Review of The Science Of The Dogon (by Laird Scranton)

Inner Traditions, Rochester (VT, USA)
2006 xv + 208

This article first appeared in The Skeptical Intelligencer 10 (2007/2008), pp 23-25

Reproduced with permission

In this book, originally published in 2002, Scranton develops further the ideas of Griaule and Temple (see The Sirius Mystery, 1976, 2nd edition1998) about the cosmological knowledge of the Dogon tribe in Mali. Like Temple but more systematically, he argues that the conceptual and symbolic cosmological system of the Dogon is largely shared with that of ancient Egypt. He goes on to claim links with Buddhist thought and symbolism and with other non-local ideas on these matters. Still more dramatically, he holds that spoken Dogon word-forms (the Dogon are illiterate), corresponding Egyptian words, Dogon symbolism considered as a system and aspects of Egyptian hieroglyphic writing collectively demonstrate awareness of cosmological truths discovered only recently by modern science (e.g. string theory). Scranton believes that the observed similarities are too close, numerous and systematic to have arisen by chance. If he is right, the upshots are obviously major.

The upshots of Temple’s claims about the Dogon were also major (notably, extrasolar visitation in ancient times); but skeptical reviewers were able to explain the symbolic and mythological data in other less dramatic terms. One of the skeptical reviews was prepared by me in conjunction with Groves (The Skeptic [Australia] 19:4 (1999), pp 56-60; short version on Among other objections, Groves and I showed that Temple’s linguistic evidence, specifically, is much weaker than he suggests. He argues extensively for contact/shared origins on the basis of superficial and unsystematic similarities between words in Dogon, Egyptian and other languages not generally believed to be linked (notably Greek). This method is very typical indeed of work in this area by non-linguists but is the best part of 200 years out of date in terms of linguistic scholarship. All qualified/well-informed historical linguists, whatever their differences on other issues, would agree that this method can easily be shown to be wholly unreliable. The likelihood of accidental unsystematic similarity is much greater than such writers realise, and in cases where the cross-linguistic correspondences between phonemes are incorrigibly unsystematic (no specific explanation for the lack of systematicity) the likelihood that they involve common origin or contact is actually very small. I have rehearsed this argument in detail in many publications; see e.g. ‘Linguistic reconstruction and revisionist accounts of ancient history’, The Skeptical Intelligencer 7 (2005), pp 22-33.

Scranton argues in similar ways, and the same objections apply again. He also treats as authoritative earlier authors such as Higgins (pp 174-175) who made the same mistake or wrote before these matters were understood. And he compounds his problems by insisting on using Budge’s older transliterations of Egyptian words, which – as he well knows – are deemed outdated but which suit his case marginally better than those transliterations which are now accepted by Egyptologists. (His case is still impossibly weak, however, even using Budge.)

And, unlike Temple, who apparently believes for some reason that he understands historical linguistics, Scranton proceeds as if the discipline did not exist. In fact, whether or not he has a case firmly grounded in semiotic and other non-linguistic similarities, the linguistic evidence per se offers him no support at all. And, in some cases, he himself admits in so many words that his equations are speculative, e.g. where he discusses some Hebrew roots on p 25.

In fact, Scranton’s amateur view of cross-linguistic relationships between words is particularly obscure and strange. Without actually being aware of the objections that a linguist would raise to his equations, but perhaps concerned about this aspect of his case, he covers himself (p 3) by saying that when he proclaims an equation between e.g. a Dogon and an Egyptian word for (allegedly) related concepts he is not necessarily referring to ‘a strict linguistic lineage for the words’. But it is not at all obvious how any other kind of relationship might be valid. Scranton goes on to say that he regards such pairs of words as related within a larger symbolic system. Now that might sometimes be the case, and as I note below his viewpoint might be somewhat strengthened thereby; but, even if this were so in a given case, that would not absolve him of the need to explain in what other (non-‘strict’?) way the words of a pair are themselves related to each other. Neither would it absolve him of the need to demonstrate that they themselves (as opposed to their senses) really are related.

Scranton also has a somewhat strange perception of contemporary mainstream scholarly views on the written Egyptian language itself. For instance, he seems to think (correspondence) that he is being radical in regarding some hieroglyphs as logographic (read for meaning, not for sound). (He then goes on to link this idea with his own specific theories.) But the script as a whole clearly developed out of a logographic script, and Egyptian language scholars already know very well that in dynastic times some glyphs – including e.g. one referring to periods of time, upon which Scranton focuses – were still read for their meanings in various contexts.

When apprised of some of my objections, Scranton was gracious enough and talked as if he was taking them on board, but in fact he recanted nothing. He holds that his overall case is overwhelmingly strong regardless of such objections. I will comment here mainly on the linguistic material, where my own expertise lies. I do not myself find many of Scranton’s cross-cultural equations of non-linguistic items persuasive either, but I leave it to scholars of semiotics and the cosmologies of early cultures to deal authoritatively with those aspects of his work. Overall, my view is that the individual equations, of all kinds, are too weak to sustain a case, however numerous they may be. And this is certainly the case for the linguistic equations.

There are several more specific objections to Scranton’s linguistic equations:

1) Many of the pairs of words which he regards as significant are very short; this obviously increases further the chance of accidental similarity. Examples include his discussion of words beginning with te- (p 87), Dogon po and Egyptian pau- (p 93), Dogon dada and Egyptian dd (p 97), etc.

2) Many of the alleged similarities are very approximate only. Without a significant degree of systematicity across many such pairs, cases of this kind cannot be taken seriously. Eg, Egyptian hpr, or even Budge’s version khepr, and Dogon ke (which in any case have only roughly similar meanings) share only the vowel (reconstructed in the case of Egyptian; vowels are not represented in the Egyptian hieroglyphic script) and a very approximate area of oral articulation for the initial consonant (p 7).

3) In comparing linguistic forms, it is necessary to examine known etymologies and to use the forms (known or reconstructed) from the relevant period. In this case the relevant period is clearly ancient. This has the consequence that Spanish or German forms (and indeed, very probably, current Dogon forms) cannot be used in comparison. Spanish ojo and German Auge (p 154; both mean ‘eye’) must be traced back at least to their Latin and Proto-Germanic origins, respectively, and if possible further back, before any comparison with Dogon or Egyptian words can be usefully carried out. Similarities between Dogon or Egyptian and current Spanish or German words (especially if approximate only, as in the case of Egyptian aakhu and ojo/Auge; see above) cannot be of any significance.

4) In some of Scranton’s examples, it is actually known that the words in question are unconnected and the (unsystematic/approximate) similarities accidental. E.g., Spanish dios and diez are unconnected (p 25); the former is from Latin deus, the latter from Latin decem, and the Latin words have separate Indo-European sources.

5) Like most non-linguists, Scranton focuses almost entirely on word-forms (lexical phonology and spelling); he ignores not only phonological systems but also grammar. Unless the posited links between e.g. Dogon and Egyptian are supposed to involve only contact between the peoples, not ‘genetic’ relationship, these more structured aspects of the languages must be considered in assessing claims of relatedness.

If it were clear that many of the Dogon and Egyptian words are really implicated, systematically, in cosmologies that take distinctly similar forms, I grant that the linguistic aspect of Scranton’s case would thereby be somewhat strengthened (I do not think it would be decisively strengthened). But, as I have noted, it does not seem to be at all certain that this is so. As things stand, unless linguists are persuaded that these words are implicated in cosmologies with significant similarities, they will regard Scranton’s approach not merely as ‘not a strictly linguistic approach’ (as he calls it) but as an approach to correlating words which is not sound in any way.

In face of all this, a linguist’s provisional conclusion must be, as noted, that there is no worthwhile specifically linguistic evidence for a link between Dogon and Egyptian.

Scranton is publishing a second book in which he will make (and try to defend) further dramatic non-standard claims about the Egyptian language and script (including the notion that written Egyptian was somehow prior to spoken Egyptian) and the links between Egyptian and Dogon. I have corresponded with Scranton on these matters and so far I find his case unconvincing and his view of the Egyptian language and script implausible; but serious comment must await publication.