The language we understand and speak

Part One

By Eli Rejwan

December 28, 2000

We may learn a second language and get the impression that we speak it

while thinking in the native language. Upon reflection, it appears that we all think in an altogether different language, with no similarity to the native language or any newly acquired language.

We make use of our ability to think throughout our life, yet we do not know how we are doing it. An investigative probe is well deserved.


1. Thought, speech and evolution
2. Memory is a group of items
3. The 'intensity' qualifier: quantitative estimates
4. The time qualifier
5. The affinity qualifier
6. The human logic
7. The spoken language: a new addition
8. Learning the spoken language
9. The memory maze
10.The two levels of memory

1. Thought, speech and evolution

Let us start with the sequence of evolution. The speechless thought evolved first. The early creatures needed memory and thought long before speech came into practice. It is safe to assume that those early organisms could see and memorize vision details, hear and memorize sounds, feel fear and use their logic to find ways to protect themselves, even make simple sounds to attract attention, all before developing a language, which is the ability to attribute meaning to variation of sounds or other signals, and learning how to produce them.

Evolution has the habit of building on old capabilities more often than inventing new ways. Since the language is a late add-on, the ability to learn words must rely on the previously developed ability to memorize sound and on previous ability to think and associate sounds with events. Even the ability to speak must somehow rely on the older ability to produce warning sounds.

Did language require a new way of thinking particular to speech ? Most unlikely. Thinking is undoubtedly a very complex process. It evolved gradually. No abandonment of one mechanism in favor of another. We do not see signs of a reorganization. We would see a clear dividing line in mentality between species which make use of communications between its members and species which do not.

The impact of languages on our thoughts is not denied. Life cannot afford leaving such a tool not exploited to the maximum, and language is clearly a good resource. We also get help from the words and from whole sentences that we can speak to help in memory recall. This is because words and sentences are often associated with the memory of past events. Obviously this should not lead us to assume that the mechanism behind thought and memory is dependent on language. Humans born deaf and mute can think and remember. The mechanism of thinking must be acting within our own minds while doing the translation to and from the spoken language. The benefit from using the language is real but indirect.

Let us examine the properties of the language of thoughts and concepts. I will call it the thought language. It is a language, not a transient state of mind as some may tend to assume. It is the language that our memory uses. Unlike spoken languages, this language lacks the sound equivalent of words.

2. Memory is a group of items

Memory objects of the thought language do not have a one-to-one correspondence with words of the spoken language. An object in the thought language my require a large number of words to describe it. It may be made up of more than one of a wide variety of information: vision, hearing, thoughts, pain, fear, hunger, satisfaction, desire and the like. It can be a mixture of sensory information and abstract notions. This is the first of four major differences between the spoken language and the thought language.

3. The intensity qualifier:quantitative estimates

Any of the item ingredients in a thought object may range in clarity from vague to crisp clear, or in quantity from little to plenty. This is to say that a memory object may contain numerous items that we treat as one thought and each of the constituent items has qualifications associated with it to describe its quantity or extent or intensity or variety. Sound can be loud or faint, light can be strong or dim, pain can be mild or strong, fear can be of varying intensity, and desire can be weak or strong. Or anything between the two extremes.

We remember these variations even though we make no attempt to observe them. We hear about a person we never met, and instinctively associate some vague values to how tall, how friendly the person is. If these details are missing we make them up. Vague qualification will do.

This quantitative association with most memory ingredients is the second of four major differences between the thought language and the spoken language.

4. The time qualifier

The third aspect that is peculiar to the thought language is the time attribute that we associate with most items. It is not just the jet-lag phenomenon associated with our internal clock. We make extensive use of timing for coordination: movements, speech and, many other functions and, most important, the internal functions of the cell. Therefore each cell must have a timing mechanism. Timing description is attached to most items stored in our memory.

The timing information is itself qualified with shades to denote an approximation of the length of time and how reliable is the approximation.

We can notice that an object we are observing for the second time is moving faster than in the previous observation, although we may have paid no attention to its speed during the first observation.

Timing information must have been introduced in the early development of life when only simple single cells existed.

5. The affinity qualifier

The last and most important aspect of the thought language relates to what I will call the affinity qualifier.

An indication of 'how this relates to me' is attached to each and every element in the thought language. The intensity qualifier applies here.

We may mark things as dangerous, useful, friendly, pretty, feared, to hold onto them and protect them, to avoid them.. These are useful information for better chances of survival. The information is relied upon in every action we take: we walk, talk, think. Our affinity processor is always in control.

The affinity processor is in use with or without our awareness. Acting on genetic and life time acquired information, it causes us to shuts our eyes if we try staring directly at the sun, and makes us conclude too heavy when we try to move a weight beyond our capacity, preventing us from tearing our tendons.

Efficient affinity information offers better survival rate. Indeed it is essential to survival. It must have existed early in evolution, at the early stage of single cell organisms. It is the area that must have received the most attention by evolution, making it very complex and difficult to study.

A biological function is necessary to puts to use the affinity information. This function directs the retrieval of more related information where the affinity value is high, and may start other biological actions where needed. Without this biological function the stored information is of little use. This matter is taken up again in part 2 with more details, but here are some examples of the actions of this biological function:

We walk by a crowd when suddenly we notice a familiar face. We made no attempt to examine faces in the crowd.

In a noisy background we hear a familiar voice which attracts our attention. We may be surprised how our sense of hearing could detect a voice among much noise.

As we discover an insect crawling along our arm we move away that arm in a reflex, before we fully realize what has happened and assess the risk.

Everyone notices many instances where such responses are surprisingly fast and amazingly efficient. We say "it grabbed my attention". This is because our senses are continuously at work, simultaneously retrieving related information from memory. If the associated affinity qualifier is strong, additional related information is retrieved and either an action is taken or the continued information retrieval gain our awareness.

6. The human logic

Our mind relies on the thought language for logical thinking. It is an inaccurate tool made very efficient by the use of a number of guiding supports: past experience, logical rules such as mathematics, and information conveyed by others.

Evolution attempted to produce human minds with excellent logical capabilities. At the same time evolution produced safeguards to stop our minds from canceling out the influences of the affinity mechanisms supporting feelings such as pride in ourselves, possessiveness, egoism, greed and the like, that are important for better chances of survival. Any thought that goes against this self preservation mechanism is likely to be made to fade away quickly from our memory or be excluded from our thoughts in the first place.

In particular, the affinity mechanism does not let us doubt the accuracy of our own judgement. Otherwise we would use our own logic to detect all the cases where nature warps our minds and we would introduce corrections. Obviously nature objects and take over from us the ability to think in this direction. A search of this matter led me to an interesting example:

In general, humans are biased towards self interest but they do not become aware of it. It is so common that you would expect our minds cannot fail to realize that the same can happen to us. What does nature do? Simply drop it off our minds. Proof that we are barred from thinking about it: there is no word for it!

In other warp cases, encountered less frequently: we expect them to happen only to others. The thought that our own minds may be subject to the same weakness never crosses our minds. If it does, it will be made to fade away quickly. 'Myself excluded', 'my logic is always reliable'.

In contrast, words like 'pride' are encouraged by the affinity mechanism. The affinity is in control. It evolved to have the ability to prevent the logical mind from invading the affinity's own domain.

More recently the human species became social creatures. Straight expression of self praise has now a downside: others in society don't like someone indulging in frequent self praise and exaggerating his/her own successes. Evolution did not revise its mechanism. Instead, it superimposed self control to limit the expression of our true feelings. We now consciously refrain from self praise and obvious egoism. Even children show some control: we do not have to learn it, hiding our feelings about ourselves is inherited and comes naturally.

7. The spoken language: a new addition

The thought language and the spoken language are very different in nature. Speech is a later development in the history of evolution. It may be difficult to determine to what extent the advent of the ability to speak influenced our ability to think and our method of thinking.

We are aware that we begin by thinking, then find suitable words to express ourselves. We make up our thoughts by using our memory and putting our helpful imagination to work for us. We do this quickly. Then we work hard in order to put the new thoughts into words of a spoken language. At times we struggle to find the right words.

We listen to the response and immediately translate it into elements of the thought language. Indication to that: we can later remember the thoughts more clearly than the spoken words. We work with the thought language and use the spoken language as a temporary process for the purpose of communicating with other individuals.

In real life complexity prevails. It is not always clear that we think exclusively in the thought language. At times we do the translation to spoken words just to help probe deeper into our memory. We are aware of the fact that knowledge of words can influence our thoughts. This may be considered to indicate that the thought language adapted to accommodate spoken words as elements of its own memory and logic. But these are not reliable indicators, it may well be that we rely solely on the thought language and make use of spoken words to provide additional links to a wider range of memory items.

The written language represents a major complication in our use of languages. For simplicity I will restrict my opinions to the status prior to the introduction of writing.

We can gain some insight into our use of the thought language and the spoken language by having a close look at the aspects where these two differ. The time attribute and the intensity attribute are usually attached to most items of a thought, but may be completely lacking in the spoken words. When we do the translation we fill in some arbitrary values in the thought language. We all remember cases where we face reality and it is substantially different from what we assumed. This would not happen if our retention of the initial information was in terms of the spoken language.

8. Learning the spoken language

Our understanding of the words of a language and our ability to select and use words vary depending on whether we learn the words from hearing or reading. I will concentrate only on the spoken language.

Let us consider how children and illiterate people learn new words: it is always from the new word's context within the sentence they hear, sometimes helped by sensed information such as vision. When the meaning is not clear they may ask what the word means and get in response an explanation: the same word in a different context, such that guessing its meaning is easy.

Here the mind's ability to remember vague knowledge is helpful because it allows us to make efficient use of doubtful knowledge for integration into additional information in the future. A collection of vague understanding can make up a fairly reliable knowledge of the meaning of a word. Nothing is wasted in the world of the thought language. This is how we learn new words. That is how children learn the whole language. It applies also to learning in general, other than learning to recognize the sound of words. This is how babies familiarize themselves with the world around them, with the movements they can make, with what they see and sense. All these source converge to support our knowledge in the thought language.

We may attribute the ability to use the context information and to guess the meaning of any new word to a miraculous ability of the mind. The working of the mind is very efficient but this would seem less of a miracle if we remember that the evaluation of context occurs in the thought language, not in the spoken language. It appears that filling the blanks in the thought language is much easier than it is in the spoken language. There are plenty of hazy ingredients in a thought. We have adapted to manage obscurity.

9. The memory maze

The nature of the thought language is altogether different from the spoken language. We make use of both, hardly aware of the duality and the way the two relate.

How do we accumulate our knowledge about words ? This is an important question. Since we know a word from the context in which we encountered it in the past, we must be either accumulating in our memory all knowledge about a word as one entity dedicated to the spoken word, or completely relying on the recorded memory of the original contexts. In the latter case the word leads us directly to past memories of each past encounter of the word, including the substantial information on the circumstances when the word was heard.

For an analogy, we can think of the index in a book. No information is presented, only the page numbers, 'pointers' to where the word is mentioned. This provides access to all the information.

The alternative method is to compile an alphabetical dictionary associated with the word, and improve the contents of the dictionary as we scan the text of the book and re-encounter the word. This method results in some duplication, and will limit the contents of the dictionary to information deemed informative at the time it is compiled. This is an important point, because our own memory of words does not appear to be limited to the information deemed relevant at each occasion the words are encountered. We do extensive referring back to the instances where we learned the word.

Observation: we can detect a word being used outside the range of its normal meaning. Such knowledge requires an assessment of all previous encounters of the word, paying attention to its meaning in the various contexts, and the establishment of a normal range for the meaning. Problem is: those uses of the word were vague to us. This makes it next to impossible to organize a practical way to maintain the information in conjunction with the spoken word. Furthermore, we do it routinely, since the out of range comes to our attention when we had no intention to verify the range.

In theory, we could be accumulating memory for both the learning instances and the spoken word, but such a duplication is wasteful and very unlikely.

Therefore our memory of words consists of pointers to the original instances where we encounter them. Unlike the book index, where the word points to a sentence in the book, the memory of a word points to the circumstances where it was heard expressed in the thought language. The conversations we hear is translated to thought memory before storage. Words point to the translated information. That is an enormous amount of information to retrieve. We do it every time we encounter the word.

We can hardly contain ourselves from interrupting to say 'this is absurd': relearn the word from its sources every time we encounter it, then neglect to remember the meaning, just start all over again when we hear it the next time. This is contrary to our perception that after we learn the word we know it and need not search for its meaning.

It is like saying this: we hear a word, pick up off the shelves a book with sound markings that relate to the sound we heard, refer to chapters in a dozen other books, each chapter gives us some notion about the word, make use of the collected notions, then we put the books back on the shelf. We never learn the word. We do this every time we hear the word.

Definitely not the way we humans do things. However, we ourselves would adopt such a design if speed was not a concern. If it can be completed in close to no time, then going back to the sources of information has some very important advantages. Less storage, no duplication. Allows the use of up-to-date sources in a way no cumulative process can allow. Many aspects become obvious without specifically checking for them, such as range of meaning and variants in meaning. It truly mirrors what we got used to expect from our memory and take the result for granted. It is a very effective way to make use of a language.

This method can explain satisfactorily what we see happening in real life, but we have to coerce ourselves to accept the incredibly fast speed of action. When access speed is close to immediate, we can afford to reassess the meaning every time we hear the word.

10. The two levels of memory

We know that our memory is slow, far from being instantaneous. Memory is in two levels. This topic is discussed in more details later in Part Two and in my article Cell knowledge. We are aware of the slow level only. The slow level relies on the fast level for all its functions.

We intuitively consider learning to be a trivial achievement. Even a mouse can learn. It is fairly safe to say that learning relies on memory. Memory is the repository of all our knowledge. However, if we are told that learning a word of the spoken language is no more than setting up leads to many memories, each containing some hazy knowledge of the word, while the leads themselves are meaningless to us, we may object. Our intuition tells us that learning is immediate knowledge.

Learning, this magic of making memory meaningful to us, is not feasible. It is an illusion. We do not have the complete picture because most of memory activities are hidden from us in the fast level of memory.

The fast level of memory can be illustrated by visualizing a secretary who looks up the words in a dictionary, consisting of pointers to the cases where we originally encountered the word. If the word is listed, the secretary reads all the indexed material at the source, do the complex translation to the basic biological signals, and deliver to us the meaning. If this happens immediately, we get the feeling that we know the word. If the word is not listed in the index, the secretary has to discuss the matter with us. The discussion occurs at the slow level of memory. We are aware of this activity, the thinking activity in an attempt to determine the meaning of the word. The same applies to images, like looking at a face and recognizing the person.

So far language considerations were centered on listening and interpreting the spoken language. Similar processes applies to producing speech. When speaking, we select the word by picking up one past usage, access the set of leads, reach all other usage, and confirm to ourselves that it is the right word which we want to use.

Continued in: 'The language we understand and speak ' Part Two '

Any comments ? Yes, click here