Re: Vocabulary concept mismatches

On Thu, Mar 04, 2004 at 07:48:27PM -0600, Herman Miller wrote:
> One of the things that's really interesting about languages, but not
> well documented in dictionaries, is the fact that the meanings of words
> don't match precisely from one language to another. The most obvious
> examples are cases where a single word in one language corresponds to
> two or more words in another, like we've been discussing recently; example:
I face that problem every time I add an entry to the Ebisédian lexicon.
Ebisédi culture and con-universe are so different from Terran culture (and
especially English concepts) that sometimes simple words require
cumbersome periphrasis. For the lexicon, I've chosen to err on the side of
excessive glosses to give a better sense of the nuance of a word, than to
choose a single, closest English word which fails to adequately express
the word.

To take a simple example, consider the English verb "to hold". In
Ebisédian, you can use the verb _je're_ to describe how a merchant might
hold up an item of merchandise to display it. On the other hand, if you
picked up an object from the ground, you'd use _ta'ge_ instead. The verb
_je're_ implies a sort of holding up, or even the wielding of a tool or a
weapon; whereas _ta'ge_ means "to finger" something, to pick something up
to examine. A third verb is _laa'ma_, to carry something. After you pick
up (ta'ge) something, you take it with you (laa'ma) and show it (je're) to
your friend. In other contexts, all three verbs may be translated
"to hold":
	- To hold something you just picked up - ta'ge
	- To hold something for display - je're
	- To hold something as you walk - laa'ma

Not to mention that _laa'ma_ also means "to bear a heavy load", which is
an entirely different nuance from _ta'ge_.

[snip]
> Some cases are slightly more complicated. Tirelat for example has two
> words meaning "eye", but they can be translated more precisely as
> "eyeball" and "iris". So here we have three English words corresponding
> to two Tirelat words.
Ebisédian terminology for different parts of the body do not always map
nicely into English. For example, _Ka'jim_ means "head and shoulders" as a
single unit, and _ta'grim_ refers to the lower part of the face that
includes the mouth, lips, and chin. These terms are also inflected for
gender; a word like _faTui'_ (fem. of _fatui'_), referring to the part of
the face around the eyes, conveys the expression of a woman's eyes, and
its masculine form, _fa'tui_ conveys the expression of a man's eyes. Much
richness of nuance is available here; whereas in English, it's just plain
and bland, "eyes".

[snip]
> But if I actually sat down and started documenting each word to that
> level of detail, I'd never have the time to finish a reasonably large
> part of the vocabulary. So for most of the vocabulary, I'll have to
> settle for brief definitions and a few example sentences.
Personally, I prefer to be stuck with a small lexicon than to lose this
richness of nuance that better expresses what the Ebisédian really says.

[snip]
> Now there are some other interesting things about this sentence, such as
> the fact that the Sangari say "Zara's eyes are green" rather than
> literally translating "have". But then there probably ought to be whole
> pages describing the equivalents of troublesome English words like "have".
Ebisédian completely lacks verbs of being (because the state of being is a
state and not an event) and verbs of possession (although it does have
verbs of acquisition, since those are events). This makes it difficult for
beginners to translate English prose that use these verbs extensively.
Possession is signified by the idiomatic juxtaposition of the conveyant
case (for the thing possessed) and the receptive case (the possessor). In
this instance, the receptive case is a lot like the dative (cf. Attic
Greek usage of the dative of possession).

However, the picture is clouded by other idioms that employ the exact same
construction; e.g., to indicate intimate relationship (esp. romantic or
spousal relationship). So when you have:

	<inanimate object>-CVY <person>-RCP

it means <person> owns <inanimate object>; but when you have

	<personA>-CVY <personB>-RCP

it means <personB> has romantic interest in <personA>. Furthermore, if you
have:
	<quality>-CVY <person>-RCP

it means that <quality> is an adjective modifying <person>. E.g., "a tall
person" is translated as _th0't33 bis33'du_ (tall-CVY person-RCP).

This makes it non-trivial to translate between the two languages.

> That's not even getting into all the metaphorical uses of "eye", like
> the eye of a needle or the eye of a storm. These really are different,
> unrelated concepts, and should be listed specifically in the dictionary.
> Maybe they speak of the "mouth" of a needle, or even the "nostril" of a
> needle! Or maybe they just use a plain unpoetic "hole", or don't even
> have anything like needles in their technology.
In Ebisédian, words like "eye" (zoji') or nose (ngisi') cannot be used
metaphorically.

Nevertheless, Ebisédian does have a few words which are literally
impossible to capture with any English word in a general sense. One such
is _gii'j3li_, whose literal meaning eludes direct expression in English,
although it can be described in a multitude of ways, eg.:
- it can mean "troublesome" or "bothersome" (e.g. an expression of
  frustration, as in "why is this so complicated (gii'j3li)?!");
- it can refer to a series of tasks e.g., in an obstacle course, which one
  must solve in steps in order to reach the goal;
- it can refer to rowdiness or uproar (e.g. a mob);
- it can refer to the events in a perpetually unfortunate life;
- it can refer, often perjoratively, to the things someone does which the
  speaker does not approve of (eg. "what is this nonsense (gii'j3li) that
  you do every day?!");
- it can refer, in a jocular way, to the antics someone did way back when
  in highschool;
- etc.

Nonetheless, the Ebisédian meaning is really only one.

Another aspect that makes it difficult to map English to Ebisédian (or
vice versa) is the fact that Ebisédian typology has no concept of
active/passive. A verb like _le's_ can mean "to go" or "to come" or "to
move about". A verb like _be'jh_ can mean "to give" or "to be given", or
"to deliver a gift". A verb like _dhel0'se_ can mean "to strive towards"
or "to surpass", depending on which noun cases are used. Eg:

	<personA>-ORG dhel0su'e <personB>-CVY

means <personB> surpassed <personA>; whereas

	<personA>-CVY dhel0su'e <personB>-RCP

means <personA> is striving to beat <personB> but is still behind.


T

--
Ruby is essentially Perl minus Wall.

Reply