| Paratactic paraphrase has its greatest value as an aid in helping the analyst see more clearly the relationship that might hold between the dependent element and the rest of the sentence. Grammatically dependent structures are hypotactic. Here we introduce three kinds of dependent clauses and investigate their paraphrastic analysis in a cursory manner. This will give a flavor of the analysis involved in the rest of this study. |
Paraphrasing a temporal clause.The sentence in (1) contains a dependent clause of time. |
| (1) | The performer hardly sees the crowd, while he is performing. |
| Now, to construct a paratactic paraphrase the analyst following the first principle of paraphrase reformulates the sentence so as to express its content in two co-ordinate and grammatically independent sentences. The first paratactic sentence expresses the import of the dependent clause as an assertion. Usually a faithful paraphrase will require that the predicate be marked as non-factual with the auxiliary may. The second sentence recasts the main assertion of the original sentence in such a way that there is a pointing or referring back to the first paratactic sentence or to some part of that sentence. |
| (1P) | a. | The performer may be performing at some time. | |
| b. | He hardly sees the crowd then. |
Temporal anaphora.This reference back from the second sentence to the first is anaphora. The adverb then in the first paragraph (1Pb) means at that time and refers to the same time as the phrase at some time does in (1Pa). In other words the analyst has paraphrased the adverbial clause of time in (1) as a separate assertion. In this paratactic analogue there is a reference in the formerly dependent clause to an indefinite time, and a reference in the independent clause back to the same time. Presumably the hypotactic structure with the conjunction while obviates the need for any such explicitly anaphoric devices. |
Possible semasiotactic structures.Paraphrase helps the analyst get at the possible semasiotactic structures in the sentences content. Ideally the only thing that changes is that a covert proposition whose truth may or may not be implied in the original is expressed as the assertion of a possibility. |
| The expression is a semasiological statement expressing a good part of the meaning of while. Both sentences contain adverb phrases (AVP). The logician might describe their reference to a common time as: |
| (AVP1) = (AVP2) = |
Paraphrasing a condition clause.Consider some further examples of sentence paraphrase. Here is a sentence containing an adverbial clause of condition: |
| (1) | No one can come in unless one pays first. |
| The analyst may also reformulate this particular sentence paratactically with explicit anaphora. |
| (1P) | a. | One should pay first. | |
| b. | One may not pay in some case. | ||
| c. | Then one cannot come in. |
| The adverb then in (1Pc) means in this case and refers to the same condition as the phrase in some case does in (1Pb). |
Paraphrasing an adjective clause.The sentence in (1) contains an adjective clause. The analyst may paraphrase this sentence as the two sentences in (1P). When we introduce adjective clauses, we divide the restrictive kind into classifying and identifying. In this example the clause is serving to classify a man as one of a particular kind as shown by his/her characteristic behavior. |
| (1) | That man is most honored who honors others. | ||
| (1P) | a. | A man may honor others. | |
| b. | That (kind of) man is most honored. |
A place for complexity.Good English style encourages a healthy mixture of sentences structured both paratactically and hypotactically. In parataxis the assertion of a rather complicated thought may take the form of several independent sentences as in (1). The number of sentences depends on the level of analysis the linguist chooses to stop at. I have taken one of the sentences (by John Borroughs, cf. exercise 13-3.20) from Kimball 1900, and paraphrased it into nine assertions and a question. I then altered most of the content words so as to express the content of Parnias Paradox without, however, altering their grammar and syntax to arrive at (1P): |
| (1P) | a. | The body may die. | |
| b. | The soul may continue to live after that. | ||
| c. | This is a common belief among the religiously inclined. | ||
| d. | It probably has some foundation in fact. | ||
| e. | Some people have been revived. | ||
| f. | Such people sometimes vividly remember events after death. | ||
| g. | So something continues mans consciousness. | ||
| h. | What is it? | ||
| i. | Yet scientists cannot explain it. | ||
| j. | Yet philosophers cannot explain it. |
| Depending on the authors audience and disposition he might well combine these nine sentences hypotactically into a single sentence. The syntax of English comes equipped with a number of different devices for formulating certain clauses as equivalent in meaning to sentences standing in different relationships. Here is what an author does in mixing different kinds of hypotaxis: |
| (1) | The common belief among the religiously inclined, that the soul continues to live after the body dies, probably has some foundation in fact, though scientists and philosophers cannot explain what it is that continues mans consciousness so that sometimes people who have been revived vividly remember events after death. |
Diagramming complex sentences.In an attempt to capture the syntax of complex sentences, I developed a method of using colored circles with lines joining their key parts as given in figure 1 for the sentence above. There will be more of this kind of diagram, which is intended to be pedagogical. Desiderata include strict formality, yet difficult challenges arise in using them consistently. |

| It may be helpful to follow the system of MultiNet. Figure 2 takes only the main relations between sentences, their nominalizations, and other elements relating the sentences hypotactically. As demonstrated previously, a logical form may derive directly from this part of the semantic net. Some of the objects derived from situations are placed in bright yellow boxes as semantic features. These are the ones that seem to be directly related to the use of abstract objects to represent the situations that are structured hypotactically. |

| Perhaps a few comments are in order beyond the definitions found in the glossary. The theta-role which I call TOLATIVE is called SCAR in standard MultiNet meaning it is the passive carrier of a state, the entity that sustains it. The distinction between this and the active AGENTIVE is sometimes neutralized as in cling on to, where both relations appear to hold simultaneously. The distinction between this kind of carrier and an EXPERIENCER is in the former being more clearly passive. ADVERSATIVE is called OPPOS as the entity opposing another. As a standard relation I place a situation node upon it to make it the focus of the main assertion of the sentence. In MultiNet there is no use for a CONJUNCTIVE, because its use is seen as fully syntactic. The semantic net implicitly represents the conjunction of all the semantic elements on it. However, there are two conjunctive functors on the pre-extensional level: *ITMS for unordered and *TUPL for ordered elements of a set. The function *ITMS-I is defined as the counterpart of *ITMS on the intensional level. MultiNet defines only eight DERIVATIONS (called CHxx) which relations have a major syntactic component. This makes them useful for treating the lexical change of a particular part of speech when it changes to fit the structure of another. As mentioned before, the MultiNet strategy is not to change the categories of entities but rather to expand the sort signatures of the appropriate relations. |