To a neuroscientist, said Hugo Spiers, a psychologist from UCL chairing a meeting at LSE last week, memory is just a physical and chemical arrangement of synapses. That’s a supremely reductionist view, the view of a NeuroNazi, said Sebastian Groes, a professor of literature from Roehampton University. Although chaired by the neuroscientist, it was the views of the literature professor, a futurologist, a science fiction writer, and a philosopher that dominated the meeting, but they couldn’t even agree on what memory is.

The NeuroNazi redeemed himself by quoting the writer Michael Crichton on the centrality of memory to human experience: “In a sense all we consist of is memories. Our personalities are constructed from memories, our lives are organised around memories, our cultures are built upon the foundation of shared memories that we call history and science.”

Memory to the neuroscientist is not one thing but several: episodic memory (of what I had for breakfast and the last girl I kissed), semantic memory (of general knowledge and how to ride a bicycle and insert an endotracheal tube), and short and long term memory. One might work well when another has faded.

Groes, who has edited a book on memory that will cost £67, talked of the “complexification” of memory, how it has moved beyond the brain. For him memory can be in machines: the photo of my car number plate in my iPhone is part of my memory, so my memory is hugely expanded. It can even be said to include most of human knowledge as I can access it through my iPhone. The professor believes too in embodied memory, the way that a concert pianist remembers a Schubert sonata without having to use her mind: the memory is in her hands.

We are all used to the idea that our computers and phones have memory, but Jessica Bland, a futurologist (well, that’s what I call her) with Nesta, described how machines may be developing memories of their own. With the Deep Mind project three entrepreneurs have created a machine that learns to play computer games given only the most rudimentary information on the rules of the game. The machine starts badly but learns, presumably memorising what it has learned. Deep Mind has been sold to Google for hundreds of millions of dollars, sparking assumptions that it must be something very significant, perhaps even a machine with consciousness that will quickly become superior in every way to the mortal faulty machines called humans. “Killer robots” will take over. (I’ve blogged on this before, speculating that once the robots take over they will switch themselves off as unlike us with our urge to reproduce they will have no reason to continue.)

The science fiction writer Adam Roberts described how science fiction writers have been fascinated by the idea of the extended mind. H G Wells imagined “the man of the year million,” who had a huge head but no arms or legs as they were no longer necessary. The man was the prototype of the Martians who invaded Earth in the War of the Worlds.

Roberts himself proposed that another way to remember was through dreams. Events that had and hadn’t happened to you both feature in dreams, and he imagined a shop where you might buy dreams. Presumably, he thought, you would buy exciting, vivid dreams, and, as dreams have a function of clearing your mind, you might wake from such a first class dream not only revitalised but rejuvenated.

Barry Smith, a professor of philosophy at Birkbeck, played the traditional role of the philosopher and declared that none of these memories in iPhones, machines, and dreams, embodied memories, or collections of synapses were really memories. A memory had, he said, two components: an event that had happened where the person remembering was present; and a memory of that event by the person remembering.

A false memory is not a memory because the person with the false memory was not present at the event that he or she is remembering. So Oliver Sachs’s memory of being at a bomb explosion when he was not there (but was remembering his brother’s vivid description), is not a memory to Smith even though it felt exactly the same as true memories to Sachs. Memories can, course, be faulty and embellished, but for Smith are still memories because they meet his criteria. With the picture in the iPhone the memory is not what is in the phone but rather the memory in the mind that there is something in the phone.

(Smith threw into the discussion the intriguing observation reported to him by a neuroscientist that the experience of dreaming may be the closest normal people can come to knowing what it’s like to be brain damaged. Thoughts [perhaps not the right word] and images are disordered, and you are essentially observing, unable to influence events.)

But what if the memory was implanted not in the iPhone but in the mind? Roberts talked of a Philip K Dick story called We Can Remember it For You Wholesale, in which people could buy memories. Spiers described how scientists had managed to implant into rats’ brains skills that they hadn’t learnt. Smith was not impressed: implanted memories did not meet his criterion of the person (or rat) remembering having been present at the event remembered.

Smith, who has written about Marcel Proust (surely the supreme poet of memory), was, however, impressed by memories evoked by smells. For Proust the taste (close to smell) and touch of a Madeleine is the starting point of his huge novel, Remembrances of Things Past. “No sooner had the warm liquid mixed with the crumbs touched my palate than a shudder ran through me and I stopped, intent upon the extraordinary thing that was happening to me.” For Smith smells can evoke unadulterated memories, these not adulterated by having to be described in words. Smell can bring back the event as if you were there again, as happened to Proust.

It is, of course, profoundly human to forget. The meeting heard of people who supposedly have Superior Autobiographical Memory (or hyperthymesia), who can remember everything from every day of their lives. For at least one of these people the condition was unbearable, with the past overwhelming the present. I remember, or at least think I do, reading a short story about a person who remembered everything. For him too life was impossible: there was no pleasure in listening to a Mozart concerto because he remembered all of it; fewer and fewer of life’s experiences were new, so those who cannot forget face a similar problem to the immortal, appalling crushing boredom.

Where was I left at the end of what at times was a confusing discussion? I recognised the “complexification” of memory but felt more comfortable with Smith’s pragmatic view. I reflected on my mother, who has had no short term memory for eight years. She is undoubtedly diminished from what she was, but she is, like a 60s hippy, living in an eternal present. Most of her memories are gone, but she is still very much her. We are perhaps more than our memories.

Richard Smith was the editor of The BMJ until 2004. He is now chair of the board of trustees of icddr,b [formerly International Centre for Diarrhoeal Disease Research, Bangladesh], and chair of the board of Patients Know Best. He is also a trustee of C3 Collaborating for Health.