Version classiqueVersion mobile
OpenEdition Books

Arguing about justice

Axel Gosseries
Philippe Vanderborght

English or Esperanto: a case for levelling down?

Marc Fleurbaey


En te papel es defendito que, contra distrivuo de bueneter, distrivuo de statu pode seldo doner cases en que tira-baso es, per tot considero, desirabil. Adopter Esperanto sor Angles com ”lingua franca” en context de justes lingust de Philippe Van Parijs similes un exampel, com si considero de bueneter alsi influes e suipes preferenso por Angles.

Note de l’auteur

I wish to thank Axel Gosseries for comments on an earlier draft of this paper.

Texte intégral

1In Linguistic Justice for Europe and for the World (2011) Philippe Van Parijs refines his theory of justice as real freedom in order to accommodate ”parity of esteem” in the context of competition between languages for recognition and use. He proposes to enforce such parity by a territorial regime imposing the use of the local language for administrative operations jointly with the dissemination of a lingua franca so as to make global communication accessible to all. Regarding the choice of the lingua franca, he considers that English now has reached a level of prominence that would render any other choice much less efficient. In particular, Van Parijs examines in detail the alternative of choosing an artificial lingua franca, such as Esperanto, or a multiplicity of linguae francae.

2There is much to be said in favour of Van Parijs’ approach to linguistic justice, and I shall not quarrel with it here. What this paper is about is whether the choice of Esperanto rather than English would involve a form of levelling down, i.e., a loss to some with no gain to others, and whether this is an argument against Esperanto or an argument in favour of levelling down.

Two effects of an artificial lingua franca

3If Esperanto (or any similarly artificial language) were chosen as the lingua franca, this could be viewed as a form of levelling down in two ways. First, the distribution of individual well-being might be adversely affected because an artificial language needs to be learnt by all, whereas a language that is the mother tongue of hundreds of millions of individuals saves these people the trouble of learning an additional language (at least those who would not do it for any other personal purpose). Of course, one can find other people who gain in the operation, if, as it is the case with Esperanto, the artificial contender is simpler than English. They might also gain (and the Anglophones lose) because competition for communication and access to the labor market would be more equal. But one might perhaps argue that the gains are very small compared to the cost endured by the Anglophones who must learn Esperanto. Van Parijs gives reasons to doubt that the gains just described would remain significant if Esperanto did acquire a prominent status and were to become the mother tongue of a sizeable population and develop irregular forms that would shorten words but would also diminish its simplicity. Conversely, a globalized form of English would almost become a foreign language to Anglophones, who would then lose part of their competitive advantage in communication.

4Second, and more interestingly perhaps, the replacement of English by Esperanto as the lingua franca would imply a form of levelling down among the natural languages themselves. If English is adopted as the lingua franca, it acquires a superior status above the other natural languages. If instead an artificial language is adopted, all natural languages share the same lower status as local tools of communication with no official cross-border function. No elevation of status is obtained by the other languages when English is downgraded from the status of lingua franca to that of a local language. The first form of levelling down had to do with the imposition of additional learning costs on many without any reduction of learning costs for anyone. This second form of levelling down associated with an artificial lingua franca consists in the status reduction of one natural language, with no status gain for any other language.

5As far as the distribution of status is concerned, this is a rather unusual configuration. In societies with unequal status among individuals, it is typically the case that eliminating the top status somehow elevates the lower statuses, or at least the immediately inferior statuses. Very often, the elimination of inequality of status actually takes the form of suppressing the lower status. The elimination of masters is described as the elimination of the status of slave. The case of nobility is less clear-cut perhaps, but certainly it goes hand in hand with the elimination of serfdom. The key problem concerning castes has to do with the untouchables, not with the Brahmans. For individuals, status is largely a relative notion, a positional good.

6In the case of the lingua franca, in contrast, it is possible to downgrade one natural language without upgrading the others. This is really a case of levelling down in status. It is as if the citizens of a country replaced their king, a fellow countryman, with a foreign, or extraterrestrial, monarch. The former kind is downgraded to the status of a subject but all the former subjects remain so. This exotic example confirms that levelling down in the distribution of status is unlikely to occur for a standard notion of social status.

7One could argue that levelling down among languages in terms of status is not as dramatic as levelling down among individuals, be it in terms of status or well-being. But insofar as the status of languages has a strong symbolic value, one can at least admit that it conveys a form of recognition to the native speakers. In fact, how could one understand the violence of linguistic conflicts if the symbolic value of recognition for the population itself, as distinct from its language, was not so great? This issue of symbolic recognition is the main motivation for Van Parijs’ approach to the ”parity of esteem”. Therefore, even at the level of populations, it is a form of symbolic levelling down for individuals, not just languages, if no individual can any longer feel the pride that his mother tongue is the world's lingua franca.

8We have seen above that levelling down in status is uncommon for individuals, because the lower status is generally raised when the higher status disappears. But, here, we have an interesting example in which the special pride of being a native speaker of the lingua franca can disappear without being replaced by a new pride for the others. Negative feelings of envy or jealousy, or simple frustration, may recede, but this has nothing to do with an elevation in symbolic status. While English seems to ”dominate” the world currently, its replacement by Esperanto would not make the native speakers of other languages ”freer” or more respectable in any sense.

Implications for egalitarianism

  • 1 An excellent recent collection of essays on this topic can be found in Holtug & Lippert- Rasmussen (...)

9What are we to make of this observation? The literature on levelling down in political philosophy1 makes a great deal of the supposedly repugnant possibility to find a single reason for a levelling down operation. Egalitarianism is criticized and often rejected in favour of prioritarianism because the latter does not find any good feature in levelling down whereas the former does find such a good feature when levelling down brings about equality. A not infrequent worry is that if egalitarianism finds it a good feature of levelling down that it may bring about equality, there must be cases in which egalitarianism considers it justified, all things considered, to implement a levelling down operation. This worry is obviously misplaced because egalitarianism can be combined with the Pareto principle which says that it is always good, all things considered, if the situation is improved for some individuals and worsened for none. However, it is indeed an interesting question whether a levelling down operation may in some cases produce an improvement, all things considered, and it is a less pressing but not uninteresting question whether it may produce an improvement in at least one respect. I belong to the crowd of those who are sceptical about the former possibility, believing that in general it is not a good thing to worsen the situation of some without improving the situation of others, but do not feel any quiver at the thought that levelling down may sometimes improve things in some limited respect (i.e., the distribution).

10Van Parijs rejects the option of taking an artificial language as the lingua franca, but does admit that it would have some advantages, in particular because it would be fairer, symbolically more neutral than English. This advantage in terms of fairness is not big according to him and can be overridden by efficiency considerations. But it is an advantage. This is a rather convincing point and appears to give at least an argument in favour of symbolic levelling down. This is not an all things considered judgment, because efficiency considerations point in the other direction, but this seems to provide a clear-cut case in which a pure levelling down in a symbolic dimension of people's status (via the status of their native language) displays some good feature.

11The possibility of finding such an example is not totally new, of course. If one considers equality to be a desirable feature of a distribution, one must consider that levelling down, when it produces an equal distribution (or a less unequal one), has at least a reason in its favour. What the linguistic example provides is another vivid example in which levelling down is good in some respect.

12Moreover, what is different between the classical egalitarian case in favour of levelling down and the linguistic example is that the latter deals with symbolic status instead of well-being. While egalitarianism in terms of wellbeing is challenged by prioritarianism, it is not common to think of a similar challenge as regards status and respect. It is very common in political philosophy to take equality of status and respect as basic values for a just society. The reason why equality instead of priority is usually invoked in this context may perhaps be that in the standard case of social status, it seems that a problematic egalitarian justification (at least in ”some” respect) of levelling down cannot arise because equality of status is always achieved at to the benefit of the lower status. In other words, equality is then never achieved by levelling down but rather by equalization at a middle level (the joint abolition of nobility and serfdom, masters and slaves) or by levelling up (the abolition of categories such as "the untouchables"). The example of the artificial lingua franca is an unusual case in which equality of status can only be brought about by levelling down one of the ordinary members and raising an external entity, the artificial language, to the higher status. If equality of status is desirable and can be obtained only by levelling down, this implies that levelling down, in this case, is good in at least some respect.

13At this juncture prioritarians may wonder if equality of status should be replaced with priority of status. Let us leave aside this issue, as there is nothing frightful, in my opinion, about considering that levelling down can be good in some respect. Let us instead examine another possibility. If equality rather than priority of status is a fundamental value, could it justify levelling down as being sometimes desirable, all things considered?

14In the case of levelling down in well-being, there is no doubt that efficiency considerations preclude the judgment that a levelling down operation improves the situation, all things considered. When efficiency is understood in the sense of Pareto, it is inefficient to worsen the situation of some without improving the situation of others when this can be avoided. All things considered, levelling down is therefore always bad.

15Equality of status and respect, in contrast, is a more basic and consensual value than equality of well-being or similar advantage. It is usually thought that it is obtained by equalization of status, or by emancipation of the lower status group. But what if it is obtainable only by levelling down? The Pareto principle does not seem as appealing in this case as in the context of distributions of well-being. Would it be an improvement to raise one member above the others if this could be done without lowering the status of the others? Would it be better to give some countries the pride of seeing their language play a global role, just for the sake of giving someone this advantage? Imagine that history had produced a situation in which Esperanto would be the lingua franca and would be the mother tongue of no particular people of the world. In this hypothetical world, imagine a philosopher arguing for the replacement of Esperanto by Flemish on the grounds that it would give the people in Flanders a special symbolic status and a unique pride, without changing anything to the status and lack-ofpride of any other people. Would he be taken seriously? It seems definitely harder to defend an inequality-generating Pareto improvement in the case of symbolic status than in the case of well-being.

16Why do we resist the application of the Pareto principle in this case? One possible reason is that it is hard to get rid of the intuition that if one is raised above the others, the others’ status is somehow diminished. In the linguistic example, this is really not the case. A local language does not lose any status just because another natural language, rather than an artificial one, is chosen as the lingua franca. The existence of a lingua franca, as such, may imply a symbolic reduction of status of the ordinary languages because they are thereby deprived of much of their potential role as cross-border communication tools. But once the principle of using a lingua franca is in place, it seems hard to defend the thesis that a local language suffers a loss if another local language, rather than an artificial one, is raised to the upper status. For instance, it is hard to claim that the Basques suffer a loss in symbolic status if English rather than Esperanto is adopted as the lingua franca, given that Basque is not the lingua franca in any scenario.

17Although the status of a local language is not diminished by the adoption of a natural lingua franca, there is still a specific loss to the speakers of the local language when they interact with native speakers of the lingua franca and when the latter use their superior communication skills to take advantage in debates and negotiations. But similar situations may occur with an artificial lingua franca. At least in the case of a natural lingua franca the native speakers cannot claim any merit in their superior skills, whereas with an artificial language inequalities in proficiency may be more reflective of unequal talents and therefore more dangerous for equality of respect. All in all, this specific loss does not seem to matter much.

18Another loss is obviously suffered by the defenders of Esperanto. If Esperanto were chosen, the Esperanto community would take great pride. This, however, is a transitional phenomenon that would disappear with the generation that defended Esperanto. And one can always think of creating another language that would be more neutral than Esperanto with respect to all existing communities.

19If one can get rid of any doubt about a possible loss to some people when a natural language is elevated to the ”global” status, does it become plausible that the Pareto superiority, in terms of the distribution of status, that goes with such elevation may provide an argument against an artificial lingua franca? I am not sure. Equality of status and respect seems better than unequal status in this case.

20This is not always true. One can find contexts in which the Pareto principle seems to apply. Levelling down cannot be accepted at all costs. For instance, if a society has the choice between everyone being the slave of an external master and an unequal society like the Western countries, in which differences of status are important but most people are above the condition of slaves, although some remain very close to it, the latter option seems preferable. In this case raising most people to a superior status seems desirable not just because of the effects on the distribution of well-being but also because the superior distribution of status appears preferable, even though it is less equal.


21We cannot yet conclude that, all things considered, Esperanto would be preferable to English as the lingua franca in a linguistic regime similar to what Van Parijs proposes. The conclusion is rather that the Pareto principle, which provides a definitive argument against levelling down in an all-things- considered judgment, does not seem to be as powerful as usual in the case of linguistic status. Raising English or any other natural language above the others by granting it the unique status of the lingua franca does create a Pareto-superior distribution of status, compared to a situation in which an artificial language is adopted. But that does not seem to constitute a knockdown argument in favour of this solution. Other arguments are needed, such as those provided by Van Parijs in his book, and related to the cost of adopting an artificial language and the well-being aspects alluded to in the beginning of this paper.

22Not only is equality of status a value in itself perhaps, as opposed to simple priority to the lower status, but there may be cases in which equality of status is preferable, all things considered, to unequal superior statuses. While the issue of Esperanto versus English does not seem to provide such an example because other considerations than status matter, it comes close to it.



HOLTUG, N. & LIPPERT-RASMUSSEN, K. (2007), Egalitarianism. New Essays on the Nature and Value of Equality, Oxford: Oxford University Press.

VAN PARIJS, P. (2011), Linguistic Justice for Europe and for the World, Oxford: Oxford University Press.


1 An excellent recent collection of essays on this topic can be found in Holtug & Lippert- Rasmussen (2007).


Marc Fleurbaey is Laurance S. Rockefeller Professor of Public Affairs (Princeton), after having been successively an economist at INSEE, a professor of economics at the Universities of Cergy-Pontoise and Pau (France), and a Research Director at the CNRS (Paris-V). He is a former editor Economics & Philosophy and a managing editor of Social Choice and Welfare. He is the author of Fairness, Responsibility, and Welfare (Oxford U.P., 2008), a co-author of A Theory of Fairness and Social Welfare (Cambridge U. P., 2011, with F. Maniquet), and the co-editor of several books, including Justice, Political Liberalism, and Utilitarianism: Themes from Harsanyi and Rawls (Cambridge U. P., 2008, with M. Salles & J. Weymark,). His research focuses in particular on the analysis of equality of opportunity and responsibility-sensitive egalitarianism and on seeking solutions to famous impossibilities of social choice theory.

© Presses universitaires de Louvain, 2011

Conditions d’utilisation :