Organizing Constructs
Most concepts fail when we try to coordinate action and theory across levels. They either flatten into vibes or fragment under pressure. Organizing Constructs are tools that don’t.
Disclaimer: This is novel work with minimal empirical grounding. I’m fleshing out an idea that’s been useful in my own thinking, but I don’t yet know how far it extends.
Two decades ago, empowerment was everywhere. It was an idea you could see in a grant application, a grassroots zine, and a Nature article in the same day. For a while, that was its strength. It allowed researchers and practitioners to talk about personal, community, and systemic change in a single breath, and offered energizing, accessible language to community members.
Eventually, though, the term collapsed under its own weight. Critics rightly called out its vagueness, its co-optation, and its use as rhetorical cover for power imbalances and chauvinism. Empowerment became, too often, a feel-good floating signifier. Able to mean almost anything, it stopped meaning much at all1.
But the reason it spread in the first place was important: it met a vital, underappreciated need. In social life, big problems span many levels at once. Therefore, our attempts to solve them need to hold for individuals, communities, institutions, and sometimes even societies. When a concept does this well, and when it is backed up with the sort of structure that keeps it from drifting, I call it an organizing construct.
What does it mean to hold in this way? What sort of structure keeps meaning stable? I’ll start with an example, then discuss what makes an organizing construct work in theory and, speculatively, in practice.
Mattering as an Organizing Construct
I developed the idea of organizing constructs while trying to solve a specific problem in my dissertation: why did mattering feel so much more practical than related concepts like empowerment?
Mattering is just what it sounds like: the sense that we are significant, that we are valued, and that we have an impact on the world around us. It’s a construct that has a significant, if relatively quiet, history in sociology and psychology. It has been connected empirically to everything from suicidality to overcoming stigma.
Multidimensional Mattering (MDM) is the name I gave to the comprehensive version of mattering measured by Isaac Prilleltensky’s MIDLS (Mattering in Domains of Life) scale2. This scale assesses an individual’s sense of mattering in different areas of their life. Do you feel significant in close relationships, at work, in your community, and to yourself? This scale assesses mattering in each of those areas, but it also measures a general sense of mattering. MDM is the total construct comprised of both domain-specific and overall mattering.
You may have already realized that MDM is inherently cross-level3. This is part of why it makes a good organizing construct. Because it exists naturally at the level of the self, relationships, and communities, it can naturally connect to ideas and theories at each of those levels. The same construct can connect at the individual level to, say, internalized shame, while engaging with sense of community at the neighborhood level.
But there’s also another reason. Baked into the definition are two distinct, yet related, experiences: feeling valued and adding value. Something only truly counts as mattering, under this formulation, if it includes both4. As I’ll talk about below, I suspect this bifurcation is key to the stability that characterizes a strong organizing construct.
Defining Organizing Constructs
In my dissertation I suggested that mattering maintains meaning across levels without dissipating. But what makes it different from empowerment, which also stretched across contexts? Why does mattering hold up when similar concepts drift into vagueness? The answer lies in the specifics of organizing constructs.
Let’s start with a definition. An organizing construct is a conceptual tool that preserves meaning across individual, organizational, community, and/or societal levels by bridging theoretical frames in accessible language. Its power lies in its ability to align action, communication, and decision-making across domains.
What kind of decision-making? Think of an initiative entitled Healthy St. Louis. Polls show health is a popular cause across the city. Studies suggest there’s huge opportunity to improve health across the population. Even better, analysts find that better health could boost economic outcomes. It feels like a win-win-win… until everyone gets in a room and tries to figure out what to actually focus on.
The mayor thinks people should eat more vegetables. The hospital advisory council cites data suggesting that reduced alcohol consumption would have the biggest impact. Meanwhile, ward leaders argue that pollution and chronic stress are the real threats, and that without addressing those, everything else is just window dressing. They’re all pointing at health, but they’re doing it at different levels, with different levers.
They argue for a while. They eventually land on a single metric that nobody can really oppose: people empowered. It’s vague, but it’s uplifting and signals intention, and it gives them something to track while they figure out the rest. Twelve months later, they report a robust number of lives empowered. The only problem is: nobody is really sure that people are healthier5.
This example illustrates what happens when we mistake shared language for shared structure. Healthy St. Louis had rhetorical force, but without an organizing construct to stabilize its meaning, it couldn’t hold across levels (individual, community) or stakeholder perspectives (mayor, health expert). As a result, it ended up stranded between conflicting interpretations.
This type of half-resolution is common in the face of cross-level problems. “Empowerment” becomes a placeholder for alignment, not a functional operating logic. The stakeholders agree not because it clarifies action, but because it defers conflict. It’s vague enough that everyone can see their goals reflected6. But ultimately, the framing does very little to connect their perspectives.
What’s the problem?
Part of why finding something better is so challenging lies in the nature of scientific expertise. Science knows a ton about what makes people healthy, but scientific constructs are often designed for rigor within a narrow theoretical frame: a particular level, domain, or method. That’s what gives them reliability, but it also makes them brittle. When you move them outside their frame (say, from individual psychology to neighborhood dynamics), they leak both meaning and utility.
This is far from a new problem. Whole social science subfields have emerged to address it:
Composition models in organizational studies try to build coherent bridges between micro- and macro- by specifying in detail how aggregation works.
Boundary objects in sociology help us translate across communities via shared anchors.
Ecological metaphors in community psychology helps us acknowledge the relationship between levels and the complexity of their dynamics.
These solutions are sophisticated and often genuinely useful. However, in complex situations, they risk meaning loss. The rigor that makes them credible also strips away key details. They’re not wrong, exactly; they’re just general enough to pass. So it still looks like the theory is working, but it doesn’t clarify meaning, guide action, or connect levels.
Further, the more rigorous they are, the less portable they become. It’s like bringing a full studio production rig to a jam session. In theory, you have everything you need: perfect signal routing, multichannel mixing, and fine-grained control over every input. But in practice, it’s too rigid and too slow. You’re dragging around gear cases while everyone else is already playing. You optimize for precision instead of participation, and by time you’re ready, we’re on to the next song.
What’s missing is an accessible way to understand how meaning holds across levels simultaneously. Organizing Constructs do this because they exist naturally at multiple levels and make the relationships between those levels visible and actionable.
But what do they actually organize? They coordinate meaning across levels. This happens through four dimensions:
Shared language: The construct offers a shared term that means just enough of the same thing across levels to allow coordination without false consensus. When a youth social worker and a city planner both talk about mattering, they’re not necessarily saying the same thing, but they’re close enough to orient action.
Connective logic: Organizing constructs act as cross-level bridges. They don’t operate through clean cause-and-effect chains so much as through conceptual resonance. Take mattering: when people feel undervalued, that isn’t just the result of low community trust, or even its cause. It’s part of what low trust is. These constructs help us understand cross-level relationships that aren’t always simple, linear, or discrete but are still structurally meaningful.
Social recognition: Different stakeholders can locate themselves and their concerns within the construct. The mayor, hospital council, and ward leaders in our St. Louis example all need to see their worlds reflected.7
Normative direction (or relatable values): Organizing constructs carry normative energy without being values themselves. People mattering is good resonates from individuals to institutions, but the construct itself doesn’t dictate how that good should be achieved or just what it will look like. Mattering serves as a scaffold for values like dignity, fairness, or belonging, allowing them to travel without distortion.
A good organizing construct8 balances these four dimensions, allowing it to function not just as an idea, but as a bridge between levels and actors that can orient action and support theory.
Organizing Constructs and Overarching Values
But recognizing one organizing construct doesn’t tell us much about the rest of them. What separates an organizing construct from a concept that just seems like it should work? To answer that, we need to look at some other candidates: concepts that span levels but face different challenges.
If you’ve read the introduction or TOC of this blog, you’ll recognize care as an overarching value that guides my work. My concern with narrow self-interest, my interest in stewardship, my plodding introspection - all these are manifestations of care. It’s no surprise, then, that when I was writing this piece, care was one of the first constructs I looked at. What I learned was illuminating, both in how it fit and how it didn’t.
At first glance it seems promising. Care has clear multi-level relevance: it begins with self-care and showing up for each other. A similar spirit inhabits the community level: a community cares for its members by attending to their needs and providing support; and at the institutional level, this can easily be translated to policy. Even at the societal level, it’s possible to formulate an overarching ethic of care, as several scholars have done. At each of these levels, care takes a slightly different form: nurture, intimacy, inclusion, distributive justice. But at the core, each form involves some combination of attention and responsibility.
So far, so good. But when you try to coordinate a multi-level initiative around it (say, St. Louis Cares), you quickly discover stakeholders mean substantially different things: self-care, sympathy, redistribution, health access. Like empowerment, the concept dilutes rapidly. The relationship between interpersonal care (tending to a loved one) and systemic care (infrastructure and policy) requires constant translation.
Coordination becomes negotiation: which care? whose care? care as what? wait, is care a metaphor?
Care has many intellectual homes (feminist ethics, public health, mutual aid), each with slightly different meanings. That diversity is often a strength in scholarship, but it’s a liability when coordinating across roles or levels. Natural language associations pull too hard in too many directions.
Justice provides another example. It’s been the rallying cry of struggles and movements for millennia, but this broad resonance covers for a surprisingly complex and fragmented ideal. The study of justice has splintered into multiple forms (distributive, procedural), distinct cultural frameworks (Western rights-based, Indigenous relational models), and competing philosophical traditions (utilitarian, deontological). As a result, justice is invoked across radically incompatible political projects, from decolonization to white supremacy.9
Like care, justice functions primarily as an orienting value rather than an organizing construct. It tells us what matters morally, but doesn’t provide the operational stability needed to coordinate across levels. Ask three stakeholders to define just outcomes and you’ll get three reputable frameworks that can’t easily reconcile. The term is too contested and philosophically sprawling to serve as a practical coordination tool, even though it remains essential as a guiding value.
With mattering, on the other hand, you don’t need to check whether you’re talking about the same thing. I’ve tested this in dozens of conversations across wildly different contexts: with youth workers, evaluators, educators, therapists, drunk guys at soccer games. It almost always clicks quickly, whether we’re talking about an individual feeling unseen in their family or a neighborhood feeling ignored by city hall. That immediate portability is the hallmark of an organizing construct. Nail down the two-part structure and you can immediately use it to diagnose problems, guide design, and evaluate outcome10.
Does that mean care isn’t important? No, not at all. A thousand times, no. Rather, I think it’s so important that it comes in prior to organizing constructs, as an overarching orienting value. We’re only interested in mattering if we have an orientation of care, or some similar value that spurs us to acknowledge the significance of others.
So what’s the difference?
Overarching values tell us what matters. They orient us morally and normatively. Care, justice, belonging: these concepts guide our commitments and shape our priorities. They’re essential for alignment around purpose.
Organizing constructs help us coordinate action and analysis. They provide operational clarity across levels. Mattering, capabilities, and resilience help us understand problems, design interventions, and evaluate outcomes without constant translation. They’re essential for alignment around practice (including situated theory).
I think we get into trouble when we fail to distinguish here.11 We ask values like justice to do the work of organizing constructs and are disheartened when peers have different ideas, or when coordination challenges prevent us from getting to the heart of the work. We try elevating something like capabilities to the level of an orienting value and are shocked that people tune out.
Both are necessary. Values without organizing constructs leave us inspired but unable to coordinate. Organizing constructs without values can’t engage the world—and, in the situations where we’d reach for an OC, that’s exactly what we need them to do.
So that’s the theory. Now, let’s turn to practice.
Using Organizing Constructs: Preliminary notes
To be clear: organizing constructs are an idea I developed for my dissertation and which I’ve tested in ad hoc conversations and private thinking. It’s a tool I’m now trying to elaborate, but it doesn’t have an empirical tradition to draw from.
What follows is my current thinking on how organizing constructs function in practice. It’s a considered starting point for testing and refinement, not a final blueprint.
Where Organizing Constructs help
I suspect OCs are most useful when three conditions are present: (1) the work spans multiple levels at once; (2) participants hold different types of knowledge or authority; and (3) time, attention, or trust is limited, but the situation isn’t urgent.
My day job is in evaluation, and some of the most challenging initiatives to evaluate are those that operate across levels and domains — for instance, a ten-year initiative to improve student career outcomes starting with elementary-school community support. In that context, organizing constructs offer a shared scaffolding for understanding: sharper than proxies and resistant to indicator sprawl.
In multi-stakeholder systems-change efforts, this matters. We often need to link subjective experience (how people feel, what stories they tell) to structural leverage points (funding, policy, institutional norms). OCs help hold those layers together without forcing them into false alignment. In participatory work12, they can create enough shared resonance that people using different frames can still stay in conversation together.
The capabilities approach offers a partial model here: it scaffolds a value (justice) into an actionable framework that travels across disciplines, decisions, and domains. Not every OC aspires to that level of formalism, but the underlying function is similar: making complex meaning portable without flattening it.
From Healthy St. Louis to St. Louis Matters
Let’s return to the Healthy St. Louis example above, which struggled to move the needle because of the difficulty in mapping the general goal of “health” across levels and stakeholders. What if they had decided, instead, on St. Louis Matters as a rallying call for their health push?
With mattering (feeling valued and adding value) as the organizing construct instead of empowerment, the conversation would have looked different. The mayor might ask: Do residents feel valued by the healthcare system, and do they feel their participation in their own health makes a meaningful difference? The hospital council would need to examine whether patients feel providers appreciate their input, and healthcare workers feel their contributions actually improve outcomes. Ward leaders could investigate: Do residents feel city leaders value their neighborhoods, and do community members feel they can effectively advocate for environmental health?
These aren’t identical questions, but they’re legibly related and easy to map. Each points to specific, actionable levers. More importantly, when someone doesn’t feel they matter to city hall, their sense that personal health choices matter erodes too. The construct reveals those connections instead of obscuring them without requiring a formal theory.
Where Organizing Constructs aren’t the move
OCs are for alignment. If the work is about healing, justice, conflict, or reconciliation, other tools may serve better. I expect they will have limited utility when power, and not shared understanding, is central to the problem13. If people are talking past each other, reproducing inequalities, and playing status games, OCs are likely just another board onto which those dynamics will map. Leaping to coordination in a case where one set of interests is underrepresented, for instance, can actually exacerbate power dynamics.
Similarly, organizing constructs don’t resolve fundamental disagreement about what’s important. Remember, values come prior. And timing matters: in times of crisis or catastrophe, OCs may orient us towards a less-than-helpful theoretical orientation when we should be leading with care and relationality. Finally, just because something can be mapped and coordinated doesn’t mean it should be. Mattering is a great organizing construct, but it’s not necessarily suitable for all tasks. And all the warnings cross-cultural theory has given us around social context still apply.
Some remaining questions
This is just an introduction. The idea is still taking shape, and there’s plenty left to explore. Here are a few questions I think are worth chasing down next.
Is a two-part structure important for stabilization?
I’ve gravitated towards constructs with a two part structure in my search for OCs. I suspect that concepts that bifurcate around complementary tensions (e.g., subjective + structural, rights + responsibilities) carry built-in constraints. They define not just what must be included, but what must be held in relationship.
This stabilizes meaning while allowing flexibility across frames.
That said, I don’t see any reason, prima facie, why this couldn’t be true of constructs with three elements or even those with one really sharp one. I imagine as the number of elements grow, the risk of fragmentation does, and as it shrinks there is less to map to. But I don’t really know.
How do OCs actually relate to values?
The distinction between OCs and values is the easy part. What’s harder is mapping their relationship in practice. I assume values come first, but don’t have much guidance to give on how to connect an OC to a value. Are some naturally better paired, or are they agnostic with respect to one another? Is the relationship one-to-one, one-to-many, or tangled and interlaminated? Is it all contextual? Probably.
One well developed example that we can probably learn from is the relationship between justice or equity (values) and capabilities (an organizing constrcut). Capabilities don’t exhaust what justice means, but they offer a portable, structured, and cross-level way of acting on it. That relationship feels like a model worth studying.
Can general constructs become OCs over time?
Are broad or leaky constructs like care, belonging, or empowerment beyond repair? Or can they be refined and constrained into functioning organizing constructs?
In some cases, I suspect that natural language associations poison the well, making it difficult to retrofit sharpness onto a word that’s already bloated with cultural weight and the baggage of failed campaigns. But maybe that’s not inevitable. Maybe with the right internal structure, even soft concepts can stabilize.
Bonus: Three Questions to Tell If Something is an Organizing Construct
Does it appear intuitively meaningful at multiple levels of analysis?
Does it contain internal structure that resists semantic drift without rigid definitions?
Can it guide inquiry, coordination, or design without excessive translation?
I’ve also built a more thorough framework (TASC) for evaluating candidate organizing constructs in practice. You can find that in the technical companion.
Where This Leaves Us
Empowerment collapsed because it tried to do too much without enough stable internal structure and public legibility. But the need it responded to didn’t go away. We still work on problems that span individuals, organizations, and systems. We need better ways to talk across those levels without losing clarity or getting bogged down in methodological complexity.
I don’t know yet how widely organizing constructs apply, or whether they are as agile in the field as I believe they can be. They could be a skeleton key that unlocks conversations and pulls together theory. Or we could learn the difference is just that mattering hasn’t been co-opted yet. If you try using it somewhere, I’d be curious what happens.
Some more academic details and rigor-related thoughts:
Organizing Constructs: Technical Companion
This isn’t to say empowerment failed only because of conceptual fuzziness. Co-optation, political capture, and power’s intransigence all played roles. But I think the lack of stable internal structure made it particularly vulnerable to these forces.
Isaac was my PhD advisor, and remains a good friend.
Sharp readers may object - isn’t this still just an individual measure applied to different domains of life? Yes. But the domains prefigure a more multilevel treatment, which I have begun to explore in published pieces and which I hope to expand upon in this blog. In the natural language register where OCs shine, mattering holds up across levels.
See the Technical Companion for a brief discussion of how Multidimensional Mattering (MDM) compares to other mattering factor structures, and what this implies for its use as an Organizing Construct.
Okay, but why don’t they just all report on the thing they care about? Vegetables, pollution, drinking - each improves health, right? This is the default I tend to see in the wild. But it can’t handle cross-level dynamics, mismatched time scales, or the challenge of comparing fundamentally different units Instead of one metric that nobody can interpret, you get a bunch. Each individual can report what they care about, but we still can’t say how (or whether) these efforts add up. And generally, for efforts of this scale, somebody’s going to want the big picture to make sense.
And, indeed, each stakeholder often maps privately from what they care about to the shared index. What organizing constructs do is allow this mapping to be public and shared.
This sounds like boundary objects, and maybe this facet aligns. I say a bit more on this in the Technical Companion.
While this is a very different usage than the psychometric one, I’m using construct rather than concept intentionally. Organizing constructs must be bounded enough to survive translation. Without specificity, they dissolve into platitudes. But if they’re too rigid, they freeze in place. Their usefulness depends on being tight enough to hold meaning and loose enough to move across frames. That balance is what separates an organizing construct from a vague buzzword or a frame-bound Theory fragment.
I remember playing Final Fantasy Tactics at 13, watching pixelated revolutionaries argue in poorly translated Shakespearean diatribes about who really represented justice. Even then, it was clear they were mostly talking past each other.
For any development nerds: It seems to me that Capability, in the Sen and Nussbaum sense of functionings + freedoms, is a great organizing construct. Like mattering, it also has a two-part structure, which seems to be important in preventing drift without ballooning. If you were trying to operationalize a Thriving America-type initiative, for instance, I think it’d be more effective than something like Well-Being or QoL. Which makes sense, since this is more or less what it was designed for.
In fact, I think this kind of error can be a vector of misdirection that keeps us squabbling over meaning rather than moving towards action. I hope to write more about this someday.
Though there are reasons for caution here. See the last section of the technical companion (10/8 update) for more here.
If you’re thinking structurally about power, you might argue that’s everything. You wouldn’t be wrong. Power shapes everything, but I think a line can be drawn between where alignment is useful enough to foreground and where it is not.