Can a case system in a language help resolve gramatical ambiguites?

Question

One time in a linguistics class I sat in on, we were discussing ambiguous sentences such as "I killed the man with the spoon". In English, as written, it is unclear if the subject is using a spoon for the killing, or if the object being killed simply had a spoon in his possession. Learning Polish, I started to wonder if ambiguities like this could be resolved by varying declension of the noun "spoon". For example, using the genitive case to say that the object possessed the spoon, as opposed to the instrumental case to indicate the the subject used the spoon for the killing, i.e: "Zabiłem człowieka z łyżką." (łyżką -- spoon, inst.) v.s. "Zabiłem człowieka z łyżki." (łyżki -- spoon, gen.)

First off, is this the correct way to resolve this ambiguity in Polish, or even grammatically correct? Also, do languages with richer case systems have less of these kinds of ambiguities in general?

Answer 1

Being a native speaker of Czech, which is quite close to Polish, I think ambiguity can be avoided, or at least reduced, to some extent.

As for your Polish sentence, it would probably be wrong to use the genitive. In Czech, we would simply leave out the preposition, and I thought the same could be done in Polish (which I can speak a little, coming from a region close to the Czech-Polish border in the east of the Czech Republic), but I was wrong - it could not, as another answer below, apparently written by a native speaker of Polish, explains. Hence:

• "Zabiłem człowieka z łyżką." (unambiguously "a man holding/possessing a spoon")
• "Zabiłem człowieka łyżką." (unambiguously "a man killed with a spoon")

Thus, I tought that, in Polish just like in Czech, the instrumental can be used precisely as its name suggests, i.e. to denote instruments, but as the example above suggests, its use is rather comitative or even possessive.

Compare the following Czech parallels showing both the differences and similarities between Czech and Polish in this respect:

• "Zabil jsem člověka se lžící." (ambiguous, where grammatical)*
• "Zabil jsem člověka lžící." (unambiguous: "killed with a spoon")

*Note that the "prepositional instrumental" is not grammatical in all varieties of Czech, certainly not in Standard Czech. This means that with some, maybe most, varieties, no ambiguity ever arises, whereas with others, it may be avoidable. I suppose languages may vary quite a bit in this respect.

Answer 2

The answer is, the case system can help avoid ambiguity, but it is not a silver bullet.

From the functional standpoint, the example sentence is actually two distinct phrases:

1. killed (a man who has a spoon);
2. (killed using a spoon) a man;

Each of them convey a different message and therefore they can be represented with two distinct syntax trees. It is only a coincidence that they look identically in English.

Now, if you try to express these two phrases in an arbitrary language, you need to obey the language's Government and binding rules that would bind your subject, verb, and object into a phrase. These rules may or may not involve noun cases, particles, articles, bound morphemes, word order, or whatever language tools existing in the language.

For every language, sooner or later you'll find pairs of messages that produce two identical phrases. Noun cases can make your search a bit longer, no more.

Also, it worth mentioning that such ambiguity has produced an entire type of humor (sometimes, anti-humor), called Pun.

Attempt to find an unambiguous phrase structure grammars is one of the major reasons for inventing constructed languages. For example, Lojban has been deliberately designed for eliminating syntactical ambiguity. We can naturally assume that if there was a natural language with no ambiguity, Lojban would have been never created.

Answer 3

In Polish, ambiguity in that case is not existent.

I killed the man with a spoon (man with spoon) - Zabiłem człowieka z łyżką

I killed the man with a spoon (using a spoon) - Zabiłem człowieka łyżką

Both of the polish sentences are completely unambiguous. Instrument of an action is expressed always in the instrumental case, with no preposition at all. The preposition "z" means "with" and provokes instrumental but only syntactically.

There's one case when a prepositional phrase can be used to express an instrument but it occurs only in slang language.

That's from a Polish native speaker.

Answer 4

Some languages would have ambiguity between instrument and attributive possessum (and possibly also comitative `I went with a friend'), and some won't. This is not strictly dependent on case. For instance, the sentence you give will be ambiguous in German

(1) Er ermordete dem Mann mit der Gabel.
he.NOM killed DEF.ACC man with DEF.DAT fork
`He killed the man with the fork.'

but will be disambiguated in Russian, where the case is the same, and difference is in whether the preposition is present

(2) a. On ubil chelovek-a vilk-oj.
 `He [[killed a man] [with a fork].'
he.NOM killed man-ACC fork-INSTR
 b.  On ubil chelovek-a s vilk-oj.
he.NOM killed man-ACC with fork-INSTR
 `He [killed [a man with a fork]].'

As you can see from bracketing, the ambiguity is structural, so it will be present in any language where these particular adjuncts coincide in their coding, be it case, adpositions, or something else.

That said, I can't help with Polish, unfortunately.

Answer 5

The answer by Dr Jetušek is great but I would strengthen it a little bit: the declension indeed helps to fight the ambiguities and whenever they survive, which is a very small percentage of cases, it is almost deliberate.

Both Czech and Polish use declension with seven cases which were inherited from the Old Slavic system. The declension often plays the same role that require additional prepositions in other languages.

The seven cases, as ordered by the Czech rules, are

1. nominative: Who? What?
2. genitive: Of whom? Of what?
3. dative: To whom? To what?
4. accusative: [verb, object follows] whom? What?
5. vocative: We call, we shout.
6. locative: About whom? About what?
7. instrumental: With whom? With what?

I added some translations of the usual "questions" by which the Czech schoolkids actually learn them – almost no children are taught the Latin names.

In some cases, the translation was hard, e.g. in the case of accusative, because English doesn't have seven cases.

The vocative is exceptionally alive and well in Czech and perhaps Polish: to call a person, we have to use a different case. A big portion of the languages have had it but in most of them, it went extinct.

The genitive (2nd) and the accusative (4th) often have the same form of the word; the dative (3rd) and locative (6th) tend to be similar, too.

The cases help to distinguish lots of different possible meanings of the sentence. Some sentences with the word "wave at" ("mávat") and "fork":

1. Vidlička mává (A fork is waving, the fork is the subject.)
2. Mává bez vidličky (He is waving without the fork.)
3. Mává vidličce (He is waving to the fork – as if it were a human.)
4. Zvlnil vidličku (He made the fork wavy, he rippled the tool.)
   • (also: Mává skrz vidličku (He is waving through the fork.))
5. Mává: "Vidličko!" (He is waving and shouting: "Miss Fork!")
6. Mává o vidličce. (He is waving, as if waving were a language, and his story is about a fork.)
7. Mává [s] vidličkou. (He is waving with the fork.)

And others. You may see that the 3rd case automatically translates with an extra "to" while the 7th case almost always translates with "with" to English. So in these two cases, the role of the case for preserving the right meaning is essential, even in the absence of any prepositions.

As Dr Jetušek wrote, the 7th case does not need an explicit "with" (in Czech: "s" or "se" if the following object starts with "s" or a similar consonant). When "with" is explicitly written down, it sounds like the object possessed the spoon or fork; when "with" is omitted and one only uses the pure 7th case, it almost always means that the word in the 7th case was the tool to achieve the process given by the verb.

The 2nd case, genitive, often represents ownership when no preposition is added. For example, the following sentences represent different case (1-7) of the word "[Old] Hungary":

2. Král Uher daroval koně. (The king of Hungary donated a horse [to someone].)
3. Král Uhrám daroval koně. (The king, probably a different one, donated a horse to Hungary.)
4. Král Uhry daroval koňům. (The king donated Hungary to the horses. The 3rd and 4th cases are reverted relatively to the previous example but the order of the words may be kept. The declension is enough to distinguish what is the gift and what is the recipient. Languages with declension unavoidably have a much greater freedom in the ordering of the words.)
5. Král Uhrami daroval koně. (The king donated the horses, through or using Hungary in some way. The Czech sentence isn't quite comprehensible but it is arguably fine from the grammar viewpoint.)

The seven cases (of declension) are useful to change or preserve the meaning but so is the letter case. The word "uhry" in lowercase means "pimples" which is something else than "Hungary". ;-)

Answer 6

This isn't about a case system; this is about a pattern in many languages where in many cases it is unclear whether a phrase modifies the adverbial phrase it is in proximity to, or the master verb that said adverbial phrase itself modifies and this can appear analogously in many cases. For instance “the man with a spoon with a car”; it is ambiguous whether “with a car” modifies “the man” and it simply has two adnominal adpositional phrases that modify it, or whether it modifies “a spoon”.

In English, adpositional phrases such as “with a knife” can modify both nouns and verbs. One might indeed have a different case for each usage, but that doesn't solve the later issue I illustrated.

This is a very common problem in nested grammar; another related problem is that the extent of subordinate clauses is ambiguous, such as in “I fed the man that slept under the bridge.”, it is ambiguous whether “under the bridge” applies to the main clause, or to the embedded relative clause that modifies “the man”.

The only way to truly disambiguate this that I see is introduce parentheses of some form. I knew of a conlang that actually had a clitic that closed subordinate clauses.

Answer 7

Interestingly, "Zabiłem człowieka z łyżki." has yet another, meaning in Polish, colloquial though. It would mean "I killed man with opened hand". "To kill a man using a spoon" would be, as stated above, "Zabiłem człowieka łyżką." Answering Your question: Yes. I have the same feeling: some syntactic ambiguity is removed with declension.