The returns on homogeneity

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

A few days ago on Twitter I wondered if economists had calculated the costs of the world having a diversity of languages, instead of one language. The logic is that unintelligibility naturally throws up barriers to communication, and the flow of ideas and labor. This is one reason why the European Union necessarily has less labor mobility than the United States all things equal, the vast majority of Americans can understand each other to a very high level of clarity, while Europeans can attain this only within their nations. Even in societies where English is widely understood the level fluency variance and lack of uniform idiom may produce a suboptimal lingua franca across these English semi-fluent societies.

But Christopher Burd challenged me on this. He suggested that semi-permeability across cultures may actually be a feature, not a bug. As an example he offered the relative isolation of the Russian academic community, and so its insulation from the orthodoxies of English-speaking linguistics, allowed for the wider acceptance of heterodox models such as the Nostratic superfamily. I asked Burd if he was making an analogy to the shifting balance theory, and he stated that he was.

In 20th century evolutionary genetics the importance of population substructure in shaping the arc of adaptation was at the heart of some of the Fisher-Wright controversies. In short R. A. Fisher conceived of the adaptive landscape as a relatively simple geometry whereby the total census size could be thought of as a breeding population and favored alleles scaled a topography characterized by a single adaptive fitness peak. Sewall Wright’s adaptive landscape exhibited a “rugged” topography, with multiple peaks. The relative genetic isolation between the groups could allow for exploration of the non-additive space of epistatic interactions, as well as temporary shifts through low fitness valleys via random genetic drift.

To some extent the nature of this argument was obscure and nearly intractable in an analytical sense. Will Provine in his biographical treatment of Wright’s ideas has argued that Wright himself had a vague and incoherent conception of his fitness landscapes. In evolutionary genetics fitness landscapes may serve more as illuminating metaphors useful for the general public as opposed crisp models which trigger empirical research programs (see Sergey Gavrilets for a contemporary attempt to make sense of fitness landscapes).

But culture is different from genetics. What Christopher Burd was implicitly alluding to I believe is the problem of herd behavior among humans; positive feedback loops which swing too far and may be counterproductive. Fisherian runaway sexual selection is probably the evolutionary genetic analog to the cultural phenomenon of irrational herds. Eventually natural selection kicks in and the positive feedback loop will run up against limits. If a whole species is subject to runaway sexual selection its census size may start decreasing if sexy traits have lower ecological utility than unsexy traits (those with unsexy traits may survive, but may not be able to find mates during the initial stages). Eventually the process will “self-correct” and the runaway will exhaust itself.

By analogy with irrational herds if the whole world shifted toward a crazy idea which spread through positive feedback loops it could cause a great deal of damage before people came to their senses. In contrast, if the world was broken up into distinct segments then the irrationalities would differ, and perhaps some of the societies would benefit from having less destructive irrationalities.

In evolutionary biology the importance of structure beyond the level of the individual has generally been dismissed until recently, when the debate has reemerged. But in the social sciences the methodological individualism of the rational actor has had less of a stranglehold, excepting economics. In regards to my question about language I suppose the bigger framing question is whether institutions and organizations really matter on a deep level as more than the sum of their parts. I suppose I lean toward the idea that they do matter, because humans are profoundly social creatures, far more than the vast majority of the world’s organisms. Pluralism of institution and organization, the existence of nation-states, is I think useful for human flourishing. So I suppose I will retract my assumption in the question I posed, because I now think it is an open matter as to whether one world language would truly be beneficial, rather than an open and shut case.

This question also is embedded in more abstract or general issues. Some economic historians have argued that one reason that the United States flourished despite the rise of tariffs in the 19th century was that it was in and of itself a huge free trade zone. In other words there were diminishing returns to trade, and much of what was to be gained in comparative advantage was already gained via the emergence of canal and rail networks within the continent United States. The extreme linguistic diversity of less developed regions of the world, or even 18th century France and Italy, is probably detrimental to economic growth and economies of scale, but do diminishing returns kick in at some point? And once diminishing returns kick in is there a point where linguistic difference as a way to foster some distinction between nation-states or groups of nation-states is a net utility positive? I still think that the European Union, for example, would benefit from fewer languages. In particular, speakers of languages with less subscription tend to suffer in my opinion, and they would benefit from “trading up.”

Even more generally what about other markers of cultural distinctiveness? Would the world benefit from one higher religion, or are we in the best of all worlds, where half a dozen religions dominate almost all societies (Christianity, Islam, Hinduism, Buddhism and East Asian religio-philosophy)? It is true that the narcissism of small differences would still be problematic (in Pakistan Sunnis kill Shias and Shias kill Sunnis, in India the conflict is Muslim vs. Hindu, and so forth), but it does seem one world religion would at least result in a commensurability of metaphysical presuppositions (more concretely, for all the differences between Roman Catholics and most Protestant groups, the two groups accept each other as Christians in good standing).


  1. In my high school/lycee history/geography class in France, the fact that France has one common language was specifically taught as one of France’s strengths. (concrete reason to feel superior to and stronger than belgium and switzerland.) It obviously makes a lot of things easier, but I thought it was weird that it was taught as an absolute and an obvious strength, instead of presented as a question…

  2. I recall Freeman Dyson arguing for the advantage in intellectual productivity of a world with multiple semi-permeable languages back in the 1970s.

  3. but I thought it was weird that it was taught as an absolute and an obvious strength, instead of presented as a question…

    well, very few people living in france in 1790 spoke “french,” and france has traditionally tried to eliminate all local languages or dialects in the interests of state formation. part of the ideology of the post-revolutionary republican france. though it would be strange from an american perspective where language is more of a bottom up affair.

  4. The very reason Europe came out ahead was because of the heterogeneity in language. You had multiple centers communicating and competing with each other. The opposite of that, China, was the reason why Asia lagged behind. The reverse of that, the multiple Chineseness of now, is why China is in a renaissance now. Most of the initial investment that kick-started the Chinese behemoth was from HK and Taiwan, and most of the innovative ideas behind this great leap either were from Taiwan, HK-Macua, Singapore or via these place. And they do speak different languages. At a certain point of the curve, having multiple languages become a productive thing, if we ignore the cultural benefit of multi-culturalism.

  5. I am reminded of James Scott‘s theory that illiteracy is a tactic of “hill people” to avoid the state, with official records being a major target in their periodic uprisings. In “Seeing Like a State” he also suggests that an incomprehensible geography (with the layout of Bruges being a prime example of his) gives locals leverage against outsiders.

  6. And they do speak different languages.

    aside from hong kong standard mandarin was/is the elite language in those polities.

    btw, catholic & protestant european elites were intelligible to each other up until the 18th century (and often later) in latin. this made it easy for university students from hungary to study at oxford, for example, if they didn’t have english.

  7. Abstracted rational thought and motivations also allow for greater internal selection within the human “herd”. This is a feature that no other species has demonstrated yet – not to mention the rapid transmission of ideas via the Internet. Any culture’s values can come under the microscope and be deemed “wrong”. Now, a cultural value has to have logically reason behind it to gain support with the younger generations. Logic and reason are taking over, in my opinion, rather than elitist cultural objects.

  8. If I had a thousand scientists or artists in one institute and wanted to foster innovation, I probably wouldn’t hold a single series of gigantic workshops: I’d split them into many separate workshops but have them report their most interesting results at a general seminar every so often. This is obvious, right? A single big workshop would no doubt produce an excellent product, but we know intuitively that there would be hierarchies, gatekeepers, and so on, that would keep some ideas out of discussion, and even from being thought in the first place. (It may be these hierarchies also play positive roles by fostering *coherent* development, but I’m not convinced.)

    If, as a God-like being, you wanted run this experiment on a continental scale, you’d create something like post-Mediaeval Europe: separate but adjacent cultures, partially isolated by geography and language. In Europe, mountain ranges and seas would have encouraged this “good balkanization” anyway, but the North European Plain, for example, is divided into three major (and several minor) cultural divisions on the basis of language alone. So even in the Middle Ages, you have Occitan troubador culture in S. France and a related, Middle High German courtly culture in S. Germany.

    Think of smaller groups too. Standard German and Standard Dutch both emerged from a West Germanic dialectal continuum, the latter because the Netherlands achieved early independence. Otherwise Dutch might have gone the way of Plattdeutsch. Arguably, having their own literary language helped foster the Netherlands’ high level of innovation. Maybe Dutch scientists and artists would have done as well in a society like the Buddenbrooks’ Lübeck – a Hochdeutsch-speaking upper class with a Low German-speaking lower class – but I’d bet against it: Dutch science might have become a provincial branch of German science; Dutch scientists emigrating to the German scientific metropolis might have found themselves buried in larger academic hierarchies.

    The same may be true of Denmark (vis-a-vis the Scandinavian language bloc) and Scandinavia in general vis-a-vis the Germans. (Certainly the standard Scandinavian views sees the period of Hansa dominance as a one of cultural stagnation.)

    A look at the distribution of Nobel Prizes suggests that the smaller national languages hold up pretty well, but the counter-example of Switzerland: mostly Hochdeutsch + Swiss dialect, with all the serious work done in Hochdeutsch. But Switzerland is clearly its own culture, even if only partly built on language (and partly against language!).

    I think you can overestimate the role of Latin as an interlanguage. Universities were Latin-speaking (even the student-quarter prostitutes, apparently) but a good command of Latin was a rare, if respected accomplishment. Increasingly, books were written in the vernacular, and cultural cross-pollation was largely via translation, especially outside of academics.

    ON the other hand, it’s not quite true to say that there were hundreds of languages. True, at the village level, each of the big language blocs (Romance, Germanic) were finely grained mosaics. But there were never hundreds of literary languages. In the middle ages, there were generally several courtly standards in each country. The modern period saw the rise of single national standards. In the France of 1789, this was standard French; nothing significant was being done in the dialects, although most French spoke dialect.

Leave a Reply