How do thoughts bubble up into the stream that are spotlighted by the conscious discriminator? If one looks at life then it is apparent right from the start that competition and conflict are an inherent facet of its behaviour.
Each cell is unique and fights for the right to exist. Sometimes cooperation forms larger groups that can sacrifice individual members but generally it is an inbuilt character. Neurons have to fire to exist and well connected and successful ones can last a long time. Unsuccessful ones die and are spirited away by glial cells. This phenomena happens at all levels so the desire to get ‘picked’ as a winner drives colonies of neurons to get rewarded for choosing a winning strategy. Neurons recruit more like minded neighbours to pass their stimuli further up the chain. The saying “neurons that fire together, wire together” emphasises that dendrites thicken and multiply to strengthen their connections and vitality. This causes more and more neural mass to accumulate similar aims and this bubbles up the pipe. Many areas such as the hippocampus use ‘rehearsal’ techniques to enrich and update qualia. The act of getting chosen is the ultimate aim and competition is fierce.
Beginning with Karl Lashley, researchers and psychologists have been searching for the engram, which is the physical trace of memory. Lashley did not find the engram, but he did suggest that memories are distributed throughout the entire brain rather than stored in one specific area. Now we know that three brain areas do play significant roles in the processing and storage of different types of memories: cerebellum, hippocampus, and amygdala. The cerebellum’s job is to process procedural memories; the hippocampus is where new memories are encoded; the amygdala helps determine what memories to store, and it plays a part in determining where the memories are stored based on whether we have a strong or weak emotional response to the event. Strong emotional experiences can trigger the release of neurotransmitters, as well as hormones, which strengthen memory, so that memory for an emotional event is usually stronger than memory for a non-emotional event. This is shown by what is known as the flashbulb memory phenomenon: our ability to remember significant life events. However, our memory for life events - autobiographical memory - is not always accurate.
The meme is an evolutionary replicator, defined as information copied from person to person by imitation. I suggest that taking memes into account may provide a better understanding of human evolution in the following way. Memes appeared in human evolution when our ancestors became capable of imitation. From this time on two replicators, memes and genes, coevolved. Successful memes changed the selective environment, favouring genes for the ability to copy them. I have called this process memetic drive. Meme-gene coevolution produced a big brain that is especially good at copying certain kinds of memes. This is an example of the more general process in which a replicator and its replication machinery evolve together. The human brain has been designed not just for the benefit of human genes, but for the replication of memes. It is a selective imitation device.
When qualia are created they also carry the environment of their creation. Those born in a dangerous or critical situation have been saturated with hormones and neurotransmitters that flag them as definitely worth taking extreme notice. This ‘choosing’ of a thought reflects the mood of the person at any one time. In reflective mood, a relaxing thought is picked if possible; in high alert an action similar meme can be combined with the actual threat. This not only hyperdrives the defences but provides history to combat the threat.
A potential mechanism for qualia involves chaining together memories for discrete events at the time of choice. In this scenario, an inferred outcome is predicted by internally simulating the short-term consequences of each memory in the chain. Retrieval mechanisms of this kind may be described by a family of theories known as model-based reinforcement learning that involve a learned model of the world. By constructing predictions for decision outcomes on the fly, such mechanisms capture a hallmark of flexible decision-making. However, this comes with the computational cost of searching through a potentially large number of memories.
To reduce the computational demand associated with qualia, events that have not been encountered together in space or time may be linked to form cognitive “short-cuts.” Together with prior memories, such higher-order relationships may form a “relational” or “cognitive map” of the world. One possibility is that memories for distinct experiences are linked together or even fundamentally restructured during awake rest and sleep. During these quiet periods, hippocampal local-field potentials (LFPs) are characterized by sharp-wave/ripples (SWRs): short-lived, large-amplitude deflections accompanied by high-frequency oscillations. During SWRs, hippocampal cells fire synchronously and their temporally structured spiking can “replay” previous waking experience to support memory and planning. Growing evidence suggests SWR activity also extends beyond replay of directly experienced information. For instance, hippocampal SWR spiking can anticipate upcoming experience reorder events according to a trained rule or even stitch together spatial trajectories In this manner, we hypothesize that hippocampal SWR activity generates spiking motifs that provide a cellular basis for novel higher-order relationships, thus breaking the constraints imposed by direct experience.
During qualia, the hippocampus engages a prospective code that preserves the learned temporal statistics of the task. In addition, during rest/sleep in mice, hippocampal SWRs show increased co-activation of neurons representing inferred relationships that include reward. During rest/sleep the hippocampus appears to “join-the-dots” between discrete items that may be profitable. The mechanism provides a means to build a cognitive map that stretches beyond direct experience, creating new knowledge to facilitate future decisions. This process of “joining-the-dots” between logically related events is consistent with evidence that SWR spiking is not only determined by prior experience, rather the intrinsic connectivity of the hippocampus , self-generated sequences, forward planning, structural knowledge and stitching together of spatial trajectories all play a role. Hippocampal SWR spiking represents a non-spatial, second-order mnemonic link between items not experienced together, over and above simulating an internal model that draws on direct experience. The increase in SWRs nesting inferred relationships suggests hippocampal spiking activity during SWRs may build higher-order relationships to integrate knowledge into a coherent schema. This new understanding of hippocampal SWRs may explain why sleep/rest facilitates behavioral readouts of insight and inferential reasoning in humans. Consistent with studies showing that reward-related activity influences the spatial content of SWRs, findings suggest SWR content can be skewed toward events that are more salient, have greater future utility, and/or generate larger reward-prediction errors. Changes in neuronal co-activation in hippocampal SWRs are suitable to influence wide-spread cortical and subcortical targets, directly or via intermediate relay regions. Specifically, hippocampal SWR spiking may broadcast value information to relate reward information received at the end of a sequence to earlier events. Reverse replay in awake SWRs occurs during reward-motivated spatial behavior), while data show an inverted temporal order in non-spatial inferred relationships. SWR-nested spiking may therefore facilitate retrospective credit assignment or value updating of sensory cues represented by the medial prefrontal cortex (mPFC) and midbrain, even if those cues are not directly paired with an outcome. Such cross-region coordination may explain why functional coupling observed between hippocampus and mPFC during post-encoding rest predicts measures of memory integration in humans. In this manner, SWR-related hippocampal training signals may alleviate the computational cost of qualia by building a model or “cognitive map” of the external world that spans multiple brain regions.
In addition to this SWR-related mechanism during rest/sleep, dorsal hippocampus first region pyramidal neurons (dCA1) are necessary for inferential choice. During qualia inception, the hippocampus represents a veridical copy of learned associations in temporal sequence . These findings were not explained by mere spatial location, yet these temporally structured mnemonic associations may be analogous to spatial sequences of place cells Sequential firing of this kind may be a necessary requirement for a brain region evolved to support memory. Previous studies suggest that during learning, memories for past overlapping events can be evoked and associated with newly encountered information to link memories across experiences. This integrative encoding may even assign value to stimuli not directly paired with an outcome , alleviating the need to recall intermediary cues at the time of choice.
There is a high possibility of hippocampal mechanism where mnemonic sequences are recalled “on-the-fly.” This mechanism may further depend upon extra-hippocampal regions representing the relevant sensory cues. The hippocampus may draw on learned experience, while other downstream circuits may use the hippocampal output to reinstate an integrated or overlapping neural code.There a division of mnemonic labor between the hippocampus on the one hand, and the mPFC and (putative dopaminergic) midbrain on the other: whereas the hippocampus draws on learned sequences , the hypothetical inferred outcome , rewarding or neutral, is represented in the mPFC and midbrain, potentially inherited by integrative encoding or spiking activity during SWRs. qualia therefore involves a memory recall mechanism that spans multiple brain regions. This differs from computational models that propose associative information is integrated locally within the medial temporal lobe via recurrent loops but is consistent with evidence showing representation of intermediary cues in the medial temporal lobe at the time of choice . Also there is support evidence suggesting the mPFC uses an abstract model of the environment to guide behaviour while the midbrain supports learning of relationships that extend beyond those associated with direct reinforcement. Retaining both veridical mnemonic recall and allowing qualia for higher-order relationships provides the comprehensive cognitive flexibility necessary for adaptive mammalian behaviour in an ever-changing environment.
Copyright © 2020 deepermind.co.uk | Maldwyn Palmer