I recently finished writing a book, “Legendary Locals of Beacon Hill,” to be published in the fall by Arcadia Publishing. This is no literary tome, but since they asked me, rather than the other way around, I was happy to do it. It’s a picture book with accompanying stories about the men and women who have influenced my neighborhood from colonial times until now. I already know there are 11 important people I’ve left out. There are probably similar books for all the downtown Boston neighborhoods in process right now.

I hope the one I did is accurate.

For what I’ve learned in doing the research for “LLBH” is that many books I consulted had glaring errors. Dates are wrong. In one venerated history, the aunt of one historian is confused with the aunt of another historian, making the whole event a muddle. Oliver Wendell Holmes, the doctor, is sometimes confused with his son, Oliver Wendell Holmes, the jurist. The whole story of an event is rarely told. Authors leave out certain people or incidents, sometimes for space considerations, and that can leave a story skewed.

The worst books may be accurate, but how could you tell, since they are unreadable. Several academic books about Beacon Hill suffer from this problem. I had been looking forward to reading them and was disappointed when the writing was so bad. One author forgot that it helps readers if paragraphs contain a topic sentence and the rest of the sentences in the paragraph are about that topic.

Which brings me to academic writing. There is no such thing – or there shouldn’t be. Academic writing should display all the characteristics of any good writing. Even the footnotes should be well-crafted. David McCullough’s books are a good model for other historians who think they should put something in writing. The Harvard professor, Jill Lepore, is also notable for delectable prose. I wish she had written something about the history of downtown Boston instead of many of those who did.

So if academics are mostly unreadable, and others are inaccurate, you’re usually told to go to primary sources. The pitfalls here are sneaky. Census takers make mistakes. Letter writers exaggerate or lie. Land records are incomplete. Birth certificates are wrong.

An example—my own mother’s birth certificate when we had to order it after she died. It showed a different birthday than the one we celebrated. And she had a different name.

Interviews are a problem. Three people tell the same story, but they emphasize different points, they mistakenly locate an incident on one street when they meant another, they confuse a date or a name.

Another pitfall is bias in an author. Newspaper reports are especially suspect. You may remember a Cambridge news story about which reporters jumped to a wrong conclusion. But their bias stuck. Near Harvard Square was a private school. It wanted to expand. Its student body was composed of many black children. The neighbors, including a prominent law professor, complained about the expansion. The Boston Globe reporters pegged the neighbors as white and rich. So they merrily reported that the neighbors were opposed to the expansion because they were bigots. The neighbors, tarnished and powerless now, ate crow, bending over backward to help the school expand.

Anyone living in Boston should know how ridiculous this story is. First, all Cambridge private schools have many black students, and they are welcomed. Second, anyone who has ever lived next to an institution knows that neighbors would care little about the student body. But they will go through the roof at expansion plans—traffic, noise, etc. etc. Those Cantabrigians were not bigots. They would have reacted the same way if little British Windsors were the students. But it will ever be wrongly reported that the famous law professor and his neighbors are bigots.

So what is a writer of history to do? Compare everything is a good start. Look at the primary sources and look at others that might substantiate or refute them. If you’re dealing in secondary sources, read many of them to see where they agree and conflict. Keep an open mind.

Finally, rely on Wikipedia. It gets a lot of grief from the academic community, but it is the only place I would ever read that a theory was disputed, or that academics didn’t agree on something. It is not complete but it was better than many books, whose authors’ egos sometimes get in the way.

As for me, I’ve checked and rechecked sources. When I could, I asked knowledgeable people to read parts of the book. If I’m writing a column, I can always correct a mistake the next week. A book lingers. I’ve got my fingers crossed that I’ve followed my own directions for how to write a book that readers can mostly believe.