tlhIngan-Hol Archive: Fri Jun 21 13:50:38 2002

Back to archive top level

To this year's listing

[Date Prev][Date Next][Thread Prev][Thread Next]

Re: Headers. Yet again.

Because my interest in this argument is definitely waning, I wanted to make 
this response as mercifully brief as possible, but while looking at the details 
of ghunchu'wI''s argument, I stumbled over a beach analogy that will likely 
prove amusing to most and perhaps clarifying to some.

It's certainly my last gasp at trying to explain my perspective. I'm sure 
that's delightful news to the masses, since I do realize that since I'm the 
ONLY person on this side of the argument, when I quit, this thread might 
actually die, like it should have two years ago.
..
> >> 2) A noun with a Type 5 suffix (with exceptions) comes before the object.
> >
> >This is a descriptive rule about the grammar which happens to be true whether
> >it is in TKD or not...

> Actually, I'm starting to think that this observation doesn't really
> deserve to be called a rule.  The exceptions to it make it a lot less
> useful than one might think. There's nothing wrong as such about
> descriptive "rules" (except maybe the name "rule" itself).  But this
> particular one has one big exception ({-'e'}), a smaller one (verbs of
> motion), and at least one more little exception that nobody has yet
> mentioned in this thread: ({naDev/pa'/Dat}). 

Excuse me, but that isn't an exception at all. Like the time stamp nouns I did 
mention earlier, it is a "header" noun (your term, not mine) that does not take 
a Type 5 noun suffix. Please read the rule again.

I'm not saying that all header nouns take a Type 5 suffix. I'm saying that 
except for {-'e'} and the use of {-Daq} and {-vo'} with exceptionally explained 
verbs of motion, [that's two, count them, two exceptions] if a noun has a Type 
5 suffix, then it is a "header" noun and has to preceed OVS.

See the difference?

> The existence of several
> different *kinds* of exceptions weakens this "rule" significantly.

I would tend to agree with you, if you actually came up with a third exception, 
but you didn't.

..
> >From that, it becomes pretty obvious that the role of subject and object are
> >defined positionally, time stamps are defined by a combination of position 
> >and
> >meaning, and all other nouns have their syntax indicated by Type 5 noun
> >suffixes.
> 
> Obvious, yes (with exceptions), but it's neither universally applicable nor
> particularly predictive.  And it has little to do with "header nouns".

Again, you are confusing what I originally stated were two different rules 
relating to what you like to call header nouns. One says nouns that aren't 
subjects and objects are header nouns, and most of those have Type 5 suffixes. 
The two exceptions that forced Okrand to use the word "usually" are nouns 
acting as time stamps, and our special, three unmarked locative nouns you refer 
to.

The second rule looks at this from the other perspective. It also has two 
exceptions. To phrase it more closely to Okrand's word choice, I could say that 
nouns with Type 5 suffixes "usually" are what YOU call "header" nouns. There 
are only two known exceptions to this rule as well. One is {-'e'}. The other is 
the use of {-Daq} and {-vo'} on the direct objects of a select group of verbs 
of motion.

> >There are two exceptions to this...
> 
> I count three, actually. 

I suggest that you are miscounting. The rule you toute has two exceptions. The 
rule I toute has two exceptions. Okrand is vague about his two exceptions, 
using the word "usually" to cover for it. I am precise about the two exceptions 
that I observe.

If I'm being inaccurate about this, please let me know. I'm seeking accuracy 
here.

> And we have precedent for unexpected revelations
> of additional exceptions.

No problem. We change the rules whenever we observe revellations. That's what 
descriptive rules are all about. Otherwise, they'd be prescriptive rules, and 
the Klingon language doesn't HAVE any prescriptive rules. Remember the 
discovery about word order when verbs get {-jaj} tagged on to them? That didn't 
make us toss out the OVS rule. We just marked an exception and moved on.

> >This is a descriptive rule of the grammar, whether it is explicit in TKD or
> >not. Do you really want to argue to say that it is not accurate? Are you
> >really
> >so afraid to wander beyond the explicit wording in TKD, even when Okrand 
tells
> >us that the rules as stated in TKD are less important than usage in canon?
> 
> I'm perfectly happy to derive things from observation, as long as they
> simplify things.  This one *sounds* simple, but it's got to be augmented
> with several exceptions in order to work, and that makes it at least as
> complicated as the *real* rules it ostensibly seeks to combine. 

Nope. It's not seeking to "combine" the rule.

One one hand, you have Okrand's rule. He made one valid statement. Nouns that 
are not subjects or objects always come before the OVS part of the sentence.

He could have left it at that, but he didn't. He tagged on the somewhat vague 
statement that these nouns "usually" have Type 5 suffixes. If he'd wanted to be 
more explicit, he could have pointed out the only two exceptions we've ever 
seen:

Time stamp nouns
Three unmarked locative nouns (naDev, Dat, pa').

So, there are only two observed exceptions to the general rule that 
all "header" nouns have Type 5 noun suffixes. What about looking at it from the 
other side? If (with two exceptions) all "header" nouns have Type 5 suffixes, 
then what can we say about all nouns with Type 5 suffixes?

Well, from observing them, we can tell that, with two exceptions (that's {-'e'} 
and the use of {-vo'} and {-Daq} on the direct object of certain verbs of 
motion), all nouns with Type 5 noun suffixes are "header" nouns.

If you are so afraid of spelling out the observed exceptions, I could be vague 
about it and just extend Okrand's vague statement thusly:

"Any noun in the sentence indicating something other than subject or object 
comes first, before the object noun. Such nouns usually end in a Type 5 noun 
suffix (section 3.3.5)." [my addition:] "Also, nouns with Type 5 noun suffixes 
usually appear before the object noun."

That's really all I'm doing. I'm looking at the second statement Okrand made 
and looking for the exceptions that made him use the word "usually", and then 
I'm additionally looking at it from the opposite perspective. "Header" nouns 
usually end in Type 5 suffixes. He said that. Nouns ending in Type 5 suffixes 
usually are "header" nouns. My observation has exactly as many exceptions as 
his does. I'm more explicit about the exceptions, and that apparently disturbs 
you.

Like our rules about word order survived modification when Okrand explained {-
jaj}, these rules about "header" nouns can survive future modification as canon 
provides us with more exceptions. That's how descriptive rules work. Deal with 
it.

There's nothing wrong with us coming up with new descriptive rules about the 
grammar so long as they are strikingly consistent with canon. Okrand keeps them 
vague so he can change them later. We can be a bit more specific if we want to, 
so long as we are willing to deal with it when canon shows our reasonable 
assumptions to be inaccurate.

My assumptions here are reasonable.

> It is less
> an observation of regular patterns as it is a reshuffling of ideas from TKD
> into a form that's more restrictive than what TKD says.

Yes. It is more restrictive. It is more specific. It is a more accurate 
description of canon. There's nothing wrong with that.
 
> >But, as I said, it's a free country. If you want to just not notice that
> >except for {-'e'} and except for special verbs of motion, all Type 5 suffixed
> >nouns go before the direct object, go for it. Keep your eyes peeled for all 
> >those canon examples where Okrand puts Type 5 noun suffixes on subjects and 
> >objects. I'm sure we'll see lots of them any day now.
> 
> With the usual exceptions, all Type 5 suffixed nouns are *seen* before the
> direct object.  When I think hard about it, I tend to conclude that it's a
> result of other underlying factors, rather than a rule in and of itself.

That is not important. I'm not commenting on the CAUSE of the grammar. I'm 
merely describing the consistent behavior of words as they are used in canon. 
That's what descriptive rules are.
 
> The argument would likely go a lot easier on everyone if we could find the
> magic phrase which would make the "header" idea clear to you in a way that
> doesn't cause you to misunderstand and think that it claims Type 5 noun
> suffixes can go anywhere.  I'll try again:
> 
> TKD says that objects come before the verb, subjects come after the verb,
> and (here's the relevant part) nouns that are anything other than subject
> or object come before the object noun.  Such nouns can be called "header"
> nouns, and by definition they come before the object. 

Please note that your choice to call these "header" nouns is no more canon than 
my observation that with two exceptions, Type 5 suffixed nouns consistently 
belong to this group. If you want to restrict me from making accurate 
observations about the language because that's not what Okrand said, then you 
should restrict yourself from using terms he never uses.

So, I'll stop talking about my descriptive rule if you'll stop using the 
term "header noun". We'll restrict ourselves to canon terminology and rules. Is 
it a deal?

> TKD also says that
> nouns of this sort usually have a Type 5 noun suffix, but it also provides
> a number of examples where a non-subject, non-object noun lacks a Type 5
> sufffix.

These exceptions come in two groups. Time stamps and unmarked locatives.

> Your focus seems to say that when you put a Type 5 suffix on a noun, that
> makes the noun have to go before the object (with exceptions).  That works
> (with exceptions), but it doesn't really have anything useful to do with
> the meaning behind the words, which is why the suffix is there in the first
> place.

Type 5 noun suffixes are described by Okrand as "syntax markers". Each one 
(except for {-'e'}) has a syntax which applies a role to that noun which 
fundamentally conflicts with the role of subject or object. A certain specified 
group of motion verbs have a special locative relationship with their direct 
objects, so those verbs can optionally use {-vo'} or {-Daq} on their direct 
objects. That makes the second exception.

We can study this "cause" of the coincidence that with these two exceptions, 
all nouns with Type 5 suffixes come before the direct object, or we can just 
observe that it always happens. I really don't care. Either way, if you make up 
a descriptive rule of the way things work, the conclusion is the same.
 
> A "header noun" focus on semantics would put the non-subject/object verbs
> first anyway, without using the suffix as a trigger.  The meaning would
> most often simultaneously require an appropriate suffix, but the Type 5
> suffix isn't what puts the noun in its place.

I don't care WHY the noun is always there. What I care about is THAT the noun 
is always there. I understand the reason the noun is there.

Okay. Let's say we go to the beach. We notice that people (subjects and 
objects) on the beach (before and after the verb) never wear shoes (Type 5 noun 
suffixes). We observe that most people ("header" nouns) on the pavement (the 
beginning of the sentence, before the direct object) do wear shoes. A couple of 
guys (time stamps and unmarked locatives) consistently refuse to wear shoes on 
the pavement, but then, those guys almost never walk on the beach. They are 
special. They have very tough feet. This is just what they do.

Now, please note that we are ignoring flip-flops (the suffix {-'e'}) because 
people wear these on the beach or on the pavement. You can't make any 
generalization about flip flops. They are exceptional to this argument. Krankor 
thinks people only wear flip flops on the beach and wants canon examples of 
somebody walking on the pavement in flipflops. Sightings are pretty rare of 
flipflops on the pavement, so I'm not sure I can name names.

Now, Okrand has said that everybody who is not on the beach is on the pavement, 
and people on the pavement usually wear shoes. I want to name names and talk 
about the only guys we ever see barefoot on the pavement. That's apparently not 
politically correct, even though it is actually more accurate than just saying 
that people on the pavement usually wear shoes.

Then, there's the boardwalk [where the direct objects of special verbs of 
motion hang out]. It's not part of the pavement, so it's sort of like the 
beach, but not quite. People there (direct objects) sometimes wear shoes and 
sometimes they don't. That's an exceptional place to this observation of a 
generality about the beach and the pavement.

So, now, I want to simply say that, except for flipflops and except for the 
boardwalk, everybody wearing shoes is on the pavement. This observation happens 
to be true. Nobody EVER walks on the beach with non-flip-flop-like shoes. Now, 
you want to argue that the REASON people wear shoes on the pavement is because 
the pavement is hot.

I don't care about the REASON people wear shoes. I'm just making an observation 
about where the people are who wear shoes. This is merely a different 
perspective on the generality Okrand made when he said that people on the 
pavement usually wear shoes. He has to say "usually" because of those two guys 
(unmarked locatives and time stamps) who keep walking around on the pavement 
barefoot.

You then say, "But maybe someday, somebody will wear shoes on the beach."

I respond that I don't care. I've been looking around for more than ten years, 
and I have never seen anybody on the beach with shoes. People in shoes are on 
the pavement. If I see new canon this generality doesn't cover, I'm happy to 
modify the observation to make it more accurate, just like we did with OVS and 
{-jaj}.

Do I make myself clear yet?

> A: "nouns with Type 5 suffixes go before the object"
> B: "non-subject, non-object nouns go before the object, and usually have a
> Type 5 suffix"
> 
> Please try to see the difference between A and B without thinking that
> someone asserting B is saying *anything* about what suffixes can go on a
> subject or object.  B is what TKD says, and it's very productive.  A needs
> lots of tweaking before it's a useful rule.

What I'm really saying with A: is "nouns with Type 5 suffixes usually go before 
the object" and then I explain the two exceptions that make me use the 
word "usually". If you look at B: and get specific about its two exceptions to 
the second part, requiring the word "usually", then both rules are equally 
productive and require equal amounts of tweaking.

> -- gnunchu'wI'

Will

Back to archive top level