I've noticed that in some (all? most?) languages, ordinal for 1 and 2 are completely different (i.e., not derived) from corresponding cardinals:

English One/Two/Three vs First/Second/Third is a bad example, because "second" is a loanword; and in other German languages, only first one is different: Dutch: Een -> eerst, twee - tweede

Russian: odin, dva, tri -> pervyi, vtoroy, tretiy

Estonian: üks, kaks, kolm -> esimene, teine, kolmas

Has any research been done into it? I wonder what this condition could imply about proto-languages.

  • 4
    Not in Chinese (including most of its major varieties), nor in either of the Japanese ordinal systems. Korean has just one exception: the lexeme "1st" in its native ordinal system (첫 번째, not *한 번째); its Sino-Korean ordinal system is completely regular. Vietnamese has thứ nhất rather than expected *thứ một - suppletion of native system by the Sino-Viet system for "1st", and thứ tư instead of *thứ bốn for "4th". – Michaelyus Jan 10 '19 at 9:50

This question was indeed extensively studied in linguistic typology and a high level summary of results can be found in WALS chapter 53. It gives a good overview of possible systems of ordinal numbers (no ordinals at all, ordinals are the same as cardinals, ordinals are completely regular, special word for "first" only, special words for "first" and "second").

Languages with an independent word for "second" are concentrated in Europe and rare in other parts of the world.

| improve this answer | |
  • I am so happy to learn about this resource!! Thank you very much! – 62mkv Jan 11 '19 at 18:10

In Germanic language the word “other” (or a cognate) has often been used as an alternative for “second”, also in Dutch as can still be seen in the relic word “anderhalf” (for one and a half): there used to be a ordinal based system of “half cardinals” where “derdehalf = two and a half (literally the third half: 1/2 is just half, 1 1/2 the second half, 2 1/2 the third half etc.) The system is still used in West Frisian (Frysk) e.g. “fjirdel” is just 3 1/2 (from “fjirde heal” = fourth half) and 1 1/2 is “oardel” also containing “oar” = other = second. Also Danish uses “anden” (other) as second and I believe it’s more common in Scandinavian languages. As you know in Dutch and in English also “third” is not regular, but does derive from “three” (“derde” in Dutch is a fossilised form of “drie + de” in a way). I think the low ordinals are used far more frequently than the higher ones and as such it’s well-known that higher frequency items are more prone to keeping irregularities that historically arose, because of easier learnability, in a way. So this way, the irregular lower ordinals have persisted.

| improve this answer | |
  • »the word “other” (or a cognate) has often been used as an alternative for “second”,« well, second is a Romance word. I have read that other is the Germanic word for second, whilst tweede, zweit- in Dutch and German are new formations. »“derde” in Dutch is a fossilised form of “drie + de” in a way« Why fossilised? It looks like a metathesis to me but I don't see how it is fossilised. – tobiornottobi Jan 10 '19 at 20:36
  • @tobiornottobi because it hasn’t been replaced by the regularised “driede” ( which my kids do say sometimes) but because of its relative frequency the metathesised form “derde “ has been kept. – Henno Brandsma Jan 10 '19 at 20:38
  • Okay, I'll have to think about your argument. – tobiornottobi Jan 10 '19 at 21:18
  • Cp. Ger. viertel "quarter" to "fjirdel", appearing regular among x -tel*=1/x. Very interesting, thanks! Also for the interested reader, we say about wallclock time that it's either *drei-viertel "three quarts" or viertel vor "quarter before/to" in an east-west isogloss. ander may compare to anti, hinter "behind", in weird ways, and we even say gegen 6, "to/towards/around 6" (viz anti, against, also gen Italien "to Italy"). prim- also compares to pro. – vectory Mar 26 '19 at 19:14
  • anderthalb is synonymous with ein-ein-halb (one-an-half) and their anlaut and ablaut are close enoug to latently lexicalize together. I hadn't figured out why it does not match though, and in fact I cannot accept "other half" quite yet. I wonder what -dert could mean and my closest match would be dreh- "turn", thus giving "a turn and a half". Since it's three halfs, the root of "tri" can fit, too. After all, another person (ein anderer) is the third person, and not strictly the one behind, though either fits 3rd-p.-perspective. anders "different"nagain fits "changed, turned". – vectory Mar 26 '19 at 19:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.