It’s the last post of the month. To date, I think that I’ve been sceptical about the merits of the SOLO taxonomy, so I thought that it was time to switch perspectives and argue its case.
It is certainly better than starting to write learning outcomes with nothing but an empty head and an empty page. Writing learning outcomes is hard. I’ve supported a number of academics in doing so, and if you have no framework it appears impossible to capture years worth of stuff into 10 or so neat phrases. The most tempting thing to do is to describe those things that are easiest to measure, which can mean that learning outcomes sink to the lowest common denominator. By framing knowledge in this way, the SOLO taxonomy provides a valuable way of reminding teachers that outcomes should match rather than than limit ambition.
Equally, if we stick with the tried and tested binary of surface and deep learning, we may be in danger of over-simplification. Learning either becomes good learning or bad learning. At least with the SOLO taxonomy we get a sense in which new knowledge builds on and is connected to old knowledge. It may still be an abstraction, but at least it has enough shades of grey to give educators more nuance in how they describe what it is that they think that a student should be able to demonstrate at the conclusion of his or her studies.
Biggs and Collins should also take no responsibility for those who use it dogmatically. The SOLO taxomony, as with the Bloom taxonomy before it, should be judged on its usefulness. It is there as a handy guide to educators rather than a statement of absolute fact. As such, the true test of its worth will be its ability to stand the test of time. If people are still referencing it in 20 years time it has clearly served its purpose.
It’s too darn hot. The thermostat has hit 30. Science says – yes, all of it – that high temperatures can lead to distorted thinking. So, let’s address one question before I melt into a puddle on the floor. Are taxonomies useful in education?
In the natural sciences, taxonomies are an attempt to classify the world, to map out who is most closely related to whom. A duck-billed platypus may have a bill that looks like, well, a duck but that does not make it a duck. It is an example of convergent evolution and sits in an entirely separate branch of the evolutionary tree. Very few natural scientists that I’ve met have ever argued that classifying the world is not a useful exercise, even if they had disagreed about how things are classified. Is it equally the case in education?
I would argue that taxonomies play two principal roles: to spark a debate about something and to make complex things more understandable. Debating is good, its teaches us to present and hone arguments, listen to counter-arguments and helps to ensure that whatever agreement is reached is more robust. A lack of debate can lead to bad things, like being swallowed by a whale. Making things more understandable is also good, to a point. The world is an impossibly complex place, and it seems a reasonable thing to create ways in which to navigate through this infinite mound of stuff with our sanity’s intact. However, do taxonomies such as the SOLO example help us to do this?
Firstly, let’s address the labelling. The names are awful. Surface and deep learning, threshold concepts and rhizomatic learning make some immediate sense to us because they use a commonly understood metaphor to help us understand the point they are making (in the same way as online retailers use the notion of a ‘basket’ to help us understand how the transaction will progress). The SOLO labels are not so intuitive. Pre-structural, Unistructural, Multistructural, Relational and Extended Abstract are terms that obfuscate rather then illuminate. They act as a barrier to understanding. Contrast that with the labels used with Bloom’s taxonomy (Remembering, Understanding, Applying…) which can be more readily understood.
Secondly, there is a danger that some people may use these taxonomies in a literal way. Instead of using them merely as a reflective tool, people may assume a literal progression between one state and the next. No such progression exists, at least in my view. We can not observe someone ascending from the multistructural to the relational, nor can be definitively prove that such distinct states exist. However if someone taken by the model uses it in dogmatic ways (in the same way that Mumford’s learning styles have been misused) then it could have the opposite effect (i.e. encouraging people to follow – and accept – the model rather than think for themselves).
For now, I think I’m going to stick with surface and deep learning as my taxonomy of choice. It is simple to explain, simple to understand and simply profound.
I have a copy of John Bigg’s ‘Teaching for Quality Learning at University’. It defines the extended abstract as:
“Going beyond what has been given, whereas the relational response stays with it. The coherent whole is conceptualised at a higher level of abstraction and is applied to a new and broader domains” (Biggs, pg. 40).
So, in order to qualify as an extended abstract you would need to create new knowledge, to make – in their terms – a breakthrough. I have a slight problem with that. Biggs acknowledges that in time the extended abstract becomes the relational inasmuch as new knowledge inevitably becomes new old knowledge. Isn’t this an example of rationalisation?
For example, Alfred Russel Wallace came up with the somewhat heretical notion that animals were subject to variation according to the environment in which they lived. He authored an essay on this issue, which he passed to Charles Darwin. This could be described as an example of the extended abstract. Of course, Darwin had been harbouring similar notions and was encouraged by Wallace’s essay to publish his own thoughts on the matter, titled ‘On the Origin of Species’. Does Darwin lose credit for this because Wallace had similar thoughts? Is getting there first important?
Perhaps I am doing Biggs and Collins a disservice. In talking about going beyond what is given, I am not clear whether they are talking about an individual making new connections for themselves – irrespective of what other people have done – or making connections that no one has made before. Either way, is it reasonable to ask people to create new knowledge as part of a set of learning outcomes? Should PhD’s have learning outcomes?
Ultimately, is newness a helpful criteria for classifying learning? Innovation, after all, is not always a good thing. For me, I think that it is positive to remind learners that knowledge isn’t static, and that they can help to discover new knowledge. I just wouldn’t call this extended abstract thinking. I would call it creativity.
Not Han Solo. Not Napoleon Solo. Not the terrible 1996 Mario Van Peebles film. In this case SOLO stands for the Structured Overview of Learning Outcomes, the 1982 taxonomy proposed by John Biggs and Kevin Collins.
Educational theorists like taxonomies. They like ordering the world so that this is connected to that or so that this is better than that. Perhaps the most famous educational taxonomy is Bloom’s taxonomy devised in 1956 by a panel of educators chaired by the Educational psychologist Benjamin Bloom. That progresses from knowledge at the bottom through to synthesis at the top (or creativity if you follow the 2000 update).
In the SOLO taxonomy there are five levels:
- Prestructural: I know nothing
- Unistructural: I know one thing
- Multistructural: I know several things
- Relational: I can make connections between the things that I know
- Extended abstract: I can take the things that I know and make new knowledge
So, unlike most taxonomies facts aren’t at the bottom. However, this is only because Biggs and Collins have introduced knowing nothing as one of the levels. At first glance, it does not seem entirely clear why Biggs and Collins have felt it necessary to distinguish between knowing one thing and knowing several things, nor what the difference between the relational and extended abstract levels are. In a sense, there is an immediate danger that you can simply the SOLO taxonomy by reducing five levels to two:
- Multistructural: I know some things
- Relational: I can make connections between the things that I know
Or, to put it another way, surface and deep learning. For my next post I think I’ll have to focus on making sense of the ‘extended abstract’ level to see whether the SOLO taxonomy really has something new to say about how we design and describe learning.