cognition(redirected from Cognitive function)
Also found in: Dictionary, Thesaurus, Medical, Legal, Acronyms.
The internal structures and processes that are involved in the acquisition and use of knowledge, including sensation, perception, attention, learning, memory, language, thinking, and reasoning. Cognitive scientists propose and test theories about the functional components of cognition based on observations of an organism's external behavior in specific situations.
Cognition throughout life can be broadly described as an interaction between knowledge-driven processes and sensory processes; and between controlled processes and automatic processes. Over time, there is a trade-off between the amount of surface information that is retained in the internal representation of objects or events (bottom-up processing) and the amount of meaning that is incorporated (top-down processing). Following exposure to a stimulus, a sensory representation (sometimes called an image, icon, or echo) is constructed that encodes nearly all the surface characteristics of the stimulus (for example, color, shape, location, pitch, and loudness). The information is short lived, lasting less than a second. Much evidence suggests that extraction of information from this representation takes place in two stages, a feature analysis stage and an object recognition stage. It is during the latter stage that attention (controlled processing) and previous knowledge come into play. See Memory, Perception
Conceptual knowledge is needed to classify objects and events in the world. Some aspects of conceptual knowledge are innate or emerge very early in development, while others are acquired through learning and inference.
A primary cognitive function of all social species is communication, which can be accomplished by a combination of vocal, gestural, and even hormonal signals. Of all species on Earth, only humans have developed a communication system based on abstract signs. This evolutionary development is closely tied to the greater reasoning capacity of humans as well. All reasoning can be broadly described as pattern recognition and search. Conceptual knowledge base are searched for relevant information in order to draw a conclusion, solve a problem, or guide behavior. Thinking often takes the form of a chain of associations among concepts in long-term memory, with one thought retrieving others to which it is related. The most common reasoning strategies include direct retrieval, imaging, means-ends analysis, analogy, classification, deduction, and formal procedures.
Reasoning by direct retrieval involves retrieving a known fact from memory to solve a problem. Reasoning imagistically involves constructing or retrieving images from conceptual memory and examining or manipulating them to solve a problem. For example, individuals reason imagistically when they determine how many windows there are in their living rooms by retrieving an image of the room and counting the windows in the image.
Means-ends analysis is typically employed when solving problems in unfamiliar domains. When a solution is not immediately apparent, reasoners typically compare the goal to the current situation and select means with which to reduce the differences between the two situations.
The restructuring of a problem representation that allows an available means to be used in a novel way or a seemingly unrelated bit of knowledge to be accessed to solve the probem is called insight.
Reasoning by analogy is used when a current situation allows an individual to recall another, similar situation that has a known solution or other information relevant to the task at hand. It is a technique that is powerful but error prone.
Reasoning by classification involves making inferences about an object or event based on its category membership.
Deductive reasoning involves drawing a conclusion based on its logical relation to one or more premises. A second common use for deduction is testing hypotheses.
Formal procedures for reasoning and for solving problems include logic, mathematics, probability theory and statistics, and scientific investigation. Understanding of the behavior and properties of physical, biological, and cognitive systems has been greatly enhanced through the use of these techniques. See Psycholinguistics
By using noninvasive techniques such as positron emission tomography (PET scan), magnetic resonance imaging, electrical skin conductance, invasive surgical and chemical investigations of animal brains, and data from clinically observed syndromes associated with brain injury, cognitive neuroscientists have pieced together information concerning the role that specific brain regions play in the processing of emotional and cognitive events. High-level visual processing, such as object recognition, takes place in the occipital lobes of the cortex, although recognition of certain highly complex visual stimuli, such as faces, is handled by the right cerebral hemisphere. Auditory stimuli in general are processed by the temporal lobes of the cortex, and written and spoken word recognition and syntactical components of language processing are handled by certain regions of the left hemisphere of the cerebral cortex, notably Broca's and Wernicke's areas; while emotional, idiomatic, and prosodic aspects of language are handled by corresponding regions in the right hemisphere. Higher cognition, such as reasoning and problem solving, involves the frontal lobes of the cortex. Memory and the processing of emotional stimuli are handled by the combined effort of the cortex (notably the anterior and frontal regions) and subcortical structures (notably the limbic system).
One particular subcortical structure—the hippocampus—plays a major role in the formation of new explicit memories. It is believed that an intact hippocampus is needed to temporarily bind together distributed sites of activation in the cortex that together make up a whole, explicit memory for an event. See Brain
Theories of cognition are often tested by building computer models that embody the theories and then comparing the model's performance with human performance on selected tasks. These models tend to be of two types. Rule-based models consist of a long-term memory containing rules which specify actions to take in the presence of particular input patterns, a short-term memory that encodes input patterns and temporarily stores data structures constructed by the rules, and a control structure that guides the process and resolves conflicts when more than one rule applies to the current input. Neural network models simulate cognition as a strengthening and weakening of associations among cognitive events. They consist of a network of interconnected nodes, a mathematical formula for modifying the connections, and a mathematical formula for propagating activation through the network. See Intelligence