When trying to make streets better for cycling, there are a range of options. At one extreme you could close the street to through motor-traffic, at the other you could make separate paths alongside. With lots of road-sharing options in-between.

The classic approach is to look at the current speed and volume of traffic, read off the type of provision required, and then see if it fits. If it does, fine; if it doesn’t, try to reduce the speed or volume of traffic so you can use something that does fit.

The minimalist approach is to turn this on its head: given the space that’s easily available, how much do we have to reduce speed or volume of traffic to make it reasonably comfortable? And can we reduce speed or volume even more, to make it more comfortable?

In some ways these are just versions of the same approach, but they can have quite substantial differences. If you’ve got space and money, perhaps building a new town or suburb, the classic approach is easier, and has been widely implemented in new developments in the Netherlands. Traffic is either fast or slow, and there are either tracks or nothing. The classic approach also works in tight-knit city centres – just exclude the traffic and the job’s done.

But in older suburbs, especially ones that were laid out in the late nineteenth century, space on main roads is often at a premium. And in our larger cities, even though the streets are wider (courtesy of the Great Fire of 1666), the press of traffic can be quite severe. The classic approach would often call for segregation, but the space won’t readily be made available, especially at junctions.

The minimalist approach is to start with the simplest intervention that gives clear space for cycling – the painted cycle lane, and look at how this can be made comfortable for a good number of cyclists. That’s initially achieved by keeping the speed of traffic down to 30mph or preferably less. It’s quite hard to get speed down if traffic has more than one lane for any distance, so this only really works for upto about 20,000 cars/day. But that’s plenty enough traffic for most main roads in most towns.

There are various things that can be done to slow main road traffic. It helps if the road is enclosed by buildings, but putting in cycle lanes, narrowing the traffic lanes to 3m or less, removing the centre line, putting in raised crossings of side roads and frequent pedestrian crossings all help. Refuges can work, but should really be part of a continuous median, otherwise cyclists will tend to get squeezed. Zebra crossings are better.

It usually takes some time – years – to implement all these measures and get traffic to slow down. But when the traffic has been slowed, cycle lanes are fairly comfortable, and acceptable to a wide range of adult and teenage cyclists.

If you’ve got the space, bus lanes (or wide cycle lanes) can be effective, but it’s harder to get the traffic to slow down. Generally, bus lanes are a good idea if there is peak-time congestion – buses and drivers don’t come cheap, and we need a good service to provide for longer journeys. But wide cycle lanes seem to cause as many problems as they solve.

The minimalist approach also applies at junctions: try to give an obvious, easy route for cyclists going straight-ahead, but don’t try to separate cyclists. Instead get the traffic to slow down, and work around the cyclists. In low-speed environments, drivers are perfectly capable of sharing safely, as long as they can understand what the cyclist is doing. The trick is to get everyone going slowly, so bikes only get in the way briefly.

Similarly at roundabouts, there are Swiss designs that slow the traffic down to cycling speed, so that the road can be comfortably shared. This takes up much less space than the Dutch segregated roundabout designs. It won’t be comfortable for family cycling, at least not for quite some years, but it is very effective at making roundabouts tolerable for a lot of adult and teenage cyclists. However, these roundabouts don’t work for bigger junctions (more than about 25,000 cars/day). Bigger junctions have to use traffic lights.

Why do I call this “Dutch” minimalism? Well partly because it’s about cycling, and “Dutch” is almost a synonym for cycling. Partly also because some of the most famous minimalists were Dutch – painters like Mondriaan (an early exponent of the creative uses of white paint). But also because the minimalist approach is very much based on the principles for bicycle planning that have been developed in the Netherlands. The facilities are very much aiming to be cohesive, direct, attractive, safe and comfortable, in classic Dutch style. It’s just that we are also actively trying to restrict ourselves to the simpler forms of facility, to leave more space for pedestrians, and to take space from the road as much as we can reasonably manage. The bicycle is a very efficient form of transport; we should aim to be efficient in its use of space, as well. If you are trying to fit in cycling from scratch in an existing congested city, it’s probably the only way.

What does Dutch minimalism ultimately look like? The end result will probably be 20mph main roads with cycle lanes. Pavements will be wide, and parking limited. Cyclists will have the freedom to use all roads without being constantly threatened by passing traffic. Pedestrians will be able to cross roads pretty much at will. That’s what I mean by a Transport Paradise.

(and Earth, and Venus, and Mercury)

In an important 2010 report which deserves a wider audience, the UK Department for Transport (DfT) identified four different types of adult cyclist, based on how they cycle, and in particular how they feel about sharing the road with other traffic.

The four behaviours that DfT identified were Assertiveness, Guardedness, Avoidance and Opportunism. But those descriptions are quite hard to remember, so I’m going to translate it into something more memorable: Cyclists are from Mars, Earth, Venus and Mercury.

Martian cyclists (assertiveness) are those that have done the training. They look behind regularly, they pull out early to pass parked cars and to turn right. They know how to tackle roundabouts.

Earthian cyclists (guardedness) are comfortable enough about other traffic that they’ll cycle on busy roads, as long as they can keep to the left, out of the way. They’re only happy turning right if the traffic is going really slowly. They hate roundabouts.

Venusian cyclists (avoidance) don’t venture onto main roads at all. They only cycle if there’s a quiet route, preferably off-road. They will cycle on the pavement of main roads, and use crossings, but are likely to switch to an alternative mode, if they can’t find a reasonable route.

Mercurian cyclists (opportunism) are typically young men in a hurry. They don’t really care about the rules of the road, and just go fast by whatever route is available – inside, outside, on the pavement, whatever.

As always with attempts at segmentation, these four types actually lie on a continuum, and some people will adopt different behaviours in different circumstances. But thinking in terms of these four types helps to ensure we cover the range of cyclists. The research was qualitative – it was based on small group discussions in a range of different towns. No attempt was made to quantify the numbers of current or potential cyclists that are in each group. In many UK towns, with hostile road environments, the Mercurian and Martian cyclists are likely to predominate.

However, one thing this does highlight is the problem with talking about cycling: different types of cyclist might as well be on different planets, for all the understanding they’ll get from other types of cyclist.

To give a practical example, where I live, there are two roads you can take into town: Hythe Bridge Street and Park End Street. One of the standing debates in my household is: which is the better way to cycle into town.

To get to Hythe Bridge Street you go straight-ahead at the big junction outside the station, and then join the heavy traffic along Hythe Bridge Street as it narrows. There are no cycle lanes and a couple of pinchpoints. At the end of Hythe Bridge Street, the traffic is forced left round a tight bend, there’s a Pelican crossing, and there’s a gap for cyclists to go straight, and into town.

To get to Park End Street, you have to pull out into the traffic approaching the big junction to fork right into the bus area. You then have to get past the bus stops and through another set of lights. There’s not much traffic on Park End Street, and there are cycle lanes, but there’s often loading, and some buses go a bit fast.

My partner prefers Hythe Bridge Street (and I think she’s mad). I prefer Park End Street (and she thinks I’m mad). The reason is that I cycle like a Martian – pulling out and watching a few buses seems the easiest. Whereas my partner cycles like an Earthian – keeping straight ahead and turning right in slow-moving traffic is easiest.

Almost all cycle campaigners (and cycling officers) are Martians. Most of them can understand that other cyclists are different. But they generally fail to understand how they are different, and how prevalent the different types might be.

Our experience in Oxford is that Earthians predominate. Oxford has about 20% of commuters cycling to work (28% in some wards), and 30% of adults using a bike at least once a week. Most of them cycle on main roads, but there’s very little take-up of adult cycle training. The reason that cycling has become normal for Oxford’s middle classes is that the roads have been adapted to make “Earthian” cycling a reasonable thing to do. We no longer have any gyratories, and our only major roundabout has been tamed (mostly). You can go straight-ahead at almost all main junctions without difficulty, and rarely get forced into making an uncomfortable right turn.

Oxford’s approach to infrastructure has always been quite pragmatic. A lot of painted cycle lanes and bus lanes were marked out in the eighties, and they worked well-enough (for Earthians). Separately, a number of quiet routes were developed, focused on schools. These also worked well – for children, and for some adults (a mixture of Venusians, and others who found them to be useful short-cuts). So these routes have been extended and linked up to make a quiet network, mostly by providing crossings of main roads. It has developed into a dual network.

The initial idea of the dual network was that the two parts were equally important. We thought that new cyclists would be Venusians. But it has become clear over the years that the main road network is considerably more important for adults. This is partly due to the circuitousness of the quiet network, but it is also because Earthians appear to outnumber Venusians, by some margin.

It may be that there are a large group of potential Venusian cyclists that aren’t being catered for. But for whatever reason, they aren’t speaking up. Perhaps they can see for themselves that the space isn’t really there. Perhaps they’re settling for the circuitous quiet routes. Perhaps they’ve converted themselves into Earthians. Or perhaps they walk, or catch the bus, or drive.

Instead, the local pressure is to fill in the gaps in the main road network – to deal with the remaining parking on the main roads, if possible, and provide more cycle lanes. There is also support for 20mph limits on side roads, and a range of techniques for slowing traffic on main roads.

For other cities wanting to improve conditions for cyclists, and struggling to find the space and funds to do so, it would be sensible to think carefully about the type of cyclist you are trying to provide for. By understanding and catering for the specific needs of Earthian cyclists, you are likely to be able to make cycling viable for a substantial portion of the adult population. Trying to provide comprehensively for Venusians in congested main road corridors is much more difficult.

So what are the key things that make the roads tolerable to Earthians? The main thing is that the default route at all junctions needs to be straight-ahead. After all, for most journeys, you mostly go straight. This is actually quite simple, and just-about permitted by the regulations: paint the cycle lane across the junction. If it is clear where the straight-ahead cyclists are going to go, then motorists are perfectly capable of finding a way round them. That can be made to work at most urban traffic lights (the Germans have a range of designs). Roundabouts and gyratories are harder, but the principle is the same: always give priority to bikes going straight-ahead over left-turning traffic. Dutch-style roundabouts meet this criterion. Gyratories generally don’t, but it may be possible to adapt them so that they do (or even better, get rid of them).

Earthians are less particular about what happens between junctions. The various degrees of segregation don’t matter very much to them, as long as they have a clear path, and traffic is keeping to 30mph (or preferably a bit slower). The main improvement is to regulate parking, and paint continuous cycle lanes. Gaps in the cycle lanes for zig-zags and bus stops aren’t ideal, but they’re not critical.

If cities want to make real progress towards getting a good proportion of people onto bikes, they should focus on providing for Earthian cyclists – there are plenty enough of them – and focus on the key things that make Earthian cycling tolerable: cycle lanes and straight-ahead priority at junctions.

In contrast to Oxford, most UK cities are quite hostile to everyday cycling. There are no cycle lanes, so the traffic has too much space and goes too fast. There are roundabouts, where you would worry about cars pulling out on you, and gyratories, where would you get cut-up. For example, here’s a map of Worcester:

Hardly any solid red lines (cycle lanes). Lots of black (roundabouts) and blue (gyratories). It’s clear that Worcester is the sort of place where young men might be brave enough to cycle, and some others will venture out on the quiet routes, but there will be little “everyday” cycling; the main roads don’t even begin to be cycle-friendly.

Map data (c) OpenStreetMap contributors, Cartography (c) Richard Mann