Invocation c. 1889 Fredrick  Lord Leighton (1830-1896)

Definitions of "Clause"

     Nothing demonstrates the chaos in the teaching of English grammar (and the current uselessness of it) more than the lack of agreement on definitions of clauses. Most people, for example, have heard of main and subordinate clauses, and of independent and dependent clauses. Although these four category names are usually used to distinguish two types of clauses, at least one practicing English teacher thought that they refer to four different types of clauses. Grammarians tend to snicker at the teacher's ignorance, but the teacher is not at fault. Many grammarians do not know what they are talking about. Or at least they do not explain it very well to future teachers.
     At the 1998 ATEG conference, I distributed a questionnaire in which one of the questions asked people to underline the words that are in the main clause in the sentence He thought she would make a good president. ATEG conferences are small, and only 21 people filled out the questionnaire. But these were college professors (who teach teachers how to teach grammar) and teachers (who, by the fact that they attended the conference, expressed interest in teaching grammar). Interestingly, four people did not answer this question. Of the seventeen who did, eleven, or 65%, (primarily teachers) said that "He thought" are the words in the main clause. Six, primarily teachers of teachers, said that the whole sentence is the main clause. As one of the linguists noted, "He thought is not a clause."
     What is most interesting, and most important, is the sheer disagreement. It means that English teachers "instruct" students in conflicting ways. A student, for example, may have a seventh grade teacher who, in talking about "main clauses," has in mind simply the "He thought." In eighth grade, the same student may have a teacher who considers "He thought she would make a good president" to be the main clause. Is it any wonder that students -- and teachers -- find grammar confusing, boring, and useless? And what is even more frustrating is that the teachers of teachers seem to be perfectly content with the confusion. Most of these college instructors appear to take the attitude that they want to teach what they want to teach. If it does not help the teachers, even if it confuses the teachers, that is the teachers' problem. Such lack of concern is my primary reason for saying, "Don't leave it to the experts."

     Textbooks are not written by the experts -- they are compilations of what sells. And the textbooks are no more help than are the professors. Probably the most widely used grammar textbook is Warriner's, so for my primary example, I'll use Warriner's English Grammar and Composition, Fourth Course (Harcourt Brace Jovanovich, 1982). Although I have seen more current versions of this text, the basic problems persist, and I want to use this one because it was loaned to me by a local teacher as a sample of the texts that they are using. This text is used with tenth through twelfth grade students. The main problem with all of these texts is that the definitions are poorly thought out and usually illustrated by simplistic examples which (perhaps intentionally?) leave out real examples which don't fit the poor definitions. To be fair to Warriner's, the problems in some other textbooks are discussed at the end of this essay.

     The Warriner's text provides a definition at the beginning of Chapter Four: "4a. A clause is a group of words that contains a verb and its subject and is used as part of a sentence." (81) Here we are told that a clause is "part" of a sentence. But on the following page, after the definition of independent clauses (which we will get to below), we are told "Each of the following sentences is the same as an independent clause. . . ." (82) This tells us, if nothing else, that definition 4a was, at best, incomplete. There we were told that a clause is part of a sentence. Here we are told that it can be the whole sentence. In itself, this is a minor problem, since these special clauses have been labeled as "independent." 
     But it reflects a major problem in whole/part logic. Is a clause a "whole," or is it a "part"? In the Warriner's text, we are told that "A subordinate clause, like an independent clause or a simple sentence, may contain complements and modifiers." (83) This is the first that the student reads about complements in independent clauses, and the examples given all point out complements within subordinate clauses. The fact that some of the subordinate clauses in the examples are themselves complements of the verbs in the main clause is NOT pointed out. In fact, if "He thought she would make a good president." were in one of the exercises in which students are asked to identify main clauses, the teachers' answer key would probably indicate that "He thought" is the main clause. (Thus the response of many of the teachers to my questionnaire.) Simply put, the textbooks are not clear about whether or not the "independent" clause includes the subordinate clauses. As a result, some teachers think they do; others think that they do not.

     There is another major problem that is nowhere addressed in Warriner's. A clause contains a verb and its subject, my emphasis. That should mean that the following sentence contains two clauses:

It has, after all, two verbs, and each verb has a subject: He asked and Mary to help. Those who already understand clauses, in the traditional sense ^{# 1}, (and in the sense used in this textbook) know that "Mary to help" is not the core of a clause. But the students have no way of knowing this. And the only reason that students don't throw the books out the window is that we rarely teach students to identify verbs in the first place. Thus most students couldn't see the problem in the definition, and those who do, have learned simply to grin and ignore it.

     Getting a handle on the problems here is not easy, but let's begin by looking at Warriner's definitions:

"4b. An independent (or main) clause expresses a complete thought and can stand by itself." (81)

"4c. A subordinate (or dependent) clause does not express a complete thought and cannot stand by itself." (82)

We'll ignore the fact that the text strangely pairs "independent" with "subordinate" and "main" with "dependent." The fact that it does so is simply another reflection of the lack of thought that went into the definitions. 
     The primary problem here is that "complete thought" is never defined -- and is probably indefinable. Suppose, for example, we have two sentences: "It is an animal." and "It is big." According to the definition, each expresses a "complete thought." But if we combine them, we get "It is a big animal." We now have the same information, but only one "complete thought." Linguists, moreover, have shown that most adult sentences are massive combinations of simpler sentences. What, then, is a "complete thought"? 
     The textbooks not only never address the question, they also pitifully attempt to address the problem by pointing to some examples that begin with subordinate conjunctions. Warriner's, for example, "explains," by stating that "Subordinate clauses are so called because they need an independent clause to complete their meaning." (82) One of the three examples give is "because no students have applied for them." Students are then told, "Note that each of these subordinate clauses has an incomplete sound when read by itself. Each one leaves you expecting more to be said." 
     Every sentence, when read by itself, has an incomplete meaning:

Out of context, every main clause must therefore be a subordinate clause. And, from the opposite perspective, isn't "because no students have applied for them" a complete thought when read in a transcript of a conversation?

My point, of course, is that the definition of an independent (main) clause as a "complete thought" is meaningless unless one can already identify independent clauses -- in which case the definition is not needed.
     The situation, however, is worse than that. A thoughtful student will actually be misled by the definition. In the sentence "He thought she would make a good president.,' the student, using the preceding definitions, will see two clauses:

Which one of them is "complete"? "He thought" cannot be, for it has a complement [she would make a good president], and complement means "completer." But "she would make a good president" makes perfectly good sense as it stands. Thus, in this sentence, "she would make a good president" is the main clause. Right?
     Some teachers object that "there is an ellipsed 'that' in there." But the objection begs the question. First of all, there is no proof that such an ellipsed "that" exists, and second, they claim the existence of this phantom "that" because they already know what a subordinate clause is! But such foreknowledge is exactly what the student of grammar does not have.  ^#2

     One ought to wonder what the first person to use the terms "independent" and "dependent" in relation to clauses was thinking about -- if he was thinking at all. In terms of meaning, these two terms are used backwards. Let's start with our usual example, "He thought she would make a good president." "She would make a good president" in no way depends on "He thought" for its meaning, but "He thought" certainly depends on "she would make a good president" for its meaning. "Thought," here, does not mean "to think in general." Thus, although "he thought" may "make sense," the sense it makes is not the sense that it expresses in "He thought she would make a good president. In this sentence, "He thought" depends on its complement. Thus, for the thinking students, "He thought" is the dependent clause.
     The same can probably be said for most adjectival and adverbial clauses. In "The man who got away was not the thief," "The man was not the thief" depends on "who got away" to clarify its meaning. Similarly, in "While it is raining, they will stay inside," "they will stay inside" depends on "While it is raining" to clarify its meaning. 
     This problem with meaning does not apply to "main" and "subordinate." "Main" does not connote "separate from," and whatever is subordinate is part of whatever it is subordinate to. In the grammatical sense, every subordinate clause functions as a noun, adjective, or adverb ^#3 within another clause. Thus, in the sense that it is a part of, and has a function in, a clause that is larger than itself, it is subordinate. Main clauses have no such function.

Grammar and Composition. Fourth Edition. Prentice Hall, 1990.

     Once again students find a poor definition of an independent clause: "An independent clause has a subject and a verb and can stand by itself as a complete sentence." (147) And again they find simplistic examples: "The bus arrived."  About subordinate clauses, students are told that "A subordinate clause has a subject and a verb but cannot stand by itself as a sentence. It is only part of a sentence." (148) It apparently never crossed the writers' minds that students might have to deal with a sentence such as "He knew the bus arrived."
     Although the definition of subordinate clauses states that they are "part of a sentence," the text implies that they are not part of the main clause. In the discussion of complex sentences, students are given the example, "This is the street that he describes in the book." (157) Top brackets are labeled to indicate that "This is the street" is the main clause, and that "that he describes in the book" is the subordinate clause. Although this text is used as a resource book for twelfth graders, there is NO explanation or discussion of noun clauses! The omission enables the writers to avoid the question of whether or not the subordinate clauses in "What children see is what children do." are part of the main clause.
     The tone of this text is probably perceived by students as insulting. It treats them as brainless, blank computer disks, waiting to be programmed. Students are told, for example, that "This chapter will show you how to combine subordinate clauses with independent clauses to make complete sentences." (148-149) The writers don't seem to realize that students have been doing this, and generally doing it correctly, since before they even entered school.

Writers Inc. D.C. Heath, 1996.

     Writers Inc is primarily a composition text and does not pretend to be a grammar book. It is highly probable that the only reason that it includes a section on grammar is that the publisher demanded it -- to sell more books. The few pages devoted to sentences and clauses are thoughtless clippings from traditional verbiage:

An Independent clause presents a complete thought and can stand alone as a sentence; a dependent clause does not present a complete thought and cannot stand alone as a sentence. Dependent clauses do, however, add important detail to a sentence. (755)

Writer's Choice: Composition and Grammar. Glencoe, 1993.

     I was loaned the "Teacher's Wraparound Edition" of this text. It is glossy, probably very expensive, and extremely heavy. It also suffers from the same problems as do the texts discussed above. The fragmentary nature of grammar instruction in almost all of these textbooks can be illustrated in this one by the fact that, according to the index, sentence fragments are explained on pages 504-505. But students are "instructed" in how to correct them on pages 378-379.  There is, in other words, no logical sequence to the book.
     In the one of the sections on sentence fragments, students are told  that "A sentence fragment is an error that occurs when an incomplete sentence is punctuated as though it were a complete sentence." (504) I looked in the index, under "sentence," to see if I could find a definition of a "complete sentence." I couldn't." The teacher's edition offers some additional thoughtless advice which is, unfortunately, also offered by many of the grammarians in ATEG:

"A quick test for fragments is to try to turn them into questions. Since fragments lack a subject or a complete verb, they will always fail this test." (504)

(1) I also think that a person's attire should be appropriate for their position. (2) Such as a secretary, who is constantly in contact with customers, should dress professionally. (3) As opposed to postal worker who is rarely seen.

"A main clause has a subject and a predicate and can stand alone as a sentence." (493)
"A subordinate clause has a subject and a predicate, but it cannot stand alone as a sentence." (493)

"In the preceding examples each noun clause forms an inseparable part of the sentence's main clause (the entire sentence)." (501)

A Short Course in Grammar, by Paul J. Hopper. W.W. Norton, 1999.

     Unlike the preceding texts, A Short Course is for college students, and it is specifically about grammar. It is, therefore, the type of book that would be used in a course for future teachers. As in  most grammar books, the material to be studied is chopped up into linguistic units, and little, if any attempt is made to enable students to analyze real texts. Linguistic terms abound, and students get to Chapter Twelve (of Seventeen) before they get to compound main clauses. I include it here simply because of its definition of a clause "to qualify as a clause a phrase must have a verb phrase or a recognizable form of a verb phrase, such as an infinitive or a gerund." (213) As I noted previously, future teachers are often subjected to this kind of confusion in their preparation for teaching. It's no wonder that they hate grammar.
 

NTC's Guide to Grammar Terms, by Richard A. Spears NTC Publishing Group, 1997.

     If I understand it correctly, this book was meant as a resource for teachers -- to help them make some sense of the muddle of grammatical terms that they find in the textbooks. Unfortunately, it only adds to the confusion. Thus, under "main clause," we find "the primary clause in a multiclause sentence; the primary independent clause in a multiclause sentence." (106) Thinking about this stuff is enough to give one a headache, but if one does, one eventually comes to realize that this definition conflicts with all the textbooks (including most of those discussed above) which equate a main clause with a simple sentence. According to this book, a main clause can only occur in a sentence that has more than one clause. The first example given is also interesting:  "I wanted to sing, but I didn't have the voice for it ." According to the text, the main clause is in bold. This means that the "but" does not (as most textbooks claim it does) separate two main clauses.
     If we go by the examples given, this text also suggests (unlike most textbooks) that main clauses and independent clauses are not the same thing. We are told that an independent clause is "a clause that can stand on its own as a complete sentence (Usually used in reference to a sentence containing more than one clause. . . .)" (91)  Among the four examples, one finds "I have several books, but none is by James Joyce." (91-92) Thus, it would appear that in the example under main clauses, "I didn't have the voice for it." would be considered an independent clause, but not a main clause?

     Perhaps the best way to end this survey of the confusion over the definition of "clause" is to quote, in full, Spears' entry under "clause":

any group of words that has a subject and a verb. (Many uses of clause refer to a clause that functions as a specific part of speech or other specific grammatical structure within another sentence. Some grammarians use phrase to include clause. Others use clause as if it had the definition found at phrase. Terms for various types of clauses are additive clause, adjective clause, adverb(ial) clause, causal clause, clause of concession, clause of condition, clause of degree, clause of manner, clause of place, clause of purpose, clause of result, clause of time, collateral clause, comment clause, comparative clause, complement(ary) clause, concessive clause, conditional clause, coordinate clause, defining relative clause, dependent clause, descriptive clause, determinative clause, essential clause, head clause, independent clause, infinitive clause, main clause, nominal clause, nondefining relative clause, nonessential clause, nonrestrictive clause, noun clause, parenthetical clause, principal clause, relative clause, restrictive clause, restrictive (relative) clause, subordinate clause, subordinate clause fragment, superordinate clause, verbless clause, wh-clause.) (27)

     Isn't it time to burn the expensive, confusing grammar books, and start using the free, simple, and systematic KISS approach to clauses?

     The KISS definition of a clause is very simple: A clause is a subject / finite verb / complement pattern and all the words that modify it. This definition avoids all the nonsense about "complete thoughts." Its usefulness depends, of course, on the students' ability to identify subjects and finite verbs, but that is one of the reasons why the KISS curriculum is sequential. Students who spend the grammatical part of fourth, fifth, and sixth grades leaning to identify subjects and finite verbs will have no trouble with clauses in seventh, eighth, and ninth -- for each S/V/C pattern they have found in a sentence, they will have one clause.

Two Objections to the Definition -- Refuted

     Two objections have been raised against this definition, neither of which looks at the question from the students' point of view. The first is an objection against the term "finite." Some grammarians see the term as unnecessary, and many English teachers have never heard of it. But as noted above, without it, "Mary to help him with his math homework" fits the definition of a clause. In terms of function, all verbs act either as finite verbs (the core of a clause) or as verbals (verbs which act as nouns, adjectives, or adverbs). ^#4 When students are learning to identify finite verbs, they do not need to (indeed should not) learn the details of verbals. They do, however, find it very helpful to know that what they are looking for has a name, a name which distinguishes it from those verbs that they are not currently to consider. Here, as so often in the current teaching of grammar, the problem arises because teachers do not have students analyze randomly selected texts. Instead, they use the sentences in the grammar textbooks, and in the textbooks, exercises devoted to identifying "verbs" do not contain verbals. Then teachers wonder why students who did so well on the grammar exercises cannot apply what they learned to their own reading and writing.
     The second objection to the KISS definition of a clause is that some clauses do not have complements. That is true, but the definition does  not state  "a subject / finite verb / complement," but rather a "subject / finite verb / complement pattern." To explain sentences such as "He runs every day," some linguists refer to what they call a "zero complement," i.e., a complement that is a complement, but does not exist. The best way to explain this to students is to use an analogy to the blueprints of a house. Blueprints are a pattern. If someone has blueprints of a house, and the blueprints include a garage, but the person does not build the garage, does that mean that the house is not a house? Whether or not a clause will have a complement depends on the verb. Some verbs require complements, others do not. Students understand this.
     There are two reasons for including the complement in the definition, one logical, and one purely practical. The logical reason involves the subject / verb / predicate noun pattern, as in "Lincoln was president." The traditional bifurcation of a clause into subject and predicate gives us

Such a representation, however, misrepresents the almost mathematical meaning of the S/V/PN pattern:

     The practical reason for including the complement in the definition of a clause involves helping students learn to identify subjects and verbs in any sentence that they read or write (and not just in those sanitized -- and thus sterile? -- sentences in the textbooks). Most people remember being taught how to find the subject of a verb -- "To find the subject of a verb, make a question with 'Who" or "What' before the verb." The problem here is that the process is incomplete. It ignores two additional, necessary rules:

1. Objects of prepositions cannot be subjects of verbs -- ever.
2. The complement of one verb cannot be the subject of another verb -- ever!

Subordinate and Main Clauses

     Most grammar books work backwards, defining a main clause and then defining subordinate. The basic KISS definitions are:

A subordinate clause is a clause that functions as a noun, adjective or adverb in another clause.

A main clause has no such function. It includes all the subordinate clauses attached to it.

1. In some linguistic grammars, infinitive phrases such as "Mary to help" are called clauses. Attempts to import these grammars into the classroom simply add to the confusion.

2. Just to make it interesting, some grammarians exclude the subordinate conjunction from its clause. Thus, in "when it rains," the subordinate clause is considered to be "it rains."

3. Occasionally, of course, a subordinate clause can function as an interjection. 

4. The objection to the term "finite" is often most strongly voiced by grammarians and linguists who fiercely propound the necessity of having students distinguish between "form" and "function." It seems ironic, therefore, that they object to a term which would help students make just such a distinction.