A theory of censorship
A ban must be understood through a framework that takes into account the reason for the ban, the stakeholders involved and its impact on the stakeholders.
Censorship and bans are as old as the civilised world. Although the two terms are different, the meaning they convey is often similar. The instances of censorship or bans occur if an idea is perceived to be harmful to a person or a group’s sensibility and this is not a contemporary phenomenon. History is rife with such examples, from the Ancient Greece to Modern Togo. Authorities, governments and seats of power have always dissuaded citizens from expressing uncomfortable ideas and will continue to do so.
Banning is prohibition of materials or ideas that might be objectionable to certain people. Censorship, on the other hand, is selective removal of the said materials and ideas. This essay attempts to arrive at a framework to analyse the two. (For simplicity sake the word ban/censorship is used interchangeably through the essay.)
Politicians and policy makers often encounter vehement requests for bans. Often, the requests are self-serving and do not merit government intervention, but there are several cases where policymakers are forced to take a decision with the view of protecting the larger social interest — banning of paedophilic content on the internet is an example of this. Censorship is viewed as regressive, and almost always rightly so, but there are cases where states or institutions impose bans to uphold certain values that are important to them irrespective of its implications to the general populace. Moreover, the fact that a ban restricts the citizens’ fundamental right has not stopped the authorities, even in democratic republics, from imposing them.
A ban must be understood through a framework that takes into account the reason for the ban, the stakeholders involved and its impact on the stakeholders. Any ban imposed has to be viewed from the perspective of two primary variables. First, the effectiveness of the ban. This relates to the effectiveness with which subject of the ban is actually prohibited. Second, the discoverability of the ban. This indicates how much of banners’ actions are discoverable. There are instances where the ‘banner’ would want the actions to be discoverable widely, and the inverse is also true.
Similarly, every ban comes with its own reasoning (flawed or otherwise) and stakeholders. The motives can range from punitive action on an opposing group to merely suppressing information. Therefore, if the primary variables of a ban are effectiveness and discoverability, the output variables can be of three types. The first is support. Bans pronounced with the intention of expressing support to key allies or interest groups.
The second is deterrence. Bans imposed to deter individuals or group of individuals from repeating things that led to the ban.
The third is retribution. The classic tit-for-tat strategy, where a ban is imposed is merely as an act of retribution.
When a ban is imposed – with any percentage of completeness and effectiveness – the intended output is often a combination of one of the three output variables mentioned above. Under the assumption of working in a society where social contracts are adhered to, the stakeholders involved are the authorities imposing the ban, the interest groups seeking the ban, people involved with the subject/object of the ban and ordinary citizens.
As a consequence of these assumptions, any form of ban/censorship can be viewed through quadrants in a Cartesian plane where the axes, X and Y, represent effectiveness and discoverability, as shown in the figure. Based on the quadrants, bans can be broadly classified into four.
The first type is a public ban. This is a ban that is effective and discoverable. The topic of the ban leaves popular consciousness, and the actions of the authority enforcing the ban are clearly visible. Examples of such a ban include removal of a book by ordering the pulping of it.
The second type is a signalling ban. Signalling bans are ineffective and discoverable i.e., the topic of the ban fails to leave popular consciousness, but the actions of the authority is visible. Typically this kind of ban is promulgated to signal disassociation from a particular topic. With the advent of internet, most bans acquire a signalling characteristic except those that necessitate physical presence.
The third type of ban is a futile ban. These are neither effective nor discoverable. This case is presented for theoretical completeness for it is hard to observe examples that fall under this category and if there are any, as the name suggests, they are futile.
The fourth type of ban is a perfect ban. The primary variables will naturally lead to us defining an ‘ideal’ case. An ideal ban or a perfect ban is one that is completely effective (100 percent) and has zero discoverability i.e. the subject is completely removed and citizens are not aware of the existence of the subject. Idealised conditions allow us to arrive at simplified models devoid of ‘noise’.
Idealisation, though unreal, helps us understand extreme cases and provides a lighthouse to analyse intermediate cases. Like Max Weber argues, an ideal type is an analytical construct that serves the investigator as a measuring rod to ascertain similarities as well as deviations in concrete cases. It must also be noted that ideal type examples can acquire tautological traits. The example for a perfect ban is one such case–“several classified documents whose existence is not known fall under this category”– by stating example the ban ceases to be ideal but there are several documents whose contents and existence is available only to a handful of people and in a sense constitute a perfect ban.
Depending on the outcomes expected (support, deterrence and retribution) and desired impact the person pronouncing the ban, has the following options
It must be reiterated that this kind of analysis provides a picture of the edge cases – (in)complete effectiveness, discoverability etc. – which are far removed from real world, but as discussed earlier, these idealised cases help us in simplification of analysis and understand the intermediate cases better.Assuming a binary model — that is input or output can be yes (1) or no (0) only – the authority promulgating a ban has 2^5 = 32 ideal cases. Because each variable can be either yes or no (1 or 0) and there are 5 such variables. An interesting exercise for future study can be to arrive at the complete matrix with relevant real world examples.
Craig Nelson’s book Thomas Paine: Enlightenment, Revolution, and the Birth of Modern Nations, narrates this interesting incident from history: “When the Viennese government compiled a Catalogue of Forbidden Books in 1765, so many Austrians used it as a reading guide that the Hapsburg censors were forced to include the Catalogue itself as a forbidden book”. In the internet age there is possibility of spawning several such catalogues for every ban tends to become a signalling ban with diminishing marginal utility. The framework provided helps policymakers analyse scenarios where they are expected to or want to censor.
Varun Ramachandra is a policy analyst at the Takshashila Institution.