1937 Paris: The Peril of Elegant Variation

Yesterday evening I launched an eBay auction for a curious map of the Paris Metro from January 1937. It reminded me of one of the articles in Henry Watson Fowler's classic work, Modern English Usage (1926), entitled "Elegant variation".

In Fowler's terminology, elegant variation is the practice of deliberately using synonyms (or near-synonyms) to avoid using the same word several times.

Fowler begins his article in the following uncompromising way:

"It is the second-rate writers, those intent rather on expressing themselves prettily than on conveying their meaning clearly, and still more those whose notions of style are based on a few misleading rules of thumb, that are chiefly open to the allurements of elegant variation."

The Wikipedia entry suggests that the word "elegant" carried a pejorative overtone in 1920s, which it since lost. But I rather think that Fowler was being sarcastic, as he often is in the Modern English Usage.

What does this have to do with maps? Well, when a map is printed in a single colour of ink, the cartographer must find some way to differentiate the lines. Some cartographers just don't bother, and leave black-and-white maps a jumbled mess. Others create variations in the linestyles and, in doing so, some succumb to the temptation of elegant variation.

One way to create differentiation is to use different styles of breaking the solid line. But after exhausting the four basic forms -- the solid line, the dashed line, the dotted line, the dot-dashed line -- the designer must invent some novel linestyles. Some, driven by the graphical equivalent of what Fowler identified as 'elegant variation', create a baroque variety of lines. Which is what one finds in this 1937 map of the Paris Métro by P. Babey.





A design discipline that was always observed in official black-and-white maps of the London Underground (and indeed in some Parisian maps) is to keep the linestyle within 'tramlines'. That is to say, to keep the dots, dashes, ticks, and whatever else -- all within the longitudinal boundaries of what would be the normal width of a line. We can see this in the following sample from a 1975 Underground map.





P. Babey, on the other hand, does not observe this discipline, but allows his linestyle to explode sideways in the form of ticks and crosses.

But what is so wrong with doing what Babey did? The answer comes from the psychology of cognition. Essentially, the mind reads detail as information and tries to interpret it. By littering his lines with ticks and gaps that compromise the integrity of the line as a single visual entity, Babey creates visual cues that the mind automatically tries to inerpret before realising that they are just noise. Hence the map is slower and harder to read, and leaves a bad aesthetic feeling.

By contrast, in Garbutt's London map of the 1960s and 70s (shown above), the various linetsyles are encapsulated and glide effortlessly into the mind's automatic cognitive workings.

Finally, let us take a look at the grand master of metro maps, Harry Beck. The whole world and his dog is nowadays familiar with Beck's colour maps, because of Ken Garland's book, Mr Beck's Underground Map (1994), but the black-and-white maps are often neglected. In the following sample, we see that Beck successfully differentiates the lines just by leveraging the careful use of angles and curves. Here is a war-time sample of Beck's black-and-white map:





Thus the lines look different without being different. They criss-cross in complex ways but never with ambiguity. This, I submit, is the cartographic parallel of the sound of one hand clapping.

© 2008, Peter B Lloyd

1913 London: Transgressing the UndergrounD map

After closely looking at the 1914 map of the London Underground yesterday, I couldn't resist selecting its predecessor, the beautiful 1913 map, for my next eBay auction.

This pre-war pinnacle in railway cartography represented a big leap forward in the evolution of the London Underground map. When Albert Stanley joined the London Underground Group of companies in 1907, he began the process of creating and selling the 'UndergrounD' brand. New, improved maps of the underground railway network were an important element in that process.

So what was wrong with the old style of railway maps? Essentially, they were tolerable for the widely spread out mainline rail services, but hard to read in dense, highly connected areas such as central London. You can see some dense knots of mainline lines in the base map of this 1913 edition, such as here:





You can trace which line is which if you follow the geometry of the curving lines carefully with your eye. But trying to decipher this monochromatic style in central London just doesn't work. With all lines looking the same, you can't tell which line is which when the lines attain a certain level of density and interconnection.

The idea of differentiating the lines in at least some way can be seen in the District Railway's maps from about early 1890s, when the first deep tube lines opened. Following the development of tunnelling technology from sub-surface cut-and-cover to deep tubes, London saw the opening of The City & South London Railway on 18th December 1890 (now part of the Northern Line), and the Waterloo & City Tube on 8th August 1898. Here is a detail from a 1898 District Railway map:





What we see in this detail are: first, the mainline railways (thick blue lines); next, the subsurface Underground lines - both Metropolitan and District (thick red lines); and then they have added as a thin blue line with round as opposed to rectangular station symbols the deep Tubes.

By the way, in this particular map, we can see the deep tube of the City & South London Railway (CSLR) crossing under the Thames to its terminus in the King William Street Station (unnamed on the map), alongside the Monument Station on the District Railway. And, running a few hundred yards to the east of it is the new cut of the CSLR, which sweeps past Monument Station and terminates at Bank Station. This is shown as another thin blue line, of course, but broken to indicate it was under construction. The old cut was abandoned at the close of business on 24th February 1900, and King Wiliam Street Station became one of London's ghost stations. (For many decades, there was talk of reopening it, which idea was finally scotched when the extension of the Jubilee Line was built right through it, bisecting the ghost tunnel.)

Anyway, my point in this digression is to show that, even before Stanley's cartographic big bang of 1908, London's map-makers had acquired the concept of using different graphical styles to differentiate underground railways.

Now, what happened in 1908 was that each Underground line was assigned a colour and the non-Underground lines either disappeared from the map or were shown in subdued hues. In the 1913 map, they are pallid beige lines in the background. Here's the 1913 colour key:





and here's a detail showing how the coloured lines looked in central London:





The importance of this conceptual leap cannot be under-estimated, as it was one of the strands that were brought together in forming the network concept of 'London' -- when the city itself was exploding into a monstrous sprawling conurbation that swallowed up surrounding villages. The city of London was placed amid a dense network of railway lines that spread continuously out from the centre. Where in that continuum was a Londoner to draw a boundary to London? It was Albert Stanley who drew the boundary for the rail commuters by creating and promoting the branding of the London 'UndergrounD' and semiotically exporting this brand through maps, signage, and carriage livery.

On the ground, the boundary was not so well-defined. Overground railway companies ran inter-city and regional commuter services deep into the heart of London, and yet they were not included in the "UndergrounD" brand and were dropped from, or subdued in, the maps. Some decades later, Ken Livingstone forced London Transport to include the North London Link on the Tube Map, but that was an exceptional and bitterly resented aberration. The long-lasting purity of the Underground map -- the principle that it should show only Underground lines -- was like an article of faith, and it had its genesis in Stanley's cartographic revolution of 1908.

Nevertheless, material reality does intrude, and we see it doing so in the following detail from the 1913 map, showing the extension of the Bakerloo Line.





The electrified deep-tunnel Bakerloo line was being extended north-westwards via the important mainline interchange at Paddington toward the nascent suburban commuter lands to the north-west of London. What we see is a broken line from Edgware Road to Queens Park, and then -- and this is the intriguing bit -- a skinny brown line continues north-eastwards from Queens Park and off the map towards Watford.

What on earth is this? And how is the humble passenger supposed to decode this part of the map? The answer has to do with the venerable London and North Western Railway (LNWR), which had many decades earlier opened the world's first passenger railway service -- between Liverpool and Manchester (in its previous incarnation as the Liverpool & Manchester Railway). (By the way, check out Simon Garfield's excellent book "The Last Journey of William Huskisson" for an insight into the momentous opening of that first passenger railway.)

The LNWR as such was incorporated in 1846 and expanded to serve the great industrial cities of England -- Birmingham, Leeds, Liverpool, Manchester ... and of course London, where its terminus was at Euston Station. But the LNWR's lines swept through the upcoming commuter lands of north-west London. To exploit this potential, the LNWR initiated, in 1909, a massive project to build additional tracks alongside the existing lines (to allow slow commuter services in parallel with inter-city expresses), to electrify its suburban lines, to build more stations within the commuter belt, and ... to create a link with the Bakerloo Line at Queens Park, so that Bakerloo train services could travel on LNWR tracks as far north as Watford -- thereby giving passengers from LNWR stations seamless connections into the underground system of central London.

The Bakerloo extension shown in this map was a long time coming into reality.

After Edgware Road opened as a temporary terminus of the Bakerloo Line in 1907, there was a hiatus before the line was extended to Paddington, owing in part to a dispute over money. (The single-stop segment from Edgware to Paddington was held up as the Bakerloo Railway Company and the Great Western Railway -- the main users of the Paddington mainline terminus -- tried to bluff each other into stumping up the money to build the extension of the Tube.) But work started in October 1912, and that initial segment opened in December 1913. (So, this map was issued while that first segment was actively being built.) The further extension of the Bakerloo from Paddington northwards to Queens Park was then delayed by the war-time shortage of labour. It eventually opened on 11th February 1915. A few months later, on 10th May 1915, Bakerloo trains started running on LNWR's tracks -- initially only as far as Willesden Junction, but reaching Watford in 1917.

The Bakerloo thus became the first "UndergrounD" Tube train service to be physically connected with a mainline network. This presented a challenge to the conceptual isolation of the "UndergrounD" brand. (By the way, although these were the first deep level Tube trains to transgress into the mainline network, transgressive sub-surface trains had done this some decades earlier, when the Metropolitan Railway ran underground trains to the town of Windsor, via Ealing Broadway, between 1883 and 1885.)

What we see in this map is the cartographer trying to deal with this transgression of the conceptual boundary of the Underground -- trying to stop the 'brand leakage' as we might say. His unsatisfactory response to this problem was the skinny brown line to the north-west of Queens Park, denoting the 'new cut' line of the LNWR, which by June 1913 physically existed (hence the map-maker could not drawn it as a broken line, as he had for the extension from Edgware Road to Queens Park). But he nevertheless needed to indicate somehow that this line was not open to Bakerloo passengers, as it was not yet electrified and anyway the Bakerloo trains could not reach it yet. His solution, which is not explained in the key, was to draw a skinny brown line. It looks like the Bakerloo line, and yet is visibly different.

Needless to say, this was not a very good design decision. Subsequent cartographers had a more robust approach to policing the boundaries of the Underground: active "UndergrounD" lines were shown in solid colours; and "UndergrounD" lines that were proposed, under construction, or temporarily close, were shown as broken lines -- irrespective of whose tracks they ran on. Overground services, when they occasionally transgressed onto the "UndergrounD" map, were shown as colourless lines.

1914 London: The Poverty of Xenocartographism

When I loaded up the Edwardian London maps to eBay a fortnight ago, I skipped the weird "Met" map with the blue text on the white background.

.

(Admittedly 1914 is outside the reign of King Edward but it's within most people's notion of the Edwardian epoch.)

I decided to load it up for auction this weekend. So far, it's not attracted a huge amount of attention (48 hits, no bids), which suggests that the collecting community haven't twigged to how interesting this item is.

Let's see why this map is a key specimen for any student of the history of metro maps ...

The Met, the District, and the "UndergrounD" Group

For the half a century preceding this map, the metro system in London was dominated by the Metropolitan Railway Company and its rivalrous offspring, the Metropolitan District Railway Company (whose name is usually shortened to the "District Railway" for obvious reasons). In August 1900, the maverick US financier Charles Tyson Yerkes landed in the traffic-clogged metropolis, whose medieval street-plan was not conducive to getting anywhere quickly. Legendary in his own lifetime for his brazenly ambitious transport schemes in the USA, Yerkes later served as the model for the character Frank Cowperwood in Theodore Dreiser's novels. Yerkes bought up the District Railway and in five years came to be the owner of four new deep tube railway lines -- the Bakerloo Tube, the Hampstead Tube, the Piccadilly Tube and the outsize Great North Railway. He died in 1905, but the juggernaut he created -- the London Underground Group of companies -- was unstoppable and still is, in its current incarnation of Transport for London. Back in 1914, however, there were still three underground railway companies working outside Yerkes' behemoth Underground Group, by far the biggest of the three outsiders being the extensive Metropolitan Railway ("the Met") -- which had actually been the first to build underground railways anywhere in the world.

The Metropolitan and the District Railway had a bitter long-standing rivalry, which was later to show up in maps issued by the Met that systematically omitted any services not run by the Met themselves. Their arch-rivals in the District Railway, however, had a cartographic policy that was rather less blinkered by narrow commercial concerns. They had, since the 1870s, produced maps that showed both their own services and the Met's. This approach was broadened further in the Edwardian period. In 1907, Charles Yerkes' Group recruited the visionary genius of public transport, Albert Stanley (later Lord Ashfield). Stanley's big insight was that the value of any given railway line to its passengers is increased combinatorially by its connections to other railway lines. He instigated the branding of the "UndergrounD" as a distinct and unified system even though large swathes of underground railway network belonged to commercial competitors. From 1908, London's travelling public started carrying around experimental new pocket maps, with colour-coded representations of the underground lines. What we now take for granted as the pocket Tube Map had its clumsy and tentative birth in 1908.

The Met had never issued its own maps. From the 1870s to 1908, the travelling public had relied on the District Railway's maps, which of course also carried promotional text and timetables for the District Railway -- but not for the Met or any other railway. From 1908, with the emergence of a complex network of subsurface lines and deep tubes, the value of the "UndergrounD" colour-coded map increased. At about 1914, it seems that the Publicity Manager's Office at the Met started thinking outside the box, and initiated the huge marketing drive for 'Metro-Land' -- the paradise of garden suburbs to the north-west of London that was served by the expanding Met lines. And, in the same year, the Publicity Manager decided tentatively to dip his toe into the cartographic waters.

The 1913 "UndergrounD" map

Let us step back a year. In June 1913, the Underground Group had produced the most successful map of the London Underground seen so far. It took on board the inventiveness of graphic devices (such as the white-bridge connector at interchange complexes) that had gone into the maps since the "UndergrounD" system maps started appearing in 1908, and worked them into finely crafted lines.

.

With a print run of 150,000, this was a massive symbolic statement of the Undeground's branding of itself as the rapid transit system of London. On the back are listed the times of first and last trains of all the lines in the London Group (which, in 1913, had swollen to include the Central London Railway and the City & South London Railway) ... but does not include the times for the Metropolitan Railway or the other two unaligned companies. More insidiously, the list of places to visit -- "Sights of London", "Places of Amusement", "Parks and Gardens", "For Sports", "Country Outings" -- omit any sites that would require the Met to get to. The semiotic message was that the Met was of use only as a connecting service to get onto one of the Underground Group's lines. This was the subtext of the copy line on the front cover, "Stations Everywere": that is, everywhere that is worth visiting in London is served by the "UndergrounD" system.


.

The 1914 Met map

My guess is that the new thinkers at the Met's Publicity Manager's Office must have been disturbed by this porrayal of the Met as just another line in a multi-coloured web of lines, and one that is subservient to the "UndergrounD" Group's services. For, in the following year we see the first pocket map issued under the Met's rubric, consisting of what is essentially the Underground Group's map of 1913, reduced to only two colours for the underground lines: red for the Met, blue for everyone else. The hubris of this map is breathtaking. It is an utter denial of Stanley's ethos for public transport, that the network is more than the sum of its parts.

This 1914 map, issued under the Met's name but being a dumbed-down de-colourised version of the "UndergrounD" map, was an evolutionary dead-end. There was never a second edition. The Met did not issue any more pocket maps until 1921, when it needed something to help it promote Metro-Land, and then it produced a series of about eleven ugly maps until about 1932, which drew no inspiration from the "UndergrounD" maps.


.