From Oho to Siri (Introduction to Language Psychology)

Author: vamei, http://www.cnblogs.com/vamei. Reprinted. Keep this statement.

This post was posted on the Sina Blog.Article. I am about to give up my Sina Blog and post my post as a backup. This article is about language psychology. I am very impressed by many experiments in Language Psychology, especially the two mentioned below. Many arguments about psychology are based on Behavioral experiments. Therefore, psychology is a very useful scientific discipline. Psychology andCodeVery far away, but very close to people, hope to be useful to friends in the garden.

 

Meaning of Language

Language is an important tool that evolution has given to humans. Its basic function is to share information through communication. We use voice and tone changes to distinguish the basic syllable (phoneme) and further combine them into meaningful words. The elementary language is further combined into complex sentences under the guidance of syntax. With language, humans can communicate with each other with a large amount of information. Such communication promotes human collaboration and greatly increases the chance of survival of species. Animals also have language, but they are relatively simple. For example, they can see the ROAR and roar, and then the simple "Oho" sound from the ape. In addition, compared with animal languages, human languages not only have complex structures, but also have the ability to transmit complex emotions. We can appreciate love, hate, loneliness, and emptiness through literature, or tell friends about their feelings, which is also a charm of human language. This emotional function eventually promotes the formation of rich interpersonal relationships. Many languages also have corresponding texts, which are recorded in a graphical manner. Text can store information and allow human experience and knowledge to be accumulated.

Language also helps individual thinking. A language is actually a symbolic system composed of speech or graphics. Without language, it is difficult for the brain to abstract concepts. Abstract concepts are important to logical thinking and memory. When we think rationally, we use some nouns to replace common things, such as "Elephants ". If we re-define the shape and other features of an elephant when thinking about "hunters killing elephants", it will greatly increase the attention system and working memory) system Burden. Language abstraction can also help us remember. For example, we often organize our own experiences in a language-based manner. When we recall, language is a good index. When we recall, the language information (such as the name of a person or hometown) quickly and effectively finds relevant memories. At this time, the language information seems to provide a set of directories for searching for memory.

 

Language Lab

Because it carries important social and individual functions, the use of language has also been enhanced in the long evolution process. People's language systems are jointly implemented by multiple brain functions, such as hearing, vision, motion area, attention, and working memory. However, language-related actions are oftenModular(Modularity), the relevant neural circuits are enhanced, and independently responsible for language-related functions. Modular functions have the advantage of fast response and processing, but are easy to mislead. We can perform a small test. Please immediately state the color shown in the figure:

Although we already know that the experiment requires true colors, the relevant text information will still interfere with our judgment. This becomesStroop Effect. When we see the image, the human brain is actually performing two tasks to identify the real color and read the text. The first task is a conscious process controlled by attention, which is relatively slow. The second task is Fast Modular action. The attention has not been reflected yet, and the information has been read. Although we only want to know the true color, the Text Comprehension function constantly interferes with our conscious process.

During the auditory process, we can also observe similar language modularization functions, such as Garret, bever, and foder experiments in 1966. We play the following sentence to the participants and play a tick at the same time on:

Even though the National Day has a long holiday, the congestion around the world has collapsed.

However, many people will report the following:

Even though the National Day has a long holiday, * congestion everywhere is still crashing.

We can see that the ticking sound is considered to be a break (comma) of the meaning. Language modularization allows us to quickly form a situation (schema), that is, to understand the main meaning of the entire sentence, and then add details from top to bottom. The ticking sound is not considered as a major part of the meaning, so it is added later, so it is easy to place it in such a broken meaning. It can be seen that the human brain does not receive and understand the language in sequence.

 

Essence of Language
There are great similarities among several mainstream modern languages. They all have a wide range of vocabulary and are quite compatible with new information, which is very important for the information age. For example, the word "diaosi" has been given a new meaning in the Chinese text and quickly accepted by people. Modern languages also comply with similar processes. For example, a conversation between two people usually revolves around a topic and cannot jump freely. If a person asks, then another person's subsequent words will be treated as an answer. The similarities of modern languages have often led to the misunderstanding that languages are very similar and can all implement the functions we mentioned above. However, with the study of small languages, especially the closed indigenous languages on the island, linguistics have found that these languages can be very different from mainstream languages. For example, if some languages do not have the word "Laugh", then the local people will not be "Laugh. This reveals that many functions of modern languages are constantly evolved, rather than exist at the beginning.

Traditionally, grammar is the key to understanding a language because it is a logical system. In the computer era, because of the needs of search and language processing, people hope that computers can understand the meaning of a piece of text according to syntax. Natural language understanding is the research direction for this problem. However, researchers gradually discovered that the results obtained through syntactic analysis are not reliable due to the ambiguity of human languages. This method often leads to many errors. Later, statistical methods were introduced into natural language understanding. The computer first trains a large number of texts to create statistical samples. When a computer encounters a new sentence, it can "Guess" the meaning of the sentence based on previous statistical experience. Apple's Siri product is an application of human-computer language interaction that has gained widespread attention. When natural language understanding succeeds again, it seems that it also implies the essence of language: the brain may be able to understand the language based on statistics rather than grammar. Specifically, the brain builds a huge text library with experience. Sentences in this library are grouped based on their meanings. When we hear a sentence, find the most likely meaning of the sentence. It can be said that while Siri is amazing, it also makes people think about the essence of language.

Maybe one day, we will learn and use the language in a brand new way.