Mandombe.

Frequent commenter Y wrote me as follows:

I happened to see the Wikipedia article on the Kimbanguist church of the DRC (formerly Zaïre), a messianic Christian movement. It’s quite interesting in itself, but what caught my attention is the Mandombe script, said by its inventor to have been revealed to him through Simon Kimbangu himself. The script’s appearance and logic are to me spectacularly strange, like nothing I have seen before except maybe some ciphers. A Unicode encoding is underway (the proposal has various examples of the script in use).

He also sent me a Wayback Machine link to a pdf of Helma Pasch’s 2010 article “Mandombe” for Afrikanistik. It really is a remarkable script; I don’t think I could learn to use it myself, but I’m glad it exists.

Comments

  1. David Marjanović says:

    Well, I think we can say that Kimbangu was not dyslexic.

  2. Maybe that’s a way to do spelling reform? Just tell that god ordered you to spell something this or that way and who can tell differently?

  3. @D.O.: Written norms, written language itself, and “high” languages under diglossia have a history of being treated as divine or sacred (whence Jean-Noël Robert’s “hieroglossia” moniker). I find it interesting to speculate on the connection between writing and religion/mystical revelation.

  4. I’ve looked at descriptions of Madombe before and been unable to understand how it works. After looking at the 2016 Unicode proposal I think I get it, and will attempt to summarize very briefly here.

    There are consonantal bases and vowels which attach to them to form a glyph representing a CV syllable. So far a fairly standard abugida, but the twist is that the glyphs for the (first twenty) consonants are formed from only five consonantal bases, which are rotated and reflected so that each can appear in four different forms representing different consonants. This is reminiscent of Canadian Aboriginal Syllabics, but here changing the consonant instead of the vowel. And there doesn’t seem to be any systematic phonological relationship between the consonants represented by transformations of the same base; e.g. the first one is b-/d-/g-/f-, the second m-/k-/p-/l-. The whole glyph undergoes this transformation, including the vowel marker; moreover most of the consonantal bases are rotationally symmetrical, so often the position and orientation of the vowel appendage distinguishes pairs of syllables with different consonants, e.g. be vs de or gi vs fi. A fifth, null consonant base allows bare vowel syllables to be expressed; its transformations represent h- and -h (???) syllables.

    Then there are consonants for gb, kp and kt, which are formed by modification of the b, k & t bases respectively.

    Beyond this, there are a set of vowel diacritics which can be added to syllabic glyphs to represent diphthongs; a vowel nasalization diacritic; and elements which can be added to the consonantal base to add prenasalization and form consonant clusters with intervening l and r. A null vowel marker allows glyphs representing bare consonants.

  5. Thanks, that’s very helpful.

  6. Of course, only now do I notice that I mistyped “Mandombe”…

  7. David Eddyshaw says:

    Written norms, written language itself, and “high” languages under diglossia have a history of being treated as divine or sacred

    Perhaps there’s something of a chicken-and-egg thing going on: I suppose that “high” languages in a diglossic situation are generally (always?) linguistically archaising, and given the sheer intellectual difficulty of maintaining a consistently archaic form of language, you’re going to need a powerful motivation: religion is an obvious one, especially if your religion is one where the ipsissima verba are necessary for your hymn/prayer/spell/whatever to be effective.

    Perhaps secular linguistic peevery is the poor orphaned remnant of this religious motive in an unbelieving age? It might account for some of the disproportionate emotional investment so often associated with it.

    Then there are consonants for gb, kp and kt, which are formed by modification of the b, k & t bases respectively.

    On the face of it, that’s a bit odd: k͡p g͡b are common as single phonemes in Africa (though not so much in this area) whereas kt isn’t. I suppose it just shows that despite its deliberate oddness the system is really parasitic on the Latin alphabet (unlike Cherokee, say.)

  8. David Marjanović says:

    (always?)

    On average, probably, but not in every detail. Standard German has a more conservative grammar than my dialect, has less syn- and apocope, and lacks a pervasive consonant lenition process; but each has kept vowel distinctions the other has lost, and the dialect keeps a few vowels the standard has syncopated.

  9. That’s probably not true for many languages which got their literary standards in 19-20th centuries.

    Cantonese (and most other Sinitic varieties) is more archaic than Mandarin Chinese, all Tibetan dialects are archaic compared to modern literary Tibetan based on Lhasa dialect, etc…

  10. January First-of-May says:

    Supposedly, many Russian dialects (…to the extent that any survive) still have a sound for ѣ separate from е, which is something that the literary language lost centuries ago.

  11. David Eddyshaw says:

    Come to think of it, whether there are archaic tendencies in the “high” language is not relevant to the more general point; the mere fact that it is different imposes a cognitive cost.

    Modern nation-building with its secular ethos of cultural homogeneity has presumably supplanted religion as the driver of diglossia, so that the effort of maintaining diglossia is now imposed on citizens by economic necessity or indeed by straightforward state coercion.

  12. You lost me there.

    Why knowing and using two languages is a bad thing exactly?

  13. David Eddyshaw says:

    It isn’t, unless it leads to loss of the “low” language. This isn’t an issue if the motivation for knowing the “high” language is purely religious (I can’t think of any instances where that has ever happened); when the motive is nationalist that is not only common but often a deliberate plan.

  14. Almeida Samo says:

    Mandombe 52 is a very advanced science unlike anything the world have ever seen. The inscrip MASONO is just a part of a wide range of other studies such as architecture, maths and KIMAZAYI(not translated) all based in Kikongo language, the language of the Gods.

  15. Trond Engen says:

    Oh, so it’s Kikongo that is the language of the gods? Finally! All the other claims I’ve seen have seemed so dubious that I almost doubted there were such a thing at all.

  16. Stu Clayton says:

    We have the Ancient Greeks to thank for showing how spiteful and ego-trippy the gods can be, including the language they used. Not a good example for the kids.

  17. John Cowan says:

    Modern nation-building with its secular ethos of cultural homogeneity has presumably supplanted religion as the driver of diglossia, so that the effort of maintaining diglossia is now imposed on citizens by economic necessity or indeed by straightforward state coercion.

    Well, in diglossic nations, yes: Scotland and Jamaica, e.g. But the result can also be a collection of mesolects, as in (white) America, England, or Germany.

    This isn’t an issue if the motivation for knowing the “high” language is purely religious (I can’t think of any instances where that has ever happened)

    See my writeup of the weird case of modern Burmese: one language, two grammars. This is distinct from the village of Kupwar (eheu fugaces!), with its three languages but only one grammar (2008, 2014).

  18. there doesn’t seem to be any systematic phonological relationship between the consonants represented by transformations of the same base

    There’s however a sign of a relationship with the Latin alphabet. b d f g are the first four consonants that have an exact equivalent in the languages written in Mandombe. The next four common to all of them would be k l m n, though for some reason n gets skipped and we get instead the set k l m p. (Tshiluba does have j /ʒ/, and Swahili h /h/ & j /dʒ/ ; if Wikipedia’s phonological inventories are to be trusted, Kikongo and Lingala lack these though.) After this would be r s t v; now r gets skipped and instead n is added in this set. Lastly we wrap up with w y z and add r in this set.

    I have no idea though about the logic behind group and family numbering. /b d f g/ as group 2 still at least has the families in alphabetic order, but the others are /k m l p/ as group 3, /n v s t/ as group 1, /w r z y/ as group 4. (A “pure” application of the alphabetic theory would probably rather predict group 1: /b d f g/, group 2: /k l m n/, group 3: /p r s t/, group 4: /v w y z/.)

  19. Almeida Samo says:

    What you see as consonants in mandombe are called MVUALA(Not translated) They are consonants if used for writing. In mandombe mechanics they can be the motor parts, and in architecture they are the bricks. Mandombe 52 is a multi dimensional science revealed from the spitual world by Kimbangu to properly codifi all the existing science and that to come, actually hidden in the African languages.

  20. Almeida Samo says:

    Mandombe starts from two elements that looks like 5 and 2 named PAKUNDUNGU and PELEKETE in Kikongo language. These 2 elements were found in the brick wall, they ‘re simetric and obeying the optical laws(if you look 5 in the mirror turns into 2. Vice versa)
    These two elements are added a third element called HIKAMU which can turne a 5 into 9 or 6, and 2 into a backward 9 or 6.
    Here Biggins the TWIST.
    The HIKAMU is rotated in various angels not non to the modern science, starting from 0 degree, 45, 90, 135, 180 degree. The 180 degree is not used for writing. Only in mathematics.

  21. John Cowan says:

    In Greece, the end of diglossia took place because the H variety lost too many domains (starting with literature), became associated with political conservatism, and finally expired altogether. I keep wondering if that might happen in Haiti too,

  22. ə de vivre says:

    I have nothing to add except that I once worked with a Congolese guy who was thinking of moving to Quebec City because that’s where many members of his church had ended up. At the time, I wondered if he knew what he was getting himself into and thought that Quebec City and Kinshasa, where he was originally from, might be the two places most unlike each other in la francophonie.

  23. J.W. Brewer says:

    One twist on David E’s observation that nationalism has replaced religion as a primary motivator of continued diglossia: as I understand the situation in Greece, ecclesiastical circles are one of the few remaining cultural redoubts in which katharevousa remained (and remains) in active use after the secular authorities threw in the towel and went all-demotic-all-the-time.

    If by the “H variety” in Haiti, John C. means “something more or less resembling standard French”, that’s still a rather useful language to know if you’re going to leave home and travel in the wider world, which is not a feature katharevousa possessed.

  24. David Marjanović says:

    In b4 “all Greek to me”.

  25. John Cowan says:

    Certainly French is very useful outside Haiti, but Haitian extended to H domains would be far more useful inside Haiti, since it is the best language of all but epsilon of the population.

  26. After the 2010 disaster is Haiti, I remember Wyclef Jean announced he was going to run in the Haitian presidential elections that fall. I heard him interviewed on the radio, and he started off with a really strong Haitian accent. However, as the reporter asked him pointed questions about why he, a rapper, who had no public sector experience, and had never lived in Haiti as an adult, was qualified to lead Haiti, he slipped fairly rapidly back into his relatively normal African-American accent, as he floundered about trying to justify his candidacy. (He was, soon thereafter, ruled ineligible.)

  27. Almeida Samo says:

    Mandombe 52 is a science. The writing system is just part of it, demonstrating how sounds are geometrically created, end their application.
    Mandombe writing system only function properly in science if writing Bantu languages or any other syllabic languages, in which the the syllabus are the crystalized cosmic sounds like in Kikongo language obeying the principal of rotation and combinatory simetry. The Kikongo language already had the attributes to be written with mandombe.
    For instance: in Kikongo to say ELEFANT is NZAMBA with 2 syllabus NZA and MBA, in progressive order. If written in regressive order is MBANZA or BIG CITY.
    Can’t do this with the word ELEFANT.
    There are only few words in English language capable to be written in both directions, and have simetry like: OK/KO the state of the boxers before and after fight.
    In Kikongo the all language is created like this.

  28. David Marjanović says:

    English simply has a lot more different syllables than Kikongo. The rest is statistics.

  29. Almeida Samo says:

    The syllables in Kikongo are like rubic cubes when written in mandombe, they become real structure. Before mandombe the Kikongo syllables where used only for magic powers where words become living things. I know this sounds like hoohoo. But with mandombe words become structure as well.
    With mandombe we can design buildings and compex computer circuits using Kikongo words

  30. January First-of-May says:

    English simply has a lot more different syllables than Kikongo. The rest is statistics.

    Pretty much; you’d probably get effects similar to the described Kikongo scenario in, say, Japanese.
    OTOH, English also almost certainly has a lot more different words than Kikongo, which somewhat counteracts the effect.

    There are of course a lot of English words that become other words when spelled backwards (the classic example pair is “live” and “evil”); there are probably also many English words that become other words in the verlan-esque transformation described by Almeida Samo, though offhand I can’t think of any.

  31. Almeida Samo says:

    The concept of syllables of the Kikongo language is hard to be understood by the western science that only focus on material side of things. Kikongo is not created in alphabetical order, but in meanly 16 syllables called MAZITA, the same as the Japanese MORA. It’s like this 16 sounds came from another advanced rhelm of existence and their are the building blocks of everything in the physical rhelm.
    Therefore the words are created using this same syllables.
    Exemple: NZAMBA(elefant), NZAMBI(God), the same NZA and MBA is used for both God and Elefant. In Kikongo only the vowel (a) as true existence. The other 4 vowels are the manifestation of (a) in 4 other degrees of sound. Therefore the MBA of NZAMBA is the same as the MBI of NZAMBI. In MBA is absolute fire and MBI interiorized fire

  32. Almeida Samo says:

    The MAZITA or syllables of Kikongo are composed with a consonant and a vowel. The consonant is an open angle that is closed with a vowel. All the words end with a vowel so that no angle is left open. We call the vowels BISIMBA and the consonants MVUALA. The BISIMBA are attached to MVUALA to form the MAZITA(ties) to create sound, where the consonants are like envelopes that conceals the messages in the vowels.
    When we write with only consonants and fewer vowels like in English, the MANDOMBE becomes useless scientifically.

  33. Jen in Edinburgh says:

    Compound words should do it easily enough – there’s the famous houseboat and boathouse, and overturn and turnover also come to mind.

    No doubt there are words that just do it by chance, but they’re harder to think of!

  34. David Eddyshaw says:

    IIRC, some of the Japanese Kokugaku

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

    scholars were given to a similar mystical interpretation of the syllables of Japanese, presumably because – as with Kikongo – there aren’t all that many distinct ones. This enables idiosyncratic ideas about how the favoured language works, and (of course) its privileged relationship to Life, the Universe and Everything.

    Mind you, I just linked to a paper elsewhere which suggests that there is at least one language in Africa in which “syllable” (as opposed to “mora”) is not a useful concept at all. And Randy LaPolla would probably approve …

  35. Almeida Samo says:

    The strange thing with MANDOMBE presumably just discovered, complement the Kikongo language that exist for millions of years. Looks like MANDOMBE is the way Kikongo was written from the beginning.
    Kikongo use a simetry and rotation in it’s words and meanings that now comes back to life.
    Exemple: in Kongo science of finance the words
    NTALU(price) is the rotation of the word NLUTA(benefits), and NKUTA(merchandise) is the rotation of NTAKU(capital). Capital-merchandise-price-benefits rotate continuously.
    This rotation is also reflected in MANDOMBE MAZITA of 0 degree BA GA DA FA they are symetric and rotational

  36. David Marjanović says:

    I just linked to a paper elsewhere which suggests that there is at least one language in Africa in which “syllable” (as opposed to “mora”) is not a useful concept at all.

    I’ve now read it: the author disagrees with his earlier work that had that conclusion, and states that “syllable” is either useful for one very narrow purpose, or at least he hasn’t been able to explain that one fact otherwise. In other words, “all languages have syllables” and “all languages have CV syllables” are probably true at such a general level that it’s downright boring.

    the Kikongo language that exist for millions of years

    No. Speaking humans probably haven’t existed for more than a single million years, quite possibly less. Kikongo is so closely related to so many other languages that it cannot have been a separate language until a few hundred years ago.

    Looks like MANDOMBE is the way Kikongo was written from the beginning.

    No. Like almost all languages, Kikongo wasn’t written at all until very recently (the 16th century in this case).

  37. Trond Engen says:

    … and not in Mandombe script until 1978. But I don’t think that matters at the mystical level.

  38. Yes, we’re not dealing with a grubber of facts here.

  39. Almeida Samo says:

    Mandombe is both mystical and scientific, it correct the science as we know it. Mandombe has already a bunch of inventions including an astronomical clock with 28 hours day and 70 minutes instead of 60 minutes to be in cynch with a 7 day week. What amaze me is that all the scientific names are in Kikongo and even in their chemistry day mean what they are, and they can be demonstrated by numbers and letters.

  40. David Marjanović says:

    all the scientific names are in Kikongo and even in their chemistry day mean what they are, and they can be demonstrated by numbers and letters.

    I have no idea what you mean; please explain.

  41. Almeida Samo says:

    It’s quite difficult to explain, but mandombe Biggins by learning to change letters into numbers using a scientific method that uses only a 25 letters alphabet. Thereafter it’s letter of the alphabet represented by a number being the letter A number 1 and Z 25. Now I can change any given word into number.
    When we look let’s say the names of the numbers in Kikongo, let’s say ONE in Kikongo is MOSI, if the word MOSI is changed into number, it’s results in 1. Then a pettern develops, where the next number 2 in Kikongo ZOLE, the word ZOLE is 3, next word is 5, 7, 9 . Then we get to number 6 in Kikongo SAMBANU and this word if turned into number is also 6. Just to mention few things.

  42. Almeida Samo says:

    I would like to mention that the 25 letters mandombe alphabet, only the letter Q is excluded because is not used in Kikongo due to it’s place placement wich is at 10th place from Z to A. (Kikongo functions from wright to left and left to wrigh)
    According to MANDOMBE, since 10=1+0=1, the letter Q is a repetition, and also geometrically is the same as the letter O, that is why is not used in Kikongo to bigin with.

  43. David Marjanović says:

    …That’s not science, though. That’s just counting.

  44. Almeida Samo says:

    I am not counting but chaging words into numbers because numbers doesn’t lie, and explaing that the letter Q is not used in Kikongo because it’s the 10th letter from the wright. The 10th letter from the left is J also avoided in Kikongo, and both Q, J fall under the angle of 180 degree with the letters C and X riquiring special consonants named MAZITA MAZINDINGA or migratory sounds.
    According to the true history, the world started in Kongo the center of the world, also known in mandombe as the SINGINI, where all races and languages departed from Kikongo. As people migrated, developed other languages and sounds.
    Since mandombe is a spiritual science, in Kikongo or even Japanese certain sounds are not to be used due to their spiritual power or misfortune.

  45. Almeida Samo says:

    Therefore, the words that describe numbers in Kikongo, corresponde to the number described or follow a pettern of add numbers first from 1 3 5 7 9.
    Its like saying English ONE and use a formula to change the word ONE into a number and the number is also 1.
    Let’s try to change the word ONE into number:
    ONE= O(15), N(14), E(5)
    ONE=15+14+5=34=3+4=7
    Now let’s do it in Kikongo. In Kikongo the number 1 is MOSI
    MOSI=M(13), O(15), S(18), I(9)
    MOSI=13+15+18+9=55=5+5=10=1+0=1

  46. Trond Engen says:

    But that’s the (English version of) the Latin alphabet, and the number and order of letters weren’t finally settled until after the invention of the printing press. Surely Kikongo is older than that? How does the number magic work in Mandombe?

  47. John Cowan says:

    Which reminds me of this amazing (tour de) force:

    Choose a number from 0-99 (actually some numbers like 77 and 87 will not work). Write it down in English words (without hyphens).

    If you have written more than 10 letters, then start over. If not, count how many letters you have written, and write that number down in English words after the first set of words. Count how many letters you have written altogether. If the result is 13, stop.

    If not, count how many letters you have written and write that number down in English words after the first and second sets of words. Count how many letters you have written altogether. It will be 13. So either way you get to 13.

    Examples:

    If you choose 87 you will write EIGHTYSEVEN, which is eleven letters, but EIGHTYSEVENELEVEN is 17 letters, so the trick fails. (There are 18 such numbers.)

    Choose 13 and write down THIRTEEN. You have 8 letters, so write down EIGHT and you have THIRTEENEIGHT, which is 13 letters.

    Choose 7 and write down SEVEN. You have written 5 letters, so write down FIVE, and now you have SEVENFIVE. You now have written 9 letters, so write down NINE and you have SEVENFIVENINE, which is also 13 letters.

    As far as I know there is no other written language in which the trick works.

  48. Trond Engen says:

    Heh! And “Heh!” again. If I’m not mistaken, there are only six numbers that don’t work for TOLV (12) in Norwegian. I’m sure there’s some law involved here.

  49. Trond Engen says:

    And the six meet the rest at 23.

  50. Trond Engen says:

    ENTOFIREÅTTETOLVSEKSTENTJUETRE
    TRESEKSTISEKSTENTJUETRE
    FEMÅTTETOLVSEKSTENTJUETRE
    NIELLEVESYTTENTJUETRE

  51. Trond Engen says:

    I made a stupid error as usual and was very wrong about the six. But all 15 numbers that don’t meet at 12, meet at 11, and they all meet at 23.

  52. David Marjanović says:

    In b4 Scandi-Congo.

  53. David Eddyshaw says:

    Curses!

    Actually, what I was going to say is that we now have another exciting tool for investigating long-range linguistic relationships: Cowan-Engen Number.

  54. Trond Engen says:

    I was too tired to do this last night. I also corrected myself incorrectly. There are 18 numbers not meeting in 12:

    33 35 37 – 44 46 48 – 54 56 58 – 63 65 67 – 74 76 78 – 94 96 98

    They all have 11 letters. 11 and 12 meet in 23. This covers all numbers up to 110.

    111 has 14 letters. Numbers with 13 and 14 letters meet in 21, and they meet the group with 15 letters in 53. But these don’t converge with group 23 for the visible future.

    All numbers up to 1100 are included in these two groups.

  55. Trond Engen says:

    Me: I’m sure there’s some law involved here.

    No law yet, but the chance of two groups meeting at any one number is inversely proportional with the average number length. It’s also postivitely correlated with variance, less clearly with skewness, and likely with kurtosis (“flatness”), but except for blips at the exact hundreds and thousands, average number length will increase while the higher order moments are near constant. Hence, if two series don’t merge before the numbers reach the tens and hundreds, they may follow eachother in lockstep for a long time. Probably not indefinitely, though, since length is basically logarithmic, and the blips come at a constant rate — unless all higher-order number lengths happen to be multiplums of the same number.

    David M.: Scandi-Congo

    My son started looking at Swahili, but he says there are too many four-letter numbers.

  56. Trond Engen says:

    They all have eleven letters

    Aargh! They all have nine letters, Nine and twelve meet in 23. But 9 + NI is 11, and it sounds weirder that 11 and 12 meet in 23.

  57. John Cowan says:

    You can only call them Cowan-Engen numbers by applying Stigler’s Law of Eponymy (not, of course, invented by Stigler). I found out about them by googling for simple examples of forcing in magic.

    In this case the prop would be an analog clock dial with just an hour hand, originally set to midnight but then invisible to the magician. The victim follows instructions which limit the original number to 1-12, all of which meet at 13 in three steps. The magician then reads their mind and announces that the hour hand points to 1, which of course it does.

    Most forces are less reliable than this, of course; they merely raise the odds rather than being a dead certainty.

  58. David Eddyshaw says:

    You can only call them Cowan-Engen numbers by applying Stigler’s Law of Eponymy

    I already had a bad conscience about this. They ought, of course, to be Samo numbers.

    (My error was as bad as my home town’s monument to the Scot who invented television. Any self-respecting Russian can set you right on that point.)

  59. Trond Engen says:

    simple examples of forcing in magic

    Yes. when I presented this for my son, the maths student, he immediately came up with a bunch of examples of how to improve your odds in a seemingly fair setting. His chief lesson is “Never make a bet with a mathematician!”

  60. since length is basically logarithmic

    Change in length is logarithmic, length grows at slightly more than the constant rate.
    If the rule is not Merge, but Replace it will definitely come to a few distinct cases. In Russian, there are 2 obvious point attractors ТРИ(3) and ОДИННАДЦАТЬ(11) and one loop attractor ЧЕТЫРЕ(4->6)->ШЕСТЬ(6->5)->ПЯТЬ(5->4)->ЧЕТЫРЕ. And probably that’s it. English, the short-word language it is, will come to FOUR all the time.

  61. David Eddyshaw says:

    “Never make a bet with a mathematician!”

    Never play cards with a man called Doc.

  62. Trond Engen says:

    D.O.: Change in length is logarithmic, length grows at slightly more than the constant rate.

    Yes. I didn’t define “length”, but meant it as the length of any number spelled out. If we instead use L for the length of the total string and ΔL for the length of the added number, then (the mean of) ΔL is logarithmic, and the chance of merging with another string is inversely proporsional with ΔL.

  63. John Cowan says:

    Never play cards with a man called Doc.

    Never eat at a restaurant called Mom’s.

    And never sleep with a [POTAS] whose troubles are worse than your own.

    —Nelson Algren

    Note that the first two lines constitute a snowclone.

  64. Almeida Samo says:

    Well, if the numbers and letters are Roman or Greek, why they work perfectly in Kikongo other than any other languages? I tried that same experiment in other languages but it just doesn’t work.
    The words of numbers in Kikongo, also much with the 7 days of creation in the Bible.
    The SAMBANU(six) or the sixth day the JEWISH pray, makes more sense in Kikongo, because SAMBA mean to pray and ANU , is the supposed SUMERIAN GOD.
    SAMBANU mean…WE PRAY ANU ON THE SIXTH DAY.
    The wine they drink when they pray in kikongo is named MBASA, just the rotation of the word SAMBA(prayer)
    I can go on and on with this.

  65. David Eddyshaw says:

    [POTAS]

    I would never sleep with a Promoter Of a Tax Avoidance Scheme. I have standards. (Never kissed a Tory …)

  66. @John Cowan: When I used to drive from Boston to Cincinnati a couple times per year, I made a point to, if possible, stop for dinner at Mom’s Dutch Kitchen. They had good food, not expensive, and excellent desserts for a song.

  67. I don’t care how good the desserts are, I’m not singing in public (unless I’m drunk, which presumably I wouldn’t be at Mom’s).

  68. January First-of-May says:

    Change in length is logarithmic, length grows at slightly more than the constant rate.
    If the rule is not Merge, but Replace it will definitely come to a few distinct cases. In Russian, there are 2 obvious point attractors ТРИ(3) and ОДИННАДЦАТЬ(11) and one loop attractor ЧЕТЫРЕ(4->6)->ШЕСТЬ(6->5)->ПЯТЬ(5->4)->ЧЕТЫРЕ. And probably that’s it.

    OTOH, the Merge rule gets some weird results. I did a preliminary calculation in Notepad (not 100% confident of it), and as far as I can tell, the branches starting from 1 (ОДИН) and 2 (ДВА) don’t merge until they reach respectively length 812 and 811, both of which transform to 831.

    By this point – indeed by the time the length passes 200 – any other number up to 60 letters long would already have merged into one or the other branch. I hadn’t checked further.

     

    EDIT: for cross-checking…

    Lengths starting from 1:

    4-10-16-27-39-53-65-79-94-109-118-133-147-159-177-193-208-220-234-254-275-294-315-331-349-366-387-408-423-443-461-484-510-523-541-557-577-597-617-635-655-676-698-721-740-752-771-791-811-831

    Lengths starting from 2:

    3-6-11-22-33-44-55-68-84-101-108-117-130-141-153-168-187-205-215-231-249-266-287-308-320-334-354-375-394-415-434-457-479-503-513-530-545-561-582-603-614-634-656-678-701-712-729-750-766-788-812-831

    Just as predicted by Trond Engen, they do indeed follow each other in near-lockstep for a while until the fluctuations happen to line up just right for a merge.

Speak Your Mind

*