Here is one without the hash, using a table. It does make me think a database would be better for this application :-) but perhaps we should looks at the real application as being learning Perl...

/usr/bin/perl
use strict;
use warnings;

my @numbers = (
    [qw( Italian Spanish French English Welsh  )],
    [qw( uno     uno     un     one     un     )],
    [qw( due     dos     deux   two     dau    )],
    [qw( tre     tres    trois  three   tri    )],
    [qw( quattro quatro  quatre four    pedwar )],
    [qw( cinque  cinco   cinq   five    pump   )],
    [qw( sei     seis    six    six     chwech )],
    [qw( sette   siete   sept   seven   saith  )],
    [qw( otto    ocho    huit   eight   wyth   )],
    [qw( nouve   nueve   neuf   nine    naw    )],
    [qw( dieci   diez    dix    ten     deg    )],
); 

print "Please enter number to translate\n";
my $num = <>;
while ($num) {
    print "You typed $num\n";
    chomp $num;
    for my $row (@numbers) {
        # dereference array, and look for out number in it
        next unless grep {/$num/} @$row;

        print "I found it: ";
        print join " <-> ", @$row;
        print "\n";
        my $i = 0;
        $i ++ until $row->[$i] eq $num;
        print "It looks like it was in $numbers[0]->[$i]\n";
        last;
    }
    print "Please enter another number to translate\n";
    $num = <>;
}
[download]

The question this raises, is what happens when you type in uno? Altering it to know when a number has multiple matches, is left as an exercise for the reader.

Cheers,
R.