflect on what the tradeoffs are and who benefits. It may be that these are only valuable for fellow attendees.   Or just me. I may have to consider a) not posting, b) slowing down and doing more post-processing, or…?   Comments welcome!

I was recently thinking about grades, and was wondering what an ‘A’ means these days.   Then, at my lad’s Back to School night, I was confronted with evidence of the two competing theories that I see. One teacher had a scale on the wall, with (I don’t remember exactly, but something like): 96-100 = A+, 91-95 = A, 86-90 = A-, and so on, down to below 50 = Fail. Now, you get full points on homework just for trying, but it’s clearly a competency model, with absolute standards.   A different teacher recounted how she tells students that if they just do the required work, that it’s only worth a C, and A’s are for above and beyond. That’s a different model. There aren’t strict criteria for that latter.   And I’m very sympathetic to that latter stance, despite that it seems subjective.

The second approach resonates with my experience back in high school, where A’s were handed out for work that really was above and beyond the ordinary.   A deeper understanding.   We seem to have shifted to a model where if you do what’s asked, you get an ‘A’.   And I see benefits of both sides.   Defining performance, and having everyone able to achieve them is ideal.   Yet, intuitively, you recognize that there’s the ability to apply concepts, and then another level where people can flexibly use them to solve novel problems, combine them with other concepts, infer new concepts, etc.

I was pointed to some work by Daniel Schwartz (thanks @mrch0mp3rs) that grounds this intuition in an innovative framework based upon some good research. In a paper with John Bransford and David Sears, they made a intriguing case for two different forms of transfer: efficient and innovative, and argued convincingly that most of our models address the former and not the latter, yet addressing the latter yielded better outcomes on both.   I think the work David Jonassen is doing on teaching problem-solving is developing just this sort of understanding, but it’s on problems that are like real world ones (and yet improves performance on standard measures).   And I like David’s lament that the problems kids solve in schools have no relation to the problems they see in the world (and, implicitly, no worth).

Right now, our competencies aren’t defined well enough to support assessing this extra level.   In an ideal world, we’d have them all mapped out, and you could get A’s in every one you could master.   We don’t live in that world, unfortunately.   So we have two paths.   We live with our lower measures, and everyone gets A’s meeting them (if they try, but that’s a separate issue), and then sort it out after graduation, in the real world, or we allow some interpretation by the teacher and measure not only effort, but a deeper form of understanding.   We’ve steered away from the second approach, probably because of the consequent arguments about favoritism, social stigma, etc.   Yet the former is increasingly meaningless, I fear.

We should bite the bullet and admit that we’re waving our hands. Then we could own up that not all teachers are ready to do the type of teaching David’s doing and Daniel’s advocating, and look to using technology to make available a higher quality of content (like the UC College Prep program has been doing) to provide support.   I’d rather see a man-in-the-moon program around getting a really meaningful curriculum up online than going to Mars at this point, and I’m a big fan of NASA and the pragmatic benefits of space exploration. Just think such a project would have a bigger impact on the world, all told.

In the meantime, we have to live with some grade inflation (gee, I got into a UC with a high school average below 4.0!), bad alignment between what schools do and what kids need as preparation for life in this century, and a very long road towards any meaningful change.   Sigh.

Aaron Silvers (@mrch0mp3rs) wrote a post tying together the non-linear nature of cyberspace with the essentially linear nature of our past.   Identifying how new technologies have to establish their own natures is a familiar refrain, but he’s comparing our learning with our   new contexts, and essentially questioning the relationship between old methods and new contexts.

It was this quote which got me thinking:

communications outside of education are happening, increasingly, in hyperspace…, but our preferred methods for learning are still long-held narrative forms

I started wondering about whether we could learn without a linear narrative form, and realized we could.   That is, we really do want to train certain people (e.g. pilots) to react before conscious thought kicks in. And similarly in, say, martial arts. It gets to a point where the expert creates rationales for what they’re doing that aren’t necessarily tied to the real action.   We saw this in the problems that arose with ‘expert systems’, where experts articulated what they did, and then they built systems that did what the experts said, and they didn’t work.

Typically, we start off on component skills, addressing them with some explicit feedback, and gradually layer on more complex frameworks (“be the ball” :), and the performance is compiled into a deeper, really subcognitive level.   We’ve used stories and feedback, and then augmented with video capture and fancy feedback machines, but we’ve kept that conscious layer of description even as it gets more abstract to match our increasingly high level of control.   Our brains have evolved to process linear narrative, as we see in all the calls for incorporating story in communication. And reflection, a critical component of learning, is conscious.   Now, consciousness is a still not understood linear phenomena emerging from our parallel processing brain.   What would it mean to end-run linear consciousness?   That’s what I started trying to imagine.

And what I came up with was a simulation game/immersive environment with non-cognitive feedback, that could train your responses. You’d be performing, and the feedback would train and integrate your responses.   Yes, there’d be an explicit model of performance to guide the feedback, but the learner might not be aware of the relationships.   Yes, learners would likely create a story about what’s happening (given that we’re likely to reflect consciously on what we do), but it could be wrong or they could even have their attention drawn elsewhere. So the game might even have some layer with high or little relation to what the system is training.   But it could work, like in that possibly apocryphal story of the students who kept paying high quality attention to the instructor when he was on side of the room, and looked away at the other side of the room, and at the end of the lecture the instructor was way off in one corner of the room.

And what concerned me was that I wouldn’t want to have a parallel training system that bypassed or misled the conscious learning.   And that’s what I raised as a comment on Aaron’s post.   Not that I thought Aaron was advocating that, but it’s just where my thinking went, and he rightly queried what I was on about.   I couldn’t fit it in a comment, hence this post.

The point being, there’s a role for the linear conscious narrative and reflection in our learning, just as there’s a role for dynamic multimedia networks in our learning as well.   It’s finding out the balance, the role of activity and reflection, that’s the interesting and important challenge.

Harold Jarche (@hjarche) retweeted his prior post on “First, we kill the curriculum“, and generated some serious interest.   For instance, Mark Oehlert (@moehlert) was inspired to write “Harold Jarche is Wicked Smart and We Need to Talk about Curriculum“.     I know Harold, and he is wicked smart (see this skewering of homework), so I commented on his blog and it seems we may have a semantics difference as opposed to a fundamental one.   Still, I want to make the point.

Harold, starting from the premise that the web is as fundamental a change as was the printing press, and, as the press could foster content, so the web can foster connections.   The emergent nature of knowledge out of a network argues against a fixed curriculum and instead for contextualized knowledge.   Arguing against a fixed curriculum, he says this:

a subject-based curriculum will always be based on the wrong subjects for some people. Without a subject-centric curriculum, teachers could choose the appropriate subject matter for their particular class

and this is, I think a valid point.   There’s too much focus on rote, and already out-dated knowledge.   Making my lad continue to demonstrate his ability to do the times tables ad nauseum only kills his love of learning.   And the first year of middle school seems to be much more about turning them into manageable prisoners rather than learning much of anything.   Things are moving so fast that it’s hard to imagine that much of what we learn, other than vocabulary, math rules, and science basics are necessary.   Jim Levin argued 30 years ago that learning multiplication and long division was outdated in the age of the calculator and that estimation was the necessary skill to ensure you were in the right ball park.   Why are we still teaching long division?   In short, drill and kill is pretty dumb.

So, is there a curriclum?   I think so, and Harold really says so too:

ensuing that students have mastered the important processes. Some of the processes that readily come to mind are critical thinking; analysing data; researching; communicating ideas; creating new things

Now, I think we’re arguing over whether skills are a curriculum, and I reckon they are, that is a focused set of learning outcomes we’re trying to achieve.   Not content, but skills.   I do believe there are some fundamentals, like levers, and gravity, and the associative property, but these are frameworks and models, tools upon which we build a flexible set of problem-solving, coupled with just the sort of skills Harold’s talking about (and I’ve talked about before).

The point is, the world’s changing, and yet we’re not equipping our kids with the necessary skills.   We need a new pedagogy, problem-focused on things kids are interested in, as Harold suggests, and focusing on their information seeking and experimentation and evaluation and the self-learning skills, not on rote exercise of skills.   I don’t do long-division anymore.   Do you?   Do you graph sentences?   Do you remember formulas?   I don’t think so.   What you do is look up information, make job aids (why are stickies so ubiquitous?), or program in the formulas or use a tool.

Whether or not we call it a curriculum, those problem-solving skills what I want my kids learning, and it’s not happening.   State standards are a joke. I don’t want them learning how to bold in Word, I want them understanding the concept of ‘styles’.   I don’t want them learning how to color a square in Powerpoint, I want them to be effective in communicating visually.   I want them to be learning how to solve ill-structured problems (cf another wicked smart person, David Jonassen)!

I don’t mind the revolutionary statement “kill the curriculum”, but I might just mean it as “kill the current curriculum”, because I do believe that the most effective path to help develop those skills is a formal learning process.   However, it’s likely to be socially constructive in nature, not instructive.   Let’s kill schooling, and reinvent schools as learning labs, with curricula focused on skills and attitudes, and perhaps a minimal core of knowledge.   Which means our standardized tests need tossing, too, but then that should also be obvious. Portfolios and contextualized abilities, not rote knowledge tests.   I reckon Harold and I (and Mark, another wicked smart person) are agreeing furiously.   Anyone for a revolution?