Why We Punctuate; or, Reason Versus Rule in the Use of Marks

by William Livingston Klein

20. I have an arrangement to do a serial for Harper’s, and a series of wayside pieces for Scribner’s, Smith illustrating, is on the tapis.

Probably very few persons, in reading the above sentence at sight, would take notice from the first comma that the sentence is here divided into two shorter sentences (the grammarians call these sentence-parts clauses). In the absence of such notice, the reader goes on to “is” before discovering the real relation of the group of words following “and.” The comma before “and” does not clearly show that “and” connects the two larger groups of the sentence, and so gives notice to the reader that the first group, which is one of the clauses, is complete; nor, in the absence of such notice, is the reader told that the words immediately following “and” look forward, instead of backward, for their completion as a group.

The confusion or uncertainty of grouping is here further increased by the character of the two groups between which “and” appears to stand. Each begins with the same word (a), thus making them appear to be coördinate groups; and each group appears to be the object of “do.”

Perhaps the author would contend that the comma before “and” is sufficient to give notice of the proper grouping, just as we used a comma for a somewhat similar grouping in Sentence 1-1. Such contention would not be without merit; but Sentence 1-1 is much shorter, and the consequent liability to make the wrong grouping is much less. Our discussion might thus end in a difference of opinion without determining the degree of separation requiring a mark of higher rank than the comma. The discussion could be opened by an admission on our part that a semicolon in No. 1-1 would be better than the comma; for its warning of the change of grouping would be unmistakable by any reader.

Two reasons may be given for the use of the comma in No. 1-1, instead of the semicolon; and we give them in order to emphasize the fact that we cannot always have in language one degree of separation that calls unmistakably for the comma, and another degree that calls unmistakably for the semicolon. The reasons for the use of the comma in No. 1-1 are as follows:

1. As the sentence is very short, the eye readily catches the relation requiring the grouping that carries “their mothers” forward for its connection, instead of backward to “children.”

2. The use of a semicolon in this sentence might seem to justify a rule requiring a semicolon in every sentence composed of two clauses, while convention hardly justifies such punctuation. On the other hand, if we prefer the use of the semicolon in No. 1-1, under what conditions would the comma be preferred? If what immediately follows the conjunction between any two clauses, especially short ones, does not suggest connection with what precedes, a comma before the conjunction gives sufficient notice of the grouping of the language into two clauses.

Thus we have forced upon us, at least apparently, the necessity of making choice in many sentences between the comma and the semicolon. The problem is further complicated by the need of the different classes of readers for whom the marks in language are used, and still further by convention.

The proper punctuation of No. 20 is as follows:

20-1. I have an arrangement to do a serial for Harper’s; and a series of wayside pieces for Scribner’s, Smith illustrating, is on the tapis.