Skeptical Humanities

Hi everybody!  I’ve been away with my fiancée in Northern Ireland and then Wales, hence the lateness of this blog-instalment!  Thanks for the continuing interest.

I’ll return later to further cases involving specific languages which have been the focus of ‘fringe’ historical linguistic attention: Basque, Hungarian, Sumerian, Zuni, etc.  If anyone is especially interested in a given language as discussed in this context, please let me know!

I turn here, however, to non-mainstream ‘epigraphic’claims involving scripts rather than languages themselves, or at any rate involving the written rather than the spoken forms of (usually ancient) languages.

It must be noted at this point that the very familiar term alphabet is restricted in technical discourse to scripts which employ (at least roughly) ‘one symbol per phoneme’.  Of course, there have been (and still are) many important non-alphabetic scripts: logographies (one symbol per morpheme/word, as in Chinese script), syllabaries (one symbol per syllable, as in Cretan Linear B or Japanese kana), abugidas (see my earlier comments on Ethiopic script), abjads (alphabets displaying only the consonants of words, as in Hebrew script), etc.

There are statistical features which can provide some indication of which script-type is represented in what appears to be a given mysterious or doubtful body of writing.  These include the sheer total number of different symbols (low for an alphabet or abjad, higher for a syllabary, still higher for a logography), the ratio of tokens to types (much higher for an alphabet with its small number of symbols than for a logography; most alphabetic letters occur many times in a given text, whereas most symbols representing entire words will obviously occur much less frequently), the typical text-length (longer for an alphabet, where several characters are required to make up each word/morpheme; systems involving only very brief texts probably represent logographies, or indeed non-scripts; see below), the complexity of symbols (individual alphabetic letters tend to be simpler, requiring fewer ‘strokes’ than logographs), etc.

Naturally, it is also important to distinguish clearly between scripts and the languages which they are used to represent; a script is not itself a language.  A given language may be written in a variety of scripts and a script may be used (perhaps with modifications) to write various languages.  The Roman alphabet used to write English, French, Malay and many other languages is an obvious example.  And the fact that, for example, Hebrew script is used mainly to write the Hebrew language and vice versa does not mean that the script is the language.  In the same way, Chinese names written in the Roman alphabet are not thereby ‘in English’.  Sometimes a language and the script which is usually used to write it are very closely associated (as in the case of Hebrew, and also that of Greek), and the script may even be especially well-suited to the language (as in the case of Chinese); but even in these cases it is always possible to write any given language in a different script (Mandarin Chinese is now often written in a modified Roman alphabet), and it is also quite possible for a script associated with one language to come to be used for another, perhaps very different language (Japanese kanji are Chinese characters).

Indeed, one script can sometimes be altogether abandoned for another by users of a given language.  The motivation for the change of script is itself often partly non-linguistic, as in the case of the Turkish move from Arabic to Roman script in the 1920s (Turkey was moving from an Islamic to a secular/‘western’ self-image and cultural framework); however, there are also important linguistic factors, involving the degree to which each relevant script can readily handle the phonological structures of the language (in the case of Turkish, the new script is actually better suited to the language than was the old).  Even if a language and its script genuinely emerged simultaneously, as is sometimes implausibly claimed for Hebrew) and as does occur in the cases of invented languages, this would/does not prevent the later adoption of a different script.

The set of claims I’m introducing here saliently involves several well-known ‘scripts’ and individual documents lacking any authoritative decipherment: the Phaistos Disk from Crete, Easter Island Rongo-Rongo, the Indus Valley Script (IVS), etc., and also the medieval ‘Voynich Manuscript’ (many decipherments have been advanced for all of these).  Some of these items, such as the Phaistos Disk, are believed by many scholars to be undecipherable, because the texts available are too short to permit analysis in the absence of a bi- or multi-lingual ‘Rosetta Stone’.  Others, again including the Phaistos Disk and also for example the tablets found at the mysterious Glozel site in France, are considered by some scholars, on linguistic and other grounds, to be modern fakes.

Some purported decipherments in this area involve systems such as Gimbutas’ ‘Old European’, some African systems, and the Panaramitee ‘rock art’ found in Aboriginal Australia, which appear in fact to represent non-linguistic semiotic systems (comparable with traffic lights and such) rather than written language (or, in the last of these three cases, simply art rather than writing of any kind).  The Glozel tablets may also be of this nature.  In fact, as I noted earlier, it has recently been suggested by some mainstream linguists that IVS too, with its very short texts, is non-linguistic in nature.  (This is not to say that such systems, if genuine, are not interesting.)

Another sub-set of this group of cases involves claims to the effect that currently accepted decipherments are incorrect.  For instance, some writers reject the standard decipherment of Egyptian hieroglyphs; such authors (notably, for Egyptian, some Latter-Day Saints) often offer rival decipherments.  Some of the non-standard claims regarding Hebrew involve the view that Biblical Hebrew was grossly misinterpreted by the Masoretes, the scholars who developed the originally purely consonantal Hebrew abjad so as to display the vowels.

The best known ‘fringe epigraphist’ is the late Barry Fell, who was also the most prominent recent exponent of the hyper-diffusionist claim that transatlantic and/or transpacific voyages brought representatives of many cultures to the Americas before the now firmly established Norse settlements of around ten centuries ago.  Working extensively from allegedly inscriptional material, Fell and his many and varied associates identify the cultural and linguistic influence of Chinese groups, Japanese-speakers, Indians, ‘Celts’ from Ireland or Wales, Egyptians, Phoenicians, Hebrews, etc., etc., and also that of Africans (some ‘Afrocentrists’ are involved here, notably Ivan Van Sertima).  There are similar claims for Australasia and other regions.

In some cases, these claims involve the contentious decipherment of what are alleged to be inscriptions (on rocks etc.) as being in known scripts, adopted or adapted to write local languages which are generally deemed to have been unwritten until modern times.  Other claims in this general area involve ‘out-of-place’ written languages in essentially familiar scripts, for instance the Yarmouth Runestone from Nova Scotia (interpreted as written in a variety of languages including Hungarian), the Kensington Runestone from Minnesota (allegedly written in runic Scandinavian in medieval times, but more probably a 19^th Century forgery), and the inscriptional Chinese, Mongolian, Malayalam etc. allegedly found in various unexpected locations as reported by Gavin Menzies.

However, many of the alleged inscriptions appear to be natural markings, plough-marks, etc.  Others (for instance some Egyptian hieroglyphic material ‘found’ in Australia) appear to be recent forgeries (often displaying learner errors).  Even where an inscription appears genuine, the specific scripts and linguistic forms identified are typically so unfamiliar that their provenance is partly a matter of speculative reconstruction.

It will be seen from the above that most claims of all these types are highly dubious, to say the least.  Those without the relevant expertise should probably accept the negative linguistic consensus on the non-standard ideas about each such case.

I will comment on some specific cases of this kind next time.

Mark