Learning together: How a Human Learning Systems approach helps services work with complexity

Whether someone can find decent work depends on where they live, their age, what education they had, whether there is transport, whether they have a disability or chronic condition, the changing nature of work and advancement of AI, and the prevailing economic conditions. No single department owns that question, and no single service can answer it, which is a reasonable description of most of what public services deal with (1).

What Human Learning Systems means

The Human Learning Systems approach is a response to this, and the name carries the argument in three words. Human means treating people as individuals rather than as a category with a standard response attached. Learning means accepting that we cannot know in advance what will work, so the work is to keep finding out. Systems means recognising that no service can achieve much alone, because people's lives run across all of them at once (2).

The learning cycle

In practice this takes the shape of a cycle. We try to understand what is producing the current outcomes, design an experiment with the people involved, carry it out, make sense together of what happened, and adapt practice and resources in light of it, and then the whole thing starts again (3).

Each turn of it is meant to create slightly better conditions for people to do well, and to connect what one person learns to what a team and an organisation learn. In reality trust, data quality and institutional pressure all determine how much genuine experimenting is possible, so what progress looks like is usually partial and uneven.

Whose learning counts

There is a question underneath all of this, which is whose learning we are talking about. Citizens, public-facing workers, managers and system leaders all need structured opportunities to look together at what is happening, what is changing in people's lives, and what any of it means, and it rarely works if only one of those groups is in the room (4).

Front-line workers exercise wide discretion under conditions of high demand and limited resources, so their everyday sense-making becomes public policy in practice, regardless of what the written policy says (5). That makes reflective learning at the point where the work happens more consequential than most organisations treat it.

An Irish example

F:ACES, a pilot on child poverty in families with young children, runs across Bray, North Wicklow and Dún Laoghaire-Rathdown (6). By mapping how the agencies involved interact, it surfaced gaps and duplications that no service could see from inside its own remit.

Building a community of practice around that map has let professionals co-develop responses together. The emphasis has moved from how each service was performing to the quality of the connections between them.

Making room for learning

Learning of this kind asks something of organisations that is genuinely uncomfortable, which is a tolerance for not knowing yet, and protected time for people to think together when there is always something more urgent to do.

We will be working through what this looks like in practice at the EPEx executive programme, Harnessing Relational Public Services for Better Social Outcomes, at UCD on 23 and 24 September, which I am co-facilitating with Professor Toby Lowe. To join us, contact epex@ucd.ie.

Dr Joan O'Donnell is an Irish systems thinking expert, consultant, lecturer on the Open University's MSc Systems programme, and an advanced facilitator of Social Presencing Theatre (MIT). She is a consultant on the Empowering Policy Excellence (EPEx) Executive Education initiative at the UCD Geary Institute for Public Policy and a fellow of SCiO, the global systems practitioner organisation.

This piece draws on Reframing Public Services in Ireland as a Relational Practice, written with Toby Lowe, Gabriel Eichsteller and Adele O'Neill and published by UCD Geary Institute for Public Policy in March 2026 (7).

References

1.  On messy situations and interdependence, see Ackoff, R. L. (1974). Redesigning the future. Wiley; and Boulton, J. (2024) and Reynolds, M. (2020) on working with complexity in practice.

2.  Lowe, T., French, M., Hawkins, M., Hesselgreaves, H., & Wilson, R. (2020). New development: Human learning systems, public service for complexity, uncertainty and citizen-centred public services. Public Money & Management, 40(3), 193-196.

3.  Human Learning Systems. www.humanlearning.systems. See also Centre for Public Impact. (2022). Human learning systems: A practical guide to doing public management differently.

4.  Lowe, T., & Centre for Public Impact. (2021). Leading public service systems: Human learning systems in practice.

5.  Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the individual in public services (30th anniversary ed.). Russell Sage Foundation.

6.  F:ACES (Family: Adult and Child Enhanced Services), Bray and North Wicklow Area Partnership. www.brayandnorthwicklowareapartnership.ie

7.  O'Donnell, J., Lowe, T., Eichsteller, G., & O'Neill, A. (2026). Reframing public services in Ireland as a relational practice. UCD Geary Institute for Public Policy. https://publicpolicy.ie/papers/reframing-public-services-in-ireland-as-a-relational-practice/