Events, Lives, and Systems: The Subject Matter of Public Services
The previous post established what kinds of objects public services engage with - patients, cases, assessments, entitlements, relationships - and the different ways those objects can be defined, from formal entities in conceptual models to autonomous agents in Promise Theory. Objects provide the things that occupy state spaces; their properties define the axes of those spaces; their persistence conditions determine what it means for something to be tracked across time.
This post turns to the other half of the ontological picture: the events that happen to those objects. Services are fundamentally about happenings - events that unfold over time, changing the states of people and systems. The broader question motivating this series remains: how does one formalise the representation of dynamic social situations so that they become amenable to structured analysis, and ultimately to algorithmic reasoning? The Pathway Generator encodes patients health situation as vectors and computes recommended or suggested pathways, but it presupposes a representation of the problem that someone had to construct. The question is how such representations get built - and what conceptual infrastructure is required.
This post develops the event side of that infrastructure. It draws on cognitive science (how humans perceive and represent events), life-course theory (how events accumulate across a lifetime), and ecological systems thinking (how events occur in nested contexts). These are not the only frameworks available - Iqbal's service grammar and Harel's statecharts, which appear later in this series, offer complementary formalisations - but together they provide the conceptual apparatus for understanding what public services engage with, and what any state-space representation must be capable of capturing.
From Objects to Events
The previous post surveyed what counts as an "object" across several theoretical traditions. Objects are the things that exist - with properties, persistence conditions, and (in Promise Theory) intent. Events are the things that happen - things that unfold over time, involving change. As the Handbook of Conceptual Modeling puts it, a process "is typically a transient, temporal thing. It 'happens' or 'occurs' to an object rather than something that 'exists' in the same sense that a physical object exists".
Services are fundamentally event-based. They are not just collections of objects (forms, systems, offices) but patterns of events (encounters, assessments, decisions, transitions). Understanding services requires understanding events - what they are, how they are structured, and how they accumulate across lives and contexts.
Gärdenfors: One Cognitive Model of Events
There are several productive ways of formalising events. Peter Gärdenfors, whose conceptual spaces theory provides a geometric foundation for meaning, offers a cognitive model that is useful for understanding how people mentally represent what happens to them - and therefore what any service representation must account for. It is not the only such model, but it provides a vocabulary that connects naturally to the force dynamics of service encounters.
Gärdenfors proposes what he calls the "two-vector condition on events":
"An event must contain at least two vectors and one object; these vectors are a result vector representing a change in properties of the object and a force vector that causes the change".
An event, in this model, has:
- A patient: the object (or person) to whom something happens
- A force vector: what causes the change (an action, a natural force, a decision)
- A result vector: how the patient's properties change
Consider a rehabilitation event: a caseworker refers a patient to vocational training. The patient is the person receiving the service. The force vector is the caseworker's action of making the referral. The result vector is the change in the patient's situation - now enrolled in training, with different possibilities ahead.
Gärdenfors extends this basic model to include:
- Agents: who or what generates the force vector (sometimes the patient themselves)
- Instruments: tools or intermediaries that modify the force
- Goals: what the agent intends to achieve
- Counterforces: resistance from the patient or environment
This gives us a rich cognitive vocabulary for events. A service encounter isn't just "something that happened" - it's a structured event with identifiable forces, agents, patients, and results.
Events Have Structure
Gärdenfors shows that events aren't atomic - they have internal structure:
"Events can be decomposed into co-occurring or parallel subevents using the dimensions of the patient space... events can be segmented sequentially by path subcomponents".
An assessment event, for instance, might involve:
- Gathering information (force: questions; result: knowledge acquired)
- Making a judgment (force: analysis; result: classification)
- Communicating a decision (force: speech act; result: patient now knows their status)
These are nested events within a larger event. Understanding this structure matters for service design - different components might require different design attention.
Events also have aspectual properties:
- Telic events have endpoints (completing an application)
- Atelic events are ongoing without definite endpoints (receiving support)
- Durative events extend over time (rehabilitation)
- Instantaneous events happen in a moment (a decision)
Services mix these aspects. A rehabilitation journey is durative and (hopefully) telic - it takes time and aims at an endpoint. But within it are instantaneous events (decisions) and atelic processes (ongoing support).
Semantic Roles: Who Does What to Whom
Linguistic analysis of events identifies "thematic roles" - recurring relationships between participants and events. Gärdenfors maps these onto his cognitive model:
"I therefore turn to a description of further elements of the model that may be added to account for various aspects of the semantics of verbs. First, one should consider the role of counterforces. The force vector of an event construal will change one or more properties of the patient".
Key thematic roles in service contexts include:
Patient/Theme: The person or thing whose properties change. In vocational rehabilitation, this is typically the person receiving services - but might also be a case file, a benefit status, or a placement.
Agent: Who or what generates the force that causes change. A caseworker making a referral. A system generating a recommendation. A policy mandating an assessment.
Instrument: Tools or intermediaries used to exert force. An assessment instrument. A digital platform. A funding mechanism.
Recipient: Who receives the result of the event. Sometimes distinct from the patient - a report might change a patient's status, but the recipient is the decision-maker who reads it.
Goal: The intended endpoint of intentional action. Return to work. Stable housing. Reduced symptoms.
This vocabulary helps us analyse what services actually do. A referral event has an agent (caseworker), a patient (person being referred), an instrument (referral form, system), and a goal (access to specialist service). Making these roles explicit clarifies who has agency, what's being changed, and what the intended outcome is.
Life Events: The Citizen Perspective
From the citizen's perspective, public services are encountered at particular moments in life - often moments of transition, vulnerability, or need. These are "life events" - significant happenings that trigger the need to have interactions with public services. Examples include:
Life events are the natural organising principle from a citizen perspective:
- Having a child
- Losing a job
- Becoming ill
- Moving house
- Retiring
- Bereavement
Each life event triggers cascades of needs that span multiple government agencies. Designing around life events - rather than around departmental structures - is a way of aligning services with how people actually experience their lives.
Lou Downe's Good Services takes a similar perspective:
"It's services that are the interface to so much of our experience of the world. From having children, getting married, moving house and, sadly, dying - all these are services".
The language of "life events" connects cognitive event perception (Gärdenfors) to public service design (Downe). Both are about events - happenings that unfold over time, involving agents and patients, causing changes in people's properties and possibilities.
The Life Course: Events Accumulate
Life events don't happen in isolation. They accumulate across a lifetime, with earlier events shaping the possibilities for later ones. This is the domain of life-course theory.
The Handbook of Research Methods in Health Social Sciences describes life-course research as:
"Life course theory brings together Bronfenbrenner's conception of multilevel social context with sociological understandings of the life course as a sequence of socially defined, age-graded events and roles".
Key concepts from life-course theory include:
Trajectories: Long-term patterns of stability and change. A career trajectory. A health trajectory. An education trajectory.
Transitions: Changes in state or role. Becoming unemployed. Entering rehabilitation. Returning to work.
Turning points: Events that substantially alter trajectories. An injury that prevents work. A training programme that opens new possibilities. A diagnosis that redefines what is possible.
Cumulative disadvantage: How early events create conditions for later disadvantage. Childhood health problems → educational disruption → limited employment → poor adult health.
For vocational rehabilitation, this perspective is significant. The person presenting for rehabilitation has a history - a trajectory of events that led to their current situation. The Pathway Generator's patient vector is one attempt to encode this trajectory computationally: it represents a person's situation as a set of variables (health status, employment history, psychological measures, functional capacity) that the algorithm uses to recommend pathways. But those variables are themselves codifications of life-course events - each score on a depression inventory or work-capacity assessment is a snapshot of where a trajectory currently stands. The question of which variables to include, and how to encode them, is a design decision that shapes what the algorithm can see and what it cannot. My earlier review of rehabilitation conceptual models documented the range of frameworks that different research traditions use to codify these trajectories, and the significant variation in what each treats as salient.
Ecological Systems: Nested Contexts
Events don't just occur in time - they occur in context. Urie Bronfenbrenner's ecological systems theory provides a framework for understanding the nested contexts within which development and change happen.
Bronfenbrenner identifies levels:
Microsystem: Immediate settings where the person directly participates. Home, workplace, rehabilitation clinic, caseworker relationship.
Mesosystem: Interactions between microsystems. How the person's family relates to their treatment programme. How their workplace relates to their health condition, or the articulation of it.
Exosystem: Settings the person doesn't directly participate in but that affect them. Local labour market conditions. Policy decisions. Organisational restructuring.
Macrosystem: Broader cultural and societal patterns. Attitudes toward disability. Economic conditions. Welfare state structures.
For service design, this means recognising that interventions at one level interact with conditions at other levels. A micro-level intervention (caseworker support) operates within meso-level conditions (coordination between services), which operate within exo-level conditions (funding regimes, local employment opportunities), which operate within macro-level conditions (cultural attitudes, economic climate).
The SCÖ project failed to deliver the changes it promised or claimed it would - mostly because it ran out of funding - but partly because it didn't account for these different layers of systemic complexity, and attempted to heavily leverage a micro-level invention (patient information) while ignoring, or failing to grapple with meso-level realities (coordination barriers between agencies) and exo-level constraints (funding structures, political conditions). Or, put another way, it failed because it leveraged the assumption that a reified algorithm could solve coordination problems without addressing the underlying systemic issues required to unlock or gather the data that the algorithm required to function, or even be speculatively tested.
The Biopsychosocial Model: Integrated Understanding
George Engel's biopsychosocial model, now foundational in many healthcare and rehabilitation settings, provides another framework for understanding the complexity of what services engage with. I surveyed these models in detail in an earlier post synthesising the vocational rehabilitation literature - here I want to situate them within the cognitive science of events:
"The biopsychosocial model of health, as conceptualised by Engel (1977), recognised that illness and ill-health are influenced by a person's psychological state as well as their social environment".
Health and functioning aren't just biological - they're the result of interactions between:
Biological factors: Body structure and function, disease processes, genetic predispositions
Psychological factors: Cognition, emotion, motivation, coping strategies, self-efficacy
Social factors: Support networks, employment conditions, housing, economic resources
The WHO's International Classification of Functioning, Disability and Health (ICF) operationalises this model:
"Within the ICF model, disablement is conceptualized as a difficulty encountered with any or all interconnected functional areas: impairments at the level of body structure and function, activity limitations at the individual level, and participation restrictions at the social level".
The ICF is itself a codification scheme - it assigns codes to body functions, activity types, participation domains and environmental factors, creating a structured vocabulary for what are, in Gärdenfors's terms, different aspects of the patient's state. Note that some of what the ICF encodes are durative conditions (chronic pain, reduced mobility), some are telic processes (completing a training programme, achieving stable employment), and some are atelic ongoing states (receiving support, participating in community life). The aspectual properties of events discussed earlier are not merely cognitive abstractions; rehabilitation practice already distinguishes between them, because different event types demand different interventions. A durative condition requires sustained management; a telic goal requires a planned sequence of actions; an atelic process requires conditions that allow it to persist. The rehabilitation models I reviewed earlier each codify these distinctions differently - the Swedish work-capacity assessments, the Icelandic JANUS variables, the BIP framework - and the choice of codification shapes what becomes visible and what remains hidden.
For vocational rehabilitation, this means that "returning to work" is not a single event but a complex of events across multiple dimensions - biological (physical capacity), psychological (motivation, self-efficacy), and social (job availability, employer attitudes, welfare structures). Services must engage with the person as a biopsychosocial whole situated in ecological context.
Events as the Fundamental Unit
Pulling these threads together, we can see why events are the fundamental unit for understanding public services, and particular health and social care services like vocational rehabilitation:
Cognitively: People perceive and represent their experiences as events - structured happenings with agents, patients, forces, and results. Services are experienced as events.
Temporally: Services unfold over time. They're not static objects but dynamic processes involving sequences of events, transitions between states, and accumulation of effects.
Contextually: Events occur in nested ecological contexts. What happens in a service encounter is shaped by - and shapes - conditions at multiple systemic levels.
Biopsychosocially: The events that services engage with - illness, unemployment, crisis - are complex phenomena involving biological, psychological, and social dimensions simultaneously.
This is why state-space thinking matters for services, and why the question of how one constructs a state space is not merely technical but ontological. A state space captures the possible configurations of a person's situation. Events are transitions between states. Goals are target states. A service is a system for generating events that move people toward goal states. But the choice of what to include in the state representation - which dimensions, which event types, which codification scheme - determines what the system can reason about. The Pathway Generator encodes patient situations as vectors; the ICF encodes functioning as a matrix of codes; the BIP framework encodes work-capacity assessments as structured indicators. Each is a different formalisation of the same underlying reality, and each makes different things computable at the cost of making other things invisible.
Implications for Design
What does this ontological analysis mean for service design?
Design for events, not just objects. The physical and digital artefacts of a service (forms, systems, spaces) are important, but they're instruments. The fundamental design question is: what events do we want to create, and what events do we want to enable?
Understand event structure. Events have agents, patients, forces, results, goals. Make these explicit. Who has agency? What changes? What's the intended outcome? What resistance exists?
Respect life-course context. The person encountering a service has a history and a trajectory. Design should understand where events fit in that trajectory and how they might alter it.
Consider ecological levels. Micro-level interventions operate within meso, exo, and macro conditions. Design should consider what conditions at other levels enable or constrain what's possible at the service level.
Acknowledge biopsychosocial complexity. The events services engage with are multiply constituted. Interventions must address biological, psychological, and social dimensions - or at least understand how they interact.
Connection to What Follows
This post has explored the notion that services are fundamentally about events - cognitively structured happenings that occur in life-course trajectories within ecological contexts, engaging with biopsychosocially complex human situations. Together with the previous post on objects, it establishes the two ontological categories that any state-space representation must accommodate: the things that exist and the things that happen to them.
The next post develops Gärdenfors's conceptual spaces framework more fully - showing how both objects and events can be understood geometrically, with implications for how people represent meaning and achieve (or fail to achieve) shared understanding. Together with the object and event ontologies developed in these posts, this provides the cognitive foundations for understanding why "making things visible" doesn't automatically produce shared understanding - and what's required for coordination across difference.
Next: "From Cognitive Frameworks to Geometric Formalism" - why the series needs a geometric vector-space formalism, and what the competing cognitive science frameworks do and don't provide.
References
Gärdenfors, P. (2017). The Geometry of Meaning: Semantics Based on Conceptual Spaces. MIT Press.
Gärdenfors, P. (2000). Conceptual Spaces: The Framework of Thought. MIT Press.
Pope, R. (2024). Platformland: An Anatomy of Next-Generation Public Services. Ash Center, Harvard Kennedy School.
Downe, L. (2020). Good Services: How to Design Services That Work. BIS Publishers.
Bronfenbrenner, U. (1979). The Ecology of Human Development. Harvard University Press.
Engel, G.L. (1977). The Need for a New Medical Model: A Challenge for Biomedicine. Science, 196(4286), 129-136.
World Health Organization (2001). International Classification of Functioning, Disability and Health (ICF). WHO.
Elder, G.H. (1998). The Life Course as Developmental Theory. Child Development, 69(1), 1-12.
Schultz, I.Z. and Gatchel, R.J. (eds.) (2016). Handbook of Return to Work: From Research to Practice. Springer.
Embley, D. and Thalheim, B. (eds.) (2012). Handbook of Conceptual Modeling: Theory, Practice, and Research Challenges. Springer.