The original 1830 "as dictated" English text of the Book of Mormon was entirely devoid of punctuation, it seems. That is sort of odd: punctuation is not some optional decoration, it is an essential component of clear written text, establishing the proper relations between various sentences, clauses, phrases, and words that make up the text. If the Urim and Thummim didn't do punctuation, it was a faulty or incomplete translation instrument. So where did the punctuation come from? Meet John H. Gilbert, punctuator and thus unacknowledged "co-translator" of the Book of Mormon.
Here's a paragraph from an article by Robert L. Mathews discussing Gilbert's work:
The first edition of the Book of Mormon was printed in 1830 in Palmyra, New York, by the E. B. Grandin Company. The principal typesetter and compositor was John H. Gilbert, who also provided most of the punctuation and paragraphing. Production was slow and fraught with the possibility of making errors, both of sight and of judgment.
Marquardt and Walters provide more information on Mr. Gilbert, noting that he was 27 years old and working in Grandin's print shop when the Book of Mormon was printed (setting the type, printing and correcting a proof sheet, then printing final sheets). According to Gilbert's own account published years later, he eventually persuaded Hyrum to leave the manuscript sheets with him at night rather than taking them back home at the end of each day; Gilbert then read the manuscript text in the evening and added punctuation in pencil. His recollection: "Names of persons and places were generally capitalized, but sentences had no end."
At first blush, this justifies the wholesale re-punctuating of later Book of Mormon editions — if some guy named Gilbert was the real punctuator, the thinking seems to be, then there is nothing wrong with changing his punctuation choices. The problem with this is that Joseph and his colleagues reviewed the proofs and adopted Gilbert's punctuating. As Gilbert himself explained, "I punctuated it to make it read as I supposed the Author intended, and but very little punctuation was altered in proof-reading."
I suppose one could claim that God worked through Joseph and the Urim and Thummim for the text, then worked through Gilbert for the punctuation. But no no one is making this claim, least of all Gilbert, who had no Urim and Thummim. He just took the sheets home at night and added punctuation, an entirely natural process. He was a printer, not a translator.
The Encyclopedia of Mormonism article "Translation of the Book of Mormon by Joseph Smith," by Welch and Rathbone, contains no references to either punctuation or Gilbert. The article contains this singularly unenlightening statement: "Little is known about the translation process itself. Few details can be gleaned from comments made by Joseph's scribes and close associates. Only Joseph Smith knew the actual process, and he declined to describe it in public." However, the article "Editions (1830-1981)," by Royal Skousen, does state that "Gilbert added punctuation and determined the paragraphing for the first edition."
Finally, to illustrate all this talk about Book of Mormon punctuation with an example, consider the first paragraph of the 1830 edition of the Book of Mormon, or what is now versified as 1 Nephi 1:1-3. In verse 1, the dash after "my days" is now changed to a comma, and a period shows at the end of verse 1 where a comma appeared in the 1830 text. The first word of verse 2, "yea," is now capitalized (since a period was added at the end of verse 1). In verse 3, a comma was removed and "to be true" was conjugated to "is true."
Personally, I kind of like the paragraphing in the 1830 edition. I think it's unfortunate the text was chopped up into individual verses in the 1879 edition done by Orson Pratt. More importantly, the present versification invites misreading, since it presents as complete sentences sections of text which were only parts of longer sentences in the 1830 edition, such as we just observed with 1 Nephi 1:2. In 1879, Orson Pratt also reformatted the text into the shorter chapters we now see in the modern edition. In the 1830 edition, First Nephi, for example, was divided into just seven long chapters.