The remarkably well-connected street grid of central Portland, Oregon.

“Whatever city neighborhoods may be, or may not be, and whatever usefulness they may have, or be coaxed into having, their qualities cannot work at cross-purposes to thoroughgoing city mobility and fluidity of use, without economically weakening the city of which they are a part.”
— Jane Jacobs, The Death and Life of Great American Cities

Both Jacobs and Alexander saw the city as a kind of tissue of human activity and interaction. That spatial structure was in part generated by that activity, and in part shaped and limited the activity itself.

Both of them recognized that the physical arrangements themselves were not enough — neither Jacobs nor Alexander was an “environmental determinist” — but both also recognized that bad physical arrangements could disrupt the critical relationships between the parts of the city, with devastating knock-on effects.

In particular, as Jacobs noted, a formerly viable urban area, fragmented by intrusive structures like freeways or superblocks, could find its economic and social vitality devastated. Freeways are an all too common example, and Jacobs recounted the sad history of New York City. However, parks and other large single uses can also be equally damaging to the urban fabric at their edges. As Jacobs pointed out, the trouble arises when these areas are bisected or fragmented by the borders of these uses, leaving weak fragments. The borders of these uses create what she termed “border vacuums” — something very destructive to viable neighborhoods.

To counter this kind of urban damage, the goal she advocates is to maintain a continuous walkable urban fabric, as a key ingredient along with other ingredients. If the fabric is walkable, the streets are more likely to be populated with pedestrians, and the neighborhood has a chance at sparking some real life. This wasn’t just theory: this is what she observed in the successful regenerating neighborhoods.

Alexander too wrote about the continuous spatial network of the city, so different from the neatly segregated parts of a “tree” diagram — as is so visible in many suburban street patterns. He pointed to many patterns for pedestrian-focused networks, including “Web of Shopping,” “Pedestrian Braid,” “Paths and Goals” and others.

We can think of urban land, Jacobs said, as being of two types. One type is what she called “general land” — the land on which people can move freely on foot. It includes streets, small parks, even building passages.

The other type, which she called “special land,” is all those uses that are not commonly used as thoroughfares by pedestrians. This includes most buildings, such as housing, office, retail, schools, civic and so on.

She argued that there needs to be a well-balanced spatial distribution of these two kinds of land, with a level of permeability for the “general” land, the land for pedestrians. Clearly it was a bad thing if these areas were fragmented by, say, freeways, into small fragments that are no longer functionally viable. And this was what happened to many older neighborhoods that were cut up by motorways.

But another key point was that even the beneficial uses of special land, things like housing and schools, could actually cause a problem if they simply got too big. They would fragment the general land at the borders, and create again the problem she termed “border vacuums.” The dead ends that result would cause a die-off of human activity.

For those who know the work on Space Syntax, Jacobs’ is a very similar kind of analysis — if streets lose connectivity say, because a motorway blocks off local streets, then the network connectivity is affected in a much wider area surrounding this fragmented element.

Again, the space itself has a structural effect on what can happen within it. (As the saying goes, if you don’t believe that, try walking through a wall!) We can readily appreciate the structural consequences of greater distances, say — we will have to walk or drive farther, and we will have to use more resources. But since we are dealing with a complex system of connections, the properties of interconnections, as well as the distances, are also important. “Connectivity” becomes an important property in and of itself, along with scale.

But this suggests that any large single use — any superblock scheme that tends to remove itself from the surrounding fabric — can be problematic. That probably goes for the superblock schemes of so-called Neighborhood Units, like the famous 1929 Clarence Perry scheme. Jacobs was very critical of this kind of unitized scheme, in large part because of its lack of connectivity to the surrounding fabric, and the resulting damage that she observed.

But what about situations where you have no alternative — say, large parks, or rivers, or other uses? Here she quotes the great urbanist Kevin Lynch, who argued that such border vacuums can be converted into “seams,” if they were given depth on either side of them. She gave the example of Central Park and some of the uses that penetrate on both sides of its edges — for example, the Metropolitan Museum of Art extending in, Roosevelt Park extending out, and so on. The same thing happens at the edge of the East River — the South Front Seaport extends into the water and helps to knit the edge back into the city. Bridges of course do the same thing.

My own home town of Portland, Oregon also did the same thing quite successfully. We took down our Harbor Freeway, which fragmented the city, and put up a linear park, and then we added uses along it, so that it does not remain a “border vacuum.” And there are quite a few bridges across the river too, so many that one of Portland’s nicknames is “Bridge City” — but they do help to knit the city together across what would otherwise be a very fragmenting border.

And of course, when taking down such a freeway, we always have to ask if we’ve compromised the mobility that a city needs. And I’m happy to say that Portland has had very good mobility — certainly on a par with other cities that have built more freeways. In fact, Portland has freeways that run right into the city — and of course that’s another kind of structure that can cause great fragmentation and damage to urban fabric. Portland however has a very noteworthy feature: significant parts of the freeways through the center of town are actually submerged, and the walkable street grid continues above it, with pedestrian sidewalks, light rail and other modes of travel.

Again, this follows Lynch’s formula for stitching together a seam across what would otherwise be a border and a barrier, and thereby activating the areas around it.

Portland also demonstrates some other surprising examples of otherwise large uses that would typically create “border vacuums” that are in fact integrated into the walkable urban fabric to a surprising degree. The campus of Portland State University is not in fact segregated in a typical isolated superblock campus setting, but instead is integrated right into the walkable urban grid of the downtown area. The same is true of many of Portland’s inner-city industrial districts, also integrated into walkable urban fabric. These are quite appealing, walkable neighborhoods, and they support significant mixed use, including high tech and office. An example is the well-known Pearl District, also a former industrial district that still has industrial and office users (including Microsoft and Google), and it remains very walkable and appealing.

What about shopping malls, “big boxes” and so on? In the very center of Portland is a shopping mall called Pioneer Place, and it is one of the most popular in the region. It spans over four blocks — but instead of taking out streets, it preserves the street grid and uses tunnels and bridges to spread into a larger complex, right over the walkable urban fabric, including the light rail line.

What about hospitals? Portland has a major hospital complex called Good Samaritan, again spanning over a number of blocks, and using bridges and tunnels to do so. Again the surrounding urbanism is very beautiful and walkable. These are successful examples from a “modern” city with a successful economy.

Indeed, Portland has managed to keep a kind of continuous carpet of walkable urbanism, right across the city. It maintains these walkable connections within a network of principal through streets that’s about 1/4 mile, or 400 meters. That number seems to be closely related to the optimum balance between pedestrian mobility and vehicular mobility, or the scale at which pedestrian-dominated areas give way to vehicle-dominated areas. It is not that these larger through streets are not pedestrian-friendly — indeed they can be, and must be — but that this is the point where pedestrian-dominated streets (such as narrow lanes, “woonerfs” and the like) give way to longer, straighter avenues and boulevards where vehicles have more free movement.

Our colleagues Sergio Porta and Ombretta Romice in the Urban Design Studies Unit of the University of Strathclyde, working with their students, have shown that this 400 meter number seems to be surprisingly invariant across many cities. We can see for example in Bologna, where the major through streets average about 400 meters on center (see illustrations below). In Oslo, we can see again the same pattern. Or in Paris, once again we find same thing. We note how Paris, like Portland, has grade-depressed railways and motorways, and the urbanism continues very beautifully overhead. In fact London does the same thing — a typical example is Oseney Crescent in Camden Town. This is part of the essential railway service to London. Notice again that the street level still offers a very walkable, attractive streetscape without “border vacuums”.

So we can see that it’s possible to build cities this way, and to maintain this walkable urban carpet, even in a thriving modern urban economy like London or Portland. It’s not necessary to chop them up in the name of mobility, as we did in the United States, very much to our regret. In fact, if we’ve learned anything, it’s that the more we try to build for mobility, the more we tend to lose it — the paradox of “induced demand.”

Another implication is that it’s not necessary to push arterial highways out to the perimeters either, where they too often become generators of sprawling “out of town” facilities. As we have seen in the United States, this kind of planning only serves to trigger the growth of even more low-density, low-connectivity, car-dependent urbanism. Instead we can take these important arterials, and all their movement and all their people, right into the heart of cities, as Portland shows, so long as we keep them grade-depressed, like railways. (Tunnels are also an option, though more expensive still.)

But one may ask, is vehicular mobility still accommodated at the finer scales? Yes, indeed, there is a remarkably effective permeable network, graduated in a progression from the local streets to the walkable avenues, like Portland’s Hawthorne Boulevard. At the next level are multi-way boulevards, like this example in the Willamette District, just outside Portland — with slip lanes and accommodation transit. Notice also the significant on-street parking that is provided, with four lanes during peak periods, and additional on-street parking during other times. This avoids the urbanism-killing big parking lots of American urbanism. In this kind of boulevard layout, the travel lanes could be up to six or more lanes, as it is in Paris and other cities, and remain pedestrian-friendly with an additional median. And as these examples show, the pedestrian realm here still has some generous features.

Going up the scale of street sizes and the vehicular mobility they provide, we finally arrive at the fastest kind of throughway, the grade-depressed freeway system that we discussed before. The result of this mobility, along with other factors, is that Portland’s inner core areas are remarkable success stories of regeneration over the last few decades.

I am pleased to report that on carbon metrics, Portland is also doing well relatively speaking, although there is much more to do. By a recent per-recession assessment, Portland was 14 percent below its 1990-level target, which as far as I know, has not been duplicated by any other US city. We all have a long way to go, but this is an encouraging indicator.

Portland has also come a long way from its declining urban core of the 1960’s, and I think we can begin to see the importance of taking down the elevated motorway and re-establishing these other links. I think we’re seeing confirmation of what Jane Jacobs observed, confirmed by other research — that when urban areas are fragmented by freeways and other barriers, it creates other kinds of damage to the urban tissue around them for some distance. People become isolated. Businesses shutter their doors. A whole series of spiraling negative conditions kick in.

Jacobs also suggests that it’s possible to reverse-engineer these phenomena, and “unslum” the damaged places, by reconnecting them to the wider urban fabric. That connectivity to the wider city and its diverse economy helps to diversify the neighborhood itself, and bring more opportunity. That’s a very hopeful prospect.

Photo Essay FOR CHAPTER V.1:

A typical day living in Portland’s continuously well-connected, walkable urban fabric

Figure V.1.1. Walking, biking, transit and car are all viable choices for travel in Portland’s well-connected urban grid. This intersection, next to the celebrated Powell’s Books, includes streetcar, bikes, walking, cars, bus, and a light rail connection three blocks to the south.

Portland is a case in point of the renaissance of which I speak. It is a modern American city with a prosperous economy, and yet it also demonstrates quite well the urban structure that Jacobs and Alexander championed. It thereby demonstrates that sustainable urban development is not antithetical to a prosperous economy with ample opportunity for employment. Indeed, as this book argues, these things are increasingly tied together.

In the 1960s and 70s, Portland was a fairly dreary town, not so different from Pittsburgh or Chattanooga in those days. But as in those other cities (and countless more), city leaders here were inspired to change things. Jacobs and Alexander are often mentioned as inspirations in part for these changes, and Jacobs herself came here occasionally to comment on and encourage the work. Alexander’s “pattern language” methodology was also used to develop much of the planning for the city in those years (as well as the better-known University of Oregon, recounted in his book The Oregon Experiment).

Figure V.1.2. Downtown Portland in the 1960s, full of freeways and parking lots. The riverfront freeway to the left was later removed to create the much-loved Tom McCall Waterfront Park. Photo: Oregon Historical Society.

Figure V.1.3. Pioneer Square, often described as the city’s “living room,” occupies the site of a former parking lot. It now has two light rail lines passing by it as well as a streetcar line nearby.

Luckily, the city had great urban bones — its small walkable blocks, streetcar grid and diverse mixed-use fabric. The city’s remarkable renaissance was begun with some added public spaces (Pioneer Square, Ankeny Square etc.), appropriate contextual infill (Pioneer Place, RiverPlace, etc.), adaptive reuse of existing buildings (EcoTrust’s Vollum Center, University of Oregon Graduate Center, etc.), and a mix of light rail, streetcar, bikeways and better bus service.

FIGURE V.1.4. Portland’s remarkably well-connected walkable grid, stretching continuously across the river, freeways and other obstructions. Principal through streets are spaced at roughly ¼ mile, or 400 meters, which is a common pattern in walkable cities. The author’s apartment building is at the circled dot on West Burnside, to the center left of the map. Image: Google Maps.

Figure V.1.5 Portland’s walkable grid stretches right across the 405 freeway in West Portland (at bottom of photo), with light rail, streetcar and bus lines, as well as bike pathways and wide walkable sidewalks. At right is the Northwest neighborhood, and at left is Goose Hollow. The author’s apartment is at center top. Image: Google Maps.

To give a flavor of the urban renaissance that happened here (and can happen elsewhere), I will offer a personal report on life in one particularly livable part of town, where I have my own home and office.

I live and work in an area that adjoins Portland’s well-known “Northwest” neighborhood, also called the Alphabet District or Nob Hill. I actually live just across the borderline, in an area called Goose Hollow. Geographically, However, Goose Hollow and Northwest are part of a continuous urban area bounded by the West Hills on the west and south, the river on the north, and Downtown and the more famous Pearl District on the east.

Figure V.1.6. Farmer’s Market in Portland’s Northwest neighborhood, a beautiful and functional place to live for the author and many others.

This is a remarkable place — close to a textbook example of well-connected, mixed, walkable, multi-modal urbanism. It’s economically diverse, with people of very modest incomes as well as wealthy people and those in between. It’s also very beautiful and livable. I live in a historic 1911 “courtyard apartment” — very typical for the area — with a balcony overlooking the neighborhood. It’s one of 45 units on a 10,000 square foot lot, or about 930 square meters. For planning geeks, that’s a net density of 196 units per acre, or 480 to the hectare — unusually high by the standards of most American cities. The gross density of the neighborhood is above 20 units to the acre — about 24 or so depending on how it’s measured — which is one of the highest in Oregon. (Yet the neighborhood has very few tall buildings — a point that fans of tall buildings should note, along with other cautionary evidence. My very livable yet high net density apartment is 6 stories.)

Figure V.1.7. The author’s 1911 courtyard apartment building, typical for the neighborhood, with a net density of 196 units to the acre (about 480 to the hectare). The author’s apartment is on the far side, with a balcony overlooking the neighborhood.

I live and work in a small one-bedroom unit with quite reasonable rent by US West Coast standards, about $1.55 USD per square foot per month, or about $17 per square meter. That includes heat, water and sanitation. It should be noted however that rents have been soaring here in recent years, a result of the phenomenon I have called “voodoo urbanism” (see chapter IV.5). I also pay about $25 per month in electric bills, using the local utility’s renewable “green” energy option. I don’t own a car, and I bike or take transit to most locations, and use car-share for others I can’t get to. The location itself is a major factor in affordable (as well as very pleasurable) living.

Figure V.1.8. Although Northwest Portland is one of the densest neighborhoods in Oregon, its diverse mix of housing, including single-family detached, duplexes, rowhouses and apartments, makes it remarkably livable and attractive.

On a typical day here I wake up, shower and get dressed, make some coffee, answer some emails, then walk or bike to a nearby deli to read the paper, say hello to friends and acquaintances, and eat a bagel or some huevos rancheros. Then it’s back to the office (in my apartment) to make some calls, do some work, or prepare for a meeting. On some days I teach an on-line course from my home at a university where I am an adjunct. Frequently I’ll head off to a meeting, most often on my bike, which is often faster than driving or transit would be. On the rare occasion that I need to drive to a meeting, I check out a ZipCar from just around the block, or I carpool with colleagues.

Figure V.1.9. Elephant’s Deli, where the author often walks to have breakfast, read the newspaper and visit friends.

For shopping I can bike to three different grocery stores within about 4 blocks of here. One of them is a high-end specialty store, another is a warehouse-type organic store, and another is a more generic and affordable grocery store. My bank, dentist, laundry, print shop and other routine services are also in the same area. I can also bike easily to the famous Powell’s Books (see Figure V.1.1), or other downtown department stores and specialty stores, for office supplies, clothes, or household goods.

Figure V.1.10. The Portland Streetcar passes in front of Good Samaritan Hospital and a row of shops offering neighborhood services.

I do have to drive to visit my two daughters and five grandkids (or they have to drive to visit me) a few times a month. Both live in smaller nearby towns close to their workplaces. One family lives in a beautiful small town up in the Columbia Gorge called White Salmon, and another in an exurban town called Wilsonville, farther south in the Willamette Valley. For those visits I check out a ZipCar, since neither family is on a reliable transit connections to here. (This is a significant problem for the suburbs here, although the city itself has quite good transit.) For most trips, including regular business meetings, I am able to bike or take transit. When I head to the airport for a business trip, I use the light rail line, which has a station a few blocks away from here.

It’s true that this neighborhood is not as practical for families with more than one or two children. When my wife and I were raising our three daughters, we briefly considered living here. However, we opted to live in a nearby suburban town, and we did own cars — although the neighborhood was also walkable and bikeable, with well-connected mixed-use streets and reasonably good transit. (The core of Portland and other cities is not the only kind of place that is experiencing a renaissance of urbanism.)

Figure V.1.12. Comparison of Portland’s connected, walkable urban form to 20th Century models criticized by both Jacobs and Alexander. Top left: typical isolated “drive-to” shopping mall surrounded by parking. Top right: Pioneer Place, a multi-block complex connected by tunnels and bridges, and allowing the walkable street grid to continue at grade, and serving pedestrians, bikes, cars, and two light rail lines. Upper left: A typical suburban hospital “supercampus”, and upper right, Portland’s Good Samaritan Hospital, again a multi-block complex connected by tunnels and bridges, allowing the street grid to continue at grade, serving pedestrians, bikes, cars, and a streetcar. Lower left, a typical isolated university campus, and lower right, Portland State University, once again a multi-block complex with a continuous walkable street grid throughout, connected also by bridges and tunnels, and featuring a streetcar. Bottom left, a typical suburban industrial supercampus, and bottom right, Portland’s Inner Eastside industrial area, including a light rail line. Photos: Google Maps.

Figure V.1.11 A number of former residences have been converted to retail uses at the ground floor along Portland’s NW 23rd Avenue forming a complex and spatially attractive streetscape.

To be clear, Portland is far from perfect: it has its share (or more) of problems with over-gentrification, loss of affordability, displacement, inequality, homelessness, traffic congestion, poor-quality over-development, loss of livability and heritage, and other common urban ills in the USA. But it has those problems in spite of, not because of, the near-perfect urban form in its core. Bigger problems occur in the sprawling suburbs, where over two-thirds of the region lives, and where car dependence, traffic congestion, poor transit service, and other related ills occur, as they do in so many other cities, new and old. However, another more recent kind of problem is also occurring in the core of the city — a destructive kind of hypertrophic growth, resulting from ill-conceived policies that amount to “killing the city with kindness.” This is the phenomenon I referred to as “voodoo urbanism”, discussed in more detail in Chapter IV.5.