Old-time religion, old-time language: Jesus buried in Japan? Literal belief in the Tower of Babel? Hindu influence in the American Southwest? In the context of nonstandard claims by fringe scholars and revisionists about the remote past, religion and language are often intertwined

In the context of nonstandard claims about the remote past, the fields of religion and language are both highly significant but might superficially appear largely unconnected. However, there are a number of cases in which religious and linguistic issues are intertwined. This is not entirely surprising, since both religion and language are core elements in human culture and thought. Language is both a determining factor for human thought and its most articulated vehicle of expression. Many human groups regard their language as an identifying characteristic. Folk-linguistic beliefs often center on the origin of the language, treated as a key aspect of mythological/religious accounts of the origin of the group and its world. Positions on the two fronts are thus frequently connected.

Starting with Genesis

One instance of this phenomenon that is well known in the West is the Tower of Babel story in Genesis. This Hebrew myth explains the diversity of human languages in terms of an initial state involving a single language being ended by divine intervention in the relatively recent past. Of course, this is quite contrary to the modern scientific/linguistic position that humans have had language for at least 70,000 years and that diversification (and convergence) of languages have proceeded by way of "cultural evolution" throughout this period. As one might expect, those who still accept the story as literally true are motivated by fundamentalist Jewish or Christian belief in the literal inerrancy of Genesis, i.e., they are "creationists."

More surprisingly, some of these modern believers in the Tower of Babel are trained in linguistics. There are various branches of the Summer Institute of Linguistics (SIL) located around the world. SIL trains linguists in fieldwork methods, so that they can analyze unwritten languages worldwide, develop writing systems, prepare dictionaries and grammars--and translate the Bible into each such language, for this otherwise worthy enterprise is linked with Wycliffe Bible Translators, an arm of fundamentalist Christianity! Indeed, some of its qualified linguists and instructors are creationists. An example of their work is an article by Kevin May (2001), which essentially upholds the Babel story. For a qualified linguist, May is remarkably ill-informed on historical linguistics, and his summaries of orthodox views are wildly outdated. (This is, of course, often true of scientifically trained creationists.)

At a time when Genesis was generally interpreted as historical (which was before the development of historical linguistics), it was often assumed that the single pre-Babel language was Hebrew, the language of the Pentateuch. This idea is, in fact, far from dead. One current manifestation of it is the work of the Jewish creationist writer Isaac Mozeson (see Mozeson 2000 and his Web site at www.homestead.com/edenics). Mozeson claims that virtually all the words of all languages derive from "Edenic," which is basically early Hebrew with some (Proto-)Semitic roots not attested in Hebrew itself. (Hebrew is a member of the Semitic language family.) As is very common in such cases, the main problem with Mozeson's proposal involves the methods of comparative linguistics that he adopts. These are long outdated and are now used only by fringe amateurs. The probability of pairs of superficially similar words in apparently unrelated languages having very similar or the same senses by chance is in fact much higher than Mozeson suggests. In this particular case, most of the alleged correspondences between phonemes are unsystematic and arbitrary; each correspondence is invoked as it is needed to "explain" specific forms, but there is typically no good explanation for why different correspondences apply in different cases, or even an admission that this is an issue that needs to be addressed. It has long been known that language change does not occur in this unsystematic way: there are often exceptions to a given pattern of correspondences, but these are relatively few; and, where information is available, they can generally be explained. Using the methods adopted by Mozeson and other amateurs, one can "prove" (spuriously) that almost any two languages share large amounts of vocabulary. The statistics involved here have recently been formalized by Ringe (1992) and other historical linguists, and, while there is some debate about specifics, the overall case is overwhelming.

In addition, in many of the cases cited by Mozeson, other etymologies are already known or proposed with good evidence. His theory also contradicts a large amount of well-grounded information about the "genetic" relationships of entire languages (in language families). Further, the analysis ignores the fact that "genetic" relatedness (as opposed to influential contact) always involves specific elements of grammar and phonology as well as shared vocabulary. In fact, it is clear from a range of major errors that Mozeson simply does not understand historical linguistics.