There were many attempts to overcome inconsistencies of the method of members of sentence. One of the most systematic approaches is known as the valence theory. It is usually considered that the founder of this theory is the French scholar Lucien Tesniére.Analysing structure of sentences in Russian and French he came to the conclusion that the meaning of a sentence resembles that of a drama, the verb in this theory being the predicate is the name of the action of a scene, while the nouns are names of the participants or actants. Alongside with the action and the participants a sentence may include indications to the circumstances of the action. If so, the structure of the sentence depends upon the action described in it. Since the verb is the name of the action the structure of the sentence depends upon the lexical meaning of the verb. The verb determines the number of the participants (the nominal elements) and the formal indicators of their functions. This ability of the verb to predict the number and the functions of the nominal elements of the sentence was named valence of the verb. The outcome of all this is that a sentence might be reduced to its predicate, which is most often a verb. Different semantic types of sentences are isolated on the basis of verbs having different valence.
Lucien Tesniére insisted that valence of verbs may vary from zero to three. So we may have four types of verbs and four semantic types of sentences.
Verbs with zero valence usually name natural phenomena like It snows, It is cold, etc. The appearance of zero valence verbs and the examples given by Lucien Tesniére are very significant. It is natural that being a student of Russian he isolated zero valence verbs having in mind Slavonic structures like Ñâåòàåò, Õîëîäíî etc. But he illustrated these verbs not only with Russian sentences but also with their translations into French. The French sentences had the subject a sort of phoney actant which is predicted by the verb. And yet Lucien Tesniére considered them as zero valence structures because in them the subject il (as in English it) does not name any real thing. Here we see a very important step toward constructing a radically new theoretical framework for describing sentence meanings — the sentence is independent of its form. What in Lucien Tesniére’s theory seemed to be an occasion became a postulate in some later theories in sentence meaning.
One valence verbs usually name states, processes or general activity, as John is clever, A candle burns, A dog runs, etc In other theories these types of meanings are not only divided but opposed. Some of the authors of other theories criticise Lucien Tesniére and his followers on this point. Still if we take into consideration the number of participants he is quite correct.
Two valence verbs usually name relations and actions as Jane has a cat, John reads a book. At this point again Lucien Tesniére and his followers were severely criticised by representatives of other semantic theories. But we have to take into consideration that the valence theory only counts the number of necessary nominal elements.
And finally, three valence verbs name actions that are performed with the use of instruments or materials as John cut meat with a sharp knife. At this point the valence theory might be criticised because actions presupposing three elements do no always presuppose material or instrument, for example John gave Jane a book, still all these verbs presuppose that a sentence should have three nominal elements.
After Lucien Tesniére the valence theory was developed by German and Soviet linguists mostly. The first step in its development was differentiation between those nominal elements that have to be used in a sentence (they are named obligatory arguments) and those that might appear in a sentence (these were named facultative or optional). In the example given for three valence verbs the third component (with the knife) was considered to be optional because it can easily be omitted without radical change of the correctness of a sentence.
The next step was closely connected with the first. Introduction of optional elements permitted to find verbs having more than three nominal elements associated with the verb, for example John sold Mike his car for a good price. The number of actants (arguments) increased in some descriptions to seven. That produced certain difficulties in classifying sentences because some actants (arguments) could be not only omitted (as optional), but cannot even appear in the sentence if other are used. For exam[ple: if the sentence has an actant with the function of an instument it cannot include an actants with the function of a medium and vice versa. The difficulty was in the fact that at the earlier stages the arguments were numbered, but not named. As a result we may receive descriptions with such statements: the verb is two valence and may have actants (arguments) 1 and 3 as The wind tore the roof off.
That made the linguists substitute numbers by names indicating the function of the thing named by the noun in the situation described by the sentence.
Further development of the valence theory was caused by observations of actual co-occurence of verbs and nominal elements of sentences. Already Lucien Tesniére suggested that there might be discrepance between the form and syntactical function of nominal elements. He showed it by insisting that french sentences with a formal impersonal il should be treated as having zero valence. In this case we may observe that the form of a sentence ( two-member sentence with the subject) does not coinside with its semantic interpretation. Further analysis has shown that actants functioning as patients or instruments may have different formal representations in th sentence, for example such verbs as read and wait demand different formal structures of the nominal elements (read sth, wait for sth), yet the nouns following these verbs must be treated semantically as patients. It was found that some verbs demand that adverbial modifiers of place or time should be in the sentences. Take for example such verb as to reside, sentences with which are impossible without a prepositional phrase.
To cope wiyh all theseproblems and inconsistencies scholars imploying the valance theory for semantic interpretation of sentences split the notion of valence into three different notions: formal, or syntactical valence responsible for the formal features of the nominal elements of the sentence (this notion coincided with government of traditional grammar), logical valence responsible for the number and semantic functions of the nominal elements (this notion coincided with the idear of valence defined by Lucien Tesniére), lexical valence responsible fo limitations of lexical classes appearing in the positions predicted by the logical valence (this notion coinsides in its force with the notion of selection categories imployed by N.Chomsky in his generative grammar).
The third type of valence needs special treatment. Lexical valence meas that if we take such verb as to walk we expect that in the position of the first (agentive) actant animate nouns should be used, e.g. Mike walked, A cow walks, etc But it’s hardly possible that inanimate niuns can be used in this position, e.g. The house walked or An orange walks seem hardly possible. Yet, these restrictions might be broken on two occasins. Restrictions on lexical combinability may be broken in metaphoric use of verbs and nouns. In poetic texts we may find sentences stating that moutains, houses, islands, trees and other inanimate things are described as walking. This ability of metaphoris use can not be explained by the valence theory and is usually considered as a specific feature of human speech generally.
Restrictions of lexical combinability may be also broken when no metaphoric senses can never be traced. If we take such verb as to pour we expect that in the position of the agent a name of a human being should be found and in the position of the patient – a name of some liquid: Jane poured the tea into our cups. But we know that the first position, the position of the agent of this verb can be filled in with a name of a liquid:Water poured into the basin. It is important to note that in this occasion not only lexical valence but also syntactical and logical valences are altered. Moreover certain lexical features of the verb are changed – the verb becomes intransitive. English demonstrates a wide range of verbs that may have two types of lexical valence each of which modifies syntactic and logical valence and yet naming virtually the same process. If all transitive verbs in English could be used like that we would simply say that this is a specific feature of the English language. But very many verbs can not be so easily made intransitive. If we take such verbs as to go and to build we see that sentences The bridge builds or The books bring are incorrect.
The examples given in the previous paragraph can lead us to two conclusions. We may first state that though the three types of valence are autonomous they are not absolutely independent of one another. Secondly, we have to consider the valence theory only as a descriptive but not explanatory one. As we shall see it below there are statements that are at least partially true for two other theories of the semantic sphere of sentence.