If sound, the previous argument shows that at least in some instances, censorship is incompatible with the democratic commitment to free and fair elections. Can we generalize this conclusion to all cases of censorship? I think not. Again a simple example can serve to illustrate the point. Imagine a group of fifty philosophers and lawyers attending a conference on freedom of expression decide to go out for dinner. Let us assume that it is unanimously agreed that these are the only two options worth considering, Chinese or Mexican food, and that in order to expedite matters—imagine the verbosity of a group of philosophers and lawyers—no one should make any speeches or otherwise try to persuade the others. Without such a rule there is a significant concern that the debate might go on well into the next day. If anyone speaks on the matter, it is decided that they will not be allowed to go for dinner with the group, so a silent vote is to decide the issue. Is this a democratic decision? To say that there must be free speech on the dinner issue in order to make the vote democratic seems entirely implausible. Indeed, I do not see any non-question-begging way to insist that such a vote is not democratic. It should be emphasized here that we are asking whether the absence of censorship is necessary for a democratic vote, not whether it is good. So, perhaps a better decision might be reached by allowing free discussion; however, our question is not how to arrive at the best democratic decision, but whether free expression is necessary for a democratic decision to be made. And unlike the don’t-contradict-Rex law, the silence rule does not give preferential treatment to the preferences or point-of-view of some. So, while free expression might well be a good accretion to the democratic decision process, it is not necessary. If this is correct, then some forms of censorship are consistent with a commitment to democracy.11
So we have two quite different examples: one where censorship usurps democratic government, and one where censorship is consistent with democratic government. How do these two cases differ? I suggest the difference is related to the distinction between unilateral and multilateral censorship. Unilateral censorship is where only one side of an issue is censored; multilateral censorship is where all sides of an issue are censored.12 Thus, the don’t-contradict-Rex law is a form of unilateral censorship; whereas the conference delegates chose a multilateral form of censorship. The position I am recommending is that unilateral censorship is inconsistent with democracy.
To see this we can make a similar point that was raised in connection with the don’t-contradict-Rex law. Suppose instead of the multilateral censorship the group had instituted with the silence rule, they had voted to allow only the pro-Mexican food lobby an opportunity to speak before the vote. This would be unilateral censorship, and it seems to me, incompatible with the claim that a democratic decision was reached. Imagine the pro-Mexican food contingent painting mental images of the ambrosia served at the Mexican eatery, while denouncing the poor service and the unsanitary conditions of the kitchen at the Chinese restaurant. (Also, they cannot resist mentioning the rumour that an inordinate number of neighbours of the Chinese food eatery have reported missing cats). With no chance to respond, it would hardly be surprising that a majority might vote for the Mexican option. Even if the Chinese option had won out, it would not show that the decision was reached in a democratic manner any more than if a majority of the castaways had sacrificed their lives and voted against Rex when there was a death penalty for doing so. In either case, what this would show is that sometimes it might be that those who hold the coercive power in elections that are not free and fair may fail to achieve their goal.
While the censoring of positive views on Chinese food shares with the don’t-contradict-Rex law the feature of being a form of unilateral censorship, notice that with the law involving Rex, the content to be censored was tied to Rex’s preference, whereas in the latter case the content to be censored was defined in terms of a subject matter (namely, Chinese food). The don’t-contradict-Rex law may sin to a greater extent against a commitment to a democratic vote precisely because all political speech is tied to an individual’s preferences in this way, but this does not undermine the general point that unilateral censorship undermines the democratic obligation to a free and fair vote by privileging some positions and excluding others from public debate.
To underscore this point, let us think in more detail about the possible effects of unilateral censorship on the outcome of the vote. Suppose that psychologists have developed a means to scan a person’s brain and it was determined that prior to any discussion, the food vote would have been 30 for Chinese and 20 for Mexican.13 However, the arguments by the Mexican food camp were effective in swinging ten voters, meaning that the Chinese vote is reduced to 20 and the Mexican food vote increased to 30. Allowing speeches in favour of the Chinese option is one way to think about reintroducing equality in the decision making process, but again we are imaging that this option is not allowed. So, how could we best address the issue of equality given that the group is working under conditions of unilateral censorship? It seems the Chinese food contingent would be in a good position to complain that their preferences were in effect tallied at a discount rate compared with their opponents, for this was the effect of introducing unilateral censorship. One way to attempt to reintroduce equality is to pro-rate each preference for Chinese food at 150%, so 20 voters @ 150% would equal 30 votes. The Mexican food vote could be reduced to 66.6%, so 30 @ 66.6% equals 20 votes. This way of pro-rating the votes would restore the original numbers of the vote before unilateral censorship was invoked. The fact that we might have to weigh some voters’ preferences more than others in order to reintroduce equality underscores the fact that unilateral censorship preferentially treats the preferences of some voters at the expense of others.
It may be demurred that in this case we know that unilateral censorship had an effect on the vote. What if the speeches in favour of Mexican food had had no effect? At least in terms of the outcome of the vote, it seems that the Chinese food lobby would have little reason to complain. So, let us concede, at least for the sake of the argument, that unilateral censorship might, in such cases, not usurp the democratic process.14 It hardly matters, since as a practical matter we are almost never going to have knowledge about such counterfactuals, e.g., how voters would have voted if unilateral censorship had not been in place. What this means is that even if unilateral censorship does not affect the outcome of the vote, most likely we will not be in a position to know this, so we will not know whether a vote was democratic or not. Furthermore, unilateral censorship could have a cumulative effect that would be even harder to track. Suppose speeches in favour of Chinese food do not influence voters to change their vote this time. Perhaps tomorrow’s vote on dinner might be affected by previous and ongoing campaigning by the Chinese food lobby. Since democracies typically are ongoing concerns, those that find themselves under conditions of unilateral censorship may worry about the cumulative effects of such censorship. The fact that it would be impossible to calculate the effect that suppressing one point of view will have on future decisions shows that this is not a workable means to reinstate equality. To emphasise how unworkable it is, imagine unilateral censorship had been enforced at the beginning of the 19th century on those advocating the view that women should be allowed to vote. Suppose 10% of the (male) voters were of the opinion that women should be allowed the vote. The authorities grant that suppression of the universal suffrage view will likely skew the vote so they generously agree to count every pro-universal suffrage vote at 200%. (That is, every ballot in favour of ending the exclusion of women from the political process would be counted as 2). At least in the short-term, this probably would have been a generous amount to concede: opinion on this issue changed very slowly. However, in the long-term this seems to underestimate the influence of the power of ideas to change public opinion. True, it took more than a century in many democracies to allow women suffrage, but I think it would have been to the long-term disadvantage of the universal suffrage movement if they had traded the opportunity henceforth to publicly argue for their position in exchange for counting votes in favour at 200%. In any event, the point of thinking about this means of attempting to reinstate equality is to underscore how unilateral censorship preferentially treats the political preferences and voices of some at the expense of others. In other words, unilateral censorship introduces a significant risk that the votes of some, in effect, count for more than one, and that the votes of others count for less than one.