How speaking works

Exactly how does speaking work? What happens in human brain? Surely it requires a lot of processing power? To be able to speak meaningful sentences out fluently, one must be able to plan what to say next, organise the words of the current sentence into correct order and pronounce the current word correctly. And speaking can not take your whole attention, you must be able to for instance walk at the same time.


According to studies, following parts are most active when we speak:

• auditory cortex 
• lateral sulcus (producing spoken language)
• temporal lobe (meaning of heard words)
• Wernicke's area (language comprehension)
• Broca's area (speech generation)
• motor region

What does this mean? How do these areas work together to form speach? I assume it all goes like this: in the brain there are multiple processes. Lateral sulcus is working - together with memory - as a preparatory process: it works in advance and prepares the abstract ideas that we are about to speak out. These ideas are not yet at this stage connected to words. The next process, The next process works in temporal lobe, and it connects abstract ideas with words. This must be done together with language comprehension process, which is done to form correctly formed sentences. The next phase is forming speech mechanically, which is done in motoric area, controlled by Broca's area, where speeck is actually produced.

These processes all work in parallel, without planning the whole sentence beforehand. When you are speaking a certain word of a sentence, you will probably know what you are going to say in that sentence in abstract level, but you don't necessarily yet what the words will be, and in what order they will come, until they do.

Let us look at some everyday examples, for how they fit to the model we have created:

Synonyms.

Sometimes it happens that when one speaks, there are two equal synonyms to choose from, but unable to choose quickly enough, the spoken word will come out as a funny combination of the two synonyms. The speaking processes feedback mechanism will quickly detect this error and the speaker will correct the mistake. But what happened? An easyexplanation is that the word-choosing process was not quite finished when it had to output the net word to the speech generation process. This caused the word to change in the middle of being pronounced.

Lost for words.

The same applies when a suitable word can not be found even if one has the idea ready. The database query just does not seem to produce any results, or the results are in wrong language. In this situation the speaker will either use some fill words to buy some time for extended database search - or return the control to the preparatory process and re-form the sentence so that the specific word would not be needed.

Blackout.

This has happened to me, and I don't think I'm the only one. When completing a sentence,  there is no next sentence available in queue, or the next sentence does not seem to have anything to do with the one you just finished. I can imagine this as a synchronisation problem in Lateral sulcus: it is working too fast. the human communication transmission method has a very limited bandwidth, and the speed of thought must keep in sync in order for the opponent to understand the speech. I assume the blackout is a result of the monitor, which is observing our own speech and providing feedback to the processes of what our own speech sounds like, real-time. If the lateral sulcus notes that the speech in the feedback channel does not have anything to do with the idea or sentence currently being prepared, the process will halt.


A great article: Connections between thought and language