For the past couple of months, I had been exploring the Mongolian language, learning about the rather interesting phonology system, the vowel harmony it sort of shares in common with some of the Turkic languages, and perhaps a bit of the writing system, Mongol bichig, which is probably one of the few writing systems today that write from top to bottom, then left to right (alongside scripts like Manchu and Oirat). There is the Cyrillic alphabet which entered major use since 1946, but there is just something unique with Mongol bichig that rings Mongolian.

But there was this curious part of Mongolian which captivated me. The writing system and some pronunciation bits aside, the concept of the “fleeting-n”, or more technically speaking, the “hidden-n declension”, was a grammatical rule which stumped me when learning about how case endings worked in Mongolian. Poking around coursebooks revealed few details, but there was this paper which I found that made me realise this rule has its origins rooted in history.

So, what is this “fleeting-n” I have been mentioning? Take ямаа (/jama/, goat) for example. An example of that word in the genitive is ямааны мах (/jamaniː max/, goat meat, or meat of goat). Looking back at the genitive declension patterns reveals something weird, however:

1. -н (n) is added to all words which end with a diphthong or ий (ii).
2. -ы (i) is added to back vowel words ending in -н (n).
3. -ий (ii) is added to front vowel words ending in н (n).
4. -ийн (iin) is added to front vowel words ending in short vowels or consonants (except those ending in н), and to back vowel words ending in ж, ч, ш, г, ь, и, and the short vowel will be dropped.
5. -ын (in) is added to all other back vowel words ending with short vowels or other consonants (except those ending in н).
6. -гийн (giin) is added to all front and back vowel word ending with long vowels.

We find that ямаа follows the second rule, but yet it does not reveal the -н until the root word is declined. This phenomenon is thus termed, the “hidden-n declension”. So what is going on here?

My initial guess was something in the dialect of Mongolian most learners are exposed to. After all, Mongolian is generally split into many different variants spoken throughout Mongolia and Inner Mongolia. Some digging around lead towards realising that Khalkha Mongolian is the version of Mongolian used in schools, and the dominant form of Mongolian. The first result I landed on came to be this paper:

It discussed the various words ending in the nasal consonants like /n/, which also mentioned how this fleeting-n appears before the case markers for the genitive, dative, and ablative cases, as well as a plural suffix, and other postpositions. It seems that even recently, no linguists have agreed on the functions of this fleeting-n (mentioned in the paper as an unstable n-stem). Some have suggested that it does not really serve any grammatical function, but as an empty morpheme. Others have suggested that this “n” plays some grammatical function which may be too complex for my understanding. However, while this paper discussed at length the significance and possible grammatical theories behind this rule, it did not quite explain why it happens.

Another avenue I tried to explore was the history of Mongolian. There have been several well-written texts in Mongolian throughout history, giving historians and linguists leads to piece together and reconstruct Mongolian as it was probably spoken in the centuries past. This might have generated a bit more leads, like:

Similar patterns may pop up in words for say, “partridge” and “crane”:

However, some words did not appear to show this sort of etymology, but yet follow this hidden-n declension. Consider words like that for “bull”, and even some loanwords like “video”:

There are some 200 words listed on Wiktionary which follow the Mongolian hidden-n declension, all from different noun classes (like animals, objects, concepts etc.), and from various sources like Russian, some Mongolic languages, Persian, and Mandarin Chinese. However, as far as digging goes, this is the deepest I can see, and understand. It seems that the pattern of hidden-n declining words may suggest a more Mongolian-centric pronunciation rule for some cases, but things past this point would just be speculation.

This has undoubtedly been an interesting dive into a specific grammatical rule in Mongolian, one which has lead me down a little road trip on the web, and in some academic papers. I hope this summarises some bits about the rule which some coursebooks do not really go at length with, and also I hope I have summarised my thought process when approaching some curious questions like these.