The mice were first placed in a tank of water and trained to find their way to a platform just below the surface. Next, the team ensured that the task was stored in the mice's long-term memory by waiting for a few weeks. Then they induced the mice to produce p25, leading to loss of neurons, learning ability and memory. When the mice were replaced in an environment rich of various stimuli the memories were restored.
The first interpretation is that the memories are stored in the ordinary sense but not to synaptic contacts but somewhere else, say to RNA inside cell nuclei known now to be coded in large quantities by the intronic portion of DNA (see the recent article in New Scientist). Stable storage of memories in the conventional sense of the word seems however to require that these RNA molecules remain in the nucleus which need not make sense.
TGD based model of long term memories for which zero energy ontology provides a justification, provides an alternative explanation. The basic ideas are following.
- Long term sensory or episodal memories making possible memory feats correspond to either sharing of mental images of the geometric past by time-like entanglement.
- This storage mechanism is not very efficient and a more efficient mechanism would be based on communication with geometric past. Memory recall would be represented as a signal sent from magnetic body at appropriate level of onionlike hierarchy of magnetic bodies to the brain of the geometric past and realized as phase conjugate negative energy dark photons. Memory would be communicated to the geometric future by using analogous positive energy signal. This time mirror mechanism is also the key mechanism of remote metabolism and mechanism of intentional action and explains the findings of Libet about strange time delays of consciousness.
- The problems due to the extremely low photon energies are circumvented if photons are dark and thus correspond to so large value of Planck constant that their energies are above thermal energy at room temperature. EEG represents only a small portion of frequencies involved and corresponds to time scale which is fraction of second. Much longer time scales are involved with what we are used to call long term memories.
In fact, almost any quantum dynamical event of the geometric past in the living body could serve as a memory storage. Nerve pulse patterns would serve only as symbolic "digital" representations of memories. This representation would be much more economic and than the representation as sensory memories localizable at the level of primary sensory organs.