Ancient Cryptography

December 08, 2022, 04:55:36 AM
Welcome, Guest. Please login or register.

Ancient Cryptography » Ancient Texts » Voynich Manuscript » Computation approach to Circular Astrolabes

Author Topic: Computation approach to Circular Astrolabes  (Read 4969 times)

0 Members and 1 Guest are viewing this topic.


  • Newbie
  • Posts: 41
Computation approach to Circular Astrolabes
« on: May 20, 2013, 12:16:06 PM »
Here is a possible approach to a solve:

Task 1:
Subtask 1.1: Match up all the words. Run these in a column in excel.
Subtask 1.2: On the header row, list the months
Subtask 1.3: In binary, fill in the rest of the cells, adding a 1 where a "celestial object" appears in a given month, and a 0 when it does not appear.
Subtask 1.4: Pull out the binary strings and convert to decimal.

What you are left with here is a unique numerical description of an object over time.

Task 2:
Do the same thing with astronomical objects from other charts.

Task 3:
Subtask 3.1: Bring all the numbers together on an excel page. Column one contains the source data, column 2 contains the numerical identifier, column 3 contains the object itself.
Subtask 3.2: Sort the page by the numerical identifier (column 2).
Subtask 3.3: Start tabulating correlations where numerical identifiers from other sources match the numerical identifiers from the VM Astrolabes.

In this manner, we don't have to worry about missing information (lack of 12 astronomical objects) or instances of mislabelling, as you are matching the objects themselves.


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #1 on: May 21, 2013, 02:15:33 PM »
Where possible, the word at 12:00 or the first word immediately to the right of 12:00 starts the list, and follows in a clock-wise rotation. In some cases where the circle was incomplete, words are listed starting at the left-most. These instances are denoted (l-r).

As concerns the rings, ring 1 is to be considered the center-most ring, ring 2 the circle that encloses this, and ring 3, if present, is the outermost ring.

The number of words in each list is noted at the start of each line.

My draft transcription:

70r Pisces
ring 2 (19): oty ar, okaly, otody, otalg, otaldar, okody, opysam, chekchy, otaly, otalarar, otaldy, okeoly, okyly, okees, otalalg, okaly, otar, oty, oky ody
ring 1 (10): otalar, otalam, dolaram, okaram, oteosal, salols, okaldal, ykolaiin, ataram, ota*al

70v Aries 1
ring 2 (10): opchey sal, otakaiman, okalal, otaly, oalcheeg, oteeodals, okolShy, otShShdy, otal ypSharal, otalchy tar am ly
ring 1 (5): otalaiin, oteo alalS araly, oteoeey otal okeal ar, otear araydy, okoly

71r Aries 2
ring 2 (10): otcoaldy, oteolar, okcoaly, otaleky, opald?ar, cheary, oteotey sary, otalaly, oteos arar, okldam
ring 1 (5): otolchdy, otoloaram, oteeol, otolchg, otal dar

71v1 Taurus 1
ring 2 (10): okolar, otchody, alcphy, otaiin, okaraiin, otar araly, opalar am gain, opalor ar, char orom, chpaly
ring 1 (5): otalody, otalaiin, otar Shar, Shol Shdy, opacpom

71v2Taurus 2
ring 2 (10): oaiin ar ary, okalam, ytalShgy, char alep, otaraldy, otaiin otain, otalp as ainam, ochol Sharam, oShodady, chdaiirdairy
ring 1 (5): otchoShy, otchdal, okeey ary, otainy, oparalar

71v3 Gemini
ring 3 (l-r) (5): opchdajy, oklairdy, okaram, okairy, okealar
ring 2 (15): okal, okaly, okal, okeey ary, oteeary, otairgy, okaircham, okeal, otars?, okady, orary, okyd, otalam, otaraldy, okalar
ring 1 (9): ralal, okam, otalShy, okaldy, chasar, *tam, ainaly, okarcham, otal

71v4 Cancer
ring 3 (12): oraiinam, osarSheee*, otaShaiin, opoiiis?oin al aes, ypoiin alaly, oteey daiin, oeeodaiin, opSholdy, t** okaiin, ykalairol, olkalaiin, olalsy
ring 2 (11): ytairal, oeeeSaiin, o*, och*y pady, p**eeeS*ly, ykairaiin oiral, okalar, orar*, olaiin olackhy, olpSheoral, aralkam
ring 1 (7): ypary, oraiiral, ytoar Shar, octho, aral, oletal, opalal

72r1 Leo
ring 2 (l-r) (12): orchey, ykam, okey, otcheody, okyeeShy, **opaiin, akaly, otoly, opalals, She** ocphey, daiin chepy, oteey dche*l
ring 1 (l-r) (9): choleey, oky, chcthey, laly, oeegy, odairan, okary, oeeShaiin, oteeShd

72r2 Virgo
ring 2 (18): okeoram, airaim, oShesy, okeosy, cheockhy, choloy, Sheoeky, oteadar, opchdysd, oteead, yteod, okeody, okeoldy, ykeeor, opaiinam, cheoldy, oeegey, oeeodaiin
ring 1 (12): ockhol, oeejy, opalg, okeeog, olkar, occdaly, cheosy, opch*y, yteedy, oraiin, okeorar, okaiin

72v Libra
ring 2 (20): ocpShy, oees, okeeoly, odal, aiinod, odady, oteod, yteod, okchody, okal, ytaly, oeeeos, oteopy, cheoepy, ykeedy, oeeaty, octhy, oteoly, okeoly, oeees
ring 1 (10): oiiny, oeey, ote*y, ypail, okalair, occody, okeal, occoly, okygy, okeey

73r Scorpio
ring 3 (l-r) (4): otaly, chockhy, okegy, yteeody
ring 2 (16): oteedyl, akeedy, okeey, okeedy, ykeeary, oteeosy, Shekal, oteegyg, okedal, chdy, dalShey, opaiin, okeos, otey, otaly, okeody
ring 1 (10): okary, okeeody?, oyteedy, oky, chepy, otal, Shek, ar, kar, okeos

73v Sagittarius
ring 2 (16): ogees, echsy, okeor, *opals, ykeody, oteogy, yp**, ypaiin y, opcheesy, araiiny, ykeeody, okeos, ykeear, qokeoly, okeogy, Sheol
ring 1 (10): cheg, oty, ara*, okal, ykey, ykche, oteog, ykey, opal, otegy


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #2 on: May 24, 2013, 10:21:59 AM »
Ya, you know, I haven't quite returned to this as I have a lot of deadlines coming due, but I noticed that the frequencies of starting letters is REALLY close to Arabic, especially if k and t are taken to be the same letter.

Of course, there's a problem there also, in that one would presume "ok" would then transliterate to "AL", but there are a small amount of Arabic stars starting with "L", where there are no VM stars that start with K. Doesn't mean the Arabic "L" stars are included though.

I was finding that otherwise, the frequency distributions seems very close indeed, both in first letters, as well as in the largest chunk of starting bigrams, which are those starting with "o", with "ok" leading the charge.

The uniqueness of words could possibly be explained if the VM vocabulary is based on a phonetic rendering rather than a lexicon. In that case, ks and ts might be presumed to be approximately the same sound, but modified based on what follows the character (that is, the t shape for L might be an emphasis if followed by, say, an H or something).

Trying to address the first problem, which is that I didn't expect QUITE so many unique words in the list. Obviously, without a harmonized star catalog, the approach listed in step one is useless.
« Last Edit: May 24, 2013, 10:46:37 AM by Phil_The_Rodent »


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #3 on: May 26, 2013, 12:27:08 PM »
Thanks for the links!

Ya, if the VM was staight up Arabic, I'm sure it would have been solved a long time ago. As I mentioned Feb 10 of last year, the language seems like it might be some sort of "localized non-Semetic Afroasiatic language which exhibits some Semetic tendencies". But if Arabic astronomy had some sort of influence on the VM nomenclature, then there remains some hope (however small) that the stars are not in pure Voynichese but are transliterated (in whole or in part) from Arabic names (or sourced from another such catalog).
« Last Edit: May 26, 2013, 02:07:45 PM by Phil_The_Rodent »


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #4 on: May 26, 2013, 11:45:35 PM »
Just loose work, hoping for something like this:

okody      <>      Al-Haq'ah
okoly      <>      Al-Han'ah
otolchdy      <>      Al-Alyah
okolShy      <>      Al-Arnab
okolar      <>      Al-Anf
otolchg      <>      Al-Hamal
otoloaram   <>      Al-Athafi


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #5 on: May 27, 2013, 01:23:41 PM »
Ya, it was rough just to give you a kind of idea on approach. Was kinda looking at it like this:

o = a
k/t = l
d = q
y = ah
l = n
chd = y
Sh = r
ar = f
chg = al

ok-o-d-y      <>      al-a-q-ah      <>      Al-Ha-q'-ah
ok-o-l-y      <>      al-a-n-ah      <>      Al-Ha-n'-ah
ot-o-l-chd-y      <>      al-a-n-y-ah      <>      Al-Al-y-ah
ok-o-l-Sh-y      <>      al-a-n-r-ah      <>      Al-Ar-nab
ok-o-l-ar      <>      al-a-n-f      <>      Al-Anf
ot-o-l-chg      <>      al-a-n-al      <>      Al-Hamal

Obviously not perfect, but, okay for the purposes of demonstration. I should mention, and probably in caps, that this is almost certainly not correct. It's just a rough-up.

I've downloaded the paper and done an OCR pass on it. When I compile the lists out, I'll post them up.
« Last Edit: May 27, 2013, 04:52:33 PM by Phil_The_Rodent »


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #6 on: June 02, 2013, 03:47:26 PM »
Some more thoughts in the interim.

A culture without a standardized lexicon of written words is not a writing society, but rather an oral tradition. In such a case, we might suppose that there would be a low number of written works (which is obviously true of Voynechese), and they would be based on phonetic transcription. I realize a fundamental assumption that seemingly everyone has made may, in this case, very well be wrong -- that is, that the subject matter experts were also the people who wrote it down.


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #7 on: June 03, 2013, 02:10:58 PM »
Hmm. Ya, I can see your point.

English is still being modified, but today it's most obvious with text messages. In this case, modifications are still based mostly on phonetic transcriptions (eg. " r u srs?"). You're right that it doesn't follow that the culture is largely non-literate (although the lack of other Voynechese texts might support that). In today's English there is a sort of language schism between formal written language (lexicographic) and informal written language (based on the oral language).

Still, I think it's good to recognize that the assumption that the subject-matter experts were the people writing is an assumption. Could be true, but, could also not be true.


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #8 on: June 04, 2013, 11:41:39 PM »
Cool. I'll have a look about for it.

And yes, I was actually a Communications major, so the writing stuff and the development of a literate culture is nothing new to me (though as time passes, things take on a rosier glow).
« Last Edit: June 04, 2013, 11:46:31 PM by Phil_The_Rodent »


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #9 on: June 05, 2013, 12:37:44 AM »
Here's a start on these.

Names of Stars of 1st Magnitude
1α Tauceyn ha-shoral-dabarān
2β Oriregel te’omimrijl jawzā
3θ Eriaharit ha-naharakhir al-nahr
4α Orisad te’omim yeminimankib al-jawzā ayman
5α CMiha-kelev ha-qatanshicrā ghumaysā
6α CMakelev gadolshiccabūr
7α Aurmoshekh ha-resenal-cayyūq
8α Booha-nitmakh ha-romeahsimāk al-rāmih
9α Virnitmakh lo’ mezuyyansimāk aczal
10α Lyrnesher nofelnasr wāqic
11ζ Sgrsof zenav ha-susasl dhanab al-faras
12α Leolev ha-aryehqalb al-asad
13α Cenregel ha-sus ha-mequddamrijl al-faras muqaddama
14α PsApi ha-dag ha-deromifūm al-hūt al-janūbī

Names of Stars of 2nd Magnitude
1α Andha-yad ha-sevucaal-kaff al-khadīb
2γ Orisad semol te’omimmankib al-jawza
3β Perrosh ha-shedra’s al-ghūl
4β Leozenav ha-aridhanab al-asad al sarfā
5α CrBnezer sefonial-fakka
6α Scolev ha-caqravqalb al-caqrab
7η UMaha-me’ir mi-benot cayish asher be-sof ha-zenav
8α Aqlha-nesher ha-mecofefal-nasr al-tā’ir
9α Cygzenav ha-tarnegoletal-ridf
10ζ Cygarkovet ha-tarnegoletrukbat al-dajāja
11β Pegsad ha-sus ha-yeminimankib al-faras
12δ Leoshidrat ha-arifaqār al-asad
1315(c) Comme’ir beyn zenav ha-ari veha-nitmakh kokhaval-dhu’āba
14α Carkesil vehu’ meha-cerekh ha-rishonsuhayl


  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #10 on: June 05, 2013, 06:33:50 PM »


  • Guest
Re: Computation approach to Circular Astrolabes
« Reply #11 on: June 06, 2013, 02:38:35 AM »
Totally brilliant of you to find this! Thanks for the link.



  • Guest
Re: Computation approach to Circular Astrolabes
« Reply #12 on: July 01, 2013, 08:50:30 AM »
"localized non-Semetic Afroasiatic language which exhibits some Semetic tendencies".

Would a minor dialect of Judeo-Persian do?

- just asking. Don't have a candidate.



  • Newbie
  • Posts: 41
Re: Computation approach to Circular Astrolabes
« Reply #13 on: July 05, 2013, 02:24:06 AM »
Could be. I'm mostly thinking North Africa and related regions (middle east). There is, I think, around 3000 African languages currently, many with few native speakers. Any sort of tribe that would have some amount of influence from both the Afroasiatic and Semetic groups could be candidates.

Again, I think it's a dead language. Hence my interest in the comparison to proper nouns from other languages in the thread.


  • Guest
Re: Computation approach to Circular Astrolabes
« Reply #14 on: July 08, 2013, 10:27:40 AM »

This comes from someone without any natural talent for languages, near-total ignorance of cipher-things, and only the faintest residue of training in  linguistics ~  so cum grano.

Earliest influence in the imagery c.5thC BC, latest apart from written text I put about the 3rdC, before recension in the 12th-13th.  So this will range, just a bt.

If there's ever a sweepstakes, put me down for early Hellenistic texts, written in glossary style, taking characters (glyphs) from as many as  seven languages to form a set by which all are represented. Basically a Yemeni-influenced Greek script, I'd guess.
If enciphered, probably by using some wheel system.. a la Alberti, but not his.

Seven languages turn up as a common number spoken by traders, so I'm opting for seven different languages that form the glossary, phonetically rendered using the one set of glyphs.

This script was made by a group who so very strongly avoided any 'x' shape that they use it neither in their script, nor in imagery, nor does the manuscript use the Roman 'x' for ten.

 I read this as tabu to do with the southern 'hell' and perhaps Argo ratis constellation.  Not Byzantine Greeks - they also place the 'x' for the southern limit of the world, but draw it there and keep the Chai in their scrpt.

 A related tabu in the manuscript is avoidance of interlace - *very* rare tabu for any Mediterranean peoples.  Cross-shapes and knots being avoided is typical of some  mariners and some peoples across North Africa (see Westermarck, for example). And, I almost forgot, it is characteristic of dynastic Egypt, especially before Alexander. So a mariner-people is possible.

 One of the languages  among the 'glossary's seven may well be New Persian, but I strongly suspect an old Yemeni dialect there somewhere, unless the text was newly rendered in the fifteenth century.  In that case, I'm tempted to include instead Genoese and perhaps even English (Old- or Middle).

Judging by those stylistics which highlight distinctions between the various sections,  some of the basic Hellenistic matter derived from a community in Egypt; more from northern Syria and the elbow between Asia Minor and the Syrian coast.  Interaction between the two regions was constant, from the ancient to the modern era, so 'national' limits don't really apply, although regional affect is still evident in the imagery.

A distinctive 5-element system is found one folio. It's not the European system of elements  and points (I think) to Manichaean influence of a type   pervasive through the southern Mediterraean to about the 3rd or 4th C AD.  It then travels as far as China, just to make things trickier.

There are some areas where all these influences might be found in close proximity in the earlier centuries AD -  Lycopolis in Egypt is one. Pehaps Canopus, or Alexandria two more.

I confess that I'm putting all this here mainly for the fun of it; after five years I've no illusion that I'll be heard.

Unlike some Voynicheros excited by owning a theory they've developed for their own, I have only 'my conclusions' and presently fairly uninterested even in them. 

Voynich'd out.

« Last Edit: July 08, 2013, 10:37:22 AM by dodonovan »