Consider Deuteronomy 28:30, in the Vulgata:

Uxorem accipias, et alius dormiat cum ea. Domum ædifices, et non habites in ea. Plantes vineam, et non vindemies eam.

So uxorem, domum, vineam, and eam are all in accusative. This is, they decline pro/nouns acting as direct objects to a transitive verb. However, because the verb is already declined (and thus indicating the person/subject "acting" the verb), further declining the direct object seems to me redundant.

For instance, if we use the nominative instead, we would get:

Uxor accipias, et alius dormiat cum ea. Domus ædifices, et non habites in ea. Plantes vineus, et non vindemies ea.

Would experienced Latin users (so not me) see an ambiguity in the above phrase? Accipias is clearly second person singular, so it should be clear that "you" are performing the verb, and not a wife. Clearly the house is not building something (what would it be building? as far as I know, to build is not an intransitive verb). Thus, it is "you" who is building houses. And etc.

My question arises because, even as a total beginner in Latin (so I can be wrong), the verb accipias already tells me who is performing the action, so its clear its not the wife. Similarly, it is evident that a house cannot be the subject of the verb to build. Thus, in a sense, I find redundant that I need to learn another grammatical case for pro/nouns. Even if I would be et non vindemies ea I think I would have understand the meaning. This is, the nominative might do a pretty good job replacing the accusative.

So, to rephrase my question, what potential ambiguities is the accusative mode clearing out, so that to make it necessary? I know this mode is inherited from Greek, so maybe is there something peculiar to Greek that required such mode?

  • 1
    I am not fluent in Latin, but I think there is some ambiguity in uxor accipias because uxor is also the vocative case. It could mean either "you take a wife" or "you take, wife" (I'm not sure whether that word order would be allowed, though)
    – b a
    Aug 3, 2018 at 10:40
  • 2
    Uxorem accipias 'you should take a wife' vs. uxor accipias 'you who are a wife should take [something]'.
    – Anonym
    Aug 20, 2018 at 19:06

3 Answers 3


Languages are full of redundancy. So I think the premise of this question—that the accusative case is "needed"—is problematic. For example, there isn't a need for English speakers to use the plural in contexts like "three plates": the word "three" already tells the listener how many plates there are. But the use of the plural in this context is nevertheless obligatory in English.

The existence of the accusative/nominative case distinction in Latin certainly allows ambiguity to be avoided in certain situations—for example, when both the subject and direct object in a sentence have the same number, person and (for periphrastic conjugations) gender, the conjugation of the verb doesn't tell you which is which—but as you've said, the grammatical role of a noun can often be inferred without knowing whether it is nominative or accusative. In fact, the nominative and accusative cases are always identical for all neuter nouns in both numbers, and sometimes for some plural non-neuter nouns. The distinction between accusative and nominative case was also lost in many of the languages descended from Latin.

The case distinction in Latin between nominative and accusative was inherited from the common ancestor of the Indo-European languages. We don't have any direct evidence of this common ancestor; rather, we can infer some of its characteristics based on our knowledge of its descendants. Greek is another Indo-European language, but Greek is not itself the ancestor of Latin; it is like a "cousin".

  • 2
    Thanks! The "three plates" example (in Spanish, tres platos, as opposed to tres plato) is very revealing. Indeed, languages are full of redundancies! An interesting question which surely Linguistic theory has dealt with, i.e. why do they arise, and to what extent are they "needed". Maybe something about "efficiency" or optimal degree of uncertainty...
    – luchonacho
    Aug 3, 2018 at 11:04
  • 3
    @luchonacho, on that broader question a starting point might be linguistics.stackexchange.com/q/25711/2394 Aug 3, 2018 at 12:39
  • @PeterTaylor Thanks. Quite interesting discussion.
    – luchonacho
    Aug 3, 2018 at 13:23

Sumelic has already done a great job explaining the "necessity" of certain grammatical constructions. What I would like to address is the changes you have made to those sentences. Although the subject of the sentences remain clear ("you"), changing the grammatical case of the direct object to also be that of the subject changes the meaning of the sentence. For example,

Uxorem accipias = May you accept a wife


Uxor accipias = May you, a wife, accept

In this way, the subject has become a wife to whom the speaker is speaking. It does not really matter that accipio is a transitive verb. Transitive verbs often do not take direct objects, just like in English sometimes. A native Latin speaker would most likely not treat uxor as a direct object upon a first reading or even a third or fourth reading because to them the declined form of the word was inherently tied to its part of speech.

Consider, also, when one could abandon the accusative form. Could you only abandon it when the subject is implied by the verb? Could you abandon it in all cases? This would quickly create problems if nothing else about the language changed. For example,

Puer puellam amat = The boy loves the girl

This sentence is completely clear about the meaning. However,

Puer puella amat

is unclear as to who is loving and who is being loved. We would have to change something about the grammatical rules of Latin to account for this. Perhaps word order would become more rigid. Perhaps new prepositions would have to be added. This is exactly what happened to Latin actually. Over time, the cases began to merge with one another, making it difficult to tell what grammatical use each word had. To assist with this, word order became more and more rigid and new prepositions were added or old ones gained new meaning (see de, which came to mean "of" in Spanish via vulgar Latin).

What you are describing was a process through which Latin actually went. As to why the language started off the way it did requires knowledge of the proto languages preceding Latin and the languages preceding those, as Sumelic mentioned. I hope this helps clarify why exactly, using Latin grammar, one could not simply change the case of the direct object.


Latin is not a designed but an evolved language. It's in the Indo-European language family and some of the properties of its ancestors can be deduced by comparing with other languages in that family that have left written traces such as Greek (which actually falls apart into dozens of different island languages few of which had written records; but some, like many works of Homer, have been preserved with an oral tradition until they came in contact with writing systems).

One feature of the language family is that sentence structure is evolved from a complex case system that conveyed a whole lot more information via a case system and nominal phrases than later on: it's not unusual for an antique Greek sentence to run over a single page in a loose connection of nominal phrases and subordinate clauses and not even contain a top-level verb.

Also, in archaic language variants, prepositions tend to be sort-of optional and are just used as a reinforcement of information already conveyed by the case, with cases including more obscure ones like locative case. As they became non-optional, the number of different cases in active use was reduced.

That process happened to different degrees in different members of the language family.

Similarly, English has a much more rigid sentence structure than other Germanic languages (though most of the modern ones usually call for complete sentences centered around a single verb) and thus was able to rationalize a lot of its case system away (and people have become unable to recognize when "I too" instead of "me too" would be necessary and mess up "thou" and "thee" when trying to emulate Early Modern English in order to sound grave).

But the end result is the same: when learning an existing language, the rules of how the language works are not yours to make. There have been a few synthetic languages (like Esperanto) that were actually designed to be logical within the confines of being appealing to speakers of existing languages, and you might have more chances arguing with proponents of those languages about what may or may not be a good fit (it's just quite likely that you'll tell them nothing new and your arguments have been taken into consideration long long ago without having the effect you desire).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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