And indeed, some hours or days later, the word which we failed to recollect, “recurs to our memory” by itself. Most often, this happens to be a name or a place-name. A similar case was correctly described by Anton Chekhov in his story “A Surname Associated with a Horse”. Such state is called “memory blocking” by psychologists, and a rather precise definition of it exists in the English language: “tip-of-the-tongue state”.
The researchers from the Institute of Cognitive Neurology of Contemporary Humanitarian Academy thought about the reasons for such spontaneous memory blocking and tried to scrutinize it in objective ways. Having realized the reasons for and mechanisms of this defect, it will be possible to find the way of fighting against it, as failure of recall something at the most necessary point sometimes complicates life significantly. And the emerging informatization epoch sets forward higher requirements to memory work. Besides, the share of people of advanced age is increasing in the society, and their memory suffers to a greater extent.
Observations of the patients can bring us to the conclusion that especially categories of words pass out of mind in case of injuries of some areas of the brain. Thus, in case of injuries to the left temporal lobe of the cerebral cortex the patients are unable to recall people’s names. Indeed, as functional magnetic resonance brain scanning shows, when recalling names it is the left temporal lobe that is active.
Psychological testing also allows to reveal the difference between recognition when there is realisation that the word (image) is familiar, and reproduction when the word is fully recollected. For example, if the probationers are shown several hundreds and even thousands of pictures and then are presented one more, the probationers nearly always responded correctly to the question if the newly presented picture had been among the earlier shown ones or not. However, they could more or less correctly describe only a small part of the seen pictures. It may be said that the volume of the “memory of recognition” significantly exceeds the volume of the “memory of recollection”.
Neurophysiologists assume that different regions of the brain are responsible for these two processes: the medial parietal region is to a larger extent connected with recognition, but the hippocampus – with recollection.
Another approach to investigating the brain is computer modeling of its work with the help of neural networks. To investigate the memory blocking phenomenon, Anatoly Terekhin and his colleagues created a neuronet model of memorization and recall processes.
One of the neural network variants, the so-called Hopfield network, consists of some quantity of neurons that are connected with each other by synaptic bonds and at each moment of time they change their state according to certain rules.
The researchers built the recognition model based on such network, having introduced a special variable they called “familiarity”, and having included in the network an additional “recognition neuron”. Without going into detail which can be precisely described only in the language of mathematics, it can be stated that upon a familiar input signal, the recognition neuron comes into one state, but upon an unknown input – into another state.
To determine abilities of the model network, it should be tested. When testing networks with different number of neurons, the researchers have determined that the network capacity for image recognition is by dozens of times higher than that for image reproduction. “The capacity for reproduction is proportionate to the number of neurons in the network, but the capacity for recognition is proportionate to square number, explains Anatoly Terekhin. For example, the network of 700 neurons can reproduce 100 images and recognize – 9,000. Besides, image recognition is performed in two steps of net work, but reproduction as a rule requires more steps – three to six for undistorted familiar images and even more steps – for unknown or heavily distorted images.”
This is where the researchers see the reason for the observed paradoxical contradiction between the person’s confidence in the word’s familiarity and complete inability to reproduce it. Simply speaking, the brain’s neuron network potential is different for recognition and reproduction– the brain recognizes easily but reproduces with much more difficulty.
The authors assume that the brain’s phenomenal ability for recognition is its reserve, which will help to cope with increasing information flow, and they believe that people in the future (when looking for some object) will increasingly head for its recognition but not for precise reproduction. Especially because detailed information will always be easy to find quickly in the Internet.
Nadezda Markina | alfa
Amazingly flexible: Learning to read in your thirties profoundly transforms the brain
26.05.2017 | Max-Planck-Institut für Kognitions- und Neurowissenschaften
Fixating on faces
26.01.2017 | California Institute of Technology
16.08.2017 | Event News
04.08.2017 | Event News
26.07.2017 | Event News
21.08.2017 | Materials Sciences
21.08.2017 | Health and Medicine
21.08.2017 | Materials Sciences