Russian historian Alexandre Nemirovsky proposed a hypothesis that Classical Black Speech was based on extinct Hurrian language. This idea is controversial but it became popular among Black Speech researchers. Nemirovsky's analysis of Ring Inscription was made prior to publication of Parma Eldalamberon issue 17 (in 2002 or earlier, while PE 17 was out in 2007).

Nemirovsky provides following arguments (taken from his personal site, since 2007, translated and commented by Un4givenOrc):

1. There is a case system exists; cases are expressed with postpositions, i.e. not with regular morphemes but particles with case meaning (this is emphasized with hyphen in romanization). Example: “burzum-ishi” where -ishi is locative marker. Compare with “Saruman-glob” (“about Saruman”? - it seems Nemirovsky was not familiar then with the History Of Middle-Earth vol. 12 which had one of translations of this phrase, nor with journal Vinyar Tengwar which had another one.) said in conversation of orcs that captured Hobbits after Boromir's death.
2. Comparison of Ring Inscription's lines 1 & 3 with lines 2 & 4 shows that format “-ûk” complies with “all”. As normalized transliteration shows off that “-ûk” is not a separate word, and not even a clitic particle, but just a morpheme, then it should be counted as verb's perfective formant of completeness (meaning “to do completely”). Surprisingly such rendition was confirmed in PE17: “ulûk, verbal ending <…> 3rd person pl. “them” (ul) in completive or total form “them-all”.
3. Correspondence of 4-times repeated formant “ul” to 4-time repeated pronoun “them” proves their identity. Therefore, -ul-, is a marker of transitive action, again used as formant adjusted into verb's form not as separate word. It means that Black Speech can be defined as a language with ergative alignment (i.e. language in which both subject and object are expressed with special markers inside verbal form). 1) “durbatulûk” and other verbs in RI do not have markers of subject. 2) it is not definition of ergative language. Ergativity means that subject of transitive verb is marked differently than subject of intransitive verb. Verbs must differ only with transitive marker while suffixal agreement with subject and object is optional, they may be expressed with separate words. 3) From PE17: ”ulûk, verbal ending expressing objects (particles indicating 'subject' were usually prefixed)“, this show very unusual structure and word order for ergative languages.
4. Rendering of verbal forms with infinitive seems reflect personless character of Mordor's form (sic!), i.e. absence of verb's subject. One the other hand, some marker must express modality of verbs somehow. Considering that -at- is the only repeated marker left, it should be treated as a formant expressing modality (in wide sense). Once again, while explanation is not very clear, Nemirovsky was closer to Tolkien's own definition which was unpublished then, while others treated -at as simple equivalent of English infinitive, subject marker, or just generous verb's derivational suffix. PE17: ”at, verb ending (like a participle) (durbat = constraining, of a sort to constrain)“ - thus -at is participle form absent in English that may express intention. See also remark to point 3 about markers of subject's grammatical person.
5. As we can see, verbal forms in Black Speech are built with agglutination, i.e. by joining a chain of markers to the root always in the same order, with each of them carrying defined grammatical meaning. In sampled forms, modal formant is first after the root, personal formant of verb's object is the second, formant of aspect is the 3rd. Absence of subject's marker doesn't show that 3rd person was zero-marked (otherwise it would be translated with 3rd person English verb), and means that probably, subject's grammatical person was not expressed at all at least in some modalities - as you can see from Tolkien's definition of -at, person is not marked because it is modifier participle with modal meaning.
6. Mordorish word “ring” do not have postpositions, in other words it has Absolutive (Nominative) case and is the subject of sentence, as in translation. As you can see in this example, absolutive case in Mordorish is used for logical subject of action. Logical object of action isn't expressed as stand-alone word in our example, but it may be explained by it's being a pronoun.

Mordor's bilingual inscription allows us to find meaning of all words and morphemes used there, and define Black Speech as ergative (not this - see remark for point 3) agglutinative language which is probably started to shift into active (Nominative-Accusative) alignment. Comparison with some personal names allows us to define few more roots and morphemes: Uglûk probably means ”(do something)-completely/all“, whence “ugl-” is active verb; comparison of “Lugburz” (Dark Tower) with “burzum” (darkness) allows us to define meaning of ”-um“ as formant of abstract substation, and also word order of descriptive construction (modified noun “lug” = “tower” before modifier “burz” = “dark”). We can't find any specific gender endings, therefore Black Speech highly probable do not have grammatical gender. Nominal and verbal roots looks similar and are constructed by the same scheme CVC(C). Of course all this cannot be called a substantial result.

However accidental comparison of Black Speech and Hurrian language of real-world Hurro-Urartian subgroup of Nakh-Daghestanian branch of East Caucasian sub-family gave surprising results (even criticized Altaicists now consider Hurro-Urartian as distinct language family, Nakh-Daghestanian is usually called Northeast-Caucasian language family). There were discovered not only structural but also material similarities between Black Speech and Hurrian language (spoken in 3rd - 1st millennium BC by ancestors of modern Armenians and Kurds, who changed the language to Indo-European): (very controversial claim, as there were also Indo-European Hittites and language-isolate Hatti in nearby territories around the same time, the connection to later is also controversial though)

1. Hurrian language is exactly agglutinative ergative language with rich case system expressed by postpositions. Modal verbs in Hurrian do not agree with subject in person, but may include object marker. Hurrian do not have grammatical gender (even some of Indo-European languages do not distinct gender in nouns, not mentioning other language families).
2. Nominal and verbal roots look the same and are composed on similar [phonological/phonotactical] model in Black Speech and Hurrian.
3. Below is comparison chart of known Black Speech words and morphemes with known Hurro-Urartian correlations:

1. Percentage of similarities is surprisingly high for such small corpus
2. Unlike other theories concerning Black Speech, pointed out similarities look legal

1. We don't know for sure if J.R.R. Tolkien was familiar with E.A. Speiser's “Introduction to Hurrian” (1941), the first dedicated book on the subject.
2. There are too little attested words both in Black Speech and Hurrian/Urartian to make any conclusions
3. There is no trace of ergative alignment in known Black Speech or Orcish corpus.
4. Nemirovsky cites some words not attested in other sources
5. Half of words are actually Urartian, a related language of the same language family

• Original Russian text of hypothesis (Archive)
• Orkish and the Black Speech at Ardalambion. See “APPENDIX: Was the Black Speech based on Hittite/Hurrian?”.