A modest proposal from a language teacher: let's stop using b*llocks learning outcomes

Any discussion among teachers about good practice will get around, at some point, to learning outcomes. Conversation will often turn to Bloom's Taxonomy and the attendant verbs connected to the "lower" and "higher" order "thinking skills". And then someone will raise the idea of SMART targets, which, in conjunction with Bloom's, enable us to develop learning outcomes such as (in a language learning context like mine) "use the present simple tense to write four simple sentences about your daily routine" or "identify six details from a text".

The trouble is, these are – and I'm sorry to put it like this – b*llocks. They are. For a number of reasons. The first and the biggest problem with them is down to SMART targets. I've long been critical of SMART as a paradigm for any kind of learning aim. When teaching discrete language items, they are simply used as proxy evidence of learning – what the teacher is really thinking, (and what the student is most likely thinking), is that "use present simple to write four sentences" means "learn about and practice present simple tense".

If the outcomes are taken at face value, then the tightly drawn nature of the outcome suggests that this language learning is now complete – that the student can now read any text and identify any four details from that text, and use present simple in any context or setting automatically.

Bloom's taxonomy makes this even worse. Gianfranco Conti gives a brilliant explanation of why Bloom's is problematic at best, misleading at worst in the context of language teaching. I strongly recommend you read it, but his argument is that language learning doesn't sit at one level of the hierarchy suggested in Bloom's, but spreads across a whole range.

In my lesson on present simple tense, for example, I would not only be expecting students to carry out "lower order" tasks such as identifying and categorising the way the present simple is formed. I would also be expecting them to hypothesise about the grammar structures involved, and perhaps synthesise this awareness, perhaps comparing the form with a language they already know. I would later ask students to demonstrate their hypothesis through practice activities, and later to apply that language knowledge in a less controlled way. Here we have visited several layers of Bloom's taxonomy, sometimes simultaneously.

In the very best case, the duality of Bloom's and SMART makes for a set of "outcomes" which are no more than descriptions of the tasks students will do in the class, albeit obscured by removing reference to the context of the lesson. At worst, the whole business misleads the teacher and students into a false representation of what is happening in the classroom, and the processes taking place there.

But let's not throw out the baby with the bathwater. We do need to keep a focus for the lesson, we just need to get rid of the performativity and accountability that SMART and Bloom's imply. Today we are going to practise reading for gist/detail; we are going to learn about the present simple tense; we are going to find out about words to do with travel and transport. No need here for measurable outcomes, because the benefit of the measurability, as I've already said, is either pointless or spurious.

This is not just an academic argument. The use and abuse of learning outcomes is wielded as a stick by observers and inspectors. Are students aware of the outcomes, are they displayed for all to see? Can the teacher assess the learning based on these outcomes? Are there opportunities for stretch and challenge in the learning outcomes? While the first point is simply a question of practice, the latter points are rendered obsolete by pointlessness and spuriosity. If the outcome doesn't represent the nature of the learning, as a Bloom's-SMART outcome would suggest, then assessment is similarly unrepresentative, and differentiation is merely a question of luck and instinct.

Do we need to lose outcomes? Yes, and no. But we need to lose Bloom's and SMART, yes, absolutely. They don't help us in developing the focus for a lesson. Instead of a single, meaningless outcome, a better model might be the division of the outcome into the intention and the assessment. The assessment doesn't belong in the outcome at all: it belongs in the lesson, and in the mind of the teacher. The intention of the lesson is a much more valid concept (with a nod to Dylan Wiliam), and allows us both honesty and accuracy in our representation of learning. Students will be exposed to language, and will have had a chance to practice that language, or develop the skill of reading for gist, listening for detail, whatever. This is the learning intention.

The added bonus of this separation is that there is now more room for emergent language. If language arises, as it does, then there is capacity here for capturing, sharing and evaluating that language, which a closed list of learning outcomes on the Bloom's-SMART model simply does not allow for.

As a model, the separation of aim and assessment allows for aims to evolve and grow, and reminds us to think about how we are checking for learning, insofar as this is possible, of both planned and unplanned language. Which, really, is no bad thing at all. 

Sam Shepherd is a languages teacher in further education in the UK. This is an edited version of a piece that first appeared on his blog