Richard Bushman describes how the Book of Mormon might have been translated.
Richard Bushman, Joseph Smith: Rough Stone Rolling: A Cultural Biography of Mormonism's Founder (New York: Alfred A. Knopf, 2005), 71-73
Neither Joseph nor Oliver explained how translation worked, but Joseph did not pretend to look at the "reformed Egyptian" words, the language on the plates, according to the book's own description. The plates lay covered on the table, while Joseph's head was in a hat looking at the seerstone, which by this time had replaced the interpreters. The varying explanations of the perplexing process fall roughly into two categories: composition and transcription. The first holds that Joseph was the author of the book. He composed it out of knowledge and imaginings collected in his own mind, perhaps aided by inspiration. He had stuffed his head with ideas for sermons, Christian doctrine, biblical language, multiple characters, stories of adventure, social criticism, theories of Indian origins, ideas about Mesoamerican civilization, and many other matters. During translation, he composed it all into a narrative dictated over the space of three months in Harmony and Fayette.
Composition is the naturalistic explanation for the Book of Mormon-the way books are always written-but it is at odds with Joseph Smith and the historical record. The accounts of the neighbors picture an unambitious, uneducated, treasure-seeking Joseph, who had never written anything and is not known to have read anything but the Bible and perhaps the newspaper. None of the neighbors noted signs of learning or intellectual interests beyond the religious discussions in a juvenile debating club. To account for the disjuncture between the Book of Mormon complexity and Joseph's history as an uneducated rural visionary, the composition theory calls for a precocious genius of extraordinary powers who was voraciously consuming information without anyone knowing it.
The transcription theory has Joseph Smith "seeing" the Book of Mormon text in the seerstone or the Urim and Thummim. He saw the words in the stone as he had seen lost objects or treasure and dictated them to his secretary. The eyewiynesses who described translation, Joseph Knight, Martin Harris, Oliver Cowdery, and David Whitmer, who was in the whouse during the last weeks of translation, understood translation as transcription. Refering to the seerstone as a Urim and Thummim, Knight said: "Now the way he translated was he put the urim and thummim into his hat and Darkned his Eyes then he would take a sentance and it would apper in Brite Roman Letters. Then he would tell the writer and he would write it. Then that would go away the next sentance would Come and so on."
Joseph himself said almost nothing about his method but implied transcription when he said that "the Lord had prepared spectacles for to read the Book." Close scrutiny of the original manuscript (by a believing scholar) seems to support transcription. Judging from the way Cowdery wrote down the words, Joseph saw twenty to thrity words at a time, dictated them, and then waited for the next twenty to appear. Difficult names (Zenoch, Amalakiah) were spelled out. By any measure, transcription was a miraculous process, calling for a huge leap of faith to believe, yet, paradoxically it is more in harmony with the young Joseph of the historical record than is composition. Transcritpion theory gives us a Joseph with a miraculous gift that evolved naturally out of his earlier treasure-seeking. The boy who gazed into stones and saw treasure grew up to become a translator who looked in a stone and saw words.