"I invite you to eat steamed lamb, steamed bear paws, steamed deer tails, roasted duck, roasted chicks, roasted goose, brine pig, brine duck, sauced chicken, bacon, pine blossom belly, dried meat, sausage, assorted Su pan, smoked chicken white belly, steamed eight treasure pig, rice stuffed duck, can pheasant, canned quail..." Friends who have heard cross-talk must be familiar with the name of the dish, but at the same time of drooling, how many dish names do you remember? If the restaurant doesn't have seafood, what else can you eat?
Accept the Yangtian ridicule from Xiaoyue Yue (picture from the Internet)
Compared with people with ten lines at a glance, we not only take time to read, but we don't remember much in the end. Why, exactly? In fact, at this time, it is our working memory that is at work.
What is working memory?
Working memory is a system with limited resources for the temporary storage and processing of information during the performance of cognitive tasks. We can understand it as a temporary psychological "workbench." At this workbench, people put temporary storage of information on the platform to manipulate and process to help us understand the language, solve problems, etc. In daily life, we often read a piece of content or hear a long paragraph. In order to better understand the content in the language, people often integrate the information they currently hear or see with the information they have previously processed, so that they can fully understand the information contained in the language. From the semantic content of a single sentence to the semantic content of a larger paragraph, people integrate information. In the process of semantic integration processing, the participation of working memory is required. Specifically, semantic integration requires working memory to store contextual information and update it based on current content.
Semantic integration in working memory and language
Whether there are differences in the psychological representations of semantic integration when understanding individual sentences and paragraph discourses is an important concern for researchers. Previous studies have suggested that semantic integration at the sentence and discourse level is instantaneous integration, that is, integration occurs around 400 milliseconds after reading the current word. The hypothesis in these studies is that all participants have the same working memory ability, but in fact there are individual differences in working memory ability.
Recently, the academic journal Neuropsychologia published the research paper "How working memory capacity modulates the time course ofsemantic integration at sentence and discourse level". In this study, to investigate the effect of working memory ability on the time process of semantic integration, the researchers screened out two groups of participants with high and low working memory breadth from a large number of participants. The language material presented to the participant consists of a single sentence and a discourse consisting of four sentences. There is a key word in a single sentence that is consistent or inconsistent with the previous contextual semantics. In the discourse, the keywords in the fourth sentence are semantically consistent or inconsistent with the first sentence in the discourse. The above materials were presented to the participants for reading comprehension tasks, so that the effects of different working memory abilities on semantic integration at the sentence and discourse level could be compared.
The behavioral results of the study showed that all participants had a reading score of more than 80% under all conditions, indicating that the participants had read carefully. EEG results showed that the high working memory group induced the N400 and P600 effects at both the sentence and discourse levels (Figure 1), while the low working memory group induced only the P600 effect (Figure 2). The results show that whether the reader can immediately detect unusual words related to the context depends on the reader's working memory ability, not the size of the context.
Figure 1 Semantic inconsistencies induced at the sentence and discourse level in the high working memory group
Figure 2 Semantic inconsistencies induced at the sentence and discourse level in the low working memory group
The emergence of the high working memory group N400 shows that individuals with high working memory can integrate current information with contextual information in a single sentence or in a larger discourse in a timely manner. Individuals with low working memory, even at the sentence level, failed to detect semantic inconsistencies in time, and the same was true at the discourse level.
It can be inferred from this that the temporal course of semantic integration does not depend on whether the context is a sentence or a discourse, regardless of the reader's working memory ability. Therefore, in the time process of semantic integration, it is the individual's working memory ability that plays a greater decisive role, rather than the size of the context.
In short, this study tells us that although you can understand the meaning of sentences by combining the words that appear in the moment with the previous context, people with poor working memory will understand it more slowly. Working memory plays a more important role in the time process of understanding sentences than contextual size. But that's not to say that context is completely ineffective. Only relatively speaking, the impact of working memory ability is more prominent.
How to save my working memory?
From scientific research back to our real lives, since working memory is so important to us, what is the way to improve our working memory ability? There are many ways to train working memory, such as training for a specific function, such as training for newer N-back tasks, or Cogmed training for working memory as a whole. The N-back task is to determine whether the current letter or position is the same as the previous N occurrences. The training results are mainly based on the correct rate of judgment and reaction time. Cogmed is a widely used working memory training program (www.cogmed.com), namely Cogmed JM, Cogmed RM and Cogmed QM. Cogmed JM is aimed at children aged 4-6 years, while CogmedRM is aimed at people aged 7-18 years, and Cogmed QM is more suitable for adults. Either way, the training content is relatively rich.
For ordinary people, it is possible to search for some software about cognitive training on the Internet, such as the more comprehensive Cognifit, IQMindware, Lumosity and other paid software. BrainWorkShop is a freeware for working memory only, including a series of N-back task training. Of course, no matter what kind of training, if you want to improve your ability in this area, you still need to train continuously. With better working memory, maybe you can also see ten lines at a glance!
bibliography:
Yang, X., Zhang, X., Zhang, Y., Zhang, Q., & Li, X. (2020). How workingmemory capacity modulates the time course of semantic integration at sentenceand discourse level. Neuropsychologia,107383.https://doi.org/10.1016/j.neuropsychologia.2020.107383.
Source: Institute of Psychology, Chinese Academy of Sciences