The Hypothesis Dump

Hypothesis

(This article is not peer-reviewed)

Our memories can be highly unreliable. We easily forget and false memories can form very rapidly or be complete fabrications. The main hypothesis describing the possible evolutionary mechanism, giving rise to this effect is: Memory formation is a highly energy draining function and only memories that increase fitness might be prioritised by the nervous system. This hypothesis is likely, but offers inadequate explanation as to why memories can be so inaccurate and unreliable. Here I propose a hypothesis that offers an explanation for the phenomenon. There is a trade-off in nature between stable/generalised memories and unstable/adaptable memories that increase cognitive plasticity. The mechanism behind this, is that unstable engrams can be adapted to be incorporated into novel networks, forming novel ideas/memories in different contexts. While, stable/rigid engrams are less likely to adapt and be incorporated, instead, they are generalised across contexts for specific evolutionary purposes.

The Observation

The nervous systems energy efficiency is linked to the fitness of the individual. The reasoning is that memory is a highly energy draining function, as seen in many higher functions (Rae et al., 2024). However, this does not fully explain the phenomenon of inaccuracy and the dynamic aspect of memory. For example, false memories do not objectively represent actual events (Fandakova and Dennis, 2024). Instead, they have been changed significantly by outside factors, so much so that they can be completely different from how they initially started. They could also be complete fabrications that never started as a representation of an environmental phenomenon (Arce et al., 2023). This means that forgetting something is not only losing the memory, but that it might change into something different. Would it not be more evolutionarily beneficial for memories to be more stable?

Not if there are some other factors missing in this picture. There might be some evolutionary benefit to forgetfulness and memory instability.

The Premise

The nervous system does not only need to process the environment or react to stimuli, it also needs to adapt, modify and recreate its responses, to be able to predict the most beneficial responses to an environment that is highly dynamic. It needs to be as dynamic as its environment. It is not beneficial for the system to be rigid; instead, it needs be plastic so that the organism can easily adapt its responses to changes in the environment, to ensure high fitness. That means, that networks need to shift functionality or engrams need to rewire to adapt, modify or re-create novel responses. This might be why they are so dynamic and fragile. I believe that the older an engram is, the higher the risk is that it represents a false memory. However, not all engrams are equal. There might be a spectrum of stability depending on some factors. For instance, it has been proven that memories with higher emotional salience tend to be more easily retained (Tyng et al., 2017). The other factor is likely repetition which is known to affect consolidation. Again, this solidifies the hypothesis regarding energy efficiency. However, it seems there is an aspect which is often neglected, the generalisation problem. There must have been selection pressure for memories to be generalised and rigid. If an animal is to predict the most beneficial response to a recurrent stimulus, it needs to have an accurate record of what previously occurred, otherwise the system will be more susceptible to mistakes. However, it also should generalise the reactivity of the rigid engram in order to increase fitness. In other words, anything in the environment that looks like the fear-inducing object, should elicit the same response, as this increases fitness over being highly adaptable.

Unstable memories are, according to this hypothesis, adaptable and susceptible to change depending on the context. Factors that would predict where a memory falls on this spectrum, is biological relevance indicated by emotional salience, and repetition (Tyng et al., 2017). These memories are unstable and can morph into and integrate with other networks, forming novel engrams. That means that they are re-wired and re-purposed to serve a new role in another network. These would be adaptations dependent on environmental changes, cues and context. This in turn means that there are possibilities for the networks to produce novel outputs because they are highly dynamic and can be re-structured. Could this be a factor in adaptation to environmental changes?

This is a factor in cognitive plasticity, because it allows for cognitive processes to be dynamic and adaptable as a reflection of the network. The instability facilitates plasticity, while stability facilitates rigidity and generalisation because it is more hardwired for a specific purpose.

Evidence thus far suggests that synapses are highly plastic, and that engrams are highly susceptible to change through Hebbian processes (Turrigiano, 2017). Whenever an engram is recalled, the concurrent context is associated with it. However, depending on the stability of that said engram, it will determine how well the engram can rewire and integrate according to this hypothesis.

Take for instance hippocampal place-cells, they are hippocampal neurons that fire in a specific location in a particular experimental arena. However, these cells can represent one place in one context and another if you change the environmental context
(Leutgeb et al., 2005). They are highly dynamic and it is essential for adapting the behaviour to the different contexts. This repurposing and rewiring of networks is a requirement, if the network output is to adapt to changes in the environment. If the cells were generalising their response to one particular position in all rooms, the animal would be less able to adapt. Generalisation is an aspect of fear memories for example (Asok et al., 2019). It is an evolutionary benefit to respond with escape to a grass field anomaly and be mistaken, rather than mistaking an actual predator for a grass field anomaly and be eaten. This aligns with the premise of this hypothesis, that rigid engrams are less prone to contextual adaptations and more prone to generalisations due to their stability.

The Predictions

For animals, there would likely be a distribution of plasticity that is dependent on where the memory-type falls on the spectrum of stability. That in turn would be predicted by emotional salience of the memory type. For example, engrams relating to what a predator is, are highly stable and not plastic. Each species of animal would in turn have differential environmental and salient cues that categorise those stimuli, dependent on biological relevance.

Furthermore, there would be a distribution of general cognitive plasticity predicted by the ability to form stable memories. For example, an individual’s general ability to consolidate memories will predict overall cognitive plasticity. This would result in individuals with adept memory being less likely to handle or adapt to environmental changes.

Based on the above arguments, it means that in nature, there is an evolutionary push-pull between cognitive plasticity and accurate memory. I would predict that animals who use active hunting as a strategy would fall more on the plasticity side of the push-pull, compared to the animals that use ambush. It may be that active hunting requires higher and faster adaptive capability than ambush. Cognitive strategies that deal with different types of diet could also predict the plasticity requirements of that strategy. For example, if your diet is mobile animals, you will need to adapt rapidly while planning, predicting and responding during an active hunt. Meanwhile, if your diet is immobile vegetation, it’s likely you mainly need to remember where it is accurately. Furthermore, if you are escaping, what mostly is happening is reflexive and innate responses. The adaptation is only necessary to react to the environment to stay mobile so as to not get caught. However, this is a simplification as escape behaviours also might require high adaptability.

In humans, we can observe some interesting findings that could point towards this mechanism. Individuals with Attention deficit disorder (ADHD) may have a decreased expression of brain-derived neurotrophic factor (BDNF), while children with autism have an increased expression of BDNF (Galvez-Contreras et al., 2017). Taken together with the trait of adaptability to changes in the environment, comparing Autism with ADHD symptoms, we find a compelling parallel to this hypothesis. Individuals on the autism spectrum have a higher difficulty dealing with environmental changes while individuals with ADHD are more prone to change their environment (Lage et al., 2024; Le Cunff, 2024). In addition to this, individuals on the Autism spectrum tend to be detail focused rather than globally focused, meaning they easily remember specific words in a text compared to the meaning and concept (Happé and Frith, 2006). I believe, that is because the stable engrams formed are less likely to adapt and integrate, with the network of engrams that represent context and meaning, while the more unstable engrams in the average individuals are not as rigid and can integrate. This could be why you forget the specific words, since those engrams do not exist anymore and have integrated, but you remember the context and meaning. While individuals with stable but rigid engrams remember the words, they have a harder time to understand the context and meaning.

The Gap

Testing this hypothesis, requires the investigation of predicted effects. The engram adaptability and behavioural effects on the individual. For instance, by investigating the effects of differential expression of BDNF in the neurodivergent individuals. Specifically , can we see that there is a correlation between mnemonic consolidation and contextual adaptation that is predicted by BDNF expression levels. However, this is not enough as we would need complementary intervention studies on animal models, to see if manipulation of BDNF expression in mice can induce similar behavioural effects.

Declaration of AI use: The author used Lumo AI for spelling, grammar-checking and, and literature search assistance during the preparation of this article.

Disclaimer: This hypothesis peicehas not undergone formal peer review and is written you be founder of the website. They represent the author’s perspectives and analyses of published scientific literature. Content should not be interpreted as medical advice or established scientific consensus.

References

Arce, R., Selaya, A., Sanmarco, J., Fariña, F., 2023. Implanting rich autobiographical false memories: Meta–analysis for forensic practice and judicial judgment making. International Journal of Clinical and Health Psychology 23, 100386. https://doi.org/10.1016/j.ijchp.2023.100386

Asok, A., Kandel, E.R., Rayman, J.B., 2019. The Neurobiology of Fear Generalization. Front. Behav. Neurosci. 12, 329. https://doi.org/10.3389/fnbeh.2018.00329

Fandakova, Y., Dennis, N.A., 2024. Introduction to the special issue: the neuroscience of false memory. Memory 32, 1267–1270. https://doi.org/10.1080/09658211.2024.2418768

Galvez-Contreras, A.Y., Campos-Ordonez, T., Gonzalez-Castaneda, R.E., Gonzalez-Perez, O., 2017. Alterations of Growth Factors in Autism and Attention-Deficit/Hyperactivity Disorder. Front. Psychiatry 8, 126. https://doi.org/10.3389/fpsyt.2017.00126

Happé, F., Frith, U., 2006. The Weak Coherence Account: Detail-focused Cognitive Style in Autism Spectrum Disorders. J Autism Dev Disord 36, 5–25. https://doi.org/10.1007/s10803-005-0039-0

Lage, C., Smith, E.S., Lawson, R.P., 2024. A meta-analysis of cognitive flexibility in autism spectrum disorder. Neuroscience & Biobehavioral Reviews 157, 105511. https://doi.org/10.1016/j.neubiorev.2023.105511

Le Cunff, A.-L., 2024. Distractibility and Impulsivity in ADHD as an Evolutionary Mismatch of High Trait Curiosity. Evolutionary Psychological Science 10, 282–297. https://doi.org/10.1007/s40806-024-00400-8

Leutgeb, J.K., Leutgeb, S., Treves, A., Meyer, R., Barnes, C.A., McNaughton, B.L., Moser, M.-B., Moser, E.I., 2005. Progressive Transformation of Hippocampal Neuronal Representations in “Morphed” Environments. Neuron 48, 345–358. https://doi.org/10.1016/j.neuron.2005.09.007

Rae, C.D., Baur, J.A., Borges, K., Dienel, G., Díaz‐García, C.M., Douglass, S.R., Drew, K., Duarte, J.M.N., Duran, J., Kann, O., Kristian, T., Lee‐Liu, D., Lindquist, B.E., McNay, E.C., Robinson, M.B., Rothman, D.L., Rowlands, B.D., Ryan, T.A., Scafidi, J., Scafidi, S., Shuttleworth, C.W., Swanson, R.A., Uruk, G., Vardjan, N., Zorec, R., McKenna, M.C., 2024. Brain energy metabolism: A roadmap for future research. Journal of Neurochemistry 168, 910–954. https://doi.org/10.1111/jnc.16032

Turrigiano, G.G., 2017. The dialectic of Hebb and homeostasis. Philos Trans R Soc Lond B Biol Sci 372, 20160258. https://doi.org/10.1098/rstb.2016.0258

Tyng, C.M., Amin, H.U., Saad, M.N.M., Malik, A.S., 2017. The Influences of Emotion on Learning and Memory. Front. Psychol. 8, 1454. https://doi.org/10.3389/fpsyg.2017.01454