The "good services are verbs, bad services are nouns" heuristic - first articulated by Lou Downe at GDS in June 2015 - has become a powerful design principle across UK government and beyond. Yet this seemingly simple framing rests on deep theoretical foundations in linguistics, philosophy, and sociology, while also obscuring important dimensions of service reality that theories of materiality, practice, and institutions reveal.
Duncan Stephen's recent post, "Services are verbs combined with nouns", articulates a distinction I have been circling since writing about states and services last year. The original problematic service names ("RIDDOR", "Immigration Health Surcharge") failed because they were jargon, not because they were nouns. Every task is actually a verb combined with a noun: "Apply for a Blue Badge", "Check your National Insurance record".
The insight is deceptively simple, but it opens onto deeper questions about what verbs and nouns actually reveal - and what they systematically obscure. This post attempts a theoretical excavation: what intellectual traditions support the verb heuristic, and what do they leave out?
Origins: From GDS Culture to International Influence
Lou Downe's (2015) articulation emerged from practical GDS work addressing a specific problem: government services named for bureaucratic convenience that confused users attempting to accomplish goals. The blog post on GDS Design Notes framed the solution elegantly: "To a user, a service is simple. It's something that helps them to do something - like learn to drive, buy a house, or become a childminder. It's an activity that needs to be done".
The conceptual groundwork preceded Downe. The GDS Design Principles (April 2012) already stated that "a service is something that helps people to do something", and service-dominant logic from Vargo and Lusch (2004) had reframed services as "the application of operant resources for the benefit of another actor" - processes and activities rather than goods.
Lynn Shostack's (1984) service blueprinting had already structured services around "customer actions" as a core swim lane, and the Jobs to Be Done framework emphasised that people "hire" products to accomplish activities. Downe (2020) synthesised these into a memorable heuristic, later codified in Good Services: How to Design Services that Work.
Linguistic and Philosophical Foundations
Speech act theory provides the deepest theoretical support for verb-focused design. J.L. Austin's (1962) How to Do Things with Words established that language is fundamentally action - performative utterances do not describe reality but create it. When a service says "Your application is approved", that is not description but declaration, creating the institutional fact it names. John Searle's (1969) taxonomy of speech acts - assertives, directives, commissives, expressives, declarations - maps directly onto service interactions: requests from users, commitments from providers, confirmations that change status.
As I explored in the promises and affordances piece, services can be understood as bundles of speech acts creating and discharging commitments. Winograd and Flores (1986) applied this directly to systems design through their "Conversation for Action" framework, modelling work as networks of speech acts - anticipating service design's focus on promises, requests, and satisfactions.
Charles Fillmore's (1968) case grammar contributes another lens: verbs encode semantic relationships between agents, instruments, and affected entities. "Book a flight" implies an agent (user), an instrument (booking system), and an affected object (the reservation). Understanding these argument structures helps designers identify what roles services must support. Indeed, as Clark and Clark (1979) demonstrate, nouns routinely surface as verbs through a productive convention that exploits precisely these case relationships - a process that dissolves the sharp noun/verb boundary the heuristic presupposes.
Yet philosophy also reveals what verb-focus loses. Hannah Arendt's (1958) The Human Condition distinguishes three fundamental human activities: labour (cyclical, consumption-driven activity tied to biological necessity, leaving nothing behind), work (fabrication of durable objects constituting the human-built world), and action (activity between people revealing identity, inherently unpredictable and political).
When service design reduces everything to "action" (verbs), it obscures whether an activity contributes to mere survival-maintenance, creates durable outcomes, or enables genuine interpersonal encounter. A healthcare service might include elements of all three - and treating them equivalently as "actions" misses their qualitatively different meanings for human flourishing.
Sociological Theories: What Surrounds the Verb
Max Weber's (1978) typology of social action - instrumental-rational, value-rational, affective, and traditional - immediately complicates simple verb descriptions. The "same" action of "paying taxes" differs fundamentally when performed as strategic calculation, civic duty, fearful compliance, or unthinking habit. Service designers who focus only on what users do miss why and how that matters to them.
Harold Garfinkel's (1967) ethnomethodology goes further, showing that actions are indexical - their meaning depends entirely on context and cannot be abstracted into context-free verb descriptions. "Checking your record" means radically different things in a benefits context versus criminal justice. Actions are also sequentially organised, as conversation analysis demonstrates: each action is produced in light of prior actions and anticipation of responses. Service encounters unfold through turn-taking, adjacency pairs (question-answer, request-grant/decline), and repair mechanisms that simple verb lists obscure.
Practice theory offers perhaps the most developed critique, drawing on a tradition running from Bourdieu's (1977) theory of practice through Schatzki's (1996) Wittgensteinian approach to Shove, Pantzar, and Watson's (2012) The Dynamics of Social Practice, which argues that practices - not individuals or behaviours - are the proper unit of analysis. Practices are constituted by three interconnected elements: materials (objects, infrastructures, tools, technologies, and the body), competences (know-how, shared understandings, skills), and meanings (social significance, emotions, aspirations).
The ABC model (Attitudes to Behaviour to Choice) that underlies much policy assumes individual attitudes drive behaviours. Practice theory inverts this: behaviours are embedded in socio-material configurations that transcend individual choice. Showering practices changed not because attitudes changed, but because piped hot water, bathroom design, and cleanliness norms co-evolved.
For service design, this means attending to the materials and competences alongside actions - and understanding that services recruit practitioners into careers of increasing commitment.
Activity Theory and Cognitive Work Analysis
Activity theory, rooted in Vygotsky's (1978) cultural-historical psychology, offers a more structured framework than flat verb descriptions. Leontiev's (1978) hierarchy distinguishes three levels: activities (oriented to motive, answering WHY), actions (oriented to goals, answering WHAT), and operations (oriented to conditions, answering HOW). A "verb" typically captures the action level - but understanding requires connecting to the motivating activity above and the automatised operations below. Leontiev's hunting example illustrates: the beater who scares game away performs an action seemingly opposed to the goal, but it only makes sense within the collective activity of hunting where beaters drive prey toward ambushers.
Engeström's (1987) activity systems model expands this to collective activity, adding community, rules, and division of labour to the subject-tool-object triangle. Services can be analysed as activity systems with contradictions driving development - tensions between elements that create design opportunities.
Cognitive Work Analysis, building on Rasmussen's (1983) skills-rules-knowledge framework for human performance, offers perhaps the most directly applicable framework. Vicente's (1999) Abstraction Hierarchy represents the work domain at five levels: functional purpose (why does this exist?), abstract function (what principles govern it?), generalised function (what processes occur?), physical function (what can components do?), and physical form (what does it look like?).
The critical insight is that Work Domain Analysis represents invariant constraints, not tasks. Moving down the hierarchy answers HOW; moving up answers WHY. This fundamentally transcends verb-focused design by revealing the means-ends structure that any particular action serves. "Book an appointment" (generalised function) only makes sense connected to functional purpose and constrained by abstract functions and physical resources. As Vicente (1999) emphasises, it is easy to confuse Work Domain Analysis with task analysis, but WDA captures what remains true regardless of specific tasks, events, or operator strategies - the "stage" rather than the "play".
This connects directly to the state space thinking I have been developing. Statecharts foreground what condition something is in at any given moment, and what events can trigger transitions - a perspective service design rarely adopts explicitly, despite the fact that services are fundamentally about transforming the state of things.
Material and Relational Theories: What Verbs Obscure
Where the sociological critiques above attend to the social context surrounding action, a different family of theories questions the assumption that action is exclusively human in the first place. These material and relational approaches argue that agency is distributed, that objects are active participants, and that the capacity to act emerges from configurations of humans and things rather than from human intention alone.
Actor-Network Theory fundamentally challenges human-centred verb descriptions. Bruno Latour (2005) and Michel Callon (1986) argue that agency is distributed across heterogeneous networks of human and non-human actants. A speed bump "acts" as effectively as a police officer to slow traffic. For service design, this means services are not just human actions but assemblages where technologies, documents, physical spaces, and regulatory systems all have agency.
When services fail, it is often because non-human actants resist or destabilise the network. Callon's (1986) four "moments of translation" - problematisation, interessement, enrolment, mobilisation - describe how networks form and stabilise around shared objects, and how they can unravel when actants refuse their assigned roles.
Jane Bennett's (2010) Vibrant Matter pushes further than ANT's symmetrical treatment of humans and non-humans. Her "vital materialism" argues that matter itself has vitality, efficacy, and a kind of agency that exceeds human intention. Things possess what she calls "thing-power" - the capacity to impede, enable, or redirect human action in ways that cannot be fully predicted or controlled.
The implication for service design is that the materials constituting services are not inert substrates for human action but active participants with their own tendencies and recalcitrance. A database does not just store data; it shapes what queries are possible, what connections are visible, what patterns emerge. An interface does not just present options; it invites certain actions while discouraging others through its material configuration. The verb "book an appointment" presupposes a booking system with its own materiality - its response times, its error states, its affordances and resistances. The action emerges from the encounter between human intention and material configuration, not from human will alone.
James Gibson's (1979) ecological psychology inverts the verb question entirely, asking not "what do users do?" but "what does the environment offer?" Affordances are action possibilities that exist in the relationship between organism and environment, directly perceived rather than cognitively inferred. Donald Norman (1988) adapted this for design: good design makes affordances discoverable. Services must invite appropriate actions through their material qualities - not just respond to user intent. This connects to the triadic model of affordance I explored in the promises piece: affordance as the conjunction of physical availability, social accessibility, and perceived invitation.
Graham Harman's (2002, 2011) object-oriented ontology offers a different critique of relational thinking - including, interestingly, a critique of ANT itself. Where Latour defines objects entirely through their relations and effects, Harman argues that objects always withdraw from their relations; there is always more to an object than what it reveals in any particular encounter. This is not mysticism; it is recognition that our service models are always partial, that the objects we manipulate have properties and potentials our designs do not acknowledge. As I explore more fully in the objects post, this withdrawal is a useful corrective to the instrumentalism that can creep into service thinking.
Susan Leigh Star's work on boundary objects (Star and Griesemer, 1989) reveals how objects mediate between communities of practice with different understandings. A patient record means different things to clinicians, administrators, and patients - yet serves as a boundary object enabling coordination without consensus. As I explored in the boundary objects post, the distinction between boundary objects (enabling coordination through productive ambiguity) and exposure devices (forcing confrontation with difference) matters for how design artefacts function in practice.
Bill Brown's (2001) thing theory emphasises that objects typically recede into background use until they break down, malfunction, or surprise us - at which point they "assert themselves as things". This is precisely where design opportunities emerge: the moments of breakdown reveal hidden assumptions and dependencies that smooth verb descriptions obscure.
Noortje Marres (2012) extends material thinking into explicitly political territory. In Material Participation, she argues that objects do not merely mediate human politics - they actively constitute what participation means and who can participate. A smart meter does not just measure energy use; it configures a particular form of participation - the householder as rational economiser, the home as site of optimisation. For service design, Marres suggests that the material infrastructure of services is not neutral scaffolding for human action but actively shapes what kinds of action, what kinds of users, and what kinds of outcomes are possible. A benefits application form does not just collect information; it constitutes the applicant as a particular kind of subject. The materials are political.
Institutional Analysis: What Constrains the Possibility of Verbs
Elinor Ostrom's (2005) Institutional Analysis and Development framework identifies seven types of rules shaping action arenas: boundary rules (who can participate), position rules (what roles exist), choice rules (what actions are allowed), information rules (what participants know), aggregation rules (how individual actions combine), payoff rules (how benefits and costs distribute), and scope rules (what outcomes are possible). These rules do not just constrain actions - they constitute what actions are even conceivable. "Apply for housing benefit" is only possible within an institutional ecology that defines eligibility, establishes application procedures, and allocates decision authority. Verb-focused design risks taking this institutional infrastructure as given background rather than design material.
What the Heuristic Gets Right - and What It Obscures
The "good services are verbs" heuristic remains valuable as a corrective to bureaucratic noun-naming that obscures user purposes. It successfully reoriented government digital services toward action and outcomes. Services do help users accomplish activities, action-oriented naming aids findability, and verb framing cuts through institutional jargon to user goals.
But it obscures the material infrastructure persisting between actions, the institutional rules constituting what actions are possible, the practice configurations bundling materials, competences, and meanings, the multiple abstraction levels connecting actions to purposes, the non-human actants with their own agency in service networks, and the states that entities can be in and the transitions between them.
The strongest critique is not that verbs are wrong, but that they are incomplete. A service is not just what users do - it is what objects afford, what institutions enable, what practices recruit into, what networks assemble, and what purposes different abstraction levels serve. Duncan Stephen's conclusion resonates: "by codifying our verbs as service patterns and nouns as information models, we have the potential to unleash great efficiencies". But the fuller picture is something like: verb plus noun plus adjective plus preposition plus institution plus practice plus network plus means-ends hierarchy.
The elegance of "good services are verbs" was always a simplification - useful for cutting through bureaucratic jargon, powerful as a reorienting heuristic, but incomplete as an analytical framework for complex sociotechnical systems. This connects to the grammar of services I sketched in an earlier post. The question is not just what users do (verbs) or what things exist (nouns), but how these combine with states, transitions, channels, and evidence into well-formed service specifications.
References
Arendt, H. (1958). The Human Condition. University of Chicago Press.
Austin, J.L. (1962). How to Do Things with Words. Oxford University Press.
Bennett, J. (2010). Vibrant Matter: A Political Ecology of Things. Duke University Press.
Bourdieu, P. (1977). Outline of a Theory of Practice. Cambridge University Press.
Brown, B. (2001). Thing theory. Critical Inquiry, 28(1), 1-22.
Callon, M. (1986). Some elements of a sociology of translation. In Law, J. (ed.) Power, Action and Belief. Routledge.
Clark, E.V. and Clark, H.H. (1979). When nouns surface as verbs. Language, 55(4), 767-811.
Downe, L. (2015, June 22). Good services are verbs, bad services are nouns. GDS Design Notes.
Downe, L. (2020). Good Services: How to Design Services That Work. BIS Publishers.
Engeström, Y. (1987). Learning by Expanding: An Activity-Theoretical Approach to Developmental Research. Orienta-Konsultit.
Fillmore, C.J. (1968). The case for case. In Bach, E. and Harms, R. (eds.) Universals in Linguistic Theory. Holt, Rinehart and Winston.
Garfinkel, H. (1967). Studies in Ethnomethodology. Prentice Hall.
Gibson, J.J. (1979). The Ecological Approach to Visual Perception. Houghton Mifflin.
Harman, G. (2002). Tool-Being: Heidegger and the Metaphysics of Objects. Open Court.
Harman, G. (2011). The Quadruple Object. Zero Books.
Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford University Press.
Leontiev, A.N. (1978). Activity, Consciousness, and Personality. Prentice Hall.
Marres, N. (2012). Material Participation: Technology, the Environment and Everyday Publics. Palgrave Macmillan.
Norman, D.A. (1988). The Design of Everyday Things. Basic Books.
Ostrom, E. (2005). Understanding Institutional Diversity. Princeton University Press.
Rasmussen, J. (1983). Skills, rules, and knowledge: Signals, signs, and symbols, and other distinctions in human performance models. IEEE Transactions on Systems, Man, and Cybernetics, 13(3), 257-266.
Schatzki, T.R. (1996). Social Practices: A Wittgensteinian Approach to Human Activity and the Social. Cambridge University Press.
Searle, J.R. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge University Press.
Shostack, G.L. (1984). Designing services that deliver. Harvard Business Review, 62(1), 133-139.
Shove, E., Pantzar, M. and Watson, M. (2012). The Dynamics of Social Practice: Everyday Life and How It Changes. SAGE.
Star, S.L. and Griesemer, J.R. (1989). Institutional ecology, 'translations' and boundary objects. Social Studies of Science, 19(3), 387-420.
Stephen, D. (2025, June 26). Services are verbs combined with nouns. Duncan Stephen. https://duncanstephen.net/services-are-verbs-combined-with-nouns/
Vargo, S.L. and Lusch, R.F. (2004). Evolving to a new dominant logic for marketing. Journal of Marketing, 68(1), 1-17.
Vicente, K.J. (1999). Cognitive Work Analysis: Toward Safe, Productive, and Healthy Computer-Based Work. Lawrence Erlbaum.
Vygotsky, L.S. (1978). Mind in Society. Harvard University Press.
Weber, M. (1978). Economy and Society. University of California Press.
Winograd, T. and Flores, F. (1986). Understanding Computers and Cognition. Addison-Wesley.