Assumptions and triple loop learning

Thomas Aston
6 min readDec 29, 2020
Cy Twombly, Untileld (Bacchus)

Triple loop learning

When looking at complex change systems, it’s not only important to reflect on actions and results. As Chris Roche put it to me recently, teams also need to periodically reflect on their assumptions about how change happens, and their identity and values, what some call triple loop learning.

Theories of change aren’t just a tick box log frames on steroids. They are about the process of achieving a shared understanding and about reflecting on our assumptions. Learning is process of reflection and gaining (new) knowledge and understanding. We’ve all seen this, right?

Even without pretensions to wisdom, we can’t go from data to insight without questioning our assumptions. Making those connections also works at different levels. And this is about different types of learning loops. There are a number of different ways to represent these learning loops, given that there’s little agreement on the origins of the concept and what its centre of gravity should be. However, borrowing from my colleagues at the Partnership to Engage, Reform and Learn (PERL) programme in Nigeria, I think we can see it like this:

1. Single loop learning: This is geared towards reflecting on the question: are we doing things right? This is about whether planned activities are being achieved (or not) and what immediate course corrections may be required to get back on track. It’s often thought of as a thermostat (too hot or too cold). For example, did our activities get conducted as planned this month or quarter? Here we commonly reflect upon detecting errors and enabling short-term improvements in working practices and results. So, we’re talking mostly about WHAT we’re doing and fixing day-to-day operational problems.

2. Double loop learning: This is about asking: are we doing the right things? This is a slightly deeper reflection on emerging patterns or trends, often over a longer period of time in a single location, or in comparison across different locations. This can be about how we interact with partners and communities and whether we’re doing this in the right way. It’s therefore a reflection on causal assumptions, pathways of change, organisational norms, practices (or processes) and policies. So, we’re not only talking about the WHAT, but also about the HOW.

3. Triple loop learning: Triple loop learning is even deeper still and tends to be about longer timeframes, considering organisational principles and goals (values, mission, vision). Beyond a reflection on patterns of success and failure, it’s about asking what is right? In other words, it’s about considering why we do what we do and reflecting on who we should be as a programme and even the organisation itself (our identity). This is therefore also about the WHY & WHO.

We can see one good representation of the different levels and learning loops below:

See: Argyris, 1977; Argyris and Schön 1992; Sinek, 2009; Ramalingam et al. 2009; Roche, 2010; Ørnemark, 2015; Prieto Martin et al. 2017; Cabaj, 2019 for some inspiration

Matching learning loops and assumptions

However, I think we can move the conversation beyond loops. As Chris Argyris (1977) pointed out:

If people are unaware of the propositions they use [i.e. theories of action], then it appears that they design for themselves private assumptions that are not genuinely self-corrective. They are thus prisoners of their own theories.

One realisation of mine recently was that these learning loops map reasonably well to different types of assumptions we have in our theories of change. There are a number of different types of assumptions too, but this is how I prefer to distinguish them:

  1. Paradigmatic assumptions: Structuring assumptions we use to order the world into fundamental categories. Most commonly, these are belief systems, world views, and philosophies. So, this is primarily about the WHY we do what we do and WHO we want to be.
  2. Prescriptive assumptions: Assumptions about what we think ought to be happening in a particular situation. Such assumptions are about how we believe we and other actors should behave. It’s about our view of what is deemed appropriate in a particular context. These assumptions underlie WHY we do what we do based on shared norms and values. So, this is also chiefly about the WHY and the WHO.
  3. Causal assumptions: These are assumptions about how different parts of the world work and about the conditions under which these can be changed. This is about the events and conditions needed (i.e. necessary) for the associated causal link to work — for the cause (or set of causes) to lead to the effect. This can be idiosyncratic to a particular location or may hold across locations. So, this is more about the way in which the WHAT connects to the HOW.
  4. Operational assumptions: These are assumptions about the operating environment for delivering a programme. This is about what it takes to implement. These assumptions might includes external context, such as issues of political stability, freedom of expression or movement, environmental factors, but they more commonly relate to operational logistics (resources, access, participation). So, this is chiefly about the WHAT.

If we flip the learning loops upside down, it quickly becomes clear that these broadly (although imperfectly) match, as we can see below:

So, when asking are we doing things right? we’ll mostly be reflecting on operational (or implementation) assumptions, questioning issues like access, logistics, and what’s impeding us or our partners from delivering activities. We’re asking this kind of question and looking at this kind of assumptions very regularly (whether consciously or not). This tends to be more the tick box log frame on steroids kind of thing so often criticised. As Chris Argyris famously put it:

“You cannot openly confront norms that tell you not to confront policies and objectives.”

For are we doing the right things? we’ll mostly be reflecting on causal assumptions; questioning the links between events, and whether the approach taken is the right one (or not) in the areas we work. As we don’t want to change our approach or strategy too regularly (or we’ll go round in circles), we should be asking this less frequently. In my view, this should happen in line with the rhythm which we expect significant changes to happen. While this can be highly structured (e.g. biannual), it can also be less structured — reflecting on learning after an intense period of effort, such as through an intense period debrief or an after action review following a key moment in your programme. Alternatively, if we’re looking for trends, then we can ask questions about why change may be happening in one context and not in another, or why a strategy was fit for purpose in location A, but not in location B.

For what is right? we have the opportunity to reflect on what type of programme or organisation we want to be. Given that this is ultimately a question about organisational identity, it’s pretty rare to do this more than once a year. We should be questioning our values, and normative assumptions about what is and isn’t appropriate to do (Are we the right organisation to do this? Does working with these partners contradict our values? Does taking this kind of approach align with our mission?).

Hopefully, I’ve convinced you how it can be helpful to be more explicit about your assumptions, and how this may help you learn better.



Thomas Aston

I'm an independent consultant specialising in theory-based and participatory evaluation methods.