How many grammatical cases does Telugu have?

Question

I can't figure out how many grammatical cases Telugu has:

• Wikipedia says 8 (Telugu grammar)
• Telugu itself says 8, but I'm not sure if they map 1-1 to linguistic cases (విభక్తులు/viḅaktulu)
  • I found a blog post that maps them to the ones in the Wikipedia article, but I'm not sure if it's accurate (Telugu Nouns: Grammatical Case)
• Another source says 7 (lists a few different cases than the Wikipedia article), but I'm not sure if I'd count it as 8, since it does merge two using a hyphen, but those two cases do exist separately (Telugu)
• Another one (probably not very reliable) says 25, which seems a bit high to me, considering what the more reputable sources say (Telugu language)

Since all the others mostly match up in terms of count (7/8), concerning the 25, looking through it, it doesn't seem inaccurate. Telugu does have noun endings and inflection for all of those listed. It may have more endings than listed there, but I'm not sure, since I speak English better than I do Telugu.

So how do you count cases, and specifically, why not many of the ones in the last source? To me, 25 cases seem like too many, especially given the answers of more reliable sources, but I have no idea why it shouldn't be 25.

This is especially since Wikipedia does count these cases for other languages, e.g. Finnish (which has 10 cases in common with Telugu if we take the 25 to be valid), where the -ssa suffix is used for the inessive case (-లో/-lō in Telugu). Is there anything different that makes Finnish's inessive case (and other cases common between the two according to the last source) valid and not Telugu's?

Answer 1

For Telugu, due to agglutination, the line between postpositions and case-affixes is blurred to the point where it's useless to count how many cases there are. The 8 number comes from trying to map Telugu grammar onto Sanskrit grammar iirc.

Maybe a better way of analyzing case affixes is by taking all of the ones that don't transparently come from stand-alone words. For example, tō comes from tōḍu meaning companion. Using this analysis, nominative, genitive/oblique stem, -ni, -ki, and -na are the true cases. Also note that some nouns have a locative case which is made by changing the last vowel of the oblique stem to -a. For example, cēyi->cēta, bayalu->bayaṭa, pēru->pēra, etc. So counting these up, we get an answer of 5 (6 if you count the weird locative for some nouns and maybe 4 or even 3 in most cases because the nominative and genitive are usually the same and accusative is often dropped before inanimates). Coincidentally, these 5 are the least likely to be written by native speakers as seperate words, unlike most of the affixes in the 25 list that are usually written seperately from the noun.

The instances where the oblique stem is different from the genitive is for -am/-em nouns that change to -āni or -ēni before -ni and -ki, but for these nouns they stay as -am/-em before other case affixes like -tō or -lō. So maybe we can analyse the genetive and the oblique stem as the same but undergoes sandhi before certain affixes. I just wanted to clear this up.

Another case is with -na as in illuna. Honestly as a learner of Telugu I have no idea whether to put the oblique stem before -na or not. -na in spoken Telugu is most used to refer to the time of an event and for most of those words the nominative/genitive are the same so this problem doesn't come up. (Instead of "at the home" people tend to say "in the home" or "near the home" or "outside the home" or stuff like that.)Also a contracted form of -na may be the source of the weird -a locative but idk.

I also wanted to clear up that iḷḷu and iṇḍḷu are both correct plurals of illu and the difference is just dialectal variation. Generally, speakers who don't have a l/ḷ distinction will use iṇḍḷu.