Thomas M. McCann
J. Sterling Morton High School
A review of research in the teaching of writing indicates that one of the classroom practices continuing to show positive results is sentence combining (Haynes, 1978). It is not clear yet how sentence combining affects the thought processes of students, nor is it clear that the increases in syntactic maturity will be maintained over a long period; but the research in this field suggests that sentence combining is an instructional tool with great potential for the teaching of writing.
One is tempted, then, to link this instructional method with a new teaching tool, but we must first think of the problems one might have in attempting to write a sentence-combining program for the microcomputer. For many programmers, the problem has been the realization that a group of related sentences can be combined in several ways. The problem for the computer is that there is no single correct answer. Unless one manages to identify all the possible combinations and writes these into the program, the computer does not know how to check the student's response so it will know what to do next. Anyone who has prepared sentence-combining activities and listened to students' responses will recognize that it is practically impossible to think of every acceptable combination. Students will surprise you with a perfectly sound combination you never thought of. One might solve the problem by keeping the sentences very simple and allowing for only one possible combination. For example, one might write a program dealing only with adjective embeddings. An adjective from one sentence must be placed before a certain noun in the other sentence. Consider the following example:
This sort of computer program would not be difficult to write; unfortunately, it would also have limited value as an instructional activity. How, then can one prepare a sentence-combining program that allows students to form the mature sentences found in the work of professional writers?
For help in writing a more useful sentence-combining program, we can turn to Francis Christensen. Many teachers are already familiar with Christensen's Notes Toward a New Rhetoric (1978) and the Christensen rhetoric program. The sentence-combining exercises for use with the microcomputer described below are structured around some of the principles in the Christensen program. The following examples illustrate the type of sentence-combining problems that would appear in the program:
The exercises are directed toward subordinating ideas through the use of free modifiers. The kinds of free modifiers found in these exercises are participial phrases, relative clauses, appositives, prepositional phrases, absolute phrases, comparisons, and certain types of adjective phrases. Before the students are turned loose on the computer, the teacher instructs them in all these methods of subordination. It is not essential that students be able to define the grammatical terms. It is essential, however, that students be able to recognize these terms when the computer gives them instructions to perform a specific syntactic operation (BASE CLAUSE, PARTICIPIAL PHRASE). The cues for combining sentences would be removed in some programs after students become proficient in these exercises.
The program is able to provide useful combining exercises because all exercises follow the same general pattern: (1) no words are changed or eliminated in the base clause, and (2) the subordinate idea is changed and attached to the base clause. Since the free modifier is a nonrestrictive phrase or
clause, it must be set off with punctuation when it is attached to the base clause. In completing the sentence-combining exercises, students are directed to follow the order in which the elements appear. If the subordinate idea appears first in the exercise, students must make it an INITIAL free modifier when combining the sentence similarly, if the subordinate idea follows the base clause, students must make it a FINAL FREE modifier. The terms initial and final refer to the position of the free modifier relative to the base clause.
By using free modifiers, which are formed in specific ways, and the method of placement described above the computer can check for the correct response. One could also design the program so that before each problem students would have the option of reviewing a particular syntactic operation. The sentences put into the program can be of various lengths, with several possible free modifiers. The programmer must identify the number of parts separated by punctuation. If, for example, the sentence has three parts, students would in a sense input three answers, each separated by a comma. The computer will check three separate strings rather than one long sentence. Each time the programmer enters a problem, he or she must specify the number of strings to expect in the answer.
The computer program for sentence combining with free modifiers does have some problems. First, the program requires that students be precise when typing in an answer, one extra space will result in a negative response. Another problem is that most students are not accomplished typists. It will take some time before students are familiar enough with the keyboard that they can move rapidly through the exercises. In addition, one criticism of the Christensen rhetoric program has been that there is too much emphasis placed on the writing of narrative-descriptive sentences, when most academic assignments are expository. Many teachers, on the other hand, may see the ability to write mature narrative-descriptive sentences as an important objective in itself.
There are some clear benefits to using a sentence-combining computer program with free modifiers. As with other sentence- combining exercises, students are building a repertoire of ways to subordinate ideas. Moreover, students' syntactic abilities develop in a structured fashion. By using free modifiers, students will explore methods of combining ideas that they never thought of using before (e.g., appositives, absolute phrases). Finally, the use of free modifiers allows the computer programmer to write a sentence-combining exercise in which the user is constructing mature sentences.
Christensen, Francis and Bonniejean Christensen. Notes Toward a New Rhetoric: Nine Essays for Teachers, 2nd ed. New York: Harper and Rowe, 1978
Haynes, Elizabeth F. "Using research in preparing to teach writing." English Journal, 68 (January 1978), 82-88.