Okrand explains to us about null prefixes. I’m dimly beginning to suspect that there is also at least one null suffix.
My vague memory of an interview with Okrand more than a decade ago, includes times when I used the word “transitive” and Okrand corrected me to say “takes an object”. This happened repeatedly because I was stupid enough to fail to recognize my stupidity.
Also, in TKD, it talks about word order as “Object Verb Subject” and uses the word “Object” quite a bit, but I can’t find an instance of him using the term “Direct Object”.
I think I’ve been inserting the idea of a Direct Object into a language that doesn’t distinguish Direct Objects from other kinds of Objects. In other words, Klingon doesn’t have a Direct Object or an Indirect Object. It just has Objects. Some are supplemented in syntactical information by Type 5 suffixes, and some aren’t.
Hey hey! You got there! Woo hoo! But don't confuse the nouns with type 5 suffixes on them as "objects." They're not. Each Klingon clause has at most one object. Those other words are just other words. The ones with type 5 suffixes I tend to call /syntactic nouns./ A linguist might call them nouns in an oblique case, though that would require a clearer understanding of what /oblique/ means to a Klingon case system than I am prepared to figure out. There are also noun-based time expressions, which don't take any kind of marking. In Klingon, the object is its own syntactic role. Just as a noun like *DujDaq* has the syntactic role of /locative/ and *yuQvo'*//has the syntactic role of /ablative,/ the object has the syntactic role of "noun the subject performs the verb on." This is purely syntactic, not semantic: being an object doesn't impart any /meaning/ to the noun or the action. This is why we can say both *puq vIghojmoH*/I teach the child/ and *QeD vIghojmoH*/I teach science:/ being the "object" doesn't have any semantic meaning, but being the "recipient" *(puq)* or "theme" *(QeD)* or whatever does.
This would explain the “prefix trick” and the {-moH} sentences I have despised for so many years. In other words, a noun might optionally take {-vaD} and have the same meaning in a sentence whether the {-vaD} is present or not.
Yes! Of course, it depends on the verb. Different verbs expect certain semantic roles as their objects and only accept words that mean certain things. We have been told explicitly that the object of *jatlh* must be the speech-event and cannot be the addressee. You can say *SoHvaD jIjatlh,* but you cannot say *SoH qajatlh.* However, we /also/ have the prefix trick, which says that you can make a verb prefix agree with an unstated indirect object if it's first- or second-person. So you /can/ say *qajatlh*/I speak to you./ It's not that *SoH* is the object of *jatlh;* it's just that the prefix is allowed to agree with an elided indirect object /instead/ of the direct object.
What I’ve been taking as a specific Direct Object might instead more simply be an unspecified kind of object, and what I think of as a Direct Object is more simply an object that perhaps takes a null Type 5 suffix, or more simply, it’s an unspecified type of object that has no possible Type 5 suffix that could be applied to it in the given sentence.
You could think of objects (and subjects) as having null-suffixes, but they wouldn't be type 5 suffixes, because subjects and objects can have the suffix *-'e'* on them. And I don't see why Klingon would have these hypothetical subject and object suffixes anywhere but type 5. I don't know if this is a helpful thing to imagine. But the general idea is correct. Klingon inflects all its various noun syntactic roles /except/ subject and object.
So, in {be’nalwI’ vIghojmoH}, {tlhIngan Hol vIghojmoH}, and {be’nalwI’vaD tlhIngnan Hol vIghojmoH}, {be’nalwI’} is never a Direct Object. It’s just an object, and in the first sentence, the {-vaD} isn’t necessary and is unstated, and the verb prefix doesn’t indicate that it is the Direct Object of the verb. It simply agrees with its status as an Object, not explicitly identified by a Type 5 suffix. In the last sentence, the {-vaD} is present for clarity’s sake.
You're almost there! It's not that *-vaD* is added to the last sentence for clarity. You can't say *be'nalwI' tlhIngan Hol vIghojmoH.* A Klingon clause can have no more than one object. That seems to be an ironclad rule of Klingon. The rule appears to be that if more than one noun or noun phrase /could/ be the object, then whatever noun or noun phrase is in the role of recipient (i.e., indirect object) gets put into the beneficiary noun case, marked explicitly with *-vaD.* The other noun or noun phrase becomes the object and is not marked. And the two nouns or noun phrases must go in the correct order: beneficiary before object.
In other words, {-vaD} is akin to a plural suffix. If you already know the noun is plural, you don’t need a plural suffix. If you know that a noun is a beneficiary, you don’t need {-vaD}. It’s not wrong to explicitly use it, but it’s just not essential. It’s a language, not a computer program. Just because you CAN use a Type 5 suffix doesn’t mean you have to.
It's kinda-sorta like *-vaD* is optional, but only where we know that a verb can use a recipient as its object. Can you say *HoD vInob* to mean that you give the captain something unspecified? I don't think so, any more than I think you can say /I give the captain/ in English to mean you give something to the captain. In English, the word /give/ doesn't work that way. But the word /tell/ does: /I tell the captain; I tell the captain the information./ So it is in Klingon: each verb will potentially have a different relationship with its object than others do.
Similarly, that also explains why certain verbs don’t require {-Daq} for nouns that obviously have a locative function. Meanwhile, it makes sense that over time, verbs that commonly have locative-nouns as objects might more formally decide when to use the {-Daq} and when to not use the {-Daq} in order to more give more nuance into the nature of the use of the locative with that verb, so you can differentiate whether you are going to a river (destination) or going in a river (route, or vehicle), etc. This kind of nuance doesn’t work for other verbs because location isn’t a core part of the meaning of most verbs.
Remember, though, that it's specifically the verb that imparts a locative sense to its object. *Dab* is a locative-object verb; *nob* is not. I can say *yuQDaq vIDab */I inhabit the planet,/ and I can say *yuQ vIDab*/I inhabit the planet,/ and I can say *yuQDaq vInob,* but this can only mean /I give him/her/it/them on the planet,/ and I can say *yuQ vInob,* and this can only mean /I give the planet (to someone)./ It's different for every verb.
Of course this idea of more generic Objects also leaves open the following possibilities for which we have no canon, but until such canon appears, this is only unverified conjecture:
tlhIngan Hol’e’ be’nalwI’ vIghojmoH.
tlhIngan Hol’e’ be’nalwI’vaD jIghojmoH.
In that last example, the prefix indicates no object because the Type 5 suffixes already identify nouns that are objects, so there’s no need to redundantly point to them with the prefix.
Well, no. Type 5 suffixes don't identify objects. It appears to be impossible to mark a noun with a type 5 suffix, except for *-'e',* or if it's *-Daq *on a locative-sense verb.
Also, the “prefix trick” isn’t really even a trick. If I’d understood the optional nature of Type 5 noun suffixes, then the following would already make sense:
chab qanob.
The prefix indicates one object, {SoH}, and positionally, there’s another object {chab}. Context tells you that {SoH} is the beneficiary, so the {-vaD} is wholly unnecessary, and {chab} doesn’t need a Type 5 Suffix to explain its grammatical relationship to the verb.
The prefix trick simply lets you make the prefix agree with an elided indirect object, whether or not there is an object. If the indirect object is present, you cannot use the prefix trick. *-vaD* is not dropped because of context.
I’m guessing that SuStel has been trying to explain this to me for years, but I couldn’t begin to understand that which I had misunderstood because it was an unknown unknown, as a former member of the military industrial congressional complex once put it.
You are VERY close now. You have it right that Klingon doesn't have SYNTACTIC positions called direct object and indirect object. It has the syntactic position of "object." Actually, it does sort of have a syntactic case of "indirect object": the suffix *-vaD.* When the noun indicates the semantic notion of recipient ("indirect object"), it can be marked with *-vaD* to indicate this semantic role syntactically. *-vaD* doesn't only mean "indirect object," though; it means "beneficiary." (E.g., *Qu'vaD lI' De'vam.* *Qu'vaD* is not an indirect object. Indirect object is a subset of beneficiary.)
Or perhaps more likely, he’s about to explain how I have once again gotten it wrong, and someone will be entertained by the ensuing thread. In any case, it gives one distraction from The State Of The World at the moment, and anything that can do that is worthwhile.
I think you've gotten over the conceptual hurdle. Now it's just a matter of avoiding the stones along the track. -- SuStel http://trimboli.name