For me, drafting a novel is only partly about getting words down. It’s also a chance to experiment with writing habits. Recently, for example, I experimented with using existing prompts, like photos or music CDs, to help create characters. It is a new technique for me, and I was curious what the results might be.
My usual approach is begin with the needs of a plot, and keep asking questions. What kind of character might fill the necessary role? What relevant back story does the new character have? And, in some ways, most important of all, what other roles might the character play so I don’t have to invent another one? Many of the answers only come as I write, sometimes after several rewrites. From my feedback, this technique can produce colorful characters, but it’s not quick or efficient. If an alternative technique could save time, I was ready to try it.
In this experiment, I wanted a love relation for a heroic figure. More – I wanted a lover who could stand up to him, and was maybe heroic herself. Immediately, I thought of a tentative character for my next novel attempt, an older woman, a widow, who is a major figure in the Sisters, a group of vigilante women who punish predatory and abusive men. I based this character on a publicity shot of folksinger June Tabor:

In this picture, Tabor has a “don’t mess with me” look. She looks sardonic, with a flare for the dramatic in her black. She may be in her sixties or thereabout, but she is someone who can take care of herself. The only difficulty is that she is thirty years too old for the character I want. I easily found pictures of a younger Tabor, but none of them had the same look (these are, after all publicity shots, and may have little connection with Tabor herself). So I set about the task of imagining the Tabor of the original photo as a younger woman.
Tabor herself has traces of a North Country accent, and I might have given my character the same voice. However, while I developed the character, I came across a documentary about the folksinger Kate Rusby. I was especially interested to learn that Rusby lives in a village near where my father grew up – and before I knew it, my character had a Yorkshire accent. Not a thick one – I wanted to have mercy on readers – but some of the cadence and a few of the words.
Having a good sense of what the character looked and sounded like, I went to work on her back story. While I already knew she had been in the militia, I wanted some heroic proportions in her. Inspiration came in the form of “Thomas the Rhymer,” a song I first heard at fourteen. The song tells of how Thomas of Ercildoune accompanies the Queen of Elfland to her realm. The song was my introduction to medieval ballads, and what has always fascinated me is that its hero was a historical figure. Go to Scotland, and people will show you the tree Thomas stood beneath when he met the Queen of Elfland, centuries ago.
During the trip to Elfland, Thomas is given “the tongue that will never lie,” although he begs the elf queen to take it back. Upon returning home, he becomes a prophet – and some of his alleged prophecies still exist. Perfect, I thought. With such a gift (or curse), my character would speak carefully. Like Cassandra of Troy, prophecy has warped her life, except that, instead of never being believed, my character would have suffered because she tried to keep a prophecy or two from coming true.
At this point, I had enough to start writing the character. When I completed the chapter in which she first appears, I was pleased with the result, and the first critique was reasonably positive. In terms of results, the experiment of finding prompts worked.
However in terms of efficiency, I am not so sure. I did not save any time, and if I had not already had a photo as a starting point, I could have spent much longer developing the character. More disturbingly, my critique partner warned me that I seemed to have become obsessed with the character, which was quite true for a few days, and in a way that couldn’t be justified by the character’s relatively minor role.
So, a mixed result overall. While I might use prompts if my usual plot-driven character development fails, I can’t see any reason to switch to it exclusively. You might have different results.