how we understand and reproduce speech and how this knowledge will help in life
Hi! I'm Zhenya, CPO at the corporate messenger Compass. In my free time, I like to dig into something new. Recently, I decided to delve deeper into the neurophysiology of communication to understand how we perceive speech, how we “create” it, and at what point difficulties with understanding each other arise. This knowledge will be useful for any communication, including work.
There is a question – it means it is time to look for answers. I found a master of linguistics in the team, tormented him with questions – and began to figure it out.
As I delved deeper into the topic, I realized that it can be viewed from two sides. The first is knowledge about speech itself: how it is organized, perceived, and reproduced. The second is the problems that are important and need to be voiced, based on available knowledge or its limitations.
I share what I got, as well as practical conclusions that can be applied to any communication in life, whether with loved ones or with colleagues. And, of course, I invite you to discuss. I will be glad to receive opinions, clarifications and additions.
Speech perception and production: the classical model and its disadvantages
The first person to make a significant contribution to the study of speech was Paul Broca, an outstanding French surgeon and anthropologist. In 1865, he published a paper on the discovery of the speech center, which was later called Broca's center. After the death of two patients with speech disorders (aphasia), the scientist studied their brains and discovered lesions in the same areas of the left hemisphere. More precisely, in the lower back part of the third frontal gyrus (both patients were right-handed).
Broca concluded that the discovered center is responsible for the motor organization of speech. If there are lesions in this area, then a disorder occurs, which was later called Broca's aphasia – that is, difficulties with speech production.
A little later, another important discovery was made by the German psychoneuropathologist Carl Wernicke. Having studied the brain of a deceased patient who had previously suffered a stroke and had serious speech problems, he discovered a lesion localized in the left hemisphere, in the posterior parietal and temporal region. Wernicke concluded that this area is involved in understanding speech. Even later, he discovered that damage to the fibers that connect Wernicke's and Broca's areas also causes aphasia.
The canonical concept of Broca and Wernicke eventually formed the basis of more detailed concepts. One of them is the Wernicke-Lichtheim-Geschwind model. It is presented as a diagram with three key points:
motor speech center (Broca's area);
sensory speech center (Wernicke's area);
“concept center” (“concept storage”), which is functionally built above the motor and sensory centers.
According to this model, words (i.e., combinations of sounds) during the perception and understanding of oral speech through hearing pass through the auditory sensory system to area 41, the primary auditory cortex. The information is then transmitted to Wernicke's area – and semantic analysis of the word occurs there.
During speech production, words from Wernicke's area travel to Broca's area via the arcuate fasciculus, which results in the assembly of morphemes in the correct order. In Broca's area, instructions for speech production are created and then sent to the motor cortex, which controls the muscles of the face and transmits information to motor neurons. These, in turn, give motor commands to the muscles involved in pronouncing words.
Sensory and motor images in this model do not have semantic content, therefore, during signal processing, an additional appeal is made to the “concept storage” or “concept center”, that is, to the semantic units of speech.
The diagram above also reflects the points where aphasias occur. They are caused not only by damage to specific zones, but also by disruptions to their connections. There are seven such aphasias. For example, when the arcuate bundle, which connects Broca's and Wernicke's zones, is damaged, a person understands speech, but makes mistakes. He is aware of these mistakes, but cannot correct them on his own.
The work of Broca, Wernicke, Geschwind and other scientists has made a huge contribution to understanding how speech works. However, the model described is now considered outdated. This is due to at least three problems:
Patients with specific aphasia often have different symptoms.
The model itself is not capable of reflecting the full complexity of the language processes that occur in the brain.
Classic symptoms of Broca's aphasia may be present in Wernicke's aphasia, and there are also cases where the patient has damage to one of the areas, but there are no symptoms characteristic of this.
Moreover, a 2016 meta-analysis of the boundaries of Broca's and Wernicke's areas found that there were significant discrepancies in the establishment of the boundaries of these areas.
Soviet scientist Alexander Romanovich Luria also made a major contribution to the study of speech. Among other things, he created a classification of aphasias and formulated the principles of system-dynamic localization of higher mental functions, which include speech. According to these principles, speech activity involves different areas of the cortex, each of which makes a special contribution to the overall process.
Problem #1 In many publicly available sources, the workings of speech are still explained by the classical but outdated Wernicke-Lichtheim-Geschwind model. Because of this, ideas about the organization of speech are greatly simplified. |
Methods of speech research: from origins to modern times
There are several methods of speech research. The classic Wernicke-Lichtheim-Geschwind model is based on the morphological method. This means that specialists observed patients with aphasia and made a detailed description of the clinical picture. After the death of such patients, scientists performed an autopsy and compared the obtained pathological data with what they had described earlier. Both Broca and Wernicke did this. And Luria studied the data from surgery protocols for his work.
This approach was very useful in the early stages, but today the concepts based on the morphological method are outdated. The morphological method has been replaced by electrophysiological, electrical stimulation methods, as well as the most relevant and preferred methods of functional neuroimaging.
The latter include fMRI, diffusion tensor MRI, positron emission tomography, and MEG. All of the listed studies are noninvasive, meaning they work without direct access to the brain. They also allow us to see which brain structures are functioning at a particular moment.
Problem #2 The most modern and effective methods of speech research have been used relatively recently. They give good results, but are still limited in their capabilities. And their implementation requires very complex and expensive equipment, so such methods are not available to all scientists. |
New data on the neurophysiology of speech
Current models of speech organization are based on the fact that this is a very complex system that cannot yet be described in one, most precise way. Speech zones include:
41st primary field of the auditory analyzer;
secondary parts of the temporal cortex (areas 42 and 22);
some parts of the parietal-convexital surface of the left hemisphere;
frontal lobes of the brain – when they are damaged, a person has difficulty understanding complex forms of speech and the subtext of statements.
Even those areas of the cortex that are not related to speech still influence the perception and production of speech, albeit indirectly. Deep structures of the brain are also associated with speech activity, including the basal ganglia, pathways, and the cerebellum. But that is not all. For example, the limbic system also influences the emotional component of speech.
If we consider the speech zones separately, they will have their own functions. But at the same time, all zones are united by bundles of long and short fibers, thanks to which they become a single, well-coordinated mechanism.
Problem #3 Modern science says that speech activity involves many parts of the brain. It is an impressive system with complex interconnections. This means that it is impossible to describe how a person perceives or reproduces oral speech using simple “from A to Z” schemes. Moreover, the more complex the system, the higher the probability of errors. |
How the brain works when perceiving speech by ear
When a person hears speech, the acoustic information enters the organ of Corti (in the ear), and then moves to the primary auditory cortex. In parallel, it is transmitted to the Wernicke area – in order to highlight the meaning. The area itself is located next to the tertiary fields, where abstraction operations and understanding of complex speech patterns occur.
To understand the meaning in context, significant areas of the temporal cortex are involved, and at the same time, auditory-verbal memory is also involved. This is necessary so that a person correctly understands a phrase based on the context, the data about which is stored in memory.
Problem #4 The way a person understands or reproduces speech is closely connected with his memory. And it has many of its own characteristics, including individual ones. In addition, the data stored in memory can be distorted under the influence of various factors. Because of this, in the same situation, people can perceive the context and react differently to the speech addressed to them. |
How the brain works when pronouncing a word
To pronounce a word, its image must pass from Wernicke's area and the supramarginal gyrus to Broca's area (via the arcuate fasciculus). As a result, a special articulation program is created, the implementation of which is the responsibility of the lower part of the motor cortex. The latter, in turn, controls the facial muscles. And for the statement to be emotionally charged, the cortex of the left hemisphere must be connected with the resources of the right hemisphere. The same connection is required when it is necessary to distinguish the intonation of speech perceived by ear.
When it comes to formulating a complete, complex statement, the frontal lobes (especially the prefrontal part) also come into play. They inhibit impulsive reactions and help transform speech into behavior that corresponds to a specific plan or intention.
Dual Stream Model
Since modern technical capabilities already allow us to track the pathways along which speech activity is serviced in the brain, the concepts of speech organization are also changing. One of the relatively new models is the “dual stream” model.
It suggests that the perception of speech sounds occurs through two streams of information. The ventral one covers the middle and upper parts of the parietal lobe of the brain. It is responsible for recognizing and processing sounds.
The dorsal stream includes the posterior parts of the frontal and temporal lobes, as well as part of the insula. It establishes connections between perceived sounds and the articulatory system.
If we return to the classical model, then the bundle connecting Wernicke's and Broca's areas is one of the parts of the dorsal stream.
There are also more complex schemes that reflect the structural components of the double flow. But it is important to understand that, despite all the modern possibilities, speech organization still remains a phenomenon that raises many questions. Science still does not know the exact sequence of information processing. For example, it is impossible to say what is processed first: sounds or the conveyed meaning. For now, we can only say that each level is processed in parallel, with all the data received closely interacting with each other.
Problem #5 Modern models of speech organization provide a more complete, but not exhaustive, idea of how speech works. However, for an ordinary person, not a specialist, they only complicate the understanding of speech processes. |
Meanings that are stored in memory
To process and understand a word heard, and to prepare a message, the brain must access memory, where various meanings are stored. This is not a separate place whose boundaries can be outlined, but also a whole system, the details of which science does not yet know.
The term “mental lexicon” is used to describe this “storehouse of meanings.” It is the sum total of all information about words. This includes the meaning (semantics) of a word, its roles in a sentence (syntax), and variations of the same word when it is spoken or written. That is, somewhere in the brain, the information is stored that “kitties” are the same as “cats,” but in a cuter form.
The mental lexicon cannot be compared to a reference book or dictionary – this would be a serious simplification. It has important features:
The smallest units of such a lexicon are not words, but morphemes (suffixes, prefixes, etc.). At the same time, the organization of the mental lexicon is multi-level, so at higher levels the minimal unit still remains the word.
The organization of meanings in the mental lexicon is represented by a complex network with many interconnections. Words that are used more often with each other will be “nearby” in this network. That is, the connections between them will be much stronger. It can be assumed that the connections between the words “cat” and “lie” will be much stronger than between the words “cat” and “work”.
Words in the mental lexicon are organized not only by meaning, but also by sound composition. This allows the brain to quickly distinguish words that sound similar, such as “things” and “prophetic.”
There is also an opinion that there is more than one mental lexicon. It is assumed that there may be two of them. One is involved in speech perception, and the other in its reproduction.
Problem #6 There are certain patterns in how connections are organized in the mental lexicon. But at the same time, there is a lot of individuality in this process. And this means that different people often have different associations on the same occasion. This leads to serious difficulties in communication. People are often surprised why they say understandable things, but they are not understood. One of the reasons is the individual characteristics of the mental lexicon. |
One of the important ideas for understanding how speech works is the concept of lexical access. It theoretically describes the stage of word processing that cannot be localized. But it can be assumed that when interacting with the mental lexicon, three successive tasks are solved:
Lexical access. This is the process by which a word heard is compared with what is stored in memory. Both the meaning and form of the word are taken into account. At this stage, everything that fits the meaning is activated in memory. For example, when the word “parrot” is mentioned, all images associated with parrots, bright colors, birds, etc. are activated and ready for subsequent use.
Lexical selection. At the stage of lexical selection, everything unnecessary is cut off and only what fits the context remains. Let's say a child mentions a parrot, and the parent's brain, taking into account the context, understands that it is not about a bird, but about a toy that they bought together last week.
Lexical integration. This is a process during which all individual images (for each word heard), taking into account the context, are put together into phrases, sentences, and even large texts. That is, everything that was previously separate units is combined into a clear system.
There is also a concept of utterance formation, which describes how speech is reproduced. It is similar to the previous one, but works in the opposite direction. Within the concept, such an issue as planning is considered – in this case, macroplanning and microplanning.
Macro planning is preparing a message with an outcome in mind that the speaker is hoping to achieve. For example, he wants to give instructions, ask something, keep up a conversation, etc.
Microplanning is the search for specific solutions for a given task. At this stage, all the appropriate words and everything connected with them pop up in your head. Work also takes place on formulating a sentence and the form in which it will be pronounced. To do this, the brain determines the parts of speech, extracts lemmas (initial forms of words) from memory, and then puts them in the required form, taking into account the structure of the sentence.
To name the selected words, a phonological analysis is carried out. They are broken down into syllables, with the stressed one highlighted, and then articulation occurs – the pronunciation itself.
Problem #7 There is currently very little neurobiological data that allows us to describe the work of the mental lexicon. Mostly, we are talking about theoretical concepts. We know only one thing for sure: both the processing and production of speech are extremely complex processes with a large number of “blind spots”. |
How and where can this knowledge be applied?
Even the limited knowledge about speech that science has today can already be used to your advantage. For example, it is easier to find a common language with people or easier to communicate in a team. What conclusions and recommendations will help with this:
Keep in mind that each person has their own unique mental lexicon and equally unique experience. This means that your loved ones, colleagues and everyone else should not understand you by default. Try to convey important information as simply as possible – without unnecessary metaphors, subtexts and complex meanings. If necessary, clarify, repeat several times in different words, ask again.
Remember that context greatly influences speech perception. Try to “cleanse” the communication space from all unnecessary things as much as possible. If we are talking about teamwork, then create instructions and regulations for all standard tasks, and then place them in one place. This will reduce verbal discussions of processes to a minimum, which will reduce the number of errors. If we are talking about personal life, then try to discuss all important and serious issues in private, and not in noisy places. Do this when all participants are calm, because emotions are also a context that affects the perception of speech.
Create conditions for quality feedback. Lively dialogue in which loved ones and colleagues can give each other constructive feedback is a valuable source of information. At work, it is better to summarize the information received in writing.
Where it is necessary to eliminate situations like “didn’t hear”, “didn’t understand”, “understood, but not correctly”, use technical solutions. Calendars, task managers, kanban boards and similar tools will help to structure verbal agreements, giving them a more universal and understandable form for everyone.
Use the uniqueness of your mental lexicon for brainstorming and searching for new ideas. Since people have their own individual connections between words and meanings, each participant will have their own chain of associations. The more people participate in a properly organized brainstorming session, the more interesting ideas can be found.
Share your impressions in the comments – I will be glad to discuss both the scientific side of the issue and practical recommendations.
![[Личный опыт] Vzhuh – and you are in biotech. How to get hired in an industry that inspires](https://prog.world/wp-content/uploads/2021/03/tvpl4knckbcu3lbrhcjnibe5jas-768x768.jpeg)
