• Damned Souls and Statistics

    by  • November 11, 2013 • Fiction • 0 Comments

    An essay by Allison Merton, as provided by Robert Dawson
    Art by Scarlett O’Hairdye

    “But Pascal expected a reward of infinite value. Do you?”

    His eyes, shadowy as a charcoal drawing, fixed me. His face was bleak and gray, his voice was flat and emotionless, and his dark suit looked as if it had been washed slightly too hot. No cloven hooves or scent of brimstone, but, very deep inside, I already suspected who he was.

    I stared at my drink, then tried to look away, to find a familiar face in the Faculty Pub. But somehow, that Friday, in a crowded room full of colleagues, I could find nobody to rescue me from this interrogation. Backs were turned, faces invisible behind the dark square wooden pillars that broke up the room. I recognized a few faces, but could not put names to them. I realized how rarely I had come here, in my five years teaching at Greenwald. My eyes were drawn back to his.

    “No, I don’t.” Through a gap between the long curtains I could see the warm tangerine light of the October sunset. It should have been a simple matter to excuse myself and walk out of the building. But the reason for walking out was one my modern mind would not put into words; so I stood there, at one end of the bar, my back to the wall, like a deer in a car’s headlights.

    “You say that you’re a statistician.” The voice stayed flat. “You teach probability theory.”

    Somehow I was on the defensive again. I tried to stick to the facts. “Yes, I have a PhD in statistics.”

    He said nothing, but I felt as if, somewhere inside my head, a trap had been sprung. My mind flooded with shame-filled memories of my thesis defense. Partway through, somebody had asked me to explain the derivation of a formula. I started to explain: the transformation, the differentiation under the integral sign. With my mouth dry, I mentioned the convergence criteria that made it work, a detail I’d omitted in the thesis. Then the creeping terror began. Those convergence criteria were met there all right; but I’d used the same technique in another derivation, late in the second chapter. I was sure I had not checked convergence there either. With a dreadful cold certainty, I realized that I could not guarantee more than local conditional convergence for that class of distributions.

    The proof did not work. In a moment one of the examiners would bring it up. For half an hour I died there. Finally the ordeal ended. They hadn’t found me out.

    It was six months before I dared to look at my thesis. By then I had taken my degree (in absentia, to my parents’ distress), and had been promised a job. On review, I realized that the error was minor, and could be repaired with another page of work. But my joy in my new doctorate had never been whole and never would be.

    Damned Souls and Statistics

    Like a typical client, he didn’t volunteer an explanation. Good–maybe I’d get away with this.


    To read the rest of this story, check out the Mad Scientist Journal: Autumn 2013 collection.

    Allison Merton has a PhD in statistics from Midnorthern University. Her thesis, still unpublished, is on transformation techniques in Bayesian statistics. She teaches at the Department of Statistics of Greenwald, where she received tenure last year. She prefers not to do consulting work, and keeps meaning to set up a personal web page.

    Robert Dawson has a PhD in mathematics from the University of Cambridge. Parts of his thesis have been published and possibly even read. He teaches at the Department of Mathematics and Computing Science at Saint Mary’s University (the Nova Scotian one), and writes science fiction in between times.  His web page is at http://cs.smu.ca/~dawson/Writing.

    Scarlett O’Hairdye is a burlesque performer, producer and artist. To learn more, visit her site at www.scarlettohairdye.com.

    This story originally appeared in LabLit in 2011.

    Follow us online:

    Leave a Reply

    Your email address will not be published. Required fields are marked *