User:Genezistan/Synchronization of thoughts/lecture 1
In Lecture One I am going to cover some issues in Linguistics, Logic and Philosophy combined. They are the concepts of meaning, the exploration of meaning or semantic analysis, and the issue of context with a view to finding some regularities between meaning, context and signs (forms). I am also going to address the isssue of identity and continuity, whether the actual identity of objects in real life, or a symbolic one created by calling it a name, or other unique verbal or non-verbal constructs to identify anything existing (such as an object, etc.), or being an illusion.
Then I am going to cover the issue of definition and defining the boundaries of objects, together with considering the law of identity and the practice of chunking. I am going to detail the identity of an object, a subject, and identification in 2D in general, as well as identification by using a language with space and time as final identifiers. I will cover the details of understanding objects (frozen in time and) arranged in space (2D) as hierarchies, triplets, tuples, icons and religious symbols.
I am also going to discuss the issue of semantic analysis, the use of grammar terminology in semantic analysis, the difference between content words and grammar words as well as the meaning of nouns and verbs. I also deal with the issue of word clusters (headers, titles and labels and messages) and the issue of multifaceted single words. Genezistan 07:12, 5 December 2009 (UTC)
- 1 Lecture One
- 2 The Meaning of what?
- 3 Meaning is a Relation
- 4 Meaning is verb
- 5 What are verbs good for?
- 6 What are nouns good for?
- 7 Context
- 8 Chunking
- 9 Identity
- 10 The Making of Definitions
- 11 The boundaries of Self and the difference between subject and object
- 12 Semantic analysis - summary
- 13 Lecture summary
- 14 References:
- 15 Next lecture
Subject and Object – seen from a distance
In writing about any subject you are expected to give it a title either as the first or the last thing to do. A title is usually a cluster of words, mainly of nominal phrases that either tell you what that title identifies (or refers to like if it were a tag), or what the writing is about. In the latter case the cluster of words would usually contain a verb in a noun form. This form of giving a title to a piece of writing is a matter of convention and is of course changed from time to time.
Therefore you can look at the structure of any writing of any genre and any length as being composed of a list of titles and a list of paragraphs made up from a list of sentences. Most of those sentences are made up from clusters of words again that are now called messages, because they make sense or they have meaning. As a rule messages have a verb part in it as a centrepiece of a sentence. Words, including verbs on their own, do not make sense.
Can you supply examples? Fine.
Now titles, headings, labels and even tags are used in connection to some physical objects as their incomplete, „non-exact” identifiers. As there are many similar objects around and there are many locations, anything in focus is subjected to its environment hence such an object is called a subject. If you talk about an object in focus in a kind of reflection, and as a product of your reproductive skills (like a piece of writing, a photo, etc.) then the subject is thought to be descriptive of the object itself.
/Insert a drawing here: Figure_1 : object – title – subject/
In representing object-title-subject as a diagram in 2D (a drawing in a sheet of paper) to visualize relations between them you automatically convert them to be frozen in spatial relation (and time). The next observable relation is size, not equal or bigger in the "contained in" version.
Example: Book (object) – The Fall (title) - fall (subject)
Now if you want to access that body of knowledge representation by its form (as a book) or media, then I am sure you know how to do that. But if you want to access the same by the subject (content) – then you have to use a different method of access. As a rule form and content are not separable (without a mental operation) in any object and they are noticed by shifting your attention from one to the other. The same is true about concepts which are also objects, and are physically represented by word clusters arranged as form and content.
We do not know for sure how they (i.e. what we call the concepts) are represented in our brains, but we are sure that we need to make our disciplined thoughts available to others by writing them down as word clusters. And our disciplined thoughts are various lists of titles and messages with special emphasis on using them as the definition of the meaning of the individual concepts as well as word clusters. Therefore such word clusters are usually messages, which have a verbal member in the cluster indicating some state, action or event related to the subject of the message. And the subject of a message may or would normally be an object in the real world as we have already seen.
Subject and object – seen from closer
Now we know that word clusters, especially sentences have meaning or they make sense. To make sense clusters also need a verb. Traditionally, individual words were classified into parts of speech, whereas today they are seen as divided into content words and grammar words. Content words are supposed to carry the meaning of a sentence, and therefore they are in the focus of semantic analysis. Semantic analysis is also done in logic, where words, especially those identifying concepts are classified to signify or to mean objects, properties and relations.
So in a sentence a logician would look for a message, which is assumed to be composed of a subject and a predicate, not quite the same decomposition as the parts of speech. For a logician such a message means a proposition, which is nothing else as said before, but a message, and the meaning of which is evaluated by a logician based on knowledge, experience or a number of logic operations. Meaning in that sense is the result of evaluation, the comparison of the content of the statement in the mind of the person with the relevant chunk of reality and establishing a match, or the lack of it. In a proposition the subject is an „object” – whether live or dead, real or imaginary, complemented with a predicate, which is normally a verb on its own, or the verb „exist” plus a verb complement, which again is either an object, or a property (a noun, or an adjective).
Consequently, the content of a statement or proposition (remember: the meaning), just as well that of a sentence or a cluster of words is, or may be different in our heads individually. What we may perceive the same way, however, is the form of such a statement or proposition – the actual words written or heard – here and now – that is by the people who share that piece of knowledge among themselves. Therefore, if they start analyzing that form in terms of content, they have a chance to come to an agreement, if they also agree on the approach to be applied for such an exercise called semantic analysis.
Semantic analysis or interpreting what we read or hear is not confined to a linguistic pursuit – it is part of our routine of writing and reading too. We mean to say or write something and we look for meaning in others’ communications. The very likely next reaction is a kind of agreement, acceptance, disagreement or refusal to what we have as input. The reasons are numerous, the main point is that we try to incorporate what we like, and „to canibalize” what we do not. And the criteria are usually based on some perceived contrast or contradiction that makes it prohibitive to unite antagonistic content mentally. Antagonistic content may be words, word clusters of any kind and function in relation to an object or a subject in a communication act.
It is a common experience that antagonistic content in the mind is a problem, whether the problem is emotionally, volatively or by reason triggered. But it also should be noted that most of our experience is based on antagonism, real, or imaginary, including that of human life (cognition) itself. We know what we know of life, because it is in sharp contrast with what we knew in the nine months of incubation in the womb – meaning good things and bad things (properties) alike. And our perception of contrast even drives us to experience various visual or audio illusions, which must shed doubts on our capability to act on reason, or call our exercise logical or reasonable all the time. Where or when we claim to practice reason is the spotting or locating „incompatibles”. (Notice the use of where and when as generic identifiers of conditions.)
The Meaning of what?
Meaning is assumed to be an issue in linguistics, more specifically a concept, the definition of which is claimed to be the competence of semantics and pragmatics. This claim may not be acceptable for me as I believe
a) meaning is not confined to a language, in other words meaning is not exclusive to any compositional components of a natural language;
b) when it is confined to a natural language, then meaning is associated with at least TWO specific and individual forms listed
a. as a collection of paired patterns b. split in two “equivalent parts” c. sorted on the side of the shorter patterns (headwords) d. in some (usually alphabetic) order (hence the whole called a code) e. that are not equal (in physical sense, as in a dichotomy and f. where the unsorted side usually takes the form of a definitions or minimum a list of synonyms . These may be collectively called as the context of the object of the meaning sought;
c) pursuant to b) above it is generally agreed that the verbal text (annotation) associated with an entry listed in a dictionary is also called or attributed to be the meaning or a sense of the entry or headword, despite the fact that such texts may take different forms, one of which is called a definition;
d) a code is used to interpret the behaviour of some object, where the interpreters are supposed to agree in that particular code;
e) similarly the behaviour of an object is affected by the interpretation of some object considered to be a code. And there again you have another layer of context. (The concept of context is going to be discussed later.)
In general, humans assume that everything, not just linguistic symbols in this world, has a meaning, or it makes sense (stands some reason). People tend to define the boundaries of the chunk of reality they become familiar with so that they can understand them as messages, and/or they may able to identify an object (in their mind) by using an external reminder or tag.
Should you experience the contrary, then it may be that your knowledge is insufficient to grasp the meaning of some particular chunk of reality, where such a message is harboured. Normally, you may find the beginning or the opening forms of any such pattern sufficient to remind you the rest of the phrase. Of course, the world tends to become more meaningful as you become more familiar with it both in time and in space. But all that is done within your limited range of human perception whether extended or improved by equipment to support all your senses.
It is also a question of debate what the real or true meaning of something is. It is also very common to say that something has no meaning, or makes no sense – in situations when you do not understand something. But it is also possible that you assume or attribute meaning to objects that in fact do not have any meaning in the sense that they do not relate to you at all. Understanding is therefore a match between your external environment and your internal knowledge or “image” of the former in the course of relating your experience and expectations to your perception and thinking. No contradiction or controversy, nothing to upset you as unbelieveable or impossible, etc. just a match that on itself is a source of pleasure and content. An experience that makes sense, hence rational, plausible and also lovable.
Meaning is a Relation
Meaning, on the other hand, - as an operation and a relation - is in parallel with inference. If meaning is a relation, inference or interpretation is also a relation. It is an issue of reflection where someone has an intent, whether explicit or implicit, and which is seen or inferred by somebody else as a “meaning”.
This is why linguists conclude that meaning is a “relational term”, which means that meaning is thought of as a relation. Consequently, it cannot be discussed on its own, you need to identify at least the two components that need to exist for anything to be related in a relation. The word relation is one of the most basic categories used in a natural language to identify the elements of thinking, and I am going to discuss its components in more details later. For the moment it should be sufficient that since meaning is a relation, it is a verb with at least a subject and an object contained in a clause, which should sound obvious. The next question is whether we create or accept the content of such as clause as true and how are we related as observers or critics to such a statement.
So “a means b” is an expression where a subject has a relation (verb – to mean) and an object called b of that verb called meaning. And the whole sequence is composed of three elements: where a is attributed by an external agent to act and result in an object called b (meaning) of that act and where a is said to be the same or equivalent thing as b due to the relation attributed to meaning, the verb – whether shown, indicated or marked, or not.
Meaning is verb
As we said before: meaning is a relation and relations are indicated by verbs . By specifying the collocations of verbs in terms of parts of speech and then in the form of lists containing alternative elements to be used with verbs you get the first layer of context, the immediate neighbourhood of what we believe to have a meaning. The content of the meaning of that form is usually explained as a kind of definition, which serves nothing but to set the boundaries of the object (form) that has a meaning. Having got a meaning explained, you need to specify the next layer of context, which is seen in the people involved in the communication act or situation. You have someone with something in his mind and his effort to put that in words, and at least another person to try to understand that as a message and to make sense of what he receives with the help of what he has in his mind. Then obviously, the next question is what a verb means.
What are verbs good for?
"The meaning of the verb element is multifaceted…." "A verb by itself does no more than expresses an action or a state of affairs. That is how children before age two use verbs. Gone. See. Jump. Catch." "… Bare verbs are ambiguous…" "The main verb gives a verb phrase its content – a specific action or state. It tells us what the verb phrase is “about”." (Crystal) For a verb to be complete you need top specify grammar persons (three) and time (at least three again) in relation to the verb in use in a sentence or proposition.
It is claimed that semantics or semantic analysis is the authority to deal with the issue of meaning, a subject to me apparently not confined to the body of linguistic disciplines. Formal Logic seem to have nipped the term and now considers itself to be solely competent to use it as a part of the AI terminology . I am not going into details of that development for the moment, just add that meaning should not be restricted to the meaning of individual words, especially not that are the part of the predicates.
Linguistics also fails to give you a good definition of what meaning is, because it does not want to deal with issues outside a human natural language, or taking a larger (longer and multiple) unit of analysis into consideration. Pragmatics looks as if it has made up for this shortcoming, but in fact it does not. With the emergence of interest in the connection of brain and language, however, the situation may change. Other traditional fields of study, such as Philosophy (if you like, an extension of Logic, especially that of Informal Logic) as well as Psychology, and within that Cognition in particular, may eventually come up with good suggestions.
The central issue to Semantics, the word meaning indicates two aspects, namely the action that something is intended, and the message that tells the idea or condition what is intended. In other words, it is focused on something not existing in reality as yet, but in the mind of a person or in any tangible form as knowledge representation. Therefore meaning is a mental action or an operation attributable to a living creature as well as the result of such an operation whether expressed verbally or signified otherwise.
As a linguistic as well as a logical category, or an abstraction, meaning is also a relation, a relation between a chunk of reality identified by some linguistically meaningful pattern (and which we attribute a meaning to, or infer that it makes sense), and the context of such pattern that includes the people interpreting such a pattern with the help of their knowledge. But it is not an abstract (intangible) idea or concept frozen without parameters in midair. On the contrary, it is very physical as soon as you make a meaning explicit by writing it down. And being written down it is usually a representation in 2D with the obvious spatial relationship between two parts written down next to one another. Looking at that only, such a spatial relation is self-explanatory, sometimes also marked clearly by using appropriate typography.
But this representation may also be looked at as another kind of (non spatial) relation between the headword and the annotation. Such relation may sound familar from algebra where the relation between the operands of algebraic operations are explained. In my view the properties of algebraic operations as represented in 2D are not really properties, but operations performed on forms (numbers), which, from this point of view, are the same operations as performed on symbols (words). I am not saying though that the result or the outcome of algebraic operations on numbers are the same as the result of the same operations performed on words. But the operations on symbols in space are identical
The main message is that the meaning of a segment in a human language whether spoken/heard or written/read is inclusive of context. Humans seem to have developped speech to mark objects around them and themselves as well by having their voices heard. According to Zipf’s law , the use of vocal cords was the least effort to serve that purpose.
Writing was probably also invented to solve a problem, when individual memory failed to retain the growing amout of auditory input, one use of which must have been to give an account of what was going on around the people.
In order to understand the meaning of meaning you have at least two players to visualize, one who intends to do something (or by paraphrasing it, who acts with an intention, or who is said to act on purpose), and another one who has the chance to perceive or infer such an intent or the act intended.
In addition to those players and their knowledge some other elements are also a part of meaning and they together constitute a context, which does not necessarily end where they stand, and it may not have a definite, clear-cut border either, because our capability is limited on the number of things seen or grasped at a time in our short-term memory.
Therefore meaning as perceived by humans also depends on the person’s individual personality, not just on his race/species and the working of his memory. We are born to look for meaning everywhere and to anticipate the future in accordance with our past experience, and some people do it exceedingly, others very reluctantly. The extreme cases are explained in Psychology. See ideas of reference , for example.
Notice that all that I have been writing about meaning and semantic analysis is nothing but listing words and sentences one by one, and claiming that one group of words is equal to another cluster therefore may be replaced or swapped. In fact, there is not much more that you can do in writing: you put forward some glyphs and assert that they mean the same, which means that they are in a way interchangeable (being identical). Think of algebra here as a good example of illustrating equalities. But there is a little more here than just equalities, you also have the case of identities on hand that are similarly expressed by verbal utterances. The issue is dealt with under the law of identity in logic, which is at the same time a philosophical issue of change and no change.
What are nouns good for?
David Crystal makes very good points on this subject. He points out that naming is not the right word to use to describe the function of nouns, yet he fails to use another word, such as identifying or identification as that concept comes from Logic and Philosophy (see the law of identity) so he concludes:
“Nouns do name entites, but so do verbs name actions and adjectives name atributes…”
Here, should he use the terminology of Philosophy, such as Form and Content consequently, he would see that a concrete word e.g. “a horse” (whether a noun, a verb or an adjective) is a Form, and their Content are the content categories, such as “a noun”, etc. From that on he would have no problem and his next ideas would nicely fit together. This is what he says:
“Nouns convey a specificity of reference which enables us to focus on the subject-matter of the text. When people ask, “What is that text about?, nouns capture the topic more readily than other content words do.
By saying so, he agrees with Mittins, who says that the smallest menaingful unit is word clusters, one of which is called headings, labels and titles. His definition of labels is exactly the same: …
Let us leave messages alone for the moment and have a look at the classification of nouns into common and proper nouns. Listen to what Crystal says about them:
“The function of proper nouns is to identify a unique definite point of reference in the real world. Proper nouns give us knowledge that is encyclopedic, not linguistic… Proper nouns are individual words and proper names are sequences of words. Common nouns may be used as part of the name, which therefore acquires a modicum of lexical content… (example: Oxford street)
These are brilliant ideas except for one, which is so common with linguists that even my former professor L. Antal thought that the same way. ….
Now as a practitioning translator I must refuse that point categorically as false and a pain in the neck. But before I explain my case, listen to Crystal’s points again:
“In a proper name all, some or none of the nouns may be proper…. In each case though the whole name functions as if it were a single proper noun. We learn such names as single units…Proper nouns are important in situations where uniqueness is important.
And uniqueness is important whenever there is a multiplicity (of objects, etc.) out of which a single (one) specimen needs to be selected, hence the population be first sorted. Because sorting is selection. And to sort any groups of objects you need an identity, such as an identifier, more specifically a numeric identifier, which is an integer, most probably a cardinal number. Now you may see what I am going to come out with. Yes, the issue of identity.
Now since tags are no longer single word phrases or glyphs made up in any systematic or regular way to uniquely identify anything - once they are removed from the object, they are to be learnt individually as generic knowledge, and they are to be retrieved as a single cluster composed of elements or forms. In texts however, it is not possible to identify the boundaries of such clusters without semantic analysis, which is not a procedure to disassemble a string of words into single, terminal symbols, but to
1) identify content words and grammar words, and
2) classify the content words into one of three possible classes, notably, objects, relations and properties
3) identify the origin of such classification through operations resulting in membership in a class
Instead of collecting large number of sentences published in various places and collating them, then sorting them into corpuses that are “proven examples of what is found correct” and then be used as templates or solutions, suggestions and references of morphologically well-formed utterances, I recommend a top-down approach just as the vocabulary of a foundation ontology with some difference however.
After clarifying and listing all the semantic primitives to be used in defining the rest of the vocabulary of the English language, I am going to demostrate that by introducing the concept of mental operations new relations may be defined and utilized among words and phrases, not just the spatial relations of collecting words in one location either in 2D or as a chain of data tables on digital media.
Such a chain of tables should be possible to use for the illustration of the growth of knowledge through mental operations. They should be designed with an interface or interoperability between the tables that represent various levels of knowledge (think of a hierarchy or continuum of concrete-abstract and generic-specific clasification of objects). Now for any person being aware of any concept it would be easy to check his depth of knowledge even numerically, if the number of turns are marked and record the path of progression in learning new ideas (complete with their properties, relations, etc.)
It may not be surprising that some word patterns are inclusive of more than one linguistic abstractions. You know that some words belong to different parts of speech, which is one way of looking at them, but not the only one for sure. People like guessing the origin of words, in this case whether a word first meant a verb or a noun, etc.
It says in the Bible that In the beginning there was the Verb. Verb or word is not quite clear, neither are we sure about when and what word we spoke when we learnt to speak as babies. It is also very likely that such initial word had multiple senses or meaning, in other words they were used in several contexts. At that time no differentiation of parts of speech or identities of objects in the real world is clear yet, just an effort to attract attention to some want in the child. Just an indication of relating through acting between a living human and the world, which is the same thing as to say behaving.
Now if it is possible that a word may be used in different situations or contexts and meaning is identical with the form and the context as it is known to a person, then it must be clear that there is a pattern here not just between a sign and a chunk of reality that it refers to, but a wider context as well, which is difficult to define or name, and is therefore usually ommitted. This practice has proven to be plausible, because all these elements are related and are stored as chains of associations. Sometimes it is enough to remember one link in the chain and then you can retrieve the rest one by one.
Therefore, words and phrases have multiple meaning, they are disambiguous by nature. This must be accepted for a fact. This feature is exploited in commercial jurnalism and in fiction. In science it is usually avoided though. But knowledge of multiple sorts connected to a word cannot be forgotten, it can only be marked as valid for the occassion. The obvious antidote is to define each word by using additional words taken to describe the context. Therefore the length of an expression is reversely proportional to the amount of explanation needed due to the different level of knowledge of the context. The more you are aware of the context (including the intent), the less you need to talk or write. And vice versa. In teams at work people may not talk for hours, yet they do their job without any problem in communication.
But the length of any expression or the size of a cluster of words is not affected solely by the explicite knowledge of the context of the subject, which is varied. It is also a question of your expectation and the recall of your past knowledge of the relation between reality and the verbal forms. If you decompose the cluster of words and attribute meaning to each word on its own, you may loose the point, as a cluster of worlds are meant to be used as one, a single semantic unit despite the fact that they are composed of several lexical units. This problem of finding the borders of forms is a generic problem and it is called chunking. The very same thing may be differently dissected yielding different results of significance. This applies to language as well where the problem is that for technical reasons words are used as units for sorting content related to such single forms.
Besides, in a linear sequence or order of words in a text or a sentence some words do not immediatelly follow or preceed the word they are directly related to, for instance as in a tag or title of an object, but give way to other words that are related to the one flanked by them from another point of view. Therefore in chunking and by chunking you basically define the meaning of a message as you know from the classic Roman example. Therefore punctuation, stress and other linguistic or grammar devices do have a serious impact on the meaning of anything said or written – all that lost in stemming, disambiguation or decontextualization for the sake of assembling dictionaries as sorted knowledge representations. By defining a word, by supplying its definition, seemingly we mark its boundaries, but not that of the word (the form), but that of the meaning (content) where other forms are used to limit such boundaries. We also have dictionaries of collocations but no meaning is attributed to the individual varieties of such collocations as they create further problems for establishing the identity of what is signified by words or a natural language.
One of the main issues as well as a cause of frustration is the three laws of thought, namely the law of identity ,http://en.wikipedia.org/wiki/Law_of_identity the law of noncontradiction and the law of excluded middle. Let me focus on the issue of identity and identification as relevant for our subject. Identity and identification is closedly related to continuity. In perception something is identical with itself as long as it is not changed. While something looks the same, thus it is identical with itself, it is also identiified by a tag, a verbal element that can be used "in place" of the original object. Such a verbal element is associated with the chunk of reality and the mental image or memory of the same. These verbal elements are then sorted in pairs with one element on each side claiming that those verbal elements are "identical" for some purpose, hence interchangeable.
The fundamental questions for the mind or reason is how to grasp anything in our environment by calling it a name, by using a verbal identifier, but making also sure that the chunk of reality that we identify by that name is distinct from the rest of the reality. This is to say that it has to have clear boundaries, or a pattern that separates that object from any other object(s) in its surroundings. But since we visually determine such contours of an object in focus, our perception will determine the input for the mind, which processes the visual information by reversing it, by condensing, and altering it to whatever format it needs for storage and recall.
Because of the speed of thought and the mental operations, and also because of the continuum of existence of an object at a given location, it is not clear whether it is the speed of thinking and "refreshement rate", or the very fast oscillation of an object that makes us accept the illusion of "identity" accross time. What to need to define is resolution and tuning to the proper level of perception (scale). We need an analysis of movement just like the one the coaches of athletes practise and combine to make athletes more efficient in doing their sequences of movement later at competitions in real time.
To that end definitions are needed, because we must delimit the borders of our subject/object in order to see the pattern clearly and in order to be able to compare that verbal pattern with our mental patterns that may also be of visual form. I claim that whatever object or subject we are discussing, we normally take that as a whole, a complete thing with the consequence of seeing one thing at a time. Actually, we can see more than one things at a time, but we cannot focus on more than one thing at a time. We are able to see a fairly wide picture of anything in light, and depending on our distance we can identify one or more things with relative precision. But we cannot see object A at distance A sharp at the same time as object B at distance B. Therefore identification is a function of distance of observation and the resolution of our perception, which at subatomic level fails to give us an answer to whether the identity of a very fast moving particle exists.
We are born to be far-sighted, but we learn to use our eye muscles and change focus and many other skills, but we can be easily trapped, if we need to follow a moving object against a background, because we loose control of seeing the background. The common experience is solving the maze type of puzzles where a pencil is needed to be able to follow our gaze, while our wider glance covers the whole area. And this is not limited to just the visual input, because a similar exercise is seen in when monks say their mantras and use a rosemary to keep track of the number of repetitions. It should also be noted that we seem to be limited in our capability to learn a number of new things at a time. We seem to need to proceed in a one thing at a time fashion, and when we look back at the origin or start of anything, we also assume that gradual development to have taken place starting from ground zero.
To illustrate this thought I refer to the exercise of students of Botany and Medicine who have to memorize the native and Latin names of plants and bones respectively as connected to the picture of the items in a book and as they are to be recognized in reality. They report that its is nearly impossible to make note of the two names at the same time, and they would prefer learning the names in one language first as connected to the pictures, then the second set of names.
This suggest that our learning is a sequential process carried out at whatever bandwidth by tallying the objects of the world one by one, that is in temporal order, but since the input is much richer than just one object the way we store that information partly unconsciously and we do not time code our experience of input for synchronisation with others we have a memory of the experience with a random pattern of properties that allow for retrieval by features other than the original label of experience.
The bottom line is that whatever object is out there to perceive and to identify is best seen as having a dual character. It has form and content, both of them need to be grasped and identified by shifting our attention "between them". If we are unable to focus on an object with precision (with exposure long enough to start our mental operations such as isolation and comparison), we see "content", some generic property (i.e. existence and quantity), or we "spot a spot" without specific properties. When we can shift our focus to its form, in other words, we are able to define the boundaries of the object, the pattern, if you like, then we can use that pattern as input to our mental operations to be performed with the objective of identifying the whole object.
The Making of Definitions
Definitions are the most plausible way of defining the boundaries or the domain of a word, concept or category, you name it. Within any one language however all the definitions rely on the same set of words, which are defined by using the most generic concepts in turns. Without that you do not have consistency of the architecture of a language either logically or grammatically. That is what semantics is really about.
Ontologists are very proud of their acheivements, a part of which is the creation of a meticulous system of definition of terms and axioms. Before I discuss any defintions in detail, let us see what any defintion is composed of. Traditionally it is said to display a definiendum and a definiens, both being an object – the question is what they are and how they are related.
Varieties of defintions
The classification of definitions does not seem to feature a singe aspect revealed in the act. I recommend to use as a basis the originator of the definition, who or which is not always traceable. Using John Dewey’s typology definitons may be:
1. denotative, such as a definition published in a dictionary, in which case it is also called either normative or reportative; or used in a document to clarify the meaning of the terms used within the documents, such as a legal contract, in which case it is also called stipulative definition;
2. expository, which is explanatory in character by the way of referencing, illustrating, by taking better known meaning (examples) and associating them, and may display some spatial relations, such as upper lower, opposite terms, antinoms, synonyms, etc.
3. scientific, where conditions of causation, production and generation are described in the form of some nomenclature, taxonomy, classification or thesauri, where it is usually an evolutionary or genealogical definition giving details of species, genus, etc., or the that of the originator (mother-child), or other geneology or descent (history) like in a text book or educational material, where causes, history, purposes, functions, or uses are explained; or in papers of science where a defintion may be logical, theoretical or operational; and
4. ad hoc, where anything, extraverbal (samples, models, analogies, mock up things, etc.) means or media may also be used to define or to identify various objects and properties, by direct referencing.
The boundaries of Self and the difference between subject and object
The theory of personality and its components, the inventory of human or individual personality traits, including their origin and their development is not a properly mapped terrain and is full of contradictions. What seems to be supported by evidence is that there is something in our mind that identifies a person with a number of external objects through relating, through as if type of phantasies, and internal emotions that accompany such relating. These are dynamic and likely to change over time, just as the aspects of reflecting the world as well as one’s own self, as complete with the relevant emotions. Of all that complex reality the point of interest for this paper is that an ego has flexible boundaries and a capacity to reflect the ouside world not just as a simple mirror, but as a generator of a virtual reality where the ego is also looked upon from a seemingly external point of view for self-reflection.
The concept of the ego as experienced by consciuosness is flexible, which means that we can augment the boundaries of our bodies to cover other object that we feel empathy with. Accordingly, we act rationally as if our boders were to be defended by defending other objects, usually our property, realtives, etc. whith what or with whom we identify. Such virtual boundaries seriously affect our sense of reality with regard to the identity of objects despite the fact that we may not admit or realize that our brains work that way. The most common experience is then a switch in aspects or viewpoints where once we are the subject, another time we are the object of the same relation. This is reflection or projection if you like.
Normally the fact that thinking, thought and ideas are a reflection only does not present any difficulties in understanding each other, despite the differences in personal experience. Humans have to synchronize and co-ordinate the language they use to share their experience, intents and knowledge in order to operate and co-operate harmoniously. We are born to this world with the means to live and to reflect upon existence, but we are unable to step outside the system sometimes deemed to be dual, or sometimes triple: consisting of body and soul and mind, whichever is appropriate for doing the work of an intellect called reflective thinking.
Semantic analysis - summary
Semantic analysis is an interpretation of verbal texts made up from „grammar words” and „content words ” with an aim to find verbal units, glyphs or signs that are
a) unambiguous, and
b) have „language independent” (non grammar bound) meaning. Language independent must be something independent of a particular language, hence featuring some meaning that is common to all humans speaking various natural languages.
Meaning, however, does not exist without context, therefore language independent meaning is assumed to be meaning outside the verbal form, or verbal content. In other words, this is a requirement to decontextualize content words and grammar words, and disambiguate any units that are known to be ambiguous (have more than one senses).
The underlying assumption is that a language is composed of symbols, notably terminal symbols that have meaning on their own and are therefore context-free. While this is true of non-natural, or formal languages, which may be described by BNF notation, it is not typical of any human or natural language where parts of speech may be many-faceted. On the contrary, there is a tendency in certain writings in the mass media to use ambiguous words and phrases on purpose, because it is assumed that such ambiguity creates suspension and excitement in the reader and therefore increases readership.
Nevertheless it is understandable that someone would like to confine the meaning of any verbal symbol into a single sense: it would be nice to have a one-to-one relationship that is easy to handle and turn into a machine instruction. This expectation is reflected in the practice of collection of words in dictionaries, yet obviously, dictionaries illustrate that there is no one-to-one relationship between words and meanings. On the contrary, one word may have several meanings, and one meaning may be expressed by several phrases. Moreover, a single word is no longer regarded as the best length or form of an entry in any dictionary any more.
So while meaning must be considered to be incomplete without context, and context must obviously include the people who hear, read or say a word or a phrase, the definition of meaning is not readily available in linguistic literature or elsewhere.
Semantic analysis with respect to a computer means the establishment of the job of a program, the list of instructions to be performed on the respective data in order to make sense of the number crunching.
Meaning is a relation and relations are indicated by verbs . By specifying the collocations of verbs in terms of parts of speech and then in the form of lists containing alternative elements to be used with verbs you get the first layer of context, the immediate neighbourhood of what (messages with verbs) we believe to have a meaning. The content of the meaning of that form is usually explained as a kind of definition, which serves nothing but to set the boundaries of the object (form) that has a meaning.
Having got a meaning explained, you need to specify the next layer of context, which is seen in the people involved in the communication act or situation, because you need grammar persons identified in realtion to verbs. You have someone with something in his mind and his effort to put that in words, and at least another person to try to understand that as a message and to make sense of what he receives with the help of what he has in his mind.
The only objective and tangible entity that they can agree on is the forms presented and they should arrive at an alignment of what they know along those representations of knowledge. In an effort to make such alignments possible you need to bring the individual variations of meanings together in a collection similar to the list of dictionary entries shown after search with onelook.com http://www.onelook.com/. But the resulting list would not be the same, nor would the findings be identical with what you get with Wordnet http://wordnetweb.princeton.edu/perl/webwn.
Traditionally words are attributed to have meanings as most words are used as names (nouns), which are supposed to be “fused” with objects in reality (realia). They are connected to objects as reminders, because objects are too numerous for people to remember. In this case such words are tags, or identifiers which are further completed with other texts or messages and they facilitate learning names (single units) (words). But identifiers again may be incomplete and they represent different levels on the generic-specific continuum.
Those interested in a different type of semantic analysis may find such papers also useful:
Although the premises are similar:
"Humans do not judge text relatedness merely at the level of text words. Words trigger reasoning at a much deeper level that manipulates concepts—the basic units of meaning that serve humans to organize and share their knowledge. Thus, humans interpret the specific wording of a document in the much larger context of their background knowledge and experience."
but the conclusions are different, as you will them see later.
In this lecture I have made a disctinction between objects and subjects where objects are defined as forms, natural as well as man made. Some of the objects are part of a natural language and they are attributed meaning just as other objects that are not a part of natural languages, nevertheless they can be similarly used. Language is used to identify some of the objects that we attribute meaning to with the help of relating them to other objects called context. Humans seek meaning in their environment by looking for a hit and a match in their search to identify what they encounter with in terms of what they already know. For meaningful identification we use word clusters that tell us both what objects are and what they are about. If they tell us what they are about, then they are the content of an object called subject, very often the only and a narrow object in the focus of our interest. For a word cluster to make sense it must be complete with the help of context, and therefore it is not enough to identify objects and properties, we must also discover their relations. Relations are determined by mental operations that are identified by verbs (with orientation as the most important operation done) through the establishment of the identity of objects by word clusters. It is not only the business of identification of objects, which are practically forms, but also their populace or numbers as we need to give an account of what we know. To give an account of objects by using a natural language is not just tallying and tagging what exist, or performing arithmetic on them, but also an issue of sorting those objects. Whatever we seem to identify by elements of a natural language is first found „unsorted” which we try to make sense of by sorting the elements of a language by form and content, which are glyphs indicating objects and subjects. Subject as content is used in our search for meaning and therefore the collection of forms must be made available as a repertory sorted on content too. This repertory exists both in individuals and as common knowledge documented in objects written or otherwise recorded. The content of the two repertories are constantly swapped, exchanged and extended. The purpose of this interaction is orientation at both individual and collective levels which may be improved by harmonization and resolution, allowing that this objective is not universally sought as it is parallelly practised with pursuing controversy, as part of competition for existence by defining what meaning is and what makes sense for humans.
David Crystal: Making Sense of Grammar, Pearson Education Longman, 2004 , London 400 p. ISBN 978-0-582-84863-4
William H. Mittins: A Grammar of Modern English, Methuen, 1962 http://www.chapters.indigo.ca/books/A-Grammar-of-Modern-English-William-H-Mittins/9780416698107-item.html?pticket=crvwera1ija15hzx3tqzfemrR%2fakUBn%2btDeKcF%2fj4y6VPGBbkGA%3d