Calgary engineer believes he's cracked the mysterious Voynich Manuscript

[Ahmet Ardıç]

Thank you for the replies.

As I mentioned before, I also know that artificial intelligence capabilities do not always make the right inference, and of course, in its current state, artificial intelligence has limited reliability.

However, when asking questions about the topics you want artificial intelligence to research, you can ask artificial intelligence to look only at academic publications and ignore all other sources. He can even look at academic criticisms written for those academic articles. In a very short time, artificial intelligence can accurately perform tasks that humans can do for decades, such as compiling information, classifying, listing, and summarizing information as desired. Moreover, it continues to improve day by day in terms of working speed and conveying the results correctly with correct inferences.

As a result, when used correctly, artificial intelligence is also very useful for researchers, and its increasing evolution towards more/intensive use will definitely continue in the future.

The purpose of asking these questions to artificial intelligence is; If we are to transform human speech sounds into written images, it is to show that there are a limited number of simple single sounds. Even when syllable sounds and word signs are included, writing signs (letters, etc., all sound images) are limited to certain numbers for each language, and this is a feature that comparative linguistics and statistical studies should take advantage of.

How many different sounds does the VM alphabet contain? What would you think about the numerical overlap with the sound images of the VM alphabet in terms of the number of signs of the alphabet in various languages, especially Indo-European and Semitic (if you compare them with all the languages you know)? What data or guidance would such information provide us with in terms of the VM-sound-images we examine and investigate?

Now please see the last question I asked the GPT4 artificial intelligence and its answer below:

My Question > 
   

GPT4 Answer > 
   

As of my last update in July 2023, the specific questions you're asking can be addressed with the information available up to that point:
1- Highest Number of Different Sounds in Spoken Language:

• The Khoisan languages of southern Africa are renowned for having the highest number of phonemes, particularly due to their extensive use of click consonants. Specifically, languages like !Xóõ (Taa) are often cited for this distinction. !Xóõ has been reported to have upwards of 100 to 164 phonemes, including an extensive array of clicks.
  

2- Language with the Most Text/Sound Signs or Letters in its Alphabet:

• The Khmer alphabet (Cambodian) is frequently cited as having the largest alphabet in the world. It consists of 74 letters: 33 consonants, 23 dependent vowels, and 12 independent vowels. However, it's worth noting this is for the alphabet itself; the script doesn't use capital letters, so the total doesn't change for uppercase/lowercase distinctions, as that concept doesn't apply. The Khmer script is an abugida, where each consonant-vowel sequence is written as a unit, so 0 of these writing signs are purely syllable or word images in the sense used by logographic systems. Each sign can represent a syllable, but they are built from combinations of consonant and vowel indicators rather than being single complex characters that represent entire syllables or words.
  

Question>  
   
The Answer > 
   

Drawing exclusively from academic research and focusing on your specific questions:
1- Manuscript with the Most Alphabet and/or Script-Image Characters among Medieval European Peoples:

• The Voynich Manuscript is written in an unknown script with about 170 unique characters. Although not a traditional language or alphabet per se and still undeciphered, it represents the manuscript with one of the largest sets of unique script-image characters found in medieval European documents.
  

2- Ancient European Language with the Largest Number of Letters:

• Gothic, the language of the Goths, utilized by Ulfilas in his translation of the Bible, had an alphabet that is noted for its size among ancient European languages. Ulfilas created an alphabet to transcribe the Gothic language, which consisted of 27 letters derived from Latin, Greek, and Runic sources.
  

3- Today's European Language with the Largest Number of Letters:

• The Icelandic language has one of the larger alphabets in contemporary European languages, with 32 letters. Unlike most languages, it has preserved much of its Old Norse ancestor's phonetic complexity, including letters not found in most other Latin-script alphabets.
  

4- Writing Images/Letters in the Mentioned Language (Icelandic):

• If we're focusing on Icelandic as the language with the most letter signs mentioned, the count of distinct writing images/letters is 32, as Icelandic does not differentiate between uppercase and lowercase letters in the total count of its alphabet. Each character, regardless of case, represents a unique phoneme or sound.
  

Meanwhile, the artificial intelligence gives "The Voynich Manuscript is written in an unknown script with about 170 unique characters." I must say that the information or explanation does not accurately reflect the correct result. How many different alphabets and how many different sound-images/letters do you think there are in VM texts?

Question > 
   
Answer > 
          
   

Given the specific focus on academic sources and the exclusion of general or non-academic information, the task is somewhat challenging due to the limitation of accessing real-time databases or up-to-date academic publications directly for verification at this moment. However, based on established academic knowledge up to my last update, I can provide general insights that align with scholarly consensus on the script and alphabet development of these languages through the medieval period. It's worth noting that exact numbers might slightly vary depending on specific manuscripts or regional variants, as medieval scripts were not always standardized.
European Medieval Manuscripts

1. Slavic Languages (Cyrillic Alphabet): The early Cyrillic alphabet used in medieval Slavic manuscripts, derived around the 9th century, originally had about 43 letters. This number could vary slightly over time and between regions as the alphabet adapted to the phonetic needs of different Slavic languages.
   
2. German (Old High German and Middle High German): Initially used the Latin alphabet with a few additional letters to represent specific German sounds. The medieval alphabet for Germanic languages might include up to 29 letters, accounting for characters like Æ/æ, Þ/þ (thorn), and Ʒ/ʒ (ezh), although not all were used consistently across manuscripts.
   
3. English (Middle English): Middle English used the Latin alphabet, supplemented with a few additional letters or letter combinations to represent English sounds not found in Latin, such as þ (thorn) and ƿ (wynn). The total could be around 30 letters, considering the inclusion of such characters.
   
4. Latin/Italian: The medieval Latin alphabet, also used for vernacular Italian in manuscripts, remained fairly consistent with the classical Latin alphabet of 23 letters. Over time, particularly towards the Renaissance, the use of J, U, and W became more distinct, potentially bringing the count to 26. Italian, evolving from Latin, largely followed this script but with evolving orthographic conventions.
   
5. Greek: The medieval Greek alphabet used in Byzantine manuscripts consisted of 24 letters, consistent with the classical Greek alphabet. There were also diacritical marks introduced for pronunciation guidance, but the primary letter count remained unchanged.
   

Medieval Hebrew & Arabic Writing

1. Hebrew: The medieval Hebrew alphabet contained 22 letters, as it has throughout much of its history. This does not include diacritical marks (niqqud) introduced in the medieval period for vowel representation, which are not considered separate letters.
   
2. Arabic: The Arabic script in the medieval period maintained its 28 letters. However, the script is an abjad, initially recording consonants with optional marks for vowels (harakat) added in manuscripts for clarity, especially in the Quran or in educational contexts.
   

It's important to note that manuscripts can exhibit regional and chronological variations, and scribes often employed ligatures, abbreviations, and unique script forms, complicating a straightforward enumeration of letters or signs. Additionally, the evolution of writing practices over the medieval period means that the number of letters or written signs could slightly vary depending on the time and location a manuscript was produced.


I ask the following questions to researchers:
1- While the number of letters/writing images in European Medieval Languages Manuscripts is limited in number, why is the number of letters and/or writing images in the Voynich manuscript higher than in manuscripts known to be European? (Or, in your opinion, if the number of writing signs in VM texts overlaps with the familiar European medieval scripts, how many letters were there in which language, and how many letters do you think there are in VM?)
2- How many different languages are actually in the Voynich texts?
3- How many different alphabets are used in VM texts?
4- How many different writing images or letters are there in VM texts that look different from each other? When giving a number, please indicate the upper and lower case letters separately, or state that the number you give represents the sum of both upper and lower case letters.
5- When you take into account all the transcriptions made for EVA variations and VM, which of these transcriptions and EVA variants probably contains the most accurate Latin alphabet equivalents of VM signs? (I'm curious about your personal opinion.)
6- Is there anyone who can say that the number of matches (alphabets) in the alphabet transcription table that you accept as "correct" does not coincide with the number of differences or text images in the VM?