If you have a list of the pairs you can loop through them and for each, do:

grep '[kern pair]' /usr/share/dict/words

or you can do clever things like two at once, eg

cat /usr/share/dict/words | grep '[kern pair 1]' | grep '[kern pair 2]'

so

cat /usr/share/dict/words | grep 'av' | grep 'aw'

will get you

awave
drawshave

to output words in all caps add this

| tr '[a-z]' '[A-Z]'

to output words with initial caps add this

| sed 's:^\(.\):\\U\\1:g'

The above ideas should work on most or all Unix/Linux-based systems, including System X.

But if all you have is a Windows box, you can get some wordlists from http://wordlist.sourceforge.net/ and the utility kit from http://unxutils.sourceforge.net/ and then get down and dirty with the command line. I think they'll work with MSWindows 95-XP. Beyond that, no idea at all.

I was thinking more along the lines of using python to dump a collection of word lists into one big and then iterate through a list of pairs and formatting the new split words. I'll cobble something together tomorrow.

.

I wonder how
http://adhesiontext.com
that generates dummy text following different languages and criterias, is working...