Although I draw free-hand, the cartoons don't come out of my black felt-tip pen just as you see them here. I draw a version, and then take the best part of it - say - the eyes, and then clip out that portion, put it beneath a fresh piece of paper and start drawing, say, hair.
Each of these vegetables was drawn separately, then pasted together.
When I have good eyes and good hair, I'll clip them out and paste the two together, put it beneath another piece of paper and start working on, say, clothes. If I finally draw a good set of clothes but it seems too small for the head, I'll use a Xerox machine to enlarge it, clip out the enlarged version, and start pasting it all together again. When I'm done, I make a final photocopy and a bitmap scan.
Needless to say, it takes quite a bit of time to make the drawings look effortless.
This multi-layered process caused the cartoon editor of the New Yorker, a singularly unpleasant man named Bob Mankoff, to question whether I had any talent at all. He also showed me how he could easily (if ineffectively) duplicate my style on his PC, drawing free-hand with his pen mouse.
Why the method matters if the final image is pleasing is beyond me, but Mr. Mankoff - who had invited me in for an interview for reasons now lost to time - also told me my cartoons had "too much style." What a meatball. Even his secretary detested him, I found out when I ran into her at a party.
Anyway, one of the advantages of this long process is that it produces one-of-a-kind originals - the last layered copy before the final photocopy, which can have up to ten layers and looks a little bit like a jigsaw puzzle. Over the years, some of the fans have been kind enough to purchase these originals, the only money I have ever made from the site.