The question is: If one assumes that the manuscript could perhaps be of Arabic origin, yet has no comprehension of the Arabic language, then how can one obtain any clarification at all?
Well, some characters are readable due to the fact that they are written in Latin script, ‘a’, ‘o’, and ‘y’. It is also the case that the terms in Arabic script can be converted into Latin script through a process called Romanization. In Western-language publications, some of these terms recur time and time again, such as ‘Islam’ and ‘Muslim’, etc.
There are some software tools available on the Internet these days that are capable of automating the Romanization process, as exemplified here. If an Arabic text is entered, then a conversion of the text in Latin script is returned (including the Abjad form, as well as the full vocalisation).
The results can be analyzed in more detail, particularly those words that are shorter and more common, and compared to the words in the manuscript to assess whether any similarities can be noted at some point. And in fact, it has occurred to me, at least, that the word ‘alty’ appears in a similar form in folio 58r, line 10.
This immediately led to the discovery of a new character, ‘t’.
Subsequently, this immediately is followed by another question: how can one be so sure that such words are actually Arabic?
Coincidentally, we happened upon a solution for this using the “Arabic-English” language combination in Google Translate, where Roman letter typing is enabled.
What you do is enter a string of words using Latin characters, and it will render a translation in English as well as the transliteration of the corresponding Arabic script! This only happens when Google also finds the proper term. Otherwise, it can be assumed that we are *not* dealing with an Arabic term.
So, the deciphering process becomes somewhat of a letter puzzle through trial and error.
You take a word from the manuscript, put in a new letter, enter this into the Translate window, and see if the output is recognized as Arabic. What can be particularly frustrating is when compound words (which are not permitted in Arabic) need to be processed. They will not be recognized as Arabic, at first, and it can throw you off and be a bit confusing.
If all goes well in the end, you will end up with a transcription map, where each individual glyph from the manuscript is mapped onto the corresponding Latin character found, including the Arabic letters which mostly are appropriate for translation.
It is essential that this map works consistently for the entire manuscript. It must be noted that the Latin characters merely correspond to a single (contextual) value that Google “understands” and can translate.
It is also not the case that the characters from the Arabic script are mapped onto that 1-to-1.
Thus, the text from the Voynich Manuscript is written in some form of Arabic.
It should be no secret that the words carry different connotations in certain circumstances. Hence, depending on the context, they will result in a different Arabic typeface and mean different things as well.
Every resultant translation, thus, will require a considerable amount of combinatorial analysis and a great degree of imagination.