"A new stage is reached in Nostratic theory in the work of Illich-Svitych, who is generally regarded as the founder of Nostratic in its modern version. His earliest comprehensive statement was published in 1967 in the form of a series of etymologies from the six families usually cited as 'classical Nostratic.' However, it is interesting to note that in a slightly earlier publication (1964), significantly called 'Oldest Indo-European-Semitic Linguistic Contacts,' he considered that the case for a relationship between Semitic and Indo-European was weak and that most of the resemblances were due to borrowing from Semitic by Indo-European."You have to understand that Greenberg, the proponent of mass comparison to the chagrin of the rest of academia, was trying like most Nostraticists and related enthusiasts to promote his grand vision of an Ice Age protolanguage while tipping his hat to whom he exaggerates as 'the founder of Nostratic in its modern version'. The modern version as it was in 1950, more like. In fact, Illich-Svitych's reconstructions suffer from the same flamboyant problems that Starostin's did with an over-emphasis on parentheses and an under-emphasis on rational, regular sound correspondences. The kind of things that a trained eye immediately recognizes as poorly worked out and farfetched without needing to waste one's time examining it in closer detail.
But Illich-Svitych's reconstructions aren't what amuse me in the above quote. I find it oddly interesting how Illich-Svitych went from the more conservative idea that any similarities between Indo-European and Semitic are due to borrowing (as the case of 'seven' conclusively shows) to an unlimited self-indulgent anything-goes view that would give him carte blanche to hallucinate linguistic unicorns. What went wrong? How did he go astray like so many other misguided souls that fumble about with their painfully doomed Indo-Semitic comparisons?
Much like the more recent long-ranger named Sergei Starostin with his obsession with diacritics, Illich-Svitych had a peculiar, romantic soft spot for ejectives which he marked with an underdot in his transciptions. Markedness was violated everywhere according to his lavish brand of Proto-Nostratic and even some of the most common words in a normal human language such as pronouns and particles were replete with these articulatorily taxing ejectives. To add further irony, Allan Bomhard had recently offered an excellent alternative to Illich-Svitych's ejective-rich theory to finally address markedness problems by turning ejective stops into plain stops. Finally, real linguistics at work. Yet, since no one's perfect, Bomhard too committed the same mistake as any other linguist obsessed with the big picture over meticulous details by failing to recognize the undeniable layer of Semitic loanwords in PIE. This is why I still think that the many instances in his lexical comparanda that suspiciously fail to show anything but PIE and Afro-Asiatic evidence may be, if not complete red herrings, mere post-Nostratic loanwords.
So I wonder: Is it possible at all for Nostraticists to move forward instead of running around in circles by finally recognizing the obvious? Can we all not see that PIE and Pre-IE was surely in contact with neighbouring languages throughout its prehistoric development (just like any normal language ancient or modern)? And if we can all reason that far, then can't we start taking the time to weed out those later post-Nostratic contacts first before we publish mere conjectures about a remote paleolithic language?
 Ilya Yakubovich from the Department of Linguistics in the University of California in Berkeley writes in her online article The Nostratic linguistic macrofamily: "According to Moscow scholars, he frequently confuses cognate words with later borrowings (e.g. by trying to reconstruct Nostratic numerals)."