When Does a Triple Decker Stop Being a Triple Decker?

Boston's triple decker is one of the city's most recognizable architectural types.

For more than a century, it has housed families, shaped neighborhoods, and defined the visual character of communities from Dorchester and Roxbury to Mattapan and beyond. It is not merely a building type; it is part of Boston's cultural DNA.

Yet cities evolve.

Families evolve.

Construction methods evolve.

Building codes evolve.

The question is no longer whether our neighborhoods will change.

The question is how.

For the proposed redevelopment of 106–108 Millet Street in Boston's Dorchester neighborhood, Flow Design Architects began with a deceptively simple question:

When does a triple decker stop being a triple decker?

The answer, we discovered, is not found in copying the past.

It is found in understanding it.

We don't subscribe to copying. Every site is unique.

Too often, new development falls into one of two extremes. Some projects ignore their surroundings entirely, creating objects that feel disconnected from their context. Others attempt to replicate historic buildings so literally that they become architectural imitations—buildings that borrow the language of the past without contributing anything meaningful to the future.

Neither approach felt appropriate.

Instead, the design team explored a third path: creating a building that respectfully participates in Dorchester's architectural conversation while speaking in its own voice.

The project studies the essential ingredients that make a triple decker recognizable. Bay windows. Verticality. Layered façades. Stacked living. Front porches and balconies. Strong relationships between interior life and the street.

These characteristics were not copied.

They were reinterpreted.

The traditional bay window, a familiar feature throughout Boston's neighborhoods, became an inhabitable space rather than a decorative projection. It expands living areas, captures natural light, frames views toward the Wheatland Community Garden, and creates moments of intimacy within compact urban homes.

The roofline was similarly reconsidered.

Rather than repeating the conventional forms of neighboring buildings, the design introduces a subtly tilted roof geometry that acknowledges the surrounding gabled houses while simultaneously orienting itself toward future energy demands and photovoltaic readiness. The gesture is both contextual and forward-looking.

Balconies and outdoor spaces are carefully positioned to strengthen connections to sunlight, views, and the adjacent community garden. They become extensions of the home rather than ornamental additions.

Materiality plays a similar role.

The building seeks neither to disappear nor dominate. Instead, it positions itself somewhere between familiarity and innovation. It recognizes the neighborhood's past, represents its present, and quietly suggests its future.

This balance became the central design challenge.

How can architecture respect its surroundings without becoming nostalgic?

How can a building belong to a neighborhood without pretending to have been built a century ago?

How can contemporary housing improve upon a beloved typology rather than simply repeat it?

The answers emerge through a series of deliberate architectural moves.

A reimagined bay window.

A performance-driven roof.

Outdoor spaces connected to landscape.

Clean contemporary detailing.

Natural light as an organizing principle.

Together, these elements create a building that remains unmistakably rooted in Dorchester while acknowledging the realities of contemporary life.

The project ultimately suggests that the future of the triple decker may not lie in preservation alone.

Nor does it lie in replacement.

Its future may lie in evolution.

The best neighborhoods are not museums.

They are living organisms.

Each generation inherits an urban fabric shaped by those who came before and leaves behind a contribution for those who follow.

At 106–108 Millet Street, the goal was never to erase the triple decker.

The goal was to ask what the next chapter might look like.

Perhaps a triple decker stops being a triple decker when it forgets where it came from.

This project attempts the opposite.

It remembers.

It learns.

And then it moves forward.