25 Apr Conceptual Paradigms
The different functions of language are acquired at different stages. As described by Sowa, these functions show why semantics can be considered a guide to how we learn syntax: “First, children associate words with concrete concepts used in perceiving the world and acting upon it. Next, they learn syntactic rules for mapping concepts and conceptual relations to well-formed sentences. Finally, they master the formal structures of the type hierarchy and the fine points of syntax” (1984, p. 215). Sowa believes conceptual data, or meaning, is the intelligent foundation upon which linguistic ability is built.
This proposition is similar to Pinker’s “semantic bootstrapping hypothesis” (1984, p.40). My son Steven, for example, during his single word stage, was so interested in the concept of “ball”, that it was not safe to seat him in the shopping cart when we were buying eggs. He had distinct concepts of the ball objects’ shape, and the process of using hands and arms, or legs and feet, to propel it. This has served him well in athletics, despite the early unfortunate experiments.
|Understanding Context Cross-Reference|
|Click on these Links to other posts and glossary/bibliography references|
|Prior Post||Next Post|
|What was that word?||Continuity of Learning|
|language semantics||Sowa 1984|
|syntax concept||Pinker 1984|
|associate words||Aristotle 1952|
|knowledge representation||Schank 1986|
In a conceptually structured knowledge representation schema, linguistic knowledge may be acquired as specific “paradigms” (Pinker, 1984, p. 175). As children hear words, they store activation patterns in memory. With repetition, these activation patterns become stronger. Once a child can associate the input with data in her conceptual vocabulary, a link is established, and the information is added to the domain of knowledge in context. The audio pattern may be replicated, or a new link may be the only change that occurs in the brain.
Conceptual knowledge representation schemata are described in section 7. The main idea applicable to this section is that the concepts that people (especially children) learn, are stored in the brain in a connected way. The way they are connected in the brain are most likely to be through synaptic links between neurons that favor the spread of activation between related concepts. Computer programs can imitate this technique for processing. We will investigate this line of reasoning further in Sections 7-9 of this blog.
A linguistic paradigm as proposed by Pinker (1984, p.175) consists of the sound pattern of the word linked to its meaning (“ball” in the concept graph shown at left), with other links to sensory knowledge such as image, olfactory, and tactile data. As the morphology and syntax of the word are learned, they are linked in context in the graph. Frequent repetitions or exposure initiates paradigm learning. Repeated occurrence of patterns, such as morphological or syntactic patterns, in various lexical contexts which exhibit conceptual similarity lead to generalized paradigms.
Chimpanzees and great apes have exhibited substantial ability to learn versions of sign language, indicating they are intelligent animals with a complex conceptual system, despite their lack of what Sowa describes as “the human facility for mapping concepts into a linear stream of speech.” Taking the comparison one step further, Sowa says that limitations in children’s utterances “do not imply limitations in the children’s conceptual structures, but limitations in their schemata for mapping structures into sentences” (1984, p.214). Such observations point to less advanced generalized paradigms or less complex language centers.
The acquisition of generalized paradigms may begin with the formation of associations in a person’s brain. For example, when Yorrick hears mom put an -s on the end of “dog” or “cartoon” to indicate more than one, he is forming a paradigm: adding -s to a word makes it plural. He is learning to associate objects based on similar characteristics. These associations will be strengthened by positive reinforcement from parents, siblings, and teachers, and will be extremely useful in improving his ability to satisfy his personal needs. Over time, the associations develop into a specialized language center.
This language center will naturally develop in the region of the motor-control center for the speech system. The base of the third frontal convolution of the cerebral cortex, Broca’s area, is directly linked to the main motor cortex. This area holds data for the motor elaboration or functional control of speaking. Its damage may cause inability to speak or write (expressive aphasia) without affecting comprehension. Wernicke’s area, in the upper temporal convolution, handles receptive-speech comprehension. Injury to this area degrades comprehension of speech and text (receptive aphasia).
Section 1 of this blog showed how the Wernicke’s and Broca’s areas of the brain are spatially linked. This linkage is consistent with the massively interconnected structure of the brain, but the character of the links is particularly interesting in that it may be able to tell us more about the conceptual linkages between expressive and receptive speech.
Further, the emotional centers contribute to our ability to confusticate, and to disentangle subtext. The interaction of these brain components, each with distinctly different functions, bespeaks a complex, interconnected model, with different processes to mimic different cognitive activities.
|Click below to look in each Understanding Context section|
|4||Perception and Cognition||5||Fuzzy Logic||6||Language and Dialog||7||Cybernetic Models|
|8||Apps and Processes||9||The End of Code||Glossary||Bibliography|