Your language or mine?

Barcelona GSE Focus

Having a common language enhances social and economic interactions in a society by improving communication. Hence, for the society as a whole clinging to more than one language seems redundant, an inheritance of the past, almost a punishment as the tale of the tower of Babel exemplifies. Nevertheless, “to have another language is to possess a second soul”, at least this is what Charlemagne believed.

Languages are not merely communication devices. We are emotionally attached to our languages, they contribute to the definition of our identities. If this is the case then, how important are they for different individuals? How is individual behavior affected by the emotional attachment to one’s own language? Do individuals change their economic decisions because of the identification they experience with their language? And if so, how should policies be designed to enhance economic welfare?

In their Barcelona GSE Working Paper (No. 852) “Your Language or Mine?” Ramon Caminal and Antonio Di Paolo take two important steps to answer these crucial questions. First, they provide a theoretical framework that takes into account both potentially conflictive dimensions of the problem; that is, the communicative versus the emotional dimension of languages. Second, they empirically test their model’s predictions in the particular but very fitting case of Catalonia.

Charlemagne in theory: languages, communication and identity

The authors model a bilingual society where the strong language is spoken by everyone, like, following the leading example, Spanish in Catalonia, and a fraction of the population additionally speaks its own language, e.g. Catalan. All individuals not only value the presence of a common language given it enhances social and economic interactions but also value their own language because of the emotional attachment it prompts. This means that all individuals are able to communicate and interact with each other, thus if they decide not to, it is because of the emotional attachment they experience toward their own language. For example, a native Catalan speaker might decide not to interact with a monolingual Spanish speaker, even if the interaction would be profitable for both parties and even if she knows Spanish, because of the identification she experiences with her own language. The sentimental loss the Catalan speaker would incur by renouncing to her language outweighs the gains from trade. This is the key and novel trade-off the authors are able to highlight and analyze thanks to their framework.

Since a unilateral decision not to trade may generate losses to other potential partners, the outcome is typically inefficient. This inefficiency leaves space for a policy intervention aimed to enhance the economic welfare of the society, but at what cost? What should be done to reduce the space for inefficiency and hence let the society take full advantage of profitable exchanges?

The authors focus their attention on the costs the speakers of the strong language would suffer whenever trying to learn the weak language. In a society where everyone knows the strong language, profitable exchanges can only be lost because some individuals refuse to trade because the distress they feel whenever renouncing to their mother tongue is not compensated by the gains of the exchange. Thus, the only feasible way for a policy maker to disrupt such inefficient dynamics is to make easier for the speakers of the strong language to learn the weak language. In principle, this intervention should be able to allow some exchanges that would be impossible otherwise. To see this point more clearly, suppose that there are potential gains from trade between a native Catalan and a monolingual Spanish speaker, but due to the emotional attachment of the Catalan to her own language the exchange does not take place. Now if the government reduces the costs of learning Catalan for the Spanish speaker, then it might very well be the case that the Spanish speaker is willing to renounce to his language hence making the exchange possible. This is one of the theoretical implications of the model.

In the paper the authors are able to show that reducing the costs of learning the weak language increases the frequency of successful mixed exchanges, and hence reduces segregation. This improves the welfare of the speakers of the weak language and, provided the learning costs are not too high, the welfare of the society as a whole. This is a challenging idea given that political debates usually underline the importance of economic policies aimed to standardize languages and thus the preservation of semantic minorities only happens for historical or cultural purposes. On the contrary, the authors are able to theoretically show that there might be economic gains from the inclusion of different languages in the same society through the reduction of the segregation.

Charlemagne and La Langue d’Oc: The Catalan Case

The theoretical results are confirmed with the empirical analysis of the Catalan case, using data from L’Enquesta d’Usos Lingüístics de la Població, a survey carried out by IDESCAT. Languages seem to be much more than mere communications devices.

The authors are able to draw such a conclusion because their empirical strategy exploits an exogenous change in the Catalan legislation happened in 1983, the so called “Language Normalization Act”. The act had the objective of making every child fully competent in both Spanish and Catalan by the end of compulsory education. This was a radical change with respect to Franco’s dictatorship that prohibited any language that wasn’t Spanish.

This historical case represents the perfect environment to test the model’s predictions. In fact, up to 1983, all residents in Catalonia were perfectly proficient in Spanish and only a fraction was speaking Catalan. The policy change reduced the costs of learning the weak language, i.e. Catalan, for the native speakers of the strong language, i.e., Spanish, thus making the historical event as close as possible to a lab experiment of the theory.

The authors study how the probability of having a Catalan-speaking partner and the probability of speaking only Catalan with the partner are affected by the policy change. Oral Catalan skills are measured on a 0-10 scale. They find that a unit increase in fluency in oral Catalan, due the 1983 reform, increases the likelihood of a mixed match, by 7.6 percentage points. Similarly, a unit increase in fluency in oral Catalan increases the probability of speaking only Catalan with the partner by 5.3 percentage points.

These empirical results confirm the fact that languages prompt an emotional attachment, making the acquisition of language skills that appear redundant from a communicative point of view, central for the reduction of segregation in a society and hence for the enhancement of general welfare. Thus, it is true, a second language is like a second soul.