This piece was written for a course in the philosophy of technology.

In this essay, I will discuss Walter Benjamin’s claims regarding the political influence of architecture in “The Work of Art in the Age of Mechanical Reproduction”. For Benjamin, architecture is a prototypical exemplar of a property I will call “the power to influence under distraction” (hereafter, PID), in that architecture is both (1) received by its users in a state of distraction, and yet (2) able to influence its users’ beliefs and behaviors via habit formation. Benjamin thinks that the emerging medium of film also possesses this type of power, and can therefore be influential on a political scale given its to access to mass audiences. In this essay, I will argue that this kind of power poses a threat to liberal values like individual autonomy and democracy. I will then confront two normative questions. First, is it morally permissible to exercise this kind of power (PID) to affect users’ beliefs and behaviors? Second, if the answer is “yes”, what are the right ways of using or governing this kind of power? On the first question, I will argue that the answer is “yes” in the sense that exercising PID is simply unavoidable in many cases. On the second question, I will outline what I call “technocratic” and “democratic” approaches, arguing that the latter poses a better solution to the challenges raised by PID.

In “Work of Art”, Walter Benjamin describes the consequences of technological changes that enable art—and in particular, film—to be reproduced and consumed on a mass scale. These changes fundamentally shift the value and purpose of art from something rooted in ritual-use “cult value” accessible to the few, to something intended for mass consumption (i.e. having “exhibition value”). While some aspects of art are lost in this shift (i.e. the “decay of aura”), the upside is the revolutionary political potential of this new art. Though the masses generally consume this art (film) in a state of distraction, Benjamin thinks this art can nonetheless influence the masses through the formation of habit, since “even the distracted person can form habits”. In short, film has the the power to influence under distraction.

For Benjamin, architecture is a prototypical example of the power to influence under distraction; thus, his discussion of architecture illustrates and supports the claim that this sort of power is possible for film as well. In Benjamin’s terms, architecture is received both by use (tactilely) and by perception (optically). Tactile reception is particularly important because it comes about “by way of habit” rather than explicit attention. In short, Benjamin is highlighting that the material design of a space can predictably shape the beliefs and behavior of those who inhabit it, whether they realize it or not. The design of religious spaces is a good example of this phenomenon. Even if the congregant does not consciously consider the vaulted ceilings, stained glass, intricate carvings and so on, they may still be influenced to feel religious awe or piety when inhabiting the space. Through repeated exposure, these feelings may become habitual, ultimately reinforcing a set of religious beliefs that persist beyond the church walls.

Accepting Benjamin’s descriptive claims about architecture’s power to influence under distraction, I will now discuss two key ethical questions raised by Benjamin’s analysis. First, is it ethical for designers to use this kind of power to affect users’1 beliefs and behaviors? Second, if the answer is “yes”, what are the ethical ways of using or managing this power? Concerning the first question, there is a good argument for “no”. By assumption, PID influences beliefs & political outcomes not through free, reasoned deliberation & persuasion, but through a form of subconscious psychological manipulation, influencing users without their realizing it. Framed this way, PID arguably violates important liberal values like individual autonomy2 and democracy3. However, while we might prefer to prohibit PID on this basis, I argue that we generally do not have this option: PID is unavoidable in many architectural cases, as “neutral” design is impossible. For an illustrative example, consider the design of grocery store aisles. We know from behavioral economics4 that distracted consumers are more likely to buy products that are placed at eye level on the shelf. Thus, in choosing which product to place at eye level (e.g. the broccoli or the potato chips), the supermarket designer (or “choice architect”5) has the power to influence consumer habits around diet and health. Further, whether we like it or not, the designer must put something on the shelf in this place6; she cannot avoid exerting some influential power over users of the space. There is no “neutral” option7. Similar arguments can be made in many architectural/design settings.

We are then left with our second question; if designers cannot avoid influencing their distracted users, what is the right way of wielding or governing this power? I will consider two approaches to answering this question: (1) the technocratic approach, and (2) the democratic approach. The technocratic approach holds that decisions about PID ought to be left to benevolent designers, guided by some vision of what is best for users or society. Benjamin implicitly endorses a version of this perspective in his enthusiasm that PID in film/architecture be used to advance his Marxist vision of the good. Likewise, behavioral economists Thaler & Sunstein advance a version of this approach, holding that designers ought to use “nudges” (their term for PID) to “make choosers [users] better off, as judged by themselves8. This principle attempts to address the concerns about individual autonomy outlined above by aiming to positively influence users on their own terms.

The technocratic approach may be persuasive if we believe that it is possible for designers to indeed know what is best for the world or users. However, the approach is little help if our moral / political concern about PID is rooted in its threat to individual autonomy and democracy. It is still manipulation even if I am being manipulated into behaviors or beliefs (that some designer would say) that I would say are in my own best interest. The democratic approach does better on this front. This approach attempts to place decisions about PID (and other aspects of technological design) back into the hands of users who are affected by them. Participatory design is one version of this, seeking to explicitly bring perspectives of affected users into the design process. Likewise, policy prescriptions that seek to bring major digital technology architectures under public democratic control (e.g. as advanced by Ethan Zuckerman and Safiya Noble) are also in this spirit, albeit on a larger scale. In my analysis, PID poses moral concerns due to its threat to individual autonomy and democracy. Democratic approaches to governing architecture and design are thus a compelling solution in that they directly reinsert free and reasoned deliberation back into the design process, giving users an explicit say over the spaces that influence them.


  1. I use the terms “designer” and “user” in attempt to be generic; in practice, the designer could for example be an architect, urban planner, or UX designer etc.; the user could be a homeowner, congregant, citizen, or web browser etc. 

  2. As it is defined in the Stanford Encyclopedia of Philosophy: “Individual autonomy is an idea that is generally understood to refer to the capacity to be one’s own person, to live one’s life according to reasons and motives that are taken as one’s own and not the product of manipulative or distorting external forces, to be in this way independent” 

  3. As defined in the Stanford Encyclopedia of Philosophy: “Democracy, properly understood, is the context in which individuals freely engage in a process of reasoned discussion and deliberation on an equal footing.” 

  4. Behavioral economics is essentially the science of influence under distraction. 

  5. This is the term used in behavioral economics 

  6. Or in an equivalent position of priority — i.e. it is not compelling to argue that supermarkets should simply get rid of eye-level shelves. This merely passes the buck to some other situation which will surely have its own biases (e.g. waist-level shelves will still be preferred over foot-level shelves). Even the argument that supermarkets should wholly overhaul their design so as to render all products equally likely to be chosen embed a normative preference 

  7. This discussion is based on a similar argument about school cafeteria design in the “Introduction” of Thaler & Sunstein’s Nudge 

  8. Nudge pg 5