Sim — Soft City
David Sim, Soft City: Building Density for Everyday Life (2019). Foreword by Jan Gehl.
The Argument
Most density is hard — monolithic, top-down, designed at the scale of the building rather than the person. Towers, parking structures, corporate campuses, shopping malls. Hard density can be efficient by certain metrics, but it produces anonymity, disorientation, and isolation. You don’t know your neighbors in a 200-unit high-rise.
Sim — who spent decades at Gehl Architects, the Danish firm that essentially invented people-centered urban design — argues for soft density. Fine-grained, mixed-use, adaptable, built at a scale where people can see and know their neighbors. Copenhagen side streets, Barcelona courtyard blocks, Tokyo neighborhood clusters, Melbourne laneways. Dense enough to support community infrastructure and incidental contact. Soft enough that people actually want to live there.
The distinction between hard and soft is not about building height. It’s about grain. Soft density has many small units, many entrances, many ground-floor uses, many paths, many surfaces that face the street. Hard density has a few large units, a few entrances, a lobby, a parking garage, and a wall facing the street. The soft city is porous. The hard city is sealed.
Incidental Contact by Design
Gehl’s foundational research insight, which Sim translates into specific design guidance: the physical environment determines how much unplanned social interaction occurs. This is not a correlation — it’s a design variable that can be controlled.
Wide streets with parking lots on both sides produce near-zero incidental contact. People drive from garage to garage. No one walks. No one lingers. No one encounters anyone they didn’t plan to see.
Narrow streets with benches, stoops, ground-floor windows, and active edges produce high incidental contact. People walk. People sit. People pause. The barista knows your name because you pass every morning. The neighbor’s kid recognizes you because you’re visible on your stoop in the evening. None of this was programmed. It was designed — through spatial decisions about width, setback, transparency, and furniture.
This is the spatial infrastructure that makes the desire path principle work physically. The desire paths can’t form if people never cross paths. Sim provides the design vocabulary for ensuring they do.
The Privacy Gradient
Sim handles one of the hardest problems in dense community design: how do you produce enough contact to build community without making people feel surveilled or exposed?
The answer is the privacy gradient — a concept shared with Alexander’s pattern language but developed here with more contemporary examples. Spaces are layered from fully public (the street, the park) through semi-public (the shared courtyard, the community garden) through semi-private (the front stoop, the balcony visible from the path) to fully private (the interior of the home, the backyard).
Each layer has different behavioral norms. On the street, you nod. In the shared courtyard, you chat. On the stoop, you linger. Inside your home, you’re invisible. The gradient gives residents control over their level of social engagement — and that control is what makes dense living tolerable. Without it, density feels intrusive. With it, density feels connected.
This is directly relevant to the Goldilocks problem: too little contact and community never forms; too much and people feel trapped. The privacy gradient is the spatial solution to the Goldilocks problem.
Density Without High-Rises
Sim’s sweet spot is the density range Wellspring is targeting: 30–80 units per hectare, achieved through row houses, courtyard buildings, low-rise apartment blocks, and stacked townhouses. This range is dense enough to support shared facilities (a tool library needs enough users to justify the inventory) and generate the critical mass of incidental contact that community requires. But it’s not so dense that anonymity takes over.
The specific building forms Sim advocates — perimeter blocks with shared interior courtyards, row houses opening onto pedestrian-priority streets, ground-floor common rooms at key intersections of paths — are achievable at Wellspring’s scale and within typical Durham zoning (depending on the site).
Everyday Life as Design Metric
Sim’s most useful framing for the project: urban design should optimize for the mundane. Not the grand civic gesture — the walk to the mailbox, the view from the kitchen window, the path between your door and the nearest shared space, the sound level in the courtyard at 8pm.
This is the design equivalent of the Worn Path’s insistence that community emerges from conditions, not commands. The manifesto is about the texture of a Tuesday afternoon, not the ribbon-cutting ceremony. Sim provides the design language for producing that texture: where do windows face? How wide is the path? Is there a bench at the point where two paths cross? Can you see the community garden from your kitchen?
Cars, Parking, and Pedestrian Priority
Sim addresses the car question directly: soft density requires reducing car dominance. Not necessarily eliminating cars — but ensuring that the pedestrian experience, not the driving experience, shapes the community’s spatial logic. Cars at the perimeter. Pedestrian paths at the center. Parking that doesn’t consume the shared space.
In Durham’s car-dependent context, this is a design challenge. Residents will have cars. The question is where cars live relative to where people live — and whether the path from your front door to the community kitchen goes past your neighbor’s stoop (good) or through a parking lot (bad).
Limitations
Sim writes from a Scandinavian/European context where public transit, cycling infrastructure, and pedestrian culture are established. Durham is different. The municipal planning sophistication and regulatory flexibility he assumes may not exist locally. The book is heavy on inspiration and light on financial mechanics — it doesn’t address how you pay for soft density. Pair with the CLT-LEHC financial model for the economics; use Sim for the spatial design.