The strange but true science of memory deletion

Memory isn’t written in stone: it’s more like a photocopy of a photocopy

When I decided to write a speculative fiction novel, I knew I wanted to keep the scope personal and the science plausible. I enjoy sci-fi epics with space empires and aliens as much as the next humanoid - but what I really love is real-feeling sci-fi, character-focused, and ideally set in a recognisable world. Our own lives, in short: with one crucial tweak. When I scroll through TV shows I’m looking for something that I won’t be able to resist pausing and annoying my partner by starting a debate (okay: monologue) about how I personally would react to this or that situation.

No, I would absolutely not sever my consciousness to create a work self and home self. But I might upload it into a digital paradise after my death. Would you?

My own novel, Tell Me An Ending, occupies the same ground: that of the ‘what if’ and ‘would you’. What if we could delete the memory of short periods of time? And then, once it was discovered that these memories weren’t actually gone forever but could be restored - would you want to know what it was you paid to forget?

I thought the science side of this would be easy; that I was in the realm of the thought experiment. The day I googled ‘memory removal’, I thought it was impossible to remove memories in real life. I was just checking which other fiction had covered the same ground so I didn’t accidentally plagiarise anyone.


What the search results actually showed was a huge amount of real-life research - apparently showing that not only is memory removal possible, it’s a lot closer than I’d realised. I called my academic friend and wept until she gave me her NCBI log-in details. When I searched the scientific journals for memory alteration, I found pages and pages of studies.

A big shift in how we understand memory happened relatively recently, with the research of Karim Nader in the late ‘90s. Ever since Plato, it was thought that a stored memory was a stable thing. Nader proved that it wasn’t: that the very act of recall renders the memory itself vulnerable to loss or change. To explain it very simply: every time we remember something, the memory is effectively rebuilt out of proteins in the brain. But if something happens to disrupt or colour the memory at the point, it will be different the next time we remember it. Memory isn’t written in stone: it’s more like a photocopy of a photocopy.

Nader and various other scientists have since demonstrated that if the formation of these memory proteins are blocked, or destroyed, the memory vanishes. Rats can be trained to fear a musical note, before an injection removes that fear. We don’t know how the rat feels about this experience, which is why the technique hasn’t yet been trialled in humans. But studies have successfully used anti-anxiety medication to remove the emotional component of patients’ traumatic memories, years after the original events. The patients were able to remember what had happened to them, but the intense fear and panic that accompanied the flashback had been stripped away.

I read through all of this not in awe of human achievement, but with the panic of a writer with a concept to protect. I was worried that by the time it was published, my book wouldn’t be speculative fiction but just... fiction.

Also, I couldn’t simply freewheel my way through some science-ish sounding explanations of the molecular structure of memory because this – unlike, say, aliens (why didn’t I write about aliens?) - was not uncharted territory. I ended up spending a long time on research. With the result that, excluding the development of my fictional company Nepenthe’s own tech, all the science referenced in the book is real. When a psychologist friend of mine read an early draft, he highlighted one of the studies as implausible. It is implausible: a doctor in the Seventies using electroshock treatment to remove his patients’ memories. It is also real.

Several readers and reviewers have called Tell Me An Ending ‘dystopian’, which is funny, because it might not be long before we’re living in this dystopia. But I don’t think there’s necessarily cause for alarm.

First of all, the manipulation of memory could offer a real chance of recovery to PTSD sufferers.

And secondly, while the idea of losing a paragraph from your life story might be disturbing, it might comfort you to know that your memories are mutating all the time: merging with other memories, losing parts, gaining new elements, disappearing altogether. The scientist Elizabeth Loftus showed how easy it is to manipulate the memory of eye witnesses, by introducing new suggestions at the point of recall. And a study carried out after 9/11 found that people’s memories of what they were doing when they heard about the attack changed significantly in the following months and years. These weren’t small changes either: some of the surveyed subjects remembered being in a different place altogether, or in the company of different people. Whether or not this comforts or disturbs you, it does show that memory has never been a reliable reference point.

However, there are potential downsides to memory removal. The tech itself is neither good nor bad. Just like nuclear fission, it’s an inherently neutral concept. But given how humans tend not to mind exploiting or harming other people for personal gain, you might wonder if the ability to erase memories would be good thing in the hands of repressive regimes - or large western corporations.

In Tell Me An Ending I tried to avoid moral judgements or simplistic answers when it came to the characters, and I’ve taken the same approach with the technology. There are a range of consequences in the book, topias running from ‘u’ to ‘dys’ and various states in between. As to whether any of these predictions will actually come true: I’ll be watching with interest.

Tell Me An Ending by Jo Harkin (Hutchinson Heinemann) is out now.