When part of a book has been set up in type, in what is called "galley form," an impression is taken, technically known as "first proof," and this proof is handed to the proof-reader. This long-suffering individual lives in a chronic state of warfare with the compositors on the one hand and the author on the other. His first duty is to see that the proof agrees with the author's manuscript, that nothing has been omitted, and nothing inserted that is not in the copy. He must see, further, that the spelling, punctuation, capitalization, grammar, and so forth, are correct, and the book set according to the "style" ordered. He first of all, therefore, compares the proof with the manuscript, or an assistant reads the manuscript aloud, the proof-reader listening intently for any variation from the proof before him and marking any errors he may find.
Now this seems easy enough, and if every author prepared his copy carefully, so that there could be no possible mistake as to his meaning, nothing would be easier; but in practice a number of questions arise which would never be thought of by an outsider. On a new work being put in hand, some half-dozen compositors are given a few sheets of copy apiece, and if the proof-reader happens to be readily accessible he is bombarded within the first half-hour or so with, "How am I to spell centre?" "Has travelling one or two l's?" "Shall I capitalize the word State?" "Shall I spell out two hundred?" "Do you want ships' names in Italic?" and so on and so on. As to punctuation, every compositor thinks he knows better than proof-reader and author combined and follows his own sweet will. As every error on the first proof must be corrected by the compositor at his own expense, here arises the cause of war mentioned in our opening paragraph.
Much has been written about printers' errors and the mistakes of "the intelligent compositor." Aside from those caused by illegible manuscript, mistakes arise from faulty "distribution," that is to say, the type has been thrown into the wrong boxes. Thus we get c for e, h for n, y for p, etc., these boxes being contiguous and the letters of the same thickness; if, for instance, the compositor picked up u instead of t the difference in thickness would at once be noticed by him and the mistake rectified. Then letters are sometimes set upside down and we find letters of a different "face" which have got into the case by mistake. In type set on machine, errors arise from striking adjacent keys, or some matrix will stick in the channel and make its appearance later, sometimes even in the next line. But the chief source of error is illegible or carelessly prepared manuscript, and to the author's slips of the pen must be added in these days the slips of the typewriter.
It is quite possible for a man to be an expert in astronomy, medicine, or natural history and yet have hazy ideas on spelling and punctuation. "When in doubt use a dash" is an old standing joke, but some authors use dashes all the time, making them do duty for commas, semicolons, and periods. They will write indifferently 4 or four and frequently their capital a's c's, m's, and n's cannot be distinguished from the small letters. They will commence a story telling that the "Captain" did so and so, and lo, on the next page the "captain" sinks into a common noun; and so with "Father," "mother," "Aunt," "uncle," etc. Just see what the story would look like if set according to copy!
Now the proof-reader is expected to rectify all this, thereby drawing on his head the wrath of the compositor, who says "he followed the copy," and occasionally incurring the wrath of the author as well for departing therefrom. Sometimes instructions are given that the author's spelling, punctuation, etc., are to be carefully followed, when of course no question can arise; and the proof-reader will query on the proof submitted to the author anything which does not seem to him to be correct.
The great newspapers and magazines have what they call a "style sheet" for the guidance of their compositors and proof-readers and insist on its being faithfully followed. Only by this means could uniformity in the appearance of the paper be secured. In this style sheet careful and minute directions are given for the use of capital letters, the use of Italic, spelling out of numbers, compound words, etc. In the Government printing-office in Washington they have a style book of some two hundred pages. Some book printing-offices have what they call "the style of the office," which will be followed if no instructions are received from the author to the contrary, while some publishing houses with connections in England insist on English spelling being followed in all their books, as books with American spelling will not sell over there.
Here is an outline of an "office style":—
"Spell and divide words according to Webster's dictionary.
"Capitalize President and all Secretaries of State, Senator, Congressman, Governor, Government (of U.S. or other country), King, Emperor, Republican (and all political parties), all pronouns relating to the Deity, Legislature, State, Nation, Street, Avenue, (Hudson) River.
"Use small capitals for B.C., A.D., A.M., and P.M.