Marty Rosenberg wrote:
> I forgot to mention a few things that I could especially use help with,
> although everything is pretty much a blank slate for me.
> First, I have difficulty coming up with words. I tried using roots and
> particles to do so, but then I realized something else: I have
> difficulty coming up with those too. Is there any system I should know
> of that could help me along these lines?

My current method (algorithmic--but hey, if you've looked at my
dissertation, you'd know my preference for such things... :) ):

I went to and generated a bunch of random numbers between
1 and 100 formatted into 6 columns (for 2-syllable (CVC) roots). Then
applied the following procedures to the numbers that resulted:

(symbols: ā - a macron; ē - e macron; ō - o macron; ð - eth; þ - thorn)
ONSET CONSONANT (* means this position is not filled)
b (1-6)    l (34-39)   t (67-71)
d (7-11)   m (40-44)   þ (72-76)
f (12-16)  n (45-50)   ð (77-80)
g (17-21)  p (51-55)   v (81-85)
k (22-27)  r (56-61)   z (86-90)
h (28-33)  s (62-66)   * (91-100)

a (1-15)   ā (71-78)
e (16-28)  ē (79-85)
i (29-38)  ō (86-91)
y (39-49)  ay (92-96)
o (50-58)  ey (97-100)
u (59-70)

FINAL CONSONANT (* means this is empty)
l (1-20)   * (93-100)
n (21-38)
r (39-54)
m (55-68)
þ (69-81)
v (82-92)

1. z is always dropped unless it is the final consonant of the root
2. for all syllables but the first: if the final consonant of the
previous syllable is:
    l then drop b, f, p, ð as initial consonant
    n then drop b, f, h, l, m, r as initial consonant
    r then drop l, ð as initial consonant
    m then drop d, f, h, n, r, t, þ as initial consonant
    þ then drop all initial consonants
    v then drop b, d, f, h, m, s, t, þ, ð, v
3. double consonants (between syllables) are allowed except for þþ and vv

EXAMPLES (these are roots, the fully formed words have class endings)
66  89  45  46  54  22  sōrnon (sōr - 'cold'--I didn't like the '-non'
second syllable, so I removed it)
5   91  80  44  1   21  bōþman -> bōþan ('island')
64  3   99  4   86  30  sa*bōn -> sabōn ('to learn')
94  8   94  64  26  91  *a*sev -> asev  (not yet assigned)

The trick is to set up your rules so they reflect the phonetic rules of
your language.
James Worlton
          "We know by means of our intelligence
          that what the intelligence does not
          comprehend is more real than what it
          does comprehend."
                           --Simone Weil