How Many Languages Do We Need?: The Economics of Linguistic Diversity

NEED?

INTRODUCTION

THE TITLE OF THE BOOK, How Many Languages Do We Need?, reflects a difficult choice that was pervasive in many multilingual societies over the course of human history, and still is today. The problem is very much on the agenda of a large number of countries, regions, and international unions that must address various aspects of multilingualism. The desire to avoid an excess of societal fragmentation in our rapidly globalizing environment narrows the focus to a relatively small number of core languages. But such restrictions inevitably disenfranchise speakers of non-core languages. In this book, we analyze the trade-off between the quest for efficiency that a small number of languages is thought to foster and a reduction in disenfranchisement, which calls for more languages.

Linguistic policies vary across the globe. The European Union adopts the policy of twenty-three official languages,² India has instituted the famous three-language policy, a variant of which was also considered in Nigeria. The range is wide, and obviously much depends on the way policies are implemented. The EU’s policy has its problems, but has not led to wars. Nor do the 207 languages in Australia, or the 364 languages in the United States. In Sri Lanka, however, two languages were one too many, and thousands of people died as a consequence.

The search for an optimal number of languages is implicitly linked to a second question about which languages to choose. Languages differ from each other; some are close, others are distant, but we can easily admit that however we measure distance, Spanish and Italian seem closer than English and Greek. That fact will play an important role in our analysis.

DIVERSITY COULD BE GOOD, BUT IT IS NOT FREE

Economists are two-handed. One hand immediately recognizes that the diversity of existing cultures and languages, which often cannot be dissociated, is important. Limiting the number of languages creates a disenfranchisement that even developed and democratic countries sometimes cannot support. History, including that of the contemporary era, shows how oppression or suppression of languages or cultures may lead to bloody wars. It is shocking and horrifying to think of the cost of linguistic policies in terms of human lives, but failure to do so ignores the passion and violence generated by the defense of or attack on one’s own culture and language. Languages have evolved over the course of human history, sometimes for smooth reasons, at other times for brutal ones. The following few questions (there are many others) are extremely interesting: (1) Why did some languages survive and even grow? (2) Why are some languages spoken by millions or even billions of people, and others by a few hundred or even less? (3) How do languages change, and why (for linguistic reasons such as phonetics or syntax; for imposed reasons such as invasions, migrations, or colonization)? These questions have been dealt with by linguists, historians and archaeologists, and more recently by biologists. We brush up against some in the chapters of the book, but space considerations unfortunately prevented us from a more detailed examination.³

The importance people and countries attach to their native language and culture can also be illustrated by the substantial resources devoted to advertising and promoting their culture, language, and literature. Some (Great Britain and its British Council, France and the Alliance Française, the German Goethe Institut, the Italian Società Dante Alighieri, or the Spanish Instituto Cervantes, to cite a few) have institutes in many other countries to teach their language and culture. Countries subsidize translations both into and from their own language.⁴

For many, the protection of linguistic diversity is especially important in our globalized environment, often dominated by English. This issue is discussed later, but here we would like to point out that what seems to be the relentless march of English can breed strong feelings of disenfranchisement. Here is what the American–Chilean literature professor and author Ariel Dorfman (2002, 92) has to say:

The ascendancy of English, like so many phenomena associated with globalization, leaves too many invisible losers, too many people silenced. … Do you come from a place … that does not control a language that commands respect? Do you reside in a language … whose existence does not have the kind of value in the marketplace that can get you a good job and help you in [the] everyday struggle to survive?

While one hand supports the virtues of linguistic diversity, the other hand would discreetly point out that they do not come for free. A host of economic studies suggests that too much diversity, whether linguistic, ethnic, religious, cultural, or genetic, breeds institutional wastefulness, bureaucratic inefficiency, and corruption. This so-called fractionalization may reduce political stability and hamper economic growth, as shown by the sad and painful example of postcolonial Africa. Cameroon has 279 languages, the Democratic Republic of Congo has 217, Nigeria 527, Sudan 134, and Tanzania 129. These staggering numbers—especially if they also represent ethnic groups—have often led to the argument that to accelerate economic progress, a country or a region may have to consider limiting the number of languages used for official, commercial, or legal purposes to a manageable and reasonable few.

The costs of linguistic diversity are not always direct and are sometimes the consequence of poor translations and miscommunication. History is full of episodes in which these factors played an essential role in various international conflicts.⁵ A well-known example is Article (i) of UN Security Council Resolution 242 of November 22, 1967, which requires the withdrawal of Israeli armed forces from territories occupied in the recent conflict. The omission of the definite article before territories has been often interpreted to imply that the withdrawal refers to some but not all territories covered by the resolution. This delicate but crucial distinction was lost in the French and Russian versions. Indeed, no definite article exists in Russian, whereas French requires inclusion of the article. The Arabic version includes the article, but Arabic was not an official UN language at the time.

WHAT THE BOOK IS ABOUT

The purpose of our book is to use economic and to some extent linguistic and sociological reasoning to balance the benefits and costs of various linguistic situations and policies. While extolling diversity, we must recognize that it may come with institutional wastefulness and inefficiency, corruption, legal and communication barriers, and lack of interest on the part of citizens about what happens in their country or region, or may lead to unrest, riots, and even wars. So also does restricting diversity, which results in disenfranchisement. It is a difficult challenge to decide on a sensible number of languages. That number is generally not one. It is probably not forty-two, the answer given by the computer in The Hitchhiker’s Guide to the Galaxy to the question, What is the answer to life, the universe and everything?⁶ And it is certainly not 527, the number of languages spoken in Nigeria today. Our contribution is to show that in the recipe, linguistic distances, which the reader will encounter in many chapters, are an important ingredient.⁷

The two opening chapters discuss the importance people attach to their native language and culture, and describe how linguistic policies have led to the alienation and disenfranchisement of various groups or individuals restricted in their linguistic rights.

Some very basic notions about linguistic and other types of distances are introduced in the next chapter and used throughout the book, especially in chapter 4, which illustrates the importance of linguistic or cultural distances in issues as diverse as bilateral international trade, the choice of immigration destination, translations of literary works, and the voting behavior of countries along linguistic and cultural lines. Smaller distances (or a common language) between countries have a positive impact on their trade. In many cases, immigrants are attracted by countries with which they had former contacts—through colonization, for example—and so gained some exposure to the culture and language. More literary works are translated from source languages (and cultures) that are close to the destination language. And the winners in a popular song contest in Europe are elected by those with whom they share cultural traits, including language.

But languages also evolve over time. In particular, learning foreign languages may alter the linguistic repertoire of various communities. To evaluate the patterns of investment in human capital and private returns from learning foreign languages, in chapter 5 we use the theory of communicative benefits introduced by Nobel Prize winner Reinhard Selten and American linguist Jonathan Pool (1991) and link it to linguistic diversity. Distances again play a prominent role, since the decision to learn a new language strikes a balance between costs (the difficulty of learning it, which correlates with distance) and benefits. Will it pay to learn a language?

The next two chapters address the issues and the consequences of linguistic and ethnolinguistic diversity, fractionalization, and polarization, as well as disenfranchisement. Diversity, fractionalization, polarization, and disenfranchisement indices that measure these phenomena have a long history in linguistics and sociology, starting with the contribution of Joseph Greenberg (1956) and its later extension by Stanley Lieberson (1964), two distinguished representatives in their respective disciplines. Such indices have more recently been used by economists, who have underlined the important role they play in explaining economic growth, redistribution, income inequality, corruption, political stability, and bureaucratic (in)efficiency in many developing countries and regions in which democracy is little more than a buzzword. We also stress the fact that in empirical work, indices that account for linguistic distances perform better than those that do not.

The last chapter is devoted to a case study of linguistic policies in the EU. Linguistic policies should achieve a delicate balance, taking into account efficiency considerations without forgetting the will of the people. We suggest that though the English-only solution may yield greater efficiency and save on costs, it could generate too large a degree of disenfranchisement that leads to less political unity and citizens’ loss of interest in a true European Union. We point to several alternative solutions that include the political feasibility of altering the extent of translation to some core languages, the provision of compensating transfers for countries that would be ready to cover their own translation needs, and some fair methods of sharing the global cost of translation and interpretation, which could be of use in other parts of the world.

SOME HINTS FOR A PIECEMEAL READING

Though we have tried to make the presentation as simple and accessible as possible, it is unavoidable that some chapters of the volume are more technical than others. Linguistics is a difficult topic, and so are economics and sociology, even when discussed in isolation. Here, all three disciplines are combined. Chapters 1 and 2 are of a purely descriptive nature. Chapter 3, on distances, uses linguistic concepts, and the reader, unless familiar with the field, should devote some time to reading this chapter, which lays out the basic insights used in all the other chapters. Chapter 5 uses game-theoretic concepts, and chapter 6 defines diversity, fractionalization, polarization, and disenfranchisement indices, which need some mathematical formulation. The reader who is not interested in technicalities may skip the first section of chapter 5 and chapter 6. Chapters 4 and 7 are devoted to the many empirical studies whose object is to estimate the impact of linguistic distances and ethnolinguistic diversity on economic outcomes. They include many tables devoted to econometric results, but these results are also discussed in simple terms, and both chapters can thus be read without relying on the detailed tables. The second part of chapter 7 and chapter 8 are devoted to a case study of linguistic policies, and most of it is easy reading. The last part of chapter 8, devoted to showing how the financial burden of the more than one billion euros spent every year by the EU on translation could be shared in a fair (or fairer) way, is a little more demanding.

Having given these warnings, we briefly address the question of the potential readership of the book. Both authors are economists, and it is obvious that their fellow economists who have had some basic training in game theory and econometrics, even if it is somewhat rusty, should be able to read the book without any difficulty. Students with a background in sociology or linguistics will probably face more problems, though they may skip the more technical chapters without losing the flavor of our arguments. A very basic economic training will permit following most of the discussion. The same is true for the lay reader. After all, as economists, we were also—and still are—lay readers in sociology and linguistics but had to grapple with some difficult concepts that we felt were necessary for our thesis.

²One per country, with the exception of Luxembourg. Some countries—Germany and Austria, for example—share a common language.

³The interested reader is referred to chapter 7 of Cavalli-Sforza and Cavalli-Sforza (1995), as well as Cavalli-Sforza (2000) on prehistorical times and ancient human diasporas, Janson (2002) and Ostler (2005) on antiquity and more recent times, and McWhorter (2001) on pidgins and creoles. Renfrew (1987) examines links between archaeological discoveries and languages.

⁴The list of countries that offer subsidies for translations is large. See http://portal.unesco.org/culture/en/ev.php.

⁵See Lewis (2004) for more examples.

⁶We owe this reference to Richard B. Thank you, Richard, for this, and for so much more.

⁷Our goal has nothing, or very little, in common with the book by Ariel Rubinstein (2000), Economics and Language. Although the title is somewhat similar to ours, Rubin-stein is mainly concerned with the links between the formal language of mathematical models and natural language. As important as they are, those issues are far removed from our main interests.

Chapter 1

MY LANGUAGE IS MY HOMELAND

Minha pátria é minha lengua.¹

—Fernando Pessoa

Russian and Swahili. The authors of this book have two different native languages, Swahili and Russian. A few years ago, the Russian native speaker accompanied the Swahili native speaker on a trip to Belarus in search of traces of the latter’s father, born in Vitebsk (as was Marc Chagall, at approximately the same time). The hotel receptionist was totally confused by the fact that the traveler with a Russian–Jewish name (the Swahili speaker) could not utter a word of Russian, while the other, with a German name, spoke fluent Russian. This was without the receptionist’s knowing that the visitor with whom she was unable to communicate was a native Swahili speaker. The fruits of globalization, as we say.

We live in a world where individuals, families, and communities often uproot themselves, voluntarily or involuntary, and move to another place, country, or even continent. In fact, both of the authors have gone through this experience at different stages of their lives. One of us was given less than a week to collect his belongings before leaving his native country. What does one take on a journey to a distant, unfamiliar land? Books, memories, records? A very moving response to that dilemma is given by Ariel Dorfman, whose work was mentioned in the introduction. Dorfman (2002, 89) recalls Gabriel García Márquez’s story of entire Colombian villages that migrated:

Fleeing from catastrophes, plagues perhaps, or recurrent floods, or merely the desolation of being caught in the middle of civil wars, inhabitants of these villages decided, at some point in history, to uproot themselves, moving to a remote location in search of peace. As they packed every belonging that could be transported, they did not forget what was most important to them: their dead. According to García Márquez, these villagers, on the verge of becoming nomads, dug up the bones in the cemetery and, in effect, carried their ancestors into the unknown, probably animated by the need to defy the fluctuations of time and geography with the illusion that something from the past permeates the present, forming a hard physical link to memory at a time of devastating change.

Even though the desire to sustain links with the past is not always expressed in such a dramatic form, an important part of cultural history and heritage is invariably linked with language. The sensitivity of linguistic identity is forcefully underlined by Henry Bretton (1976, 447):

Language then may indeed be the most explosive issue universally and over time. This mainly because language alone, unlike all other concerns associated with nationalism and ethnocentrism … is so closely tied to the individual self. Fear of being deprived of communicating skills seems to raise political passions to a fever pitch.

When moving to distant lands, migrants take with them their language and begin a never-ending romance with it. Dorfman (2002, 90) continues:

That language, which contains the seeds of their most intimate identity, will be put to the test when the voyage is over, especially if the migrants are moving to a foreign land. This is because those who await them at the new location have their own dead, their own ceremonies and cemeteries, and, of course, their own tongue.

The voyage, even if accomplished successfully (whatever this means), leaves deep marks and scars. We learn different languages, assume different linguistic identities that may confuse others and sometimes ourselves.

The threat to the survival of values, practices, and history that is posed by migrations is not limited to individuals and communities. History is full of examples (and we discuss some of them in subsequent chapters) of linguistic groups and regions whose experiences include language conflict, inherent incompatibilities between language communities, reluctance of the majority group to concede linguistic rights to minorities, the power of civil servants to protect their linguistic privileges, and the important and unpredictable material and symbolic consequences of linguistic choices (Pool 1991).

Even entire countries can feel insecure about their linguistic survival under threats coming from outside. Consider Iceland, one of the few monolingual countries in the world and practically the only surviving example in Europe of a linguistically homogeneous nationstate. Almost every Icelander speaks Icelandic as a first language and uses it as the dominant language in all spheres of life (Vikor 2000). Even without an internal linguistic issue, however, the country still felt compelled to protect its own identity, closely linked with its language. For about four hundred years there were numerous attempts to purge the Icelandic language of foreign, mostly Danish, influence. More recently, the dominance of English has become a serious threat to Icelandic identity, which developed through centuries of isolation, literary works, and the fostering of independent cultural values. The protection of their language included the successful attempt by Icelanders to persuade Microsoft to add Icelandic to the list of languages supported by Windows. It took more than six years of persistent lobbying by Icelandic authorities to overturn an initial refusal by Microsoft, but in 2004 the company produced an Icelandic version of Windows XP and the Office Packet of associated programs (Corgan 2004). In the beginning of the twenty-first century Iceland was still spending about 3 percent of its gross domestic product on activities related to the promotion and defense of its language, including translating and printing an astonishing number of foreign textbooks and fiction. Recently, Iceland engaged in serious attempts to become a member of the EU. While the survival of the language is not usually listed as a reason for the European shift in the country’s foreign policy, one can notice, nevertheless, that the language’s protection would receive a tremendous boost from the EU’s multilingualism policy.

Iceland notwithstanding, almost all countries are multilingual. Indeed, the 2009 edition of Ethnologue lists 6,909 distinct living languages spoken all over the world, and since there are only 271 nations, dependencies, and other entities, the issue of linguistic diversity appears almost everywhere. Even though many of these nearly seven thousand languages are spoken in small and often remote and isolated communities, ethnic and linguistic heterogeneity is not an exclusively Third World phenomenon. Despite a long tradition of nation-states in Western Europe, most countries there are multilingual, with a plethora of indigenous regional languages such as Welsh in the United Kingdom, Catalan, Basque, and Galician in Spain, Provençal, Breton, and Occitan in France, and Frisian in the Netherlands. Obviously, in the study of societies that consist of several groups of citizens, one has to take into account various aspects of diversity, including wealth, geography, history, religion, and ethnicity. Each of these factors has an effect on the strength of the country’s economic and political institutions, its growth, its stability, and its patterns of wealth redistribution.

The preservation of linguistic diversity and respect for the cultural heritage of members of a society is an important and much-needed task. However, in making policy and operational decisions, one has to recognize two caveats in protecting an excessive degree of diversity:

There is no free lunch for diversity. Sustaining a high degree of societal diversity could require allocating substantial resources to creating educational institutions and developing communication and coordination between groups. The financial cost and communication barriers may have a negative impact on national unity and, in general, may heavily burden the proper functioning of the society’s institutions. It is difficult to assign a precise cost to sustaining diversity, but this cost nevertheless exists and is, in some cases, far from negligible. As Harvard economist Martin Weitzman (1992, 363) points out:

Often there is an implicit injunction to preserve diversity because it represents a higher value than other things, which by comparison are only money. Yet laws of economics apply to diversity also. We cannot preserve everything. There are no free lunches for diversity. Given our limited resources, preservation of diversity in one context can only be accomplished at some real opportunity cost in terms of well-being forgone in other spheres of life, including, possibly, a loss of diversity somewhere else in the system.

When is the meal too salty? How much salt do we want in our meal? No salt is not good at all for most customers, but by gradually increasing the amount of salt, we may quickly move from tasteless to uneatable. Shouldn’t there be an optimal amount of salt, not too little, not too much? Also, how many spoons of sugar are needed for our morning coffee or tea? Some do not use sugar at all, Russians put three or four spoonfuls in their tea and drink the tea with the spoon in the cup, continually stirring their sticky beverage. Again, an optimum has to be found somewhere between zero and four, five, or six. Some art lovers may find black-and-white paintings rather dull, but adding an excessive number of colors does not necessarily make a painting more attractive. In all these examples—salt, sugar, colors—there exists an optimal degree of diversity that ensures a tasty meal, a pleasant cup of coffee, or sometimes a beautiful painting.

A study of diverse societies may suggest that there is an upper bound to the degree of heterogeneity that still guarantees proper and sustainable functioning. If it is too diverse, the society may fail to develop and implement policies acceptable for large groups of its citizens, and even the basis of societal existence could be questioned and threatened. The challenges of making everybody happy or even obedient are tremendous, and the eventual collapse of the Macedonian, Roman, Ottoman, and Mongolian empires was in large part caused by the inability of their rulers to keep all diverse and distant parts of their kingdoms under a unified rule. This happened not that long ago with the Russian empire, and it is happening now in China with Tibet and Xinjiang. One can wonder (as many do) whether the potential inclusion of countries such as Turkey and Ukraine in the EU would increase diversity within the EU to a degree that would rule out a reasonable compromise among its members.

Naturally, the determination of linguistic policies that reflect an optimal degree of diversity is a very challenging and delicate task. After all, language, like religion, does not lend itself easily to compromise (Laponce 1992, 599). Since language is an intimate part of