“It’s Hungarian.” It is probably the one answer you would come across the most, online and in person. After all, being part of the Uralic languages, members of this family are generally known for their extensive grammatical case system. Estonian has 14 or 15, Finnish has 15, and Hungarian has 18 of these. But there is something more to that. Depending on if you include other case-like suffixes, relaxing the grammatical case definition a bit, Hungarian could have as many as 34 of these cases.

However, even within this language family, the most well-known languages here do not really have the most number of grammatical cases. The Veps language, another member of the Uralic language family, has 23 or 24 grammatical cases, with some grammatical cases not found in any other counterpart. And this does not include case-like suffixes that help to bump the count for Hungarian up to 34.

You might want to take this count with a pinch of salt, though. This is taken from an uncited segment in the English Wikipedia article for Veps, and Riho Grünthal’s Vepsän kielioppi (2016), in Finnish, seems to suggest that Veps has 18 grammatical cases, or at least, that is as far as my understanding of the Finnish language gets me. There could be more that I am missing out, so that goes into further reading for my Finnish-speaking readers. Looking up the Russian version, however, it seems that this figure of 23 or 24 cases was derived from I.V. Brodskiy’s publication in 2014, or Zajtseva’s publication in 1981, both of which are entirely in Russian. So, there is some debate on what is, and what is not a grammatical case here.

But what if I told you that there are some languages purported to have even more grammatical cases?

If you look this question up on the Guinness Book of World Records today, you would find nothing. But if you did it in 1997, you would find a curious language called Tabassaran, a Northeast Caucasian language spoken primarily in Southern Dagestan in Russia. Reading on, you would see that Tabassaran “uses the most noun cases, 48”. Further literature searches would lead you to the Hjelmslev paper on the Tabassaran case system, published in 1935. This publication gives a different figure — 52, or if you exclude the two that only apply to adjectives, 50. It is still a lot, but it has drawn some ire to some linguists, who critiqued that these claims were based on “sloppy analysis of ‘case'”.

This brings us to the paper published by Comrie and Polinsky in 1998, called “The great Daghestanian case hoax”. This publication aimed to debunk this claim of Tabassaran having the most number of grammatical cases, by applying the case analysis used for that language onto Tsez, a language they were studying at the time, and showing how many more grammatical cases Tsez would have.

Applying the same method to Tsez, they found that Tsez would have a total of 126 grammatical ‘cases’, or more appropriately speaking, ‘case combinations’. Broadly speaking, a grammatical case is a category of nouns, or noun modifiers, that correspond to at least one grammatical function for a nominal group in a wording. It is a system of marking nouns to give detail to the relationship they have in a phrase or clause. Non-local, or core cases, largely pertain to the more commonly encountered cases we see in many languages. These can include nominative, accusative, absolutive, ergative, genitive, and instrumental cases. Local cases, on the other hand, largely pertain to the noun’s spatial orientation in relation to other elements in a phrase or clause.

For languages like Tabassaran and Tsez, or languages with generally more complex grammatical case systems, most of these cases pertain to local cases. Tabassaran operates on 4 non-local cases, namely the absolutive, ergative, genitive, and to some extent, the dative. Tsez has 6 non-local cases, the absolutive, ergative, instrumental, dative, genitive 1, and genitive 2. The authors of the paper also noted the possibility of the 2 equative cases to be non-local as well, with the equative-1 being able to be used in combination with the 6 previously defined non-local cases.

It is the local cases that brings the most controversy on how these cases are enumerated and counted. In Tabassaran, a total of 7 or 8 local cases denote a noun’s spatial orientation, “in [hollow space]”, “on (horizontal)”, “behind”, “under”, “at”, “near, or in front of”, “among”, and “on (vertical)”. These cases, on their own, are described as “in absence of motion”. Basically, how a noun is related to, say, another noun, is described assuming that both are stationary. To give further detail to motion, 3 more cases are added to the mix — “essive (no motion, not moving)”, “ablative (motion from a place)”, and “allative (motion to a place)”. The dative, thought to be a core case, could also be used here to denote the translative, which describes the change in state of the noun it is attached to.

If these cases are counted separately, Tabassaran would have a total of 14 or 15 grammatical cases depending on the dialect.

But if you include some combinatorics, largely due to how many different combinations of local cases one can theoretically form, one would yield a number of 7x3x2, or 8x3x2 local cases alone. That is, 42 or 48 local cases. Adding in the 4 non-local cases, and the translative use of the dative case, gives us a total of 47 or 53 total cases. So it is this combination of inflectional forms that gives a total that is closer to the touted number in these previously published literature.

The authors then applied the same combinatorics principle of counting cases to Tsez. In Tsez, there are 7 ways to distinguish a noun’s spatial orientation, similar to that in Tabassaran but without “behind”, and 4 ways to denote direction. This includes the aforementioned essive, ablative, and allative, and also includes the lative case, that is, a more general form of the allative, but with different contexts for usage. Here, the allative is subsumed under the dative case, giving us one fewer unique case suffix. This is combined with the distinction of distality, if it is near or far from the speaker. These suffixes can theoretically be combined with the equative-1 case mentioned earlier, which denotes a comparison of one noun to another.

This gives us 7x4x2x2 = 112 local case combinations. When added to the 8+6 non-local case combinations, this gives us a total of 126 case combinations.

However, when taking case suffixes alone, without combination, we get a much more modest number: 8 non-local cases, plus 7 spatial orientation local cases, 2 directional local cases (subsuming the allative under the dative), and 1 distal case, to give us a net total of 18 cases, which is, the same as Hungarian.

Thus, it seems that these counts of grammatical cases are inflated primarily due to the factoring of case combinations. Learners of Tsez or Tabassaran do not necessarily learn each of these combinations by heart; they are more likely to learn each suffix and the principles or rules of combining these other suffixes.

But it seems that people do not really learn from history. The Wikipedia article for “Grammatical case” claims that Tsez has 64 grammatical cases, and does not cite any literature making that claim. Checking up on the Wikipedia article for the Tsez language gives us a total of 39, with, once again, a lack of cited sources. If we were to take the case analysis from Comrie and Polinsky paper as a reference, this number is pretty different, as you have seen earlier.

Nevertheless, this dive into literature shows us the rather conflicting nature of how grammatical case is defined, and how different sources could provide different figures for grammatical cases, depending on the methodologies used to assess this.

So, which language has the most grammatical cases?

It’s probably Hungarian, but also probably Veps, and depending on how you define grammatical case, you might say Tsez.

Further Reading

Debunking paper

Comrie B, Polinsky M. The Great Dagestanian Case Hoax. In: Case, Typology, and Grammar. Amsterdam: John Benjamins ; 1998. pp. 95-114.

Veps language, Finnish

Grünthal, Riho: Vepsän kielioppi. Helsinki: Suomalais-Ugrilainen Seura, 2016.

Veps language, Russian (you might need the Wayback Machine for this)

Бродский И. В. Первая часть // Vepsän kel’: Openduzkirj täuz’kaznuzile. — СПб., 2014. — С. 8—9.

Зайцева М. И. Грамматика вепсского языка. — Л.: Наука, 1981. — С. 186—187.