A little while back, Love, Sex & Merlot, our sister imprint, posed an intriguing question: Can a reader consistently tell the gender of an author, solely by reading their dialogue?
The preliminary answer was too thought-provoking to not share…
You CAN tell the gender of an author solely by their written dialogue.
This is the list of books they used, keeping the same order as the snippets listed in their earlier poll:
Snippet #1: The Shadow in the North, by Philip Pullman (M)
Snippet #2: One for the Money, by Janet Evanovich (F)
Snippet #3: An Arranged Marriage, by Jo Beverley (F)
Snippet #4: The Eyre Affair, by Jasper Fforde (M)
Snippet #5: Potatoes Are Cheaper, by Max Shulman (M)
Snippet #6: The Dressmaker, by Posie Graeme-Evans (F)
Snippet #7: Grania, by Morgan Llywellyn (F)
Snippet #8: The Wedding Officer, by Anthony Capella (M)
Snippet #9: Victory of Eagles, by Naomi Novik (F)
Snippet #10: Nina September, by Bron Zeage (M) … (P.S. This is LSM’s first erotica – mark your calendars for the July release!)
These were the guesses:
This, then, is the percentage of correct guesses:
They had unequal numbers of answers on most of these, since they ended up posing half the questions on social media as well as on the website.
But the results were rather astonishing.
Of the snippets with more than two answers (the ones with only two answers should be likely throw out for science’s sake), there was a remarkable lean towards one solid answer: Yes, you can tell the gender of an author by their writing style.
A few respondents were thrown off track by the gender of the narrator, but for the most part readers guessed correctly every time. This was made even more remarkable by the fact that these snippets represent books of many genres, published between the 1960s and the present day. It’s not always the case, clearly—Naomi Novik threw people for a loop. But 4 out of 5 leaves a far more obvious result than anyone was expecting.
Snippet #1 (Pullman) is a male writing a female lead. Snippet #2 (Evanovich) was specifically chosen to be tricky and seem masculine. Snippet #3 (Beverley) is a female writing a male conversation. Snippet #3 (Fforde) is, again, a male writing a female lead. And snippet #9 (Novik) is a female writing a male lead.
The main distinguisher that readers mentioned, throughout the polls, was basically brevity (in existence or lack thereof). The male authors tended towards shorter, brusquer sentences with fewer adjectives and little variation in verbs; the female authors tended towards longer or more varied line lengths with more varied adjectives and verbs.
The other interesting piece of the puzzle was that respondents who are Readers—i.e. read voraciously, and/or read equally in various genres—almost always guessed correctly. Respondents who don’t read daily, or who don’t read much fiction, had a harder time guessing.
We suppose the takeaway from this experiment is that regardless of how good a writer you are, some of your own personality still shines through. They wanted—and, quite honestly, they and we were expecting to be able—to say that those ridiculous people who maintain that males and females write dialogue differently are full of it. But LSM’s brief results speak to the contrary.
It’s possible to fool a reader, but seems to require either a large amount of proofreading (and understanding what giveaways you’re looking for), or working very hard to cultivate a particular style (Novik is well-known for her crisp dialogue, and had been using William Laurence as her main narrative character for five books at the point of this snippet). Many of these authors are wholly famous, and their writing beloved, yet they still can’t fool a reader with a watchful eye.
Nevertheless, we were intrigued by the results of this experiment, and are thinking about experiments we might implement in coming weeks. And, much like the Mythbusters, we’re thinking of ways LSM might improve this experiment. Maybe all ten books could be chosen from the same genre? Maybe they could choose ten different snippets from the same author (without telling the reader, of course)?
In light of these results, how do you make sure that your dialogue fits your characters? Let us know in the comments below!