Relying on unreliable memories
Or why I love reading author's notes
I was recently hanging out with a group of old friends from school. Conversation randomly turned to 11 September 2001, when we were in Year 10. One friend recalled that we were called into a special assembly and sent home early. She remembered it distinctly: the feeling of walking home on a sunny afternoon, delighted to be out of school early, not quite registering the significance of the attack. But no-one else remembered this at all. In fact, the rest of us were absolutely certain that it was a normal school day: no special assembly, no early departure. I have an equally specific memory of getting the bus home at normal time and being mildly annoyed that the news was interrupting my usual post-school TV viewing. This kind of dispute comes up all the time between friends with a long-shared history - someone remembers it one way, someone else another way. But this stayed in my mind, because it made me think about my work. The basic fact of what reporters do is talking to people to find out what happened to them. But this means relying, to a large extent, on memories, and memory is fundamentally unreliable.
Of course there are ways of verifying facts - checking that someone’s account of events corroborates with other witnesses, or asking to see written records, like texts or emails sent at the time, or seeking out official or legal documents. The extent to which you might do this kind of fact-checking is situationally dependent. If you’re writing something investigative, or making legally contentious allegations, it can’t be based on individual, unverified stories. But what about reconstructing someone’s memories in order to create a “scene” in a long-form story? The aim there might not be to uncover an injustice or make an allegation of wrong-doing, but to illuminate someone’s character, or make the story more compelling and vivid for the reader. How do you check, in those instances, whether something actually happened? And what about the points where something is fundamentally unverifiable through documentation? A personal recollection of how someone reacted as a dramatic event unfolded; an emotional interaction between two people five years earlier?
When I started writing long-form narrative pieces in around 2015, lots of people recommended that I read Katherine Boo’s Behind the Beautiful Forevers: Life Death and Hope in a Mumbai Slum. It’s about South Asia, and at the time I was primarily writing about Pakistan, so it was instructive for me on a number of levels. The book is a masterwork of narrative nonfiction; the storytelling is extraordinarily cinematic and intimate. It reads like a novel, and I found myself wondering frequently “HOW could she possibly have known this detail?” Things like what someone was thinking or feeling in a particular moment, who stood where during a neighbourly dispute, or the exact smells and sights someone experienced during a conversation. I found some answers to these questions in Boo’s detailed author’s note at the end of the book. (This kickstarted my ongoing obsession with reading author’s notes in nonfiction books - they are often such a fascinating insight into people’s reporting processes and priorities).
Boo talks about the (literally) thousands of official records she searched through to verify and corroborate things where she could - no mean feat, given the nightmare of South Asian bureacracy. But I was particularly interested in the things that can’t be corroborated by documentation - feelings, interpersonal disputes. Here’s one part that really stood out to me:
“From that day in November 2007 that I walked into Annawadi and met Asha and Manju until March 2011, when I completed my reporting, I documented the experiences of residents with written notes, video recordings, audiotapes, and photographs. Several children of the slum, having mastered my Flip Video camera, also documented events recounted in this book.”
I found this fascinating and impressive (and it answered some of the questions I had while reading about how she could have known certain things), but as a toolbook it was almost entirely unhelpful to me. I can’t imagine ever having the time or resources to spend years researching a single story, to video record my subjects’ daily lives, or to give people cameras to record what’s going on when I am not around.
The two obvious ways to verify someone’s memories are to seek out documentation or corroborate their account with other people (so in the case of my friend remembering getting sent home from school early for 9/11, the fact that no-one else remembers it would discount her version of events - sorry pal). In an ideal world we would always be able to do this, but fundamentally this work is about dealing with people, and sometimes it’s messier than that. There are times when it might risk alienating the source to ask to speak to everyone else who was there at the time to see if they’re telling the truth - or when you simply might not logistically be able to do it, for reasons of time, money or practicality. (This type of fact-checking is time- and resource-intensive in a way that lots of people, especially in British journalism, don’t always have time for, but that’s a subject for another day.)
This came up while I was researching my book Karachi Vice: Life and Death in a Contested City, which tells the story of multiple urban conflicts through the lives of five ordinary people. (You can read more about it, incuding nice things reviewers said, here, or buy a copy here!) The research process involved speaking to those five people over the course of months. I was going back and forth between Karachi and London, so there were often long gaps between interviews, gaps in which I’d usually realise I needed more detail on a particular anecdote or memory to include it. I’d ask about it on the next trip, thinking that my interviewee would just fill in the missing details - but frequently, in relaying the story again, they’d fundamentally change key aspects. This isn’t unusual. Memories are not fixed - they shift and fracture, particularly when it comes to traumatic events.
One example I grappled with in particular was a crime reporter’s recollection of the night Karachi airport was besieged by Taliban terrorists. Sometimes he told me he was having dinner with three police officer friends when he heard the airport was under attack, and sometimes that he was alone at home. How could I check that? Asking him for the names and numbers of the three professional contacts he may or may not have had dinner with on a particular evening five years prior? I would feel weird if someone asked me to provide that information, and this was someone who was cagey about his movements and contacts at the best of times - with good reason, as he was on a terror group’s hitlist. At the time I was writing, I happened to be reading Patrick Radden Keefe’s brilliant book Say Nothing: A True Story of Murder and Memory in Northern Ireland, and found the author’s note (have I mentioned that I love author’s notes?) really helpful:
“Memory is a slippery thing, so I have sought wherever possible to establish corroboration for individual recollections. In instances where there are discrepancies among different accounts, I have used the most plausible version of events in the main text of the book and elaborated on alternative accounts, or other nuances, in the notes.
“This is not a history book, but a work of narrative-nonfiction.”
After agonising over this for weeks, I realised I could simply leave out aspects of the story that were really inconsistent or unverifiable. I definitely have a tendency to go down rabbit-holes on random details, but taking a step back, it doesn’t always matter that much to include every single piece of information or colour. And there’s a dual duty at play: a duty to the reader, to get as close to the truth as possible, and to the source - to be respectful, and mindful of the fact that no-one actually has to speak to a journalist. To my mind, writing carefully, truthfully and compassionately is always about balancing those two things.
Reading/listening
Jon Ronson has released a new series of his podcast Things Fell Apart, which explores the culture wars in nuanced, genuinely surprising ways. I listened to the entire series over a few days and thoroughly recommend it.
States across India are increasingly criminalising interreligious marriage, sometimes referred to as “love jihad”. This story, published by the Dial, follows one couple’s long legal battle to be together.
It’s from a few months ago, but this New Yorker story by Margaret Talbot is well worth your time. It’s about a woman revisiting her traumatic childhood in Austria to find out why she was kept in a mysterious “child observation station”.
Melissa Denes wrote an excellent piece for the Guardian this week, following a rape prosecutor through two trials. It’s a sobering look not only at the challenge of prosecuting sexual assault cases but Britain’s crumbling criminal justice system.
This New York Times Magazine story is a cautionary tale of dreadful journalistic ethics. It tells the story of a woman who sought help from true crime podcasters to investigate her sister’s murder, and ended up getting on their wrong side. I read it when it came out last month, and keep thinking about it.
Some news from me - I’ve been shortlisted for the Press Awards! (The full shortlist - full of brilliant journalists - is here. I am slightly bemused about why I’ve been nominated in the health journalism category, but I will gladly take it).
And finally, this Substack is still brand new! If you enjoyed it, please hit subscribe and share this post with anyone who you think might be interested. I really appreciate your support, and thank you for reading this far. I’ll be back in a couple of weeks.