The Ed Techie

Reductionism is your friend

<management of complexity - http://www.flickr.com/photos/michaelheiss/3090102907/>

Plenty of better qualified people than myself have made this point, but it's cropped up a few times in both Twitter and face to face conversations, where 'reductionist' has been used in a generally dismissive way, as if to be identified as such signifies the end of a discussion. So, partly because I want to clarify for myself, and also to get feedback in case I'm missing something, I thought I'd set down my take on reductionism.

In general, I think reductionism is a sound approach. Anyone with a child will know that we are natural reductionists. When your child goes through the 'why?' phase what they want is a sub-explanation to a seemingly simple question ('Why is it cold?'). That nearly all children do this naturally should tell us it's a pretty good knowledge gathering approach. You look for answers to a higher level phenomenon by breaking it down into its constituent parts.

Reductionism works well - all around you are the products of reductionism: that computer you're using could have a sticker on it 'Powered by reductionism'; that book on evolution 'Authored by reductionism'; that medicine that is curing your disease 'Brought to you by reductionism'. 

So why are people so down on reductionism? Dan Dennett puts it well in his book Darwin's Dangerous Idea:

"The term that is most often bandied about in these conflicts, often as a term of abuse, is 'reductionism'... [reductionism is made to look] philistine and heartless, if not downright evil... There are bland and preposterous readings of [a reductionist] claim. According to the bland readings, it is possible (and even desirable) to unify chemistry and physics, biology and chemistry, and yes, even the social sciences and biology... According to the preposterous readings, reductionists want to abandon the principles, theories, vocabulary, laws of the higher-level sciences, in favor of the lower-level terms. A reductionist dream, on such a preposterous reading, might be to write "A Comparison of Keats and Shelley from the Molecular Point of View"... Probably nobody is a reductionist in the preposterous sense and everybody should be a reductionist in the bland sense, so the 'charge' of reductionism is too bland to merit a response"

The problem is that of 'greedy reductionism' as Dennett terms it, whereby too much is claimed, too quickly and overly simplistic explanations are given which miss, or ignore some of the complexity. In reaction to this it seems that sentiment has swung too far the other way and declared all reductionism wrong.

There are also phenomenon that reductionism is not good at explaining, for example chaos theory, and emergence. Some scientists will argue that emergence can still be explained in reductionist terms we just don't have sufficiently complex models, and others that it can't and there are emergent properties that are greater than the sum of the parts. This is a debate worth having and whether different approaches are viable to different phenomenon, but surely the conclusion is not that reductionism never works (we have plenty of evidence to the contrary).

Other people have a general mistrust about reductionism, that it is mechanistic and they want to retain some sense of the whole. This can be a misunderstanding about what reductionism seeks to do, as Dennett suggests. As both Marr and Pylyshyn have argued there are different levels of explanation: if you want to know why I was running down the street you expect a psychological explanation ('I was late for a meeting') not a biological one ('my muscles were contracting causing my legs to move in response to messages from my brain'), or a chemical or physical one.

The concern I have about those who wish to dismiss reductionism in favour of more holistic responses is that it is often a pseudo-explanation or even a dead-end. For centuries the church provided this pseudo-explanation along the lines of 'It's God's will', which effectively means don't look any further. But being good little reductionists, people did look further and we found subsequent explanations at every level we looked. This didn't stop people feeling emotions, producing art or making jokes. So I am always wary when reductionism is rejected but nothing offered in its place other than vague ideas of complexity or emergence (or even mysticism). We've been here before. (There is also a slight implication that reductionism is in some sense 'right wing'. I'm probably naive in this sense, but I simply don't understand this at all.)

I fully concur that reductionism is a) not always the best way to explain things b) may miss some subtlety between the components and c) may not be the correct approach as we look at emergent properties in complex systems. But that doesn't mean it offers nothing, so rather than the rejection of reductionism being an endpoint in an argument ('that's so reductionist!') it should be the start, in that you need to have a good reason why it won't work and what you are replacing it with.

A conspiracy of sentiment

Yesterday Paul Chambers lost his appeal against his fine and conviction for posting a joke on twitter which was prosecuted under the anti-terrorist legislation.

The case was so obviously ridiculous that everyone thought common-sense would prevail, but eschewing humour and reality, Judge Jacqueline Davies deemed the tweet "menacing in its content and obviously so. It could not be more clear. Any ordinary person reading this would see it in that way and be alarmed."

Any ordinary person would have seen it as a joke. I had previously thought this was simply a case of the law getting itself in a mess, escalating something and then being unable to climb down. Added to that a healthy dose of people not understanding twitter. 

Now though I'm convinced it's more than that. I don't believe in conspiracy theories, I know all those involved haven't got together at some late night meeting and decided to make a random citizen's life a misery. But when something so nonsensical progresses through multiple stages and possibilities to put an end to it, then you have to look for some explanation. I think although there is no outright conspiracy, what we do have is a conspiracy of sentiment. All those involved at various stages: politicians, the police, CPS, judges, media are all acting from the same unspoken emotional base. This can be summarised as: they hate you.

They hate that you undermine their carefully crafted messages and turn them into jokes. They hate that you are forming new methods of entertainment that they don't understand. They hate that you can organise yourselves without them knowing about it. They hate that power has been democratised. They hate that you get at content for free. They hate it, hate it, hate it. So when the opportunity arises to stamp on one of you snivelling social media types, they grasp it with both hands.

And that's why the Twitter Joke Trial is important - because it reveals the mindset of those in power, and because it will be the first of many attempts by them to control or at least seek revenge on what they don't understand and loathe.

The Lamb Formula = Good enough?

In my previous presentation one of the slides I didn't get to displayed the following Venn Diagram for what I term 'The Lamb formula'

(Image created by crappygraphs)

This is from Brian Lamb, who borrows the title from Errol Morris' documentary, to describe the kind of technology he likes, and thinks is useful. My take on these three elements (which may be different to Brian's) as they pertain to education is:

• Fast - Technology that is easy to learn and quick to set up. You don't need to go on a training course to use it, or put in a request to central IT services to set it up. This means you can experiment quickly.
• Cheap - tools that are usually free, or at least have a freemium model so you can buy any extension out of your own pocket. Animoto is a good example - you can create free 30seconds videos or pay a small one-off fee to create a longer one. Slideshare is an example of a free service. This means you don't need to get authorisation to use them from a budget holder. It also means you don't need to worry about the size of audience or return on investment, which is liberating.
• Out of control - they are outside of formal institutional control structures, so they have a more personal element, and are more flexible. 

Overall, this tends to encourage experimentation and innovation in terms of both what people produce for content services, and the uses they put technology to in education. If you've invested £300K in an eportfolio system, say, then there becomes an obligation to stick with it over many years. If you've picked a free blog tool and told students to use it as a portfolio you can switch if you want to and also put it to different uses.

There are, of course, many times when this approach may not be suitable, but that doesn't mean it is never appropriate. Thinking about this also brought to mind a couple of conversations I've had with Liam Green-Hughes recently, who is very interested in the notion of 'good enough'. Liam is referring to hardware often when he talks about this, the Flip cam being a good example. It is good enough at recording to make the lightweight device and the ease of use be determining factors.

I have wondered what 'good enough' means in education terms. In his Open Ed keynote Erik Duval suggested that we can afford to disrupt education, because we don't do a particularly good job anyway. I think we do sometimes become scared of experimenting in education and become obsessed about every possible bad scenario, while ignoring some of the current deficiencies (in fairness this is partly a reaction to the litigious environment universities now have to exist in).  The good enough approach suggests that using fast, cheap and out of control technologies in education is at least good enough, and may be superior in that offers new ways of working. 

Let's play OER Roulette!

(The academic version of sending your child up a chimney)

This is the last in my trilogy of Open Ed 2010 posts.

For my presentation at Open Ed 2010 I wanted to do something a bit different. I had the last slot on a long second day (ie ‘the graveyard shift’) and I was speaking after David Wiley (who gave an excellent presentation as always), so the odds were stacked against me. In addition my daughter accompanied me so I wanted to include her in some way.

I remember seeing Elvis Costello in concert once and he had a large wheel at the front with a number of songs listed on it, which they would spin to select the next number to play. I wanted to recreate this, but transporting a big wheel was unrealistic, so I got my daughter to get members of the audience to select cards. My talk had the hypothesis that by considering OERs as big and little, it provided a lens on a number of issues around OERs, including:

• Sustainability
• Working methods
• Production time
• Context
• Academic status
• The role of content
• The nature of projects
• Sites vs Portals
• Aggregation vs Adaptation
• Implict messages

This gave the presentation a non-linear structure and because I had more subjects than could be covered in the allotted time, a degree of uncertainty. It also means I can give the same presentation again and it will at least be slightly different.

The presentation is below, which I haven't had time to turn into a slidecast and is now linear, so it loses some of the interactivity. I quite liked using this approach and given my previous post on not doing presentations, it at least allowed me to play with the format to a degree, but I would have preferred it to be more participatory.

OER Roulette

View more presentations from Martin Weller.

Am I done with conferencing?

This is no reflection on Open Ed, I’ve been moving away from conferences for a while. I enjoyed giving my presentation at Open Ed, and I managed to make it interactive to an extent. But I’m coming to the conclusion that I’m done with the traditional conference format. I had a couple of discussions with Brian and Scott around this and talked about how difficult it was to break away from the traditional format. The reason comes down to money essentially, although it’s tied up with a host of other issues.

There is a circular logic in organising a conference that is difficult to escape, let’s call it the Conference Conservatism Circle, which goes something like this: In order to make a conference viable you need people to attend and pay fees. To attend most people need to get funds from their university or a research project. To justify this they need to give a presentation. A presentation needs to be peer-reviewed so they can include it on their CV. People only attend conferences that offer this. You are organising a conference and need to make it viable…

 

Cycle

View more presentations from Martin Weller.

This is another example of where academic practice is continuing despite the alternatives now available. Before we had a digital, open, network then getting people together so they could present to each other was a reasonably pragmatic solution to knowledge sharing. We have largely continued with this practice despite it no longer being the best use of the time.

I’m not arguing for all online conferences here, I think getting together with people is valuable. Indeed it’s precisely because I think it’s valuable that I think we should stop wasting it giving presentations.

The barcamp and unconference models have experimented with this, but this more fluid, practical or discussion based approach has not been widely taken up in academic circles, largely because of the economics I outlined above (and a rather innate sense of conservatism amongst academics). Here are some questions I’ve been asking of myself:

• If peer-review is still important then how about a more open model, where your proposal for a suggestion is voted on by all delegates prior to the conference?
• Why don’t we use the net for the information dissemination function (eg make our presentations live beforehand, as video or slidecasts) and then use the face to face segment for discussion?
• Can we have more practical sessions that aim to develop something? This could be a site, a learning resource, some code, a set of guidelines, etc. But actually have functioning outputs from conferences, instead of just suntans and hangovers.
• Some people are worth seeing as presenters, so is there a proportion of the conference (but a minority) that is given over to these?
• Would more open presentation formats work, eg having to speak on a random subject for two minutes?
• If conferences are really about forming networks can we facilitate people forming groups (based on prior discussions and interests) around developing research projects?

All of these seem to me good uses of the conference time and having people co-located, yet all would struggle to get acceptance and funding as proper academic activity. We do have, for instance, research proposal sessions, or writing workshops, but these are both rare and distinct from the normal conference activity. I know some conferences have been experimental in their use of the format, but the majority of conferences across disciplines are still quite traditional and are locked into the Conference Conservatism Cycle.

So my modest proposal is one of direct action. As with my stance on only reviewing for open access journals, if we all insisted on moving away from the traditional presentation format, would conference organisers, and more importantly, funders shift practice also?

I know I’ll probably renege on this, but I’ll try it out: If you want me to come to your conference, then I will only do so if I can try something akin to the formats above, and not a standard presentation.

 

The strange case of the unconnected conference

(An unexpected appearance by the Two Ronnies went some way to alleviate frustration about the lack of wifi at OpenEd 2010 - Richard Hall and Joss Winn in full flow).

I was at the Open Ed 2010 conference last week. First off, I want to say that it was great to finally meet Brian Lamb and I really enjoyed presentations from Richard Hall, David Wiley, Erik Duval, Joss Winn and others. So this post is not a comment on the quality of the conference or the discussions I had there.

The conference was held in the science museum CosmoCaixa. This is a great museum and my daughter loved it. It is, sadly, not a great conference venue. Although they had wifi in every room which you could connect to, the bandwidth coming into the museum was so low that in reality no-one could get online (I had visions of a 14K modem in an office somewhere which then got fed into all these lecture halls with great AV facilities and their own routers).

This made me reflect on how connectivity, and the amplified conference has become the norm for me. Being unable to connect had a serious impact on the nature of the conference I think, in the following ways:

1. Loss of archive -  It couldn’t be streamed, or tweeted, or live-blogged, so the more holistic, distributed record of a conference that has become the norm was lost.
2. No backchannel – within the conference itself not being able to chat with other participants and share thoughts meant some of the immediate conversation around a presentation was lost.
3. Stunted presentations – many presenters wanted to at least demonstrate a site. I foolishly wanted to play some YouTube vids (although thankfully an Ethernet connection had been found for presenters by the time I came to present which gave some limited access). Being unable to do so made it seem like we had been sucked into a conference in 1995.
4. No wider participation – at least with those present you could have conversations over coffee, but I am also accustomed to have dialogue with those in my network who aren’t present. There were many time I would have tweeted during a talk and I know I would have had a response from someone in my network who wasn’t present.
5. No amplification – if one of the measures of success for a conference is its impact in its field then to lose connectivity for its duration results in it effectively being isolated. Not to be able to amplify the content and discussion at a conference to a global network severely restricts its significance.

Before someone jumps in and says ‘perhaps you should try talking to the person next to you’, yes I did plenty of that. I’m capable of doing both. And I think there is increasing value in hosting ‘disconnect’ events, where you deliberately don’t have access, such as writing retreats. But these types of events are geared up for this and it’s what you sign up for. What the experience here brought home to me is that amplification is not an interesting extra or a techie nice-to-have – it’s essential.

 

An unbundled publishing business proposal

For my digital scholarship book I have been reading about the tenure process a lot (I'll blog it soon). One of the issues that arises is that peer-review is always perceived as the gold standard. Promotion committees in universities, faced with the complexity of judging research in highly specialised fields, revert to using peer-review publication as a proxy. This is becoming problematic in the humanities in particular where the academic book is seen as the main evidence of excellence in research. Such monographs were often published by university presses. But in financially straitened times, many presses are closing, and if they are staying in business then they need to make publishing decisions with very strong financial incentives - they're only going to publish those books which they can be sure will be profitable.

Now, I think we should be finding new ways of achieving what peer-review provides for us, but in the meantime here is a business proposal. Publishing a book is easy - anyone can do it without the need for approval from a publisher using services such as Blurb and Lulu. But in academia we still want the peer-review filter prior to publication.

So here's my publishing business proposal - a service which sits on top of an online publisher such as Blurb. The model would work something like this:

1. The author pays a modest fee (let's say £100) for peer-review which is organised by the company. They then either pass the book, get modifications or reject it. 
2. Once accepted the author publishes via Blurb (or other service), using the publishing company's style sheet to mark it as approved, and with a set price.
3. An additional proof-reading service can be provided by the company, for a fee.
4. The author gets to keep all profits from the sale of the book, hopefully recouping the cost of the initial peer-review process, and making more than they would through the small percentage from traditional publishers.

What is missing is the distribution channel into libraries and book stores here, but that might be an additional service that the author could again opt for at a fee.

What this model does is unbundle the various functions of a publisher and allow authors to pay for them selectively, but to keep all of the profit from the book. At the moment these services are all bundled together and the author pays by only getting a small percentage of the royalties. The unbundled model allows greater flexibility and also means that an author can take a risk if they really want to publish something but it may be unprofitable. In the current model we have conflated academic worthiness and profitability - when you review a book you are really passing a judgement on how well it will sell, not on how academically sound it is. By separating out these elements this model brings clarity (I think) to the process, and still allows academics the peer-review filter they cherish.

Is someone already doing this and if not, anyone want to fund it?

Tenure, publishing and Tony

As Tony Hirst has blogged, his recent promotion case was unsuccessful. I'm obviously disappointed by this, for his sake, and because it was our first attempt at pushing through a digital scholarship case.

We don't know why it was unsuccessful yet (detailed feedback will follow I suspect), but today I was reading an excellent report from the Center for Higher Education Studies at Berkeley titled "Assessing the Future Landscape of Scholarly Communication: An Exploration of Faculty Valuesand Needs in Seven Disciplines". There are several points it raises which I think reflect on Tony's experience and others like him. Here are some key quotes with my interpretation of each of them:

"the peerreviewed journal article is the primary mode of scholarly dissemination in the sciences and the quantitative social sciences, while the more interpretive, historical, and qualitative disciplines rely heavily on the university press monograph with a varying mix of journal articles, critical editions, and other publications. These traditions, which rely heavily on various forms of peer review, may override the perceived “opportunities” afforded by new technologies, including those falling into the Web 2.0 category."

So, even if we acknowledge there are benefits to using new technologies, the cultural traditions may prevent us taking advantage of them. This was a theme I explored in my earlier presentation on researchers and new technologies and which I go into detail in for my forthcoming book.

"In most fields, however, a stellar publication record in prestigious peer-reviewed outlets usually counts significantly more in advancement decisions. At some institutions, scholarly contributions such as data curation or multimedia websites are considered to be forms of “service” or “teaching” in a scholar’s academic portfolio, or they may receive credit when presented in a peer-reviewed publication that “discusses” the resource or data set."

Even when other outputs are recognised, the peer-reviewed journal trumps them all to the extent that they may placed in an ill-fitting separate category.

"The degree to which peer review, despite its perceived shortcomings, is considered to be an important filter of academic quality, cannot be overstated."

This stuff is deeply entrenched.

"enthusiasm for the development and adoption of technology should not be conflated with the hard reality of tenure and promotion requirements (including the needs and goals of final archival publication) in highly competitive and complex professional environments. Experiments in new genres of scholarship and dissemination are occurring in every field, but they are taking place within the context of relatively conservative value and reward systems that have the practice of peer review at their core."

This reinforces the point I made in my research talk - the environment we have created works in direct opposition to innovation and adoption of new technologies. This is a dangerous situation.

If we take Tony as an example the way he works raises some interesting questions about what we mean by research. He typically explores new technologies and data visualisation in particular. A random sampling of recent posts include:

• An analysis of twitter connections between UK politicians
• A representation of online communities who use the same hashtag
• An interrogation of the Mendeley software to show users by institution
• Sharing his own promotion case
• A presentation on ‘data driven journalism’

Each of these is intended to promote discussion, and has suggestion for implications, for example in how higher education can make effective use of data. None of the posts arise from a specific research project, and each of them is fairly small in terms of time and resource. The use of a blog though (instead of publications) allows Tony to engage in this ongoing experimentation, as it has an outlet, and it simultaneously encourages it also since discussions will arise on the blog (or in other places such as Twitter). Taken as a whole then, the blog itself represents the research process, and in this context it is difficult to say that it is not demonstrating the REF definition of research: ‘a process of investigation leading to new insights effectively shared.’

A paper on scholarship has recently been approved by the OU Senate and there is undoubtedly a desire to engage with and recognise new forms of scholarly activity. But what the report demonstrates is that cultural practices are almost inextricably bound up with publishing. And when you unpick what publishing means it leads you to some difficult considerations. This is Bellow's Law in action. When a case like Tony's comes across it simply doesn't fit well with the existing framework. This is, I guess, one of the perils with being a pioneer. Not that this is much consolation.

The way I'm working is (kinda) working

I read Tony Schwartz's The Way We're Working Isn't Working recently (and after my last post, yes it was on the Kindle). Some of it irritated me and much of it I couldn't relate to (not being a high-powered AMERICAN executive who works 15 hour days), but after Scott blogged about the meditation app he found useful, I thought I'd post some thoughts on what I did take away from it.

The first is that napping is good! I like an afternoon nap, but in an open plan office the sight of someone sprawled across their desk with a small puddle of drool congealing by the keyboard is, for some reason, deemed unacceptable. He advocates a 90 minute full on sleep, which is pretty much impossible to fit in, but I do find when I'm working at home a quick 10 minute nap is enough to revitalise me. I don't suppose in these financially straitened times a campaign for more napping in the workplace will find much support but he has some strong evidence that it is much more productive to do so than to carry on through.

Secondly, he proposes working in 90 minute creative bursts, without distraction (we are nowhere near as good at multi-tasking as we think we are he argues). I tried this recently when I had a week of study leave to work on my digital scholarship book. I parcelled the day out into 90 minute chunks and during these I turned off twitter and email and wrote. At the end of 90 minutes I would make myself stop, go away from the desk for half an hour and do something else (walk the dog, have lunch or nap). It worked very well, I wrote a lot in that week and I maintained a high level of concentration and felt more focused.

The rest of the book has advice on eating, self-esteem, taking holidays, etc which are things I don't worry too much about anyway. But it did make me appreciate that we don't often reflect on our own working practice, it is too often something imposed on us by meetings, physical environments and deadlines.

However, there are times when I don't mind working in the distracted, channel-hopping mode he deplores. For example if I'm doing a task that is tedious (completing a research proposal say), then I like to be dipping in and out of twitter, following links, reading other bits in between - it makes the task more bearable. Whether it makes me more productive is a different matter, but if it means I get the job done with sanity intact, that is good enough for me. And it also helps to mix up something mildly creative (even crafting a jokey tweet) with something laborious, possibly even making the output of the tedious task better as a result. 

So I don't think it's a matter of having one way of working, but rather of having ways that suit different requirements. And with that platitude I shall take my nap...

 

Kindle & the iTuning of book purchasing

Probably my favourite app on my iPad is, errm, Kindle. Yes, getting my £600 device to behave like a £100 one is where it's at.

That aside, what is interesting about the Kindle is the way it is beginning to alter my book buying behaviour. In short it is iTuning my purchasing behaviour, in that I am tending to buy more on impulse - a recommendation or thought comes to me, and I don't add it to a wishlist, I buy it. This means I am adopting a more exploratory approach, buying books I wouldn't do otherwise, or ones I know I only want a section from.

But this behaviour is hampered by the publishers and arcane pricing policies. iTunes had the advantage that music (in the form of CDs) was evenly priced, ranging from around £3.99 to £13.99. This meant that getting consensus from recording companies on a set price of 99c for a track was relatively easy. Books on the other hand (and particularly academic books) range wildly from £1 to £100. Reaching agreement on a uniform price is therefore more difficult. But, for the sake of the industry itself, it is essential. At the moment the Kindle version is usually the same price as the physical one - and laughably, sometimes it is more expensive.

Amazon is probably the only player big enough to push this through. Publishers are proprietary and closed by instinct, so want to believe e-books are just another outlet but with higher profit margins. There are two reasons why I think they need to alter their behaviour:

i) iTuning buying behaviour leads to greater sales - if the standard price for a Kindle book was sufficiently low, say £1.99, then it leads to higher sales, particularly in the long tail. Brynjolfsson, Hu & Simester (2007) demonstrate that when a product moves online the concentration of sales becomes more distributed: “the Internet channel exhibits a significantly less concentrated sales distribution when compared with the catalog channel, even though these two channels offer the same products at the same set of prices.” Being online encourages a more ‘long tail’ oriented set of behaviours. They further argue that as search costs reduce, then so sales concentration becomes more skewed towards niche products. This means selling more of those special interest books, but only if the price is sufficiently low to encourage 'I may as well' purchasing behaviour. Maintaining the price point artificially high loses the benefits they might gain from this new channel.

ii) Resentment breeds piracy - by maintaining physical object prices, buyers become irritated at what they perceive as both greed and a lack of understanding around the new channel. Music piracy was driven partly by getting something for free, but also out of frustration that music companies were either ignoring the internet or treating it just like an infinite shopfront. One of the reasons Apple got agreement for iTunes was the fear of Napster. By repeating these mistakes publishers will drive potential customers into piracy, and then spend ages trying to lure them back. This is a key moment for book purchasing, they shouldn't ignore the very recent lessons history has to teach them. I want my book purchasing behaviour to be iTuned, don't turn me away now.