From gauss!mycroft.rand.org!jim Sat Jan 02 12:04:16 PST 1993
Received: by gauss; Sat Jan  2 15:10:40 EST 1993
Received: by inet.att.com; Sat Jan  2 15:10 EST 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA14446
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 2 Jan 1993 12:04:13 -0800
Received: from localhost by mycroft.rand.org; Sat, 2 Jan 93 12:04:18 PST
Message-Id: <9301022004.AA01741@mycroft.rand.org>
To: voynich@rand.org
Subject: Anybody need a copy of Newbold?
Date: Sat, 02 Jan 93 12:04:16 PST
From: Jim Gillogly <jim@mycroft.rand.org>
Status: OR

I noticed a copy of Newbold's "The Cipher of Roger Bacon" in a bookstore
the other day for $35.  Anybody want me to pick it up for them?  It appears
to have all the Voynich plates and is in fair condition, no dj, covers
reinforced with book tape.  Bookplate from a previous owner, a Mr. Kelley
(no, not Edmund).

If you'd like me to get it for you (and if it's still there), the tax here
is about 8% and I'd expect postage to be a couple of dollars unless you're
out of the US.

	Jim Gillogly

From gauss!trl.oz.au!j.guy Mon Jan  4 13:42:31 EST 1993
Received: by gauss; Sun Jan  3 21:43:28 EST 1993
Received: by inet.att.com; Sun Jan  3 21:43 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA00179
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 3 Jan 1993 18:42:52 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA16840; Mon, 4 Jan 93 13:42:31 EST
Date: Mon, 4 Jan 93 13:42:31 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301040242.AA16840@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: Voynich editor countdown
Status: OR


Still remains to do:

1) write the documentation
2) zip the lot
3) uuencode it
4) send it to Jim Gillogly

In other words, yes, it's just about there. The funny thing is, I am a
bit disappointed, after so much effort. Possibly because the font still
doesn't look that awfully good. 3705 lines of Pascal, packed. The
EXE file is 62K


I have taken some short-cuts. For instance, I'm using a batch file that
loads the Voynich font, runs the editor, then reloads the standard
font.  Early (4 am) this morning too, I realized that the *&$^@#$@
Turbo Editor Toolbox on which I'd based my work did not have a "Directory"
and a "Change directory" command, so I implemented them, quick and dirty,
and they don't look so integrated with the rest. So it won't be really
finished. A thing I haven't done is pick a colour scheme for monochrome
monitors. As it stands, it assumes a color monitor, VGA or SVGA, and
won't work on monochrome. Perhaps I will make the (slight) effort of
adapting it to monochrome just before zipping it all and sending it.
The hardest part will be to get hold of a machine with a monochrome
monitor. We don't seem to have any around here. It will work on color
EGA, except that the fonts will look a bit wrong. (I'll have to figure
out how to adapt Harald Thunem's font editor to the 14-byte EGA font
definitions, some day).

Rick ... Rick who? Can't find a name in that message -- e-mail unwashed
as I am -- RJYANCO@maherst.edu had suggested I implemented the block-marking
method of WordPerfect. But that isn't trivial (got to keep track of the
scan codes), so I decided on implementing the next closest thing:
double-click the left mouse button to mark the start of a block, the
right button to mark the end. That's on top the other commands and the
shortcuts (F7 and F8), of course.

Enough said. Now to write the documentation.



From gauss!trl.oz.au!j.guy Wed Jan  6 09:54:31 EST 1993
Received: by gauss; Tue Jan  5 17:55:35 EST 1993
Received: by inet.att.com; Tue Jan  5 17:55 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA21958
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 5 Jan 1993 14:54:40 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA18970; Wed, 6 Jan 93 09:54:31 EST
Date: Wed, 6 Jan 93 09:54:31 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301052254.AA18970@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: We're there.
Status: OR

However, I shall refrain from sending the long-awaited Voynich editor
to Jim for putting in pub/jim at rand.org for just a week or two yet.
You know how it is: the very moment I send it, I am bound to
discover a dreadful bug that thrashes your hard disk, fills it with
lost chains, destroys your FAT, and parks your heads with a
case-splitting Big Bang.

To whet your appetites, however, here is the documentation -- proudly
written using VOYED.



                    THE VOYNICH EDITOR -- Version 93.01



This package should contain the following:

LOADFONT.EXE       Harald Thunem's font loader.
FE.EXE             Harald Thunem's font editor.
VOYNICH.FNT        The font file you must load. Contains
                   both Roman and Neo-Frogguy Voynich fonts.
VOYNICH.MAP        The keyboard map of the Voynich fonts.
VOYED.EXE          The Voynich editor proper.
VOYEDIT.BAT        A batch file that loads the Voynich font,
                   runs VOYEDIT, then restores the default
                   character set.

GETTING STARTED

Copy those files either into the directory that contains your Voynich
data, or into a directory in your path, e.g. C:\BIN or C:\UTILS or
whatever. When you run VOYED, it first identifies the directory in
which it resides; next it looks for VOYNICH.MAP in that directory. So
it is absolutely essential that VOYED.EXE and VOYNICH.MAP be in the
same directory.

As I have not been able to locate a PC with monochrome VGA yet, I
haven't allowed for those. So find yourself one with a color monitor,
VGA or EGA, but preferably VGA.

Now fire the editor up, either running the batch file VOYEDIT, or
loading the Voynich font (LOADFONT VOYNICH), then running VOYED.

I have tried to make everything easy to use and, at this stage, I'd
say: just play around with the editor and see what it does.

Type something in English or whatever. Now press Alt-K: the text you
have been typing moves down four lines to make room for a map of the
keyboard, green on black.

Press Alt-A: the keyboard map switches to Neo-Frogguy and you can type
in some Voynich: 4oqpc89 4oCQPtc89 etc. Alternatively, if you really
cannot be bothered remembering which keys lurk behind those Voynich
squiggles, use the mouse to type. Move the mouse onto a letter on the
keyboard map, click the left button and voila! by a trick of Enochian
magic, that letter appears just where the cursor was in the text
(thank you, Dr Dee and Mr Kelley).

If you click the mouse in the body of the text, the cursor teleports
there.

To mark the beginning of a block double-click the left button there.
To mark its end, double-click the right button just after the end.
(Why not on the end-of-block itself? you may ask. In that case you
could not distinguish between a line, carriage-return included, and
the same line, carriage-return excluded.)

Be warned that the TAB key moves in eccentric ways. Probably nothing
like your favourite word-processor. When overwriting, it moves
the cursor to just under the first letter of the next word on the
previous line. When inserting text, it inserts just enough spaces so
that the cursor and the rest of line on its right are aligned on the
next word on the previous line. If the previous line is blank, nothing
happens. I found this behaviour of the TAB key far more useful 
than fixed tab settings.

That is about all you need to know to get started. If you are as
adverse as me to reading instructions, you can skip the rest, I think,
without much harm done. In that spirit, rather than boring you now
with a description of the commands and menus (I shall leave that for
last), let me explain a few things. And first...


BUGS?

I am certain that bugs still lurk in this piece of software. Why, as I
was writing this, I found one that messed up the "reformat paragraph"
command when in Auto-Indent mode. It took me five minutes to locate
and two to fix, but still... The truth is: this program is a mess. It
is computationally far, far too expensive, redrawing the whole screen
when just one line needs to be redrawn, and other sundry horrors. I
sometimes notice a sort of shooting star scanning a line from right to
left. I do not know what it is, beyond something to do with screen
refresh, but that's one of the things that irk me, and I certainly
shall try to get rid of the wretched phenomenon... in time.

The next thing I ought to do is rewrite the whole sorry lot. That will
probably shrink it by another thousand lines and make it faster and
more reliable. So I beg your forgiveness, and assure you that I will
be the first not to put up with any nonsense from my brainchild, for
my infinite natural-born laziness simply doesn't allow it.

In the meantime, save your work often (pressing the F2 key does it).


THE KEYBOARD MAP

A bit of a kludge, really. The editor does not test for what keyboard
you have, simply because I do not know how to do that. It assumes that
you have the same as mine, on my 386 at home. The assumption may well
be wrong, for the backslash key is in a different place on my 486 at
work. And if you have a German or, worse still, French keyboard, all
is lost! I promise that I will learn how to recognize keyboards and
will amend the Voynich editor accordingly, but do not hold your
breaths for it yet.


MAKE-YOUR-OWN ALPHABETS

If Neo-Frogguy is not your view of a proper font for the Voynich; if,
God forbid! you're not interested in the Voynich at all, only in
typing text in mixed alphabets, you can use Harald Thunem's font
editor and alter the VOYNICH.MAP file for your foul designs.

Let us look at the beginning of the VOYNICH.MAP file:

2 128
^ 129

Each line contains one character, followed by one space and a
number. The number is the ASCII code (decimal) of the Voynich
letter assigned to that character. So:

1. Design your own font using Harald Thunem's font editor, keeping
   your alternative alphabet in the extended ASCII range. In other
   words, recycle as many as you like of those mathematical symbols
   and letters with accents, but leave what you see on your standard
   keyboard unchanged, and do not touch the six special characters
   that make up a single-border box, that is: (illegible on this 
   Unix box).

2. Using Harald Thunem's font editor again, make a note of the
   decimal ASCII number of each of your customized letters.

3. Write a keyboard map file with, on each line:
   a) a character
   b) a space
   c) the ASCII number of the special letter you want to correspond to
      that character.
   If you want the Voynich editor to handle a special letter as a
   punctuation mark, make its ASCII number negative, e.g.: & -129

4. Rename the keyboard map file VOYNICH.MAP.


ERROR MESSAGES

The original editor out of the Turbo Editor Toolbox featured many
error messages. I have removed all but a few, and these I have
replaced with little warning tunes. There is only one to worry about
and that is the out-of-memory dirge. Do not ask me what its three
notes are, I am hopeless at musical dictation. But know that its first
note is 432 Hz, 0.1 second long; its second note 216 Hz, i.e. one
octave below, and twice as long (0.2s); its third note is 54 Hz and
0.4s, i.e. two octaves below and twice as long again. It should never
be heard, unless you are editing some mammoth monster of a file, 400K
or thereabout and shifting huge blocks around. However, this editor
does a lot of dynamic string allocation and deallocation and perhaps,
perhaps, it does not deallocate everything it should when it should.
In which case you may run out of memory after a long editing session.
(That, incidentally, is the kind of programming error that plagued
Windows 3.0 and caused it to crash so inordinately often, I am told).

Having given you those dire warnings, I shall turn to the dire stuff,
namely, the detailed instructions for use.


EDITOR COMMANDS

The top line of the screen gives you a status report: file name, a
hint to press F10 to enter the menu, the current line and column where
you, or rather, the cursor, is. It wakes up in "Insert" mode, and that
is why you see the three letters INS in the top right corner. The
shape of the cursor (a half-block) also shows you that you are in
insert mode. Press the "Ins" key on the numeric pad to toggle between
insert and overwrite mode and see how the cursor changes shapes, and
INS disappears and reappears.

The second line of the screen is occupied by the menu bar, white on
blue:


  File    Block    Replace/find   Delete    Go to    Edit    Toggle   Quit

If you have a mouse installed, there will be a grey block in the top
left corner of the screen. If you move the mouse to the menu bar
you'll see the grey block disappear. Instead, one of the menu items
will light up, white on red. To open the corresponding menu, click the
left button, or press RETURN. To close it, move the mouse outside the
menu box, or on its border, and click, or press Escape, or F10.

Whether or not you have a mouse installed, you can enter the menu bar
in two other ways.

1. Press F10 or Escape. That lights up a menu item (in white
   on red). To open the menu, press RETURN. To move to another
   menu item, press the left-arrow, up-arrow, right-arrow, or
   down-arrow keys. The corresponding WordStar Ctrl keys have
   the same effect (Ctrl-S = left arrow, Ctrl-D = right arrow,
   Ctrl-E = up arrow, Ctrl-X = down arrow).

Or:

2. Press Alt and the initial letter of one of the menu items
   (those letters that are capitalized and highlighted in cyan on
   blue). That directly opens the corresponding menu item.

To execute one of the commands in the open menu, type the highlighted
letter in it (the capital letter in cyan on blue), or move the
white-on-red highlight onto it using the arrow keys or the mouse and
press RETURN (or click the left button). Careful! Whenever part of the
menu bar or part of a menu is lit in white on red, all the input from
the keyboard is directed to the menu handler. I am writing this
documentation file with VOYEDIT and, at the beginning, I panicked once
or twice: I kept typing away merrily, and lo, but no behold, nothing
appeared on the screen and the cursor wasn't moving. What, a bug? I
despaired. But no, I had moved the mouse by accident, and some part of
the menu bar was lit red. So keep that mouse at the bottom of the
screen, or keep an eye peeled for a red light on the menu bar.

If menus are not your cup of tea, each command has a WordStar-like
equivalent, which we shall see presently.

The File Menu
-------------

Item:          Shortcut             WordStar-like commands:
                                             mnemonics
Load file      F3                   ^K^O     O for Open
Insert file                         ^K^R     R for Read
Save           F2                   ^K^S     S for Save
save As    Alt-F2                   ^K^N     N for New
Directory                           ^K^D     D for Directory
Change directory                    ^K^J     J for Jump

"Load file" clears everything in memory and loads it with the contents
of a new file (it will ask you for a file name). When you run the
Voynich editor it wakes up blank, and you see, in the top left corner,
*NEW FILE*. Normally, the first thing you would do is load an existing
file. Why, will you ask, didn't you allow us to pass the name of the
file to edit on the command line, viz. VOYED MYFILE? Because, later,
when I have got around to adapting Harald Thunem's font editor to my
purposes, I intend to allow passing on the command line, not the name
of the file to edit, but that of a font-cum-keyboard-map file, so that
VOYED will in fact become a multi-lingual word-processor.


The Block Menu
--------------

Item:                    Shortcut      WordStar-like commands:
                                             mnemonics
mark Start of block       F7           ^K^B  B for Beginning
mark End of block         F8           ^K^K
Copy block at cursor                   ^K^C  C for Copy
Move block to cursor                   ^K^V
Delete block                           ^K^Y
Hide/show block        Alt-H           ^K^H  H for Hide
Write block to file                    ^K^W  W for Write
Read block from file                   ^K^R  R for Read

"Read block form file" is equivalent to "Insert file" of the File menu.

With a mouse, you can double-click the left button to mark the start
of a block, double-click the right button to mark its end.

Marked text, that is, text belonging to a marked block, is yellow on
blue, plain text is yellow on black. I first had marked text in
white on black, but you could not see the leading and trailing
spaces of a marked block, if any. I eventually settled on blue
for the background color, as it seemed easiest on my eyes. A pleasing
color scheme is a vexed problem. I started with text in black on
light grey, but soon found it very hard on the eyes. I am very
open to suggestions there, in particular from those of you who
might be using monochrome laptops: how do I distinguish between
text, marked text, keyboard map, menu bar, lit menu items in
monochrome?


The Replace/find Menu
---------------------

Item:           Shortcut           WordStar-like commands:
                                                mnemonics
Find             F9                 ^Q^F     F for Find
Replace      Alt-F9                 ^Q^A
Next          Alt-N                 ^L


When selecting Find, you'll be asked for options, namely, B(ackward),
W(ord) and N(o case). "Backward" means, look for the given string
scanning backwards from where you are. "Word" means search for the
given string as a whole word, not a substring of a word. "No case"
means make no distinction between lower and uppercase letters. In your
answer, use only the first, capitalized, letter of an option. Thus, if
you want a backward search of a string as a whole word, case ignored,
answer BNW.

When selecting Replace, you get two more options: G(lobal) and A(sk).
Global means scan through the whole file. Ask means ask before you
replace. If you select Global and don't select Ask, every string
in the file will be replaced automatically, after which the number of
changes made will appear on the status line for a short while (1
second). If you select Ask, VOYEDIT will ask you every time: Replace
so and so (y/n)? and wait for you answer.

"Next" is "play it again, Sam", that is, find or replace the next
occurrence of a string.


The Delete Menu
---------------

Item:               Shortcut      WordStar-like commands:
                                                mnemonics
Word                               ^T
Line                               ^Y
to End of line                     ^Q^Y
Block                              ^K^Y
All                                ^K^A         A for All

Delete all is a dangerous command. You will be asked for confirmation.
Anything but YES as an answer will delete nothing.

Delete block will delete the current marked block, when highlighted
yellow on blue, whether visible on the screen or outside it. Nothing
happens when the block is hidden.

The editor distinguishes between two types of characters: letters and
punctuation marks. "Delete word" deletes the character under the cursor
and all the characters to its right until it encounters a character of
a different type.

In addition, the backspace and Del keys will do what you normally
expect them to do. Backspace deletes the character left of the
cursor, and Del deletes the character at the cursor.


The Go to Menu
--------------

Item:                Shortcut       WordStar-like commands:
                                                mnemonics
Start of block                         ^Q^B     B for Beginning
End of block                           ^Q^E     E for End
Line #...                              ^O^N     N for Number
Previous page          PgUp            ^R
Next page              PgDn            ^C
Top of file       Ctrl-Home            ^Q^R
Bottom of file     Ctrl-End            ^Q^C


For "page" read "one screenful". So "go to next page" really means "go
down one screenful".

There are a few commands there that cannot be accessed through the menu
(they seemed too intuitive):

The arrow key will move you left, right, up, down. Ctrl-Left-arrow
and Ctrl-Right-arrow move you one word left or right. The Home key
takes you to the first column of the line you are on; the End key
to the end of the line (^Q^A and ^Q^D in WordStar parlance).


The Edit menu
-------------

Item:                 Shortcut         WordStar-like commands:
                                                  mnemonics

reformat Paragraph                     ^B
Center this line                       ^O^C       C for Center
set Left margin                        ^O^L       L for Left
set Right margin                       ^O^R       R for Right

I called it Edit because I was running out of inspiration for short
descriptive words all starting in different letters.

The procedure for setting the left and right margins is awkward, I
must admit: you are asked for the number of the leftmost and rightmost
columns. So for no left margin and a page 60 characters wide, you
answer 1 to left margin and 60 to right margin. I didn't think that
one would often reset margins when typing Voynich text, so I left
it at that.




The Toggle Menu
---------------

Item:           Shortcut         WordStar-like commands:
                                             mnemonics
Overwrite/insert  Ins            ^V
Alphabet          Alt-A          ^O^A        A for Alphabet
autoIndent        Alt-I          ^Q^I        I for Indent
Keyboard display  Alt-K          ^O^D        D for Display
Shift             Alt-S          ^O^K

Autoindent is one of the more useful settings. To know whether it's on
or off, look at the status line, top right corner of the screen. If
you see "AI" above the "Quit" of the menu bar, autoindent is on. What
it does is align a lign on the first non-blank character of the lign
above. Useful for keeping things neat without fiddling with the left
margin setting.

The "toggle shift" command causes the character under the cursor to
switch between shifted and unshifted. For instance, small a become
capital A, and vice versa; 1 become ! and vice versa. A letter is
affected if and only if its shifted and unshifted forms both belong
to the same alphabet. For instance, when you are typing in Voynich
letters, small g represents part of a Voynich letter but capital G
produces a Roman letter; in that case the toggle-shift command has no
effect. Small i and capital I, however, both represent Voynich
letters, and the toggle-shift command causes one to turn into the
other. So it is quite different from uppercase and lowercase. It ought
to be very useful to switch between connected and unconnected forms of
the same letter, or the free-standing and "struck-through" forms of
the gallows.


The Quit Menu
-------------

Item:           Shortcut         WordStar-like commands:
                                             mnemonics
Not yet
Yes, quit now   Alt-X            ^K^X        eXit

If you have made changes since your last save, you will be asked if
you want to save them before quitting.


                       ----------------------
                           --------------
                               -------
                                  |






From gauss!MAX.U.WASHINGTON.EDU!RJB Thu Jan  7 15:38 PST 1993
Received: by gauss; Thu Jan  7 18:38:17 EST 1993
Received: by inet.att.com; Thu Jan  7 18:38 EST 1993
Received: from uwavm.u.washington.edu by rand.org with SMTP id AA03373
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 7 Jan 1993 15:38:06 -0800
Received: from MAX.U.WASHINGTON.EDU by UWAVM.U.WASHINGTON.EDU (IBM VM SMTP V2R1)
   with TCP; Thu, 07 Jan 93 15:37:24 PST
Received: from U.WASHINGTON.EDU by MAX.U.WASHINGTON.EDU; Thu, 7 Jan 93 15:38 PST
Date: Thu, 7 Jan 93 15:38 PST
From: RJB@U.WASHINGTON.EDU
Subject: editor counted down
Sender: RJB@MAX.U.WASHINGTON.EDU
To: voynich@rand.org
Message-Id: <9BD5BEDEC21F2184CA@MAX.U.WASHINGTON.EDU>
X-Envelope-To: voynich@rand.org
X-Vms-To: IN%"voynich@rand.org"
Status: OR


Certainly a vote of congratulations and thanks is in order?

--rjb

From gauss!mycroft.rand.org!jim Thu Jan 07 16:47:47 PST 1993
Received: by gauss; Thu Jan  7 19:48:22 EST 1993
Received: by inet.att.com; Thu Jan  7 19:48 EST 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA05766
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 7 Jan 1993 16:47:49 -0800
Received: from localhost by mycroft.rand.org; Thu, 7 Jan 93 16:47:47 PST
Message-Id: <9301080047.AA13627@mycroft.rand.org>
To: voynich@rand.org
Subject: Re: The Voynich Editor Countdown.. 
In-Reply-To: Your message of Fri, 08 Jan 93 10:01:49 -0500.
             <9301072301.AA21553@medici.trl.OZ.AU> 
From: Jim Gillogly <jim@rand.org>
Reply-To: jim@rand.org
Date: Thu, 07 Jan 93 16:47:47 PST
Sender: jim@mycroft.rand.org
Status: OR

I moved Jacques' editor to the archives at

	rand.org:pub/jim/voyedit.zip

and <boy> am I impressed!  This looks like your professional-quality
buy-it-in-the-store editor, with pull-down menus and the whole bit...
except that it'll edit civilized (i.e. Voynich) files!

Sorry about that, all ye without PC clones...

Well done, Jacques!

	Jim Gillogly
	Trewesday, 17 Afteryule S.R. 1993, 00:47

From gauss!trl.oz.au!j.guy Fri Jan  8 10:01:49 EST 1993
Received: by gauss; Thu Jan  7 18:02:38 EST 1993
Received: by inet.att.com; Thu Jan  7 18:02 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA01712
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 7 Jan 1993 15:01:58 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA21553; Fri, 8 Jan 93 10:01:49 EST
Date: Fri, 8 Jan 93 10:01:49 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301072301.AA21553@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: The Voynich Editor Countdown..
Status: OR


... is over.

I have sent the uuencoded file (96062 bytes) to Jim Gillogly a few
minutes ago. 

You ought to find it soon in pub/jim/voyedit.zip at rand.org.

It contains, zipped in 69700 bytes, the editor proper, the font and
keyboard map files, 26K of documentation, written using Voyedit, and
Harald Thunem's font editor and loader, the latter slightly modified by
me so that it recognizes in which directory it resides.

You will need a PC with a color VGA, but a color EGA will do, in
a cinch. It may work on some laptops with monochrome VGA screens.

The largest file I have managed to edit with it was 293K long
(9,814 lines).




From gauss!trl.oz.au!j.guy Sun Jan 10 11:55:29 EST 1993
Received: by gauss; Sat Jan  9 19:56:30 EST 1993
Received: by inet.att.com; Sat Jan  9 19:56 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA28203
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 9 Jan 1993 16:55:39 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA23634; Sun, 10 Jan 93 11:55:29 EST
Date: Sun, 10 Jan 93 11:55:29 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301100055.AA23634@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: EGA monitors, anyone?
Status: OR


I have discovered the trick, or at least, a good part of the trick,
for modifying Harald Thunem's font editor to cater for EGA too.

But I am, as you know, lazy. So here is my question:

Has anyone in the group got a PC with an EGA display only?

 

From gauss!trl.oz.au!j.guy Tue Jan 12 15:18:06 EST 1993
Received: by gauss; Mon Jan 11 23:18:54 EST 1993
Received: by inet.att.com; Mon Jan 11 23:19 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA09160
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 11 Jan 1993 20:18:15 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA26449; Tue, 12 Jan 93 15:18:06 EST
Date: Tue, 12 Jan 93 15:18:06 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301120418.AA26449@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: vowels
Status: OR

Sneak preview.

I have received from Brian Winkel, the editor of Cryptologia,
an article on, again, Sukhotin's vowel identification algorithm.
The author had this idea of keeping decrementing the sum of the
rows of the letters identified as vowels, then dividing those
residual sums by the absolute frequencies of the corresponding
vowels, to get a number in the range 0..1. Well, I tried it
on English, and it worked! I used VOYED.DOC as my input file,
and got H and K as the last "vowels" identified. But, whereas
the true vowels had score above 0.5, K and H were both below 
0.1. I tried it on Croatian, on which Sukhotin's algorithm
failed miserably, identifying S as a vowel, and... S got
a score just under 0.2, way below what the true vowels got.
I tried it on unpointed Hebrew, and I didn't get much sense,
EXCEPT THAT the two "vowels" with the clearly highest scores
were yod and vav. Zayin came third, with 0.333, but that score
is bogus, because zayin occurred only 6 times in the sample
text. 

Now, just guess what I am going to try on, as soon as I ...

From gauss!trl.oz.au!j.guy Wed Jan 13 11:28:57 EST 1993
Received: by gauss; Tue Jan 12 19:29:57 EST 1993
Received: by inet.att.com; Tue Jan 12 19:30 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA20478
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 12 Jan 1993 16:29:14 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA27442; Wed, 13 Jan 93 11:28:57 EST
Date: Wed, 13 Jan 93 11:28:57 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301130028.AA27442@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: Latest Voynich vowels
Status: OR


I have tinkered a bit with the still unknown author's "confidence factor",
and submitted the whole of VOYNICH.NOW to the revised algorithm. I had, first,
replaced every occurrence of CC with a single character, since it is certain
that CC is a single letter. Here are the results (Currier's transcription):


Letter       Confidence       Frequency
  O             0.873          17,885
  A             0.942          10,094
  9             0.600           8,636
  C             0.532           8,577
  CC            0.432           3,079
  0 (zero)      0.630              27

"Frequency" is not the absolute frequency of the letter, but the number of
times it is found adjacent to another, different letter.  E.g.
if the whole text were AAT the "frequency" of A would be 1.

Now, for comparison purposes:

CROATIAN
Letter       Confidence       Frequency
  A             0.892           683
  I             0.914           557
  E             0.877           537
  O             0.806           547
  U             0.784           259
  S             0.145           297  <--- low "confidence"; not a vowel after all.
  0 (zero)      1.000             2  <--- low frequency: meaningless "confidence"
  4             1.000             1
  7             1.0000            1

UNPOINTED HEBREW
Letter       Confidence       Frequency
  yod           0.471           280
  vav           0.560           273
  lamed         0.211           142
  he            0.114           149
  resh          0.264           106
  ayin          0.118            68
  zayin         0.286            14  

However, on certain texts, I get very strange results indeed:

LEVITICUS, King James version
Letter       Confidence       Frequency
    E           0.632            18,149
    O           0.521            13,301
    A           0.496            11,895
    I           0.481             9,872
    T           0.179            11,974
    D           0.243             4,180
    U           0.145             4,172
    G           0.174             2,226

U fails to make it as a vowel, but D almost does. Why? Perhaps because
of the very high frequency of YOU and YOUR; and of AND. So, careful!

        


From gauss!trl.oz.au!j.guy Wed Jan 13 14:25:05 EST 1993
Received: by gauss; Tue Jan 12 22:26:39 EST 1993
Received: by inet.att.com; Tue Jan 12 22:26 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA25467
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 12 Jan 1993 19:25:30 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA28062; Wed, 13 Jan 93 14:25:05 EST
Date: Wed, 13 Jan 93 14:25:05 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301130325.AA28062@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: Two Voynich editor bugs.
Status: OR


I've received one bug report from nelson@reed.edu, and I've found 
one myself!

1. If your display mode is set to anything but 25 lines and 80 columns,
   the Voynich editor crashes very nastily and hangs your system
   (press the reset button). Mea culpa, I never thought of it, because
   I always use the standard, 25 line, 80 column mode.

2. If you do a global replace, the stupid thing doesn't think the file
   has changed at all and will not ask you if you want to save it
   when you quit. That is unpardonably stupid of me, and so easily
   fixed.

I'll wait for more to show up before I send a better version to Jim
Gillogly.

Please try that editor out as much as you can and let me know 
what other stupidities it commits.

Meanwhile, I have learnt the meaning of the last word of the 7th last
line of folio 78r (which is olpa2ax in Frogguy). It's Latin, and reads:

Ut sint in sacco, rimantur pulpam de via ecstra.
Iste currat in angulos ac in sulco, ecce, ova aspicit.

The author, Roger Bacon, was obviously a top-flight expert in data
compression.

I am going to have a go at writing a font editor that will also 
work for designing fonts for dot-matrix printers -- at least
the more common, Epson-compatible one.

I only have to locate a manual, and, hopefully, a printer!

From gauss!trl.oz.au!j.guy Fri Jan 15 11:03:34 EST 1993
Received: by gauss; Thu Jan 14 19:05:20 EST 1993
Received: by inet.att.com; Thu Jan 14 19:05 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA04821
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 14 Jan 1993 16:04:08 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA00854; Fri, 15 Jan 93 11:03:34 EST
Date: Fri, 15 Jan 93 11:03:34 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301150003.AA00854@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: Voynich miscellanea. Call for volunteers. Special Offer.
Status: OR

I had received from nelson@reed.edu questions about my "Latest
Voynich vowels", and had answered directly, not to voynich@rand.org.
This gave rise to other questions, so that now, I feel I might 
do worse than answer them publicly. It seems that there are a few
newcomers to this group, and perhaps they would no mind joining in?

Here is the latest e-mail from Nelson with answers interspread.

---------Start of quote --------------------------------------------
(you'll have to forgive me if I make a mistake somewhere. The only
alphabets I have to work from right now are the ones transcribed in
D'Imperio. Isn't there a postscript Frogguy and Currier alphabet
somewhere on the ftp site? I couldn't find it. I don't even have
Frogguy handy, although I remember it's structure more or less.

>CCC is quite common, and you see CCCC occasionally.

OK, but does this work like the way i stacks up? It seems perfectly
reasonable to consider K to be IJ, L to be IIJ, and 5 to be IIIJ,
especially seeing how uncommon K, L, and 5 are. Same goes for some
other letters in the alphabet, as Frogguy fluent people no doubt know.

I'm not trying to hash out the One True Alphabet of the VMS, I'm just
expressing my frustration at not knowing what it is :-)


 My answer:  Identifying what constitutes the true letters of a
connected script such as Voynich, in the absence of any information
about the language, is a problem that has not yet been solved. The
closest that came to it was a series of articles by a Russian
scientist, that same B.V. Sukhotin, who proposed and algorithm for
segmenting a continuous text into its component morphs. I tried it in
1977 and, well, it leaves to be desired. It's terribly expensive
computationally, too. Currier K, L, and 5 are almost certainly letter
groups, not single letters, for the reason you say: they are very
rare. His N and M, on the other hand, are almost certainly single
letters, again judging from their frequency. I suspect that such
strings as 8AII89, which you see occasionally, are variants of
8AN89 or 8AM89 (I don't know which).


>Further, the two strokes of CC are lightly connected

How is this character different from S?

My answer: CC *is* different from S. S only superficially looks like CC.
The second C in S has a flat top, and looks like an Uncial t.  The
medieval script known as Beneventan, by the way, has a letter identical
to Voynich S, and the value of that letter is t. Sukhotin's algorithm
identifies S as a consonant. Coincidence?

>Finally, when you count CC as a single letter, Sukhotin's algorithm
>*always* identifies it as a clear vowel. The other letter always
>identified as a vowel is O, which look like a Roman (or Greek) o. All
>in all, I think that it makes for a very strong presumption.

If you want to accept that there's a one-to-one correspondence between
things on the page and letters or phonemes. Which isn't too bad of an
assumption, probably, but it makes me unhappy.


Answer: There need not be. The correspondence can be very loose, as in
English.  You will remember, from my "Latest Voynich vowels" message,
the revised algorithm identified the vowels in a standard English text,
and no longer misidentified any consonants. Same in a sample Croatian
text where the original Sukhotin algorithm kept misenditifying S as a
vowel.  And finally unpointed Hebrew, in which the two letters that
often function as vowels came out clearly (in Hebrew writing all short
vowels and many long vowels are expressed points and dashes; only some
long vowels are regularly represented by a point or dash *and* a letter)
So, as long as the Voynich manuscript is written in an alphabetical 
system, Sukhotin's algorithm ought to help start sort things out.



>The history of Frogguy is this: being lazy as I was, I couldn't be
>bothered learning, and above all, remembering Currier. So, after much
>trial and error, I designed a font and a transliteration system such
>that Voynich text would look a bit like Voynich when written in Roman
>letters, and that English text would still be readable when written
>in Voynich letters.

So the "guy" in "Frogguy" is you?                               Yes!

>S  =  ct

or cc with a cursive embellishment to make it one letter? Is there any
other use of t other than in making ct?

Answer: Frogguy t, identical with an Uncial t, is always connected to a
preceding letter. It is also the right-hand half of Currier Z, and of a
number of rarer letters, of which Jim Reeds made a list out of
D'Imperio's, assigned them numbers, and produced a postcript file to
print them. Those rarer letters -- some of which, I found, are not so
rare -- are not distinguished by Currier (e.g. Neo-Frogguy It, which`
is fairly frequent, is, in Currier, the same as ct, i.e. S). Currier
has discussed how he thought S and his letters (Q etc) composed of
"gallows" intruding inside S were probably written (stroke order). The
file is somewhere in pub/jim but I don't remember its name off-hand
(incidentally, perhaps it's time for a clean-up and a compilation?
Things seems to have accumulated there and it must be hard for a
newcomer to find his way through).
 


(again, I'm not trying to solve the problem - I just don't understand
the decisions made in the transcription system.)
------------end of quote---------------------------------------------


CALL FOR A VOLUNTEER, PERHAPS

I have a suggestion.  We ought to prepare a GIF of at least Currier,
Frogguy, and Neo-Frogguy.


Come to think of it, we have OCR scanner here, but I never inquired
about it. Perhaps it can scan into GIF or whatever format. If so,
I'll draw up those alphabets and produce a GIF file. If not, I'll
just draw those alphabets and call for a lucky volunteer to scan
my artwork into GIF.


POST SCRIPTUM AND SPECIAL OFFER

Working with a computer or two, or three, every day of my life has
addled my brain. A pox and seven curses on all this high-tech nonsense!
I'll just draw those alphabets, make a stack of photocopies, and send
them old-style mail to those who ask for them.

 


  

From gauss!trl.oz.au!j.guy Sat Jan 16 14:07:04 EST 1993
Received: by gauss; Fri Jan 15 22:08:25 EST 1993
Received: by inet.att.com; Fri Jan 15 22:08 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA14333
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 15 Jan 1993 19:07:26 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA02171; Sat, 16 Jan 93 14:07:04 EST
Date: Sat, 16 Jan 93 14:07:04 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301160307.AA02171@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: About "michiton oladabas"
Status: OR

"michiton oladabas multos portas" is what many authors 
have read on the last page of the manuscript. 

I had a look at the very good photographic reproductions in Newbold's
book "The Bacon Cipher", and I beg to disagree with everybody. Instead
of "mich..." I read "a(nu)ch..." i.e. "auch" or "anch". Or perhaps
"auth" or "anth". That comes out very clearly in Plate XVI which shows
the "key" enlarged. The leftmost part of what various authors have
taken to be an "m" is an "a", very much like to the last "a" of
"oladabas", and a very common medieval form for "a". The next two
strokes, connected to the "a" and to each other, and which they have
interpreted as the rightmost stroke of an "m" followed by an "i", are
also a common medieval form for "n" and "u". The letters that have been
interpreted as the first two a's in "oladabas" are very different from
the third "a" and more like what I read as an "a" in "a(nu)ch...", and
perhaps they are not a's (but I know too little palaeography).



Above that famous line is (Plate XV) another line, very difficult
to read, but clearly in the same alphabet, where I read in the 
last word, or rather, imagine that I can read:

              ?ntzifer

in which ? is a letter I can't identify, (it looks a bit like 'v' of
modern German gothic cursive). I'm not sure at all about the 'r'
either. I can't help thinking of "entziffern" = to decipher in German.
But perhaps it's my imagination. I do not even know when the word
"entziffern" was first attested.

The last line of the "key" starts with two Voynich words, AROR ZCC9
in Currier, a2o2 c'tcc9 in Frogguy, followed by what some read

      valsch ubren so nim gaf mich o

First, the "n" of "ubren" is probably not an "n", but rather a "y".
Word-final "n" in medieval scripts of this style, did not have a
separate form with a flourish. On the other hand, "y" often was the
shape of the last letter of that supposed "ubren". The last letter of
"gaf" is not an "f". It is similar in shape to the first letter of
"so". In other words, it is an s-longa.  That does not necessarily mean
that it is an "s". The "s" often took a different shape word-finally,
similar to our modern "s". "R", too, was often written differently
when word-final, and I have seen manuscripts where its final form
looked like s-longa. So this word might be "gar", a good German word...
if the letter after the "g" is an "a", and I'm not convinced of that.

What is strange is that the German-like text is well-aligned on the
opening Voynich words. It makes one think that that line was written
by the same person, one very fluent in writing Voynich letters.
 


From gauss!gauss.att.com!reeds Sun Jan 17 12:44:25 EST 1993
Received: by gauss; Sun Jan 17 12:44:57 EST 1993
Received: by inet.att.com; Sun Jan 17 12:45 EST 1993
Received: from research.att.com by rand.org with SMTP id AA09091
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 17 Jan 1993 09:44:44 -0800
Message-Id: <199301171744.AA09091@rand.org>
From: reeds@gauss.att.com
Date: Sun, 17 Jan 93 12:44:25 EST
To: voynich@rand.org
Subject: Voynich progress by Jacques 
Status: OR

Jacques recently posted several interesting letters.  This is
in reaction to them.

1.  Nelson@reed.edu wanted to know about availability of postscript
fonts for Currier and Frogguy letters.  I have been spending about
a year tinkering off and on with a Currier font, a reasonably 
up-to-date version of which is publically ftp-able from rand.org, 
with pathname "pub/jim/voyfont.ps.sh.Z".  It codes up the 36 letters
in Prescott Currier's alphabet and about 100 extra symbols which occur
with varying probablility in the VMS and in secondary literature.
I also made a stab at the Frogguy letters, but Guy says my version
of his alphabet stinks.  

My Currier font is "work in progress" and you get what you get.  I, of
course, am eager for feedback about usability, elegance, suggestions
for changes, etc.

There is a possibility that James (Kibo) Parry will finish it up
professionally, after I have done tinkering with it.  This will
happen in the indefinite future.

And of course Jacques has his Frogguy bitmaps.

2.  Jacques suggests that pub/jim is in need of cleaning up & organizing.
I agree.  I think each contributer should review his contributions,
and suggest to Jim Gillogly which can be discarded, which can be replaced
with better versions, and so on.  As a major offender I will of course
do this soon.  

3.  I think a FAQ file might possibly be in order.  On the other hand,
whoever starts to write such a file is well on his way to writing a book!

4.  I agree with Jacques that "michiton oladabas..."  seems to start
with "a" not "m".  I am not so sure about "entziffer", though.

Another clear picture of the last page is given by
Bennett:  have you looked there recently?


Jim Reeds


From gauss!world.std.com!kibo Mon Jan 18 01:08:10 0500 1993
Received: by gauss; Mon Jan 18 01:08:40 EST 1993
Received: by inet.att.com; Mon Jan 18 01:08 EST 1993
Received: from world.std.com by rand.org with SMTP id AA20229
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 17 Jan 1993 22:08:13 -0800
Received: by world.std.com (5.65c/Spike-2.0)
	id AA08138; Mon, 18 Jan 1993 01:08:10 -0500
Date: Mon, 18 Jan 1993 01:08:10 -0500
From: kibo@world.std.com (James "Kibo" Parry)
Message-Id: <199301180608.AA08138@world.std.com>
To: voynich@rand.org
Subject: Re:  Voynich progress by Jacques
Status: OR


Jim Reeds wrote:
> My Currier font is "work in progress" and you get what you get.  I, of
> course, am eager for feedback about usability, elegance, suggestions
> for changes, etc.
> 
> There is a possibility that James (Kibo) Parry will finish it up
> professionally, after I have done tinkering with it.  This will
> happen in the indefinite future.

Very indefinite;  I seem to get busier and busier each passing month.
(I'm currently trying to digitize enough letters of a humana antiqua
from 1467 to have a specimen to show at the graphic arts show here next
week...)

Now, what would be *really* great would be if someone had photostats of
the Voynich of enough quality so that the characters could be enlarged
to at least two inches tall... then they could simply be digitized
directly...
							-- K.


From gauss!trl.oz.au!j.guy Mon Jan 18 11:07:25 EST 1993
Received: by gauss; Sun Jan 17 19:07:54 EST 1993
Received: by inet.att.com; Sun Jan 17 19:07 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA14830
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 17 Jan 1993 16:07:29 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA03562; Mon, 18 Jan 93 11:07:25 EST
Date: Mon, 18 Jan 93 11:07:25 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301180007.AA03562@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: more on "michi ton..."
Status: OR


In Joyce Whalley's "The Art of Calligraphy -- Western Europe and America" (Bloomsbury
Books, London 1980) there is p.79 a full-page reproduction out of a medieval manuscript
the writing of which strikingly resembles that of the "key" on the last page of the
Voynich. It is German, ca 1460, written in Latin and treats of things botanical. Trees,
on this page. There is a letter that looks a bit like Voynich 8 and what most authors
read as "s" in the "key". But that letter stands for "d". Final "s", on the other hand,
looks like a Greek beta, of the modern German double s. The last word of line 7, in the
left hand column, is "valde" ("strongly, very"). Its five letters are identical to the
first five letters of the six-letter word that Newbold and others read "valst" or "valsch"
or whatever.

There are, in paragraph 2, two occurrences of an "x" exactly like the x's in the "key",
the second one being an "x", the first one an abbreviation. I suspect "us", since final
"us" has been long written x, but I cannot decipher the Latin for sure. It seems to say:

Avornus est arbor parva
quae ?? ??? Alpes oritur
cinis (?) medicinis coctex Datus  <-- a false x there
in cibis aut potibus mirabiliter
ventrem laxat.                    <-- a true x

The x of "coctex" (I am not sure about the letter before x) does not represent an x.

So much for the readings of the key that go  "six  + marix"  etc.

I don't think one has much chance of getting anywhere without paying very close
attention to the calligraphy of the manuscript. I am going to be very careful 
drawing up a table of Voynich, Currier and Frogguy equivalences. Writing that
table off the top of my head isn't good enough. I must go back to the best 
reproductions I have and work as objectively as I can from there.


From gauss!trl.oz.au!j.guy Mon Jan 25 10:14:57 EST 1993
Received: by gauss; Sun Jan 24 18:16:27 EST 1993
Received: by inet.att.com; Sun Jan 24 18:16 EST 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA09253
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 24 Jan 1993 15:15:34 -0800
Received: from hal.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA13961; Mon, 25 Jan 93 10:14:57 EST
Date: Mon, 25 Jan 93 10:14:57 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9301242314.AA13961@medici.trl.OZ.AU>
To: voynich@rand.org
Subject: Be patient...
Status: OR


... I'm working on it, that is, tables of the transliteration
systems. It takes time if it is done properly, and it's hard on
the eyes. Spotting the less usual characters in those negative
enlargements I have of the Yale microfilm can be time-consuming
and painful. Patience, patience...

From gauss!ucrmath.ucr.edu!baez Sat Feb 20 11:28:34 PST 1993
Received: by gauss; Sat Feb 20 14:27:33 EST 1993
Received: by inet.att.com; Sat Feb 20 14:27 EST 1993
Received: from ucrmath.ucr.edu by rand.org with SMTP id AA14033
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 20 Feb 1993 11:26:57 -0800
Received: from guitar.ucr.edu by ucrmath.ucr.edu (4.1/SMI-4.1)
	id AA26703; Sat, 20 Feb 93 11:28:34 PST
Date: Sat, 20 Feb 93 11:28:34 PST
From: baez@ucrmath.ucr.edu (john baez)
Message-Id: <9302201928.AA26703@ucrmath.ucr.edu>
To: voynich@rand.org
Subject: Rosicrucians to the rescue
Status: OR

I have long lost hope regarding the mysterious Voynich ms, but
maybe we can consult the font of all wisdom...

Article 2116 in alt.religion.computers:
From: aa438@cleveland.Freenet.Edu (John Perko)
Subject: The Rosicrucian Fellowship BBS
Date: 20 Feb 93 04:10:53 GMT
Organization: Case Western Reserve University, Cleveland, OH (USA)
Lines: 45
NNTP-Posting-Host: slc5.ins.cwru.edu


10:40:48 PM  2/19/93

THE ROSICRUCIAN FELLOWSHIP
ELECTRONIC BULLETIN BOARD
BBS

With your personal computer and a modem,
The Western Wisdom Teachings of the Brothers of the Rose Cross
are now as near as your telephone line!

Online services offered include:
--A free download library containing the entire written legacy of Max Heindel
--A continually updated calendar of events at Mount Ecclesia (the Fellowship's
Headquarters)
--Electronic mail services--send and receive messages to and from other 
students of the Rosicrucian Teacings on a worldwide basis.

The number is:

(619) 721-1828

For more information, contact the Telecommunication Team at Mount Ecclesia

The Rosicrucian Fellowship
2222 Mission Avenue Avenue
P.O. Box 713
Oceanside, CA  92049-0713
USA
Telephone:  (619) 757-6600
Facimile:  (619) 721-3806



Thanks very much!
And May the Roses Bloom Upon Your Cross!

Johnny Perko


-- 
Melesselva termaruva tennoio. ["They will love eachother forever."]           
"All your past except its beauty is gone, and nothing is left
   but a blessing."  [ACIM, T, 176]
"This is the purpose of life"  [Johnny and Tril: Garden Center.]

From gauss!romulus.reed.edu!nelson Tue Apr  6 22:11 PDT 1993
Received: by gauss; Wed Apr  7 01:11:46 EDT 1993
Received: by inet.att.com; Wed Apr  7 01:11 EDT 1993
Received: from romulus.reed.edu (reed.edu) by rand.org with SMTP id AA28362
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 6 Apr 1993 22:11:05 -0700
Received: from local by romulus.reed.edu (/\==/\ Smail3.1.25.1 #25.21)
        id <m0ngSPf-0003CTC@romulus.reed.edu>; Tue, 6 Apr 93 22:11 PDT
Message-Id: <m0ngSPf-0003CTC@romulus.reed.edu>
Date: Tue, 6 Apr 93 22:11 PDT
From: nelson@reed.edu (Nelson Minar)
To: voynich@rand.org
Subject: sci.crypt FAQ
Status: OR

Haven't heard much out of the mailing list in a few months, and the
ftp site's empty now.

The long awaited sci.crypt FAQ is out and it has no reference in it to
the Voynich manuscript. I asked and the FAQ maintainers are interested
in including a paragraph or two. They also implied that they already
had some information on the VMS (I imagine they've read Kahn, too :-),
but I was going to tell them about the mailing list and the ftp site.

Anyone have any strong opinions? What's up with the ftp site?

From gauss!mycroft.rand.org!jim Wed Apr 07 08:51:35 PDT 1993
Received: by gauss; Wed Apr  7 11:52:31 EDT 1993
To: gauss!reeds
Received: by inet.att.com; Wed Apr  7 11:52 EDT 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA11716
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 7 Apr 1993 08:51:38 -0700
Received: from localhost by mycroft.rand.org; Wed, 7 Apr 93 08:51:36 PDT
Message-Id: <9304071551.AA14089@mycroft.rand.org>
Cc: voynich@rand.org
Subject: Re: sci.crypt FAQ 
In-Reply-To: Your message of Tue, 06 Apr 93 22:11:00 -0700.
             <m0ngSPf-0003CTC@romulus.reed.edu> 
From: Jim Gillogly <jim@rand.org>
Reply-To: jim@rand.org
Date: Wed, 07 Apr 93 08:51:35 PDT
Sender: jim@mycroft.rand.org
Status: OR

Sorry about the ftp site -- it's still available for use, but I haven't
re-filled it since our disk crash; we don't back up the ftp volume, but I
still have everything that went in there on my personal disk, so nothing
is lost.  We were due to re-think what we wanted in there anyway -- any
suggestions?

	Jim Gillogly
	16 Astron S.R. 1993, 15:50

From gauss!eecs.umich.edu!kckluge Thu Apr 22 15:51:57 0400 1993
Received: by gauss; Thu Apr 22 15:52:37 EDT 1993
Received: by inet.att.com; Thu Apr 22 15:52 EDT 1993
Received: from krusty.eecs.umich.edu by rand.org with SMTP id AA05634
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 22 Apr 1993 12:51:28 -0700
Received: by krusty.eecs.umich.edu id AA17649 ; Thu, 22 Apr 1993 15:51:57 -0400
Date: Thu, 22 Apr 1993 15:51:57 -0400
From: Karl Kluge <kckluge@eecs.umich.edu>
Message-Id: <199304221951.AA17649@krusty.eecs.umich.edu>
To: voynich@rand.org
Subject: Vatican Library exhibition on-line
Status: OR


The Library of Congress has JPEG compressed images and the caption
text for the Vatican Library exhibit (along with some shareware
viewer software) available via anon. ftp to seq1.loc.gov, in the
directory vatican.exhibition. Thought this might be of some interest
to the group, particularly those who can't make it to DC to see the
exhibit.

Note the new email address. I don't have mail forwarding set up from my
CMU accounts, but I do read my email there periodically.

Karl

From gauss!world.std.com!kibo Fri May  7 01:11:16 0400 1993 remote from alice
Received: by gauss; Fri May  7 01:19:28 EDT 1993
Received: by inet.att.com; Fri May  7 01:19 EDT 1993
Received: from world.std.com by rand.org with SMTP id AA23819
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 6 May 1993 22:11:18 -0700
Received: by world.std.com (5.65c/Spike-2.0)
	id AA14103; Fri, 7 May 1993 01:11:16 -0400
Date: Fri, 7 May 1993 01:11:16 -0400
From: kibo@world.std.com (James "Kibo" Parry)
Message-Id: <199305070511.AA14103@world.std.com>
To: voynich@rand.org
Subject: Is anything happening re the Voynich?
Status: OR


I haven't seen any messages to the Voynich list recently, and I was
wondering if interest has slackened or something.

As for myself, I still want to finish up the PostScript Voynich calligraphic
font I've started, but it's still on a very back burner...

							-- K.

From gauss!andrew.cmu.edu!kc3u+ Sat May 15 19:10:18 0400 1993
Received: by gauss; Sat May 15 19:12:31 EDT 1993
Received: by inet.att.com; Sat May 15 19:12 EDT 1993
Received: from PO5.ANDREW.CMU.EDU by rand.org with SMTP id AA24128
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 15 May 1993 16:11:47 -0700
Received: by po5.andrew.cmu.edu (5.54/3.15) id <AA11191@X> for voynich@rand.org; Sat, 15 May 93 19:11:45 EDT
Received: via switchmail; Sat, 15 May 1993 19:11:42 -0400 (EDT)
Received: from pcs8.andrew.cmu.edu via qmail
          ID </afs/andrew.cmu.edu/service/mailqs/q000/QF.QfxLVQm00WBMA0U1Ui>;
          Sat, 15 May 1993 19:10:21 -0400 (EDT)
Received: from pcs8.andrew.cmu.edu via qmail
          ID </afs/andrew.cmu.edu/usr25/kc3u/.Outgoing/QF.wfxLVOe00WBMA3=1EB>;
          Sat, 15 May 1993 19:10:18 -0400 (EDT)
Received: from mms.0.1.873.MacMail.0.9.CUILIB.3.45.SNAP.NOT.LINKED.pcs8.andrew.cmu.edu.sun4c.411
          via MS.5.6.pcs8.andrew.cmu.edu.sun4c_411;
          Sat, 15 May 1993 19:10:18 -0400 (EDT)
Message-Id: <sfxLVOa00WBM83=19s@andrew.cmu.edu>
Date: Sat, 15 May 1993 19:10:18 -0400 (EDT)
From: Kendrick Kerwin Chua <kendrick+@cmu.edu>
To: Outbound News <outnews+netnews.alt.magick@andrew.cmu.edu>,
        Outbound News
    <outnews+netnews.alt.horror.cthulhu@andrew.cmu.edu>,
        Outbound News <outnews+netnews.alt.necromicon@andrew.cmu.edu>,
        Outbound News <outnews+netnews.alt.pagan@andrew.cmu.edu>,
        voynich@rand.org
Subject: Goodbye for now
Cc: 
Status: OR

Well, here's signing off for the summer. In the worst case, I will
probably not have any net access from now until September. In the best
case, I will be able to log on weekly from Florida. If by some chance
you need me or the Necronomicon FAQ for any reason, please do e-mail me
at this account address. I will try my best to address anyone's request
as soon as I am able. Best of luck to everyone.

Kendrick Kerwin Chua - kc3u+@andrew.cmu.edu  -or-  kendrick+@CMU.EDU
Dark Lord come forth and claim your world!  Godan Ktones!  Kibashen!

From gauss!reeds Thu Sep  2 13:48 EDT 1993
Received: by ninet.research.att.com; Thu Sep  2 13:53 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA16552
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 2 Sep 1993 10:51:15 -0700
Message-Id: <199309021751.AA16552@rand.org>
From: reeds@research.att.com
Date: Thu, 2 Sep 93 13:48 EDT
To: voynich@rand.org
Subject: Voynich Friedman transcription 
Status: OR

I recently visited the Marshall Library Lexington, VA, 
and spent most of a day in the Friedman collection, looking
at the Voynichiana.  I saw Friedman's photocopies of the VMS,
Petersen's hand copy (which was made by tracing in ink from
the photostats onto onion-skin paper), and much much besides.

One interesting thing was a pretty legibile copy of a printout
of a machine transcription made some time in the mid to late
1940s.  I typed in the first page: 

FGAG2,GDAE,AR,GHAM,SOE,SORG,0D0RC2,GDOR,SOE8G,
2ORG,DZAR,ORG,DAIR,THAM,SOR,AR,HZAR,HZAR,8ALA,
2GAIIR,SCDG,OR,GDAM,SO8,HZOARG,HZ2C,8ARM,2G,
0OM,OHCCG,OHCAR,ROEOHG,HZAAR,8AM,ODAM,OR,ODAL,
2AIRG,TCAR,HZAM,PZAR,FZAM,,G8ARAISG,

O8AR,SG,SOE,PZOG,OG8AR,S2,FZOAM,SO8ARG,
GSCG,SO8G,ODTOG,OHTOE,TOHZG,02TG,8AN,TOR,DO2,
8AM,SOR,FZOE,SO8G,,8AN,02,HCO8G

G8AN,PZC2AM,OE2,PZCG,GHAN,SOSG,PZO8AEC2,
ODSO,DSOG,OHAIRN,OHCOE,ODAL,SO8AN,2DZCG,8AM,
SOG,DZCG,DO8AM,PZG,PZ08AIIE2,HZCG,SC,OE8AN,8,
8AN,OM,TOE,O8AM,TO8AN,T8G,ODON,8OR,HZG,DO8,
8AM,SDZCG,DZOR,TOR,SCG,DOE,TOE,TOE,DOR,TOE,
SO,TOE,SO8AL,DSG,DTG,8OR,TO8AM,SO,DOCAK,
GTO,HTCG,TCDAN,SCO,PSOE,8G8G8,HZG,8AIHZG,
GHO,SOE,SC,DO8SCG,PZCAEG,8ARAN,8AN,DZG82,
8TAR,SHZAM,ODAIIR,TCG,2TG,PZHOE,HZOE2,8AOCHA,
SOD,TOR,TCG,8AN,DZCG,,OHOE,8AIM,

PZO,SAM,SODTCCG,TOE,HSO8CC2G,SCG,PG8CCG,TG,RO8,
00ON,TOE,8AN,HZAE,8AR,SCAR,DAM,8AR,SCG,HZCAR,
CTO00,DAM,SOAM,ODOE,8AM,FAR,HZOE,8AM,HZOE8AR,
GTCCG,ODCCG,ODG,8AM,ODTCG,DODAM,0TOE,DA8TG,8AE,
8TCO,SO8G,DOTCG,HZG,ODTCG,DCCG,DCCG,8AE,THOR.
0CO,TOE,TOD,TOHG,TOHCG,,8TAM

(I edited it slightly to be a bit more legibile.
I make no claims about the accuracy of my transcription.)


For comparison, here is how "voynich.now" has the same
page.



<f1r> {D'Imperio A}

<f1r.1> VAS92.9FAE.AR.APAM.ZOE.ZOR9.QOR92.9.FOR.ZOE89-
<f1r.2> 2OR9.XAR.O.R.9.FAN.ZPAM.ZAR.AR*.QAR.QAR.8AD-
<f1r.3> 29AU.ZCF9.OR.9FAM.ZO8.QOAR9.Q*R.8ARAM.29-
<f1r.4> [O82*]OM.OPCC9.OPCOR.2OEOP9.Q*AR.8AM.OFAM.OE.OFAD-{first char a plumed O?}
<f1r.5> 2AT.9.SCAR.QAM.WAR.YAM#
<f1r.6> 98ARAIZO#
<f1r.7> *.O8AR.*9.ZOE.WO9.O98AR.Z*.2.YOAM.ZO8A29-
<f1r.8> *Z9.SO89.OFSO9.OPSOE.SOQ9.O2.S9.8AN.ZOR.FO2-
<f1r.9> 8AM.ZO2.YOE.ZO89#
<f1r.10> 8AN.*OPCO89#
<f1r.11> *.98AN.WC2AM.OE.2.WC9.9PAN.ZOZ9.WO8AE.C2-
<f1r.12> OFZO.FZO9.OPA3.OPCOE.OFAD.ZO8AN.2X9.8AM-
<f1r.13> ZO9.XC9.FO8AM.W9.WO8AH2.QC9.ZO.OE8AN.8-
<f1r.14> 8AN.OM.SOE.O8AM.SO8AN.S89.OFAN.8*D.Q9.FO8-
<f1r.15> 8AM.ZXC9.XOR.SOR.ZC9.FOE.SOE.SOE.FOR.SOE-
<f1r.16> ZO.SOE.Z.O8AD.FZ9.FS9.8OR.SO8AM.ZO.FCCAJ-
<f1r.17> 9SO.PSC9.SCFAN.ZCOBZOE.80898.Q9.8AQ9-
<f1r.18> 9PO.ZOE.ZC.FO8ZC9.WCAE9.8ARAN.8AN.X9,,-
<f1r.19> 8SAR.ZQAM.OFAU.SC9.*S9.BOPOE.QOE2.8A8CPO-
<f1r.20> ZOF.SOR.SC9.8AN.XO9#
<f1r.21> OPOE.8A3#
<f1r.22> WO.ZAM.ZOFSCC9.SOE.PZO8CC29.ZC9.B98CC9.S9.RO.8*-
<f1r.23> *8ON.SOE.8AN.QAE.8AR.ZCAR.FAM.8AR.ZC9.QAR-
<f1r.24> SO.*O.FAM.ZOAM.OFOE.8AM.VAR.QOE.8AM.QOE8AR-
<f1r.25> 9SCC9.OFA9.OF9.8AM.OFSC9.F.OFAM.**SOE.F**S9.8AE-
<f1r.26> 8C*O.ZO89.FOZC9.Q9.OF.SC9.FCC9.8AE.SPOR-
<f1r.27> *O*.SOE.SOF.SOP9.SOPC9#
<f1r.28> 8SAM#



Jim Reeds


From gauss!cl.cam.ac.uk!Michael.Roe Fri Sep 03 01:27:43 +0100 1993
Received: by ninet.research.att.com; Thu Sep  2 20:29 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA02552
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 2 Sep 1993 17:27:58 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 3 Sep 1993 01:27:48 +0100
To: voynich@rand.org
Subject: A and B authors of the VMS
In-Reply-To: Your message of Thu, 02 Sep 93 13:48:00 -0400. <199309021751.AA16552@rand.org>
Date: Fri, 03 Sep 93 01:27:43 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:016130:930903002752"@cl.cam.ac.uk>
Status: OR


About a year ago, when Mary D'Imperio's and Prescott Currier's papers on the
two ''languages'' in the VMS were posted to the net, I tried to reproduce 
their results with different cluster analysis and discriminant anlaysis 
packages. The initial results I got seemed to confirm their findings, but I 
didn't got round to investigating much further. Anyway, we've recently 
installed a new version of the statistics package SAS, and this gave me an 
excuse to take a break from real work and do a few more statistics runs on the
VMS. (Just to check that the new version of SAS is installed properly, you 
understand... :-) )

Experiment 1:

Taking the folios for which the D'Imperio transcription ascribes an author,
I randomly divided them into two groups of equal size. The first group was used
to compute a disciminant function for distinguishing between the 'A' and 'B'
authors, using the per-folio letter frequencies as variables.

This discriminant function was then applied to the second group, and
its predictions compared with the authorship actually ascribed by D'Imperio.

The result was a 95% success rate in predicting ``authorship'' based on letter
frequencies alone. This is fairly strong evidence that the ``authorship''
assigned by D'Imperio is really correlated with the letter frequencies. [OK,
so there's probably some proper way of putting a confidence interval on this -
but this seems good enough] That is, the two ``languages'' are a real property 
of the data, and not a figment of our imaginations.

Experiment 2:

Using all the folios for which D'Imperio ascribes authorship, compute two
orthogonal linear functions of the per-folio letter frequencies which do best
at separating the 'A' and 'B' samples. (Canonical Discriminant). Then use
these two functions as the X and Y co-ordinates for a scatter diagram plot
of all the folios, including the ones which have not been ascrined to either
the 'A' or 'B' authors.

The plot has two quite distinct blobs, one of which only contains 'A' points
and the other only contains 'B' points. There is also a third, more diffuse
blob, separate from either of these. Tracing  these points back to the
transcription, it turns out that all of these points correspond to folios
transcribed by Jim Reeds rather than by Mary. That is, there are strong
statistical differences between transcriptions produced by different
transcribers. Of course, this is to be expected, but it was nice that just
by looking at the plot I could say ''these are different somehow'' and then
find a reason for it.

Conclusions:

(a) The 'A' and 'B' folios are genuinely different. However it does not 
    necessarily follow that they are by different *authors*. Using a different
    transcriber produces statistical differences which are just as great as
    those between the 'A' and 'B' folios. Perhaps the 'A' and 'B' folios are
    the result of the VMS being a copy of some lost original, made by two 
    different copyists?

(b) It is important to be careful when merging transcriptions produced by
    different transcribers.

(c) This sort of statistical approach seems worth pursuing.

A question:

Presumably Mary D'Imperio and Prescott Currier assigned an author for all of
the folios.  Does anyone know which author the other folios are ascribed to? 
(i.e the folios that were not in Mary's original transcription).

It would be possible to make a good guess at 'A' or 'B' authorship just from
''letter'' and ''word'' frequencies, but it would be nice to have the 
information from the handwriting analysis too.

Mike


From gauss!nyx.cs.du.edu!rcarter Fri Sep  3 00:06:35 MDT 1993
Received: by ninet.research.att.com; Fri Sep  3 02:07 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA10740
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 2 Sep 1993 23:05:49 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA09543; Fri, 3 Sep 93 02:05:46 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA12587; Fri, 3 Sep 93 00:06:35 MDT
Date: Fri, 3 Sep 93 00:06:35 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309030606.AA12587@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
To: Michael.Roe@cl.cam.ac.uk, voynich@rand.org
Subject: Re:  A and B authors of the VMS
Status: OR

RE: handwriting analysis and authorship of the VM...

Well, the argument by some (many?) has been that the author of the VM
is John Dee; someone (not on Team Voynich) had done some handwriting
analysis that they thought showed a connection...

I disagree strongly with both points, though; by looking at John Dee's
own catalog of his library, there are several points that disuade me
from any real resemblance of the Vm and Dee's handwriting; not only
that, but Dee has some peculiarities in his handwriting that do -not-
show up in the VM...

I think there is only the slimmest of chances that Dee ever knew of
the VM, the only possible link is that of Dee and Rudolph; again,
tenuous at best.

I agree that the VM is that of at least 2 transcribers; and I tend to
think that it was transcribed by dictation and translation.

Regards, Ron | rcarter@nyx.cs.du.edu

From gauss!reeds Fri Sep  3 11:08 EDT 1993
Received: by ninet.research.att.com; Fri Sep  3 11:12 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA22321
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 3 Sep 1993 08:10:32 -0700
Message-Id: <199309031510.AA22321@rand.org>
From: reeds@research.att.com
Date: Fri, 3 Sep 93 11:08 EDT
To: voynich@rand.org
Subject: Voynich A and B and Dee 
Status: OR

Mike Roe describes cluser analysis of VMS pages, looking for A and B
clusters.  In the process a third cluster arises:

	There is also a third, more diffuse
	blob, separate from either of these. Tracing  these points back to the
	transcription, it turns out that all of these points correspond to folios
	transcribed by Jim Reeds rather than by Mary. That is, there are strong
	statistical differences between transcriptions produced by different
	transcribers. 
	
One difference between my transcription style and Mary's is this:  I annotated
my transcriptions with descriptions of the page layout, variant readings, and
reliability assessments.  Is it possible, Mike, that your discriminant
functions are picking up such things?  My transcriptions have more comments,
hence more { and } characters, and more lower case letters.

Of course 'substantive' transcription bias is possible, too, and Mike's comment
that it is important to be careful when merging transcriptions produced by
different transcribers is 100% correct.

===

My colleage Ken Church, a linguist with a lexicostatistical bent, has invented
another graphical way of displaying similarities between subpassages of
an extended text.  (Unfortunately, his software is hard to use, so I cannot
yet do it without active assistance from a very busy VMS non-fanatic.
So I have not been able to play around with this technique.)  The idea
is very simple.  Suppose the successive words in the VMS are w1, w2, ... , wN.
If the i-th word in the text, wi, is the same as the j-th word, wj, put 
a speck of black ink at the (i,j) place on a big sheet of paper.  Stand back
and stare at the resulting picture.  In the case of the VMS you see a checker-
board, with a dark diagonal line (all pixels (i,i) are black, because wi = wi.)
The checkerboard comes from the alternation of A and B.  You can put all the
A pages first, then all the B pages, and repeat the process.  Then you see
a big 'quartered flag', with dark diagonal.  But inside of each block you
see more texture, some of which I was able to ascribe to change of 'topic':
herbal B versus biology B, and so on.  Similar results hold when you do the
same to the Bible:  you see the checkerboard block structure induced
by the books of the Bible, which have varying language statistics.  This
kind of dot plot is in effect a kind of similarity matrix, showing how much
like each part of the input text is like each other part of the text.

There are many practical details left out of the above:  one can only plot
a 2000 by 2000 pixel matrix on common laserwriters, so one has to use
some kind of interpolation.  One can fiddle with the definition of 'word'.
One can require words i, i+1, and i+2 to be equal, respectively, to words
j, j+1, and j+2.  One can apply the technique to any filtered version of
the text, and so on & so on.  Try it yourself.

===



Mike asks for Currier's 'global' assignment of pages to language and hand:

	Does anyone know which author the other folios are ascribed to? 
	(i.e the folios that were not in Mary's original transcription).

This can be found in Table A of Currier's seminar paper.  It is available
in both ASCII and in Postscript; if you cannot find it on rand I'll send
you a copy.  The information is copied into my 'foliation' document, which
every serious VMS scholar should have.

===

Ron Carter mentions the Dee conection:

	Well, the argument by some (many?) has been that the author of the VM
	is John Dee; someone (not on Team Voynich) had done some handwriting
	analysis that they thought showed a connection...

As I recall, A. G. Watson, one of the editors of Dee's recently published 
library catalogue noted only that the handwritten folio numbers on the upper
right hand corners were Dee's.  The folio numbers are much larger than the
letters in the text, have a different ink, and were written with a broader pen.
Since the folio numbers do not change handwriting as you go from A to B, it
seems clear that they were written after the book was assembled.  Dee as
owner, adding his own folio numbers, but not Dee as author or scribe.

The Yale catalog entry for the VMS (Beineke 408) cites these Dee manuscripts
as having the same handwriting on the folio numbers:  Bodleian Library, 
Ashmole 1790, f9v, and Ashmole 487.

===

Jim Reeds


From gauss!mycroft.rand.org!jim Fri Sep 03 09:06:44 PDT 1993
Received: by ninet.research.att.com; Fri Sep  3 12:07 EDT 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA24914
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 3 Sep 1993 09:06:45 -0700
Received: from localhost by mycroft.rand.org; Fri, 3 Sep 93 09:06:44 PDT
Message-Id: <9309031606.AA19493@mycroft.rand.org>
To: voynich@rand.org
Subject: Voy -- Reeds' foliation paper
In-Reply-To: Your message of Fri, 03 Sep 93 11:08:00 -0400.
             <199309031510.AA22321@rand.org> 
From: Jim Gillogly <James_Gillogly@rand.org>
Reply-To: James_Gillogly@rand.org
Date: Fri, 03 Sep 93 09:06:44 PDT
Sender: jim@mycroft.rand.org
Status: OR


> reeds@research.att.com writes:
> Mike asks for Currier's 'global' assignment of pages to language and hand:
> 
> Does anyone know which author the other folios are ascribed to? 
> (i.e the folios that were not in Mary's original transcription).
> 
> This can be found in Table A of Currier's seminar paper.  It is available
> in both ASCII and in Postscript; if you cannot find it on rand I'll send
> you a copy.  The information is copied into my 'foliation' document, which
> every serious VMS scholar should have.

Yes, this document is absolutely essential; besides the A/B information
with Currier's further breakdowns, it includes what pages have been
transcribed, which are available on the British Library microfilm, which
have been reproduced in other available documents (note that some of
reproductions in the literature are mis-numbered), notes from D'Imperio on
page contents, and "key-like" sequences.

It's available in the rand.org anonymous ftp area as:

	rand.org:pub/voynich/checklist

While we're on the subject please send me notes on anything you notice
missing from the archive: we had a "disk event", but all the files that
were there are still accessible to me, but not necessarily in the archive.

	Jim Gillogly
	12 Halimath S.R. 1993, 16:04

From gauss!cl.cam.ac.uk!Michael.Roe Fri Sep 03 20:01:42 +0100 1993
Received: by ninet.research.att.com; Fri Sep  3 15:04 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA03281
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 3 Sep 1993 12:01:53 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 3 Sep 1993 20:01:46 +0100
To: voynich@rand.org
Subject: Re: Voy -- Reeds' foliation paper
In-Reply-To: Jim Gillogly's message of Fri, 03 Sep 93 09:06:44 -0700. <9309031606.AA19493@mycroft.rand.org>
Date: Fri, 03 Sep 93 20:01:42 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:220830:930903190149"@cl.cam.ac.uk>
Status: OR

> This can be found in Table A of Currier's seminar paper.  It is available
> in both ASCII and in Postscript; if you cannot find it on rand I'll send
> you a copy.  The information is copied into my 'foliation' document, which
> every serious VMS scholar should have.

I had the 10th Dec '91 version of the foliation, which didn't give the 
''author'' for all the folios. I've FTP'd a more recent version of the
foliation from RAND --- thanks! I see that even the most recent 
version of the foliation doesn't give an ''author'' for the astrological 
folios 67 to 73; is this because the astrological folios contain too little
text to decide whether they're 'A' or 'B'?

Also, did we ever decide how to transcribe a ``picnic table'' (Froggy 'n')?
I guess just 'n' would do, as this doesn't clash with any of the Currier
notation. (I've just been looking at f66r, which has this character and a few
other wierdos in the margin).

Mike


From gauss!reeds Fri Sep  3 16:30 EDT 1993
Received: by ninet.research.att.com; Fri Sep  3 16:33 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA06917
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 3 Sep 1993 13:32:15 -0700
Message-Id: <199309032032.AA06917@rand.org>
From: reeds@research.att.com
Date: Fri, 3 Sep 93 16:30 EDT
To: voynich@rand.org
Subject: Voynich folios 67-73 A or B, and Picnic Table
Status: OR

Mike Roe points out that 

> even the most recent 
> version of the foliation doesn't give an ''author'' for the astrological 
> folios 67 to 73; is this because the astrological folios contain too little
> text to decide whether they're 'A' or 'B'?

I suppose so.  I will update the checklist with some minor new data gleaned
from Currier's Table A and from my new set of copies of Friedman's photostats.

Mike also asks

> Also, did we ever decide how to transcribe a ``picnic table'' (Froggy 'n')?
> I guess just 'n' would do, as this doesn't clash with any of the Currier
> notation. (I've just been looking at f66r, which has this character and a few
> other wierdos in the margin).

My latest V font (rand.org:/pub/voynich/voyfont.ps.sh.Z)  calls it X48, but
I know that that name will not catch on!  I think 'n' is fine.



Jim Reeds

From gauss!reeds Sun Sep  5 00:35 EDT 1993
Received: by ninet.research.att.com; Sun Sep  5 00:38 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA06713
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 4 Sep 1993 21:35:31 -0700
Message-Id: <199309050435.AA06713@rand.org>
From: reeds@research.att.com
Date: Sun, 5 Sep 93 00:35 EDT
To: voynich@rand.org
Subject: Voynich Friedman Collection Goodies
Status: OR

I spent much of the day looking at my new copies of Friedman's copies
of Petersen's copies of the VMS.  In many cases they are clearer than
the Yale copyflow; in many not.  Three interesting things:

	1.  Someone has carefully written folio numbers like f68r1 
	on the negative (I think) so the folio numbers are right there,
	part of the photograph.  These are the numbers D'Imperio used
	in her book, which I tried to put in my 'checklist' file
	and which I wrote in little rectangular boxes in page corners
	of my Petersen xeroxes.  It was gratifying to learn than I
	had only made one (or four, depending on how you count) mistake:
	I had interchanged f85/86r5 with f85/86r6, and similarly f85/86v5
	with f85/86v6.

	2.  Someone, perhaps Petersen, had numbered the photostats serially,
	and the numbers agree with the page numbers used by D'Imperio in
	her computer transcription.  Alas, these numbers suffer a phase
	hiccough at page 206 or so with respect to the page numbers in
	Currier's Table A.

	3. The curved chunk taken out of f114 was taken out between 1931
	and 1976:  someone trimmed a bit of worm-eaten margin away.

I also gazed at my sample pages of xeroxes of the "First Study Group"
transcription, which take up between 100 and 200 sheets of printout of
card images, 46 cards per sheet, 30 Voynich characters per card, moderately
copyedited, dated some time in the period 1946-1949.  I indeed plan to
get a xerox of the whole printout, and shop around for a commercial data-
entry shop to put in on diskette.  (What are the going rates for typing in
data? 100 pages times 46 lines times 30 chars = 138000, should be a few days
to type in, I suppose.  I will get an estimate and report back.)  Once 
typed in, it is a cinch to convert it to Currier (which is derived from 
1SG, after all), and mechanically spot differences.

Jim Reeds

From gauss!cl.cam.ac.uk!Michael.Roe Mon Sep 06 23:35:36 +0100 1993
Received: by ninet.research.att.com; Mon Sep  6 18:37 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA11512
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 6 Sep 1993 15:35:50 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Mon, 6 Sep 1993 23:35:40 +0100
To: voynich@rand.org
Cc: Michael.Roe@cl.cam.ac.uk
Subject: Identification of BL photocopies of the VMS
Date: Mon, 06 Sep 93 23:35:36 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:101700:930906223545"@cl.cam.ac.uk>
Status: OR


The British Library has two collections of photocopies of the Voynich
Manuscript; The first (Catalgoue FACS 439) is random collection of 
unidentified folios manuscript. The second (FACS 461) is folios 1r to 56r 
inclusive.

I've just been looking through the random ones (FACS 439) and trying to 
identify them from their first lines using Jim Reed's folio finder (finder.txt).

Provisional identification is as follows:

Microfilm   Provisional
Frame No.   Identification
=========   ===============
1           (Catalogue entry for FACS 439)
2           68r1/68r2 (part of 68r3 visible)
3           65v/66r
4           78v/79r
5           107v/108r
6           108v/111r
7           111v/112r
8           112v/113r
9           1r
10          116v

The folio numbers on the MS itself are clearly visible in the top right
hand corners of 108r, 111r, 112r and 113r. 

(I given microfilm frame numbers because I'm working from a microfilm of
the BL photocopies, rather than the photocopies themselves)

Mike

From gauss!trl.oz.au!j.guy Tue Sep  7 09:50:16 +1000 1993
Received: by ninet.research.att.com; Mon Sep  6 19:51 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA12966
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 6 Sep 1993 16:50:25 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA17410; Tue, 7 Sep 93 09:50:19 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309062350.AA17410@medici.trl.OZ.AU>
Subject: "First Study Group" cards
To: voynich@rand.org
Date: Tue, 7 Sep 1993 09:50:16 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 111       
Status: OR

I volunteer to type them in.

Which pledged, it's good to see the Voynich group
rising from its ashes again. 


From gauss!cl.cam.ac.uk!Michael.Roe Fri Sep 10 15:38:28 +0100 1993
Received: by ninet.research.att.com; Fri Sep 10 10:41 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA14258
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 10 Sep 1993 07:36:11 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 10 Sep 1993 15:38:36 +0100
To: voynich@rand.org
Cc: Michael.Roe@cl.cam.ac.uk
Subject: Cluster analysis and the Voynich MS
Date: Fri, 10 Sep 93 15:38:28 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:170620:930910143846"@cl.cam.ac.uk>
Status: OR


In a previous message I said that some of the recently transcribed folios
had a different letter-frequency distribution from either the ''A'' or ''B''
folios as transcribed by Mary D'Imperio. At the time, the explanation that 
came to mind was that different transcribers might tend to make different
types of error.

I now have an alternative hypothesis. Mary's transcription deliberately avoids
the pages with lots of labels, particularly the astrological ones. The pages
we've transcribed since have been in the gaps in Mary's transcription, and
hence have been for the most part pages with lots of labels. I think the
letter-frequency distribution of the labels really is different from that
of either the ''A'' or ''B'' text. 

A second observation: the letter frequency distribtion of f31r/v is also
noticeably different from the majority of either 'A' or 'B' text. Looking at
the microfilm, the handwriting on f31 is very different from the preceding
and following folios (believed to be ''A''). I'm beginning to believe that
there are at least three different authors, each with their own charcteristic
handwriting, word-usage patterns and style of art-work. (I'm sure there are
at least two).

Mike

From gauss!reeds Fri Sep 10 11:40 EDT 1993
Received: by ninet.research.att.com; Fri Sep 10 11:43 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA17193
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 10 Sep 1993 08:36:45 -0700
Message-Id: <199309101536.AA17193@rand.org>
From: reeds@research.att.com
Date: Fri, 10 Sep 93 11:40 EDT
To: voynich@rand.org
Subject: Voynich folios 31r/31v
Status: OR

I like Mike's theory that label statistics are different from regular textual
statistics, and that this effect is confounded with transcriber.  It should
be possible to test by repeating the computations on the label-rich pages 
with the label data deleted....

The observation about 31r/31v being different is interesting.  Gathering
(or quire) 4 is formed by 4 sheets, folded to make 8 leaves.  Folio 31
is part of a sheet, the other part of which is f.26.  Do you think that
f.26r and f.26v are in the same hand/dialect as f.31r and f.31v?  I don't
think Currier knew which folios were paired up with which others in the
quires, but his assigment to A and B is consistent with the attractive "one
sheet, one language" doctrine.

Jim Reeds


From cl.cam.ac.uk!Michael.Roe Fri Sep 10 18:10:19 +0100 1993
Received: by ninet.research.att.com; Fri Sep 10 13:13 EDT 1993
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 10 Sep 1993 18:10:24 +0100
To: reeds@research.att.com
cc: voynich@rand.org
Subject: Re: Voynich folios 31r/31v
In-reply-to: Your message of Fri, 10 Sep 93 11:40:00 -0400. <199309101536.AA17193@rand.org>
Date: Fri, 10 Sep 93 18:10:19 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-ID: <"swan.cl.cam.:196140:930910171027"@cl.cam.ac.uk>
Status: OR

> The observation about 31r/31v being different is interesting.  Gathering
> (or quire) 4 is formed by 4 sheets, folded to make 8 leaves.  Folio 31
> is part of a sheet, the other part of which is f.26.  Do you think that
> f.26r and f.26v are in the same hand/dialect as f.31r and f.31v?  

>From the letter frequencies, f26r/v is in the ``standard'' B dialect, while
f31r/v is different from both the A and B dialects. (Note that most of the 
folios in the range 20-30 are in the A dialect).

I'm not sure about the handwriting; it's hard to compare two pages that are
widely separated on the microfilm, as you really need to put two pages 
beside each other to see the difference clearly. Thus, f26 differs from
f24 and f25, and f31 differs from f29 and f31, but it's hard to tell how similar
26 and 31 are.

Mike

From gauss!reeds Fri Sep 10 23:56 EDT 1993
Received: by ninet.research.att.com; Fri Sep 10 23:58 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA13728
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 10 Sep 1993 20:56:56 -0700
Message-Id: <199309110356.AA13728@rand.org>
From: reeds@research.att.com
Date: Fri, 10 Sep 93 23:56 EDT
To: voynich@rand.org
Subject: Voynich transcription
Status: OR

Today my care package from the Friedman collection came in the mail.

There are about 140 pages of item 1609, a printout of an IBM card
transcription of almost all of the VMS, done in the "First Study
Group" alphabet found on D'Imperio, page 97.  I estimate 190,000
characters of data (not counting line numbers).  Some of it might
be automatically scannable.

It is somewhat awe-inspiring to see IBM card printouts made before
I was born.

About 1/4 or 1/3 has been proofread & copyedited.  I think the printout
was made in 7 Jan 1946, and the proofreading was done in 27 March 1949.
According to D'Imperio, the FSG disbanded in 1946 with the post war
demobilization; maybe WFF filled a rainy week in 1949 proofreading
the old FSG printout.

And item 1613 consists of 20 pages of another transcription (in
an extremely legible hand, writing on 1/4" graph paper), with no
date that I can read, by "Francis M. Pritchard ?", which seems to
be ff 112, 113, and 114.  It is in a strange transcription alphabet,
which I have not figured out yet.  I am part way through typing it in.
The first page is labeled "folio 112 verso", and reads:

	DJSQE,OJDFP,RWQFP,JASQJNFTG,FPDFL,SQJDFL
	,RQJD,JDQFPSQNE,GFPRE,SQQDQQE,RQJP,DQQNE
	,OJDFTG,SQUE,OJDQJP,JDQRQNE,OJDQNE,OJARQ
	NE,OJDQNFHRE,NFL,RQNE,SQNFP,JCQNE,JCQQJ,
	RQNE,OJDQQE,NFL,RUFG,JCFG,OJCFL,JCFTH,NF
	M,RQE,RJDQNFTG,FTG,FPJPSQNE,OJCQNE,OJDQN
	FG,JCQQNE,JDQNE,RQNE,PDFH,OJD0N,J,FM,JCQ
	NE,OJDQNE,RQNE,OJDQQE,OJDQQNE,OJDQJP,SQN
	E,OJCQNFP,PJP,JDQQJ,JDRFG,JCQQNE,RQFP,OJ
	DRE,PDQQE,RJP,P,DFM,JCQQN,FM,FI,FPDQQNEH
	,OJDQQRQE,OJDFM,JCQFM,BRQQJ,PDFL,JDQQE,P
	,PRQQNE,JCFP,DFL,RQNFL,FP,SJ,JCRQE,RQJP,
	DQRE,RUQE,JDFL,FP,PDQQE,OJDFL,RQWE,RQNFG
	,FH,PSQE,SQNFP,JDQFG,P,DQNE,RQJPDQQE,PDQ
	E,JDQQE,OJDQQE,OJDQE,OJDFL,FH,IJM,SQN,OJ
	DSQJ,PJG,RQJ,PJP,FM,SQE,OJDFL,RQFG,OJCQJ
	P,SUE,PNE,EPRQQE,SQFG,FM,SQQFG,JCFL,JCQN
	E,OJDFL,RQNDFL,SQNFTCFL,RFPN,OJFG,FL,SQE
	,JDRQE,OJDFM,FQWE,SJ,PRFP,SQFE,SWQE,DSFG
	CFG,SQFM,JDFM,RWE,RQJP,DQNE;RQCFG,JDFM,S

Can anybody spot just where in the VMS this is, and tell what
the transcription alphabet is?  I leave this as a puzzle, maybe
the first simple substitution puzzle cryptogram in the world whose
plaintext is in Voynicheese!

Jim Reeds


From gauss!umcc.umich.edu!bgrant Sat Sep 11 11:42:56 0400 1993
Received: by ninet.research.att.com; Sat Sep 11 11:44 EDT 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA22079
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 11 Sep 1993 08:43:00 -0700
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0obX6H-000AAjC@umcc.umich.edu>; Sat, 11 Sep 93 11:42 EDT
Message-Id: <m0obX6H-000AAjC@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: A/B Hand Letter Frequencies
To: voynich@rand.org
Date: Sat, 11 Sep 1993 11:42:56 -0400 (EDT)
X-Mailer: ELM [version 2.4 PL22]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 555       
Status: OR

	There is another possible explanation for the differing letter 
frequency distributions in the A and B hands other than errors: that
two scribes were working from plaintext, encrypting as they went and
from time to time inserting null characters.  The choice of which null
character to use could be idiosyncratic for each scribe without affecting
the meaning of the text.
	If this is the case, the different distributions could be used 
as a tool to determine which characters are nulls and to recover the text
without nulls.

Bruce Grant (umcc.ais.org)

From gauss!nyx.cs.du.edu!rcarter Sat Sep 11 11:05:46 MDT 1993
Received: by ninet.research.att.com; Sat Sep 11 13:06 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA23250
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 11 Sep 1993 10:05:01 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA11130; Sat, 11 Sep 93 13:04:56 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA28820; Sat, 11 Sep 93 11:05:46 MDT
Date: Sat, 11 Sep 93 11:05:46 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309111705.AA28820@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
To: bgrant@umcc.umich.edu, voynich@rand.org
Subject: Re:  A/B Hand Letter Frequencies
Status: OR

I am willing to rehash some (old) pet theories...  :-)

I have theorized for awhile that the VMs was the result of at least two
scribes taking -dictation- and transcribing the document in question in
a phonetic shorthand.

Other than the shorthand, which -would- act as an encoding process of
sorts, I think the VMs is -plaintext- and varies because the skills or
knowledge of the shorthand varies between scribes.

In my scenario, there might be more than one person -giving- dictation.

The next level down (of speculation/fantasy  :-) of my theory has the
original document in Nuahatl; a herbal/biological/astronomical holy grail
of sorts...

Wait a second!!!  Hmmm, the only culture that -did- have a real strength
in these areas -were- the Mayans; with all the flowing blood, the star
charts, -and- herbals, this theory might not be too fantastic; of course,
after our cloister of monk-scientists bastardized the drawings beyond any
recognition (to give a break to the VMs draftsman, he was working from
drawings of plants he was unfamiliar with, and judging from the drawings,
his expertise(?) was the human (albeit European females) form, and he
probably thought the star drawings were just random dots, etc.)  there
might not be very many clues left to point out the Mayan origins...

*sigh*  Ron | rcarter@nyx.cs.du.edu | 11 SEP 93 | Denver, CO USA

From gauss!umcc.umich.edu!bgrant Sat Sep 11 19:58:38 0400 1993
Received: by ninet.research.att.com; Sat Sep 11 20:00 EDT 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA29987
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 11 Sep 1993 16:58:41 -0700
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0obepz-000AAnC@umcc.umich.edu>; Sat, 11 Sep 93 19:58 EDT
Message-Id: <m0obepz-000AAnC@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: Re: Re: A/B Hand Letter Frequencies
To: voynich@rand.org
Date: Sat, 11 Sep 1993 19:58:38 -0400 (EDT)
X-Mailer: ELM [version 2.4 PL22]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1348      
Status: OR

If I understand Ron Carter's scenario, it involves one guy dictating
alternately to two scribes who take down his words in shorthand and then
make (a different set of) mistakes expanding it to the plaintext which
has come down to us.

>From the pattern of pages, though, I get the idea that both scribes were
working together, grabbing a few pages at a time from a master copy of the
plaintext.  For one thing, no page has writing by both scribes, and the 
same scribe writes both sides of a sheet.  After scribe A writes the first
48 pages, they alternate a few pages at a time, for instance:

	pages 49-50 scribe B
	pages 51-58 scribe A
	pages 59-60 scribe B
	pages 61-62 scribe A
	pages 63-66 scribe B

	    etc.

It makes more sense to me that they would be divvying up a stack of 
sheets of plaintext and coming back for the next batch when they finish
enciphering the plaintext and adding random nulls, rather than taking turns
listening to a guy who alternates between them.  (If the guy
is dictating it shouldn't take too long to do a page, so if the two 
scribes are spelling each other they must be dashing in and out of the 
room pretty often.)

As far as the Mayans are concerned, my guess is that they got the herbal
information from the Egyptians while arranging to borrow their pyramid-
building machines.

Bruce Grant (umcc.ais.org)

From gauss!nyx.cs.du.edu!rcarter Sat Sep 11 18:38:07 MDT 1993
Received: by ninet.research.att.com; Sat Sep 11 20:38 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA00600
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sat, 11 Sep 1993 17:37:19 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA28948; Sat, 11 Sep 93 20:37:17 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA25621; Sat, 11 Sep 93 18:38:07 MDT
Date: Sat, 11 Sep 93 18:38:07 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309120038.AA25621@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
To: bgrant@umcc.umich.edu, voynich@rand.org
Subject: Re: Re: A/B Hand Letter Frequencies
Status: OR

Not quite Bruce, but close...  My scenario has one person dictating
to one person first, who copies it down in a phonetic shorthand, and
then another person takes over for the first transciber and introduces
his form of the shortahnd, and this one time dictation/transcription
is what we have in the VMs.

An example: I dictate to you, in Mayan, the word "quexlcoatl" (winged
snake) but you, noy knowing Mayan (Nuahatl) write it down not only in
phonetic form, but also in (let's say) a shorthand phonetic form.

It might turn into (in English) "ke)q(" or some such (seemingly)
nonsensical stuff...  But someone else, trained in a similar form of
phonetic shorthand (but with less experience) might produce "kwe)qa(".

I propose that the VMs is a phonetic shorthand in abbeviated Latin
transcribed by at least two people from a person dictating in Nuahatl.

Now, just to prove it, eh?  :-)

Hmmm, UFOs. pyramids, and Mayans...  If you don't think the Mayans
produced herbals, think again; they were ahead of Europeans on herbals.

Ron | rcarter@nyx.cs.du.edu | 11 SEP 93 | Denver, CO  USA

From gauss!mycroft.rand.org!jim Mon Sep 13 10:15:55 PDT 1993
Received: by ninet.research.att.com; Mon Sep 13 13:30 EDT 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA13105
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 13 Sep 1993 10:15:53 -0700
Received: from localhost by mycroft.rand.org; Mon, 13 Sep 93 10:15:56 PDT
Message-Id: <9309131715.AA06880@mycroft.rand.org>
To: voynich@rand.org
Subject: Shorthand, Nahuatl or otherwise [Re: A/B Hand Letter Frequencies]
In-Reply-To: Your message of Sat, 11 Sep 93 18:38:07 -0600.
             <9309120038.AA25621@nyx.cs.du.edu> 
From: Jim Gillogly <James_Gillogly@rand.org>
Reply-To: James_Gillogly@rand.org
Date: Mon, 13 Sep 93 10:15:55 PDT
Sender: jim@mycroft.rand.org
Status: OR

> rcarter@nyx.cs.du.edu (Ron Carter) writes:
> I propose that the VMs is a phonetic shorthand in abbeviated Latin
> transcribed by at least two people from a person dictating in Nuahatl.
> 
> Now, just to prove it, eh?  :-)

I have lots of problems seeing the VMs as straight shorthand.  One is that
shorthand characters tend to be easier to draw than garden-variety
characters.  The "9" for "-US" is fine -- easy to write and distinguish --
but the reasonably common characters B and V, and the not-so-common but
still consistent characters Q, W, X, and Y are quite complex to find in a
set of symbols intended to speed or shorten written text.  Given that
there are only about three dozen characters, I would expect somebody
designing a shorthand for them to use letters no more complicated than the
equivalent Latin ones.

Mind you, I don't know beans about Nahuatl, and one can imagine someone
familiar with Latin abbreviations feeling he had to choose symbols not
already in use to represent the non-Latin sounds of that language.  Could
Q represent "-XT-" or something?  Who knows...

Another problem I have with it is that shorthand tends to increase
variability by squeezing out some of the redundancy in the language.
Voynich Language is <more> redundant than the European languages I've
compared it with.  I think Jacques has exhibited some minimalist
phonologies with stats that could match it, so again this doesn't
eliminate the Nahuatl conjecture... but in my view decreases the
likelihood of straight European shorthand.

If one were to assume it <is> in shorthand directly encoding some language
phonetically, could one deduce anything from the non-linguistic-looking
bits, such as words repeated too often to look natural?  For example, what
does one make of:

	<f8v.8> OFSOEFZ.SOE.SOE.SOE.QAM.8AN-

Some suggestions have been repeated adjectives for emphasis (very very
very) or pious declarations (amen amen amen); maybe a Nahuatl book of
natural history would have names with repeated elements, like the Hawaiian
reef trigger fish that I'm wearing on my t-shirt (humuhumunukunukuapua'a).

In any case, it's a testable theory... Ron, have you found any Nahuatl
text to run a statistical comparison on?  Is its phonology as rich as
European languages?

	Jim Gillogly
	Mersday, 22 Halimath S.R. 1993, 17:10

From gauss!reeds Mon Sep 13 16:52 EDT 1993
Received: by ninet.research.att.com; Mon Sep 13 16:56 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA22431
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 13 Sep 1993 13:53:44 -0700
Message-Id: <199309132053.AA22431@rand.org>
From: reeds@research.att.com
Date: Mon, 13 Sep 93 16:52 EDT
To: voynich@rand.org
Subject: Voynich Study Group, First
Status: OR

I just finished typing in 6 VMS pages transcribed by the First Study Group
more than 45 years ago, from 20 pages of key punch coding sheets, which
are item 1613 in the Friedman collection.  It was interesting because the
transcription alphabet was not the one listed in D'Imperio's Fig 19
for the FSG.  Rather, reading down in the FSG column, instead of
Roman P,F,H,G,... this transcription used A, B. C, D, ...  I conjecture
that this was an early form of FSG, used after the basic stock of V letters
had been decided on, but before the easy-to-remember equivalents like A, 2, O
had been settled on.


Jim Reeds


From gauss!cl.cam.ac.uk!Michael.Roe Tue Sep 14 13:49:39 +0100 1993
Received: by ninet.research.att.com; Thu Sep 16 11:51 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA22286
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 08:50:09 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Tue, 14 Sep 1993 13:49:45 +0100
To: reeds@research.att.com
Cc: voynich@rand.org
Subject: Pritchard Voynich alphabet (spoiler)
In-Reply-To: Your message of Fri, 10 Sep 93 23:56:00 -0400. <199309110356.AA13728@rand.org>
Date: Tue, 14 Sep 93 13:49:39 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:267390:930914124948"@cl.cam.ac.uk>
Status: OR

> Can anybody spot just where in the VMS this is, and tell what
> the transcription alphabet is?  I leave this as a puzzle, maybe
> the first simple substitution puzzle cryptogram in the world whose
> plaintext is in Voynicheese!

Those members of this group who want to try this a a puzzle, beware! This
message contains the solution.














Using a table of frequently-occuring Voynich ``words'' it was possible
out the following correspondances:

Pritchard  A B C D E F G H I J K L M N O P Q R S T U V W X
Currier    B V P F 9 A R J   O   N M 8 4 E C S Z I 

This looks very much like the First Study Group alphabet, but with the symbols
labelled A-Z instead of given mnemonic names. Filling in the gaps:

Pritchard  A B C D E F G H I J K L M N O P Q R S T U V W X
Currier    B V P F 9 A R J 2 O D N M 8 4 E C S Z I W Y Q X

The transcription is of f111v, not "folio 112 verso" as claimed.  (I've 
compared it to f111v and f112v on the BL microfilm).

Mike

From gauss!mycroft.rand.org!jim Tue Sep 14 10:21:43 PDT 1993
Received: by ninet.research.att.com; Tue Sep 14 13:38 EDT 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA28004
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 14 Sep 1993 10:21:45 -0700
Received: from localhost by mycroft.rand.org; Tue, 14 Sep 93 10:21:44 PDT
Message-Id: <9309141721.AA09399@mycroft.rand.org>
To: voynich@rand.org
Subject: VMs: New transcribed pages available
Date: Tue, 14 Sep 93 10:21:43 PDT
From: Jim Gillogly <jim@mycroft.rand.org>
Status: OR

Well, older ones, actually...  Mike Roe has converted his Bennett-notation
pages into Currier notation, and sent them for inclusion.  These are f66r,
67r1, 67v2, 68r1, 68v3, 70v2, 71v, 86v3, 93r, and 100v.  Jim Reeds has
converted some First Study Group transcriptions from the Friedman
Collection, and they're included as f111v-114r.

You can pick up the current set of text in the archive by anonymous FTP:

	rand.org:pub/voynich/voynich.now

Let us know if your usual tools tell anything interesting about the
new data.

	Jim Gillogly
	Highday, 23 Halimath S.R. 1993, 17:20

From gauss!trl.oz.au!j.guy Wed Sep 15 16:00:31 +1000 1993
Received: by ninet.research.att.com; Wed Sep 15 12:05 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA02725
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 15 Sep 1993 08:55:27 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA23836; Wed, 15 Sep 93 16:00:34 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309150600.AA23836@medici.trl.OZ.AU>
Subject: Received...
To: voynich@rand.org
Date: Wed, 15 Sep 1993 16:00:31 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 353       
Status: OR

... the photocopies from the George Marshall Foundation.
Just at the right time, too: I was going insane chasing
a bug in language-classifying program package I'd been
writing. Our HP optical scanner gave up trying to make
sense of what it saw, as I guessed it would. So, within
a minute, I'll start inputting those sheets, starting
from the end.     


From gauss!trl.oz.au!j.guy Thu Sep 16 12:16:18 +1000 1993
Received: by ninet.research.att.com; Wed Sep 15 22:23 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA28542
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 15 Sep 1993 19:16:25 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA08007; Thu, 16 Sep 93 12:16:21 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309160216.AA08007@medici.trl.OZ.AU>
Subject: Some observations
To: voynich@rand.org
Date: Thu, 16 Sep 1993 12:16:18 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 354       
Status: OR


I've now typed in "pages" 233 and 234, and compared a bit with
the last page of Petersen. The transcription system is easy
and transparent, but we'll need to proofread against the Voynich
text itself: I have seen, in a few lines, quite a few cases
of "M" which should be corrected to "N".

Logging off now, called to housecleaning duties by my wife...


From gauss!nyx.cs.du.edu!rcarter Thu Sep 16 00:32:39 0600 1993
Received: by ninet.research.att.com; Thu Sep 16 02:38 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA02570
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 15 Sep 1993 23:31:52 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA22440; Thu, 16 Sep 93 02:31:49 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA25425; Thu, 16 Sep 93 00:32:41 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309160632.AA25425@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
Subject: VMs shorthand, Nahuatel, etc.
To: voynich@rand.org
Date: Thu, 16 Sep 1993 00:32:39 -0600 (MDT)
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 3017      
Status: OR

>Jim Gillogly <James_Gillogly@rand.org> has some points/questions:

>>rcarter@nyx.cs.du.edu (Ron Carter) writes:

>>I propose that the VMs is a phonetic shorthand in abbeviated Latin
>>transcribed by at least two people from a person dictating in Nahuatl.

>I have lots of problems seeing the VMs as straight shorthand.  One is
>that shorthand characters tend to be easier to draw than garden-variety
>characters... ... I would expect somebody designing a shorthand for them
>to use letters no more complicated than the equivalent Latin ones.

True enough, however some of the existing Latin shorthand, fairly commonly
used in the 15th, 16th, and 17th C. got fairly complex in design at times.

Am I suggesting that the Carter-VMs-Nahuatl-Latin Abbrv. link was in any
way usual?  Nope, as I imagine the monks/priests dealing with the proposed
original Nahuatl (pre-translated) version of the VMs thought they had a
pretty heretical document on their hands, and made up a "non-standard"
version of their own shorthand...

>Mind you, I don't know beans about Nahuatl, and one can imagine someone
>familiar with Latin abbreviations feeling he had to choose symbols not
>already in use to represent the non-Latin sounds of that language.  Could
>Q represent "-XT-" or something?  Who knows...

And when you take into account such non-Latin stuff as atl and qtl, etc...

>Another problem I have with it is that shorthand tends to increase
>variability by squeezing out some of the redundancy in the language.

In an area beyond my ken...

>If one were to assume it <is> in shorthand directly encoding some language
>phonetically, could one deduce anything from the non-linguistic-looking
>bits, such as words repeated too often to look natural?  For example, what
>does one make of:

>	<f8v.8> OFSOEFZ.SOE.SOE.SOE.QAM.8AN-

Another oddball thing that has occurred to me, is that some of the repeat
phrases are the result of the person giving dictation losing their place,
and repeating themselves.  There aren't any cross-offs in the VMs, so maybe
they just continued writing even after making a mistake.  To extend the
heretical material line of thought, perhaps they finished the VMs in a
hurry and had no time to go back and proofread/correct it. 

>Some suggestions have been repeated adjectives for emphasis (very very
>very) or pious declarations (amen amen amen); maybe a Nahuatl book of
>natural history would have names with repeated elements, like the Hawaiian
>reef trigger fish that I'm wearing on my t-shirt (humuhumunukunukuapua'a).

Well, there -are- elements in Nahuatl, ...atlatlatl...  is a real fragment
as is stuff like ...atlqtlatl...  Granted, not terribly usual, but...

>In any case, it's a testable theory... Ron, have you found any Nahuatl
>text to run a statistical comparison on?

I will begin typing in some stuff from a Nahuatl herbal that I have access
to...  Really nice stuff...  Someone else will have to run stats on it.

Ron Carter | rcarter@nyx.cs.du.edu | 16 SEP 93 | Denver, CO USA


From gauss!mycroft.rand.org!jim Thu Sep 16 08:22:33 PDT 1993
Received: by ninet.research.att.com; Thu Sep 16 11:24 EDT 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA20764
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 08:22:31 -0700
Received: from localhost by mycroft.rand.org; Thu, 16 Sep 93 08:22:33 PDT
Message-Id: <9309161522.AA14830@mycroft.rand.org>
To: voynich@rand.org
Subject: VMs: no strikeouts
From: Jim Gillogly <James_Gillogly@rand.org>
Reply-To: James_Gillogly@rand.org
Date: Thu, 16 Sep 93 08:22:33 PDT
Sender: jim@mycroft.rand.org
Status: OR

> rcarter@nyx.cs.du.edu (Ron Carter) writes:
>                            There aren't any cross-offs in the VMs, so maybe
> they just continued writing even after making a mistake.  To extend the
> heretical material line of thought, perhaps they finished the VMs in a
> hurry and had no time to go back and proofread/correct it. 

This is an interesting observation that hadn't occurred to me.  Are
strikeouts ever seen in "fair copies" of manuscripts of the 16th century?
I suppose there are a few possibilities besides the conjecture that
mistakes are still in there: (a) errors were scraped off, like a
palimpsest; (b) the page was started over, including the illustration if
relevant to the text; (c) it doesn't mean anything anyway, so that errors
are not an issue. (a) should be testable by observation of the original,
if anybody gets over that way again.  One factor that would tend to reduce
errors (as well as reduce the likelihood that this is the result of
someone taking dictation) is that the text is evidently copied from an
original, based on the extended ligatures, which indicate the copiest
<knows> another P will be coming later in the line to connect with.

	Jim Gillogly
	25 Halimath S.R. 1993, 15:20

From gauss!reeds Thu Sep 16 11:40 EDT 1993
Received: by ninet.research.att.com; Thu Sep 16 11:43 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA21845
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 08:42:05 -0700
Message-Id: <199309161542.AA21845@rand.org>
From: reeds@research.att.com
Date: Thu, 16 Sep 93 11:40 EDT
To: voynich@rand.org
Subject: Voynich letter shapes source?
Status: OR

While reading through my copy of Joseph Galland's "An Historical 
and Analytical Bibliography of the Literature of Cryptology"
I glanced at the entry for Trithemius's "Polygraphia" (1518).

According to Galland, Books I through IV consist of lists of
"Chinese Menu" alphabets described by Kahn on p 132-135 ("conditor
clemens discernans mundana" = "cave", since conditor = c in alphabet 
1, clemens = a in alphabet 2, etc), Book V is about what we call
polyalphabets, and Book VI (this is what interests me) "contains
several alphabets of allegedly ancient writing characters which 
Trithemius offered as material for anyone who wished to create his
own secret alphabet".

Has anyone seen these Book VI characters?  Are they reproduced anywhere?


Jim Reeds


From gauss!nyx.cs.du.edu!rcarter Thu Sep 16 11:40:38 0600 1993
Received: by ninet.research.att.com; Thu Sep 16 13:41 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA28008
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 10:39:50 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA24850; Thu, 16 Sep 93 13:39:48 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA22796; Thu, 16 Sep 93 11:40:40 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309161740.AA22796@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
Subject: Strikeovers, etc.
To: voynich@rand.org
Date: Thu, 16 Sep 1993 11:40:38 -0600 (MDT)
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2133      
Status: OR

>Jim Gillogly write:

>>rcarter@nyx.cs.du.edu (Ron Carter) writes:

>>There aren't any cross-offs in the VMs, so maybe they just continued
>> writing even after making a mistake.

>This is an interesting observation that hadn't occurred to me.  Are
>strikeouts ever seen in "fair copies" of manuscripts of the 16th century?

It is not -too- unusual for 16th C. Ms. to have mistakes left in place.

>I suppose there are a few possibilities besides the conjecture that
>mistakes are still in there:

>(a) errors were scraped off, like a palimpsest...

The usual way of handling mistakes, esp. on vellum; as you point out, best
told by in-person inspection... 

>(b) the page was started over, including the illustration ...

Yikes!  And -in- Voynich...

>(c) it doesn't mean anything anyway, so that errors are not an issue.

>One factor that would tend to reduce errors (as well as reduce the
>likelihood that this is the result of someone taking dictation) is that
>the text is evidently copied from an original, based on the extended
>ligatures ...

If anyone hadn't quessed, my Mayan/VMs theory is truly meant to be
-extremely- speculative.  :-)  However, I now have in my (highly
unorganized) possession, two photocopies of herbal plates, one in
Nahuatl, of herbals that are plants associated with snake bite cures...
They both show a plant, with two snakes entwined in the plant.

In the VMs, there is a plate in which appears two creatures, that people
call "worms", to which I submit are really poorly drawn snakes, and the
plate is supposed to be a snake bite herbal cure.

Now, to find "serpentia" or "-quatl" in the text of that plate...

Why all this seemingly absurd speculation on my part?  I feel that one
potential path to break through the enigma of the VMs, is to find some
sort of clue that will give us just -one- word; I feel the drawings of
the star charts are worthless, as are the herbal plates, for the most
part; I fee the same about the (albeit cute!) VoynichNymphs; as to the
value of the medicinals; an avenue I hope to explore further...

Regards, Ron | rcarter@nyx.cs.du.edu | Denver, CO USA | 16 SEP 93

From gauss!SEI.CMU.EDU!firth Thu Sep 16 14:21:06 0400 1993
Received: by ninet.research.att.com; Thu Sep 16 14:22 EDT 1993
Received: from ts4a.sei.cmu.edu by rand.org with SMTP id AA00207
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 11:21:11 -0700
Received: by ts4a.sei.cmu.edu (5.65/2.5)
        id AA14328; Thu, 16 Sep 93 14:21:06 -0400
Date: Thu, 16 Sep 93 14:21:06 -0400
From: firth@SEI.CMU.EDU
Message-Id: <9309161821.AA14328@ts4a.sei.cmu.edu>
To: voynich@rand.org
Subject: Not marble...
Status: OR


... nor the gilded monuments/ of princes shall outlive this powerful cypher.

Yes, folks, I'm still alive.  And my photocopy of the MS has been closed
for lo these many months - from lack of time, lack of ideas, and the
(tragic) loss of my beloved Microvax, on which I had most of the analysis
tools.

It's good to see some other members of the comatose Team Voynich rising
from their unquiet graves to perturb the air with shrieks of Thai, Mayan,
Chinese, Basque, and Gibberish.  But where do we go from here?

Ron Carter's recent post wished that we had "just one word" of plain
text to crack the MS.  That would be a small start - and then I
remembered that the guy XXX person who cracked Old Persian started
with just one word: the word for king ("kshatra", supplies the unreliable
lumber-room of the cerebral cortex).

And we need a common noun, not a proper noun.  So I agree that the star
names are poor bets - so what if we found 'Aldebaran' - that would open
the script, but not the language.  But yes, finding 'serpens' or 'coatl'
(isn't that "water snake", though?) would break the code once for all.

Another thought.  Past efforts on the herbal section seem to have come
up with no good identification of the plants.  But they tried botanical
identification, ie real plants.  Maybe we should try literary identification,
ie find examples in other herbaria of similar nonsense plants.  After
all, you'd have precious little luck taking a mediaeval bestiary to
the local zoo.  And that might give more clues to the provenance of
the MS.

A final thought.  All this stuff about Mayans and Koreans and whatever
is fascinating, and should never be discounted, but we do have one
very big clue as to the provenance of the MS, and it's one any other
hypothesis must explain away: the Zodiac.

Please, post any and all ideas, however vague.  I'm pretty much at a loss.

Robert

PS: Didn't mean to be too hard on the mediaevals.  Yes, the illustrator
of the rhinoceros had clearly never seen one, but whoever designed Sonic
has obviously never seen a real hedgehog, and what's his excuse?

From gauss!trl.oz.au!j.guy Fri Sep 17 12:01:22 +1000 1993
Received: by ninet.research.att.com; Thu Sep 16 22:14 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA18915
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 19:01:41 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA12131; Fri, 17 Sep 93 12:01:23 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309170201.AA12131@medici.trl.OZ.AU>
Subject: 228 pages to go
To: voynich@rand.org
Date: Fri, 17 Sep 1993 12:01:22 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1029      
Status: OR

... I'm even starting to speak Voynich fluently.
Calling out those letters in French, rather than
English, helps a lot too: every Voynich word
translates into a sentence that makes sense!
For instance:

4ODAR  = quarante dards (forty stings)
OEDAK  = ouais, d'ac    (i.e. "oui, d'ac[cord]")

The VMS is actually a treatise on food and drink,
to wit, the many occurrences of EDAM, HAM, and
ARAK. 

Seriously, the repetitiveness of the text is
strange. It does not fit any language I know,
and at the same time is strangely reminiscent
of the repetitive patterns of the text of the
Easter Island tablets. Was it Friedman who had
come to the conclusion that it must have been
an artificial, "philosophical" language (in the
style of Bishop Wilkins' "universal character")?
It looks like it for sure. But how could several
people ever have learnt such an artificial 
language well enough to become fluent in it?
And why those two "dialects"? No, it must be
something else. No wonder Romaine Newbold went
batty trying to decipher it.

From gauss!umcc.umich.edu!bgrant Fri Sep 17 00:07:14 0400 1993
Received: by ninet.research.att.com; Fri Sep 17 00:23 EDT 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA22228
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 16 Sep 1993 21:07:18 -0700
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0odX6J-000AAhC@umcc.umich.edu>; Fri, 17 Sep 93 00:07 EDT
Message-Id: <m0odX6J-000AAhC@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: Repetitiveness of the VM
To: voynich@rand.org
Date: Fri, 17 Sep 1993 00:07:14 -0400 (EDT)
X-Mailer: ELM [version 2.4 PL22]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 585       
Status: OR

Perhaps the VM is repetitious because the scribe is making it up as he
goes.  What I mean is that the real message is carried by less than all
the letters, say by one letter per "word" or by the _length_ of the words.
It could be like the simple ciphers where the n-th letter of each word
is significant and the rest is filler.  In this case, the repetition may
be due to lack of imagination on the part of the scribe.

This might also explain why some characters seldom appear as finals (e.g.
4) and others are seldom initials, i.e. because they're just windowdressing
anyway.

Bruce

From gauss!cl.cam.ac.uk!Michael.Roe Fri Sep 17 15:04:29 +0100 1993
Received: by ninet.research.att.com; Fri Sep 17 10:06 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA05467
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 07:04:41 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 17 Sep 1993 15:04:33 +0100
To: voynich@rand.org
Subject: Re: Repetitiveness of the VM
In-Reply-To: Your message of Fri, 17 Sep 93 00:07:14 -0400. <m0odX6J-000AAhC@umcc.umich.edu>
Date: Fri, 17 Sep 93 15:04:29 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:067170:930917140436"@cl.cam.ac.uk>
Status: OR


On the other hand, the VM is extrordinarily free of repetition. From the text
I've analysed so far, *no* group of 5 consecutive ``words'' *ever* occurs twice.
Considering the amount of text, this is less repetive than normal languages.

For example, a much shorter Latin text I looked at for comparison has a 6 word
repeat of ``Cuius materie aut forme sit uas ...'' Almost any technical text in
any European language is going to have standard phrases that get repeated eg.
``... is a registered trademark of ...''

True, there does seem to be a fair amount of local repetition, but this may just
be because the alphabet is small; If you've only got a few symbols, it's
impossible to avoid short repeats.

At this point, I have a confession to make; the page of the VM I transcribed
as ``f100v'' should have been ''f101v1''. I was misled by it being labelled
as f100v on p.86 of Brumbaugh. My program for identifying repeated patterns
immediately identified a repeat of an entire page! (f101v as transcribed by
me and f101v1 as transcribed by Jim).

Mike

From gauss!reeds Fri Sep 17 10:54 EDT 1993
Received: by ninet.research.att.com; Fri Sep 17 10:59 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA07768
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 07:56:00 -0700
Message-Id: <199309171456.AA07768@rand.org>
From: reeds@research.att.com
Date: Fri, 17 Sep 93 10:54 EDT
To: voynich@rand.org
Subject: Voynich f101v
Status: OR

Mike Roe reports confusion about f101v versus f101v1.  

> At this point, I have a confession to make; the page of the VM I transcribed
> as f100v'' should have been ''f101v1''. I was misled by it being labelled
> as f100v on p.86 of Brumbaugh. My program for identifying repeated patterns
> immediately identified a repeat of an entire page! (f101v as transcribed by
> me and f101v1 as transcribed by Jim).

This mixup is understandable.  Brumbaugh (who seems to have a genius for
small mistakes) and the VMS itself are jointly to blame.

Folio 101 is a foldout (valley fold on recto side, I think) so both
sides of the leaf are divided into two panels by the fold.  The writing,
however, treats each side as a single wide page, with the writing and
pictures continuing right over the crease as if it were not there.  So
a transcription of f101r, single double-width 'logical page' makes sense.
But the crease on the verso side has become very worn (this is perhaps 
the most damaged part of the MS) so the writing on the crease is no longer
legibile. The gap is about 15 mm wide, which is wide enough for a short 
word or two of V text.  Thus it looks to a transcriber like a pair of 
pages, f101v2 on the left, f101v1 on the right.  And until we learn 
Voynichese, I suppose it is easier to transcribe it that way, leaving us
with two texts vertically sundered, rather like the Abbe Farria's letter.

So from the point of view of the geometry of folding it should be f101v1
and f101v2.  From the point of view of 'logical' pages of writing it 
should be f101v.  But from what we can see nowadays it might as well
be f101v1 and f101v2.  I suppose we should use f101v when referring to
the whole of the verso side of f101, and f101v2 when referring to that
part to the left of the crease and f101v1 to that part to the right of
the crease.  Mike and I both transcribed f101v1.  The left hand side,
f101v2, awaits transcription.


Jim Reeds


From gauss!VFL.Paramax.COM!lock60!snark!cowan Fri Sep 17 11:03:07 0400 1993
Received: by ninet.research.att.com; Fri Sep 17 11:20 EDT 1993
Received: from gvls1.VFL.Paramax.COM by rand.org with SMTP id AA08939
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 08:19:01 -0700
Received: from lock60.UUCP by gvls1.VFL.Paramax.COM (4.1/mls/4.1) 
	id AA09665; Fri, 17 Sep 93 11:18:58 EDT
Received: by lock60.Canal.Org (smail2.5)
	id AA17483; 17 Sep 93 11:18:22 EDT (Fri)
Received: by snark.uucp (Smail3.1.28.1 #2)
	id m0odhL2-000BfXC; Fri, 17 Sep 93 11:03 EDT
Message-Id: <m0odhL2-000BfXC@snark.uucp>
From: snark!cowan@VFL.Paramax.COM (John Cowan)
Subject: Re: 228 pages to go
To: voynich@rand.org (Voynich List)
Date: Fri, 17 Sep 1993 11:03:07 -0400 (EDT)
In-Reply-To: <9309170201.AA12131@medici.trl.OZ.AU> from "Jacques Guy" at Sep 17, 93 12:01:22 pm
X-Mailer: ELM [version 2.4 PL22]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 961       
Status: OR

frogguy writes:

> OEDAK  = ouais, d'ac    (i.e. "oui, d'ac[cord]")

In the Canglish language, that phrase is spelled "Sweedack!".

> Seriously, the repetitiveness of the text is
> strange. It does not fit any language I know,
> and at the same time is strangely reminiscent
> of the repetitive patterns of the text of the
> Easter Island tablets.

What about the notion that Voynichese has approximately the same entropy
as Tahitian?  You are the only resident Pacificist on this list, I think.

> Was it Friedman who had
> come to the conclusion that it must have been
> an artificial, "philosophical" language (in the
> style of Bishop Wilkins' "universal character")?

Friedman it was.

> No wonder Romaine Newbold went
> batty trying to decipher it.

Not so fast.  My father studied with Newbold (but not the VMS) and he
was by no means crazy, just wrong.

-- 
John Cowan	cowan@snark.thyrsus.com		...!uunet!lock60!snark!cowan
			e'osai ko sarji la lojban.

From gauss!cl.cam.ac.uk!Michael.Roe Fri Sep 17 17:37:09 +0100 1993
Received: by ninet.research.att.com; Fri Sep 17 12:38 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA12610
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 09:37:25 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Fri, 17 Sep 1993 17:37:14 +0100
To: voynich@rand.org (Voynich List)
Subject: Entropy and the VM
In-Reply-To: Your message of Fri, 17 Sep 93 11:03:07 -0400. <m0odhL2-000BfXC@snark.uucp>
Date: Fri, 17 Sep 93 17:37:09 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:097600:930917163717"@cl.cam.ac.uk>
Status: OR


John Cowan writes:
> What about the notion that Voynichese has approximately the same entropy
> as Tahitian? 

Interestingly, the entropy of a word chosen at random from the VM is about 10
bits --- approximately the same as for an English text of the same length.
(Of course, this measure ignores the correlation between successive words)

Mike

From gauss!SEI.CMU.EDU!firth Fri Sep 17 13:07:29 0400 1993
Received: by ninet.research.att.com; Fri Sep 17 13:09 EDT 1993
Received: from ts4a.sei.cmu.edu by rand.org with SMTP id AA14093
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 10:07:39 -0700
Received: by ts4a.sei.cmu.edu (5.65/2.5)
        id AA25808; Fri, 17 Sep 93 13:07:29 -0400
Date: Fri, 17 Sep 93 13:07:29 -0400
From: firth@SEI.CMU.EDU
Message-Id: <9309171707.AA25808@ts4a.sei.cmu.edu>
To: voynich@rand.org
Subject: Notes on the provenance of the Voynich MS
Status: OR

Notes on the Voynich Manuscript - Part 20
-----------------------------------------

Speculations on the Date and Provenance of the Voynich MS
---------------------------------------------------------

Consider this, if you will:

	Was this the face that launched a thousand ships
	And burned the topless towers of Ilium?
	Sweet Helen, make me immortal with a kiss!

Now, suppose we came upon this text in a mysterious manuscript,
and had no clue, from history or palaeography, from vellum or
ink, as to the provenance of what we were reading.  Nevertheless,
we could conclude, with a high degree of probability, that it
had originated from Europe or the Mediterranean littoral: from
one or other of the states and cultures that are the heirs to
the Hellenic civilization.  And why?  because who else would
refer so explicitly to the Iliad of Homer?

Now consider, if you can find it, Botticelli's Primavera.  This
is a picture, not a text; nevertheless, we can establish, from
the subject, style, and iconography, that it was painted during
the European Renaissance., and very probably in Northern Italy.

And so to the topic of this note: if we consider the Voynich MS
not as a puzzle, a cypher, but as a work of art, what can we
deduce about its possible provenance, from its subject, style
and iconography?  Oh, how I wish I held the real thing, and not
a miserable monochrome photocopy of a few pages!  And a miserable
photocopy of a hand-written copy, and that copy by a Westerner!

I am not an expert in these matters, and it might be a good idea
to consult one.  But I've seen a lot of art, from almost all over
the world, and can fairly reliably tell the difference between
Mayan, Chinese, Egyptian, Persian and Roman renderings of the
same subject.  And of the images in the MS, none looks to me
obviously non-European, and all that look familiar look European.

Here are some examples.

(a) the features of the women in the zodiac and plumbing sections.
    They all look caucasian.  Moreover, many of them have hairdos,
    and those mostly look mediaeval european.

(b) the pictures that seem to be allegories of the seasons.  It's
    pretty easy to identify Spring, Summer, Autumn and Winter, and
    the iconography - flowers, fruits, old man with stick &c - is
    again traditional mediaeval.

(c) the astronomical folios almost all seem based on divisons of
    12 and 24, either hours in the day or months in the year.  That's
    the Western solar calendar - a Mayan (and, indeed an Arab) would
    use a very different iconography.  If we had a lot of 18s and 20s,
    that would point almost irrefutably to a Mayan origin - I submit
    this evidence is also pretty strong.

While looking at heads, look at their angle with respect to the
viewer.  If they were Mayan or Egyptian, almost all of them would
be in full profile.  If Byzantine or Indian, they would be either
full profile or full face.  But many of them are half or three
quarter profile.  That's typical again of European art, and of
the late-Mediaeval and Renaissance period.

Here is another consideration.  As far as I can see, all the
images are two-dimensional.  There is no attempt at either
"true" perspective or the various "false" perspectives found
in Egyptian, Chinese, and late-Mediaeval art.  If the artist
wanted to give anything depth - the plumbing pictures with mass
hot tubs for instance - well, he was less skilled that the illustrator
of Les Tres Riches Heurs or the painter of the fowling scene in
the tomb of Tutankhamun.  So either he was untrained, or he lived
before Giotto's innovations had spread to his part of the world.

--------

Let's turn to the Zodiac.  It's the so-called "Western" zodiac,
the one documented by Berosus and almost certainly cast into
that form in the Hellenistic age.  That again says the MS comes
from somewhere that was once part of the Roman Empire.  And the
icons let us get more specific.  The MS has Libra, the scales,
not Zygon, the yoke, so it's latin, not greek. And Leo, the
lion, isn't a bestiary lion: he's an heraldic leopard, as
found in Mediaeval French heraldry - and hence, thanks to Duke
William, in the arms of England.  Scorpio, as already remarked,
isn't a scorpion, he's a lizard or (I think) a salamander.  That
again says Western Europe, where scorpions are hard to find.
And Cancer is a lobster, which hints strongly at France.

Look at all of them, and imagine meeting them in the street.
Taurus isn't a giant, wife-ravishing monster - he's the image
of the "little white bull", the bulls of Catalonia.  And
Aries with his lyre-shaped horns - I've seen him hopping
about the hills above Barcelona.  Gemini are a man and a woman,
and that's a late change.  To me, they look like the Lovers of
an Italian Tarot deck.  Those fish, Pisces, they were drawn
from life, and they are carp, at a venture.  From a fishpond,
not from the sea, just as a monk would have drawn them.  And
Saggitarius, the archer, he holds a crossbow.  The crossbow has
been invented twice in all history - in Ancient Greece and in
Mediaeval Europe.  That's not a greek crossbow - we have a
description by Heron of Alexandria, and it's nothing like
that drawing.  But it is very like the crossbows you see in
museums.  Now, if he'd been holding an atl-atl...

--------

The maps.  Well, maybe - there are two folios that seem to
contain an approximate map of the world.  And it is the classic
mediaeval disc-and-tau: a round, flat world divided by a T
of inner seas into three continents: Asia (eastern half),
Africa (southern quarter); Europe (northern quarter).  The
Mappa Mundi in Hereford Cathedral looks just like that, and
it dates from the thirteenth century.  By the time of Behain,
in 1460 or so, it had vanished.  So, Western Europe again, and
surely before 1400.

--------

My conclusion:  The Voynich Manuscript was created in Europe,
by Europeans, and probably in one of the lands that border the
Gulf of Lyons.  It is late Mediaeval, not Renaissanxce, and
hence dates from approximately 1350.

--------

And now for the evidence that does not fit the above!

Again, back to the Zodiac.  The number of nymphs evidently matters -
if you turn from folio to folio, you can see the scribe trying
different patterns to fit them in (10+19, 8+16+6, 10+20...), and
often having to scrunch them up, or spread them out, near the end
of a ring.  (By the way, that suggests these folios were not a fair
copy, but at least in layout were originals.)

But, what calendar has one month of 29 days (Pisces) and then nine
months of thirty days?  None at all.  Maybe Pisces is a mistake, and
all the months have 30 days.  But that's the Pharaonic Egyptian
calendar, and why would anyone revive it?  And why have it beginning
with Pisces - the Egyptian year began in our July, with the heliacal
rising of Sirius.

So, maybe this is astrological rather than calendric.  But, again,
why begin the calendar with Pisces and not Aries?  That I find very,
very odd - because, you see, it's right: the Vernal Equinox does
fall in Pisces - about 2 days in at present, and in 1350 about 11
days in.  Why, alone of all works of Western astrology, is the
Voynich Zodiac true to the stars?  Who in that age even remembered
the precession of the equinoxes, discovered by Hipparkhos of Nicaea
but, as far as I know, forgotten until Tycho de Brahe?

(No, wait a minute, there's a Middle English rime about the Great
Year - Graves quotes it in The White Goddess - so maybe this
knowledge just went underground.)

And why are Aries and Taurus divided into light and dark halves,
each with 15 nymphs, stars, or days?

--------

The plumbing.  Yes, the nymphs look like plump european wenches,
but the pipes, valves and stuff?  Not even the Romans could have
built a fluid transport system that well.  I've read Vitruvius
on plumbing, and the joining, branching, tapering, curving of
the Voynich designs simply could not have been realised with
the technology of the time.  For example, when the Romans wanted
a pipe to turn a right-angle corner, they encased the joint in
a block of solid marble, because they didn't know how else to
cope with the stresses.  Nor could they create a true Y-junction
- they drilled a hole in a straight pipe and stuck a smaller pipe
into it at an angle.

What it looks like to me, is an attempt to redraw a system of
veins and arteries as if it were a plumbing system.  But where
did that idea come from, in a time when dissection was prohibited
by law, and centuries before Harvey discovered the circulation
of the blood?  As others have remarked, these pictures don't look
Mediaeval or Renaissance - they look eerily modern.

Another example - there are nymphs sitting in small bath tubs.
Nonsense!  There were few enough bath tubs in all Europe in those
times, and they were made of stone or metal.  Nobody could cast
a pipe the size of a tub, no kiln big enough to fire it could
have been kept hot.  All right, it's an allegory - but allegory
uses familiar images to convey an unfamiliar meaning, that's its
point - it doesn't use impossible images as if they were commonplace.

(But then, what about Hieronymous Bosch?  Look at the weird machines
in his picture of Hell (right panel of the Garden of Worldly Delights).
And wasn't he supposed to be a member of a secret Gnostic cult called
the Brethren of the Free Spirit, distantly related to the Cathari?
No! no! ... that way madness lies.)

--------

The plants.  Not the leaves and stems and alchemical beasts, but the
folios that seem to show cell structure.  How is that possible before
van Leeuwenhoek?  Roger Bacon invented the microscope, you reply,
following Newbold.  But who showed him how to prepare slide samples?
Who invented the microscope slide, the microtome, the art of staining?
You can't just look at a plant and see cell structure.

--------

No: there is something else in the Voynich MS; something alien.  Images
that not only didn't exist and couldn't exist, but that nobody in that
time should even have been able to imagine.  (Orbis tertius, says the
daimon; I ignore him.)

Last conclusion: I need to see the real thing.

Robert

From gauss!eecs.umich.edu!kckluge Fri Sep 17 13:20:55 0400 1993
Received: by ninet.research.att.com; Fri Sep 17 13:22 EDT 1993
Received: from krusty.eecs.umich.edu by rand.org with SMTP id AA14664
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 10:20:32 -0700
Received: by krusty.eecs.umich.edu id AA09027 ; Fri, 17 Sep 1993 13:20:55 -0400
Date: Fri, 17 Sep 1993 13:20:55 -0400
From: Karl Kluge <kckluge@eecs.umich.edu>
Message-Id: <199309171720.AA09027@krusty.eecs.umich.edu>
To: voynich@rand.org
In-Reply-To: <9309171707.AA25808@ts4a.sei.cmu.edu> (firth@SEI.CMU.EDU)
Subject: Re: Notes on the provenance of the Voynich MS
Status: OR


D'Imperio covers some of that territory. I'd be interested in seeing 
Panofsky's letters on the subject (I think those are in the Friedman
Collection).

I seem to be getting all the messages twice, by the way. 

Karl

From gauss!nyx.cs.du.edu!rcarter Fri Sep 17 12:33:42 0600 1993
Received: by ninet.research.att.com; Fri Sep 17 14:34 EDT 1993
Received: from relay1.UU.NET by rand.org with SMTP id AA18229
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 11:33:19 -0700
Received: from nyx.cs.du.edu by relay1.UU.NET with SMTP 
	(5.61/UUNET-internet-primary) id AA03347; Fri, 17 Sep 93 14:32:52 -0400
Received: by nyx.cs.du.edu (4.1/SMI-4.1)
	id AA12103; Fri, 17 Sep 93 12:33:43 MDT
From: rcarter@nyx.cs.du.edu (Ron Carter)
Message-Id: <9309171833.AA12103@nyx.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
Subject: Repetiveness in the VMs...  :-)
To: voynich@rand.org
Date: Fri, 17 Sep 1993 12:33:42 -0600 (MDT)
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 426       
Status: OR

>Mike Roe <Michael.Roe@cl.cam.ac.uk> comments:

>Almost any technical text in any European language is going to
>have standard phrases that get repeated eg. ``... is a registered
>trademark of ...''

Oh great!  The VMs will probably turn out to be the equivalent of a
16th C. group patent filing, trademark disclaimers and all...  :-)

Oh those lawyer-monks!

Regards, Ron | rcarter@nyx.cs.du.edu | Denver, CO USA | 17 SEP 93

From gauss!trl.oz.au!j.guy Sat Sep 18 12:30:27 +1000 1993
Received: by ninet.research.att.com; Fri Sep 17 22:32 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA05597
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Fri, 17 Sep 1993 19:31:30 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA13577; Sat, 18 Sep 93 12:30:28 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309180230.AA13577@medici.trl.OZ.AU>
Subject: 223 to go
To: voynich@rand.org
Date: Sat, 18 Sep 1993 12:30:27 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 895       
Status: OR

It's fairly easy work. The repetitiveness and redundancy of
the "language" makes it all the easier. Ends of paragraphs are
marked by a double comma (,,), but ends of lines are not. We'll
have to put that in later, I guess, working from the real thing.
Occasionally, I hit upon a seemingly impossible combination --
impossible in Voynichese. Perhaps those "impossible" words are
numbers. It's a monosyllabic language, with tones indicated
rather irregularly (don't believe a word of this: it only 
feels like a Mandarin at time). At any rate, if it's a simple
substitution code, and if it's a known European language, then
it's got to be a rhyming dictionary or a treatise on spoonerisms
and alliterative poetry. Now, of course, if it's 13th century
or so, I am still attracted to the theory that it's a fairy book, written
by the Little Folk in their fairy tongue. The writing *is* tiny,
too...

From gauss!SEI.CMU.EDU!firth Mon Sep 20 13:00:15 0400 1993
Received: by ninet.research.att.com; Mon Sep 20 13:02 EDT 1993
Received: from ts4a.sei.cmu.edu by rand.org with SMTP id AA11100
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 20 Sep 1993 10:00:13 -0700
Received: by ts4a.sei.cmu.edu (5.65/2.5)
        id AA04304; Mon, 20 Sep 93 13:00:15 -0400
Date: Mon, 20 Sep 93 13:00:15 -0400
From: firth@SEI.CMU.EDU
Message-Id: <9309201700.AA04304@ts4a.sei.cmu.edu>
To: voynich@rand.org
Subject: Voynich Plumbing
Status: OR

Folks

In a recent Note, I once again perpetrated an ignorant blatant assertion:
that the Voynich plumbing could not be realised with the technology of
the time.  Rereading the note, it suddenly occurred to me that I was, as
so often, very, very wrong.

No, you could not make pipes like that with mediaeval technology, either
of pottery or of metal.  But there was one material that could be worked
by hand into those shapes, though not on a large enough scale to hold
nymphs.  And I'm an idiot to have forgotten it, since I've seen the
process myself, in La Serenissima herself, at the Murano works - which,
by a typically staggering Voynich coincidence, were founded in 1291.

Latest wild conjecture: the Voynich plumbing is made of glass.  If so,
the folios are surely alchemical, in spite of the iconography.

Robert

From gauss!reeds Mon Sep 20 20:23 EDT 1993
Received: by ninet.research.att.com; Mon Sep 20 20:47 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA02131
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 20 Sep 1993 17:27:10 -0700
Message-Id: <199309210027.AA02131@rand.org>
From: reeds@research.att.com
Date: Mon, 20 Sep 93 20:23 EDT
To: voynich@rand.org
Subject: Voynich X Windows
Status: OR

Are there any X window gurus who might be willing to answer a few
novice hand-holding type questions as I code up a VMS-related bit
of window graphics software?


Jim Reeds


From gauss!trl.oz.au!j.guy Wed Sep 22 09:33:44 +1000 1993
Received: by ninet.research.att.com; Tue Sep 21 19:36 EDT 1993
Received: from medici.trl.OZ.AU ([137.147.20.61]) by rand.org with SMTP id AA19657
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 21 Sep 1993 16:33:50 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA16864; Wed, 22 Sep 93 09:33:45 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309212333.AA16864@medici.trl.OZ.AU>
Subject: The picnic table
To: voynich@rand.org
Date: Wed, 22 Sep 1993 09:33:44 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 255       
Status: OR

Having now input about 40K of "Friedman transcription",
I did a count of the occurrences of "Y", which is the
"picnic table" in that system: ELEVEN! Far less uncommon
than I thought. 
Now to type in pages 216 and 217 (I try to stick to one
folio a day).


From gauss!cl.cam.ac.uk!Michael.Roe Wed Sep 22 01:47:46 +0100 1993
Received: by ninet.research.att.com; Tue Sep 21 20:50 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA22248
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 21 Sep 1993 17:47:57 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Wed, 22 Sep 1993 01:47:51 +0100
To: voynich@rand.org
Cc: Michael.Roe@cl.cam.ac.uk
Subject: The Voynich Zodiac
Date: Wed, 22 Sep 93 01:47:46 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:180570:930922004756"@cl.cam.ac.uk>
Status: OR


It suddenly occurred to me that there are 29 days in Pisces, so maybe the 29
figures on f70v2 are intentional, and not a mistake for 30.

The problem with this theory is that there are (I think) 31 days in Scorpio,	
but only 30 figures on f73r.

I know next to nothing of the history of astrology, so maybe there's a way to
make it fit and account for the missing days.

I have vague memories of a calendar with a 360-day year surviving into the
20th century, and used for some forms of share trading. (On the other hand,
maybe this is apocryphal.)

Mike


From gauss!trl.oz.au!j.guy Wed Sep 22 11:14:41 +1000 1993
Received: by ninet.research.att.com; Tue Sep 21 21:16 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA23079
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 21 Sep 1993 18:14:47 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA17039; Wed, 22 Sep 93 11:14:43 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309220114.AA17039@medici.trl.OZ.AU>
Subject: The picnic tables: where
To: voynich@rand.org
Date: Wed, 22 Sep 1993 11:14:41 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1742      
Status: OR


Here they are. The #nnn is the "page" number;
under it is the "line" number (each line contains
exactly 30 characters). The comma represents a 
space. I suspect that the double comma (,,)
represents sometimes an end-of-line (true 
"line", as appear in the VMS), sometimes an
end-of-paragraph. The lowercase stuff under
each line of quoted Friedman is the correspond
paleofrogguy (neofrogguy only differs in its
use of uppercase, mostly redundant, only there
to make the Voynich editor screens look prettier).
I've also found two occurrences of "9" and I
haven't chased that one up yet.


#219   folio 106v
 53 2ARY,ADCCG,OHCC8G,DSCO8G,4ODCC
       ^
    sa2n alpcc9 oqpcc89 lpc't89 4olpcc

#220   folio 107r
 53 ,2AEYAR,SG,4ODAM,ODAE,4ODZC8G,
        ^
     saxna2 c't9 4olpaiiv olpax 4oclptc89

#222   folio 108r
  2  PTG,HO8ASY,OPARG,OEPAIR,O8G,OH
              ^
     qjct9 qpo8ac'tn oqja29 oxqjai2 o89 oqp

#223   folio 108v
 13  ,ODCC8G,DC8ARYG,PTC8AM,ODC8G,O
                  ^
     olpcc89 lp8a2n9 qjctc8aiiv olpc8g o

#224   folio 111r
 46  M,ODAR,AM,HCCG,OHCCG,OYOR,SCG,
                           ^
     iiv olpa2 aiiv qpcc9 oqpcc9 ono2 c'tc9

 57  AE,ODZG,OYAR,AM,O8AM,TO8G,2OAM
             ^
     ax ocqpt9 ona2 aiiv o8aiiv cto89 soaiiv

#226   folio 112r
  1  FOETCG,4ODCCG,GDAIR,YAR,AEEG,O
                         ^
     ljoxctc9 4olpcc9 9lpai2 na2 axx9 o

#229   folio 113v
 53  SOY,ARAS,OPZCG,PZC8G,OPTCO8G,O
       ^
     c'ton a2ac't ocqjtc9 cqjtc89 olpctco89 o

#231   folio 114v
 33  ,TOG,PSC8G,4OHO*,SYAK,OS*O,8AM
                       ^
      cto9 qjc'tc89 4oqpo* c'tnaig oc't*o 8aiiv

#233   folio 115v
 33  OPTC8G,GSC8G,4ODCC8G,EYOR,YOM,
                           ^   ^
     oqjctc89 9c'tc89 4olpcc89 xno2 noiiv

From gauss!world.std.com!kibo Wed Sep 22 02:39:27 0400 1993
Received: by ninet.research.att.com; Wed Sep 22 02:41 EDT 1993
Received: from world.std.com by rand.org with SMTP id AA01870
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Tue, 21 Sep 1993 23:40:19 -0700
Received: by world.std.com (5.65c/Spike-2.0)
	id AA00320; Wed, 22 Sep 1993 02:39:27 -0400
Date: Wed, 22 Sep 1993 02:39:27 -0400
From: kibo@world.std.com (James "Kibo" Parry)
Message-Id: <199309220639.AA00320@world.std.com>
To: voynich@rand.org
Subject: Re:  Voynich Plumbing
Status: OR


> Latest wild conjecture: the Voynich plumbing is made of glass.

Rubber?  :-)

[that would suggest an origin in India, perhaps.]

						-- K.

From gauss!cl.cam.ac.uk!Michael.Roe Wed Sep 22 18:49:17 +0100 1993
Received: by ninet.research.att.com; Wed Sep 22 13:51 EDT 1993
Received: from swan.cl.cam.ac.uk by rand.org with SMTP id AA22704
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 22 Sep 1993 10:49:34 -0700
Received: from smew.cl.cam.ac.uk (user mrr (rfc931)) by swan.cl.cam.ac.uk 
          with SMTP (PP-6.5) to cl; Wed, 22 Sep 1993 18:49:21 +0100
To: voynich@rand.org
Cc: Michael.Roe@cl.cam.ac.uk
Subject: A word to unlock the Voynich MS
Date: Wed, 22 Sep 93 18:49:17 +0100
From: Mike Roe <Michael.Roe@cl.cam.ac.uk>
Message-Id: <"swan.cl.cam.:204940:930922174927"@cl.cam.ac.uk>
Status: OR


In a message a while back, Ron carter wished for just one word to provide a way
into the text.

That word could be "OPAE", with its declension OPAE, 4OPAE, SOPAE, it's
variant form OFAE, 4OFAE, SOFAE, and its line-final form OPAJ, 4OFAJ, SOFAJ.

This word occurs rather a lot in the labels in the astrological folios.
``Star'' maybe? Or ``day''? Or some synonym of these in some language...

P and F are dividing bars separating inflection from the noun. At a guess,
O- = nominative, 4O- = accusative, SO- = dative...

                   +-O--+  +-R-+
 O  --+         +--+    +--+   +----+
      |  +-P-+  |  +-SO-+  +-E-+    |
 4O --+--+   +--+                   |
      |  +-F-+  |  +-C--+           |
 SO --+         |  |    |           |
                |  +-CC-+           |
                |  |    |           |
                +--+-SC-+--9--------+------>
                |  |    |           |
                |  +-S--+           |
                |  |    |           |
                |  +-Z--+           |
                |  |    |           |
                |  +----+           |
                |                   |
                |  +-AE-+           |
                |  |    |           |
                +--+-AJ-+-----------+
                   |    |
                   +-AM-+
                   |    |
                   +-AN-+
Mike

From gauss!castrov.cuc.ab.ca!wuth Wed Sep 22 14:56:07 MDT 1993
Received: by ninet.research.att.com; Wed Sep 22 22:43 EDT 1993
Received: from fsa.cpsc.ucalgary.ca by rand.org with SMTP id AA16111
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 22 Sep 1993 19:40:58 -0700
Received: from news.cpsc.ucalgary.ca by fsa.cpsc.ucalgary.ca (4.1/CSd1.2)
	id <AA19983@fsa.cpsc.ucalgary.ca>; Wed, 22 Sep 93 20:39:31 MDT
Received: by news.cpsc.ucalgary.ca (5.67/cs1.0)
	id AA24972; Wed, 22 Sep 93 20:40:34 -0600
Return-Path: <wuth@castrov.cuc.ab.ca>
Received: from castrov.cuc.ab.ca by ajfcal.cuc.ab.ca (5.65c/Cuc2.2)
	id AA21369; Wed, 22 Sep 1993 16:17:07 -0600
Received: by castrov.cuc.ab.ca (V1.15.19{info@castrov.UUCP/5}/Amiga)
	id AA02b7w; Wed, 22 Sep 93 14:56:07 MDT
Date: Wed, 22 Sep 93 14:56:07 MDT
Message-Id: <9309222056.AA02b7w@castrov.cuc.ab.ca>
From: wuth@castrov.cuc.ab.ca (Brett Wuth)
To: voynich@rand.org
Subject: Voynich plumbing ideas
Status: OR

My apologies if this has already been gone over; I've been away from the
list for a while.  I'm also only working from memory of a photocopy
of some of the material.

It seems to me that in speculating about the plumbing there are several
questions that given different answers could result in different theories:

1. Does the "plumbing" represent physical structures?  (Yes/Parts/No)

   Could it be a state diagram?  (Of the Voynich cypher?)

2. Does the connectivity/direction/length/width/turns/joints of the "pipes"
   convey meaning?  How do they relate to each other/the other features?

   For instance modern road maps depict roads much wider than the scale of
   the map, and sometimes not in proportion to the width of the road.

   Perhaps only the connectivity/topology is significant.

3. Is there any attempt to represent the interior structure of the "pipes"?
   Is that interior solid/liquid/gas?

   What evidence is there that the contents of the pipes are liquid?

   Adjust with the scale, and use air as the contents, and perhaps
   we have a representation of a tunnel system.

   Adjust the scale again and perhaps we have something akin to the
   runners of a strawberry patch.

4. Do the "bathtubs" or their relative location/orientation/size/structure/
   number convey meaning?

5. Do the "ladies" or their relative location/orientation/size/features/
   number convey meaning?

Perhaps we need a list of all illustration features that a theory needs to
account for (even if just to say "fanciful addition").  Again, I'm working
from memory, so correct and add to this list:

"pipes", "pipe" connectivity, "pipe" topology, "pipe" direction, 
"pipe" lengths, "pipe" width, "pipe" turns, "pipe" joints,
"pipe" interior, "bathtubs", "bathtub" locations, "bathtub" orientation,
"bathtub" size, "bathtub" structures (enumerate), "bathtub" interior,
number of "bathtubs", "ladies", "lady" location, "lady" orientation,
"lady" size, "lady" features (enumerate), number of "ladies".

Cheers,
-- 
Brett Wuth  wuth@castrov.cuc.ab.ca bwuth@cuug.ab.ca wuth@castrov.UUCP
U302, 3510 44St SW, Calgary, Alberta, T3E 3R9, CANADA  Tel: +1 403 242-0848
Ask for my PGPKey(print=E4 F8 ED EE CB E0 1A D2 FA 3D 8B 2D 94 B1 A2 92)

From gauss!trl.oz.au!j.guy Fri Sep 24 12:19:32 +1000 1993
Received: by ninet.research.att.com; Thu Sep 23 22:21 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA06074
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 23 Sep 1993 19:19:49 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA19130; Fri, 24 Sep 93 12:19:39 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309240219.AA19130@medici.trl.OZ.AU>
Subject: 66k now
To: voynich@rand.org
Date: Fri, 24 Sep 1993 12:19:32 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 673       
Status: OR

I no longer know for sure where I am: the photostats no
longer give the folio number. I'm on "page" 208, which
ought to be f.101r, since p.212 was f.103r.

I have been not infrequently worried at heavily smudged
S's and 8's -- it's hard to tell the difference.

The good news is that I've typed in 45 pages of photostats
so far and that there are only under 100 left. I had
counted on spending half-a-hour on each page, stupidly
forgetting that the last pages are chock-a-block full
of writing -- typically 2500 characters each, when earlier
pages often have as few as 200 (VMS pages, that is, not
photostat pages).

One third of the way through! (Where's the champagne?)


From gauss!trl.oz.au!j.guy Tue Sep 28 15:48:32 +1000 1993
Received: by ninet.research.att.com; Tue Sep 28 01:50 EDT 1993
Received: from medici.trl.OZ.AU by rand.org with SMTP id AA05572
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 27 Sep 1993 22:47:52 -0700
Received: from tardis.trl.OZ.AU by medici.trl.OZ.AU (4.1/AIS-1M)
	id AA25535; Tue, 28 Sep 93 15:48:36 EST
From: j.guy@trl.oz.au (Jacques Guy)
Message-Id: <9309280548.AA25535@medici.trl.OZ.AU>
Subject: 83K typed in and odd things
To: voynich@rand.org
Date: Tue, 28 Sep 1993 15:48:32 +1000 (EST)
X-Mailer: ELM [version 2.4 PL20]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 380       
Status: OR


The odd things are on photocopy p.183,
second VMS paragraph (or line, as the case
may be):


PORATOE,4O8G,4OHCOE,OE8A4,6H6E,
2C69,269,6EHA4,6H36E,3C6E,
6H6EF3G,HAR,TO,EDCOPOE,OCCOR,AR,

Most unusual stuff! That was the first time
I encountered "3" and "6" in that transcription
system. Then, it becomes, and stays, normal again;
to wit:

8AM...... [oh, you know the rest!]







From gauss!reeds Sun Oct 17 19:54 EDT 1993
Received: by ninet.research.att.com; Sun Oct 17 20:04 EDT 1993
Received: from research.att.com by rand.org with SMTP id AA03878
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 17 Oct 1993 17:02:55 -0700
Message-Id: <199310180002.AA03878@rand.org>
From: reeds@research.att.com
Date: Sun, 17 Oct 93 19:54 EDT
To: voynich@rand.org
Subject: Voynich , Wilf.
Status: OR

Can anyone give me a definitive answer to this question:
How do you spell Mr. Voynich's first name?  Wilfred or Wilfrid?
Kahn and Zimansky give Wilfred, Galland, Clark, D'Imperio and
Newbold give Wilfrid.

Jim Reeds


From gauss!eecs.umich.edu!kckluge Mon Oct 18 17:00:55 0400 1993
Received: by gauss; Mon Oct 18 17:25:44 EDT 1993
Received: by inet.att.com; Mon Oct 18 17:25 EDT 1993
Received: from krusty.eecs.umich.edu by rand.org with SMTP id AA22706
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Mon, 18 Oct 1993 14:00:48 -0700
Received: by krusty.eecs.umich.edu id AA00654 ; Mon, 18 Oct 1993 17:00:55 -0400
Date: Mon, 18 Oct 1993 17:00:55 -0400
From: Karl Kluge <kckluge@eecs.umich.edu>
Message-Id: <199310182100.AA00654@krusty.eecs.umich.edu>
To: voynich@rand.org
In-Reply-To: <199310180002.AA03878@rand.org> (reeds@research.att.com)
Subject: Re: Voynich , Wilf.
Status: OR


The bib cite I have for his talk before the College of Physicians of Philly
(published in their proceedings) has Wilfrid.

Karl

From gauss!reeds Thu Nov 11 10:26 EST 1993
Received: by ninet.research.att.com; Thu Nov 11 10:42 EST 1993
Received: from research.att.com (ninet.research.att.com) by rand.org with SMTP id AA26153
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Thu, 11 Nov 1993 07:28:02 -0800
Message-Id: <199311111528.AA26153@rand.org>
From: reeds@research.att.com
Date: Thu, 11 Nov 93 10:26 EST
To: voynich@rand.org
Subject: Voynich Study Group of the 1940's
Status: OR

I'm looking for surviving members of the 1940's ``First Study
Group.''  Does anybody in our group know any such?  Does anybody
in our group know any of Friedman's colleagues?  If you do,
please ask!

Thanks!

Jim Reeds


From gauss!reeds Sun Nov 14 11:29 EST 1993
Received: by ninet.research.att.com; Sun Nov 14 11:33 EST 1993
Received: from research.att.com (ninet.research.att.com) by rand.org with SMTP id AA03795
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 14 Nov 1993 08:31:50 -0800
Message-Id: <199311141631.AA03795@rand.org>
From: reeds@research.att.com
Date: Sun, 14 Nov 93 11:29 EST
To: voynich@rand.org
Subject: Voynich essay
Status: OR


My attention was just pointed to an essay on the VMS by Terrence McKenna
in his book "Archaic Revival" that came out last year.  Has anyone read
it?


Jim Reeds


From gauss!mycroft.rand.org!jim Sun Nov 14 09:20:50 PST 1993
Received: by ninet.research.att.com; Sun Nov 14 12:22 EST 1993
Received: from mycroft.rand.org by rand.org with SMTP id AA04564
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 14 Nov 1993 09:20:52 -0800
Received: from localhost by mycroft.rand.org; Sun, 14 Nov 93 09:20:51 PST
Message-Id: <9311141720.AA21520@mycroft.rand.org>
To: voynich@rand.org
Subject: Re: Voynich essay 
In-Reply-To: Your message of Sun, 14 Nov 93 11:29:00 -0500.
             <199311141631.AA03795@rand.org> 
From: Jim Gillogly <James_Gillogly@rand.org>
Reply-To: James_Gillogly@rand.org
Date: Sun, 14 Nov 93 09:20:50 PST
Sender: jim@mycroft.rand.org
Status: OR


> reeds@research.att.com writes:
> My attention was just pointed to an essay on the VMS by Terrence McKenna
> in his book "Archaic Revival" that came out last year.  Has anyone read
> it?

I've read it.  He has a fair amount of it right, a fair amount of it wrong,
and a much better grasp than I of the cultural milieu of court life in the
Rudolf II period.  I think it's worth reading for the flavor and perspective,
but not for the detail specifically about the Voynich Manuscript.

McKenna also put out an audio tape about the Voynich some years ago.  I got
the same impressions about his depth of research in the period when
listening to it, without noticing the mistakes.  However, my recollection is
that it's largely the same material.  I believe the chapter in "Archaic
Revival" was written using the tape as a resource -- my evidence is that
there were numerous spelling errors that wouldn't have come up if the
source documents had been open in front of the writer/transcriber.

I believe his guess about the contents was similar to my current theory
(Sunday-Friday only), that Edward Kelley and an unnamed co-conspirator
constructed it, their motive being gold ducats.

	Jim Gillogly

From gauss!umcc.umich.edu!bgrant Mon Nov 15 00:08:47 0500 1993
Received: by ninet.research.att.com; Mon Nov 15 00:10 EST 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA16419
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 14 Nov 1993 21:08:50 -0800
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0oywBE-000AB8C@umcc.umich.edu>; Mon, 15 Nov 93 00:08 EST
Message-Id: <m0oywBE-000AB8C@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: FTPable Latin
To: voynich@rand.org
Date: Mon, 15 Nov 1993 00:08:47 -0500 (EST)
X-Mailer: ELM [version 2.4 PL23]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1032      
Status: OR

In order to calculate some letter and word frequencies I would like to   
find machine-readable samples of Latin and English (ideally from the
period when the VM was probably written, say 1200-1500 AD, but I'll
take anything).  Does anyone know of a source for something like this
on the net?

I've already posted a query to the LATIN-L mailing list with no results
so far, and looked through a list of online Latin projects on the
Georgetown U. system (guvax.georgetown.edu) but haven't found anything
readily accessible (e.g. FTPable).

For the English, one possibility might be the Project Guttenberg archive
of online texts, but I can't find the name of the site for it - does anyone
know offhand?

Incidentally, in looking through the Georgetown archive I came across
on of the truly great book titles of all time:

	"Thomae Aquinatis Opera Omnia Cum Hypertextibus in CD-ROM"

-Something old, something new ... 
****************************************************************************
 Bruce Grant -- bgrant@umcc.ais.org  

From gauss!unix.amherst.edu!rjyanco Mon Nov 15 00:43:08 0500 1993
Received: by ninet.research.att.com; Mon Nov 15 00:44 EST 1993
Received: from amhux3.amherst.edu by rand.org with SMTP id AA16933
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Sun, 14 Nov 1993 21:43:10 -0800
Received: by amhux3.amherst.edu id AA00432
  (5.65c+/IDA-1.4.4 for voynich@rand.org); Mon, 15 Nov 1993 00:43:09 -0500
From: "Richard J. Yanco" <rjyanco@unix.amherst.edu>
Message-Id: <199311150543.AA00432@amhux3.amherst.edu>
Subject: Re: FTPable Latin
To: voynich@rand.org
Date: Mon, 15 Nov 1993 00:43:08 -0500 (EST)
In-Reply-To: <m0oywBE-000AB8C@umcc.umich.edu> from "Bruce Grant" at Nov 15, 93 00:08:47 am
X-Mailer: ELM [version 2.4 PL17]
Content-Type: text
Content-Length: 1779      
Status: OR

> For the English, one possibility might be the Project Guttenberg archive
> of online texts, but I can't find the name of the site for it - does anyone
> know offhand?

Archie says:  gatekeeper.dec.com, nigel.msen.com, alice.fmi.uni-passau.de,
sunsite.unc.edu, ftp.uu.net, ftp.sunet.se, and unix.hensa.ac.uk all have
directories named "gutenberg."

I've been meaning to mention something, but didn't want to put it in
its own message:  earlier this semester I was presented with a sheet
showing a German script current in 1900.  (If I am correctly reading the
handwriting at the top of the page, the photocopy is from Harald Su"ss,
_Deutsche Schreibschrift_, Augsburg, 1992, p. 13)  What was interesting
was that although the letters were somewhat recognizable, when strung
together they were much less recognizable.

The problem is that the letters are broken up into components.  The
script "o" looks how I would write a cursive v, the "c" much like
an undotted cursive i, and (N.B.) the "a" looks like a cursive vi.
The "u" is merely the "n" with a mark above it; the "m" could pass
for "nc," "cn," or "ccc."  The "g" looks like "oj."

Now, if I were a scribe *copying* a document written in this script,
I wouldn't be able to make heads or tails of it (particularly if it
were in a foreign language).  My personal reaction would be to look
for repeated symbols, and consider them letters, possibly even
putting artificial space between what I considered letters.

While I would not argue that the VMS is written in 1900 German script,
I would toss out the suggestion that letters with space between them
may be components of the actual letters.  A letter which comes up as
a vowel may in actuality not be a letter at all, but a common component
of the actual letters.

Rick


From gauss!reeds Wed Nov 17 18:39 EST 1993
Received: by ninet.research.att.com; Wed Nov 17 18:41 EST 1993
Received: from research.att.com (ninet.research.att.com) by rand.org with SMTP id AA29482
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 17 Nov 1993 15:39:19 -0800
Message-Id: <199311172339.AA29482@rand.org>
From: reeds@research.att.com
Date: Wed, 17 Nov 93 18:39 EST
To: voynich@rand.org
Subject: Voynich Study Group, First, IBM Card Transcription of
Status: OR


Here is an update on the status of the IBM card transcription made
in the 1940's by the First Study Group of Army cryptanalysts under
William F. Friedman.

The Marshall Library has a printout of a transcription of the whole
VMS, which Jacques Guy and I started to type into the computer.
I have typed the whole thing in;  Jacques typed 110 pages out of 130,
and will finish when he returns early next year from an extended 
vacation.  Then we will compare our versions & correct all typing mistakes
thus found.  We are writing a paper for Cryptologia about the Friedman
transcription, and will make an electronic copy of the transcription
available in some form or other.

Just today I got formal permission from the Marshall Library to publish
this material.

After that, the next step would be to cast the Friedman transcription
into Currier and to compare it with our own transcription on rand.org,
marking (and eventually correcting) all differences.  At that point we
will have a complete transcription of the VMS, with about half of its
material checked against independent transcriptions.

(And after that the plan is to break the code and to read that mysterious
manuscript.)

Jim Reeds


From gauss!nyx10.cs.du.edu!rcarter Wed Nov 17 17:33:50 0700 1993
Received: by ninet.research.att.com; Wed Nov 17 20:11 EST 1993
Received: from mordor.cs.du.edu by rand.org with SMTP id AA01288
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 17 Nov 1993 16:34:00 -0800
Received: from nyx10.cs.du.edu by mordor.cs.du.edu with SMTP id AA00396
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 17 Nov 1993 17:33:22 -0700
Received: by nyx10.cs.du.edu (4.1/SMI-4.1)
	id AA13202; Wed, 17 Nov 93 17:33:50 MST
From: rcarter@nyx10.cs.du.edu (Ron Carter)
Message-Id: <9311180033.AA13202@nyx10.cs.du.edu>
X-Disclaimer: Nyx is a public access Unix system run by the University
	of Denver.  The University has neither control over nor
	responsibility for the opinions or correct identity of users.
Subject: Thanks and congrats...
To: voynich@rand.org
Date: Wed, 17 Nov 1993 17:33:50 -0700 (MST)
X-Mailer: ELM [version 2.4 PL21]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 347       
Status: OR

Thanks to Jim and (aaarrggghhh, who else worked on it?) for doing
the punch card transcription...  Good show people!

Perhaps you could summarize what stuff still needs to be double
entered/transcribed so as to provide the check for the half that
wasn't in the punch cards?

-- 
Regards, Ron | rcarter@nyx.cs.du.edu | Denver, Colorado | 17 NOV 93

From gauss!reeds Wed Nov 17 22:45 EST 1993
Received: by ninet.research.att.com; Wed Nov 17 22:52 EST 1993
Received: from research.att.com (ninet.research.att.com) by rand.org with SMTP id AA05383
  (5.65c/IDA-1.4.4 for <voynich@rand.org>); Wed, 17 Nov 1993 19:46:44 -0800
Message-Id: <199311180346.AA05383@rand.org>
From: reeds@research.att.com
Date: Wed, 17 Nov 93 22:45 EST
To: voynich@rand.org
Subject: Voynich transcription: don't hold your breath
Status: OR


Ron Carter asks:

> Perhaps you could summarize what stuff still needs to be double
> entered/transcribed so as to provide the check for the half that
> wasn't in the punch cards?

It is confusing, and I have not yet bothered to tote up just what
is covered already & what needs to be done.    I have a new version
of my 'checklist' which has some of this info in it; maybe Jim
Gillogly will put it in the peanut gallery for us.

Very briefly, the following transcriptions exist:

1.  The main First Study Group transcription of 1946, item 1609 in 
the Friedman Collection of the Marshall Library, which covers the 
whole VMS, minus one or 2 skipped pages, captions, and writing in 
diagrams.  This is what my previous post called the "IBM card 
transcription".  We don't know where the cards are: item 1609 is a
printout listing of the cards.  My on-line version of it takes up about 
260,000 characters, of which about 25% are line and page number info,
with maybe 190,000 transcription "payload".  A strange thing about 
this transcription is that the FSG people did not record line ends.
(Pages in this transcription marked '1SG' in my 'checklist'.)

2.  A subsidiary FSG transcription, item 1613 in the Friedman collection,
apparently done independently of the main FSG transcription effort,
covering folios f.111v through f.114r, which takes up about 16,000
characters in raw form, or about 20,000 when put into our own format
(with line numbers, etc.)  I described this about 2 months ago in
this forum; Michael Roe figured out the transcription alphabet used
(a simple substitution of the FSG alphabet listed in D'Imperio's book).

3.  The D'Imperio transcription, voynich.orig on rand.org, made ca. 1976,
which is about 100,000 characters long, with maybe 80,000 or 90,000 
characters transcription payload.  (Pages in this transcription maked 'D'
in my 'checklist'.)

4.  Our own revision and amplification of the D'Imperio transcription,
voynich.now on rand.org, which is about 165,000 characters long.  This
is a superset of 2 and 3.  I guess that it has a payload of about 130,000 
characters.  (Pages in this transcription marked either 'D' or '3SG'
in my 'checklist'.)

Hence the main FSG transcription has about 60,000 characters not covered by
other transcriptions available to us.

But this figure is misleading, because the FSG transcription alphabet is
somewhat wordier than the Currier alphabet.

When Jacques finishes his part of the task we will (1) merge our
versions into an "official" on-line copy of the FSG transcription,
(2) recast it into Currier, (3) find and mark all deviations between
it and the other transcriptions.  The final results will probably
be a much larger version of voynich.now, possibly available in March
or April.

Then and only then will the exact coverage be known, & the amount
of discrepancy between versions be clear.   

One annoying thing about this whole thing is that I don't know the names
of any of the members of the First Study Group (except for Friedman,
their leader).  The relevant records were at NSA when D'Imperio wrote
her book; they might no longer exist, and even if they did exist, and
friendly spooks located them for us, they still might not be releasable.

So we will all owe a debt to Friedman and (for now) to an unknown number 
of unknowns.

And of course to Jacques Guy, for typing all this stuff in.  And to
my colleague Henry Baird, for helping me OCR it in.


Jim Reeds


From gauss!umcc.umich.edu!bgrant Tue Nov 23 23:33:31 0500 1993
Received: by ninet.research.att.com; Tue Nov 23 23:36 EST 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA08677
  (5.67a/IDA-1.5 for <voynich@rand.org>); Tue, 23 Nov 1993 20:33:35 -0800
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0p2Bv1-000ABJC@umcc.umich.edu>; Tue, 23 Nov 93 23:33 EST
Message-Id: <m0p2Bv1-000ABJC@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: Medtextl Mailing List
To: voynich@rand.org
Date: Tue, 23 Nov 1993 23:33:31 -0500 (EST)
X-Mailer: ELM [version 2.4 PL23]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 676       
Status: OR

I have just discovered a mailing list that may be of interest
to Voynich manuscript fans:

	Medtextl is a mailing list concerning the study of medieval 
	manuscripts.  Some recent postings have discussed the proper
        way to preserve manuscripts, whether it is better to work from
	an original or a microfilm copy, and the origin of "mooning"
	(possibly in Canterbury Tales.)

To subscribe send a message to 'listserv@uiucvmd.bitnet' containing the 
text "SUBSCRIBE MEDTEXTL <your name here>".  To unsubscribe, the text
must be "SIGNOFF MEDTEXTL".

Bruce

****************************************************************************
 Bruce Grant -- bgrant@umcc.ais.org  

From gauss!umcc.umich.edu!bgrant Sat Dec  4 12:19:49 0500 1993 remote from rice
Received: by ninet.research.att.com; Sat Dec  4 12:22 EST 1993
Received: from umcc.umich.edu (umcc.umcc.umich.edu) by rand.org with SMTP id AA15336
  (5.67a/IDA-1.5 for <voynich@rand.org>); Sat, 4 Dec 1993 09:19:52 -0800
Received: by umcc.umich.edu (/\==/\ Smail3.1.25.1 #25.6)
	id <m0p60e5-000ABkC@umcc.umich.edu>; Sat, 4 Dec 93 12:19 EST
Message-Id: <m0p60e5-000ABkC@umcc.umich.edu>
From: bgrant@umcc.umich.edu (Bruce Grant)
Subject: FINDING ENDINGS
To: voynich@rand.org
Date: Sat, 4 Dec 1993 12:19:49 -0500 (EST)
X-Mailer: ELM [version 2.4 PL23]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1818      
Status: OR

Is anyone on this list aware of an algorithm for identifying possible
grammatical endings in a body of text?  Intuitively, it seems to me that 
an ending ought to be a final string which occurs frequently but can't be
extended to a longer string which is also very frequent.

For example, in Latin, given a word like AMICUS, -US would be considered
to be the ending of the word because there are lots of words ending in
"-US" but not too many ending in "-CUS".  On the other hand, in the word
"OMNIBUS" the ending would be "-IBUS" because this ending is pretty 
frequent but "-NIBUS" is much less so.

In general, longer strings will be less frequent than their shorter  
substringsi.  It seems like you could organize the possible endings of 
words in a tree, then start from the leaf corresponding to your
word and move toward the root 'til the frequency jumped sharply.)

            S    (freq)
            |
            US   (freq)
            |
            +----------------+------------------+------------ etc.
            |                |                  |
            BUS (freq)       CUS (not)          MUS (freq)
            |                |                  |
         >> IBUS (freq)      ICUS (not)      >> AMUS (freq)
            |                |                  |
            NIBUS (not)      MICUS (not)        MAMUS (not)
            |                |                  |
            MNIBUS (not)     AMICUS (not)       AMAMUS (not)
            |
            OMNIBUS (not)

Since the frequencies for shorter strings will be higher anyway, though,
the problem is how to define what constitutes a "sharp" jump.

Has anybody ever seen an algorithm like this?

Bruce
														
****************************************************************************
 Bruce Grant -- bgrant@umcc.ais.org  

