Arabic Auxlang Aborts Before Take-Off

First off, here's a provisional name for my hypothetical Arabo-Islamic auxlang, that I postulated in my post called Arabic Auxlang Anyone?

As a contrast to the name of the romanoclone auxlang name Occidental, I dub my hypothetical language Oriental. The Oriental which I am referring to isn't in the more modern sense where it's understood to mean the Far East and cultures such as Chinese, Japanese and Korean. I'm harking back to the old-fashioned Oriental of the 19th century and the Orientalist aesthetic movement. At this time Oriental included more the Near East and Middle-Eastern, Turkic, Persian and Indic cultures. This realm is pretty close the what this auxlang would cater to.

The most relevant passage of my previous post would have to be:

The source languages that I would propose are:

Modern Standard Arabic

Turkish

Persian/Farsi

Swahili

Hindi-Urdu (One or two languages depending on your political persuasion)Indonesia-

Malay (Essentially two standards of a plural-centric language) 

Perhaps the criteria for words could be that they need to have cognates present in 4/6 of the above languages. Or 3/6 if that's too stringent.

Interestingly enough, with the exception of Persian and Hindi-Urdu, these six languages are genetically all un-related. Contrast this [to] Interlingua, where all the source languages are Indo-European and the majority are Romantic.

I've bolded the really critical fact. Here's the problem. Undoubtedly, these 6 languages have borrowed, lent and shared a large quantity of vocabulary. Some through the spread of Islam, some through the military conquests of the Persians and Turks, some through the spread of Hinduism and Buddhism, some through Arabic traders on the Indian ocean. But whereas languages might borrow large numbers of words when in contact with other cultures, they generally have a core-vocabulary that tends not to be replaced with borrowings. These words tend to be the most basic fundamental words in the vocabulary, words for such things as louse, two, water, ear, die, I, liver, eye, hand, hear, tree, fish, name, stone, tooth, breasts, you, path, bone, tongue, skin, night, leaf, blood, horn, person, knee, one, nose, full, come, star, mountain, fire, we, drink, see, new, dog, sun. These are the words that are typically found in a Swadesh list. The list that I just gave was the 40 concepts that have been found least likely to change over the aeons of a language's history.

My critical stumbling block for Oriental, is that these 6 languages are unlikely to have much shared vocabulary for the most basic of words. Oriental will probably have less trouble finding words for law, religion, government, commerce, science, technology agriculture and warfare. But for the most basic concepts, there will be little common ground between the 6 languages.

Take the case of a typical romanoclone auxlang; Interlingua, Of its source languages French, Spanish, Portuguese and Italian are all closely related Romantic languages. The other source languages; English, German and Russian are still share a genetic relationship, albeit more distantly by way of Proto-Indo-European.

In my proposed Oriental, the 6 languages come from 5 different language families.

The only languages that share a distant common ancestor are Persian and Hindi-Urdu.

So what could be the selection criteria for the most basic of words, such as are unlikely to have any common ground among the source languages?

I can suggest two possible solutions: 

1. The basic vocabulary is made of words that have cognates in Persian and Hindi-Urdu.

2. The basic vocabulary is made of words from Arabic.

In the first proposal, the common basic vocabulary of Persian and Hindi-Urdu is Proto-Indo-Aryan. For convenience of research, one could base it on Proto-Indo-European. That means Oriental is built as an Indo-European language with superstrates from waves of borrowing from Arabic, Sanskrit, Persian Turkish and Graeco-Roman.. 

In the second proposal, Oriental is an Arabic pidgin with large amounts of borrowings from Turkish, Persian, Sanskrit and Graeco-Roman.

The first proposal happens to be quite close to Olivier Simon's Sambahsa, just with a tighter focus on one region of the world. I guess the wheel has already been invented (by Proto-Indo-Europeans!)