The Multilevel Metropolis

The grade-separated pedestrian systems built in the 20th century have a variety of names: skyways, skywalks, pedways, footbridges, the +15, and the Ville Souteraine. But they have one thing in common — they have radically altered the form and spatial logic of cities around the world. North American cities like Minneapolis, St. Paul, Des Moines, and Calgary have extensive skyway systems that parallel the original streets. Montreal and Toronto have subterranean labyrinths. Hong Kong has floating three-dimensional circuits that connect transit stations, shopping malls, office towers, and parks. And multilevel urbanisms continue to expand. In the last decade, Mumbai has relieved its crowded streets by building nearly 40 pedestrian overpasses. Yet even as such infrastructures proliferate, they receive scant critical attention, despite their fundamental role in the production of urban space.

They have one thing in common: they have radically altered the form and spatial logic of cities around the world.

Since the 1960s, seventeen notable pedestrian systems have developed in the United States and Canada. ¹ Whereas early experiments in multilevel urbanism belonged to social utopians and the architectural avant-garde, the systems that were built at scale were advanced by pragmatists who embraced both the public and private spheres. As cities struggled to withstand the growing economic power of suburbs, civic leaders began to “pedestrianize” their urban centers. What better way to compete with the suburban shopping mall than to mimic its enclosed form? Cities began to connect and consolidate interior spaces through skybridges and arcades, in an effort to make downtown convenient, comfortable, safe, and climate controlled for office workers and shoppers.

These pedestrian systems can be conceived as a “thickening” of the street level or a “delamination” of the ground plane into a second level above or below grade. This doubling, or sometimes tripling, of the street was described by urban critic Trevor Boddy in the 1980s as an “analogous city.” ² The spatial ambiguity created by the stacked circulation levels — which often lack suf­ficient vertical connections — can render the urban layers as fully inde­pendent realms. For that reason, skyways and tunnels are seen by many theorists as deviant or untenable urban forms.

Nevertheless, they are attractive to city planners and business leaders for their ability to concentrate transit, commerce, and real estate value. Once established, such systems tend to expand, incrementally and informally, according to private and public development interests. What begins with a few interconnected blocks in the downtown core grows over decades into a pedestrianized, multilevel central business district. The largest networks reach an autocatalytic stage, like the 69-block Minneapolis Skyway, where future developments must be connected to be competitive. At that point, these systems have to be acknowledged as durable urban forms. Among the major pedestrian networks in the United States, only the Cincinnati Skywalk has been terminated. (It is now mostly dismantled.) Many of the others are still growing after 50 years, defying conventional modes of urban development.

In the 21st century, we see a new wave of interest in the skyway as architectural form.

In the 21st century, we see a new wave of interest in the skyway as architectural form. Steven Holl’s high-wire act in Beijing, the Linked Hybrid complex (2009), features bridges stretched taut between buildings, while his Vanke Center (2009) in Shenzhen links buildings and programs in an elaborate city-scaled lattice. MVRDV has proposed high-rise towers in Seoul (2011) and Shanghai (2015) that are grafted together by pixelating the building module to form a bridge of housing and public space. Urban retail complexes by Zaha Hadid Architects, Foreign Office Architects, and Future Systems weave layered bands of horizontal circulation, extending interiorized commercial space to an urban scale. Elsewhere, planners and architects propose elevated green walkways, hoping to repeat the success of New York’s High Line (2009–2014) by James Corner Field Operations and Diller Scofidio + Renfro.

While architects of such high-profile projects knowingly or unknowingly reference the concepts of an earlier avant-garde — touting “social condensers,” “floating cities,” and “horizontal skyscrapers” — they have mostly ignored the multilevel urbanisms that have developed continuously for half a century in cities like Minneapolis and Calgary. In Parallel Cities, we aim to recover a critical history of those seemingly mundane projects, which actually are radical experiments in urban design and planning. What do urban skyways offer the city as a social and political apparatus? What does the success of their incremental growth suggest about processes of urban transformation? What are the opportunities and risks of deploying these forms on a large scale?

The Circulation of Ideas

Every skyway project seems to have its own origin story, the legend of a visionary hero or partnership of developers and bureaucrats who willed it into existence. But as it turns out, those heroes invariably got their ideas from the mid-20th century European avant-garde, occasionally filtered through the 1939 World’s Fair in New York and the World Expos of the 1950s and 1960s. Yet the main proponents were directly influenced by ideas developed at the International Congresses of Modern Architecture (CIAM) and imported to North America by Spanish architect Josep Lluís Sert, who led Harvard’s Graduate School of Design from 1953 to 1969. At Harvard, Sert continued to push the agenda of CIAM 8: “The Heart of the City,” particularly its emphasis on the social structure of the urban core and three-dimensional pedestrian network strategies derived from Le Corbusier, as well as experimental schemes by British architects. Historian Eric Mumford describes the Harvard dean as presiding over a reconciliation between CIAM’s doctrinaire modernists and the socially-oriented CIAM apostates known as Team 10. Under Sert the two initially adversial urbanisms became nearly indistinguishable. ³

The first Harvard Urban Design Conference, in 1956, ratified that merger. Among the participants was Victor Gruen, the architect who not only conceived the modern shopping mall but also proposed some of the first downtown pedestrian zones. He later published a book, The Heart of Our Cities: The Urban Crisis: Diagnosis and Treatment, that echoed CIAM’s The Heart of the City: Toward the Humanization of Urban Life. Another conference participant was Jane Jacobs, the urban theorist and activist who vehemently opposed CIAM zoning and praised some elements of Gruen’s urban proposals.

That rich intellectual activity would have remained largely unrealized were it not for the students and practitioners who started working with developers and city governments in Middle America.

In the years that followed, the architectural design and planning faculty at Harvard and MIT coalesced around a growing interest in grade-separated pedestrian systems. Members of Team 10 were in the mix, along with Metabolist architects Kenzo Tange and Fumihiko Maki. The Ford Foundation supported the MIT-Harvard Joint Center for Urban Studies (1959–1971), which explored modernist approaches to city planning, with funding directed toward research-based projects that supported larger domestic and international policy agendas. ⁴ Architecture and urban design were tied to the social sciences, information sciences, and sys­tems theory; and it was all intertwined with research from CIAM/Team 10 and from the Conference of Architects for the Study of the Environment (1964–1974), instigated by Peter Eisenman. Their experiments in multilevel urbanism were carried forward by a new generation at the Institute for Architecture and Urban Studies and in the 1967 exhibition New City at the Museum of Modern Art.

But that rich intellectual activity would have remained largely unrealized were it not for the students and practitioners who took those ideas from academia and started working with developers and city governments in Middle America. The two most influential architects in the deployment of multilevel pedestrian systems were Gruen and Vincent Ponte. They elaborated on the work of CIAM to invent new typologies that were applied in urban design proposals for cities such as Fort Worth, Dallas, and St. Paul, laying the foundation for the modern skyway and subway systems.

Victor Gruen Goes to Texas

The rapid dissemination of those CIAM 8 concepts was a direct result of how they were designed and promoted. Gruen understood the complex forces underlying urban development and renewal. His politics were an unusual combination of socialist idealism and capitalist pragmatism, and the spaces he designed were eminently practical, rooted in modernist postwar plan­ning principles and incremental methods.

Gruen had been educated in Red Vienna, where he was influenced by the work of Adolf Loos and Erich Mendelsohn. A committed socialist, Gruen embedded his politics in his architecture; he saw pedestrian separation as a way to cre­ate more active and socially oriented urban centers. Shortly after emigrating to the United States, he worked on Norman Bel Geddes’s and Eero Saarinen’s Futurama exhibition for the 1939 World’s Fair, which included elevated pedestrian walkways. The project united Gruen’s interest in urban design with his background in retail architecture, an experience that he described rather clinically as “a stimulating task in con­templating and bringing into visual form future ideas for transportation and city planning.” ⁶ Multilevel urbanism centered on retail environments became one of his passions, and in 1956 he included downtown skybridges in his proposal for a pedestrian center in Fort Worth, Texas.

That was a momentous year for Gruen. Not only did he attend the Harvard conference and publish his Fort Worth plan; he also opened the first enclosed shopping mall in the United States, the Southdale Center in suburban Minneapolis. The project had strong social ambitions and attempted to create the type of civic space that might be found in a traditional European square. Gruen envisioned the mall as a center for local art and culture, as well as a place to buy necessities. That program was implemented for a time, but later Southdale evolved into a more conventional shopping center, with renovations to accommodate ever-shifting national brands and chain stores.

Victor Gruen introduced something that could be described in contemporary terms as a form of tactical urbanism.

Gruen’s urban design proposals introduced something that could be described in contemporary terms as a form of tactical urbanism: a vocabulary of adaptable components, such as pedestrian bridges, plazas, and arcades, that could be deployed selectively. Construction could thus proceed incrementally, radically changing a city over time. Gruen’s plan for Fort Worth proposed to reorder the city around a central pedestrian plaza with shop­ping. Vehicles were relegated to the periphery, and elevated pedestrian bridges connected parking ramps to the walking zone at the urban core. Although it was framed within a compelling narrative that referred to everyday life in historic European cities, Gruen’s alternative was distinctly modernist.

Jane Jacobs praised his Fort Worth plan in her famous essay, “Downtown Is for People.” ⁷ She argued that Gruen’s work was superior to the early modernism of CIAM or Robert Moses’s brand of urban renewal because it specified a variety of activities at the street level. While she broadly endorsed Gruen’s social programming, she warned that clients and others who enacted his plans would overlook that critical point, focusing instead on his strategies for traffic management. And as she predicted, Gruen’s model for incremental development proved tenuous as his proposals — or something like them — were built in cities across the country. Local governments implemented only those components that were desirable at a given time to particular political constituencies, with little regard for the whole. The concept of a vehicle-free center was often abandoned as the cities evolved. The socially-oriented urbanism that was so crucial to Gruen’s vision demanded a more integrated and comprehensive approach.

Twin Cities: Minneapolis and St. Paul

When the Southdale Center opened, its success threatened businesses in downtown Minneapolis and St. Paul. Civic leaders looked to Gruen for concepts to reactivate their commercial cores. It was common at that time for master plans to be produced by nongovernmental entities, such as business councils and chambers of commerce. Government agencies would then pick and choose from the available plans, implementing whatever components fit their agenda. In 1956, a group of local young architects influenced by Gruen proposed a car-free zone with skyway connections for downtown St. Paul. The next year, architects at the University of Minnesota (under MIT graduate Walter Vivrett) presented a similar plan for downtown Minneapolis. Both groups were theoretically grounded and aware of the international discourse on multilevel urbanism that was being promoted through the efforts of Team 10.

Today, the Minneapolis Skyway extends for more than eight miles, the world’s longest contiguous system.

Gruen himself was hired in 1957 to work for a private company on the Capital Center project, a 12-block centralized pedestrian mall in downtown St. Paul that would incorporate principles of the Fort Worth plan. He is cred­ited as the designer of a department store and parking ramp for that project, but his larger contribution was designing an urban scheme that evolved, after he left the project, into the nation’s first comprehensive skyway proposal. In 1961, the Capital Center project was taken over by the three architects who had originally proposed it (Louis Lundgren, Grover Dimond, and Brooks Gavin). ⁸ Building on Gruen’s work, the city sought federal funding for an urban renewal project and skyway system that would cover nine blocks. ⁹

Across the river, Minneapolis followed similar planning strategies but skipped the federal funding. The first two skybridges were built in 1962 and 1963 as part of a private devel­opment. Lawrence Halprin’s Nicollet Mall, a transit-oriented pedestrian mall, was completed in 1968. The skyway system was defined in 1972 with the construction of a hub at Philip Johnson’s IDS Center, and today it extends for more than eight miles, the world’s longest contiguous system. Minneapolis adopted a private model, in which building owners negotiate with the city for air rights over the street. The city incrementally regulates the skyway system’s overall form through rules that determine the location, orientation, and basic dimensions of its components, and air rights are granted for bridges roughly centered within each block. The result is a regular pattern of bridge connections between labyrinthine block interiors. Design efforts are generally focused on the form of the individual components rather than the overall system, often resulting in chaotic paths.

St. Paul finally began construction of its skyway system in 1967 and, unlike Minneapolis, adopted a public model. Bridge placement was deter­mined by the Department of City Planning, and the designs were based on a single prototype. Construction was funded by the federal government and regulated by the state Department of Transportation. Because St. Paul has more historic buildings than Minneapolis and an irregular sloping topography, bridges were designed to adapt to highly com­plex conditions, creating a very circuitous network. Urbanist William H. Whyte criticized the system, calling St. Paul the “blank wall capital of the United States,” in reference to the system’s disorienting interior spaces and walkways that passed by the windowless side walls of buildings. ¹⁰

Gruen’s influence can be seen today in many North American cities, but in a diluted form. At the height of suburbanization in a car-crazed country, his plans for urban pedestrian zones were rarely implemented as he envisioned. Instead, city planners and developers spanned streets with footbridges, or tunneled under them, creating a more unified retail experience without threatening the primacy of automobiles. The economic competition between cities and suburbs led to the “mallification” of many downtowns, a phenomenon exacerbated by the implementation of con­venient and secure pedestrian pathways over and under existing city streets.

The Multilevel Man in Montreal and Dallas

Vincent Ponte was a very different kind of protagonist. Described in Time magazine as the “Multilevel Man,” he developed incremental mixed-use superblock plans in the 1950s and ’60s for numerous cities, including Montreal, Winnipeg, Miami, Columbus, and Dallas. Ponte was a key speaker and participant at the 1956 Harvard Urban and Design Conference, which furthered the ideas and influence of CIAM 8. As planner for Montreal’s La Ville Souteraine (The Underground City, 1962), he promoted a separation of pedestrian and automo­bile traffic. Montreal was the first case of a com­prehensive below-grade pedestrian system implemented at an urban scale.

Ponte was driven by economics more than social or environmental concerns. Working with Montreal City Transit and Eatons Department Stores, he catered to the inter­ests of downtown business communities and local politicians. Despite his affiliation with CIAM, Ponte claimed that his vision of the multilevel metropolis was inspired by Leonardo da Vinci’s ideal cities and Antonio Sant’Elia’s plans for Rome. ¹¹ Like Gruen, he was more planner than developer. He introduced skyways as a malleable urban tool and left it to others to build them.

Most skyway cities are hybrids that combine characteristics of both planned and self-organizing systems. Montreal is an excellent exam­ple of this in a subterranean network. La Ville Souteraine was conceived by I. M. Pei and Henry Cobb in the 1960s as a self-contained, fully complete underground pedes­trian zone, emulating New York City’s Rockefeller Center. It was subsequently expanded by Ponte, and it ultimately evolved into a loose network that moved beyond its original boundaries and grew like other ad hoc systems in the United States and Canada.

Most skyway cities are hybrids that combine characteristics of both planned and self-organizing systems.

Similar to his superblock proposals for Montreal, which integrated shopping malls and transit, Ponte’s vision for the Dallas Pedestrian Network encompassed the entire downtown. His clients were the corpora­tions that occupied the majority of the office space, including powerful oil and tech companies. The system he designed comprised one mile of overhead walk­ways and two miles of underground tunnel links, connecting a total of 36 blocks. Here, too, the network evolved into a more informal configuration, despite the comprehensive master plan. The combination of above- and below-grade connections amplified its discontinuities. Responding to the car-centric culture of Dallas, developers prioritized parking ramp connections and vertical links to surface parking lots. Architecture critic David Dillon wrote that “what seemed like progressive planning in the 1960s has become regressive in the 1990s.” The city’s relentless interiorization took many retail spaces off the street and into the tunnels, which created a caste system where “downtown streets belong to the poor, the homeless, and the politically disenfranchised” and the air-con­ditioned interior to a homogenous population of office workers. ¹²

Calgary: Toward an Urban Megastructure

The first multilevel urbanist projects influenced by CIAM and Team 10 appeared in England as buildings connected with open-air bridges. Soon there were more spatially complex mat buildings, mega­structures, and fully interiorized urban networks. The skyway-subway cities that later developed in North America were strikingly similar in scale and patterns of growth. Unlike rationally planned gridiron cities, they exhibited self-regulating behaviors in their organic development, which was mediated by economic, political, and bureaucratic agendas, not to mention architectural conditions, building codes, and the idiosyncrasies of property owners.

Calgary, Alberta, is the one of the few counterexamples — a project that was conceived as a complete urban system and strategically implemented over time. Calgary’s +15 is an elevated pedestrian system named for the height of its footbridges above street level. It was conceived by chief city planner Harold Hanen, who was hired immediately after graduating from MIT in 1966. Hanen had studied under Jerzy Soltan, and he was immersed in the theories of Team 10, Le Corbusier, and the Metabolists. His final thesis (a group project crediting Soltan) was a multilevel city with a fully integrated public transit system. ¹³

Within six months on the job in Calgary, Hanen had proposed a similar plan that combined mass transit with a pedestrian network. The plan allowed for acquisition of public right-of-way through private buildings and created bonus-density incentives that accelerated the development of the network throughout the city. Hanen also proposed urban amenities like the Devonian Gardens, an elevated, interiorized public park that now generates revenue for the city as an event venue. Although Hanen stayed in the position for only three years, his plan lives on. ¹⁴ Recent projects like the National Music Centre of Canada (2016), by Allied Works Architecture, leverage the city’s bonus density to extend large volumes of space over the street, creating interior connec­tions that could tie into the city’s emerging +30 and +45 levels.

A few men almost single-handedly transformed more than a dozen cities, sometimes without drawings, designs, or detailed plans.

One common thread across all of these built projects is their creators’ mastery of the tools of public policy. Gruen, Ponte, and Hanen were passionate about the multilevel city, but they were not especially driven by a need to design all physi­cal form or architectural experience. Rather, they had a shrewd understanding of urban economics. As a result, these few men almost single-handedly transformed more than a dozen cities in North America, sometimes without presenting drawings, designs, or detailed plans. They devised regulatory codes and tax incentives that could be described as programmatic strategies rather than master planning. Even the systems that were comprehensively planned at inception tended to grow informally over time, which proved the resiliency of the programming script.

Manhattan: Access Trees, Hubs, and Enclaves

One of the more radical proposals for the multilevel city was developed in New York City by the Regional Plan Association in the late 1960s. It was later published as Urban Design Manhattan, one of a series of reports that led to the Second Regional Plan. From 1960 to 1968, the effort was led by planner Stanley Tankel, a proponent of garden cities who had studied under Walter Gropius at Harvard and worked for the London County Council on the post-Blitz multilevel reconstruction. In New York, Tankel collaborated with Jane Jacobs and other Greenwich Village professionals to execute the 1956 Village Study, which examined the segregation of pedestrian and vehicular traffic. Tankel hired planning consultants Rai Y. Okamoto and Frank E. Williams to develop the proposals for Urban Design Manhattan. ^15.

Grand Central Station and Rockefeller Center were described as prototypical multilevel “access trees,” and the inter­sections of multimodal transit systems were seen as mixing chambers, or social hubs. The proposal identified specific multilevel districts where planners should concentrate multi-use public spaces, and it identified diagonal connectors to link the different zones against the generally orthogonal matrix. In Midtown Manhattan, the authors proposed connecting sites like Grand Central Station, Times Square, Columbus Circle, and Rockefeller Center. This work culminated with the Special Greenwich Street Development District, approved in 1971, which attempted to zone a complex multi­level urban design solution into an area around the World Trade Center in lower Manhattan.

The history of the World Trade Center district represents the end of one kind of three-dimensional urbanism oriented toward the street, and the transition to a future city form based on the elevated enclave.

The history of the WTC district, from the 1970s through the design competition after 9/11, represents the end of one kind of three-dimensional urbanism oriented toward the street, and the transition to a future city form based on the elevated enclave. In the 1970s, the special district promoted the orderly and vertical expansion of commercial development to accommodate multilevel public spaces. That included grade-separated pedestrian circulation improvements with strong connections to street-level retail and transportation hubs. The plan used familiar incentives like floor area bonuses, but also created a diverse set of design prototypes to achieve specific objectives. ¹⁶

The original goal was to build interconnections between the first two levels of the city, conjoining the buildings into megastructures, although the reality was more limited in scope. The first project built under this system was the multilevel Deutsche Bank Building complex, designed by the architectural firm Shreve, Lamb & Harmon, at what was formerly the Bankers Trust Plaza (1971–1973). The first elevated bridge and tower plaza was designed by M. Paul Friedberg in 1974. It was damaged on 9/11 and demolished along with three pedes­trian bridges at Liberty, Vesey, and Chambers streets. Although formal regulations for elevated pedestrianization had ended in 1998, ¹⁷ public reaction forced the installation of new bridges designed by SHoP Architects (2002, 2003) at Rector Street.

The sectional logic of New York City can be traced through three distinct epochs of urban design. The first, exemplified by the 1929 Regional Plan, was driven by a desire to accommodate higher urban densities through the layering of transportation systems and the regulation of building mass to permit light and air onto the street. The next epoch began in the 1960s and culminated with the Second Regional Plan. This era was dominated by systems thinkers who proposed to rationalize the Manhattan gridiron as a three-dimensional infrastructural problem that would be solved by increasing efficient movement between commuter hubs and workplaces. But even as these ideas took hold, alternative forms of multilevel urbanism were emerging: superblocks, megastructures, and other clustered developments. By the 21st century, it was clear that the enclave had replaced the network as a preoccupation of urban designers.

Architects who wished to create interconnected public or quasi-public space within consolidated architectural projects gravitated toward ideas that corre­sponded with the concentrated capital investments of public-private developments. Among the design submissions for the reconstruction of the World Trade Center site, eight out of nine schemes proposed elevated bridges or tangential intersections connecting building clusters high above the streets of Manhattan. The international attention focused on the competition further legitimized the three-dimensional urban enclave as the quintessential form of interiorized neoliberal urbanism. Networks that were primarily driven by public investment, outside the boundaries of individual projects, seemed to be a thing of the past.

However, recent proposals suggest that a synthesis of those three epochs of multilevel urbanism is indeed possible, if not inevitable. For example, SOM’s 2012 proposal for Grand Central Station envisions a high-density cluster of towers that would surround the station with densely layered transportation systems, including pedestrian walkways and concourses, mass transportation, shopping malls, rooftop gardens, and a circular observation deck that hovers over Grand Central like a halo. In this way, the project attempts to combine the “access tree” logic of the Second Regional Plan with the multilevel enclave and branded urban spectacle of contemporary neoliberal developments.

Hong Kong: City as Process

Space, it is said, is both socially produced and produces the social; but to this we will have to add that it has become so complex and enigmatic that it cannot be directly described. Urban spaces in particular are like black holes: we perceive them only in the effects they produce.

— Ackbar Abbas, “Between the Visible and Intelligible in Asian Cinema,” 2010

We conclude with Hong Kong, which has the most extensive multilevel pedestrian system in the world. Its form has evolved organically from a need to reconcile the city’s complex street pattern with the steep hillside terrain. Although not directly influenced by the Metabolists, the growth of the city’s 700 footbridges may be one of the best contemporary examples of the Metabolist concept of the “city as process.” Cities like Hong Kong have followed a distinct trajectory since the 1960s, using multilevel strategies reactively to accom­modate increased density and reduce congestion, pollution, and crime, while promoting economic development.

The British colonial government first proposed an elevated pedestrian net­work connecting the Central and Admiralty districts in 1961. Hong Kong’s Colony Outline Plan in 1969 described a more complete vision: a multilevel city with elevated public spaces connecting housing, businesses, offices, and parking in a single megastructure elevated above a mass transit line. ¹⁸ After the first elevated footbridges were constructed in 1963, a more interconnected system began to take form, beginning with the creation of Connaught Place in the Central District in 1970. Additional “flyover” connections across congested streets and roadways enabled greater com­mercial development. Due to government own­ership of all lands and Hong Kong’s lack of a significant historic conservation agenda, the implementation of this system proved far more successful than in other cities.

Elevated walkway systems now span the majority of the Sheung Wan, Central, Admiralty, and Wan Chai districts. The pedestrian network features a range of connector prototypes, including deck-access plazas and podiums, flyover bridges, open-air footbridges, and high-bridge networks (exterior pedestrian bridges over streets), interiorized walkways, elevated parks, and exterior escalators that scale the steep hillsides. The most significant of these is the Central Escalator, a series of moving, covered walkways more than 2000 feet long that connects Hong Kong’s central business district to the Mid-Levels residential district. Hong Kong employs “on-demand planning” to develop and refine the system and respond to changing circumstances. Planners use data-mapping to focus on specific districts and neighborhoods rather than trying to achieve a continuous urban network. ¹⁹

Since the government owns all land and leases it to developers, the system is inherently temporal and adaptable. The process for each new development is not simply about negotiating a building on a site, but about integrating each new component as part of a larger interconnected urban condition. On-demand planning supposedly ensures that the condition generates the form. Although the system is meant to serve the public, there are inevitably political and economic agendas underlying decisions about control and access.

As the population of Hong Kong and its surrounding districts grows, the city continues to evolve into greater three-dimensional complexity. Land reclamation projects have been ongoing since the 1890s, and in the early 21st century the government has created more than 6,000 hectares of new land along the waterfront. This expansion has been accompanied by massive multilevel infra­structural improvements and building development. The pedestrian network that began in the 1960s as an effort to reconcile urban density with the steep natural topography has escalated to a project of an entirely different order. Planners create complex interconnections within the new artificial topography of the autonomous urban form. Walkways emerging from underground transit hubs connect exterior terraces and podia with interior atria, shopping mall spaces, and the extended urban realm. Analogous to the ossification and resorption of growing bone tissue, the transformations of Hong Kong reveal the metabolism of the modern metropolis.