--- Haukur Thorgeirsson <haukurth@...> wrote:

> Tim hat gefragt:
> "Is this a close parallel with
> Mod German kennen/wissen?"
> While Keth wrote an informative answer to this
> I am inclined to think the the simple answer is
> "yes" :)
> kenna - kennen - I know him
> vita - wissen - I know it
> kunna - k�nnen - I know how
> Oh, well, "kunna" and "k�nnen" aren't quite the
> same.

I have a mixed message - one, Keth is right in
pointing out that the use of "kenna" as in
"spuertst Du, dass die Kaffee warm ist" (Do you
feel that the coffee is warm) is foreign to
German. One would never ever say "kennst Du, dass
die Kaffee warm ist". Other example is in his
grammatical treatise example, "Soloecism ver�a i
tveim partum" = Soloecism 'wird' in zwei Teilen
(Soloecism 'becomes' in two parts). You can't use
'become' in this sense in German ('werden') or
English.

On the other hand, the differences in usage are
so numerous that all prescriptive attempts to get
across the exact usage of any given word amounts
to a futile attempt: if you want to
prescriptively specify the usage of "kenna", then
logically you should go on and try to specify the
special usages of *all* words (and how the usages
can change completely in idiomatic juxtaposition
to other words).

That makes a page of explanation for each word
alone, and another page of explanation for word
the word can be combined with to change it's
meaning (say, on average, about 5 idiomatically
mutating words for each word in the language). At
about 30,000 words in any workable language, that
makes a 150,000 page dictionary - an impossible
work to write in the first place, much less have
any chance of learning!

That's why language training in the 'real' world
actually just 'stops' at a certain point: I can
remember my frustration as my German advanced to
the point that my German instructors could no
longer explain why certain words were used a
certain way in more sophisticated usages - there
comes a point where the rules cannot be taught by
people anymore.

This is because people know more than they know.
That's not really such a mysterious statement -
in fact, a down-to-earth comparision with a
computer system reveals the same thing: if, say,
a computer has a set of rules it has for
processessing and printing out information, that
does not mean that one of the functions the
computer can print out is a list of these rules!
To do this, the computer would need to be
programmed to print out it's own rules (esp in a
format that these rules could be processed, for
instance, by another computer). But then the
rules in the computer consist not only of the
rules it uses for processing whatever data it is
mean to process (let's say complex weather
information), but also includes the rules it
follows for printing it's own rules.

That's not so hard - a beginner can make a
program that does nothing more than print itsself
out, so what's the comparision to people? To
compare you have to add the extra feature that
people have learning processes running all the
time. We have some rudimentary learning going on
in computers today - what it basically is is a
process by which the program is built to not only
process incoming data, but also treat *its own
code* as a kind of *data* that can be processed.

Thus, when such a learning computer would be
given the job to print out it's code, it would
automatically learn during the process of
printing and thus need to change the code while
it is printing it. In a poorly thought-out
program, that would mean that it has its code
file open for output (to print it), and
simultaneously try to open it for input (to add
learning to it), causing a write-protect error
and halting the program.

Of course any half-decent programmer could get
around this, for instance, by halting the
learning routines while the printing routines are
running, or copying the code to a stored file and
printing that (thus not printing the small amount
of code changes that may have happened in the
process of printing). Note, however, that we
haven't programmed a computer that can do
something as complex as speak for itsself (I
myself think that means the computer would
'think' like you and I), so we don't really know
how complex the task of "printing its code" would
really be for a computer that can talk.

The point is that the human brain did not evolve
for the purpose of "printing" at any rate - the
capacity for language did not evolve so that we
could spend our time teaching the languages, but
so we could speak to other native speakers of the
language. That's why the brain doesn't have any
of the special programming to keep write-errors
from happening. The learning routine is running
by definition *all the time* and has no switches
to shut it off for special runs. It doesn't even
have the "code" organised into a simple thing
like a "file" that can be identified and
described by the speaker.

Humans having amazingly flexible brains, there is
something called "language rules" which we can
become conscious of. That doesn't mean that the
brain is divulging *all* the rules it uses when
devising speach (if we knew all the rules then we
could literally bring a machine to hold up a
conversation just like a person), but it *does*
divulge *enough* rules to give a person the
'ground work' with which they can progress
further.

The 'ground work' is, strangely, not necessary
for babies. But then babies spend two years of
their lives not being able to ask any other
living beings for the things they want in any
specific way, so they have a major, overwhelming
motivation to start to make sense out of the
language they are hearing. Most brains can
apparently only go through this kind of stress
once, since only one in a million peole can
actually pick up a second language just by
mimickry and trial and error. You see the
examples among the immigrants to the US in the
last two centuries, who often came with no
schooling in English whatsoever. They all 'picked
up' some English usage, but very, very few seemed
to still be able to just relearn the language
with the same automatism they learned their
native language.

So the effective way for non-babies to learn
language is to first of all give them the
ground-work (that is, grammar, word-order etc) of
a language, and a more-or-less one-to-one
correllation of the foreign words to their native
words. This gives them at least some meagre
chance of struggling through sentences.

And when they start to struggle through the
sentence, I don't care how much prescriptive
training they have gotten from instructors - they
are going to be in for a shock! *That's* when
they start to stumble over all the idiom and
special usage issues that no human instructor can
adequately bring across. Learning at that point
really *is* something like a baby's learning: it
is intuitive, and 'sinks in' in a way that can't
really be described in rules. Keth describes this
as a 'feel' for the language. They then begin to
really understand each word not in the terms of
the equivalent word in their own language, but as
a description of reality that is encoded in their
brain in much the same mysterious code that the
native language is encoded. However, the encoding
rarely, rarely installs itsself with all the
functionality of the original language (I don't
think there is enough Old Norse literature to
enable even linquistic geniuses to achieve this
level in Old Norse - and the crucial added info
delivered by *oral intonation* is lacking
entirely). It's like the original email software
you install - that email software is the one
registered with all you other programs, and if
you install a new email software, it either is
not the 'default email', or - if you make it the
default - it never meshes with the other programs
quite like the originally installed software.

So, no, I don't think Keth was correct in
pointing out the exact usage of 'kenna' - we will
have to painstakingly learn the 'real' usage of
the word by seeing it used over and over again
when we finally get enough 'ground work' to make
sense of Old Norse material at all.


=====
Kindest Regards,
- DeepStream
|'''' ''''||'''' '||'''' '':
||'''' ''''|'|||'''' '||'''' '|'''':
||'|'''|'''' ':|||''''||'''' ':|||''''||'''' ':
|||'''' '|'''' '''|':|'''' ''''||''':|||'''' '|||''||''

__________________________________________________
Do You Yahoo!?
Get email at your own domain with Yahoo! Mail.
http://personal.mail.yahoo.com/