17 Mar 2007

Pokorny lives again

Yes, my friends, I'm talking about Julius Pokorny, author of the Indogermanisches Etymologisches Wörterbuch, (ie. "IndoEuropean Etymological Dictionary") and unfortunate victim of a tram hit-and-run that ended his life in 1970. However, he lives again thanks to the miracle (or curse) of the internet.

For those new to comparative linguistics, Indo-European (or "IE", for short) is a language hypothesized to have been spoken about six thousand years ago. It is the ancestor of most languages now dominating Europe and India. English too is an Indo-European language.

I have just come across yet another site that faithfully provides us with Pokorny's work, rather overconfidently named Linguistics Research Center, a site coming from the University of Texas. Together with the American Heritage Dictionary and the late Sergei Starostin's Tower of Babel, there is no shortage of access to Pokorny's reconstructions online.

This seeming wealth of information on the internet tests the limit of how outdated information can be before it is no longer true information and starts becoming insidious miseducation. In this case, Pokorny's works are outdated by at least a half-century. I have no doubt that far better books can be found with which to build an info-site, even at the University of Texas.

The changes that Indo-European linguistics has experienced since he published his book, during those tween years of the 60s, 70s, 80s and beyond, made an irreversible impact on the rise of universal acceptance of "Laryngeal Theory". Laryngeal Theory proposes that Indo-European (IE) had three "h"-like sounds (written as h1, h2 and h3) that had disappeared in most languages. Fortunately, h2 and h3 had left signs of their original presence in various words of IE languages by altering the vowels neighbouring them before they completely disappeared.

With the discovery of Hittite and other languages such as Luwian, Lycian, Lydian and Palaic, it was shown that h2 and h3 didn't entirely disappear in all IE languages, remaining "h" in this Anatolian branch of the family. Laryngeals are now an undeniable feature of the Proto-Indo-European sound inventory. To deny their existence now amid all of the extensive data showing otherwise is tantamount to denying that the earth revolves around the sun.

Yet, Pokorny never reconstructed any of his roots with these added laryngeals despite the Laryngeal Theory already taking hold in the field at the time. His antiquated work wasn't without its critics either. So today, if one has any access to a university library at all, Pokorny's reconstructions are a pale imitation of real information on the subject. They are only useful if one is interested in the history of linguistic studies, just as Plato's works are no longer informative on modern cosmology.

The following table illustrates a sample of important differences between Julius Pokorny's reconstructions and those used today:

Let's take the malformed *ant- "front" (based on words like Greek anti and Latin ante) as an example which takes all of a minute of googling to disprove thanks to Hittite hanti with preceding "h", showing us that the true form must be *h2ent- instead.

Considering that there are many more current resources available on Indo-European reconstruction (eg: Encyclopedia of Indo-European Culture by J. P. Mallory and Douglas Q. Adams), I don't think it's harsh to find it more than a little odd that an institution of higher learning like the University of Texas can't try a little harder to educate the public with information written in this century.


  1. Another e-mail surrogate comment.

    In relation to your complaints elsewhere that there's no definitive Proto-Indo-European concensus work since Pokorny, I wanted to draw your attention to this Proto-Indo-European lexicon pulled together from Mallory & Adams, Lubotsky, Beekes and Matasovic.

    It's necessarily imperfect, but will no doubt serve as a useful resource if you don't know about it already.

  2. Thanks, that link will be useful to everyone. Kudos to SGMcCabe and the immense work put into it.

    It however still contains some pointless archaicisms and saying "Oh well, it's imperfect but it will do." is a little slothful. No rest for the wicked.

    I see that the list still contains entries with long instead of iH and a pointless distinction between hₐ and h₂. The use of parentheses in a long-researched word like *dʰuǵ(hₐ)tḗr is also antiquated. It's surely *dʰuǵh₂tḗr at the very least but the implausible, unpronounceable combination of phonemes here emphasizes the points I raised about revising the traditional phonological system to finally address the fact that and its series simply cannot have been palatalized.

    There's still this obstructive obsession with maintaining tradition and ignoring facts. This is what I continue to object to. Dare to innovate.