A nice NY Times story by Margalit Fox on an unsung hero of historical linguistics (even though she wasn’t a linguist):
Alice Elizabeth Kober was born in Manhattan on Dec. 23, 1906, the daughter of recent immigrants from Hungary. A brilliant student, she earned a bachelor’s degree in classics from Hunter College, and it was there, in a course on early Greek life, that she appears to have encountered Linear B.
Enthralled — and already confident of her own blazing intellect — she announced on her graduation that she would one day decipher the script. She came within a hair’s breadth of doing so before her own untimely death, at 43, just two years before Mr. Ventris cracked the code. […]
It was Dr. Kober who cataloged every word and every character of Linear B on homemade index cards, cut painstakingly by hand from whatever she could find. (During World War II and afterward, paper was scarce, and she scissored her ersatz cards — 180,000 of them — from old greeting cards, church circulars and checkout slips she discreetly pinched from the Brooklyn College library.)
On her cards, she noted statistics about every character of the script — its frequency at the beginnings and ends of words, and its relation to every other character — with the meticulousness of a cryptographer. Sorting the cards night after night, Dr. Kober homed in on patterns of symbols that illuminated the structure of the words on the tablets. For as she, more than any other investigator, understood, it was internal evidence — the repeated configurations of characters that lay hidden within the inscriptions themselves — that would furnish the key to decipherment.
A life well spent, if you ask me.
It’s a pity she died before the decipherment was complete, in no large part because she tends to get robbed of the appropriate credit.
On another note, Ventris himself died at an even younger age just before his work on the decipherment was published.
They both worked obsessively on an ancient script and died unexpectedly young, in separate incidents, before the meaning of the code was publicly revealed? And we still don’t know Linear A is telling us… I think Dan Brown should be told about this.
If I remember correctly Ventris had announced that they’d deciphered Linear B before he died; he just hadn’t published all his findings.
With Linear A, the odds are good that most of the characters have the same sound values, but nobody can figure out what language they’re actually encoding. It could be an isolate, or it might become clearer if enough texts are discovered. My understanding is that there is much more Linear B material than Linear A.
I once designed an introductory linguistics course with Chadwick’s book on the decipherment as one of the textbooks. It was a “unit mastery” course taught for a number of years, and to make it work, many, many test questions had to be devised. Of the questions I wrote, the one that most pleased me was based on part of the Kober/Ventris method of deducing “links” among the signs of the the script based on the conjectures that the script was syllabic and that Linear B was a case language. (The students mostly didn’t know about cases or declensions.)
The question was to deduce the pronunciations of 4 words from one word with a known pronunciation and the information that the 4 words were two case forms of two nouns — the same two cases. None of the syllabic signs of the word with known pronunciation recurred in the other 4 words.
The students had to reason that when they compared the two case forms of the same noun, the first different signs, working from left to right, must contain the same consonant, from the end of the stem, but different vowels, at the beginning of the case endings. Similarly, comparing the same case form of two different nouns told them that two syllabic signs were linked by having the same vowel, from the case ending, but different consonants, from the ends of the two different noun stems.
Putting all that information together made it possible to determine the phonetic values of all the syllabic signs.
Just varying the symbols for the syllabic signs and the phonetics made it easy to turn this into many variants of the same basic test question.
It’s a well-deserved tribute to Kober. But Fox is overstating her claim to be bringing to light a woman whose contributions “have been all but lost to time”. I know the decipherment story from two sources: The Decipherment of Linear B by John Chadwick, Ventris’s collaborator, first published in 1958; and part of the chapter on decipherment in Simon Singh’s The Code Book, 1999. I don’t have either of them to hand, but searches on Amazon and Google Books confirm my recollection that Kober is a prominent part of the story in both. Anyone at all familiar with the decipherment will be aware of the importance of “Kober’s triplets”. And she’s not consigned to a bit part, either: Chadwick, for example, says “I do not think there can be any doubt that Miss Kober would have taken a leading part in events of later years, had she been spared; she alone of the earlier investigators was pursuing the track which led Ventris ultimately to the solution of the problem.”
Chadwick
Singh
Unless she has found some additional contribution of Kober’s, unrecorded in other sources, I find it hard to swallow Fox’s various claims that she’s brought to light an unsung heroine, or – given that Chadwick, the writer of those generous words above, was English – the business about “a British masculine triumphal narrative”.
In other New York press coverage of linguists, today’s Post has a feature (not delayed six decades . . .) on “The word nerd: Meet Hollywood’s go-to guy for made-up languages.” http://www.nypost.com/p/entertainment/tv/the_word_nerd_bzA9b1kw2DPwhCPfTXZbeP It isn’t that bad as tabloid coverage of conlang stuff goes . . . And I guess it’s sort of heartwarming that someone who as a kid was bothered by the low quality of the Klingon dialogue in various Star Trek projects grew up to make a living doing similar work but to a higher standard.
Modest, self-effacing John Chadwick probably deserves more credit for the decipherment than he allowed himself to claim.
Speaking of made-up languages, this is an interesting puzzle.
“Unsung” often means “not the answer to the quiz question”. If the question is “who deciphered Linear B?” then the answer is Ventris, and Kober is unsung. Anyone who learns more about it will have heard of her, but I suspect that quite a few people know only the quiz answer. I can make myself seem marvellously knowledgeable by citing the only fact I know about a topic to someone who knows nothing at all about it.
Jonathan Lopez in the Wall Street Journal has a rather harsh review of the book on which Fox’s column is based (The Riddle of the Labyrinth) and like Breffni above questions exactly how “unsung” Kober was during her lifetime.
Thanks Ben Z.
That’s too bad. You can’t really blame an author for exaggerating the importance of her subject, but it sounds like a sloppy book. (Insert rant on the lack of decent editing these days.)
From the publisher’s description: “In the tradition of Simon Winchester and Dava Sobel, The Riddle of the Labyrinth: The Quest to Crack an Ancient Code tells one of the most intriguing stories in the history of language, masterfully blending history, linguistics, and cryptology with an elegantly wrought narrative.”
I also saw, somewhere, a blurb by Winchester for the book. I am determined not to let this prejudice me against the book.
Jonathon Lopez’ article says “Some of the finest minds in linguistics, classics, archaeology and other disciplines vied to crack the code …” So far as linguistics goes, this is over-the-top, isn’t it? When I first read Chadwick’s book, as a linguistics student sometime in the late 60s, I loved the book, however I had never heard of any of the protagonists before, and subsequently in linguistics, never heard of them them afterwards, either, apart from the Linear B story.
LH, I don’t think the issue is that Fox exaggerates Kober’s importance. Even if the new book doesn’t turn up some previously unknown aspect of her work, I’d be receptive to an argument that Kober’s contribution deserves re-evaluation. Maybe she should even share the honours equally with Ventris.
The problem is Fox’s self-aggrandising claim that she has rediscovered an American heroine who, it is implied, was airbrushed out of history by British chauvinists. Singh, Andrew Robinson (author of The Man who Deciphered Linear B) and the ghost of John Chadwick would be entitled to feel a bit hard done by.
“If I remember correctly Ventris had announced that they’d deciphered Linear B before he died; he just hadn’t published all his findings.”
Ventris and Chadwick had published an article in the Journal of Hellenic Studies (“Evidence . . . “) outlining their theory that Linear B was Greek.
http://www.jstor.org/discover/10.2307/628239?uid=3739256&uid=2&uid=4&sid=21102297546627
As the article notes, they were going against the grain. Kober herself was convinced that that Linear B was not Greek at her untimely death, and you have to wonder whether she really would have anticipated Ventris and Chadwick in deciphering Linear B had she lived. Initially, many scholars were very skeptical, which is understandable, since the Linear B syllabary is a very crude representation of Greek–it was obviously based on a script (probably Linear A) developed for a language with a radically different phonology. In fact, skepticism didn’t die out completely, I believe, until the 1970s or even later, and many interpretations are still controversial.
[Linear B] was obviously based on a script (probably Linear A) developed for a language with a radically different phonology
Actually there’s a modern counterexample to that. Buginese writing looks like it was designed for a language with a simple CV phonology (with some prenasalized consonants). But its syllables in fact can have final consonants, namely /ŋ/, /ʔ/, and gemination of the following initial consonant (as in Japanese), they just aren’t written. What is more, Buginese is part of the Indic script family, most members of which use a virama diacritic (and often a set of ligatures) to represent a consonant without a vowel. But attempts to introduce a virama into Buginese writing in modern times have been rejected by the script’s users.
The apparently severe constraints on what consonants can close syllables in Buginese makes that writing system a dubious counterexample to the general idea that syllabic scripts develop naturally only for languages with open syllables. Maybe the makers of the Buginese script thought that the stopping and nasality they heard at the ends of syllables were really attributes of the syllables’ vowels (and maybe they were right).
“Actually there’s a modern counterexample to that. Buginese writing looks like it was designed for a language with a simple CV phonology (with some prenasalized consonants). But its syllables in fact can have final consonants,”
It’s not just the failure to represent syllable-final consonants that makes Linear B ill-adapted to Greek. The script has only one series of symbols for each class of obstruents (labial, coronal and velar), where there are three series of obstruents in Greek (voiceless, voiced and aspirated). For example, /p/, /b/ and /ph/ are all represented by a single series of CV symbols. /l/ and /r/ are represented by a single series of CV symbols. CVi (where Vi is a diphthong in -i) are not distinguished from CV, etc. And we have reason to believe that these distinctions must have existed in Mycenaean because they demonstrably existed in PIE and in classical Greek. (I suppose there’s the outside possibility that Mycenaean was a dialect of Greek that didn’t survive the Dark Age after 1100, in which a large number of radical mergers had occurred, but I don’t think anyone has maintained that theory.)
I suspect that once it was realized that the non-logographic symbols of Linear B were a syllabary, it was the paucity of symbols that led Kober and others to resist the identification of the script as Greek.
It does seem likely that Linear B was based somehow on Linear A (a few symbols are shared), which so far has resisted interpretation, and probably isn’t Greek (but they said that about Linear B, too, didn’t they?).
The problem is Fox’s self-aggrandising claim that she has rediscovered an American heroine who, it is implied, was airbrushed out of history by British chauvinists.
I think that’s an unfair reading of her “a British masculine triumphal narrative.” The decipherment of Linear B has always, and rightly, been considered a British triumph (insofar as the national origin of science has any bearing, which shouldn’t be too far), and I don’t think she’s saying “those damn Brits stole our achievement!” I think her emphasis is on the “masculine triumphal narrative,” which (like the comparable Everest narrative) emphasizes the glorious achievement of a brave, lonely man going where no man has gone before, and downplays the “it takes a village” aspect of any such achievement. In any event, I don’t think one phrase in an op-ed piece should be a reason to reject an entire book.
I can see that reading now that you point it out. But then it’s still “those damn Brits with their masculine triumph narratives”, which doesn’t seem any fairer.
I’m not rejecting the whole book – I haven’t read it, and certainly would if it came my way. I’m making one specific criticism, which is not based on one sentence, it’s based on the whole thrust of NYT piece. If the Lopez review that Ben Zimmer linked to is anything to go by, the NYT piece is representative of the book in that regard – witness this quotation: “the story I knew–the only story anyone knew–was incomplete. A major actor in the drama was missing: an American woman named Alice Elizabeth Kober.” Kober is far from being missing from Chadwick’s 1958 book, and even in the unlikely event that Fox hadn’t read it, it’s plain wrong to claim that a Koberless version of the decipherment was “the only story anyone knew”.
My view remains that – barring new evidence in the book – Fox is overstating the extent to which Kober has been neglected, and that’s unfair to previous authors who’ve told the decipherment story, or certainly to the ones I’ve read.
Bill W: It does seem likely that Linear B was based somehow on Linear A (a few symbols are shared), which so far has resisted interpretation, and probably isn’t Greek (but they said that about Linear B, too, didn’t they?).
Until the same type of inscriptions were found not just in Crete but at Pylos in Greece, there was no reason to think that the language of Linear B, found in Crete, a place with a quite different culture from Greece, would be Greek. Ventris conjectured Etruscan, perhaps because of the resemblances in Cretan and Etruscan art. The idea that the language of Linear B was Greek did not arise until some of the words were identified as Greek (such as with the pictures of pots with different numbers of handles). The Cretan and Pylos Linear B inscriptions date from after the fall of the independent Cretan civilization and before the fall of the Mycenean civilization which had taken over Crete. I don’t think there is any reason to think that early Greece had much influence on the early Cretan culture (the opposite seems to be true: eg the story of Zeus born or at least raised on Mount Ida, in Crete, and other mythological and religious links).
Linear A presumably dates from the period of the original, pre-catastrophic Cretan culture. Some time after the Mycenean takeover, bilingual Cretan survivors more or less familiar with the A script must have adapted it – or what they remembered of it – for the needs of the Greek-speaking Mycenean administration, hence the B script. The fact that the script is poorly suited to Greek, merging together different Greek phonemes (eg using only one character for three different Greek consonants such as k/g/kh) suggests that the Cretan adapters did not perceive the difference between the Greek consonants as significant because their own language had only one phoneme where Greek had three. But perhaps the type of records they were keeping with this script did not require as much precision as whole texts dealing with varied matters would have done, so in subsequent generations of now Greek-speaking scribes the script may have been kept intact because it was a kind of shorthand, not needing improvements for the sake of precision.
a dubious counterexample to the general idea that syllabic scripts develop naturally only for languages with open syllables
Well, depending on what you mean by “naturally”, I don’t think it’s taken by anyone (certainly not me) to be a counterexample. I don’t know whether Buginese is a direct adaptation of some previous Indic script or the result of stimulus diffusion from Indic scripts, but it certainly did not appear ex nihilo. I merely meant that if Buginese were a dead and forgotten language (it is still living and written quite comfortably, though still without glottal stops, in the Latin script), we’d incorrectly infer that it was a CV language.
There are many other cases where ease of writing or tradition has predominated over ease of reading without using context cues. If Japanese were a dead language, and we knew only the early form of the kana syllabaries before the standardization of the half-voiced and voiced consonant marks, we probably would reconstruct a language with fewer consonant distinctions than we know older Japanese to have had.
Until almost within living memory, Turkish, with eight phonemic vowels, was written using the Arabic script, which can only handle three. Vowel harmony helped there, but not enough to eliminate ambiguity altogether, which was one reason that so much of written Ottoman Turkish is really Perso-Arabic with the occasional Turkish word thrown in for local color. (I exaggerate, but not much.)
Similarly, Old Manchu was for a whole generation written in a highly degenerate (in the technical sense) and ambiguous descendant of Aramaic, until disambiguating dots and circles were added to the letters to make it a true alphabet.
But perhaps the type of records they were keeping with this script did not require as much precision as whole texts dealing with varied matters would have done, so in subsequent generations of now Greek-speaking scribes the script may have been kept intact because it was a kind of shorthand, not needing improvements for the sake of precision.
M-L: Yes, I think that’s probably exactly right. There was no need for a very accurate representation of Greek because the script was being used mainly for inventories. But I think that when the CV grid was being worked out, but before the successful decipherment, the paucity of identifiable phonemes might have misled Kober, among others, into doubting that Linear B could have been Greek. If you were to make a grid of classical Greek CV combinations, you would come up with many more boxes. You would expect, for example, open and closed e and o phonemes, in addition to a, i and u, and many more consonant phonemes.
Also, as I mentioned before, I suspect that Chadwick was the real unsung hero–or rather, another unsung hero if you feel that Kober wasn’t given her due as the scholar who laid the basis, and established the correct methodology, for the decipherment. While Ventris had the original flash of insight, I doubt that even with a solid background in classical Greek, he could have navigated through the archaic form of Greek that Mycenaean represents, to arrive at a coherent analysis, without Chadwick’s collaboration. In addition to his background in Greek linguistics, Chadwick already had substantial experience in WWII as a codebreaker. But Chadwick tended to downplay his role in favor of Ventris.
Ill-adapted to Greek as it was, Linear B was good enough for government work.
I just finished the Fox book. Perhaps there is some exaggeration of Kober’s lack of recognition, but not really much, and I think she does make a case that Kober’s substantial contribution–in the face of many difficulties–should be given more than just a passing mention.
After reading the book, I think the Lopez WSJ review is quite unfair–to the point where I wonder whether he really read it or just traded on information that it somehow made sweeping accusations about Kober’s contribution being deliberately downplayed just because she was a woman, which it doesn’t–as if it were somehow a radical feminist tirade, which it isn’t. Fox does document difficulties Kober faced as a woman in an academic profession in the 1940s, though–difficulties which not only she but also a male colleague recognized at the time.
The Fox book seems quite accurate (based on my having read Chadwick’s Decipherment), giving credit to both Kober for what she accomplished, and to Ventris not just for his flash of insight, but also for the rigor with which he pursued the decipherment, building on Kober’s methodology. Fox tells us more of the rather sad personal histories of both Kober and Ventris than Chadwick’s succinct and elegant classic does, but he was writing while the story was still just a few years old and the primary figures were not long dead. All in all, I think Fox’s book is a sober and solid work of popular non-fiction on a scholarly subject.
Thanks, it’s nice to get a response from someone who’s actually read the book, which now sounds definitely worth reading.
test