The paradox of unlearning

"Repeated cycles of Learning, Unlearning and Relearning is a must for survival today!", thundered the sage on the stage at the HR conference. "Is it even possible to unlearn?", I wondered.

I am in complete agreement with the position that in an ever changing environment, we constantly need to learn new things. The question in my mind was only about the necessity and feasibility of unlearning. When I persisted with this question for a while, this entire matter of unlearning seemed to get increasingly paradoxical.  

So, what is paradoxical about unlearning? As we have seen earlier, a paradox occurs when there are multiple perspectives/opinions (doxa) that exist alongside (para)- each of which is true - but they appear to contradict/to be in conflict with one another. Now, let us look at some of the opinions about unlearning

1. Unlearning is as natural as learning.

2. Unlearning is simply impossible. You can't really remove something from your mind unless there is some sort of brain damage or extreme forms of mind control (like ‘brainwashing’)!

3. New knowledge replaces old knowledge as individuals learn more; much like overwriting. It is not considered to be the same as forgetting, where information is lost regardless of its usefulness.

4. Change in a particular behavior does not in fact remove the learning altogether; it simply reduces the likelihood of the behavior in certain contexts. Hence, the proposal that new learning ‘overwrites’ old learning is problematic.  

5. Existing knowledge or behaviors interfere with learning and, therefore, unlearning  needs to happen before new learning can occur.

6. Unlearning and learning occur simultaneously.

7. The distinction between learning, unlearning and relearning is arbitrary. 

8. Unlearning itself is very valuable. It allows us to see things as they really are. The essence of unlearning is about 'emptying' and not about 'emptying so that we can fill it up'.

9. Unlearning should not be viewed as an end in itself. It is just an intermediate step in learning.

10. Unlearning is not about forgetting. It’s about the ability to choose an alternative mental model or paradigm. When we unlearn, we step outside the current mental model in order to choose a different one.

11. We are usually unconscious of our mental models and that makes unlearning difficult. Also, we tend to view the new model through the lens of the old and that makes switching models even more difficult.

12. We don't have to worry about unlearning. Moving towards the new learning would automatically take us away from the old learning(and hence unlearning would happen automatically). Individuals learn new ways of choosing a response to a particular situation, rather than unlearning a particular response.

So, how do we resolve this?  

Even though there is quite a bit of discussion about unlearning these days, there are few theories confirmed by empirical evidence to identify how individuals unlearn and what factors may influence this unlearning. So we have to look at other options. One such option is to look at the underlying definition of 'learning' when we talk about unlearning.

If we define learning as 'acquiring knowledge', then unlearning is not possible (in the sense that it can't be forgotten) and also not even necessary (unless the existing knowledge was wrong or misleading, in which case it can be modified keeping in mind the new knowledge). When new knowledge is acquired,  the old knowledge is not erased, but maintained (‘in parentheses’) for situations where it is believed that the new knowledge does not apply.

If we look at learning as essentially a 'sense-making process' (where individuals interpret and create meaning of their experiences) and not as a 'fact gathering process', unlearning is about modifying the way we ‘make sense’ (see ‘architects of meaning’ for a related discussion).

If we define learning as 'sustainable change in behavior', then new behavior can just replace old behavior. The only case where  unlearning is required would be that of conditioned responses that interfere with learning new behaviors.  Some of these conditioned responses have roots in the underlying (unconscious) mental models. We are often operating with mental models that have grown outdated or obsolete.  It takes unlearning to see the model as only one if possibilities and not as the only possibility. So unlearning is not mainly about replacing one mental model with another; it is about having the ability to consciously chose from a range of mental models based on which is more appropriate in a given situation.

Unlearning is a process as opposed to a discrete event. The process of unlearning is about liberation from the conditioning. It involves intentional evaluation of self, task and the environment to determine if a change in the current behavior is necessary and possible. Mindfulness, ability to read contextual cues, openness to explore other possibilities and meta-cognitive ability are key enablers for this unlearning process.

It is also interesting to look at why there is so much interest in unlearning. It is probably because of the assumption that new learning can’t happen unless unlearning happens. As we saw earlier, this is not necessarily the case – except in the case of conditioned responses that create rigidity and impact 'learnability'.  Another key reason for the interest in unlearning is the belief that unlearning is essential for promoting innovation and for enabling organizations to respond more effectively to unanticipated change or crisis events (by to recognizing and modifying previous habits, approaches and behaviors that are no longer optimal).

Unlearning can happen at both the individual (assumptions, mental models, habits, response patterns etc.) and at the organization level (beliefs, paradigms, norms, rules, procedures, strategies etc.). While the unlearning at individual and organization levels can reinforce each other, they can also happen independently. For example, organizational unlearning can occur in the absence of individual unlearning through the removal of key influencers. Yes, it is important to look at the interface between individual and organizational learning in order to better understand and manage the interactions.

In a way, constant unlearning and relearning is a wasteful process. It is more efficient to modify or re-purpose old learning where possible(like a software 'update' as opposed to 'uninstall and reinstall'). Hence, relearning is better described as refocused learning (as opposed to replacing old learning with new learning). Individuals learn new ways of choosing a response to a particular situation, rather than unlearning a particular response. The focus is on modifying the response to be more effective and not on replacing one response with another (which might not necessarily lead to better outcomes). Hence, relearning is not antithetical to learning (it can happen without unlearning) and it is in fact more like learning that is made more appropriate to the current context! 

All adult learning involves relating new information with existing information and thereby modifying the existing understanding. Hence, there is no requirement for a ‘clean slate’ or ‘empty vessel’ (unlearning!) to enable new learning. Even in case of children, it seems a bit weird to suggest that a child has to unlearn 'crawling' before the child can learn 'walking'!

If we stick with the behavioral definition of 'learning' (as 'sustainable change in behavior'), adopting a new pattern of behavior is just 'learning' and hence we can even argue that the concepts of unlearning and relearning are not really required (except in the cases of brainwashing and cognitive impairment, respectively) and that they might be even misleading! Of course, we can examine and work on any possible impediments and enablers to learning in a particular context.

Any thoughts/suggestions?