Ancestral research is rewarding, but record access becomes more and more problematic as pursuit goes on. The more generations that are traced, the more likely it is that ancestors lived far away. The trip to the courthouse that sufficed for recent generations becomes prohibitively expensive for remote ones, a problem magnified by an ever increasing number of ancestral lines.
One alternative is to rent microfilms of original records through the Family History Libraries of the LDS Church. However, at several dollars a reel this can itself become pricey. As a result most researchers gladly turn to abstracted records to avoid the expense as well as time and trouble of consulting the originals.
Whether published in book form, left unpublished in manuscript form, or posted to the web, abstracts are a compact and convenient means of accessing hundreds to thousands of records that might otherwise be poorly accessible. However, with this convenience comes not only the good, but the bad and the ugly.
Abstracters are the saints of the genealogy community. They contribute hundreds of hours of their own labor to save others from a similar investment. Although some obtain compensation from sales of their volumes to libraries and individuals, it’s a safe guess that very few are able to trade their day jobs for full-time pursuit of poorly decipherable script on fading pages.
The good news is that abstracters are often local history experts, having spent many years pursuing their own research in the immediate geographical area of their abstracted material. This generally leads to greater accuracy. Local history experts know which names commonly appear in a given period of time and which do not, with the latter calling for greater scrutiny. They may also have extensive experience with the handwriting of the clerks who created the original records.
As another “good”, did I mention that abstracts save great time and expense? Throughout what follows, this needs to be remembered.
As in any genealogical enterprise, mistakes are made during the abstraction of public records. Some errors are simple mistakes of detail, for example, a misread place name or date. Some have greater impact, as when a personal name is misread. If the given name James was abstracted as Jacob, and you are looking for James, you may discount as irrelevant a record pertaining to your ancestor. Errors in family names are even more impactful, because as a researcher you will likely miss the records completely.
Abstracts of gravestone inscriptions have these same problems, magnified by the weathering of stone. In the 1960s my father and I abstracted the gravestone of a James Bole in Freeport, Armstrong county, Pennsylvania, as indicating that he died in 1854, age 78. Decades later it was discovered that the age was actually 73, as reported in an abstract that had been made in 1903 when the stone was less weathered. Furthermore the younger age was backed by a contemporary merchant’s diary in 1854 stating that James had just died, age 73 . In my experience, and certainly in this one, the digits “8” and “3” are frequently confused. So are “1”, “4”, and “7”, and “5” and “6”. Weathering easily obscures short segments of numbers and letters, making one resemble another.
Another high-impact error is the complete omission of records. This can happen in a moment of inattention during the abstraction of text, or when two pages are turned instead of one.
How bad can it get? Occasionally, an earlier volume of abstracts is so riddled with errors that an entirely new effort is undertaken. One example with which I am familiar is C.G. Chamberlayne’s 1937 volume, The Vestry Book and Register of St. Peter’s Parish, New Kent and James City Counties, Virginia, 1684-1786. One of the motivations for its publication was specifically to correct numerous errors in a 1904 publication, The Parish Register of St. Peter’s, New Kent County, Virginia, from 1680 to 1787.
I reserve the term “ugly” for abstraction errors so bad that they actually change the course of genealogical research. Fortunately these are rare.
Unfortunately one of them concerns my wife’s putative near-descent from Luther Martin (1748-1826), a member of the Continental Congress and U.S. Constitutional Convention, as well as counsel for the defense in the 1807 treason trial of Aaron Burr. Through the Hickey and Robinson lines, my wife descends from a Washington county, Pennsylvania, settler named Zephaniah Martin, widely claimed on web pages to have been Luther’s brother. Both were certainly from New Jersey. In Zephaniah’s case he moved to Pennsylvania about the year 1786, coming from Mendham township, Morris county, New Jersey [2, 3].
Zephaniah’s relationship to Luther seems proved by a will abstract appearing in the New Jersey Archives, partially reproduced here:
1755, July 1. Martin, Benjamin, of Piscataway, Middlesex Co.; will of. Wife, Philerato. Sons — Benjamin, Nathanael, Peter. Daughter, Zerviah, wife of Jeremiah Blackford. Grandchildren– Athanasius, James, and Luther; Jeremiah, Zephaniah, Ruben, sons of Benjamin; Mary, Isaiah and Benjamin, children of John and Hannah Blackford; Benjamin and Nehemiah, children of Nehemiah and Elizabeth Bonham; Zerviah, daughter of Zedekiah and Anna Bonham…. 
From this, family historian Louise Martin Mohler quite reasonably concluded that my wife’s ancestor Zephaniah was a son of Benjamin Martin; a grandson of testator Benjamin Martin and his wife Philoretta; and a brother of the famous Luther Martin. She published an influential book stating exactly that, namely The Martin Family of America, first released in 1983 and then revised in 1987.
However, something about the abstract bothered me. In the intervening years I have forgotten what, but it may have been simply the semi-colon between Luther and Jeremiah. I wrote to Louise, questioning the Luther Martin relationship.
She then did something remarkable. She ordered a copy of the original will — and when the results turned out catastrophic for the Luther Martin relationship, she passed a copy along to me and admitted that an error had been made. Here is the corresponding section of the original will, with all details except the names and relationships removed for clarity:
… I Benjamin Martin of Pisataway in the County of Middlesex and Eastern Division of the province of New Jersey … unto my True and well beloved wife Philerata Martin … son Benjamin Martin … to his son Nathanael Martin … son Peter Martin … Athanasius Martin son of my said son Benjamin … [unreadable] him to Jeremiah Blackford son of John And Hannah Blackford … Grandson James Martin son of my said son Benjamin … Grandson Zephaniah Blackford son of John and Hannah Blackford … my grandson Reuben Blackford son of said John and hannah Blackford … Daughter Zerviah Blackford wife of Jeremiah Blackford … Grandson Benjamin Bonham son of Nehemiah and Elizabeth Bonham … grandson Hezekiah Bonham son of Nehemiah and Elizabeth Bonham … Grand Daughter Zerviah Bonham Daughter of Zedekiah and Anna Bonham … grand Daughter Mary Blackford Daughter of John and hannah Blackford … Grandson Luther Martin son of My said son Benjamin … Grandson Isaiah Blackford son of John and Hannah Blackford … Benjamin Blackford son of John and Hannah Blackford … 
Some of the heirs were renamed later in the document. But despite the presence of a few unreadable words, one thing that is crystal clear is that Zephaniah Martin was not named in the will. It is equally clear that the fundamental error in the New Jersey Archives abstract was that “Jeremiah, Zephaniah, Ruben, sons of Benjamin” were actually Jeremiah, Zephaniah, and Reuben, sons of John and Hannah Blackford. But there was another error as well: “Benjamin and Nehemiah, children of Nehemiah and Elizabeth Bonham”, were actually Benjamin and Hezekiah, children of Nehemiah and Elizabeth Bonham.
As it happened the loss of the connection between my wife’s ancestor Zephaniah Martin and Luther Martin was not completely catastrophic, although that was not known to me in the 1980s. Eventually they did prove to be related. In fact recently acquired records, covered in , have allowed the conclusion that Zephaniah Martin was the son of a James Martin of Middlesex and Morris counties, New Jersey. James was himself a great-grandson of immigrant John Martin (ca 1618-1687), who initially settled at Dover, New Hampshire, before moving to New Jersey. Zephaniah Martin and Luther Martin were 3rd cousins, not brothers.
Without question, abstracts are valuable resources in genealogical research. They make records available that would otherwise be prohibitively expensive and troublesome to collect. In my experience they are far more likely to be correct than incorrect.
The problem is that introducing human intervention between an original and its abstract unavoidably increases the chance of error. An additional factor is that the state of the original itself may be such that readings are ambiguous. That is hinted at by the unreadable words in the Benjamin Martin will, and more definitely by the weathered James Bole gravestone. Ambiguity can likewise lead to error.
As genealogists, the way we should respond is to always maintain a healthy skepticism about abstracted sources. A misplaced semi-colon; a contradiction of other records; an implausibility as to name, time, or place; all of these may be reasons to doubt an abstract.
In such cases we should be willing to go the extra mile, and spend the extra dollar, to seek the original record and determine for ourselves what it truly says. If the original is itself problematic, then perhaps earlier abstracts or other evidence will resolve the problem. For if genealogy is a massive, interlocking, and endlessly fascinating puzzle, we should make sure we’re assembling the right pieces.
 New Jersey Archives, 1st series, v. 32, p. 216.
 Boles, D.B. (2016). The Omnibus Ancestry. Available for download through Lulu.
 Boles, D.B. (1993). Barth-Hickey Ancestry. Troy, NY: Private print. Available from Bolesbooks.
 Original signed will, Middlesex co, NJ.
Picture attribution: Public domain.