A circular subway map for NYC
Transport mapping expert Max Roberts recently unveiled a radical new take on the New York subway map. Based on concentric circles and spokes, it's the ninth 'Circle' design he has proposed for an underground transport system. He spoke to CR about how he created it and what makes a good map...
Roberts is a psychology lecturer at the University of Essex with a keen interest in schematic mapping. His site Tube Map Central is dedicated to the design history of transport maps, and over the years Roberts has designed several city maps himself – for London, Madrid, Berlin, Moscow, Paris, Tokyo, Washington DC, Chicago, and now New York. (His page on his Circle maps is here).
As anyone familiar with the NYC subway will know, Roberts' new approach is very different from the MTA's official map (both shown, below).
Above: Max Roberts' new 'Circle' map for the New York subway. Larger version, here
Above: Detail of the official MTA New York subway map, available here
Unlike London's Tube map, which has remained relatively true to the design principles conceived by Harry Beck in 1931 (and published in 1933), New York's subway map has moved around from version to version. Graphic designer Massimo Vignelli famously created a striking (and divisive) map for the subway in 1972, but it was taken out of use just seven years later.
Generating some debate of his own online, we spoke to Roberts about his latest creation.
CR: How difficult was it to make the NYC circular map compared to the other eight you've tackled in the series? Is there an area within the map itself which proved particularly tricky?
MR: New York was definitely up there with the hardest networks, Paris, Tokyo, London, but I think that Paris gave me the most trouble.
I've mapped the New York subway before, so I've got a good familiarity with the trouble-spots. Lower Manhattan is the obvious problem, with all those lines and long-name stations in close proximity. I hate station names interrupting lines, it defeats all the clarity of the map. The arrow for World Trade Centre station hurt a little bit, but the payback was a lovely straight 1-Line all the way from South Ferry to Central Park.
The hard parts of New York are the complex service patterns (e.g. B/D in the Bronx and J/Z skip stops) and how to show 'express' versus 'local' trains (there are three ways to show express/local; I opted for the compact way).
A schematic map with distorted geography has to 'double-name' the stations with the parallel street names – so that those five 23rd Streets in Manhattan can be distinguished, for example (see bottom of detail shown below). That really adds to the pressure in tight spaces.
The south-west of Brooklyn is harder than it looks because it defines the width and height of the map – you have to do that next, and get it as compact as possible. There is an annoying sequence of stations on the 2/3 line from Bergen Street to Franklin Avenue, just like Turnham Green to Ravenscourt Park on the London map. Then there are the awful tangles around Hoyt-Shermerhorn and DeKalb Avenue.
CR: Unlike London, New York has a coastal map (like Chicago). Does this present different problems such as where to 'centre' the circle? How did you identify the suitable point of origin for the NYC map?
MR: New York was never on my list of networks to do. I've only tackled those with a clear circle or loop previously. With these, you can tell roughly where the centre of gravity is likely to be, although Berlin and Paris both needed two attempts. So, I never expected New York to function well with this style at all, and there were fewer clues for a good central point.
I started thinking about the Canal Street station complex in Lower Manhattan as the centre, but that would mean that lines at points south would be spreading out again. I ended up fitting Cortlandt Street between the red and yellow lines, because no matter where the central point was, those lines could be no closer, then moving up to Chambers Street.
By the time I got up to Canal Street, all difficult points of Manhattan had been fitted, and the minimum slant of the lines necessary gave me a fix for the radiation point. To my delight/relief, the top of the map, at the Bronx had not been splayed out too much, and was actually nice and spacious. I realised I could get a lovely pure arc from Bay Ridge right over the Manhattan Bridge and that was it. The map was finished, I just had to fill in the details!
I know the system well, I've mapped it several times. I knew exactly which parts would make my life difficult, and got them right first, then the rest of the map crystallised out from the centre. I think its one of the maps I have enjoyed designing the most.
CR: Which previous NYC subway maps do you think have done the best job to date, and why? What's your take on Vignelli's approach from 1972, which lasted for just seven years on the system?
MR: I've always liked the 1958 diagram by George Salomon. It is tricky to work out what all the lines are doing but very elegant. Salomon was directly inspired by Beck, and he intended it to use more than three colours, but the TA wasn't ready for that innovation. It took them another 20 years to colour by trunkline. I've reworked Salomon's design as per his original intentions. The map might still be with us today if his original proposals had been implemented.
Vignelli's 1972 map is much maligned and misunderstood. It's an incredibly disciplined design, and apart from the famous 'errors' is actually not as geographically distorting as people claim.
[A comparison between the maps designed by Salomon, Vignelli and Michael Hertz Associates (1979) is featured in a New Yorker story, here, taken from Paul Shaw's book, Helvetica and the New York City Subway System].
His original has a much more powerful shape than the new reworked Vogue/weekender version, but the chaotic line colour-scheme of the original made it offputting to use. I've also reworked the original to recolour it. It is far more effective. Vignelli likes it too!
CR: What are your thoughts on the merging of the schematic map approach with one that references 'above ground' geography, such as Mark Noad's design for the London map? Are the two even compatible?
MR: London's geography is ugly, and Noad's map compounds the problem by having poor coherence. The broad aims for an effective diagram are simplicity (of individual line trajectories), coherence (overall shape), harmony (aesthetically pleasing elements), balance (no sudden changes in density), and topographicity (I made that word up: not distorting geography too far beyond people's comfort zones).
Coherence is the most subtle, it took me 14 years with experimenting with maps to begin to understand it, and after experimenting with circles maps, I realise I still don't, but it is the coherence that makes the difference between a good and a great design.
When I am in Paris, I usually use the excellent geographical metro map, it is a really nice piece of cartography, especially compared with their schematic map. It has smooth curves crossing Paris. Turning them into a load of zigzags, or an incoherent collection of multiple angles, doesn't seem to offer anything. Beck's map wasn't a design classic because it used straight lines, it was a classic because it had so few corners.
I actually find hybrids a ghastly monstrosity, neither fish nor fowl. A good geographical map shows where the network is, a good schematic shows how the components relate logically together. A hybrid doesn't fulfill either function effectively. My own opinion is that every transport undertaking should produce two maps, an excellent geographical map, and an excellent schematic, so that people can choose what they see as necessary for their own preferences and purposes. Paris and Berlin let you do that, but not NYC and London.
CR: You have written that "if a product is attractive and powerful for some people, so that they enjoy looking at it, that is half the battle won for the information designer". Can the 'attractiveness' of a map really enable people to consider it in more detail, or to give it more time?
MR: If someone says "I hate that, it's really confusing" they are hardly going to stick around, not when they have a handy app on the phone that removes the need to look at the horrible thing. The British love maps, and those that love them the most say "I love maps, I could look at them for hours". They must be learning something while they are looking.
Ugly maps won't get used, nice maps might be used. Pity about individual differences in aesthetic preference! You can't please all people all the time.
CR: Finally, with your London map, you incorporated the tube symbol at the centre. How naturally did this emerge? Or did you have the idea to include it and work out a way to make it happen?
MR: You need something circular in the middle, so that the centre can relate to the Overground, but trying to turn the Circle Line into a pure circle is pretty disastrous. In fact, the way the routes work together mean that once you reach the roundel shape, everything is very very easy to fit together.
The east side of the Circle Line is pinched on the current map, the north west corner is pinched in reality, and smoothing it out gives the geograpical anomaly that Zhan Guo objects to [Guo is assistant professor of urban planning and transportation policy at NYU Wagner]. Making the whole thing symmetrical gives a nice straight run for the Piccadilly Line.
Central London caused me no difficulties with that shape. So, I didn't set out to do it, but a sequence of design decisions and my hunches pushed me to the point where I realised that this was a good shape to use functionally and conceptually.
There is a map commentator on the internet who called it a cheap designer's trick. I think that if you are an expert and you are going to say something like about someone else's map, you need to have a go yourself first, just to check!
For more of Roberts' designs see his website, Tube Map Central.
Want to learn a new skill? Hone your craft? Or just switch off that Mac and do something a little less boring instead for a while? Then our August issue is for you with details on workshops, short courses and a host of ideas to reinvigorate the creative mind. You can buy the August issue of Creative Review direct from us here. Better yet, subscribe to make sure that you never miss out on a copy - you'll save money too. Details here.
Having had a go at a tube-style map a while back I have great respect for Beck et al. The circle idea solves some issues well but some ugly corners there too.
I wanted to see more of Roberts work but the link to his website didn't work for me (have CR audience overloaded the Essex Uni server?).
You can see the 'reality' problems caused by these circular maps if you understand the geography of the cities themselves. As an example, I know that St.Pauls underground station is close to the river. You can walk across the Millenium bridge to get to Tate Modern on the other side of the Thames. But to look at the circular map you have no way of knowing that St. Pauls is anywhere near Tate Modern or the Thames.
Whether that's important...?
I really like the London Underground one. I knowthat geographically it's quite a way off how London actually is above ground, but then so is the existing one. I don't think Londoners will ever accept their tube map being meddled with too much, but it might be quite nice to see some of these alternate versions dotted around various stations. just to mix it up a little.
|Sagmeister & Walsh and the robbery that never was|
|How Fredrik Bond achieved an 'epic strut' for Moneysupermarket.com|
|See the other side to homelessness|
|Keeping the Anzac memory alive|
|The making of the CR Annual cover|