Concepts as Symbols
Throughout our evolutionary history, it became crucial to go beyond what was perceptually available, for the sake of our survival. To do it, we needed to evoke the previous experiences we had interacting with the world, by default. Concepts helped us easily access these numerous experiences. Concepts can be defined as mental representations that help us make sense of the world. Through the process of generalization and discrimination learning, we grouped things that gave rise to similar consequences and represented them as concepts. When we talk about experiences, it is an event that consists of stimulus and response. Concepts are not an individual perceptual image, but an abstraction of numerous such experiences.
Concepts are represented as prototypes, which is an average of all those experiences that we have had. The most common experience of fruits that we have is that they are sweet to taste. Hence when we think of fruits, there is more probability that we might think of an apple or a banana rather than a tomato, which is also a fruit. This is because when we represent a concept as a prototype, we mentally integrate the most common features and omit not so common details. In this process of integration, the feature of a tomato, which is a combination of sour, salty, and savory gets canceled out. This is also the reason why penguins are hardly prototypes for the category of birds, except for those who may be living in the south pole!
Concepts are mental symbols that map on to the real world. They make complex objects and ideas from the world easy to use and understand. The interesting fact about these concepts is that they are results of reasoning. Computations are performed, integrating all the experiences that we have had. Concepts give you the awareness of all those experiences and we know what it means. Thus, every time we see a fruit, we do not have to process the stimuli and then infer that it is a fruit. Instead, we immediately know what it is, as the concept of fruit is a product of all those processes. Meanings and conclusions can be arrived upon efficiently with concepts. They represent events, situations, and classes of objects and ideas. Thus, they provide an easily accessible ‘compartment’ where stimuli from the world can be segregated, helping us perceive the chaotic world in an orderly fashion.
The stages of concept acquisition are observation, generalization, and abstraction. We initially start by making observations about the world. We learn that certain stimuli lead to similar outcomes and hence generalize them. Once the stimuli are grouped together, we need to derive a unified meaning from them. This is known as abstraction. It can be stated as knowledge derived from experience. When we talk about concept formation it is important to understand the role of statistical regularities in the world. We have evolved to find patterns in the environment. We are able to find out the probability of something occurring given that we know the probability of certain other things. This is crucial in making predictions about the world. Thus, with existing concepts, we can make predictions about novel entities and classify them. The ‘concept’ learning phenomena itself can be looked at as the brain’s predictive mechanism. Based on learned parameters of classification, the higher-order regions of the brain predict the class of the incoming stimuli. If there is a match between the predicted and the incoming signal from the sensory areas of the brain, the predictions are further strengthened. If there is a mismatch, a prediction error takes place. This error is the one that drives learning. It allows flexibility to modify our understanding of concepts and lets us generalize or discriminate. Based on the statistical regularities stored in the ‘concepts’ we predict that the red fruit in front of us would be as sweet as an apple. With the first bite into the tomato, we realize that it is not quite what we expected. This is a situation of a prediction error. This leaves room for the segregation of fruits as the ones that are sweet to taste and the ones that are not. This helps us realize that within the concept of fruits, there could exist sub-classes. Thus, a greater level of abstraction is required to represent concepts at the super-ordinate level (fruit) as compared to the sub-ordinate concepts (tomato). This is because as we go down the hierarchy of categories, the objects become more concrete and can be defined by a specific set of criteria.
When the predictive mechanisms of the brain either becomes lower or higher due to neural deficits, the process of concept formation goes awry. Children diagnosed with autism spectrum disorder, exhibit a lower predictive mechanism. In these children the sensory-based experience is dominant and thus they find it difficult to form generalizations. It is found that instead of concepts which are abstractions over numerous entities, these children store the sensory experiences themselves in their long term memory. To these children, the world seems unconnected as they perceive every stimulus to be a unique experience. On the other hand, those diagnosed with schizophrenia, have a higher than normal predictive mechanism. In such cases, the predictions generated by the brain override the sensory experiences and thus are unable to generate prediction errors and thus there exists no room for learning and modifying the predictions. This results in schizophrenic patients being delusional, where they live with false beliefs. For the normal population, there may be a lower predictive mechanism during the early phase of learning so that there can be higher attention allocated to errors. Once learned, the brain starts to generate internal models of the world, that is higher predictions resulting in efficient generalizations and categorization of the world.
There is another crucial mechanism of concept acquisition, which usually uses language. As kids, parents, and teachers taught us concepts, that were readily available to us without having to generalize and abstract ourselves. The linguistic representation of concepts is unique to human beings and has made communication easier. These concepts help us represent combinatorial thoughts, where we are able to combine multiple concepts through languages. They also facilitate causal inferences about the world.
The process of concept formation is quite similar to the formation of beliefs. Both take into consideration the frequency of occurrence of an entity. There exists a difference between the actual perception and the interpretation of the same in both concepts and beliefs. They are both results of abstraction over numerous experiences and are symbolic in nature. Given all of this, concepts can be regarded and equated to beliefs itself, that are used to make sense of the complex world that we live in!
