Thursday, December 22, 2011

Roosevelt Island and Insular Urbanism

Stephen Smith and Cap'n Transit have been discussing the recent news that Cornell University-Technion-Israel Institute of Technology's graduate school of applied sciences will be constructed on New York City's Roosevelt Island, a 147-acre sliver of land set squarely in the middle of the East River. 

Although Stephen's article questions the wisdom of subsidizing physical expansion of the technology sector in poor and middle-class neighborhoods, both he and the Cap'n also critique the proposed design of the new campus (below), which Stephen has called "quasi-Corbusian:"  

The proposed plan, from the Cornell website.
Citing Nathan Lewis, Cap'n instead suggests a traditionally urban form incorporating "really narrow streets" for the new campus.  As it turns out, there are abundant examples of traditional urbanism on small or narrow islands that might serve as models, as pre-industrial cities often formed on or around mid-river islands for reasons of security or ease of transportation.

Here for instance is Paris' Île Saint-Louis, a 27-acre island only slightly narrower than Roosevelt Island.  Although no high rises are present, population density here is approximately 60,000 per square mile, with room to spare for modest but well-defined park space and a river promenade no more than one minute's walk away for any resident.

This is Chioggia, a canal town on the southern end of the Venetian lagoon which, although charming in its own right, is completely overshadowed by its famous northern neighbor.  I visited this town several years ago, and in spite of the extremely high ground coverage of buildings, there was no sense of claustrophobia.  The need for park space is reduced when the waterfront is so readily accessible.

Rhoda Island, adjoining central Cairo, has an evident early or mid-20th century plan in the garden city manner, with relatively wide streets on a modified grid.  Still, as virtually all buildings are mid-rise apartments, population density is very high. The Nile is no more than 500 feet away from any building.

For contrast, scroll back up to the proposed campus and compare the arrangements of space, the density of the built form, and the networks for pedestrian circulation.  Although the site claims that the primary building will be "the largest net-zero energy building in the eastern United States," there is nothing particularly green in underusing high-value urban land adjacent to mass transit.

The Cornell plan, rather than being a departure from current planning practice on the island, actually continues an ongoing trend away from more traditional forms.  The history of post-1950 urban development on Roosevelt Island is a interesting story in itself (for more see here and here), but at least at the outset, in the late 1960s and 1970s, planners conceived of a car-free haven densely built up with apartments. In spite of rather dismal architecture, the result (at right) was strikingly urban, and moreover achieved a sense of place through 1) keeping the street fairly narrow; 2) curving the street, creating visual enclosure; 3) slightly varying the angle at which buildings face the street to create gentle variations in right-of-way width; and 4) incorporating traditional city elements such as the covered arcade, at right.  The result, in spite of the architecture, is quite good, and it is probably no coincidence that the wikipedia page for the island features this very same perspective.

More recent developments have abandoned this approach, however.  Monolithic new apartment buildings to the south are placed in the middle of broad expanses of grass lawn in an approach far more reminiscent of 1960s tower-in-the-park urbanism than is the case for the older buildings to the north.  Rather than defining the street, these apartments are objects floating loosely in space, unrelated to each other or to the topography of the island.  An aerial view shows the extremely low footprint of these buildings – an extravagant use of land even by the standards of 1960s public housing  and their lack of engagement with the waterfront across the street.  Further, does this plan leave any room for future infill or expansion?  Most unbuilt land appears spoken for as roadway or dedicated park space.

If, on the other hand, builders and planners wish to provide a traditional urban environment, there are a multitude of successful examples waiting to serve as inspiration.  If not on a narrow island, with mass transit access, adjacent to Manhattan, then where?


  1. Great post. Also, check out the video in my latest post – from a ground-level and interior perspective, it's even worse!

  2. I share your love of narrow crooked streets etc, but given that this will be a college campus, I'm not sure quite the same criteria apply.

    Of course, that doesn't mean the design doesn't suck... :/

  3. @Stephen: Thanks, and I agree, that's worse than the aerial shots suggested. The interior spaces remind me a bit of Koolhaas' Seattle Central Library.

    @Snogglethorpe: Really narrow streets would be just one potential approach here, and not necessarily the most suitable one, I agree. Still, I see no reason why a university campus need conform to the "grassy quad" model, in part or in whole. The location provides an opportunity to do something a bit more ... urban.

  4. Re: Universities and narrow streets. Oxford has pretty narrow streets, with buildings right up against them: (street view) (map)
    Cambridge is a litte more spread out: (map)
    But it still has sort of narrow streets and houses: (streets)

  5. In the USA, Yale has lots of courtyards, but narrow spaces between buildings and lots of enclosure:

  6. The unspeakable evil of Le Corbusier lives on after his death. It's a great pity that institutions of higher learning, which should know better, are promoting his urbanicidal ideas.

    1. Urbanicidal: Best new word I've heard in a long time, and a perfect description of Le Corbusier and most (sub)urban design these days.

  7. ^
    Agreed. Unfortunately the notion of the "tower in the park" lives on in the architecture profession and the architecture schools (which are organized and run as Modernist cults, really).

    The tower in the park has even been rebranded as "Landscape Urbanism" recently and it's once again being promoted by LU avatars like Charles Waldheim.

    I don't know why many starchitects/architecture schools are still wedded to the antiquated, antiurban notion of the tower in the park. (Look at Shanghai's or Dubai's skyline - many of the ugly abstract sculptures that make up those skylines are all isolated in their own islandlike "parks" and cut off from each other by highwaylike arterial roads.)

  8. Thanks, I guess, for pointing out what appears to be yet another urban mistake.

    I'd like to hear speculation on on why development teams just don't get urban design. What is it? Why such stupidity?

  9. "Still, I see no reason why a university campus need conform to the "grassy quad" model, in part or in whole. The location provides an opportunity to do something a bit more ... urban."

    It's interesting how US universities follow the "grassy quad" model* while many European universities follow the "embedded within the larger urban fabric" model. That second model - save for instances where a "grassy quad" university spilled out into the surrounding urban streets as a result of hypertrophic growth (like Columbia University) - never really caught on in the US. It's a shame really; this kind of urban university format can make for some really charming, appealing squares, streets, and other public spaces.

    *Our prewar universities follow this model, but our postwar universities are more or less designed to function and look like suburban office parks.

  10. Btw, I will offer a suggestion about why we have such problem about manifesting otherwise reasonable ideas such as building places where people can learn: the problem is the very idea of the "campus."

    Is it possible to build a "campus" along urban lines? In theory, yes. In practice, the word "campus" is so powerful as to preclude any image except Princeton or Stanford -- huge lawns, if one can get 'em and big plazas -- "yards" -- if you are a bit constrained.

    Check out an article I wrote about the problems of Seattyle's U District.

  11. A few comments:

    1) From a quick reading of these posts, it seems to me that people are mistakenly lumping together, at least to a certain extent, "tower-in-the-park" designs with "grassy quad" designs. While perhaps not the ideal urbanism of a large city (especially when repeated numerous times), grassy quads are still more humanistic / urbanistic than "tower-in-the-park" designs.

    Don't know if it qualifies as a classic "grassy quad" (it's been a while since I've visited), but Princeton has some spectacularly "comfy" quads that are just wonderful (at least in terms of a university campus adjacent to a small town).

    2) It's a special shame that they've departed from the original Philip Johnson design scheme for Roosevelt Island, which Charlie describes in his post. This design scheme gave the island a very pleasant air of distinction which residents of the island were very proud of. And for many years, this design scheme, along with the somewhat similar design scheme of Battery Park City, was the pride of all New York City.

    It's sad to see that the design approach was abandoned rather than refined and extended.

    3) The popularity of these "tower-in-the-park" design schemes today (another, among numerous examples, is NYU's original redevelopment plan for its Greenwich Village superblocks which was, and still is to a lesser degree, along these neo-Corbu lines too), and the recent attempts at rehabilitating Robert Moses' reputation, show that the "tower-in-the-park" concept and orthodox planning have never really died. They've always had their advocates and, these days, the advocates are the ones in power.

    I think the reason these design schemes have remained popular is due to a built in appeal to the architectural community -- it gives an architect more of an opportunity to create his/her "own" world and to show off. Genuine urbanism is too subtle and requires too much modesty and deference.

    4) It also helps show that another myth surrounding the work of Jane Jacobs is the myth that her ideas are universally accepted and that they've become the new conventional wisdom. As I have mentioned elsewhere, very little -- almost none -- of Jane Jacobs work has become universally accepted and the new conventional wisdom.

    So the work of Jane Jacobs (as opposed to the myths surrounding her work) is both ignored and, at the same time, criticized for being the new conventional wisdom!

    5) All that being said, it's better that this design is on an island, where the anti-urbanism of a "tower-in-the-park" design is less problematic than it would be if built in the middle of a real city. In other words, if one were going to indulge in the "fun" (and these developments can be fun) of a tower-in-the-park development or world's fair-like development, an island is probably the best place to do it. Just like the wonderfulness of a Princeton quad is exactly right for a university that is adjacent to a small town.

    Benjamin Hemric
    Fri., Dec. 23, 2011, 4:15 p.m.

    P.S. -- I haven't had time to do much commenting lately, and thus have not had time to continue the conversation re Jane Jacobs and narrow streets. (So this P.S. is meant as a kind of "place saver" for a future comment in that thread.)

  12. Among Canada's universities, I would say that Ryerson is the one that has created the most urban environment. It's centered around 42 Gould Street in Toronto so if you go into streetview there, the original building is the 3 storey brick one, which looks to be early postwar architecture. It's more or less meshed into the surrounding fabric of Toronto, although its buildings can be fairly hypertrophic. For example, there is their school of management around 81 Dundas W which is a big square building with a couple of big boxes on the ground floor (Best Buy and Canadian Tire) and maybe the second floor too.

    Their main projects at the moment are the repurposing of Maple Leaf Gardens (former home of the Maple Leafs) into a large supermarket and athletic facility:

    And their proposed Student Learning Centre, although I'm not a big fan on how their retail is set up:

  13. the whole alarm only proves a point becoming evident nowadays: there is no Architecture any more, it is self abolishing (and looks temporary). It is replaced by the peculiarities of Design (buildings look the same). We will have to reincarnate Urban Planning, as alternatives of old show.

  14. @Benjamin and others: I agree there's a distinction that should be drawn here between "Corbusian" style and the "grassy quad" (the latter representing an American, and Canadian, adaptation of the relatively pastoral medieval English universities -- the older learning institutions of Spain and Italy are entirely urban places). Still, there are some shared characteristics: a low ground coverage of buildings and an inward-looking focus where buildings do not do a good job of addressing the street. The similarities meant that it was relatively easy for the American universities of the 1950s and 1960s to transition to a Corbusian format: visit almost any large American university and you'll be able to find a cluster of dormitory high-rises set among giant plazas or vast grass lawns. Generally these are the buildings that campus planners have scheduled for eventual demolition.

    In describing this particular plan as "Corbusian" I presume Stephen was referring to the slab-like form of the buildings and the seemingly careless approach toward the open spaces, which have a disjointed, "left-over" feel to them, very different from the sensitively designed and enclosed old university quads which Benjamin mentions. Still, the low building footprints can also be seen in part as a legacy of the "campus" thinking that David Sucher talks about in his comment and the article he links (thanks for that, David). And as Marc mentions, the influence of landscape urbanism is evident as well.

    From a functional perspective, my question was whether this design allows for future expansion: a large amount of very high-value land is monopolized for a small number of buildings. A very low-density design might be less consequential in Ithaca or Tuscaloosa, but should this institution want to expand in 50 or 100 years, where will it go? The spatial constraints would, I think, demand an urban design here. The planning documents speak of a 30-year build out, but that is short-term in the time horizons of universities.

    @Nicolas: thanks for those links. Even if the buildings do have a hypertrophic character, integrating retail into campus buildings and bringing big box stores into an urban environment are impressive achievements, at least by American standards!

  15. At Columbia, there's a vast quad, with the campus around it raised to be less accessible from the street (the legacy of Eisenhower when he was the president of the university), but outside it the buildings are fairly urban. On the one hand, the six blocks that define the campus proper plus two extra blocks for a more recent expansion are urban dead zones, separating Broadway and Amsterdam needlessly and helping contribute to Morningside Heights' decline in Jane Jacobs' time.

    But away from the campus blocks, the buildings are not as bad. The dorms that are off-campus are mid-rise apartment buildings or high-rise towers on a base, more like the non-campus buildings on 110th Street or the residential two-way streets of the Upper East Side than like the project towers immediately north of Morningside Heights. There's a fair amount of urbanity centered at Broadway and 110th and a smaller bit centered at Amsterdam and 122nd.

    Brown has something similar - there's a quad, but much of the neighborhood's life clusters away from it and toward more normal streets. The heart of College Hill is Thayer Street, at the northeastern margin of the campus proper. The street has more parking than I'd like but has reasonable sidewalks, on-street parking as a buffer from the single-lane roadway, low traffic speeds, and buildings that usually abut the street directly. In other words, it's your typical 18th- or 19th-century American street.

  16. @Charlie: My university in Canada (UWaterloo) has largely been growing through intensification. It was founded in the 50s, on the outskirts of a small city, so it had a lot of greenspace. Both the city and the university have grown a lot since then, and a lot of the growth has been in the form of intensification, building on parking lots and greenspace. At this point, there is only one large site in the central part of the campus that is suitable for a new building. Most of the other sites are only big enough for extensions of existing buildings (which has already started happening). I suspect that a greater amount of new growth will be in the form of construction around (or even beyond) the less dense "fringe" of the campus though, as well as opening satellite campuses.

  17. @Alon: Interesting points about Columbia. I think there was a fairly widespread battening down of the hatches at urban universities in the 1950-1980 time period as the surrounding neighborhoods went into steep decline, although at places like Columbia or Yale, that tendency to disengage from the city went back much farther. Columbia after all made a steady northwards retreat up Manhattan during the 19th century as the city encroached on it, while Yale was founded well to the north of the commercial center of pre-industrial New Haven, which clustered by the waterfront.

    @Nicolas: The phenomenon of the formerly greenfield university now hemmed in by development seems to be very common. Nashville, which I'm most familiar with, has at least two universities (Vanderbilt and Belmont) in that same situation. Both were founded as small greenfield institutions well outside the city with only a handful of buildings, were overtaken by low-density development in the late 19th century during a period when enrollments grew slowly, then, only after they had been completely surrounded by homes and businesses, began to experience explosive growth.

    This has been a constant source of friction between the schools and the surrounding neighborhoods, which, to further complicate the picture, were historically designated in the 1970s for their Craftsman architecture. Vanderbilt has gone to the extent of repurposing an abandoned mall for some of its hospital staff, but my general sense is that these schools would prefer to maintain a single contiguous campus, even if that requires an excruciating process of house-by-house expansion.

  18. ok, I thought that would be the case, so I would think Cornell could expand onto those greenspaces.

    In the case of Waterloo though, the university has a very large property that they could have expanded onto instead of intensifying, but I guess it wouldn't have been very good to have to take the bus from class to class... Most of the growth to date has been on greenspace and parking in the core of the campus, and I expect future growth to occur mostly on the parking lots that surround the campus and maybe around affiliated institutions right next to the main campus. That's not to say the surrounded residential neighbourhoods are unaffected by growth, but that's mostly in the form of student oriented apartment buildings.

  19. The "grassy quad" idea is not a bad one. It is a variant of the traditional "town square" -- the European version, which is an entirely pedestrian area, not the U.S. variant which is surrounded by automobile roadways on all sides. You tend to see more grass in the British examples. It depends, ultimately, on use -- a heavily used "grassy quad" would become a mudpit if you have an outdoor market or outdoor gathering right after a heavy rain. But, if it is not used too much, then grass is fine and might be quite nice. The "town square" format can be scaled from very small size up to gigantic size, for example Red Square in Moscow or Tianmen Square in Beijing, all within the format of the pedestrian-centric Traditional City. The important thing, in my view, is to get rid of the automobile roadways and useless "green space" landscaping. A "town square" should be a center for human activity.

  20. Interesting discussion.
    Let me add, as a long-term resident of Ile Saint Louis, to the ideas regarding urban parks. The discussion of “room to spare for modest but well-defined park space” on (Ile St Louis) needs qualification—I personally loved the park (Sq. Barye) partly because it was mostly empty, ie. hardly ever used by residents of Ile St Louis, because it was stranded on the wrong side of the Pont Sully+4 lane road that traverses the southern nose of the island. So in fact it holds a lesson for city parks --obviously they are far more useful if within the urban fabric.

    Having said that, modern planners get it wrong, even the French when they have some of the best and densest city environment in the world (I wish today I owned that tiny little studio apartment on Ile St Louis!). Look at the Jussieu university campus opposite the southern end of the island. It is a Corbusian horroshow. The concept was to achieve the necessary density but retain an openness at ground level by building the structures as a grid and elevated on piloti. The "squares" formed by the buildings and the connecting space underneath were supposed to be a lively zone of interaction (I believe the surfacing is marble but unfortunately my memory is of stained concrete). Instead it was a uninviting wind-blown cold desert that has never responded to any attempts to remedy it. Again, totally against Jane Jacobs' prescription of as much ground-level activity as possible. I really don't know why they don't fill in the spaces under the piloti and use it for retail, restaurants etc. (The hideous central admin tower that has only recently been rehabilitated/cleared of its asbestos has a big restaurant at its base but it is about as inviting as a Stalinist collective kitchen.) The French did something similar again at La Defense and are still trying hard to make that place more people friendly.

  21. In fact an interesting comparison could be with the Île Seguin-Rives de Seine redevelopment project of the original Renault car factory (abuts the western end of the 16th arrondissement of central Paris, just downstream of the Eiffel tower). Though it is a much larger site with a big segment on the mainland as well as the island. It is not totally clear if the plans on this English website ( are up to date but there is a New York University (in fact NYU!) plus the American University in Paris on the island.
    The most recent info (though I have not looked thru the French language sites) is about the cultural precinct on the island which is being designed by Jean Nouvel:

    The site has been empty since Renault left in 1992 and perhaps the 20 year delay is no bad thing. The redevelopment of the eastern half of the 13th arrondissement in the 70s and 80s led to a series of unfortunate medium-rise (31 floors) apartment towers that suffer from everything one might imagine from that era: low quality, poor environmental planning; some of them are already looking old and tired. But by the time the last area to be developed--the area next to the Seine eg where the Bibliotheque Mitterand is--they had abandoned such towers and the quality is much better.

  22. @drmichaelrjames: thanks very much for your comments! Very interesting to learn about the park -- looking more closely at the map, I do see what you mean. Thanks also for bringing my attention to that new project. The old Renault factory had an intriguingly urban appearance on the basis of one of those photos -- almost imitating rowhouses along the riverbank. There does not appear to be much of an attempt to recapture those qualities in the Nouvel plan, though.