All of earth’s creatures face similar survival challenges – having to procure food, social networks and mating partners – and research shows that this leads to many overlaps in the cognitive features of insects, other animals and humans. This makes it highly likely that this kind of error is far from being unique to bees. As Hunt and Chittka put it, ‘[S]ystematic memory errors may be widespread in animals … memory traces for various stimuli may “merge” such that features acquired in distinct bouts of training are combined in an animal’s mind, so that stimuli that have never been viewed before, but are a combination of the features presented in training, may be chosen during recall.’ Muddled memories are arguably the norm for all kinds of creatures.
For bees, or indeed any other animal, to have evolved a memory which contains such a capacity for error seems hard to believe. After all, being prone to make mistakes which are potentially detrimental to one’s survival would hardly be favoured by natural selection. The implication must be that the same systems which leave memory open to such mistakes also confer benefits which outweigh those potential disadvantages. To understand what those may be, we need to get a sense of the bigger picture, and to look in greater depth at the physiology of memory.
Plastic brains
Exactly how we are able to store an experience or thought in the brain has been a question that has fascinated us ever since we first began to entertain the idea that there might be no such thing as a spirit or soul (or that if there is, it might not be an extension of the brain). If that is the case, all information must be stored physically in the brain. This movement away from dualism (the belief that the mind and body are separate) to monism (the belief that all thinking originates in the brain) has led to a pervasive desire to understand the physical workings of the brain.
While the philosopher Descartes, a famous dualist, believed that the soul and body interacted through the pineal gland, a pea-sized structure located close to the centre of the brain, most scientists today agree that consciousness is not related to an incorporeal spirit, but rather is the result of a complex array of physical systems. Thanks to modern technology, including brain-imaging procedures such as fMRI (functional magnetic resonance imaging) and EEG (electroencephalography), we no longer need to study these systems by dissecting cadavers or examining case studies; for the first time in humanity’s existence we can now examine active brains while they are perceiving the world.
Our brains are incredibly malleable and adaptive. They are created for a world of uncertainty and quick decision-making, having had to survive in harsh environments since their inception. So, as our bee-searchers Hunt and Chittka put it,2 ‘The pervasiveness of such false memories generates a puzzle: in the face of selection pressure for accuracy of memory, how could such systematic failures have persisted over evolutionary time? It is possible that memory errors are an inevitable by-product of our adaptive memories.’ So, arguably, the bees muddling up which flowers actually provided them with nectar was a good thing, in the sense that the ability to muddle up memories is the by-product of a brain that can change, learn and reason. Occasional memory mistakes are but a small price to pay for that.
This adaptive quality of our brain is called neuronal plasticity, and it is only due to neuronal plasticity that we can have any memories at all. The cells in our brain, called neurons, connect with one another to develop meaningful networks, and these networks change in accordance with new experiences. If we were unable to incorporate new information into our existing neural networks, we would be unable to change our thinking or behaviour in light of new evidence, and we would be hard-pressed to deal with any alteration to our environment. It is also through this process that we are able to record and incorporate both the positive and negative experiences we have with others, and this can ultimately help us differentiate friend from foe.
Every time we have an experience we can potentially form a memory of it, a memory that exists in the brain as a network of neurons. This could be a semantic memory of a fact, such as that Obama was the president of the US in 2015. It could be an autobiographical memory of, say, the time you went to London to see a show. Or, it could be a memory of a decision-making process such as how you solved a puzzle. For any kind of experience to stick around in the form of a memory, it needs to form a physical representation in your brain.
Today we know a great deal more about how this happens because technological advances such as fMRI have allowed us to take photographs from inside of the brain, allowing us for the first time in human history to directly see what live memories look like. These advances have enabled researchers to study the biological and chemical mechanisms underlying memory processes, and to test purely physiological theories of memory formation. We now know far more about memory than even just a decade ago, and are able to map memories from their birth to their decay.
Memory stamping
The process through which an experience is laid down as a physical memory representation in the brain is known as biological stamping. In order to stamp new experiences into long-term memory, a biochemical synthesis is required to make connections between the existing neurons in our brain.
Our neurons have thin arms known as dendrites that allow them to physically stretch out to other cells; spines on them act as the communication centres between these arms. Within each individual neuron, messages move mostly as electrical impulses, but neurons mostly communicate with each other through chemicals passed via synapses. A synapse is a gap (or cleft) between two neurons. Synapses are transmitters and receivers; strong memories are largely the result of a continued easy flow of information from one cell to the next. This communication happens through chemical messengers called neurotransmitters that tell neurons what to do, most notably whether they should become more or less active. We could think of the neurons as two airports with planes (neurotransmitters) flying between them. Depending on the runways (receptors) available on the synapse they reach, some of these planes will be able to land and some will not. This controls the flow of information between neurons, making sure that we are not burning out our neurons when we are highly stimulated.
I remember one of my professors demonstrating the concept of synapses and their associated cells in a memorable, and adorable, way. He stood in the middle of the lecture hall filled with about 200 students and waited patiently until he had our attention.
‘I am a neuron,’ he proclaimed, matter-of-factly. He whipped out his arms to stand like the letter T. ‘These are my dendrites.’ Then he opened his hands, flexing his fingers, which had been fists until this point. ‘These are my spines.’ He called up another student and asked them to stand beside him in exactly the same pose. He brought his fingertips to his neighbour’s, creating a tiny space between them, ‘And these are my synapses.’ Finally, he took hold of his neighbour’s hand and shook it, representing the way impulses could travel through his neuron body into the body of his neighbour.
Our brain has approximately 86 billion neurons already in place, so recording a memory is largely an act of making and adjusting connections between existing brain cells, rather than making new ones. While all parts of neuronal connections can be modified, most researchers argue that it is primarily the synapses that are important for memory formation.
Long-term potentiation is the process through which connections between neurons are increased due to a strengthening of synapses. This strengthening happens because the neurons are strongly or repeatedly activated in relation to each other. For example, you might be on a beach in Spain for the first time in years, feeling really relaxed. This will activate the neurons in the ‘beach’ network, along with those in your neural networks for ‘Spain’ and ‘relaxed’. If an experience activates these connections strongly enough, or if similar experiences do so repeatedly, a long-lasting connection between these networks will be created; an associative memory linking the concepts of ‘Spain’, ‘beach’ and ‘relaxed’, for example.
One of the most prominent resear
chers in this area, who made fundamental strides in our understanding of what memories actually are from a biochemical perspective, is Michel Baudry.3 In 2011, he and his team at the University of Southern California published a review of over 25 years of their work, in which they essentially boiled the biochemistry of memory down to two things: a process called long-term potentiation and the influence of a class of substances called calpains, calcium-dependent proteases. Baudry and his team say that calcium is needed to stimulate the proteins in our brains that allow our synapses to undergo memory-related changes that can last. When a connection between neurons is repeatedly or strongly activated, like an association between memories (‘park’ and ‘trees’, say), calpains are activated at that exact location. The calpains then change the structure of our synapses, leading to a stronger connection between the activated memory cells in the brain. It seems only when calpains come for a visit that we see the transformation from simple experience to lasting memory.
Sea slugs and rat brains
Also studying this phenomenon is Eric Kandel. I have never met this amazing winner of the Nobel Prize in medicine, a pioneer in memory research. I have, however, research-stalked him for years, following his papers, textbooks, autobiography and interviews. And through this I feel as though I know him. Kandel started an obsession with sea slugs in 1962, and along with his colleagues and students at Columbia University in New York, he continues to do research on the sea slug Aplysia. Aplysia is a portmanteau of ancient Greek words meaning ‘sea’ and ‘hare’. Apparently they called them sea hares because of the little protrusions on their squishy heads which resemble the ears of a hare.
Kandel chose the slugs as research subjects because they use a simple system of neurons to remember their experiences and react to them. For example, if you pinch a sea slug’s gill in an experimental setting, it can learn to react by withdrawing it. The neurons involved can be isolated and extracted, and grow at an astonishing pace. Neurons can be kept alive away from their host brains in vitro if they are placed into a life-supporting oxygenated fluid.
Since the sole purpose of neurons is to make connections and form a brain, isolated neurons immediately begin to search for other neurons to network with. To do this they grow longer dendrites and more synapses. According to Kandel,4 ‘new growth of synapses occurs in front of your eyes over the course of a day’. This exceptionally fast growth, which is far faster than the growth of human neurons, makes the sea slugs ideal candidates for studying how memories are formed within and between individual cells. And because humans rely on almost all of the same neuronal processes as these invertebrates, the research has direct implications for human memory.
The sea slugs have taught us a tremendous number of things over the past few decades, and have contributed much to what we know about memory today. One of the most recent findings, detailed in a series of papers published by Kandel’s lab in 2015,5 6 is that one of the proteins responsible for long-term memory is unlike most other kinds of proteins – it is called a prion.
Prions, short for ‘proteinaceous infectious particles’, can change in shape, folding and reshaping in structurally distinct ways. Another of their notable properties is that they can either exist on their own, or they can form chains. These chains can automatically trigger neighbouring cells to join the chain and therein make a physical connection. Until their image update in 2015, many of us, if we had heard of them at all, would have associated prions only with negative diseases like Alzheimer’s and BSE (mad cow disease). The reputation of prions was so bad that Kandel pre-empted people’s potential bad reaction, saying ‘Do you think God created prions just to kill?’7 before going on to explain their crucial role in memory.
The primary role of prions in memory formation seems to be to stabilise the synapses that constitute long-term memories, thereby adding permanence to the physical changes that have already taken place due to long-term potentiation and the influx of calpains. Calpains are like the architects of the synapse, planning how communications should flow, while prions are the construction workers who make the changes more permanent.
But just because a connection is fixed now, this does not mean it will remain fixed forever. Calpains and prions can come back at any time and change things once more. In 2000, researchers Karim Nader, Glenn Schafe, and Joseph Le Doux8 at New York University examined the issue of memory fragments that change on a purely biochemical level, conducting an experiment where rats were played a particular tone and then given an electric shock. When they were subsequently played the same tone, they would freeze in fear, indicating that a memory associating the tone with a painful shock had been successfully generated.
Because learning to fear a particular situation or place is inherently linked with an emotional response, the rat’s fear memory was expected to be in the amygdala, a centrally located part of the brain which looks like two walnuts (one in each brain hemisphere) and which is largely responsible for emotion. In the next stage of the experiments, the researchers gave rats the same tone-followed-by-shock treatment, but injected a compound called anisomycin that inhibits the formation of proteins – proteins like calpains – directly into part of the amygdala after the rats received the shock. What they found was that these rats did not react with fear when they heard the tone again. In other words, the rats were unable to form new long-term memories of something that scared them because the compound had stopped the proteins in their brains from working properly, further emphasising the crucial role such proteins have in memory formation. This block has to be introduced quickly, however, as the biochemical process of stamping begins almost immediately during a learning exercise or personal experience.
That isn’t the whole story. After either 1, or 14 weeks, the rats who had formed an association were played the tone again to cause activation of the fear memory, but without the shock. After either 1 week or 14 weeks, the rats who had formed an association were played the tone again but not shocked to cause activation of the fear memory. However, if they were then injected with the protein block, they subsequently stopped reacting to the tone exactly as if the original memory formation had been interrupted. In other words, the memory had been destroyed. Any time the rats were stimulated to recall the fear memory, it could be interrupted by the researchers if they introduced the compound.
What is more, the rats only forgot the association if the compound was present during the memory recall. If they were given the memory-blocking drug anisomycin in isolation, without being played the tone to make them recall the memory of the shock, nothing happened. This indicates that the drug does not inherently just make any particular kind of memory dissolve. Rather, there seems to be an interaction between an activated memory in the brain and this drug that leads to the memory being erased. By introducing a protein synthesis inhibitor immediately after or during recall of a long-term memory, the reconsolidation of the memory is stopped; the memory stops being stored in the brain.
This leads us to one of today’s most en vogue biochemical theories of memory: retrieval-induced forgetting. This theory states that whenever we remember we also forget. So, while it seems intuitively appealing that every time we recall a memory we consolidate it and form a stronger and more accurate memory, this is far from the truth. Instead, every time a memory is recalled it is effectively retrieved, examined, and then recreated from scratch to be stored again. It is the equivalent of keeping a file of index cards, pulling one out to read it, throwing it away, and then copying out a new version on a fresh card for filing once more. And this is thought to happen every time we recall any memory.
In 2013 researchers Jason Chan and Jessica LaPaglia9 at Iowa State University explored this phenomenon in humans. They conducted a series of experiments demonstrating that every time an experience stored in long-term memory was recalled by participants this cycle of encoding and storage was repeated. They used no drugs in their studies, only interviewing. In one of these studies they showed participants a video of a fictional te
rrorist attack. They then asked them to recall what happened. After the participants recalled the memory they were then given misinformation – when they correctly recalled that the terrorist had used a hypodermic needle on a flight attendant, they were incorrectly told that the terrorist had used a stun gun.
When later asked to recall the event once more, participants recalled the incorrect information (the stun gun) and were unable to recall the actual event detail (the needle). The researchers claimed that this meant the new information had actually replaced the original memory. So if bad memory interviewing introduces inaccurate information, it can actually lead to a restructuring of the biochemical stamps of memories in the brain with non-medical procedures. This is how retrieval, if interrupted, can actually induce forgetting in a number of ways. It makes every event, every time it is recalled, physiologically vulnerable to distortion and forgetting.
Another way to inhibit memory in both rats and humans on a biochemical level is through a drug many people will already have heard of –commonly known by its trade name, Rohypnol.
Roofies
The notion of memory-altering drugs is a mainstay of our social consciousness. So-called ‘roofies’ in particular have taken centre stage, with their ability to have only little detrimental impact on our current state of mind, but temporarily destroying our ability to form new memories. Rohypnol – pharmaceutical name flunitrazepam – is a member of a class of drugs known as benzodiazepines. ‘Benzos’ are used recreationally due to their interactive effects with other drugs such as alcohol and heroin. In non-recreational settings, they are most commonly used for their anti-anxiety, anti-convulsant, muscle-relaxant and sleep-inducing properties – in emergency rooms benzos are commonly used to sedate patients. In criminal contexts, they are known as a type of date-rape drug.
The Memory Illusion Page 7
