Metaphors & Mental Models: Updating Our Language for Public Sector Learning

We need to rethink what it means to learn and be learners, in a complex, dynamic and fluid world where many answers are not predetermined, but must be cultivated and co-evolved.

Metaphors and Mental Models_banner-teaser

Date Posted

4 Nov 2021


Issue 23, 29 Oct 2021

In some recent talks I have given, I’ve begun by asking the audience what the following sentences have in common:

  • “I will defeat that argument.”
  • “Children blossom into adults.”
  • “I don’t have room for this in my life!”
  • “Life has cheated me.”
  • “Scarcity has given birth to a generation of paranoid teenagers.”

After a few attempts at finding thematic unity in these very diverse sentences, someone in the audience usually figures out that none of them is a literal description: each employs a metaphor of some kind—a comparison of one thing to another—to highlight particular characteristics. Children are likened to flowers, arguments to battles, life to a physical space or game. The concept of scarcity is anthropomorphised: treated as human, e.g., in its ability to ‘give birth’.

Metaphors matter. While they are usually taken to be the exclusive tools of writers and poets, we all use metaphors more often than we might be aware, and they shape how we perceive the world and think about issues. Our choice of metaphor can subtly affect not just what we think, but also what we do. For instance, imagine how saying “I will engage that argument” instead of “I will defeat…” might change the tone and tenor of our interaction with the source of that argument.

As thinkers and practitioners have long pointed out (see box story Metaphors in Organisational Life), metaphors profoundly shape the way we conceive of and carry out life in any institution, and how we view ourselves and one another in relation to our organisations.

Our current ideas about learners, learning and work have been marked by pervasive metaphors, which in a world of increasing complexity and interdependence may prove increasingly deficient. What might more appropriate new metaphors be, and how might these change the way we think and practise learning in the public sector?

Metaphors in Organisational Life

Metaphors profoundly shape the way we conceive of and carry out life in any institution, and how we view ourselves and one another in relation to our organisations.

Read More

Learners as Receptacles: The Metaphor and Its Limitations

Arguably the most well-used metaphor for learning is to see learners as receptacles, with learning as a process that fills them with useful content: information, knowledge, experience or skills. This thinking has permeated even our daily idioms: we learn from an early age that ‘empty vessels make the most noise’, unless their emptiness is rectified by the process of being filled with new knowledge. Brazilian philosopher Paolo Freire’s “banking concept of education”1 is similarly underpinned by the idea that students simply store information relayed to them by teachers.

The receptacle metaphor is seductive in its simplicity, but makes at least eight problematic learning assumptions:

  1. Assumption of Instrumentality: Learning must lead to some organisational or individual benefit, often measured in terms of outcomes. Learning must make the vessel less empty, or else it cannot claim the name ‘learning’.
  2. Assumption of Concrete Rationality: Learning, whether for individuals or a whole organisation, must involve some analytical rigour to have value. The empty vessel must be filled by something of literal substance—with mass, weight, density—rather than something more intangible or indefinite.
  3. Assumption of Certainty: We need to know, before embarking on a learning endeavour, that the filling of the empty vessel will happen; that the learning will involve some content that imparts clear-cut wisdom to the learner.
  4. Assumption of Additivity: Learning is good, so more learning is better; the more the empty vessel is filled, the more positive the outcomes.
  5. Assumption of Fixed Capacity: Each vessel, by nature, has a limited volume, beyond which it cannot be filled further.
  6. Assumption of Linearity: Learning can be measured in terms of units, as the empty vessel is filled. Each additional unit of learning adds as much as the previous one, consistently over time.
  7. Assumption of Passive Receptivity: The empty container is filled by beneficial knowledge without much action of its own. Both the words ‘receive’ and ‘receptacle’ share common etymological roots in the Latin ‘receptare’, to ‘receive back’.
  8. Assumption of Individuality: The empty vessel stands alone as it is filled; one person’s learning is carried out, and has its effects felt, in a mostly isolated or atomised manner. Little interaction happens with other vessels in the process.

When squared off against a world characterised by more complex issues, with interdependent actors, these assumptions start to show their limitations.

For instance, the Assumption of Instrumentality is limiting when we realise that some learning can have intrinsic rather than purely instrumental worth. The gains from learning need not be functional, in terms of better skills, productivity or output, but could instead help inculcate in learners a deeper sense of meaning and purpose, autonomy, or even the satisfaction of mastery over challenging subject matter. Such benefits are inadequately reflected in the receptacle metaphor, which reflects consequentialist modes of thinking that emphasise productive outcomes and the optimisation of resources. It ignores the possibility that valuable organisational outcomes can emerge from something other than filling or being filled: such as when employees make higher quality contributions because they feel more deeply engaged after a learning experience.

Similarly, the Assumption of Concrete Rationality does little justice to learning that is non-cognitive. Some issues require approaches that are not just analytical but also linked to emotions, resilience, and being psychologically informed. These call for learning that supports understanding the vagaries of one’s feelings, recognising that some skills (like mindfulness) are physically embodied.

This is partly why so many programmes for professionals, whether at schools of business or government, or entities like Singapore’s Civil Service College (CSC), are devoting increasing amounts of time to dimensions such as self-care, as well as dealing with crises and other adversities that involve responses with cognitive, emotional, and embodied dimensions. CSC’s Applied Simulation Training Laboratory has contributed significantly in this domain, developing immersive, experiential learning platforms like serious games and policy simulations.

Some issues require approaches that are not just analytical but also linked to emotions, resilience, and being psychologically informed.

Beyond the Assumption of Certainty, learning needs to incorporate approaches that afford greater space for uncertainty—accounting for issues with no clear-cut outcomes at the start, and which could benefit from iterative and experimental approaches. Such learning does not involve simply absorbing what John Maynard Keynes termed a “body of settled conclusions”,2 but involves the learner sitting amid, and making sense of, surrounding turbulence and flux. A significant amount of ‘on-the-job training’ can take this form: especially pertaining to roles that transcend simple templates and checklists that may involve tasks that are indistinct, slippery, and unclear. Learning, in such instances, happens by doing, and not just by being filled with pre-determined information, knowledge or expertise.

Learning involves the learner sitting amid, and making sense of, surrounding turbulence and flux.

The Assumption of Additivity fails to account for situations where ‘less’ is sometimes ‘more’, when the tempo of learning needs to be varied for maximum effect. For instance, in CSC’s leadership programmes, ‘white space’ is often deliberately set aside for learners to reflect on and consolidate previously conveyed content, rather than assuming that more content is always better. Taking this idea further, there are instances where what was learnt earlier might need to be shed, as learners take on new roles. Marshall Goldsmith epitomises this idea in his book—aphoristically titled What Got You Here Won’t Get You There3—because sometimes we need to change our approaches, and not simply learn more of what we already know and are comfortable with. Rather than learners as receptacles, one might think of learners as sculptors, chiselling away and removing, rather than adding, until they arrive at what Michelangelo memorably named “the angel in the marble”.

The Assumption of Fixed Capacity is flawed because, as psychologist Carol Dweck has pointed out in her concept of the ‘Growth Mindset’, learners can evolve, progress and acquire new scope for ever greater learning as they mature, given the right guidance and feedback.4 Learners can even be changed by the content and process of learning. In contrast, the ‘Fixed Mindset’, like the receptacle metaphor, assumes that learners remain static over time.

The Assumption of Linearity is challenged by insights from complexity science. In interdependent systems where cause and effect relationships are not obvious ex ante, outcomes do not always emanate from input in the neat, linear relationships that characterise mechanistic, Cartesian situations. Instead, complex systems involve ‘phase transitions’, more popularly known as ‘tipping points’, where unexpected outcomes can emerge either very swiftly or only after extended periods of time. The study of complex systems often reminds me how some of the content I learnt as a new public officer, at a CSC programme called the Foundation Course, only matured into fulsome insights many years later. Broad concepts like ‘Whole-of-Government’ thinking or ‘Whole-of-Nation’ approaches only came to life as I experienced a range of jobs, each highlighting different operational aspects that could not have been fully explored in a single programme. In some circumstances, a new Assumption of Non-Linearity might do better justice: since the deepest learning may only occur long after the original teachable moment is over.

The Assumption of Passive Receptivity is reductive because learners ought to be active shapers of their own learning—a point emphasised by proponents of constructionist learning approaches. For them, learners are far from mere receivers, devoid of agency and with deficits to fill with knowledge. Instead, learners have gifts and assets in their own right, which they can exercise and use to enhance their learning processes. I have often found this to be true when coaching public officers transitioning into new jobs: while they certainly have new skills to acquire, they often also bring valuable insights and experiences from their previous jobs, which can help them acclimatise to their new roles. These construction processes need not be limited to the individual, but can indeed be processes of social construction, at the level of teams or some larger aggregation of individuals.

This is why the Assumption of Individuality needs to be refined along with the Assumption of Passive Receptivity: complexity and interdependence often call for learning at the team or collective level. The transformative power of socially constructed learning is abundantly demonstrated in the case of Austin’s Butterfly,5 in which a class of first-grade students at ANSER Charter School in Boise, Idaho helped their classmate Austin take a scientific illustration of a butterfly through multiple drafts, starting with a rudimentary sketch but culminating in a high-quality final product. Far from being a lonely or atomised process, learning in this case is shown to be richest when it is collective, with critique and descriptive feedback leading to exponential improvements. The social nature of the process happens through mechanisms like conversation, feedback and coaching, and deliberative discussions. A favourite part of this story is when an American first-grader points out that the feedback from Austin’s peers worked best when it was “more specific, but they weren’t mean about it”.6

Complexity and interdependence often call for learning at the team or collective level.

Better Metaphors for Learning

Given the problematic assumptions of the receptacle metaphor for learning, can we do better? Is there a metaphor that can adequately reflect the importance of intrinsic motivation, emotion and embodiment, uncertainty, subtraction and pauses, non-linearity, learner agency, as well as socially constituted learning?

One possibility is the idea of learning as a garden: a thriving and complex ecosystem in which different flora and fauna (even weeds!) have a part to play. Each learner could be a different inhabitant organism, whose role may not be obvious in a direct, instrumental sense, but nevertheless bringing a significant contribution to the garden ecology. Some learners could even be gardeners, shaping and sustaining the entire ecosystem. Without any one component, the delicate ecological balance of the garden will be destabilised. Gardening is an even better metaphor than sculpture, since it deals with ‘angels’ that are living and dynamically evolving, not trapped, immutable, in a piece of stone. Even stones in gardens are alive and evolving: subject to erosion, weathering, the ravages of nature and time. They also interact with other elements and organisms in a garden, all of which co-inhabit common spaces and co-influence one another, just as learning can be communal.

Since gardens consist of unpredictable living beings, they are intrinsically uncertain; but they can accommodate uncertainty by making micro adjustments in their overall ecological composition. Gardens are thus well suited to non-linearity, evolving as they undergo phase transitions (sometimes over highly extended time periods), rather than hewing rigidly to pre-set outcomes. Each organism in a garden, animal or plant or mineral, has scope to shape its own role and destiny, similar to agency-driven learning, and these organisms interact socially, so that the whole becomes greater than the sum of the parts. The Singapore Army captured these aspects in the idea of a ‘Learning Army, Thinking Soldier’, part of the broader transformation to a third-generation Singapore Armed Forces. Building on the foundation of rigorous training and doctrine, thinking soldiers exhibit the living qualities of a garden: situational awareness, adaptation and agility.

Gardens are not wild, untamed jungles. They benefit from pruning—some plants only bloom after, not before, pruning!—and are hence attuned to the gains from subtraction rather than uncritical accumulation. So long as they are given space to grow and are not overly manicured, gardens can be lush and vibrant. Not every contribution in a garden needs to be rational or analytically substantiated; some parts could be important as spaces for emotions to play out, or for embodied activities like exercise or quiet walks to take place.

Similarly, the learning process has its untidy and unpredictable aspects but it is not a totally whimsical free-for-all. The garden metaphor captures how learning also needs regular tending—without such deliberate attention and effort, both learning and gardens will atrophy and eventually wither from neglect.

Unsurprisingly, the garden metaphor is already part of our vocabulary for learning—kindergartens are literally ‘gardens for children’!—and it is not a stretch to see its renewed relevance to continual adult learning today.

Cultivating Learning Values

What sort of values might support the view of learning as a garden? In Singapore, the Public Service values of Integrity, Service and Excellence should continue to apply at the global level, across all agencies and functions of the Government. But some additional learning-specific values would also be instructive. These Learning Values could be derived from the Assumptions that the garden metaphor helps to refine:

  1. Learning has intrinsic worth
  2. Learning involves the entire body: not just the head, but also the heart and hands
  3. Learning is iterative
  4. Less learning can be more
  5. Learning capacity can grow over time
  6. Celebrate complexity in learning
  7. Learners have agency to construct their own learning
  8. We construct learning together as well as alone

These values seem important for all levels of public sector learners, from the most junior to the most senior. Since leadership can be particularly pivotal in articulating, shaping and sustaining organisational learning cultures, a set of complementary Leadership Learning Values could be developed:

  • Leaders should take into account both intrinsic and instrumental measures of value
  • Leaders should celebrate whole-person capabilities , not just analytical skills
  • Leaders should encourage iterative processes, not just concrete outcomes
  • Leaders should know when to slow down to celebrate pauses and deliberate reductions; leaders should model ‘less is more’
  • Leaders should find ways for their teams to grow, and not assume their capacities are fixed
  • Leaders should recognise interdependence and adaptivity
  • Leaders should exercise agency and encourage others to do so—thereby sharing and shaping the construction of learning
  • Leaders should set aside space and time for collective learning conversations—both conversations about learning, and conversations where learning takes place7

No metaphor is perfect or complete; any metaphor will invariably highlight some aspects of a phenomenon, at the expense of others. But seeing learning as a garden, inhabited and shaped by learners, can be a first step in updating our approaches: enriching our language, and eventually improving our actions and decisions as genuinely learning organisations.


Aaron Maniam is currently Deputy Secretary (Industry & Information) at the Ministry of Communications and Information, coordinating digitalisation policy, and concurrently Chief of the inter-agency Global Positioning Strategy Office. He was the first Head of the Centre for Strategic Futures (2009–2011) and Institute Director of CSC's Institute of Public Sector Leadership (2011–2013), where he started CSC's Applied Simulation Training Laboratory.



  1. P. Freire, Pedagogy of the Oppressed (London: Penguin Books Ltd, 1968).
  2. J. M. Keynes, The General Theory of Income, Interest and Money (Hinsdale: Dryden Press, 2016).
  3. M. Goldsmith, What Got You Here Won’t Get You There (London: Profile Books Ltd, 2012).
  4. C. Dweck, Mindset: Changing the Way You Think to Fulfil Your Potential (updated edition) (London: Little Brown Book Group, 2017).
  5. Video available at, accessed September 25, 2021.
  6. Ibid.
  7. Such approaches might draw from Daniel Kim’s Core Theory of Success, which sees a dynamic and mutually reinforcing relationship between the quality of relationships, collective thinking, actions and results in an organisation. Details at: D. H. Kim, Introduction to Systems Thinking (Massachusetts: Pegasus Communications, 1999).

Back to ETHOS homepage