The Voynich Ninja

I would like to bring your attention to the Austrian artist August Walla (1936-2001), who suffered from schizophrenia and spent most of his time in psychiatric institutions (or at least under medical supervision). There, he was encouraged to express himself artistically, and became a member of the “Gugging artists” who received praise and appreciation from modern painters and art collectors. His works now are on display in several museums. His life, and also his clinical record, are well documented.

Walla’s style involved the inclusion of text into his paintings. Some are pure text, but most of them combine text and figurative illustration. In those cases, the text is somewhat integrated into the picture and used to fill any voids between the figures, obviously following aesthetic principles as much as its inherent “meaning”.

Some of his texts are (kind of) understandable and some are not, some appear to be meaningless. To him, they were undoubtedly meaningful. In some sense, he used his art to communicate; in some sense, to follow an urge to express himself and his religious and cosmological beliefs (which are actually quite complex). Very likely, he attributed some magical component to his work.
 
He “chose” painting as his form of artistic expression because he knew this from his early education, and because was encouraged by his environment. In Gugging, the necessary materials were also supplied. 

In his extensive oeuvre, a small number of topics prevail – family, religion, politics. Single works as well as series of his works are extremely repetitive. He even will repeat single words or letters for a couple of times. 

He invented “languages”, using dictionaries. Also the words he invents are used very repetitively, or he combines them to create „new“ words. Walla even created  a rune-like set of characters that he used occasionally. But mostly he wrote “plain text”, using capitals and an idiosyncratic punctuation. 

Some of his work reads like this (and I will deliberately start with one that is not really representative, but really touching: #1, then introduce some “weird” ones: #2-5 and finally some “meaningless” examples: #6-8).

Quote:[Text1]

Ewigkeitvorherzeit von Ewigkeitzeit.?

WELTALLENDEFINSTERNIS UND DIE EWIGKEITENDEZEITRIESENUHR MIT ZEIT VOR DER EWIGKEIT UND MEINEM VERJÜNGERUNGSWUNSCH 11 JAHRE JUNG ZU WERDEN.?

Interlinear translation: 

Eternitybeforetime of Eternitytime.? UNIVERSEENDDARKNESS AND THE ETERNITYENDTIMEGIANTCLOCK WITH TIME BEFORE THE ETERNITY AND MY REJUVENATINGWISH 11 YEARS YOUNG TO BECOME.?

Quote:[Text2]

Walla August.!

BIN DER HEILIGE GOTT SAÄRIL.! GOTT SAÄRILL.! GOTT ALLAH.
SONNE HINTERM BISAMBERG IN DEN BEZIRK KORNEUBURG AN DER DONAU?
SARARILL UND ALLAH.! KPÖ I 1 II 2 I 1 III 3

Walla August.! AM THE HOLY GOD SAÄRIL.! GOD SAÄRIL.! GOD ALLAH! SUN BEHIND BISAMBERG IN THE COUNTY OF KORNEUBURG [home of August Walla] BY THE DANUBE? SARARILL AND ALLAH.! KPÖ [=Communist Party of Austria] I 1 II 2 I 1 III 3

Quote:[Text3]

STERN, ADOLFE. STERN, MEMMAK.? STERN, GUURU.? STERN, MERKUR. STERN, PANKI. STERN MARS.! WENUS. ERDEN.! SATURN. URANUS.! BLUTO.! SONNE. MOND.! STERNE.

Stern, pl. Sterne = star; Erde = Earth. This was written on a window pane.

Quote:[Text4]

HHH
ICH BRAVER HALBTEUFEL DANKE, DEM GOTTESSOHN JESUS CHRISTUS WENN ICH HEIMKOMME NACH KORERNEUBURG
Ich lobe solchen braven Menschen die heimbringen dem braven Burschen August Walla.!
HALBTEUFLEIN AUGUSTIN ALOIS WALLA GEORG.! TEUFEL MIT BLONDEN HAAREN, IST EIN SÜSSER STROLCH, ALS STRITZIBUBE.!
Sepp.! Sepp.!
GESPENST.! SÜSSER BRAVER STRITZIBUBE IST DA GLÜCKBRINGENTER SO DUNKELBLONDER KIRCHENEHRLOSER FREUNDLICHER TEUFEL ALS SO HALBTEUFEL HIER.!

In the following translations, I had to smooth some things out for the sake of readability: 

HHH I GOOD HALFDEVIL THANK, THE SON OF GOD JESUS CHRIST WHEN I COME HOME TO KORNEUBURG. I praise such good men who bring home the good lad August Walla.! HALFDEVIL AUGUSTIN ALOIS WALLA GEORG.! DEVIL WITH BLOND HAIR, IS A SWEET HOODLUM, A RASCAL.! Sepp.! Sepp.! [=short for Joseph] GHOST.! SWEET GOOD RASCAL IS HERE LUCKY SO DARK BLOND CHURCHDISHONOURABLE FRIENDLY DEVIL AS SO HALFDEVIL HERE.!

Quote:[Text5]

+ WALLA AUGUST + SO BILDL FÜR BRAVN BRAVN BRAVN HERR KURT HOFBAUER + BILDL EIN BRAVER BUNTER PAPAGEI + DU LIEBER GOTT ALLES BÖSE KOMMT WIEDER, UND ALLES BÖSE VERGEHT WIEDER + PAPAGEI MARIA + 1997.! +17

+ WALLA AUGUST + SUCH PIC FOR GOOD GOOD GOOD MISTER KURT HOFBAUER + PIC A GOOD COLORFUL PARROT + YOU GOOD GOD ALL EVIL RETURNS AGAIN, AND ALL EVIL PERISHES AGAIN + PARROT MARIA + 1997.! +17

The picture does not show a parrot, but a box of watercolors.

Quote:[Text6]

Jeaobek Jeaobek Adolfe, gpatb autbnaa akgn tb akgntb Adolfe, Gor. Saaahnh jmn Hego Aoxbhego Bocefskoheuheago.?

(all gibberish except for „Adolfe“, which is followed by a swastika)

Quote:[Text7]

KISI KOMISER SAIRE KISI OGUSTO LAP LASTIRMAK KISI KOMISER. SAIRE PUT GOTT, KOF SLAMNÜNYA SAIRE KOF EZELF.! Lied.! KITLEMEK ISLÄM. PUT SAAAHNHX PUT PUT PUT SAAAHNHX.! FAKAT FAKAT MEK ESYA=KOMISER. MEK ESYA, KOMISER.

(all gibberish except for “OGUSTO” which sounds like Italian/Spanish… = August; “GOTT” = God; “Lied” = Song; maybe “ISLÄM”?)

Quote:[Text8]

MELICIDIUM ETTOQUE PÜPUS + ISCRIX + GRÄMEN ET LIQUOR TRUTTANUS MELICIDIUMTRUTTANUS ISCRIX AUGUST + GRYPS + GRYPS COMPOS GRYPSCOMPOS.! CAELESSARARILLH, SARARILLH.!

(Latin. Isn’t it?)

So, if you are still with me, I would like to bring up the question: if we would examine the texts of Walla – especially those which seem to bear no meaning - the same way we examine the VMS, would the results be similar?

Very interesting.

Hi Ben. I'm a medical doctor who has long had an interest in, and experience treating, mental health issues. I'm not a psychiatrist, but I used to work as a general practitioner in a location where mental health services were not widely available, and so it's something I needed to get good at. I've also been a lifelong "word nerd" and amateur linguist, and so one subject that has always fascinated me is the specific ways language gets distorted by patients with mental illnesses, especially ones like schizophrenia and Alzheimer's dementia, which seem to directly affect the language centers of the brain.

I won't bore you with the technical details, but what I can say is that in all such cases that I've seen, there is indeed a "method to the madness". Grammar, vocabulary, prosody, and word usage can become so distorted that it's very hard to understand. Specifically, it can be very hard to pick out a point or purpose to the speech. But often, if you transcribe and analyze a large amount of the speech of a schizophrenia or dementia patient, there are noticeable themes and intentions apparent. It's just that the way they're worded is odd and not at all straightforward. Schizophrenia and dementia patients are usually not aware that their language is not readily understandable to most people; it does make a kind of sense from their subjective perspective, which is also quite distorted. The "key" to unlocking the confusing speech of a psych patient is usually a friend or family member who knew the patient well, before their mental illness became severe. Without this, it can be very hard to piece together what intentions they were trying to communicate, and of course the dead don't talk.

One theory I've heard expressed by several Voynich researchers is that the VMS was the product of an educated person from a wealthy family, who descended into mental illness. He was sequestered away out of family embarrassment, with as much writing materials as he needed to keep him busy, and the VMS was his solitary life's work. It would be interesting to try statistically analyzing large samples of writing (or transcribed speech) of patients with various sorts of mental illnesses, to see if the statistics were closer to what we see in the VMS than most written language. I'm particularly interested in the writing of You are not allowed to view links. Register or Login to view. patients — a condition I have never seen or treated clinically. Of particular interest, the writing of hypergraphia patients is usually voluminious, and the patient feels compelled to write exactly what comes to his mind, even though the writing is usually questionably coherent to others reading it. I suspect that self-made soapmaker Emanuel Bronner suffered from undiagnosed hypergraphia. Take a look at the "preaching" on his soap labels:
You are not allowed to view links. Register or Login to view.

(10-03-2020, 12:46 AM)RenegadeHealer Wrote: You are not allowed to view links. Register or Login to view.One theory I've heard expressed by several Voynich researchers is that the VMS was the product of an educated person from a wealthy family, who descended into mental illness.

 
Yes, you run into this opinion - and related theories, such as glossolalia - quite frequently. But it always comes with a tone like, "...but there is no way of knowing". That's exactly what I would like to scrutinize. 

In the examples I provided, I think one can see possible relations. #6 surely is as close to Voynichese as #8 is to Latin. 
More seriously, in the text samples #6-8 you can see some of Wallas techniques of repetition, permutation and combination of "words" that (to me) bear resemblance to the movements in the VMS. You also see the persistence of "words" over an extended period of time (cf. SARARILL(H), who of course is one of the more prominent private deities of Walla, or SAAAHNH(X), which seems to be some kind of incantation).

(10-03-2020, 12:46 AM)RenegadeHealer Wrote: You are not allowed to view links. Register or Login to view.It would be interesting to try statistically analyzing large samples of writing (or transcribed speech) of patients with various sorts of mental illnesses, to see if the statistics were closer to what we see in the VMS than most written language.

It is actually the goal of this thread to find out if there is interest for a project like this. Personally, I could imagine this would more fruitful than doing the n-th iteration of a test on Pliny, Dante or Moby Dick. And AFAIK, there has been surprsingly little research done into this direction (also regarding the drawings, but that would be an entirely different subject).

(10-03-2020, 12:46 AM)RenegadeHealer Wrote: You are not allowed to view links. Register or Login to view.I'm particularly interested in the writing of You are not allowed to view links. Register or Login to view. patients.

I don't recall I have seen works of an artist diagnosed with hypergraphia. I'm pretty sure that there is none among the Gugging artists. Walla, on the other hand, was quite manic in producing his works, not only confining himself to paper and canvases, but painting/writing on household articles, rooms, walls, houses, streets ... (which frequently got him into trouble). So there would be enough text for analysis.

(10-03-2020, 04:42 PM)Ben Trovato Wrote: You are not allowed to view links. Register or Login to view.It is actually the goal of this thread to find out if there is interest for a project like this.

... but unfortunately, that does not appear to be the case.

(14-03-2020, 08:45 PM)Ben Trovato Wrote: You are not allowed to view links. Register or Login to view.

(10-03-2020, 04:42 PM)Ben Trovato Wrote: You are not allowed to view links. Register or Login to view.It is actually the goal of this thread to find out if there is interest for a project like this.

... but unfortunately, that does not appear to be the case.

I think this is definitely worth trying.

(14-03-2020, 09:34 PM)Alin_J Wrote: You are not allowed to view links. Register or Login to view.I think this is definitely worth trying.

I think so too, obviously ... but: I lack the mathematical background to perform these tests or to interpret the results.

What I could do is the following: digitalizing text of Wallas work (ideally several corpora of different text "genres") and make it accessible to more competent researchers. But as this is some amount of work, I'd only start if it is secured that somebody would do something with it.

... "worth trying", of course it would be!

Ben Trovato

(16-03-2020, 12:52 PM)Ben Trovato Wrote: You are not allowed to view links. Register or Login to view.

(14-03-2020, 09:34 PM)Alin_J Wrote: You are not allowed to view links. Register or Login to view.I think this is definitely worth trying.

I think so too, obviously ... but: I lack the mathematical background to perform these tests or to interpret the results.

What I could do is the following: digitalizing text of Wallas work (ideally several corpora of different text "genres") and make it accessible to more competent researchers. But as this is some amount of work, I'd only start if it is secured that somebody would do something with it.

... "worth trying", of course it would be!

Ben Trovato

I agree that would be interesting, in particular for "genres" which are large enough (ideally, a few thousand words). Could you please provide a rough estimation of the sizes of these corpora? 
Could you please also attach a sample image of Walla's handwriting? I would like to see if I can read it and help you with the transcription.

The two "genres" that come to mind immediately are the "weird" ones (#2-5 in the first post) and the "gibberish" ones (#6-8), which, in my view, seem most interesting. Both are are very carefully and "neatly" done, so reading is not an issue. Availability is. Although you may find a couple of pics online, the most extensive documentations are printed (and of course, all the librarires are closed...)

But in general, there is a huge amount of material accessible. For the first group, I have no doubt that one could extract tens of thousands characters. For the second group, I honestly don't know, but my guess would be thousands, maybe much more. In fact, that would be one question I'd like to discuss - how much material it takes for proper testing.

Then there are his "runes" that I mentioned above. He used them not so frequently, and AFAIK, there has been no effort to "decode" them. 

I attach two pictures, one from which text #2 is taken and a photo of Walla presenting a table with an inscription using his "runes" (which are written in two seperate layers).

[attachment=4090]

[attachment=4089]

Thank you, Ben!
I am not sure the "runes" are readable. At least one of the two layers appears to be written in the Latin alphabet, but it seems very hard to tell characters apart in a black and white picture.

On the other hand, the colour image is quite readable.


About  sample size: You are not allowed to view links. Register or Login to view., Anton (who certainly is more knowledgeable than myself) proposed 10,000 characters. When I tried to address the task he proposed, I transcribed You are not allowed to view links. Register or Login to view.: this took me a couple of months, but I worked on both a medieval manuscript and a printed edition, carefully comparing the two sources. Walla should be more manageable.

The painting you posted contains almost 200 characters. If that is indicative of the average text quantity in a single painting, transcribing 50 or 60 paintings could be enough. I don't speak German, but the text superficially looks like "proper" German to me. It seems possible that the statistics of the Walla transcription will not be dramatically different from those of a German text, but we will see...