All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Envisioning Information
Visual Explanations
Beautiful Evidence
Paper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer
connected to the internet:
La représentation de l'information
quantitative 200 pages $12
La Representación Visual de Información
Cuantitativa 200 páginas $12
Visual and Statistical Thinking $2
The Cognitive Style of Powerpoint $2
Seeing Around + Feynman Diagrams $2
Data Analysis for Politics and Policy $2
catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Bethesda, September 28
Baltimore, September 29
Arlington, October 1, 2
Indianapolis, November 9
Columbus, November 10
Cleveland, November 12
San Francisco, December 14, 15, 16
London Underground maps (+ worldwide subway maps)

-- Jim Heimer (email)

Response to London Underground Map

London Underground Map

Harry Beck's diagram of the 7+ lines of the London Underground, although geographically inaccurate, provides a coherent overview of a complex system. With excellent color printing, classic British railroad typography (by Edward Johnson), and, in the modern style, only horizontal, vertical, and 45 degree lines, the map became a beautiful organizing image of London. For apparently quite a number of people, the map organized London (rather than London organizing the map). Despite 70 years of revision due to extensions of the Underground and bureaucratic tinkering (the marketing department wrecked the map for several years), the map nicely survives to this day.

Later European and American knock-offs did not succeed at all. The Underground Map and Minard's famous Carte Figurative of the French Army's disaster in Russia in the war of 1812 are alike in important respects: both are brilliant, and neither travels well. The Underground Map and Napoleon's March are perfectly attuned to their particular data, so focused on their data sets. They do not serve, then, as good practical generic architectures for design; indeed, revisions and knock-offs have usually been corruptions or parodies of the originals. Both, however, exemplify the deep principles of information design in operation, as well as the craft and passion behind great information displays.

There is a fine book on the map: Ken Garland, Mr Beck's Underground Map (Capital Transport Publishing 1994). The book describes the enormous care, craft, thought, and hard work of Harry Beck that went on for decades--exactly what it takes to do great information design and so in contrast to the quick-and-dirty practices and thinking of commercial art. Garland's book is also a model for writing histories of great information designs.

The map makes 3 significant design compromises: substantial geographic distortion, a complex color code for identification, and, in small sizes, it is difficult to read.

-- Edward Tufte

Daily use

For those who don't use the Tube map daily, some explanation. The main problem with the Tube map is that it is geographically inaccurate. This can lead to a very distorted view of London, particularly for those new to the city.

Many stations are geographically very close and you can end up spending half an hour in stuffy tunnels when you could have worked overground for five minutes. The same occurs in reverse.

It is also deficient in the way it describes some stations. For example, the interchange between the Hammersmith and City line and other lines at Paddington. It's a reasonably long walk but they are still in the same physical station.

Of course nobody has come up with anything better and it is still a fantastic map. After a year or so of living in London, I can work out my route in seconds.

Of interest for comparison is this map which shows the London Underground as it truly is, geographically:

geographically-accurate London tube map

-- Simon Rumble (email)

Response to London Underground Map

The beauty of the orthogonal London Underground map is that it provides exactly the information the rider needs - the name of the line (with color coding), the direction of travel (East, West, North or South - the same way that the directions to the platforms are given), and the stops on the way. Distance between stops and geographical orientation are of very secondary interest to finding the correct train. Most stops are about the same time of travel apart and the trains come frequently enough that you don't need to plan to catch the 7:04 am Picadilly Line Eastbound from Gloucester Road arriving at Green Park at 7:22 am (my old commute), for example.

Incidentally, the platforms are also well marked with a map that runs from top to bottom on the wall as you enter the platform. The station from which you are boarding is at the top, and all of the remaining stations on that route in your direction of travel are shown down the diagram. It's very handy for confirming that you really did want to go in the same direction as the train you are about to take.

On the other hand, as you could infer from the geogpahically correct map, you do miss some spectacular architecture and landmarks by traveling subsurface. For this reason, my wife much preferred the bus - until she got stuck in London traffic for an hour covering the same distance as a 10 minute tube ride. I seldom used the bus, as I found the London bus route map to be unintelligible.

A few of the varied London underground maps:

London underground maps

-- Jim Heimer (email)

Historical collection of many maps of the London tube

Here's a wonderful link, The London Tube Map archive:















-- Edward Tufte

above ground / below ground (Response to London Underground Map)

Transport for London are replacing surface street maps at bus stops with London Underground-style maps, which is just dumb, eliminates useful information about surrounding streets, alternate routes via walking, etc.

London bus map

Other kinds of information are also being shown with Underground-style maps, for example this map of independent coffee shops:

London coffee map

-- American in London (email)

Madrid Underground Map

The map of the Madrid Underground (Metro) is also very well designed. It may have been inspired by the London map, but the color coding is much better (not so many hues of brown) and you can always tell when an interchange is a short or long walk.

Madrid Metro

Also, the signs inside are excellent, it's hard to get lost. You always see a list of coming stops in the entrance, in the platform and inside the train. Geographically accurate maps are found in the exits and in every platform. Finally, each time a new station or line is temporarily closed or opened, the signs and maps in every line and coach get updated overnight. It's a huge concertated effort that should be merited more often.

-- Alex Fern??ndez (email)

Moscow Underground Map

The Moscow Metro Map ranks up there as well, in my opinion. Some things are hard to put forth in a map however; for example, some trains only go as far as the last transfer on the dark green line, and then switch automatically to blue, while others go all the way. Only the locals, and those who can listen for the announcement in Russian, ever figure this out.

Moscow Metro

Other than that, the directions are superb. The interior signs are well-placed (again, you learn as much reading Russian as you have to) in most of the former Soviet cities, as is the case in other cities in the area, too bad their maps don't really do it like Moscow's though. For example Kiev:

Kiev Metro

Note added by ET: In the Moscow map, gratuitous noise and content-free shapes are generated by the figure-ground contrast and activated negative space. Thus visual clutter. See Envisioning Information, chapter 3 on this.

-- Tom Hickerson (email)

Response to London Underground Map

Interesting thoughts about the Underground map here.

-- Edward Tufte

NYC and London map records

Some marginalia around underground/tube maps:

NYC: both Todd Glickman and Peter Lloyd maintain records of NYC subway maps, which include a series of six or so issued in the days/weeks after 911. By all accounts, that period was handled very well. These maps occasionally come up on eBay and tended to have: a box which stated when the map was reissued and referred travellers to mta.nyc.ny.us for more updates, and a speech-bubble enlargement showing the lower Manhattan area.

London: the best records of issued maps are probably Letch's London Transport Bus and Tube Maps 1920-2000 and Burwood and Brady's London Transport Maps 2nd edition, 1983.

Picking up on ET's comment that the London tube map is highly optimised for its context, can we recognise cities from the thumbnail images on these Google Images searches? (and does that actually tell us anything useful about their design?): subway map, tube map, metro map.

-- rodcorp

Response to London Underground Map

Some interesting maps from this discussion board (now sadly with many broken links).



San Francisco and Bay Area

-- Tom Carden

Response to London Underground Map

That's a wonderful use of Google Images by rodcorp--to make international comparisons across subway maps.

The forum cited by Tom Carden has a great many interesting maps.

-- Edward Tufte

Response to London Underground Map


I don't have pdf versions at the moment, but you can read the same texts below this second illustration (which doesn't use the tube map).

Invisible Cities

Why the tube map? In IC, Marco Polo visits Trude but can never leave ("You can resume your flight whenever you like," they said to me, "but you will arrive at another Trude, absolutely the same, detail by detail. The world is covered by a sole Trude which does not begin and does not end. Only the name of the airport changes"). It's as if the airports in the world have become mere stations in the transport system for a single city. So I imagined the London tube map as an airline route map, with the flight looping up and then back to the next station along, and redrew it. (Kathy Prendergast's Lost map, in which all the place names in the US are erased except those that contain "lost", is perhaps a more elegant take on a similar idea.)


And why Invisible Cities? A long-standing love of its qualities: its economy, structure, scope, its 'visual' nature, memory, signs etc. And the lack of any maps for the territories IC describes, which triggered the initial interest, though I note that Calvino's estate wanted to keep the books map-less, which makes sense for the longevity of the published book I think. Nonetheless I thought I'd start making my own, though they're not traditional maps by any means.

best, Rodcorp

-- Rodcorp (email)

Response to London Underground Map

Picking up on the geographic distortion in the map, it's clear that for some journeys it's really not worth getting on the London Underground: it takes a long time, and costs you money. Sometimes it's quicker and easier to walk.

So here's the tube map with some small additions to show which stations are an arbitrary and as-the-crow-flies 500 metres apart from each other:
London Tube Map with Walklines: sometimes it's quicker to walk

You can see the 'walklines' superimposed on the standard map, the same with a faded out tube map, or just the lines. It makes the tube map less readable (not a good thing), but it was an interesting exercise.

London Tube with Walklines

London Tube with Walklines faint

London Tube with Walklines lines only

Update: An October 2008 revision to the tube map adds a few distance markers to the map. It's 250m between White City on the Central line and Wood lane on the Hammersmith & City. And to the right, 100m between Shepherd's Bush Central line and overland. There's a similar thing at Canary Wharf Jubilee line: it's 200m from Canary Wharf DLR and 150m from Heron Quays DLR. I particularly like the angle they're drawn at - like an elegant version of the walklines I was trying out a long time ago.

London tube 2008 distance markers

But this isn't quite so successful: there's 100m between West Hampstead overland and West Hampstead Jubilee line, and the latter also has a comment that West Hampstead Thameslink is 200m away from the tube. The tube map used to simply show the tube and overland stations as a single node on the map, which I think was better.


-- Rodcorp (email)

Response to London Underground Map

The walk-lines are superb! Classics are rarely improved by tinkering, but the walk-lines are a genuine improvement to the Underground map. Bravo!

-- Edward Tufte

Response to London Underground Map

Prior to finding this thread about the Tube Map, I discovered Rodcorp's site while searching for information about the "flaws" in Beck's map. Your maps and charts are really beautiful, and I'm wondering if (this question could be in response to Jeffrey Berg's most recent post too) you followed this project completed last year in Amsterdam?

Here's a bright idea -- an e-map that shows routes people actually use. With GPS tracers and GPRS networks, the Amsterdam RealTime project tracked 75 volunteers as they walked, cycled and drove around the city. After 40 days, the cumulative data lit up this map: The moretravelled the route, the brighter it glows. (The really hot spots are red.) RealTime was part of an exhibit at the Amsterdam City Archive last fall. After receiving requests to make similar maps of Brussels, Lisbon and Paris, the project is considering taking its code on the road. Next up: licensing the technology or distributing it as freeware. --Wired

Amsterdam map

As a city with incredible public transportation and a sophisticated graphic design presence as well, it is interesting to see how human movement through the city (regardless of transportation mode) is broken down to a graphic level. The resulting maps are really quite beautiful, and are an interesting compliment to your Underground map with walking routes.


Jamie Hazlitt

-- Jamie Hazlitt (email)

Response to London Underground Map

Which neatly ties together this thread with the one about traffic signals and roundabouts.

(In particular the Swindon Magic Roundabout mapped by GPS)

Swendon roundabout

-- John Morse (email)

Response to London Underground Map

In all this discussion, I'm surprised that the Tokyo subway gets no mention! It also has a lovely map. I took at photo of the system map at Shinjuku Station when I was there recently. Note that the ticket dispensers for each line are color coded to match the line. The Tokyo Metro website versions of the maps don't do justice to the physical maps at the station. When granted adequate lateral dimension, the map comes out very nicely.

Tokyo subway map

When riding the subway, the older cars have a light that blinks on a scale map showing the next stop station. The newer ones have an LCD flat panel high resolution display that cycle between the map, the Roomaji (Roman characters) name of the station, the Hiragana (Japanese phonetic) name of the station, and the Kanji for the station. I was very comfortable getting around in spite of having minimal Japanese language ability.

Tokyo subway display

Tokyo subway display

Tokyo subway display

Tokyo subway display

I also found their city maps and printed maps to be superb. The Shobunsa map of Tokyo was indispensible, with subtle use of color, lots of detail, and a convenient carrying envelope.

Shobunsa Tokyo map

-- Phillip Remaker (email)

Response to London Underground Map

A previous post mentioned the Moscow metro. Here is the current metro map:

Moscow Metro Map

Of course, it is the design of the stations, not the map, which sets the Moscow subway apart.

Novoslobodskaya station

-- John Morse (email)

Response to London Underground Map

Here's an interesting subway map style:

Sao Paolo subway

It's for the city of Sao Paulo, Brazil. They've drawn the network in more of a geographical layout and have also incorporated some of the major streets, parks, and landmarks.

ET note: Figure-ground contrast is too strong, making white lines too active.

-- Stan Musick (email)

Response to London Underground Map

Mike asked about the version of the London tube map produced at the Delft University in the Netherlands:

I mention it because I think it was a great improvement on the original Beck version. The Delft version accepts that the underground map is divided into two areas, the inner centre and the outer regions and that each has a different purpose: the inner to allow you to navigate around the centre of London; the outer to get you successfully in and out of London. So the group hit on a simple but efective compromise. The outer area was drawn diagrammatically, whilst the inner area was drawn topographically. I occasionally travel into London from my home in Cambridge, and would find this version ideal. It would get me to London and then give me enough geographic orientation whilst I am there. It is perfect for my purpose!

Mike, thanks for pointing it out - you made me go back to the bookshelves and dig it out. In Visual Function the map is reproduced small and in b&w, so it didn't make the best scan, but here's a detail of the Delft version, and this is what he wrote about it:

Delft London tube
For metro-type systems with few stops and long distances in-between [a radically schematized map] is not so much of a problem. But in the case of fine-mashed transport systems like tram and bus networks, it is vitally important to retain a recognizable reality [...] In the Delft version, underground routes in the center of London are rendered topographically, those outside diagrammatically. In the city center the map is augmented with references to major landmarks like parks, places of interest, and museums. This enables tourists, for example, to plan their visit better and avoid bizarre detours. [p6]

I wonder if the Delft version was produced primarily with the tourist in mind - I'm not sure of the utility for the commuter traveller really. But I'd love to try it.

-- rodcorp (email)

Response to London Underground Map

Has anyone seen the London underground map in 3D rendered by Corey Clarke which takes it to a different level? Wow!

Clarke 3D London tube

-- sean (email)

Problems with large-scale 3d maps

Large scale 3D maps are quite difficult. It is revealing how 'little' of the third dimension there is when you move to a large scale. If you do some quick calculations on the London Underground and then draw a 3D map to fit an A4 (297 mm x 210 mm) sheet of paper. My figures come from a very quick web search and should be accurate enough. The longest dimension is East-West at 72 km or 72,000 m; Vertically it is -21 m to +150 m or 171 m; This gives you a vertical / horizontal ratio of 421 to 1. In other words about 0.7 mm in the vertical or about 6 sheets of paper on our A4 map. Hardly worth the effort! The world becomes two dimensional very rapidly.

ET touched on this in his criticism of the vertical exaggeration of computer generated images of Venus in Visual Explanations.

-- Andrew Nicholls (email)

Response to London Underground Map

The geometric purity and beauty of Beck's map is wonderful and is a big symbol of the Underground; the geographically accurate map is also just fine.

It is often a good idea to use an aerial photograph as a base grid for a transportation map. See, for example, pages 108-109 of Envisioning Information.

-- Edward Tufte

Response to London Underground Map

Consider this geographically accurate map overlaid onto a NASA night-time satellite image of London:

Satellite image + tube map

The original blog post with the image is no longer available, but further discussion can be found here over at Signal v Noise.

-- Edward Tufte

Map history

The NYC Transit Museum displays examples of subway maps from the system's entire history.

There was an excellent episode on the TV Series "Design Classics" devoted to the history and development of Beck's Underground map. At one moment, they showed an animated version of the Underground map as it is morphing into a map that reflected true geographical distances--a most dramatic comparison.

-- Seth Joseph Weine (email)

Metro Maps of the World

To all metro and map geeks, I commend Mark Ovenden's book Metro Maps of the World which, for some reason, doesn't seem to have been mentioned anywhere in this thread.

-- Dominic Sayers (email)

Response to London Underground Map

A substantial guide to the London Underground Map and its many many variations.

-- Edward Tufte

Response to London Underground Map

Another interesting idea to get around geographic ambiguity is to add more information about travel times. While the map of Portland, Oregon's MAX Light rail lines in the following link is somewhat cluttered, it does illustrate my point.


This map covers some 33 miles (end-to-end on the Blue line), but it is compressed in an orthagonal way.

The interesting thing is the blue bar at the bottom which shows approximate trip times between important stops. One can easily calculate that the time it takes to get from Hillsboro to the Airport is 87 minutes simply by adding things up.

I would have preferred they use increments of 5 or 10 to keep the addition quick and easy, but it's definitely a start. Exposure of schedule details for a "metro" style transit system is largely unnecessary since people aren't ever standing around waiting very long.

Overall, this map fails to meet the "London Test", but it does add the interesting temporal element.

-- Aaron Huslage (email)

Indicating service interruptions - London

London Underground uses the map on a web page to provide passengers with the up-to-date status of their network. I use it daily. It can be seen here.

Useful though the realtime information is I feel that the display is inverted. Problems are shown by the line being rendered in its conventional colour. At the moment, as LU deals with the debris from the bombs of July 7th 2005 the Circle, Hammersmith and City, Piccadilly, and part of the District are all shown in their normal colours. On the day of the terrorist attacks the entire network was shut down and presumably the display looked like a conventional map. That could have lead some people to infer that the network was functioning normally. Perhaps rendering problem line(s) in the grey currently used for working lines would be better.

ET response: Trevor Jenkins' point is exactly right. What has changed is not the lines that have continued operating but the lines with problems. Thus the non-operating lines should be grayed down in the representation. This becomes clear when looking at the disconcerting display.

-- Trevor Jenkins (email)

Response to Indicating service interruptions - London

The one advantage to note with using colours for the lines on which there are problems is that this makes it easier to identify which lines are encountering problems when more than one line is affected. For example, currently, one can see that the circle line (in yellow) has problems straight away, whereas if this were greyed out, one would need to work that out by a process of deduction. As a user of the underground, one wants to know which lines have problems and the colour coding helps in this.

-- Will Oswald (email)

Indicating service interruptions - Paris

Perhaps a better approach than graying-out lines (or blinking and flashing!) would be to have an "X" marking the problem area rather than the whole line. For example, take a look at how Paris displays service interruptions: http://www.ratp.fr/

The obvious thing, however, would be a legend that would list the lines that are not working and the delay.

Central Line      DELAYED      15min
Circle Line       DELAYED      25min 

-- Tchad (email)

Indicating service interruptions - LA

The Los Angeles Area has had a map available in various incarnations (from a real-time display on the public-access cable channel to various websites) for years featuring a map of the freeway system and mapped data points showing traffic speed. One of the better recent incarnations can be found here.

The L.A. map strikes me as better than the Paris traffic map, in that the speed and stoppage info is shown at point of measurement rather than entire road segments, the colors are less garish, and the highway and place names are more readable.

-- Matt Grimaldi (email)

Response to London Underground Map

Mexico City's subway map displays both the train lines and the city's major streets, thus avoiding the problem of exaggerated and/or inaccurate representations of the geography; certainly the fact that Mexico City is more or less circular helps make this possible. Here's a link: http://www.metro.df.gob.mx/red/index.html

More interesting from a design sense, every station in the Mexico City system is identified by both its name and by an icon representing a landmark -- a church, a famous statue, etc. -- that's visible from the entryway to that station. It's an ingenious way of conveying information to the many Mexico City residents who cannot read.

I haven't found a single page that displays every icon, but they can be viewed individually and in their linear groupings at the subway's official site: http://www.metro.df.gob.mx/red/linea1.html

Lastly, here's a page that displays many of the world's subway systems to scale, and in their actual physical forms. It's a useful reference for anyone interested in subway maps, and it's just plain interesting: http://www.fakeisthenewreal.org/subway/

-- Tim Heffernan (email)

Response to London Underground maps

See the London Underground Corporate Identity Colour Standards at http://rodcorp.typepad.com/rodcorp/2005/07/what_colours_ar.html

rodcorp provides a PDF link to the full document, which is excellent.

-- Edward Tufte

Response to London Underground maps

The full document includes excellent samples, but the pseudo-color chips seem unnecessarily repetitive-especially because they are all rendered as RGB.

A few years ago an excellent signage color scheme was put in at Chicago's Midway Airport. Within weeks it was camouflaged by some season-long event advertising that used the same colors and hung from every lamppost. At any useful distance you could not distinguish parking directions from yet another invitation. I see a parallel to ET's instruction that "the strategies of magic suggest what not to do" (VE, Ch. 3). Do these identity standards ever proscribe advertising or other signage that competes with the identity?

-- Dave Nash (email)

Time maps

I found a new map of the London Underground today that was pretty interesting. It was a project of a graphic design student in London. His concept was to base the measurement not geographically, but in time. I think it is an interesting take on an already interesting map. Real outside the box thinking. A description of the student's process and many images can be found here.

London tube time travel map

-- David Parry (email)

Response to Time maps

Time maps were done about 25 years ago for the Tokyo subway by a famous Japanese map designer. Here is a version done for the Boston subway:

Boston transit time travel map

-- Edward Tufte

Response to Time maps

There is a travel time map for Toyko designed by Kohei Sugiura featured at Visual Complexity.

Toyko transit time travel

Toyko transit time travel

It uses the same method I recently used for a follow-up to Oskar Karlin's work; a dynamic redrawing of the geographic tube map using travel time instead of distance. If you have Java enabled, try the interactive tube travel time applet. If not, screen grabs for Oxford Circus and Elephant and Castle travel times are available (click to see larger images):

London tube time travel

London tube time travel

There are lots of things I would change/improve, but alas the holidays are drawing to a close so I don't know if I'll have time to implement any of them. Suggestions are still welcome.

-- Tom Carden (email)

Response to London Underground maps

In response of the link http://www.oskarlin.com/2005/11/29/time-travel/ Here is an concept-map of the London Underground, which is inspired by the map Of Oskarlin. The number of NB's at the bottom of the site is clear that there is still a lot of work to do. LINK: http://brainoff.com/mirror/tube_map_travel_times/

-- Remy Jon Ming (email)

Response to London Underground maps

Maxwell J. Roberts has written an update of Ken Garland's wonderful book (Mr Beck's Underground Map, Capital Transport, 1996) entitled Underground Maps After Beck.

Underground Maps After Beck

I've ordered the book, and will report later.

-- Edward Tufte

Response to London Underground maps

Dorian Lynskey of the English newspaper, The Guardian, has created a version of the Underground map with each line representing a different musical genre. Each station is named after a influential artist or band. The real beauty is Lynskey's attempt to use appropriate bands for the intersections between lines/genres.

Download a pdf version here.

-- Spencer Hudson (email)

Response to London Underground maps

Often imitated, never duplicated: the Pyongyang (DPRK) Metro map. The links include some excellent Dear Leader pictures.

Pyongyang metro map

-- ET

Response to London Underground maps

The London Underground map finished second to the Concorde in the Great British Design Quest.

final Concorde flight

The evaluation depends on how various criteria for "best" are weighted. The Concorde: way ahead in its class, completely different, beautiful, amazing, and breaking the military monopoly on supersonic flight--but a service available only to the rich, environmentally obnoxious (per passenger mile, although not in terms of an overall accounting of pollution sources, and trivial compared to military aircraft). The Underground map: effective, influential, served millions and millions of riders, long-lived, no damaging side-effects, beautiful in its unique way--but not the runaway best in its field, some inept versions produced during its long history.

Both produce extraordinary, memorable design experiences. Neither ever worked very well outside their immediate context.

The main design lessons from the map: content-driven design (by someone possessed by the content) rather than designer design alienated from content, the enormous attention to detail, the persistent power of an extraordinary base diagram, the ability of the tenacious Beck to improve the map incrementally over the years, the compromises necessary in mapping a complex system, beautiful typography the idea of producing something beautiful, straight-forward, and widely used.

The Concorde lessons: the experience and intensity of being on different planet, the architectural beauty of a grand machine, the strangeness of the cabin (a long narrow tube of a rocket (2 seats, a narrow aisle, and 2 seats) with an interior surface that becomes very warm to the touch by the end of the flight), the Mach meter hitting 2.0, the miracle of putting London or Paris 3 hours and 48 minutes from New York, the economic issues raised by the whole thing, the funny stories generated by the flights. And walking up and down the aisle shamelessly scanning the passengers for celebrities.

Prompted by the contribution immediately above I went to the voting website a few weeks ago but could not choose between the two, unable to compare Concordes with Underground Maps.

-- Edward Tufte

Mr. Beck's Paris map

Beck's Paris map

London's Transport Museum recently announced that it had acquired a hitherto unknown hand-drawn map by Harry Beck of the Paris Metro system. Apparently Beck had been in contact with the Metro authorities in the 1930s about developing a design, but the the designs were ultimately rejected. There's a detailed article at the Creative Review.

An excerpt:

Like London before Beck, the Paris Metro network had almost exclusively been represented geographically: maps outside stations were (and continue to be) highly detailed topographic plans of the entire city, showing virtually every road, park and waterway with the Metro lines superimposed in all their winding glory.

Though a few examples of privately drawn diagrams have emerged (one Kandinsky-esque rhapsody in abstraction from 1939, so utterly bizarre and impractical that it was never repeated) schematics were not adopted by the city until the last years of the 20th century.

According to Ken Garland's history, Mr Beck's Underground Map, the Metro operator approached Beck to design a diagram. Garland supposes the work was begun in the late 1930s but not finished until after the War. Little survives of his first attempt except a lone copy in Garland's collection.

The Paris Metro is not as easy to simplify as the London Underground. Firstly the lines interweave with each other more (Ligne 7 being the snakiest of these customers); this gives rise to more interchanges (by 1933 around 40 in London, 50 in Paris). Also the system was then mostly hemmed-in by the old Paris walls (a distance equivalent east-west to the width between South Kensington and Canary Wharf and north-south between Camden and Brixton).

With 200-plus stations in easy reach, this is great for passengers, but more challenging for map-makers. One of Beck's greatest innova??tions was to massively expand inner London and condense the outer suburbs. This was just not needed in Paris (at that time) because the entire system was already in the `centre' and very few stations in the suburbs.

What Beck therefore tried for Paris was in some ways more radical than what he'd achieved for London. He sought, in the mass of interlinked lines, some key visual axes to give his diagram order. Seizing the east-west running Ligne 1, Beck made it his prime axis (not as in London's Central Line, running horizontally, but at that neat angle of 45 degrees). He exploited something unnoticed by previous cartographers: that Lines 2 and 6 form a rudimentary circle. Beck transformed them into a rectangle with rounded edges.

From these roots he plotted the other lines as straight as possible with impressive results: the curvaceous Ligne 10 becomes a flat line with its odd one-way loop stylised at extreme left. Kinky Ligne 14 is straightened to a single stroke. Ligne 3 - often seen on other maps with up to 11 direction changes - is reduced to just one nicely rounded alteration.

The overall appearance is clear, balanced and arguably easier to follow. But the key question was: would the French like it? The answer when Beck presented his first version was a resounding "Non!" Beck was not deterred. Indeed, his first London diagram was also rejected but he persisted and eventually its adoption, adoration and appositeness for the Underground was widely applauded. The same fate was not to befall Paris.

Beck went back to his drawing board and produced a second version. It's not known if this was commissioned but, luckily for us, it survives in full colour and was recently revealed as one of the attractions at the refurbished London Transport Museum. It was published for the first time in a book when Paris Metro Style: In Map and Station Design came out in November 2008.

Like any inspired genius, Beck did not waver from his initial concept: here again were his two original axes but Ligne 4 is simplified in its northern half. There are 15 physical direction changes in Ligne 7; Beck whittled these down to two. Ligne 8's 14 real bends went to two, Ligne 11's eight turns cut to none and Beck also, with great wit, added the River Seine.

So why did the Paris Metro (now operated by the RATP) reject Beck's clear simplification of their beloved system? One reason is visible at each station entrance; Parisians use the maps here as a free public service to help them find their way round the city - the ubiquitous geographic wall map is more than just a Metro plan.

The French adore pure cartography - laying claim to many mapping firsts, not least of which was Cassinni's magnificent Carte geometrique de la France - a topographic map of the entire country (begun in the 1670s, though not finished until a century later). The painstakingly precise 1739 Turgot map of Paris (a kind of 3D view from the air, purported to show every visible window) is legendary.

Aside from cartographic history though, Roberts argues there was a fundamental problem with Beck's Ligne 1 axis: "Paris is on a slant. Line 1 especially... is at roughly 25 degrees to horizontal. For a traditional diagrammatic map, which angle should it be snapped to - horizontal or 45 degrees? Whatever angle [is chosen, results in] at least one of the following problems: (1) uneven use of space as lines are compressed together or stretched apart more than in reality; (2) lots of kinks for trajectory correction to avoid (1); or (3) lots of geographical distortion."

Roberts suggests Beck's omissions on both versions (Gare de Lyon missing and Montparnasse drawn wrong on the first, and both Edgar Quinet and Vavin stations missing on the 1951 version), led to suspicion that the concept was untrustworthy. In his fascinating critique of Beck's work (at tubemapcentral.com) Roberts postulates powerfully that though Beck's diagram has aesthetic qualities, it distorted well-known Parisian geography too much for comfort.

Also diminishing a diagram's benefits are the closeness of the stations to each other; one can be plonked down blindfolded in virtually any Paris quarter, walk 500m in any direction and theoreti??cally bump into a Metro entrance. Although in practice there are several holes in the system, such station spacing is much denser than in any other city in the world; a feat the French are justifiably proud of. But pride may be the true reason for the operators' disinclination towards Beck's or anyone else's diagrams.

In staunchly proud Paris, despite the multi-coloured spaghetti with which most contemporary maps portrayed the Metro, there was opposi??tion to following Britain. Double-decker buses for instance were tried out in the 1960s but thought unsuitable for Paris streets partially because they looked too British. In addition, a 1934 pocket Metro map by F Lagoute introduced a style that lasted almost 40 years: though it fell short of standardising angles, its clarity and geographical reflection of the city was sufficient for Parisians not to complain.

Some rather excessive geometry on this Kandinsky-esque Paris Metro pocket map, issued by a private publisher in 1939:

Beck's Paris map

Beck's Paris map

A 1936 version of the `Lagoute' pocket map of the Paris Metro with coloured lines, little topography and some straightening:

Beck's Paris map

Beck's Paris map

Beck's first diagram of the Paris Metro. Ken Garland believes that Beck worked on this in the late 1930s, submitting the plan just after the end of WWII (only to see it rejected). Printed by kind permission Ken Garland/Capital Transport:

Beck's Paris map

Beck's Paris map

In 1951 Beck submitted this revised edition of a map he had worked on for the Paris Metro in the late 1930s. But his map for the French capital was rejected and a diagrammatic approach to the city's system wasn't employed until 1999. Used by kind permission London Transport Museum:

Beck's Paris map

Details from Beck's second attempt at the Paris Metro map. Used by kind permission London Transport Museum:

Beck's Paris map

Beck's Paris map

-- W. W. Oliver (email)

Response to London Underground maps

Compare Beck's Paris Metro map mentioned above with "Le" plan du metro, Le grand classique de la RATP many years later:

Le grand classique de la RATP

-- Edward Tufte

Response to London Underground maps

Interactive travel time map of the Underground.

I have enough trouble choosing how to display data statically.

Now I have a multitude of dynamic and interactive techniques to choose from. And bosses love moving, spinning, swrilling, and twirlling things.

-- Karl Hartkopf (email)

Response to Time maps

Another set of maps depicting time travel (as opposed to distance travel) for Great Britain, including London, here (click images to see larger).

With contours shown at ten-minute intervals:

London time travel map

Moving slightly further out, nearby tube stations (St James's Park, Westminster and Pimlico) and bus routes to the south and east are important. Further south there are islands of accessibility around mainline train stations such as Brixton and Clapham Junction. On a smaller-scale map, the tube and railway lines themselves show up as chains of such islands:

London time travel map

On this final map, of the whole Greater London area and surrounds, the contour lines are at half-hour intervals. At this scale the suburbs of London appear to be arranged along a southwest-northeast axis, a result of good rail links to Surbiton and Twickenham (at the ends of the two red tendrils which stretch away from the center to lower left). Other rail lines, such as those to Bromley and Orpington in the southeast, are also visible, as are islands of short journey time such as Watford (northwest), Hersham and Esher (southwest); these surround individual locations with fast (c. 1 hour) journey times into central London.

London time travel map

-- ET

Response to London Underground maps

Stuttgart's city rail map has a unique feature - the non major cardinal lines are set at 30deg, not 45deg as in Beck's masterpiece.

It has the curious effect of simulating a quasi-3D view.

-- John Ashworth (email)

Response to London Underground maps

For my one-day course last week in New York, Crispian Strachan (CBE, OStJ, QPM, DL, MA, FRSA), former Chief of police in Northumbria, flew in from London and brought this interesting world-wide version of the Underground map, for which he receives a KC (Kindly Contributor) award:

The credit line is to Alan Foale; the map was published in The Times in 2003.

-- Edward Tufte

Response to London Underground maps

Martin Nöllenburg investigated "Automated Drawing of Metro Maps" in his thesis last year. The approach seems to base on the works of Henry C. Beck. Besides the idea itself the document contains many sample graphs of metro maps as well as a reference list that might give hints for further reading. The work is accessible at his University Karlsruhe page.

ET's response: This is a lovely paper, particularly the graphical comparisons of various real and generated Underground-style maps. The more general issue is that much of the value of the London original comes from its intense local care and craft in design, its longevity, and how it became the way in which residents came to see London. There remain serious concerns about the export of the map to other systems. The New York subway knock-off of the London classic was widely ridiculed because of its over-stylized and geographically inaccurate character. Nonetheless, Martin Nöllenburg's paper is very intriguing, particularly in its findings about automated design work. Can other classics be so implemented?

-- Gerd Schlottig (email)

Response to London Underground maps

Thank you for referring to my thesis, Gerd.

As for applying Beck's design principles to other networks, I agree that one needs to do this with care. However, his basic principles like octilinearity, straightening lines, and concentrating on the network's topology are adopted in a huge majority of current metro maps, see Ovenden's book. These diagrams might not become as iconic for the city as in the case of London, but they still make navigation on the network easier for passengers.

What are other classics that you could think of in terms of implementing them as automated designs? Our algorithm is particularly tailored for drawing networks which is also my current research interest.

-- Martin Nollenburg (email)

Massimo Vignelli and the NYC underground

Here's a four minute video of Massimo Vignelli talking about his 1972 map of the New York underground system, an excerpt from the Helvetica interviews. He explains how the information is arranged on the page, acknowledges his debt to the London tube map, suggests his map might have been even better without some of the geographic detail, and laments that his map has been replaced by something far less satisfying.

An excerpt:

It's funny, but I realized the other day the mistake I made. So this is really the clearest kind of map I have ever seen in terms of information for the subway. It's very simple. Every subway line on the map has a color, and in reality they already have one. And every station has a dot, you know. Every stop is a dot--no dot, no stop. It couldn't be easier than that. There is nothing to fragment the legibility of this. Instead, if you look at today's map, it's a total disaster, with fragmentation all over the place. I can show it to you. And this is what we tried to avoid.

Now, 75 years ago, in London, they did the first map with a 90- and 45-degree grid like this one here, and it's been working fine in London for all this time. But New York is a different kind of a city. And now I just realized that maybe, possibly, I made a mistake by indicating, even in a deformed way, the areas--Manhattan, Queens, the Bronx, et cetera. I probably should have done what they've done in London--not have any indication of the geography. It's a completely blank, white background so that there is no suggestion of geography whatsoever.

One of the problems they had in New York is that the people, they couldn't relate the geography with the station, with the lines, and they were confused by that. But it's just because they shouldn't. There were neighborhood maps in the subway stations, so really there's no reason why this map had to be literal--it could be completely abstract. But I think that it would've been even better if I had pushed the envelope even further and not had anything, just the lines and the stops. Maybe that would have been better. Otherwise, it's perfect--I think it's the most beautiful spaghetti work ever done. It's terrific. And it's so clear, it's unbelievable.

Now, the reality is that 50% of humanity is visually oriented, and 50% of humanity is verbally oriented. The visually oriented people have no problem reading any kind of map, and the verbal people, they can never read a map. But the verbal people have one great advantage over the visual people: they can be heard. And that's why they changed this map! They started to complain, these people, opening their mouths, until this beautiful map was substituted with the junky one that you can see now by going into the subway station. It is a map that is so loaded with information, which is so difficult to retrieve, that it makes the whole point of the map useless. If I made a mistake, it was not making the geography abstract--making the water beige and the parks gray instead of green--it was just the fact that we indicated these things when we shouldn't have. We should have just made it blank. It would have been better.

Vignelli NYC map

-- Peter Marquis-Kyle (email)

Redesigning the NYC subway maps

Consider this project by Eddie Jabbour, graphic designer for Kick Design, to replace the confusing NYC subway map (below: originals on left and Kick maps on right -- click for larger versions).

alternate NYC subway map

Encouraged by the interest being shown in the project, Eddie Jabbour contacted the MTA and designers at the agency agreed to meet with him. But when he showed up at the agency's Midtown offices with copies of his work, they were quick to find fault with it. According to Christopher Boylan, the transportation authority's executive director of corporate and community affairs, who recalled the meeting, the main criticism was that Mr. Jabbour's map, like Mr. Vignelli's, was artistic but geographically inaccurate.

There's an interesting piece on Jabbour in the New York Times.

The other day, in his minimalist office, Mr. Jabbour pinned two maps to the wall, then pointed to the different renderings of the Atlantic Avenue terminal in Brooklyn, which he says is the most difficult station to represent because so many subway lines converge there. In Mr. Jabbour's map, the subway lines run parallel to one another, making the map easier to read, if slightly inaccurate. Each line is marked with a circle bearing the route's letter or number, instead of the oblong station markers used on the current map.

There are other differences. Unlike the official map, Mr. Jabbour's map does not have a single line representing all the trains in a "cluster" route, like the 1, 2 and 3 trains in Manhattan. He used the same type font throughout, and words travel left to right, rather than diagonally, as on much of the official map. The lines bend only in 45- and 90-degree angles, to create a gridlike pattern.

In the eyes of Mr. Jabbour, the New York system is too complicated to layer on information like commuter rail and bus routes, as the current map does. He would like to see a map that is singularly devoted to the subway.

That map, created in 1979 by Michael Hertz (who was officially credited with the design) and John Tauranac (who was chairman of the map committee), in turn replaced the highly conceptual map created by the Italian graphic designer Massimo Vignelli. Mr. Vignelli's map was aesthetically pleasing but geographically inaccurate; it depicted Central Park as a square rather than the rectangle it is, for example, and offered riders no sense of where they would be once they left a subway station.

When the current map was adopted, the subway system was decidedly unpopular. The trains were dirty, and riders associated the graffiti-covered cars with crime, often rightly so. Mr. Jabbour sees the current map as tainted by those perceptions. The subways improved dramatically in the 1990s, and in his opinion, the map needs a similarly ambitious face-lift.

Mr. Jabbour's interest in subway maps began when he studied abroad in London in the mid-1970s while earning a degree in design from Syracuse University. Eager to explore the city and even get lost in it, he immediately recognized how easy it was to navigate the city's Underground. The "tube" map, an acknowledged masterpiece in the field, was Mr. Jabbour's main inspiration for his current project.

He planned to continue working on a hybrid of the two previous designs until he was satisfied that his mother, "a woman from the Midwest," could read the map. Then, three years ago, Mr. Jabbour briefly acquired a larger audience. His name had been mentioned in a newspaper article about the design of the city's subway map, and his map was posted on foamer blogs like Live From the Third Rail. Hoping to seize the moment, he contacted the Metropolitan Transportation Authority, and to his surprise, designers from the agency agreed to meet with him.

But when he showed up at the agency's Midtown offices with copies of his work, they were quick to find fault with it. According to Christopher Boylan, the transportation authority's executive director of corporate and community affairs, who recalled the meeting, the main criticism was that Mr. Jabbour's map, like Mr. Vignelli's, was artistic but geographically inaccurate.

"He's a good designer and it's an interesting map," Mr. Boylan said. "The design is important, but the thing we're concerned with is the best directional guidance. We design a map for use, not solely to look good, and we think it looks good."


-- Mark Whybird (email)

Subway systems of the world

Here is an interesting and artistic collection of subway maps using the small multiple format. I like being able to compare them directly!

Subway systems of the world, presented on the same scale

-- Jennifer (email)

New Madrid metro map

In 2007 the Madrid metro was redesigned by Rafa Sanudo with a higher level of line generalisation by not presenting any 45 degrees line (original article in Spanish).


People seem to be upset by this new map. I think that a big change like this will take some time to be accepted. Although I think it will not be a big problem especially for new users, I wonder if it was really necessary.

Underground "maps" are basically diagrams that allow the reader to detect the path from A to B and they are of course a distorted representation of the real world. So distorted that they are not even maps. But, although diagrams underground "maps" have their own language made of symbols, colors, text and lines.

Is it really necessary to take out one "word" (45 degrees lines) from this language in order to simplify things? Does it really simplify things or contribute - at a first glance - to increase confusion in the old users? A new design might also stimulate the use of alternative paths different from the ones suggested by the original map.

-- maurizio gibin (email)

Living with Mr Beck's map

The Beck London Underground map is a classic icon, but it has serious shortcomings as a transport map.

It informs the casual user of the possible routes between two stations, but it doesn't convey any information about the best way to get from A to B. For example, I can travel between Warren Street and Stockwell by either the Northern line or the Victoria line without changing. But the map doesn't tell me that the Victoria line is usually faster, with more frequent trains. Of course, an indicator is that there are more stations on the Northern line route, but the map seems to indicate that the Victoria line follows a less direct path.

A greater problem with the tube map is the disproportionate spacing between stations. The Beck map encourages people to take a train between two stations when other forms of transport would be far more suitable. The most famous example is that of Covent Garden and Leicester Square - not more than two hundred metres apart in reality, but well spaced on the Beck map. In response to your question, obviously not many people do realise that the Beck map distorts distances, otherwise Transport for London wouldn't have such bad congestion problems at Covent Garden station.

So what are the solutions? Transport for London are, quite rightly, keen to protect the underground map as their intellectual property, so few other maps of the underground exist. The organisation has moved with the times to produce a journey planner which does inform the user of travelling times and better ways (e.g. walking) of moving between two stations. However, this is an Internet based service, which isn't that useful for the traveller on the ground. Street maps such as the A to Z convey real distances and tube stations, but are cluttered to look at and don't show the transport links. The only map I've found that shows tubes and streets (and buses) is the Quickmap all-on one. This is a larger map than a TfL tube map, with more information that is more time consuming for the novice user to consult. The problem is that it only covers zones 1-3 - great for the tourist, but not for the Londoner.

-- Thomas Taylor (email)

Response to Living with Mr Beck's map

Realising when there is a more suitable form of transport between two stations does require quite a bit of learning at present. Here's a map by Applied Information Group (part of Legible London) designed to speed up that learning process - it shows journeys that take less time on foot than they do by tube. I'd say it's more powerful and memorable than further annotations to existing maps.

-- Lewis Crouch (email)

Walklines again

Similar to the walklines discussion above -- I've added an element to the design rules of the current tube map to show distance between some stations in metres.

London tube walkline

The BBC recently had an item on a radio item and blog with references to an interesting alternative map by Maxwell Roberts here.

-- Alex Gollner (email)

Beyond Beck

Too much current map design is based slavishly on the London Underground map principles and not enough creative thinking goes into the interpretation of complex modern travel systems. The aim should be to make what is unclear on the ground easier to understand, yet often what is straightforward on the ground is made to look less practical.

Although iconic, the groundbreaking map design solution created by Harry Beck (London Transport) has many inherent problems. Lines that are straight often have to made crooked because of the limitation caused by the 45° maximum angle. Then, to reduce the number of bends (an enemy of clarity) interchange stations suffer by having multiple interconnected nodes that bear no relationship to the??ease or difficulty of a change - and purely for the cartographers convenience. And the dominance of the interchange symbol over the tick, distorting the importance of stations, is questionable. The map also leaves unsolved the issue of different lines sharing the same route, shown inconsistently either touching or separated. And the Beck formula doesn't work for many other systems, for example, the Paris map using his formula is no more helpful in planning a journey than the geographic one.

The map designer should not be forcing a network to follow these abstract rules but should look for clues on the ground to show the individuality of region being depicted. London has a coke-bottle shaped circle line that defines the shape of the west end and city; Europe has the Rhone Valley with its distinctive arrow shape with the point at Frankfurt; Manhattan has its own distinctive tilt; there's a tilted parallelogram that links Liverpool and Manchester together; there's the axis of a major thoroughfare - these are all shapes and angles that help the user identify with a map. Yes, complex areas need opening out, long station names have to be coped with, but geography is always there.

Please see my website ProjectMapping.co.uk to see many innovations. For example, the new UK rail maps feature 22.5° angles to enable all main lines to radiate from London and to reflect the backbone, or shape, of the country. They also develop a new system to solve the problem of repetitive place names where there are multiple stations on multiple lines. The Merseyrail map uses 30° and 60° angles which help to shrink the size of the map and more accurately reflect how the network 'looks'. The Greater Manchester map indicates the city centre by the use of a large capital M and usefully shows the platform layout at the divided Piccadilly station. Lines that are balloon loops or have natural curves are shown as such and are not forced into squares with the corners rounded off.

Merseyrail map

-- Andrew Smithers (email)

London Tube + Social Science

I thought readers of this thread might be interested in the following article in the most recent issue of the journal Social Studies of Science, which looks at the way in which the tube map shapes users' understandings of London geography. The full paper is available online.

Janet Vertesi, "Mind the Gap: The London Underground Map and Users' Representations of Urban Space." Social Studies of Science, Vol. 38, No. 1, 7-33 (2008).
Abstract: This paper explores the effects of iconic, abstract representations of complex objects on our interactions with those objects through an ethnographic study of the use of the London Underground Map to tame and enframe the city of London. Official reports insist that the `Tube Map's' iconic status is due to its exemplary design principles or its utility for journey planning underground. This paper, however, presents results that suggest a different role for the familiar image: one of an essential visual technology that stands as an interface between the city and its user, presenting and structuring the points of access and possibilities for interaction within the urban space. The analysis explores the public understanding of an inscription in the world beyond the laboratory bench, the indexicality of the immutable mobile's visual language, and the relationship between representing and intervening. It further suggests fruitful crossovers between Science Studies, Urban Studies, and Human--Computer Interaction by approaching the individual as a `user' of a city and its graphical interface, applying the technique of cognitive mapping to overlapping virtual and analog spaces, and exploring the social and practical effects of strong and standardized visual languages on further narratives and interactions with scientific, technological, or everyday objects.

-- Adam Hedgecoe (email)

Mapping London

I came across a fascinating book when browsing through my local bookstore here in London: Mapping London: Making Sense of the City, by Simon Foxell. The book includes several classics well-known on this forum, including the Snow cholera map and the Beck underground map (along with examples of what the Underground map looked like prior to Beck), but also displays many other fascinating representations of London. Well worth a read for any budding cartographers.

Mapping London

-- Will Oswald (email)

Subway maps in the bathroom

Christoph Niemann posted this picture to his blog the other day. The story of how he got there would most likely interest a few of the regular readers

-- Tchad (email)

Maps + identity

Note that in the UK, the map is the brand. A rail network map isn't just about how to get from A to B. A rail network map is about the identity of the business or organisation running services on those lines. It's what makes one operator distinct from another, it's what gives different systems their individual identity.

-- Andrew Smithers (email)

Live system-wide train maps

New displays for the NYC subway system to show upcoming trains. Report by William Neuman at the New York Times.

NYC subway display

Beginning in December 2008, officials will install a computer screen at each end of the platform showing a graphic representation of the entire L line and the location of every train on it. Waiting passengers can watch the trains move along the tracks as the data is updated every 15 seconds. That way, passengers can see for themselves if there really is another train "right behind this one."

"There's no end to the possibilities," said Wilson M. Milian, a director of technology projects at the transit agency. He said that the same information could some day be displayed on the Internet, so that riders could quickly check the location of trains before heading to the subway.

-- Edward Tufte

New London map without zones

The BBC reports:

Removing zones from a new map of the London Underground network has made it "confusing", a passenger watchdog said. Transport for London (TfL) distributed the map last weekend without the River Thames and the fare zones marked on it.
I'm not convinced the changes have really "simplified" and "uncluttered" the map. new map

-- rjp (email)

Redesign threatens Mr Beck's map

A story in the Guardian covering the new plans to re-design the 'Beck' Tube map, as a result of expansion of 'oyster cards' to include overland trains, with an accompanying history of the London tube map (with images).

-- Adam Hedgecoe (email)

London Maps on Slate

An installment of a recent series about subway signs and wayfinding in Slate focused on London. In a nutshell, wayfinding signage with carefully-designed maps are being used on the street to supplement the geographical inaccuracies of Mr Beck's map.

Wayfinding signage for city streets is very different from the wayfinding programs you find in transportation environments like Penn Station or Heathrow. In an airport, the signs boss you around: Check in, go to security, proceed to your gate. On city streets, good signs are more like valets: They give you options, telling you about nearby attractions and helping you figure out where you want to go. Transport wayfinding gets you from point A to point B, but good urban wayfinding systems help you develop your own mental map of a place. As a result, they're often more aesthetically complicated than the this-way-to-baggage-claim arrows that confront harried air travelers; they usually include extensive maps or provide an index of streets or amenities.

Legible London marks one of the first efforts to subdue a city so immense.

Why is London willing to try out new signage? Surprisingly, it has a lot to do with the success of an earlier sign--the amazing, iconic London subway map. When people picture London, they often picture Harry Beck's diagram of the tube. But the tube map--though a groundbreaking, elegant, and useful piece of information design--bears only a hazy relationship to the city's real layout. The map, which was first released in 1933, abandoned the geographically accurate style then customary among transit maps. Previous maps of the London system had been hard to read, because they squeezed the stops in central London close together in an effort to make room for the elongating tentacles of suburban lines. Beck deduced that transit maps don't have to be--and perhaps shouldn't be--straightforward geographical representations. And so he straightened out the map's lines and enlarged its scale in the center of London, enabling passengers to understand how the city's most-used stations and lines connected.

Mr Beck's map

Beck's approach has been much emulated since--so much so that it's rare, these days, to find a metropolitan transit map that doesn't take some liberties with geography. But there's one major downside of Beck's map, especially considering that it's a primary means of understanding the city for locals and tourists alike: The tube diagram gives users mistaken impressions about distance--in particular, locations that are in fact quite close together can appear far apart.

geographically accurate map

As a result, people often overestimate the difficulty of walking to nearby locations, taking the familiar tube instead. In recent years, as the tube has become more congested, that's become a problem. The tube in central London often operates at maximum capacity. At rush hour, the Oxford Circus station--located in a busy shopping district--can get so crowded it will close its doors, waiting to clear passengers from the station before it lets new ones in. As a result, the mayor's office and Transport for London (the agency that runs the Underground) have been looking for ways to encourage walking in the city. They see Legible London as a step in that direction.

Legible London

Legible London hopes to reduce tube congestion by showing Londoners how disparate pockets of the city link up. Step one is to produce a geographically accurate map of London that's designed for pedestrian use and includes key landmarks--major stores, architectural oddities, even details like the style of pavement--that will help walkers find their way. Step two is to make that map available to as many people as possible, by posting it on street signs, tube stations, and bus shelters, and by making copies available to locals and tourists in shops and hotels. The project has been developed by AIG Group, a design firm based in London. A prototype was rolled out in London's Mayfair neighborhood in 2007, and in 2009, the city opted to launch pilots in several other areas: Covent Garden and Bloomsbury, South Bank and Bankside, and Richmond and Twickenham.

Legible London is an unusually narrow project in its focus on pedestrians. Most American urban wayfinding systems, like the ones in Philadelphia and Charlotte, N.C., are designed for drivers and pedestrians, linking highway signage to major city streets to parking options to pedestrian pathways. Legible London's brief is simply to get people to walk more. Still, the London project offers an interesting window onto how urban sign systems have evolved over the past few decades.

Legible London detail

One crucial component is the map. Traditionally, maps on these sorts of signs offered a bird's-eye view of the area in question, and were oriented with north at the top. Over the years, though, designers have learned that people tend to do better with maps that detail what the facades of buildings look like--a helpful feature for users who have trouble extrapolating from a 2-D map to the 3-D world around them. Legible London's maps put this principle into practice, featuring axonometric renderings of key landmarks.

Tim Fendley, the AIG designer heading up the Legible London project, says his firm's aim is "to help people insert themselves into the map." AIG tested several orientations during the development period, and users consistently preferred a "heads-up" orientation that puts whatever the user is facing at the top. While there were a few complaints from military men who objected to the deviation from the north-at-the-top standard, Fendley says, "We kind of figured they didn't need the help."

Another hallmark of modern wayfinding design is an evidence-based approach. Fendley's team stood on street corners in Mayfair every Friday morning for a period of months, often holding up enormous prototype maps and observing user response. These tests helped determine, for example, that the signs needed an index. Although architects had suggested doing away with the index for the sake of clean design, Fendley's team found that this frustrated users: "We took [one] out there, and three people came up and looked at it and said 'Where's Swallow Passage?' And I knew it was there; I knew the map by heart. And I was thinking, It's there! It's there! But they were like, 'Oh, rubbish!' And off they went. And I thought, 'Damn. We're going to need an index.' "

Another item that tested poorly was the original "You Are Here" icon. But AIG found that another element on its map--the target circles it used to show which destinations are within a 5-minute walk, and which are within a 15-minute walk--helped snap the user's attention to the center of the map. Initially, the circle was "a lot weaker," Fendley says, but after positive user response, "we turned the volume up."

You are here

(Continue on Slate.)

-- Bill C. (email)

Subway map of the world

I think subway maps, as well as being sources of information, can also be seen as works of art. They have also spawned interesting spin-offs, such as recreations of existing designs, The Great Bear by Simon Patterson is a great example of this, or concepts based on the design style of subway maps.

Here's another version of a map of a "world subway system."

world subway map

-- sonia (email)

Redesigns of Beck's map

Genius re-designs and updates of the design classic can be found here.

London redesign London redesign

-- Brian Flicker (email)

Another London redesign

There is excellent recent paper by Zhan Guo explaining some of the shortcomings of the current London Tube map.

Guo, Zhan 2011. "Mind the Map! The Impact of Transit Maps on Path Choice in Public Transit." Transportation Research Part A: Policy and Practice, Vol. 45, 7, 625-639.

In response, I decided to try to create a version of the map that is geographically accurate but retains the clarity of Beck's original. A live version can be found here.


-- Mark Noad (email)

Errors in NYC subway map

"On the Vaunted City Subway Map, Mistakes and Phantom Blocks" written by Matt Flegenheimer of The New York Times on May 6, 2012, talks about the errors on the New York City subway map.

"I'm more than embarrassed," he said softly.

Many New Yorkers have undoubtedly noticed that the subway map has its geographic faults, from peccadilloes like a wayward street to more obvious inaccuracies like the supersize island of Manhattan.

But Mr. Tauranac's sheepish discovery of the errors has at once rekindled and complicated a long-simmering debate over who deserves credit for the watershed 1979 guide. Michael Hertz, whose firm is credited with designing the initial template for the map, has long chafed at Mr. Tauranac's calling himself the "design chief" on a project that has garnered numerous accolades, including a commendation from the United States Department of Transportation and the National Endowment of the Arts.

"We've had parallel careers," Mr. Hertz said in a telephone interview. "I design subway maps, and he claims to design subway maps."

After years of publicizing his role in print, in lectures and on his Web site, Mr. Tauranac has a new strategy that is his most inventive yet, Mr. Hertz said. By taking the blame for the blunders, he said, Mr. Tauranac implicitly assumed credit for the rest. "That's his shtick," Mr. Hertz said.

But he appeared in no hurry to take responsibility for the mistakes himself.

"He's overseeing the project," Mr. Hertz said, adding that he himself perhaps deserved some blame, but "not as much" as Mr. Tauranac. "I was not an expert on the geography of the city," he said.

Presented with Mr. Tauranac's findings, Adam Lisberg, chief spokesman for the Metropolitan Transportation Authority, said the agency would consider amending future editions of the subway map, provided the changes did not distort its clarity.

But Mr. Lisberg took care to cast the Broadway quirks, particularly those near Lincoln Center, not as errors but as the byproduct of "design decisions" on a map that is not intended to be precise. He said he was not aware of any rider complaints about the West Side geography ever having been recorded.

"This is not a street map," Mr. Lisberg said. "This is a subway map."

On the New York Public Library's blog, Cynthia Chaldekas of the Mid-Manhattan Library wrote an article about the subway's history, "John Tauranac Talks New York City Subway Map History," recounting a talk given by map designer John Tauranac in April 2010 at the NYPL.

1940 Voorhies map

1940 Vorrhies subway map sponsored by Union Dime Savings

Initially subway maps were produced by businesses, whose need for revenue produced maps that showed all the subway lines equally and major points of interest. For example, The Chelsea Hotel on 23rd St. produced a number of handsome maps for the subway rider, starting in the 1930s. Once the lines were merged in 1940, the real work of creating a map that was both useful and aesthetically pleasing began. On Tuesday April 27th, John Tauranac presented The "Unofficial" Subway Maps: A Look at How Individual Mapmakers Have Depicted the Mysterious Workings of the Subway at the Mid-Manhattan Library.

John Tauranac is a mapmaker. He was design chief of the classic 1979 NYC subway map that has formed the basis of the all NYC subway maps since that initial design. John Tauranac is also a connoisseur of maps and that was made clear throughout his talk. One of the first statements he said when he began his discussion was "clarity is king." By this he meant that a map, any map, needs to be useful. If not useful then what is the point of the map. Pragmatism should be desired over aesthetics. Keeping this in mind, John also went on to eloquently state, "Graphic design is a solution to a problem, with an overlay of aesthetics."

What followed was a tour of historical New York City subway maps, whereby the discussion veered into nuance and overall design of each map. John talked at length on the difference between a schematic depiction versus a geographic depiction of the subway system. A schematic map does not necessarily follow the geography of an area. If the information desired to be on the map distorts the area being drawn, so be it. A geographic map is the opposite; the mapmaker strives to keep the integrity of the geography in tact. He then showed examples of maps where attempts had been made to reconcile the two.

John also waxed and waned on more subtle aspects of subway map beauty. The choice of some mapmakers to use unnatural colors for natural occurrences, such as making the bodies of water throughout the city yellow or making parks gray instead of green, John felt was a poor design choice. John pointed out how something seemingly small can lead to great effect, like blackening an outer edge line, to create clarity, at the same time enhancing the overall beauty of the map by establishing a visual contrast. The thin darkened line brightens the color within the line, making it easier to see and also making it more aesthetically pleasing.

The mapping of the New York City Subway System is complicated. Our transit system has express lines, night and weekend service, as well as a complex network of transfer points. We are the only transit system in the world that has all these things. Choices about what information to include and what not to include on a map must be considered. Also, the way in which certain features will be expressed must also weigh into the design discussion. All these considerations lead to a level complexity that makes the mapmakers task very challenging.

There have been designs of the NYC subway map where, in one instance, the look of the map was more important than the overall utility. World class designer Massimo Vignelli was commissioned to design a map for the city. The result was the 1972 subway map. It is a striking document, austere, even restrained. It is a schematic map where lines of color represent the different train lines. For my eye it is simply too minimal and disconnected from the subterranean world of trains under the city, as well as the hustle and bustle that floats along the city's surface. Vignelli's map description is a disembodied relative of the complex cityscape New Yorkers experience both above and below ground.

A remedy to Vignelli's map was the 1979 subway map, whose skeletal remains have formed the basis of all the city's subway maps since. It is a geographical map that is at once understood and appreciated by riders everyday.

Tauranac map

comparison of three types of maps: the 1979 map , combination map, 1972 Vinelli map

-- Edward Tufte

Another London redesign

I regard the official London maps as design classics with a number of problems. The heritage is obvious but new lines are somewhat haphazardly added. For example the Jubilee line from Westminster to Canary wharf is rather contorted.

My own map, with a live version available here:


-- Peter Saxton (email)

Mind the Maps

"Mind the Maps," an excellent blog post by Ollie O'Brien of University College, London's CASA Institute on recent takes (post-2010) on the London Underground map, featuring a collection of various versions of the tube map marking the 150th year of the Underground.

Mind the Maps

Mind the Maps

Mind the Maps

For a comprehensive overview of alternative tube maps try this listing from The Londonist. The latest incarnation being one made from Lego.

-- Jerry Clough (email)

Updated Moscow map

The most recent version of the Moscow metro map:

Moscow metro

London 2050

I did a version of the London Underground map (keeping to the "Beck Conventions") for the London Reconnections site, at London 2050: A Cartographical Interlude.

The map takes the current one and adds in everything from the London Mayor's "The London Infrastructure Plan 2050" (Crossrail 2, Crossrail 3, new Overground lines, Met Watford changes, even Northern Line to Clapham Junction!). It has been seen on the BBC London News, as well as featured by Time Out and The Londonist.

London 2050

-- Brian Butterworth (email)

Threads relevant to maps: