It is said that a picture is worth a thousand words. What is often forgotten is that a single word can be worth many thousands of pictures. As an example suppose you were to enter the world's most comprehensive library with a single word in your mind, and that this word was "bird." You might find literally thousands of books dealing with this subject, in each of which were many images, paintings, and photographs. And when you stop to think about it, it may seem quite remarkable that so much knowledge could be accessed through just that one word.
As it turns out, words are also worth many millions of computer pixels and many hours of computer data processing. Just think, for example, of how many pixels it might take to represent an identifiable cockroach. Not only would it be necessary to use enough pixels to create the general outline of a cockroach, but it would also be necessary that the image of the cockroach include enough unique characteristics to distinguish it from other similar insects, such as certain beetles. And yet the same entity can be identified by the written word, "cockroach," or just the two-syllable sound.
Indeed, the sound of a word like "bird" or "cockroach" will start a process of image building in our minds. We may suddenly "see" a cockroach in some vague manner in our minds. Then we may remember that cockroaches have six legs, and provide the image our minds have created with these appendages. Etc.
At this level a human word is not much different, say, from a particular kind of grunt or squeel made by a baboon. Many animal species are known to use a variety of such sounds to warn of the presence of particular kinds of predators, using different sounds to indicate different predators. The mechanism is presumably the same. The sound evokes the image of the associated predator in the brain of the hearer, and the hearer takes the appropriate precautions. Thus a single grunt, bark, squeel, etc., can evoke a whole array of facts, images, memories, etc., of lions.
Thus far an identical process seems to be at work in both animals and humans, namely the animal hears the sound, recognizes the sound as a symbol, and activates the part of its brain that responds to the identification of that symbol.
In summary then it can be said that certain sounds can activate certain regions of certain brains.
Notice that in this process no syntactic element is involved, only the semantic association of a sound with some region of the brain. The direction of activation is either straight in, as described above, or else straight out, as when an animal sounds a warning to its comrades, in which the brain activates the symbol from which the animal produces the appropriate sound. No lateral activation is involved.
But it can be seen that in higher animals it is possible for certain symbols to be clustered together in meaningful ways. For example the human sounds for "eat," "monkeys," and "bananas," might be arranged in such a way as to indicate that "Monkeys eat bananas." The particular order required to represent this meaning will differ from language to language, but the meaning will remain the same. Let us call this word order "surface syntax."
Let 'S' stand for "subject," 'V' stand for "verb," and 'O' stand for "object. Then we might have the following different languages and correct syntaxes required to represent the meaning, "Monkeys eat bananas."
Language A: Monkeys eat bananas. (SVO)
Language B: Bananas eat monkeys. (OVS)
Language C: Eat monkeys bananas. (VSO)
Language D: Eat bananas monkeys. (VOS)
Language E: Monkeys bananas eat. (SOV)
Language F: Bananas monkeys eat. (OSV)
In summary we might observe that in higher animals, and especially in man, besides the pathways that go straight into and straight out of the brain, such as sound->symbol->brain and brain->symbol->sound, there also exist lateral links between symbols that manifest themselves in surface syntax. Furthermore we might observe that symbols can stand in various relationships to one another at some deeper level, but that these relationships are not completely determined by the linear order in which the symbols occur (or in other words by surface syntax) since this order may vary from language to language. And we might deduce that there exists some kind of subsurface syntax that holds symbols together without regard for surface syntax linear order.
Human language consists of symbols that translate to sounds, images, and other sensory patterns, and all the links that can exist between these symbols and the regions of the brain that handle them. Needless to say, in human language the patterns created by such symbols and the links between them can quickly become very complex. Thus if we are to understand language we will need a model with which we can work, and for this I propose the following:
Let us imagine the brain as consisting of horizontal planes, or layers, with gaps between them. We might then have the following:
1. Top layer, the phonological plane. Here sights and sounds are identified as symbols.
2. Second layer, the syntactic plane. This layer consists of individual connecting points, called nodes. The nodes of this plane are Panlingua atoms, or words.
3. Third layer, the semantic plane. The nodes of this layer are called semnods.
4. Fourth layer, the lower brain. Here reside the various regions of the brain that handle stimuli associated with various symbols.
Now let us imagine two African warriors walking across the savannah. One of them recognizes a lion lurking in the grass. The sight of this predator triggers a response from the region of his brain that deals with lions. Part of this response is a warning to his comrade. "Lion!" he says, and both men suddenly come to a standstill, their spears poised for attack. What has happened from a linguistic point of view?
1. Warrior #1 recognizes lion, which activates the region of his lower brain that deals with lions.
2. The "lion" region of the lower brain of warrior #1 sends an impulse out to the phonological plane of his brain where it is translated into the audible word that is the symbol for "lion" in his language.
3. Warrior #2 hears the word and recognizes it as the symbol for "lion," which means that the "lion symbol" region of the phonological plane of his brain becomes activated.
4. The activation of this region of the phonological plane of warrior #2 sends an impulse into the region of his lower brain that deals with lions.
In the above explanation we have deliberately ignored the semantic and syntactic planes for purposes of simplification. There is no syntactic component to the utterance of the single word, "lion." A signal passes straight up out of the lower brain of warrior #1 and straight down into the lower brain of warrior #2.
But what if our two warriors were to happen upon the freshly killed carcass of a zebra, and warrior #1 were to say to warrior #2, "This zebra was killed by a lion?" What major linguistic linkages would be involved then?
Before we tackle this problem let us set up some definitions:
A synlink (syntactic link) is a link between two Panlingua atoms, or words.
A lexlink (lexical link) is a link between a word and a semnod.
A semlink is a link between two semnods.
In general terms the processes involved are similar to those in the preceding example. Signals pass up out of the lower brain of warrior #1, get converted to symbolic sounds, get heard and interpreted as symbols by warrior #2, and pass into the lower brain of warrior #2. But this time more than one symbol is involved, and so we have the phenomenon of syntax. The symbols occur in a certain linear order (or sequence) determined by surface syntax, and the words relate to one another in certain ways determined by subsurface syntax. Much has been said about surface syntax, that is, the linear order in which a string of symbols appears. But this order is easy to determine just by listening to the symbols. On the other hand, although we know there must be various regions of the lower brain that do various things, at this point very little is known about them. So let us focus our attention upon the linkages of the syntactic and semantic planes, which we will be able to determine by deduction.
"This zebra was killed by a lion."
In the following explanations, the words in capitals represent Panlingua atoms that reflect the same words in the sentence just given above:
THIS is linked to ZEBRA by a synlink of type "determiner." ZEBRA is linked to KILLED by a synlink of type "patient." A is linked to LION by a synlink of type "determiner." LION is linked to KILLED by a synlink of type "agent."
All of the above links are between Panlingua atoms, or words, and lie completely within the syntactic plane. The following links connect words to their semnods. Only link type is given, since semnods are already identified by the words to which they are linked.
THIS has a lexlink of type "default." ZEBRA has a lexlink of type "default." KILLED has a lexlink of type "past-tense declarative transitional." A has a lexlink of type "default." LION has a lexlink of type "default."
In the above lines, "default" means "no special type in particular." A majority of lexlinks are of this type, but for the minority that are not, lexlink type can carry very important elements of meaning.
The synlinks and lexlinks given above define the Panlingua representation of the sentence, "This zebra was killed by a lion." The fact that the surface syntax was in the passive voice was not included in this Panlingua representation because the meaning is exactly the same as, "A lion killed this zebra."
In summary, thoughts are constructed in Panlingua by setting up synlinks and lexlinks between nodes in the syntactic and semantic planes. The study of Panlingua, which is carried out by deduction, is therefore limited to an examination of the links and nodes of these two important planes of the human mind.