An open prohibition at least announces itself; when a piece of design produces the same result without declaring any ban, the address for objection grows blurred. Power does not always say “forbidden” — sometimes we simply notice that we cannot proceed on the screen.
A bridge is a thing that joins two shores. An app is a thing that meets a need quickly: it shows the way, takes an order, brings people together. Seen from a distance, neither seems to have anything to do with politics. A bridge is concrete, an app is code; neither speaks, votes or declares a side. Yet the height of a bridge can determine which vehicles may pass, and the rules of an app can determine who becomes visible. Something built to connect can, at the same time, divide.
The bridge example often cited in Langdon Winner’s discussions of technology is striking for exactly this reason. It is claimed that some bridges on the Long Island parkways in New York were designed too low for buses to pass, thereby making it harder for the poor and Black populations who used public transport to reach certain areas. Even though the historical details of this example have since been debated, the question it raised still stands: when a decision to discriminate is buried in concrete, can it become an invisible and permanent instrument of government? To ask the same question of the applications in our hands today, the bridge’s story need not be flawless.
Obstructing without prohibiting
An open prohibition at least announces itself. We can object to a sign reading “you may not enter here,” look for the owner of the ban and open it to debate. But when a physical or digital arrangement produces the same result without declaring any prohibition, the address for objection grows blurred. The bus cannot pass under the bridge; the system merely displays dimensions. A person cannot complete an application screen; the software simply warns of “invalid information.” The one left outside encounters not a person deciding but a mechanism working as though it were a law of nature.
Today this matters especially in digital public services. If a transaction can be completed only with an email address, a smartphone, a bank account or a new identity-verification method, the person without them has not been explicitly banned. They nevertheless remain outside the service. The institution may say it offers “the same system to everyone”; yet an equally offered technical condition produces different results across unequal lives. This time the height of the bridge is hidden in the required fields of an entry form.
The app’s invisible city plan
When a navigation app shows us the shortest route it does not merely describe the existing city; it can also change the flow of traffic. It can turn a neighbourhood into a shortcut, direct constant traffic into a small street, and push some businesses forward while burying others at the bottom of the map. Whatever order a delivery app displays restaurants in, it distributes the flow of customers in that order too. A business whose rating falls stays physically in the same place yet has been moved to the back street of the digital city.
We do not perceive this power as we would a municipal council decision, because what stands before us is a map, a list and a recommendation. But if the conditions for reaching a place, for a business staying afloat or for a worker finding a job are increasingly arranged by applications, then a small private form of government has been established there. The rules are usually not public; changes arrive not through elections but through updates. The user, too, is treated not as a citizen but as a customer who has accepted the terms.
Whose word is a rating, whose decision is a ranking?
The app economy presents scores, stars and rankings as objective measures. Looking at five stars, we feel we are seeing society’s collective judgement. Yet who may rate, which review is deleted, whether new or old comments weigh more, and which factors besides the score enter the ranking are all design decisions. A business appearing at the top because it advertises is not the same as a business appearing at the top because it is widely liked; but the interface can easily blur the difference.
For workers the consequences can be heavier still. When a courier or a driver fears that a single low rating will affect their income, the balance of power in their relationship with the customer shifts. Rating ceases to measure service quality and can become an instrument of control that produces continuous compliance. The manager who decides is invisible; instead of dismissal, access to the account closes; instead of a disciplinary penalty, the number of orders falls. Like the bus that cannot pass under the bridge, the worker is caught at a threshold whose reason they do not fully know.
The thing called technical necessity
When these arrangements are criticised we often meet the answer “that’s how the algorithm works” or “the system doesn’t allow it.” The sentence is useful because it moves responsibility from human beings to machines. Yet it is again human beings who build the algorithm, choose its objective, set the measure of success and decide not to change it. The system may have certain limitations; but those limitations did not descend from the sky. Treating speed as more important than justice, profit than accessibility, or security than privacy is a political rather than a technical preference.
We need not assume every application is ill-intentioned. Many exclusionary outcomes may be the product of hurried decisions, narrow user testing, or existing inequalities transferred into code without question. But unintended discrimination is still discrimination. The engineer of a bridge may not have thought through every consequence; that does not change the fact of whom the bridge lets through and whom it stops.
Rethinking the right of passage
Once we accept that a bridge or an app is political, a new field of responsibility opens before us. Instead of closing design decisions off as “a matter for experts,” we can discuss the say of those affected by them. Auditing algorithms that perform public functions, preserving non-digital routes to essential services, enabling workers to contest automated decisions and making platform rules intelligible are the beginnings of that discussion.
The point is not to reject every technology with suspicion but to see the passage a technology closes as well as the road it opens. While joining two shores, whose vehicle is the bridge built to fit? While making life easier, whose life does the app leave out of account? The answers may change, but the question must stay fixed. For power does not always appear before us saying “this is forbidden.” Sometimes we simply notice that we cannot proceed on the screen — or that our head has struck the bridge.