Opinion: as a sixth cyclist dies in London in two weeks, Sam Jacob argues that roads should be designed "in a way that incorporates intelligence as well as brute engineering" and asks:" "Who is the city for?"

Roads are super complex landscapes. All those speed bumps, arrows, double yellows, zig zags, kerbs, red men, green men, zebra, pelican, puffin and pegasus crossings are both the surface over which we travel and codes that modify and instruct how we travel. They are simultaneously map and territory, abstract markings on the surface of the city that become the city.

They may often be imperfect and in a constant state of revision but roads are the fundamental product of civilisation. Roads even, it might be argued, civilise us, as infrastructure that connects both places and us one to another into a collective society. Roads are where all our multi-faceted desires and demands (literally) intersect, where they are negotiated in real time, turn by turn. Of course, sometimes these negotiations tragically fail.

After six cyclist deaths in London over the last 13 days, there is understandably a sense of panic on the streets - certainly amongst the cycling community. A friend late back to the office after cycling back from a meeting found her phone flashing with a series of panicked messages checking that she hadn’t become another cyclist casualty. Over the top, for sure, but also indicative of the heightened tensions surrounding the capital's carriageways - a tension revealed in the aggression that often characterises our behaviour on the road too.

After this spate of accidents there are, understandably, calls to do something. I don’t doubt either that there is a real desire on behalf of the authorities to do something too, but what that thing might be is much harder to identify.

The problem is first practical. How can the variety of road users - pedestrians, bikes, cars, trucks - co-exist in a safe and civilised way? But it's also a philosophical and political issue: who the is city for?

Though we might think of them as natural, streets and roads are as much concepts as things. In Britain, pre-Roman roads formed as tracks across a landscape between settlements. More desire lines than infrastructure, we could think of these as routes worn into the surface of the earth by habit, formed by the subjective behaviour of travellers. Roads here are produced by the act of travelling itself. As such they are less defined, their edges blurred. Roman roads on the other hand brought a very different conception - an abstract, as-the-crow-flies, objective inscription of intent. The Roman road organised and controlled how we crossed the landscape.

After the fall of the empire, Roman roads fell back into disrepair, their engineered surfaces collapsing back into the soil. By 1555 the country's roads were so poor that an act of parliament was passed requiring parishes to maintain their roads. Men of the parish were required to work for six days each year to maintain and repair the roads, but unpaid and under-resourced, little improved. As the industrial revolution gathered pace, parliament passed what was known as the Turnpike Act, which allowed the creation of private toll roads. Given the potential for profit, investment in the provision and maintenance of roads accelerated the quality and network of roads.

Roads are now (on the whole) public in the sense that they are owned by public bodies, but the ownership, management and maintenance of roads is shared between local authorities, Transport for London, and the Highways Agency. Perhaps this fragmented ownership is reflected in the confusions and conflicts on the roads.

The point of this historical diversion is to underline the point that roads are not static entities - they evolve in relation to the society that creates them - and that roads then alter the society they ostensibly serve. Turnpikes, for example, benefited those who invested (often members of parliament themselves), were often unpopular (the defensive pikes added to as protection and penalty of execution for anyone destroying a toll booth). The increased costs associated with transporting livestock into the city pushed up the price of meat and inevitably affected the urban poor most directly.

Back to the present moment; back to the issue of cycling in London. The problem of traffic and highways is usually thought of as an engineering project rather than a function of holistic urban design. Which is, I'd argue, itself part of the problem.

The engineering-first approach to cycle infrastructure produces a range of solutions:

First there’s the half-hearted standard bike paths. These might demarcate a route that seems sensible for a cyclist to follow, but be forewarned: you’re as likely to find one leading you into a dead-end, straight into a lamppost, or into a pile of bin bags dumped in its track. More often than not, they seem like elaborate devices dreamt up by Wile E. Coyote. They would be funny if they weren’t so laughable.

Next up there's the Barclays Cycle Superhighway. These are semi-infrastructural licks of paint whose gestural wide strips of blue attempt to form zones within the road surface dedicated to cyclists.

They represent a particularly abstract form of planning as though the fluorescent highlighter, beloved of planning officers as they mark out zones and routes on the black and white expanses of OS plans, had reached down out of the sky and simply started sketching its intention directly onto the surface of the city. This is infrastructure as intention rather than reality. Cycle Superhighways might assume the appearance of infrastructural authority but the reality is that they are often little more than a trompe-l'œil. They have an indistinct status: a name that suggests real hard wired traffic infrastructure but a reality that is little more than wayfinding. In spite of their good intentions, you can’t paint the city you want into existence.

At the most extreme end, Transport For London is testing out Dutch-style roundabouts. Frankly though, in most of London there's no way that the crooked, winding streets could be tamed into anything bearing more logic. London, born out of a singular lack of planning, seems to have a resistance to any logical planning set within its grain. Which is, of course, part of its charm: a city that’s evolved out of the lives lived within it rather than been envisioned by the mind of a Haussmann. That’s not to say that hard-wired segregated solutions aren’t either possible or desirable, but that the possibilities of their implementation are limited.

The problem of our roads seems a problem that we can't build our way out of. That is to say, it's not a problem of things but of space. Or rather, of things in space in motion.

Nowhere is this more visible than watching a giant hinged articulated lorry swinging itself expansively out at a junction, only to switch back round as though it were a particularly languorous, overweight uncle attempting a drunken hokey cokey. Even a large van has trouble making it around the corner without riding up over a curb.

The view from my saddle is this: it's the incompatible co-existence of the biggest and the smallest, the heaviest and the lightest, the most armour-plated and the softest flesh between lorry and cyclist that’s the issue. There’s nothing you can engineer to mitigate this situation.

The cyclist/lorry conflict is the most extreme of examples. In extremis, it exemplifies a crisis in the nature of our city's streets. Currently we imagine roads as a universal resource, a system that makes little differentiation between its users, or the nature of their use. Those interests, I'd argue, are not all public. Commercial deliveries might put food on the shelves of our supermarkets, materials on construction sites that we may one day work or live in, or deliver tourists to historic landmarks, but the form of these deliveries are mainly unrestricted and in volumes that suit logistics managers.

Surely in an age of smart city rhetoric, big data, and the impending possibilities that GPS and digital mapping bring to transportation (as Dan Hill discussed in his last Dezeen column) it's time to refigure the design problem of the London street. That is to say, to conceive of transport design in a way that incorporates intelligence as well as brute engineering. The kinds of data available on even consumer-end apps like City Mapper show how joining up available datasets provides new ways of configuring movement through the city.

What if, for example, deliveries were timed not to coincide with rush hours. What if large loads were the exception and goods were distributed from out-of-town depots in smaller electric vehicles. Indeed, wouldn’t Heathrow, vacated after the construction of the Boris Island Airport, provide a suitable interchange of this sort?

We focus tremendous amounts of time, money and expertise on the design of so many other forms of transport, but roads seem to be far less of a design question. Perhaps they seem too ordinary, lacking the glamour associated with cars or airports. Yet we should recognise the necessity of roads as a design project - and the huge significance that roads represent.

Just as ancient Rome could conceive of the kinds of networks that supported its imperial ambitions, we need to find ways to imagine the kind of streets that our public, accessible city of the twenty-first century demands.

The design of London's roads is not just about the tragic deaths of cyclists. It's about how we make sure our city becomes public and how roads continue to force us to negotiate a contemporary urban civility.

Sam Jacob is a director of architecture practice FAT, professor of architecture at University of Illinois Chicago and director of Night School at the Architectural Association School of Architecture, as well as editing www.strangeharvest.com.