hash of hashes

Hi.
I have a file with a word and it's part-of-speech on each line.
I have so far put all words in one array and the part-of-speech in another s
o that one word's index corresponds with the part of speech.

But what I really want to do is to check which part-of-speech one word has (
 it can have more than one) and then count how many times it has the differe
nt part-of-speech.

So I was thinking about using a hash of hashes where the parent hash has wor
ds as keys and hashes as value. The child hashes should have part-of-speech 
as keys and the number of these as values.

Like this example:

# %wordPos = (
#     cat => {
#         noun      => 30,
#     },
#     race => {
#         noun      => 6,
#         verb      => 2,
#     },
#     jump => {
#         noun      => 3,
#         verb      => 7,
#     },
# );


I don't know if there is a better way to do it. But I still don't know how t
o build this hash of hashes with the input I get.

/Maria