As repeatedly illustrated in the foregoing, the situation in which we use words and the purposes we use them for often influences their meaning and how it is interpreted. In many sorts of lingual activity words can be employed meaningfully without our being particularly aware of any accompanying concepts. In some cases, words are used without any concept, such as when we shout with sudden physical injury or in sheer joy. Semantic analysis draws attention more closely to the concepts and assertions that underlie language than we may be accustomed to. Thus far its purpose is to improve thought as well as communication. Thought is itself frequently tenuous until adequately expressed, often it is in the expression that we can observe the consistency of thought (or lack of it). Semantic analysis illustrates how language is like a fishing net - the sizes and shapes of its holes and their distribution determine what can slip through and what be caught in expressions. Alter the language net and the meanings it is capable of conveying alter too.

What is true of language is also largely true of thought: our conceptual apparatus is also selective. Some conceptual systems will select certain aspects of reality, different ones will provide different results. In French there are things that can be expressed exactly that cannot in English, and vice-versa. In the many specialised sub-languages, such as that of each of the branches of science or literature, different concepts are expressible. Semantics is also concerned with the comparison of such sub-languages and the problems of to what degree they are compatible with one another or how to 'translate between them.

A difficulty underlying interpretation is that the language always contains a great number of terms whose corresponding concepts are themselves unclear. Where a concept begins and ends will often depend mostly upon a wide context (such as a whole book provides) or on the associations of the interpreter, which are mental and subjective in nature. One way of putting this is 'concepts have fuzzy edges'. Not only do they tend to 'overlap' other concepts (such as the concepts expressed by 'river', 'stream', 'brook', 'rill') but they also are difficult to isolate as concepts distinct from emotions, sensations or intuitions. The distinction between 'emotive' and 'cognitive' is seldom easy to draw with accuracy, for in human intercourse emotion and thought are closely interrelated. Where to draw the line between concepts, feelings, perceptions and sensory things in general is a matter that can only be decided on the strength of the evidence in each actual instance.

The exact meaning of a term is sought in areas like philosophy, the sciences and arts or when we read, write or broadcast with the aim of cognitive communication, but such exactness is seldom a serious problem in ordinary conversation. As Wittgenstein has pointed out in his 'Philosophical Investigations', the degree of precision is not normally problematical in ordinary language usage. We use words roughly and can correct misunderstandings that arise as they become a problem by questioning and discussion. However, misunderstandings occur in daily life without their being discovered, due to what Arne Næss terms 'pseudo-agreement' and 'pseudo-disagreement'. Generally though, precision is an ideal only to be sought by those who require it for definite purposes, people like logicians, mathematicians, scientists, lawyers and the like. There is no reason why concepts or terms should all be precise. If they were so, much of the flexibility of thought and language would be lost. Poetic metaphor would be excluded as well as humorous 'puns' and allusions. A language with scientifically-fixed meanings of great exactness, such as the Vienna positivists attempted to develop, would be no living language, but dry, dead and boring. Rejection of such a 'universal scientific language', as its foremost investigator Wittgenstein himself eventually did, leads to the view that language ought to be many-faceted, just as world culture and life are, which it also reflects.

The verification of meaning is achieved when none of the principles of efficient communication are broken. This further requires that we consider how to decide whether the overall principle is fulfilled in any instance of communication. The overall principle states: A communication is efficient where it is likely that the recipient interprets an expression so as to understand the intended assertion of the communicator. This is the primary criterion upon which each of the five other principles are based and aim to fulfil. They consist in general maxims designed to avoid different sources of misunderstanding which disturb efficient communication. As such they provide techniques which aid interpreters in testing or verifying the intended meaning (or assertion) of various expressions. They make it possible to assess the likelihood of recipient and communicator understanding the same assertion. Note that the word 'likely' occurs in the stated overall principle. This is because one can never prove with certainty that the overall principle is fulfilled in any communication. To do so it would, have to be practicable to compare the intended and derived assertions themselves, without going via any form of expression. This cannot be demonstrated, of course, because assertions only exist as the subjective mental experiences of communicators and recipients. Incidentally, this is not a denial of the possibility of telepathic or other forms of extra-sensory communication. Such direct and immediate communication is well-documented in serious research literature. Yet it is not generally demonstrable, for it is not evident except in uncommon situations, and then only to those to whom it occurs at the moment. Also, despite the certainty with which such apprehensions are experienced, they can frequently be explained by other circumstances.

Agreement and disagreement are also cases in which one can never be definitively certain whether the assertion in question is understood with equal precision by both parties. Subsequent discussion and interpretation of the expression involved may reveal differences or similarities not previously evident. So 'real' agreement and disagreement may not turn out to be real, but only verbal.

Wittgenstein has argued that it is only in practice that the problems of meaning get sorted out and that, by and large, words work well enough for the purposes to which we use them. Certainly, the final test of intended meaning is necessarily empirical and practical. In other words, verification as to whether one has or has not understood an assertion correctly will depend upon the actual situation, persons and communications in each case (i.e. be empirical) and upon whether the purposes of communicator and/or recipient in expressing an assertion are achieved (i.e. are practicable). The methods of practical verification must therefore vary according to the situation and cannot be stated generally or in the form of exhaustive rules. Purposes vary tremendously, as do individuals. The principles explained here are only general rules and, though applicable to written and some spoken communication, are not exhaustive of all means of finding out what anyone means.

Particularly where words do not work well enough for the purposes intended, such principles give some guidelines in penetrating the problem. Also, they direct attention to the fact that words may often seem to achieve their purpose without actually doing so. This may well be far more common than generally suspected. International relations, insofar as they are based on formal written treaties, laws and communiqués are frequently plagued by incompatible interpretations of texts and terms so vague as to produce only pseudo-agreement. Political pronouncements, electioneering promises and ministerial statements are all too often of the same order. Legal documents, such as insurance policies, are almost notorious for their impenetrable 'round-about' wording and their misleading 'small print'. Even in the apparently simplest sorts of verbal situation too, establishing whether an assertion has been correctly understood and the purpose of it achieved can be most difficult when the purposes of the partakers are at odds.

A famous example of Wittgenstein's can be extended to illustrate this point: Suppose a bricklayer on top of a ladder building a wall' asks his assistant 'Throw more bricks when you see I've used up my supply'. If bricks are thrown to him before the supply is used up - or if bricks are instead thrown at him - communication has evidently broken down. There can be any number of reasons why the intended assertion is not understood or why the purpose of the bricklayer was not understood. Maybe the helper did not hear the last part of the question, maybe he did not wish to hear or to help, maybe he was angry with the bricklayer or doing it as a joke and so on. By asking the helper if he heard, what he understood, why he did not follow the request and such things the matter may be cleared up. Or it may not, even if the helper is capable or willing to answer such questions.
In the above example we see how the simplest sort of verbal situation can create problems of fulfilling the overall principle and of how to verify its fulfilment even in simple language situations.

Finally, the view of language implied by the practice of semantics as here demonstrated may be summed up as 'words have both meanings and usages'. Taking this last expression as an example, one may ask what it asserts. Interpreting it as precisely as intended by me, one must refer to the context of the whole section on semantics. A practical test of whether anyone has interpreted such an assertion in its full depth of meaning could, say, be construed in the form of a written examination. Answers to such written examinations do convey a quite thorough verification of efficient communication (as well as inefficient communication, alas). In such a situation the purpose of using an expression in an answer can be quite thoroughly established and defined through the regulations for examinations, through tutorial instruction and reference to the whole field of linguistic analysis in the process.