« confessions of a spatial navigator | Main | san francisco: the paper clip snaps »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Alex B.

My Dad got a new car not too long ago (a Subaru) with a built in GPS device. The console runs all the car's electronics, too - stereo, etc. I love it, however, simply for the map. You can fix it where North is always up. Whenever I'm back home visiting the folks and I borrow the car, it's never a matter of putting in the directions - but I love having the map there all the time.

It would be a really cool feature to have - if I needed a car in DC. I guess that's what the Google Maps app is for with the smartphones.

Multimodal Man

One could argue there are "me-centered" spaces versus "objective" spaces in the built environment. When I was a rural land use planner and had to do deal with the Public Land Survey System on a daily basis I felt confident of understanding the whole when I delved into the details of a certain parcel. One the other hand I have, in my life experience, looked for obscure addresses in a developing country only to find no order or reason to a street, except that each homeowner selected their favorite number to afix to their home. In fact, it was often the case that the house number moved when the the home owner did, much like people who retain a cell phone number from some obscure area code from rural South Dakota when they move to Los Angeles.
The strongest "objective" organization of the built environment that I know of is Salt Lake City and by extension, the state of Utah. All address numbers in the city
derive their value not just from some downtown intersection but rather the block housing the Mormon temple. The Utah PLSS is centered on the southeast corner of this block.
Could one say there are "me-centered" cities? Boston? London?


I know exactly what you mean about those mapbooks where each page is a different part of the city. I have always avoided those, as I find them very confusing even if the pages are in sequence from east to west. It is indeed almost painful to read them, despite the advantages of their smaller, pocketbook size.

Let me add though that while I favour GPS systems with north as a fixed point, it is possible for a spatial navigator to learn how to adapt to a "me-centered" navigational device. When I first started playing Flight Simulator, I found it impossible to picture the direction my plane was flying in, because all the directional devices - GPS and the instruments - are "me-centered". When you turn your plane, all that happens is the circle with the numerical dial (0 to 360) rotates. You also have a purely numeric digital indicator and a numeric compass.

Even though it took me a while, over time and with practice I was able to easily have a mental picture of my direction, purely because of the numbers. If my heading was 114 and I had to adjust to a heading of 155, I could picture in my head these directions, because in my mental image 0 degrees was always at the top. So the numbers took the place of the visual reference, because I could use the numbers to create a visual map in my head. Without the numbers however, I would have been totally lost.

Multimodal man brings up another interesting point. Two cities I have lived in - Toronto and Portland - are extremely objective (spatial) in their organization. Both cities have a grid formation. In Toronto, Yonge Street is the dividing line, and streets that cross it have an "east" or "west" designation added on depending on where they are in relation to Yonge street. So "X Avenue West" becomes "X Avenue East" as it crosses Yonge street. I believe it is actually illegal there for any street to use the term "west" if it is east of Yonge, or vice versa, so as to avoid confusion. Portland is even more spatial - the city is organized into quadrants (NE, SE, NW, SW) and street headings all begin with these identifiers depending on which quadrant they are in. In more "me-centered" cities, a street might be located in the far west section of a city, but still be named "X Avenue East" because it is east of some neighborhood park. To the people in that neighborhood, it makes perfect sense to add the "east" identifier to that street, but to outsiders it is only cause for confusion. In this sense, "me-centered" cities probably foster a greater sense of neighborhoods, but are more hostile to outsiders (and by "outsiders" I mean not just tourists but even people living in an adjacent neighborhood).

There are other, more unconventional ways of organizing a city in spatial terms. For example, a city like New York could be said to be spatially organized because the avenues and streets are numbered sequentially. So if you are on the corner of 58th Avenue and 202nd Street, you can picture visually exactly where you are on the grid, and what direction you need to go if you want to get to, say, 26th Avenue and 98th street.

The interesting question for transit is the extent to which transit systems could be made to be more "narrative" or more "spatial" (objective) in their organization. Would a transit system be made easier, or more difficult, if say the bus lines were numbered sequentially according to some organized system such as their direction, or distance travelled from a key focal point (i.e. downtown)? Some transit systems already use the line numbers for descriptive purposes. Two digit identifiers might be a "regular" bus service, three digits a rush-hour only service. An 'X' tacked on to the end means express service. Would it be possible to design a system where the line numbers were descriptive in spatial terms? And would this make the system easier, or more difficult for people to understand?

W. K. Lis

Whenever I am in a strange city, I always try to orient myself by finding north. When I had an analog watch, I would try to align my shadow with the hour hand (12 noon standard time is north, or 1 PM daylight saving time is north). With a digital watch, I have to imagine myself as a sundial to do so.


Growing up in Chicago as a spatial navigator was remarkably easy: the lake (Lake Michigan) is always east! And since it's so flat, you can usually tell where the lake is, even if you move inland a mile or more. Although I wonder if that's confusing cause and effect: I'd be interested to see how geography like this during our formative years impacts what kind of navigator we become, i.e. I'd expect to have more narrative navigators in a place like Indianapolis--which has no distinctive geography tied to a cardinal direction--than in Chicago, Milwaukee, or on either coast.

Adam Parast

Jarrett great posts, this and the last one. My masters research project is actually about issues like this. Everyones comments are right on target but I wanted to add a few new thoughts.

First abstracted system maps and geographically accurate maps. The reason that system maps like the tube are so successful is because they simply otherwise every complex tasks. Additionally, when you are underground VS navigators as well as narrative navigators are forced to navigate in more or less similar ways. There is no extra information like landmarks to guide actions. What is so important in these systems is the process of going from geographically accurate maps at the first station to a system map and then back again to a geographically accurate map at the destination station often focused on the station area.

Another related related point about buses. In most european cities (I'm studying in Stockholm right now) all stops in the bus system have a name with significant signage, and next stop reader boards in the bus (and possibly annunciation systems). The combination of bus stop name signage, and next stop information allows navigation that is similar that used in metro systems. With this type of system abstracted systems maps may work because the stop name helps to tie the stop to the geography of the city.

In contrast, the US does not do this for ordinary buses. Bus systems in the US are almost exclusively built upon the street network and very few bus stops actually have names. To me this shows a fundamentally different approach to communication of the system to users. Riders must navigate much like car drivers following along as opposed to saying, I have to get off at XX and I don't care what happens in between.

The last point that I haven't seen anyone speak about is a term that I ran across in my research call "tracking". I think it is something that many of us, both visual and narrative thinkers use subconsciously. It is essentially the process of following where one is during the journey process and check to ensure that no error has occurred. For visual thinkers it would be where they are on the map. For narrative thinkers it is which step they are in their journey. It also is the process of error checking. Donald Norman would call it "feedback". For example if you get on a subway and then notice that the next station you are coming to is in the wrong direction, you are "tracking" and the feedback (station name) allows you to correct for the error. This is why redundancy of onboard information is so important.


In psychology, these representations are called "allocentric" and "egocentric". Humans seem to use both types of representations, and they are coded in different parts of the brain.

Louis Haywood

Here in New Orleans, almost everyone uses a system of objective spatial directions, but not everyone is very comfortable with them. On any "east-west street" you are either traveling "downtown" or "uptown", or "toward the city" or "away from the city," or even more simply, "upriver or downriver". If you are travelling north-south, you are either moving "Lakebound" or "Riverbound".

All street names between Rampart Street and the river change names completely at Canal Street, and all other streets change from North to South at Canal (i.e. North Claiborne and South Claiborne). However, street portions downriver are called North, and streets upriver are called South.

What does this mean for transit? Canal street divides the network, and virtually all routes end at Canal Street.


Some cities in the UK, and even a few in North America (Toronto being one example) use locally-based address numbering. The first location on a street has the address number 1 (or some other low number), and addresses go sequentially from there, no matter where the street starts. My wife used to live in Scarborough, ON (a Toronto suburb) and her prior residence had a number of 11, despite being quite distant from downtown.

Most US cities I'm aware assign address numbers relative to some center (the LDS Temple in Salt Lake, the Willamettte River for E/W and the Willamette Stone/Baseline for N/S in Portland)--hence the first house on my street has the number 7605. We're 76 blocks south, give or take, from the Willamette Baseline, hence the number.


Older cities in the US also use addresses relative to the street. I've been living in the Boston area for years, and I've never had an address > 30, even though I've never lived near the center of either Boston or the town I'm living. In. My midwestern relatives giggle when they hear our address, because addresses less than 5 digits seem strange to them.


One other issue is that Portland and certain of its suburbs (which used to be freestanding towns before overtaken by sprawl) have conflicting street numbering schemes, in some cases mixed together in the same neighborhood. In Beaverton, one can find low street numbers (less than 30) running east-west; this is the traditional Beaverton street grid. North-south numbers between 80 or so, and 185, are the PORTLAND street grid. Likewise in Hillsboro, whose native street grid also uses numbered streets for north-south--there, the City has undertaken a program to rename streets using Portland-relative numbers to the Hillsboro grid--an action which isn't entirely popular with those having to change their address.

Portland has a small stretch of land, in the Macadam area in SW Portland, where street addresses start with zero (0123) and the zero is significant. The Willamette River divides east and west, however, it runs slightly northwest-to-southeast; hence in SW Portland there are numerous addresses east of SW 1st Avenue but west of the river; these are given addresses starting with zero. Forget the zero and you go to the wrong address.

I'm sure similar issues are found elsewhere.


To add to Multimodal Man's observation, Salt Lake City also has very few "names" for streets. A lot of street names are the blocks, like 1200 South, 2000 East, etc.

Salt Lake City's addresses are map coordinates. You could visit a friend who lives at 2000 South 1200 West, for instance.

Now, for my hometown's contribution. Here is how the L.A. system works. And it's quite complicated, which is a bad thing considering most Angelenos still have less of an understanding of precipitation than prehistoric societies. :>

The Los Angeles city and county postal compass is centered at the intersection of First and Main Streets in downtown Los Angeles. It is here where you get north, south, east and west.

L.A. also does not use 0-99 as the first block from the compass origin; blocks begin from 100.

Here's where things start getting stupid. Incorporated cities are free to adopt their own postal compasses within their own municipal limits, even on streets that run through many communities.

So you can have street addresses that break a steady pattern. For instance, Santa Monica's east-west streets continue through to L.A., but Santa Monica uses the ocean for its postal compass and numbers increase going to the eastern city limits. At the L.A. limits, the block numbers reach the 12000s but decrease heading east.


It's interesting to look at an agency which recently made a switch from a geographically accurate system map to the "spider map" style, like the Los Angeles MTA. They turned a very useful system map, which showed graphically the closeness of routes in the central city and availability of faster classes of service, into a complete piece of garbage that serves neither spatial navigators nor narrative navigators well. I would be interested in hearing your thoughts.

The old map: http://www.scribd.com/doc/21859893/MTA-System-Map-0108
The current map: http://www.metro.net/riding_metro/maps/images/System_Map.pdf

(Disregard the fact that the new map only covers half of the area of the old map, and leaves out over a million people in the San Gabriel Valley and Southeast Los Angeles County. That was most likely a cost-driven attempt to shrink the map down.)


San Francisco is the same way. It is actually helpful if you know where the origin of the street is. For example, many east-west and north-south streets originate off Market, or off the Embarcadero. Therefore, it is relatively easy to compute the number of blocks away it is from these major points. The problem is the smaller streets.


Thanks for the link, Jarrett! Related to your observations about map-books, I sometimes have a split-second delay in reading clocks, whether analog with hands or digital with numbers. Many years ago I found a little app for Windows 3.1 that displayed the time in words, like "It's almost a quarter to twelve." I found that I could read and understand those with no delay at all! Sadly, I haven't found a similar one for XP, or for my phone.

The most useful map-books I've found are the Paris par arrondissement maps, especially for the inner districts, because you can think of each district as a unit. The Michelin one I have splits the larger, outer districts over two maps to maintain a consistent scale. By contrast, the maps-books I have for Westchester, Nassau, Bergen and Hudson Counties in the NYC suburbs are hard to read for the reasons you give, and would be a lot easier if they had one or two pages for every town.

Finally, with regards to Ailurin's comment, the "allocentric" and "egocentric" concepts also correspond roughly to Witkin's "field independent" and "field dependent" cognitive styles:


Alon Levy

In most cities I know of outside the US - anywhere in Israel, Monaco, anywhere in Italy where I've looked - houses are numbered sequentially. The concept of a regular block doesn't exist, even in cities that developed recently, like Tel Aviv.

In Singapore house numbers don't move with residents, but the numbering is neither block-based nor sequential. It may have been sequential in the past but was obliterated by consolidation of multiple lots - I'm not sure. As in Israel (or France, or Italy, or Britain), the concept of block doesn't really exist.


I am not particularly familiar with Los Angeles, however I have been there a few times. And I believe the new map is almost criminal in the level of deception it propagates.

One of the things about LA is that it is almost impossible to convey on a map the sheer scope and vastness of it. It feels more like a province than a city, with identical little villages interspersed amongst miles upon miles of human and industrial sprawl, oil derries, and emptiness. It is particularly important for someone unfortunate enough to be relying on transit in that place to have a map that accurately portrays the hostile environment they are facing. No map can really portray it, but the old one at least tried. You do get some glimmer of insight on that map of the incredible distance separating San Pedro and the south bay region from West LA. But on the new map, it looks like a short little jaunt.

The old map severely underestimates the distance (and hence the amount of walking) separating bus lines from each other and from destinations. But that is simply because of the size of the place. No map can do it justice. But the new map makes it look like the whole city has a dense network of bus lines serving a compact space! What a farce. The marketing department can put as much lipstick on that pig as they like. It's still a pig.

Alon Levy

In Manhattan, both my mom (as a tourist) and I (as a resident) have found it useful that the crosstown buses are named after the streets they run on. The M14 runs on 14th, the M72 runs on 72nd, etc. I've had much more difficulty understanding the north-south Manhattan buses, which are numbered only semi-logically.


I suspect there is a strong correlation here to learning styles. The current commonly and widely used model is VAK, where we learn using all three styles - Visual, Auditory and Kinisthetic, but have a strong preference for one. A visual learner likes looking at pictures, diagrams and maps of things, an auditory learner likes text or being told about them, while a kinisthetic learner likes doing it.

In a classroom or meeting, the visual learners will draw flowcharts and diagrams, the auditory learner will write reams of notes, while the kinisthetic learner will be ones who have to have something in their hands (often a pen) even though they're not actually doing anything with it.

It seems logical to me that those who learn about a city visually need to see a map - and they like to see the whole thing. Auditory learners need detailed directions, and the GPSs that change to show the direction you are heading at the top would be designed for these people. Kinisthetic learners can use both but need to visit a place to put it into context within the whole city.


here in Aus, street numbers are sequential, and usually numbered outwards from the city centre. as for east/west north/south street names, these are only used where a street is broken, and indicate which side of the break you need to go to. so to me, the east designation meaning east of the park makes perfect sense, as the park breaks the street.


Growing up my whole life outside Toronto I had the same experience -- I came to think of the lake always being to the south, and (by extension) that northbound travel would more than likely be inclined (no pun intended) to be uphill.

This really messed me up for my first semester at school in nearby Hamilton, where the harbour is at the north and "uphill" (the Niagara Escarpment) is at the south. I spent the better part of half a year with north and south totally turned around.

Jarrett at HumanTransit.org

I'll do a post soon on the new LA map. No transit agency is interested conveying a message that transit in their city will be an awful experience.


Cecil Adams of the Straight Dope recently answered a question about the street naming and numbering origins of various cities.


I got a kick out of Tulsa, Oklahoma's naming system. It uses the alphabet in order, and for north-south streets, it names them after cities either to the east or west of Tulsa.


The comments for this thread seem to have vanished...


Oh my god!
I finally know WHY I am the way I am. :)

Severely spatial-navigator here. Delivered Pizza for 14 years, so like a taxi driver I had to be good at plotting a route to places I've never been, but I was always that way and so is my family: My stepfather's directions to our house never included things like "6th on the right", but instead gave the exact distance from the intersection.

"Don't give me directions, just tell me your address." That's my mantra. I can never follow someone else's directions (except for those very rare people who "give good directions"), but give me a map and your address and I'll get there.

Now if you'll excuse me, I have to find my shoes. ;)


"I have had the same experience, in many cities, when I ask for a transit system map. Surprisingly few agencies publish overview maps that show the key elements of the network, giving you the structure on which to hang local details. "

Heh. *Most agencies with trains do so*. *Most all-bus agencies don't.*

This is (one reason) agencies with trains can be used by visitors (at least map-oriented visitors) and all-bus agencies cannot.


I suspect there is a strong correlation here to learning styles. The current commonly and widely used model is VAK, where we learn using all three styles - Visual, Auditory and Kinisthetic, but have a strong preference for one."

Nope. I don't correlate at all; in fact, I don't even fit into that model of learning styles, which is clearly deficient. I love maps and diagrams but I hate pictures and photographs, and much prefer text. And I get text better in an auditory fashion. I prefer "doing" to pictures, but prefer maps and diagrams to "doing"....

Stefan Marks

It reminds me of something I read years ago about map design - "you shouldn't need a Masters Degree to figure it out."

Clearly - very few bus network maps rise to the elegance of Beck design for the Tube simply because of the complexity of a city's network. Any system maps anyone would recommend?

Brent Palmer

"Any system maps anyone would recommend?"
Maybe the ZVV map for the City of Zurich...


New York is somewhat unusual in regard to addresses. With the notable exception of the crosstown 'streets' in Manhattan - which use 5th Ave. as a baseline - the city is mostly a grid with numerical addressing (starting at '1' where any given street begins) and no central point. In the outer boroughs, this can be especially confusing, as Brooklyn and Queens basically consist of multiple grids jumbled on top of each other. So the streets follow some loose pattern of organization, but there is no one singular 'start' point that you can refer to (like, say, the intersection of Madison and State in Chicago). The patterns of the streets can also change somewhat randomly in parts of the city, where different grids overlap into each other.

Justin N

Best grid system ever- Sacramento, CA.

Not only is the entire city laid out on a nice, even grid, with numbered streets running east-west, and lettered ones running north-south, but the block numbers correspond to the number or letter designation of the street they're on. For example, 1800 J Street is on the corner of 18th St. and J St., while 500 14th Street is on the corner of E and 14th. It was heaven.

And I get the whole spatial vs. narrative navigation distinction entirely. (I've always called the latter "woman navigation", though I know that makes me a tad sexist. I study the map of a city before I get there, just to understand it, and I can get anywhere in my hometown with just an intersection. My wife needs step-by-step directions to anywhere.)


Every comment I've seen says something about how I feel.

One aspect I find as a spatial navigator or better yet as someone who is good a spatial orientation. Is that when ever I'm giving directions or getting directions, whether that be from a human or by reading a map. I visualize the journey in my head. I'm picturing myself turn left or right and keeping my orientation relative to everything around me.

I have found that people who get lost easily don't spend the time to visual their journey on the way and so when they make a turn they can only think of straight ahead and not where they are overall.

One thing I like about Vancouver overall is anyone who has been here or lives here knows that north is always where the mountains are. So for spatial navigator all you have to do is point your map towards the mountains.

At the same time I'm not a big fan of the city of New Westminster (a suburb of Vancouver). The problem being is that all the streets are numbered with both the avenues and streets being in the same number range. So if I say to meet me at 6th Avenue and 8th Street. You can quite easily find 6th Street and 8th Avenue. The number are correct just the avenues and streets are reversed.

The comments to this entry are closed.

the firm

Jarrett is now in ...

Related Posts Plugin for WordPress, Blogger...
Related Posts Plugin for WordPress, Blogger...