Re: OT: Dictionaries: why two sections??

Henrik Theiling skrev:
> Hi!
>
> Do you often open a dictionary, try to find a word and
> when you reach the corresponding page and line, you notice
> that all words around are in the wrong language? And then
> you notice that you started in the wrong one of the two
> sections of the dictionary, namely exactly the wrong
> direction of translation? And have to do the lookup again
> in the other direction?
>
> It happens to be very often. I want to look up
> 'schimmelig' in a Latin-German, German-Latin
> dictionary and then at sch..., there are only Latin
> words around! Oh, yes, wrong section. Use German-Latin
> instead of Latin-German.
>
> Why are there two sections?? It is unnatural! Let all the
> words be in one list, no matter what language they are in!
>
> Did anyone encounter this problem before? Did anyone have
> the same solution? :-)
>
>   **Henrik (with way too many dictionaries around him)
One English-Swedish dictionary I owned had a gray strip
printed along the outer margin of each page in the Swe-Eng
section, which made it easier to see at a glance which
section one was in. I applied this to some of my travel
pocket dictionaries by coloring the outer edge of the leaves
of one section with a colored marker. Worked quite dandy!

Personally I think it would be very confusing not to have a
separate index for each language, although it would be
easier if each language was consistently printed in
different typefaces (e.g. sans and serif, although I suppose
you've seen the old Langenscheidts dictionaries where German
is printed in Fraktur and the other language in Roman!). But
as Larry pointed out you get into trouble with different
scripts, or even with different sorting orders for different
languages printed in the same script, e.g. Swedish
vs. German ordering of _ä ö_, or Sanskrit which is sorted in
Devanágarí order even when printed in Roman. I can
assure you that looking up Swedish ö-words amidst o-
words just doesn't compute in my brain, nor does looking
up Sanskrit words in Latin sorting, although I'll admit
that it's of course an acquired habit.

The real solution is of course an electronic dictionary,
where the looking up is done by a machine, and you get all
hits and near hits -- which should ideally include possible
cognates and derivatives of the same stem -- served in a
list(*). It is perfectly possible to have the lookup program
guess the language for you as at <http://dict.leo.org/>.
Needless to say you'll also have the added advantage of
hyperlinking. My guess is that dead-tree dictionaries will
be quite obsolete in a hundred years time.

(*) E.g. I thought out an algorithm for searching an Old
English dictionary where e.g. an entered search string
"sieythea" would be converted into a regular expression
/s[iey]+([dðþ]|th)[ea]*/ with which the real search is
made, i.e.:

* Any sequence of two or more vowels is searched as one or
   more (but word-finally zero or more) instances of any
   combination of these vowels.

* Any vowel will of course cause the engine to look for both
   its short and long counterpart!

* The sequence "th" will cause the engine to look for any of
   _d, ð, þ, th_.

* Some other consonant combinations will be treated
   similarly, e.g.

         # "ch" before any consonant will cause the engine to
           look for any of _h, c, ch_ (/[ch]+/) in that
           position.

         # "x" will look for any of _sc, cs, x_.

         # "k" and "c" will be treated as equivalent.

         # Palatalization dots above _g_ or _c_ will be
           ignored.

This way a search string that has been converted into a
regular expression according to a rather simple set of rules
will find most or all words related to the word you are
looking for. Of course no word is actually spelled with
_iey_ in OE, but a user is expected to know that there is
merger between _ie, i, y_ in late West Saxon. Such cognate
words will be far apart in a traditional paper dictionary!

--
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    "Maybe" is a strange word.  When mum or dad says it
    it means "yes", but when my big brothers say it it
    means "no"!

                            (Philip Jonsson jr, age 7)


--
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    "Maybe" is a strange word.  When mum or dad says it
    it means "yes", but when my big brothers say it it
    means "no"!

                            (Philip Jonsson jr, age 7)


--
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    "Maybe" is a strange word.  When mum or dad says it
    it means "yes", but when my big brothers say it it
    means "no"!

                            (Philip Jonsson jr, age 7)