Iconicity.

Helen DeWitt at paperpools posts a quote from Pharaoh’s Land and Beyond, ed. Pearce Paul Creasman and Richard H. Wilkinson (Oxford University Press, 2017) posted in turn by Rolf Degen (@DegenRolf) on Twitter (where he “performed various arcane manipulations to come up with a quotation that blithely bypasses the 140-character limit”). I can’t copy and paste from the image at her site, so I’ll post as much as I can be bothered to get by trawling Google Books and hope you’re intrigued enough to click the link for more:

The high iconicity of the hieroglyphic script again played a key role on the stage of intellectual history at another crossroad of Egyptian and Levantine cultures. […] Unlike the case of the Minoan and Anatolian scripts, the invention of the alphabet was not born in the environment of erudite scribes, but was apparently created as a non-institutional cultural product by illiterate Canaanite miners. Though they were experts in their professional field of mining, the inventors of the alphabet were far removed from the circles of professional writing in cuneiform and Egyptian. It is precisely this naïveté that allowed them to invent something completely new, as they were unencumbered by the scripts of their day. […] Like the inventors of the Cretan and Anatolian hieroglyphs, the Canaanites borrowed the Egyptian idea of turning pictures into script. Yet, not being professional scribes and not working in the service of any official ideology or institution, they did not bother to invent a whole set of new icons. They adopted roughly two dozen icons from the hieroglyphs around them […]

I like the emphasis on non-institutional and non-professional inventors. (DeWitt says: “If you are not following @Rolfdegen on Twitter, you should, and if you are not on Twitter you could do worse than sign up and follow only the incomparable @DegenRolf.”)

Comments

  1. They must have had a scribe helping them, because they selected about two dozen sounds that they thought would be sufficient to represent their Semitic language and represented each by the corresponding hieroglyphic symbol. There must be an amazing untold story here. Some humble genius who independently invented the concept of alphabetic writing, someone bilingual and someone literate in hieroglyphics–presumably the person who kept the accounts for the mine. All these people thrown together at a remote mining site.

    As soon as the earth-shattering concept is developed, people are so delighted at the new invention that they start in chiseling inscriptions on the rocks.

  2. Trond Engen says

    (Not opening the link today, just throwing out what I’ve been thinking about this.)

    The Egyptians already had invented the system of quasi-alphabetic representation of foreign names and concepts. The leaps made by the Canaanites were 1) to settle on a discrete set of glyphs corresponding roughly to the consonant phonemes of their language, and 2) to ditch all logograms and use the alphabet (abjad) to represent free text. 1) is a straightforward adaptation for quick instruction of non-native accountants 2) is what happens when the need to record (or write accompanying documents) exceeds the simple accounting the system was devised for.

  3. Right, but like the egg of Columbus, it only seems straightforward and obvious after it’s been discovered.

  4. Trond Engen says

    Oh, sure. No little achievement, but conceivable.

  5. David Marjanović says

    to ditch all logograms

    And all two-consonant and three-consonant signs.

  6. Trond Engen says

    Yes. I meant to say that I think of all the ditching as part of leap 1.

  7. The above quotation is by Orly Goldwasser, who as far as I know (or she knows either) is the only scholar to hold this view. Here’s her monograph, which I have not yet read, though I have read a few squibs of hers. She believes that Canaanite-speakers chose hieroglyphics without knowledge of their meaning in Egyptian based solely on their shapes in accordance with the acrophonic principle, much as Sequoyah did (except he did not have pictures to work with, and so there was no acrophony).

  8. It seems somehow wrong, however technically correct, to label these hypothetical miners “illiterate” when if this account is correct they literally invented the alphabet. A sort of inverted “pleading for mercy after murdering your parents on the grounds that you’re an orphan” situation.

    (Has there ever been a more appropriate use of “literally” than in the above paragraph? I don’t think there has.)

  9. It’s a stretch in my opinion to assume that because the oldest surviving letters are found on rocks at a mine, that the miners invented the alphabet. Scratches in rock protected from the elements must be the most durable of human markings. I don’t believe Chauvet, Altamira and Lascaux are the origins of art either.

  10. Joe in Australia says

    As Trond Engen says, the quasi-alphabetic use of hieroglyphic logograms to spell out names was already established; Sumerian scribes had a similar technique. Consequently, I think the most parsimonious explanation is that scribes who had to regularly write foreign names standardised on a particular set of quasi-alphabetic signs. If you were a sergeant (e.g.) administering a bunch of foreigners you would quickly realise the benefit of learning those signs yourself so that you could mark off their hours worked without calling for a scribe. Eventually one of these sergeants (who was probably foreign himself) would have needed to send a message when no scribe was around, and written it using the quasi-alphabetic signs that then became the alphabet. And here our troubles began.

  11. David Marjanović says

    (Has there ever been a more appropriate use of “literally” than in the above paragraph? I don’t think there has.)

    *applause*

  12. Trond Engen says

    ryan: It’s a stretch in my opinion to assume that because the oldest surviving letters are found on rocks at a mine, that the miners invented the alphabet.

    That’s an important point. We never have the first of anything. Obviously, Canaanite tribes must have been heavily involved in trade with and between the Egyptians and the Levant, so it’s also reasonable to think that a handy system for transmitting orders and contracts in Semitic would spread fast. But until there’s evidence that the script was already widespread, it’s reasonable to think that the durable texts occurred in the same intellectual environment that used the script for more secular purposes.

  13. Joe in Australia’s scenario makes sense too.

  14. Trond Engen says

    I really should go back to opening links and reading before weighing in. It’s occuring to me that there are some downsides to following LH from my phone while waiting for family members outside shops.

  15. Trond Engen says

    Joe in Australia’s scenario makes sense too.

    Yes. It’s more or less the same scenario with different participants, Egyptian border officials dealing with Canaanites or Canaanite traders or miners dealing with Egyptians.

  16. Embarrassingly, I see that I need a copyeditor for my blog – I originally misremembered Rolf Degen’s Twitter handle and corrected it, but only once, having referred to it twice. (It is @DegenRolf.)

  17. SFReader says

    Life is too short to spend it on remembering Twitter handles. We could spend the time studying dialects of Yemen instead

  18. Hear, hear!

  19. “As Trond Engen says, the quasi-alphabetic use of hieroglyphic logograms to spell out names was already established; Sumerian scribes had a similar technique. ”

    There is something similar in Chinese, a set of characters normally used for these transliterations. They are usually not hard to recognize because when they appear as transliterations, they make no sense read as normal words.

    “The Egyptians already had invented the system of quasi-alphabetic representation of foreign names and concepts. The leaps made by the Canaanites were 1) to settle on a discrete set of glyphs corresponding roughly to the consonant phonemes of their language, and 2) to ditch all logograms and use the alphabet (abjad) to represent free text.”

    This is essentially what happened in the invention of kana. It didn’t progress 9or degrade) to an alphabet because there wasn’t any real need for that next step.

  20. Trond Engen says

    I’ve finally read Orly Goldwasser’s monograph. It reminds me of the fact that the alphabetic signs were given Semitic names and phonetic values independent of Egyptian, which is why I have imagined the alphabetic system to have been developed in or for a Semitic language. What I didn’t know is that signs are derived from misinterpretations of glyphs and from function elements with no phonetic value. Goldwasser interprets this as the result of invention by illiterates, but I think it could equally well be the result of consolidation and gradual nativization of a system used for some time on papyri and sheep hides by Canaaneic caravan traders. And by Canaaneic merchant sailors as well. She does not provide a convincing land-based mechanism for the transition to a North Canaaneic or “Phoenician” alphabet.

  21. David Marjanović says

    I think the most parsimonious explanation is that scribes who had to regularly write foreign names standardised on a particular set of quasi-alphabetic signs.

    But why would they use signs in their Semitic acrophonic sound values, instead of their own single-consonant signs? The consonant inventories of the languages involved were clearly similar enough for this to work.

  22. Trond Engen says

    Could it have been invented by Semitic speakers as a code? I can see three uses for a code: secretly for candid communication, semi-transparently for games, and openly as property labels on goods. They aren’t completely mutually exclusive, though I can’t see a path back to more secrecy.

  23. Stu Clayton says

    They aren’t completely mutually exclusive, though I can’t see a path back to more secrecy.

    Trond, what does that mean ? What might mutual exclusivity, whether complete or not, have to do with “a path back” ? The word “though” here absolutely confounds me.

    Perhaps you’re saying that a code, once exposed, can’t be used any longer for candid communication. Or “semi-transparently for games” (another expression that confounds me). But you can remove property labels from goods, preparatory to stealing those goods.

  24. Trond Engen says

    Yes. Muddy thoughts vaguely expressed.

    I set out to say that the three ways to use a code are mutually exclusive, but rewrote when I realised that they aren’t. Those in on a secret code can also play games or label goods with it within that closed community. What you can’t do is use something that was invented for labelling goods and understood by the wider community as a secret code.

    And a secret code can be used by a wider and wider closed circle until everybody’s part of it.

  25. Lars (the original one) says

    “A secret can be kept by three people, as long as two of them are dead.” (Steve Brust is my source (“an old Serioli saying”) but he probably stole it from somebody else).

  26. Lars (the original one): “Three may keep a Secret, if two of them are dead,” is from the 1735 edition of Poor Richard’s Almanack by Benjamin Franklin.

  27. David Marjanović says

    Could it have been invented by Semitic speakers as a code?

    Its existence wasn’t kept secret at all. Some of the oldest known inscriptions are on votive statuettes in the Egyptian temple next to the mines.

  28. Someone posted a link to new Militarev paper which deals with this exact subject.

    Basically Jewish bondage in Egypt is based on true history, there was a large Semitic diaspora in New Kingdom Egypt (some of them later became Jews), special signs for Semitic sounds in Egyptian script were invented by Semitic ethnic minority in Egypt who were significantly over-represented in the scribal profession.

  29. I don’t think anyone disputes that there were Semitic peoples in Egypt, but getting from “Semitic peoples” to “Israelites/Jews” is a leap.

    I would think that Egypt would want Semitic scribes to help communicate with Shasu wanderers, Canaanite subjects, Akkadian, Phoenician, and Ugarit rulers, and merchants from those regions.

  30. David Marjanović says

    What sounds were there that occurred in Semitic but not in Egyptian? Goldwasser mentions one example in passing, namely that the ancestor of S/Σ/ש actually stood for [θ] and is not based on a hieroglyph, but depicts a bow (in the shape that “Semites” are depicted as carrying in Egyptian paintings).

    Akkadian, Phoenician, and Ugarit rulers

    The diplomatic correspondence with all of these (and more) was in Akkadian at least in the Middle and Late Bronze Age.

  31. Lars Mathiesen says

    So how big was the difference actually between Akkadian, Assyrian, Punic and Ugaritic at that time, say 2000 BC? Egyptian and Semitic had diverged a lot by that time as I understand it, but what is ‘a lot’? Albanian vs Germanic level, or would it be possible for a layman to tell that Egyptian and Akkadian were related?

  32. Egyptian and Semitic languages are very distantly related, much more than, say, Hindi and English.

    There are a few words which layman could reasonably guess were related (for example, “water” or “die”), but overwhelming majority of words are so different that it wouldn’t be possible to suppose their common origin.

  33. David Eddyshaw says

    The ancestors of Punic and Ugaritic would have been extremely similar in 2000 BC, almost certainly to the degree of being mutually comprehensible dialects of the same language; Akkadian would have been easily seen to be related by speakers, with much core vocabulary immediately recognisable, but not mutually comprehensible with the ancestor of Canaanite/Aramaic/Ugaritic on account of the Babylonians’ tendency to talk backwards, leave out half the consonants, muddle up tenses and spout lots of Sumerian loanwords just to show off. It wouldn’t have been hard for a Canaanite (say) to pick up the spoken language, though. Perhaps something on the level of German vs Swedish (if German were written in cuneiform.)

    As SF says, the relationship with Egyptian would not have been evident, unless some genius had invented the comparative method four thousand years early. However, a Proto-Greenberg might have noticed that Egyptian was much closer typologically to Semitic than Sumerian or Hurrian were, and much more similar in phonology and morphology, sharing the tendency to reshuffle vowels at will in the verbal system, for example.

  34. Lars Mathiesen says

    OK, so the Egyptians would need Semitic scribes to send letters in Akkadian. I was not sure how big the divergence was in the second millennium. (I think we were talking about the chronology needed for some elements of what became the israelite people to have come from there).

    But would the Assyrians need scribes to read them? Would the Carthaginians? Would an Akkadian scribe be able to read Punic? (I think those scribes would probably be native speakers of Assyrian by that time, but that’s still East vs Central Semitic).

  35. David Eddyshaw says

    In fact, Canaanites wrote letters in sorta-Akkadian that they seem to have actually pronounced in Canaanite.

    https://en.wikipedia.org/wiki/Amarna_letters

    (This is a good bit later than 2000 BC, mind.)

    By the time there’s such a thing as Punic (first millennium), the Babylonians and Assyrians were speaking Aramaic, which had taken over as the Middle-Eastern lingua franca.

  36. David Eddyshaw says

    There’s a lot out there on the language of the Amarna letters.

    https://tspace.library.utoronto.ca/bitstream/1807/88977/1/Baranowski_Krzysztof_J_2014_PhD_thesis.pdf

    I came across a paper making a good case that the language is not so much bad Akkadian written by incompetents as perfectly good Canaanite dressed up as Akkadian; can’t find it just at present, unfortunately.

  37. David Marjanović says

    Perhaps something on the level of German vs Swedish (if German were written in cuneiform.)

    If you know German, you can reportedly learn Swedish in six weeks; but when you first hear it, you still understand next to nothing. Reading is a lot easier, but the divergent vocabulary and a few of the sound changes still trip you up.

  38. Stu Clayton says

    Would Swedish be easier for a German to learn if it too were written in cuneiform ? Asking for a friend who doesn’t know whether cuneiform is phonetic.

  39. Lars Mathiesen says

    If I’m reliably informed one half of Assyrian cuneiform is phonetic in Assyrian, the second is phonetic in Sumerian, and the third is logographic in Sumerian.

    And then “they” used it to spell Hittite.

  40. David Eddyshaw says

    Akkadian was actually mostly written phonetically, though the use of Sumerograms (Sumerian words to be read as Akkadian, as opposed to actual Sumerian loanwords) varies by genre. It’s just not a very good phonetic system. Unless you’re a professional scribe, of course, for whom it fulfils its purpose admirably.

    Hittite not only uses Sumerian words to be read as Hittite but Akkadian words to be read as Hittite. Got to safeguard the position of the Scribes’ Guild somehow …

    The fine people of Ugarit basically said “to hell with this” and invented a cuneiform alphabet. It’s clear they were doomed from the start.

    Hurrian is apparently written pretty much phonetically throughout, which is actually a problem for modern readers/decipherers (much as Japanese written in kana throughout is, counterintuitively, quite hard to read compared to the usual system.)

  41. Stu Clayton says

    With all this confusion, it would be a miracle that these people got anything done. *Did* they get anything done, apart from composing mysterious flowery missives to each other ? Did they have the notion of “scam” ?

  42. Large portion of the Book of Genesis is devoted to successful scams of Joseph in Egypt.

  43. David Eddyshaw says

    The text suggests that he was more scammed against than scamming. Jacob, now …

    It is now thought that scamming is the only truly indispensable part of Universal Grammar.

  44. The Ancient Afroasiatic essentialist explanations summarize the state of our knowledge as a typically Semitic piece of epigrammatic wisdom literature. In particular, the contributions known to be by Huehnergard and Häberl stand out among the obscure and even pseudonymous authors and the heterodox views of the rest.

    I can’t resist including this entry by Danny Wier from the modern AA section: “Al-Arabiyya al-essentialliyya al-lanqu’aj al-moor as-similar al-Hebruwwa adh-dhan al-Inqlishiyya.” Though formally a mere sequence of nouns, it expresses its (trivial) point more clearly than a thousand verses of the purest Classical Arabic poetry.

    Wikipedia gives a table of consonant correspondences between the various branches of AA, showing just how much restructuring went on in Egyptian compared to the relatively small steps between PAA and PS. Two points: PAA is still very controversial and may well be biased toward PS (Teeters’s law), and the Egyptian column really should have asterisks like the rest, since the phonology of Egyptian is just as much a reconstruction as the proto-languages next to it.

  45. Athel Cornish-Bowden says

    If you know German, you can reportedly learn Swedish in six weeks; but when you first hear it, you still understand next to nothing.

    If you know Spanish you can read Portuguese almost without effort the first time you try, and understand 90% of it. The hundredth time you hear it you still understand next to nothing. That’s in Portugal. In Brazil you can get a fair idea of what’s being said once you get used to it.

  46. If you know Swedish, you can read Danish and understand close to 100% of it. But when you hear it, you maybe understand 50% (depending on how Danish the Danish speaker is. The Danishest Danes will be 100% unintelligible, sometimes even to other Danes.)

  47. David Eddyshaw says

    Wikipedia gives a table of consonant correspondences between the various branches of AA

    The Wikipedia table is considerably at variance with some accounts (like Loprieno’s) I’ve seen of the development of the Egyptian system from PAA, which (for example) include the on-the-face-of-it unlikely but apparently robust change d -> ʕ (so that the Egyptian word for “hand”, written with this consonant, is actually the same etymon as Semitic *yadu), and a complete merger of the “emphatic” series of consonants with the plain series.

    Kind of proves your point, of course.

    Unequivocal Chadic cognates of Semitic and Egyptian words are actually very few in total; I don’t think the material exists in reality to draw up neat tables of correspondences. The existing dictionaries of PAA are both textbook examples of How Not to Do It with things like appeals to wholly unmotivated semantic drift and cherry-picking from single languages within subgroups; in a way, their unsatisfactory nature is hardly surprising given the mind-numbing time depth, in which a single very coherent subgroup (Semitic) is over five thousand years old at least, and the proto-language for that branch alone was already then as different from the oldest-attested member of the family (Egyptian) as Welsh is from Urdu.

    There is a remarkable degree of similarity in general modus operandi across the branches, though, like the frequently very small vowel systems highly prone to ablaut. Some proposals for Proto-Chadic (which itself cannot have been born yesterday) have no emic vowel distinctions at all.

  48. Lars Mathiesen says

    Danish vs Swedish — there are lots of false friends and inflectional gotchas, even for reading. You can make sense of 99% of a newspaper, but for something more technical like a contract there is almost sure to be something you get wrong. That goes both ways, of course. E.g.:

    Danish:
    (en) kontrakt sg.indef,
    kontrakten sg.def.
    kontrakter pl.indef.
    kontrakterne pl.def.

    Swedish:
    (ett) kontrakt sg.indef.
    kontraktet sg.def.
    kontrakt pl.indef.
    kontrakten pl.def.

  49. David Marjanović says

    The fine people of Ugarit basically said “to hell with this” and invented a cuneiform alphabet. It’s clear they were doomed from the start.

    They simply took the alphabet and invented a cuneiform font for it. It went swimmingly until the Sea Peoples swam along and burninated the city… which is probably why all those clay tablets were preserved, though.

    The existing dictionaries of PAA are both textbook examples of How Not to Do It

    In one of them, supposedly, you can see the way Arabic dictionaries are organized. Evidently, the authors went through an Arabic dictionary and looked for cognates of every entry.

  50. Another point that should have been mentioned before is that there would have been Semitic people in Egypt * because a couple of dynasties (before the New Kingdom) were Semitic (14th & 15th (Hyksos)). WikiP says that 14th dynasty was Semitic, and was invaded/overthrown by a different group of Semitic speakers.

    One thing that the articles emphasize repeatedly is that data from those periods are very sparse.

    ___________________________________________________________________
    *: Which could well have included scribes who influenced the writing system.

  51. The Danishest Danes will be 100% unintelligible, sometimes even to other Danes.

    Has https://www.youtube.com/watch?v=s-mOy8VUEBk been forgotten?

    “. . . the Danish language has collapsed into meaningless guttural sounds . . .”

  52. Lars Mathiesen says

    It has not been forgotten, but remember that it was made by Norwegians who are jealous of our subtle phonetics and ability to imbue strings of guttural sounds with meaning.

  53. The ‘missing link’ that helped our ancestors to begin communicating with each other through language may have been iconic sounds, rather than charades-like gestures – giving rise to the unique human power to coin new words describing the world around us, a new study reveals.

    https://www.birmingham.ac.uk/news/latest/2021/05/ancestors-created-'iconic'-sounds.aspx

  54. Ah, “may have been” — the universal hook of pop sci articles!

  55. David Marjanović says

    Potential cold water on the details of Goldwasser’s scenario: a short notice of Early Alphabetic inscriptions in Syria from a few hundred years before the ones in Egypt.

  56. David Eddyshaw says

    The team published their findings in Scientific Reports, highlighting that the vocalizations produced by English speakers could be understood by listeners from a diverse range of cultural and linguistic backgrounds.

    I knew it! You just need to talk SLOWLY AND LOUDLY in English.

    People get paid for this while real language documentation is out there waiting to be done …

    To be more charitable than this sort of thing really deserves, there is a good bit of evidence for cross-linguistic sound symbolism in non-onomatopoeic words; Edward Sapir himself had a classic paper on this, to the effect that (IIRC) words for “big” tend to have lower vowels, while words for “little” tend to have higher vowels, and made-up words with these characteristics are interpreted along such lines by speakers from several different linguistic backgrounds.

  57. David Marjanović says

    There’s more recent work on this sort of thing, too. I’ll look for some tomorrow. It deals with the exceptions, e.g. the fact that the Korean for “big” sounds teensy-weensy…

  58. John Cowan says

    Big, high vowel; small, low (or at least non-high) vowel.

    Among anglophone linguists, English is always either an example of the obscure feature they are describing, or a counterexample to the universal they are proclaiming. This is one of the reasons I post on academia.edu papers-in-progress.

  59. January First-of-May says

    Big, high vowel; small, low (or at least non-high) vowel.

    Indeed, though OTOH see the history of mickle (up to and including “many a mickle makes a muckle”, which is basically a poster for this symbolism).

  60. The contrasting antonyms are big and little (about the same height), and large and small (ditto).

  61. David Marjanović says

    I’ll look for some tomorrow.

    Er, no. I’m considerably busier than I realized.

  62. David Eddyshaw says

    It works in French (petit/grand) and Kusaal (bil/tita’ar), so it’s clearly a language universal.

    Not in Welsh (bach/mawr) but Welsh is an exception to everything, on account of being a conlang invented by the ancient Britons to confuse the Romans.

  63. And here was me thinking Welsh was the language of Gardd Eden!

  64. PlasticPaddy says

    @de
    I suppose P-C *bikkos > bach, so the vowel was the same as VL *piccus/pittus > Fr petit. In fact the VL word might be ex Celtic

  65. Looks like it, doesn’t it?

  66. VL *piccus

    pikku
    Finnish
    Etymology
    Cognate with Karelian pikku, Veps pikuraine and Votic pikkarainõ. Originally a diminutive form of pieni.

    Adjective
    pikku (not comparable)

    1. little, small
    2. little old
    Yksi pikku pokeripeli silloin tällöin, ei se ole uhkapeliä.
    Just one little old poker game now and then, that’s not gambling.

    https://en.wiktionary.org/wiki/pikku

    The plot thickens.

  67. David Marjanović says

Speak Your Mind

*