About Genealogical Sources


Sattelite imagery of the Middle Appalachians from Philadelphia to Pittsburgh

Before I start to describe the Perrins I want to make some comments about the historical records I have used. Much of what I say here would be old hat to a professional genealogist or historian, but may help others underrstand my reasoning in this genealogy.

Historical Records

Historians speak of primary and secondary sources. The general belief is that as a primary source was written down closer to the event it records, its information is less likely to be distorted. Of course there may be exceptions to this rule, but any time persons compile or edit original records they may introduce bias. In the summary of the types of records below I will try to identify where this happens. The types of records are listed in order of genealogical importance.

Wills and deeds

Governments do a good job of retaining information, particularly when that information aids in taxation. So it is not surprising that the sturdiest American and English documents that demonstrate family relationships are wills. When they are made, wills usually give information regarding family relationships. However, a will may not always mention those family members who have already died or are financially secure, for example daughters who have married or sons who have migrated elsewhere. Estate inventories, as recorded at various times, may also provide hints regarding personal relationships, since friends of the family, neighbors, or persons related to the deceased were typically asked to execute this task.

Property deeds can identify who was where when, although most old deeds don't describe very well where where was. The recorded deeds and surveys at least give the dimensions of the tracts involved. With old topographic maps, a good vector graphics application and some imagination I have been able to place some properties, perhaps even accurately.

Birth, marriage and death

In America birth and death certificates rarely existed at all until after 1880, and not routinely until after 1900. Even when available, a death certificate provides information given by an informant who may be up to two generations removed from the individual who died; as a result inaccuracies can occur. Clergy started recording marriages rather early, but on the frontier marriages routinely took place without clergy. Some states did register marriages early on. Newspaper accounts of births, deaths and marriages were not very complete either until the twentieth century, unless the person was exceptionally notable. Cemetery records, when available, can fill gaps in the death record.

In contrast to America, the British records for baptism, marriage and burial extend back to 1561. Currently the Church of England records are available only in microfiche and are often faded and illegible. The Quakers were much more thorough in their record keeping; they included in their documents such things as guest lists at weddings and attendants at births.


A word about the U.S. census. Started in 1790, these enumerations initially provided only the name of the head of household and the general age and sex of those living with him or her. Starting in 1850, all members of the household were listed, along with precise ages and place of birth. The accuracy of these records depended on the intelligence of both the census taker and whomever provided information to him. I have seen mistakes in the recorded age of up to 10 years from the expected age. Names are often spelled creatively, particularly in the case of families who were not literate. And there are big gaps in census information. The entire 1890 U.S. census was lost in a fire. When the British burned the U.S. Capitol in 1814, census data for many states, including Virginia, were lost. In Allegany County, Maryland, the 1790 census does not exist; similarly the first surviving census for Franklin County, Ohio is 1830, 27 years after statehood. Finally, it is not unusual for families to disappear beneath the census radar when they are no longer prospering.

In the nineteenth century it became common for cities to publish directories, listing both people and businesses. These seem very analogous to the telephone books of more recent times. When available, these records allow for more minute tracing of the people listed, although the caveats mentioned above still apply. The absence of proof is never proof of absence.

Other Government Records

Miscellaneous government records, such as tax assessments in England and early America, may provide glimpses of individuals. Like the early census they only list the head of household. For frontier America cultivated acreage and numbers of livestock may be listed as well. It sometimes is possible to identify when a son in a family reached the age of majority, and therefore is taxed separately.

Sometimes membership lists for churches may establish a person's location. Once in a while other legal matters in court may suggest relationships, family or otherwise.

Family Bibles

Family Bibles can provide fairly detailed genealogical information. Again their accuracy depends upon when the information within them was recorded, as entries may have been made several generations after the fact. Unfortunately it is unusual to be actually able to see the original record. I will mention several Bibles germane to this family history; there certainly have been others judging from the information posted on Ancestry.com and from private correspondence. In the Thomas Perrin of Town Creek section I will discuss a good example of a presumed Bible record.

Historical biographies

One fascinating source of information comes from the late nineteenth century, when it was common for publishers to go to a locale (usually a county) and prepare a book for publication containing biographies with associated genealogies that they collected from the county's residents. Generally it was those who paid who got into such a book. The writers often waxed eloquently about their client's ancestries to the point of absurdity. In this narrative a comparison of the accounts presented here will show widely divergent notions regarding the origins of the Perrin family. Still, after accounting for generation compression and event mis-attribution, these accounts may contain facts which, when corroborated, are useful, and when not corroborated, are tantalizing.

Folk tales

Finally there are personal accounts. These can be difficult to distinguish from legend. I have included these in this genealogy whenever possible, recording them verbatim, as they are most entertaining.

In the novel Mason & Dixon Thomas Pynchon states the importance of such accounts, in a dialog between his narrator (Reverend Cherrycoke) and his audience. (Of course, like most Pynchon writing, it is oblique, and furthermore he obscures the subject with eighteenth century speech.):

"Alas," beams the Revd, "must we place our unqualified Faith in the Implement, as the Tale accompting for its Presence, - these Family stories have been perfected in the hellish Fire of Domestick Recension, generation 'pon generation, till what survives is the pure truth, anneal'd in Mercilessness, about each Figure, no matter how stretch'd, nor how influenced over the years by all Sentiments from unreflective love to inflexible Dislike."

"Don't leave out Irresponsible Embellishment."

"Rather, part of the common Duty of Remembering, - surely our Sentiments, - how we dream'd of, and were mistaken in, each other, - count for at least as much as our poor cold Chronologies."

Modern published genealogies

With the arrival of the internet more people have posted genealogies, either independently, or through commercial organizations such as Ancestry.com. Regardless of their venue, most web-published genealogies do not provide sources for their information. In the case of Ancestry.com, cut and paste copies of genealogies posted by others abound. (Indeed, folks have cut and pasted portions of my Web publications into their own genealogies). Assumptions thus may be repeated to the point of being believed. With only one exception have I relied on this type of information; in that instance there was enough specific data that I could corroborate to make the entire genealogy trustworthy.

Before the internet people privately published their genealogies, and this sort of effort continues to the present. While they vary in clarity or accuracy, the best of them explain their sources and can be extremely helpful.


Several other types of soft genealogical evidence deserve mention.

Family Naming Conventions

Before the nineteenth century, families were more or less predictable in their choice of first names for their children. Typically the names of the maternal and paternal grandparents were used first, followed by the parental names themselves. Naming conventions were rarely followed precisely, but the new appearance of a first name in a generation may argue for an ancestor of the same name somewhere in the past.


During times of frontier settlement families who knew each other from one location often chose to move west together. This history has a number of cases where I know this to be true. There are also instances where co-migration seems likely, and I may rely on such a possibility when making speculations.

Handwriting Analysis

When original documents are available, the presence of a signature may sometimes be of value. On the frontier many were unable to sign their name at all; knowing if this be true may allow one to distinguish individuals who share the same name. Similarly, if a signature is identical to one found elsewhere, it may aid in following an individual on the move. I have seen instances when signatures unexpectedly difffer, such cases require interpretation as well. In this genealogy there may be an instance when an ancestor purposely disguised his handwriting. Finally, it is possible that, due to changes in handwriting style, a later author misread a signature. So. as with every other kind of evidence, critical analysis is necessary.

Linguistic Analysis

Along with handwriting, the spelling of names may provide clues regarding ancestry. For John Perrin or Perins in Maryland there is an interesting discrepancy between how the individual spelled his name versus how the authorities spelled it. In that instance some of the phonetic renderings of his name may actually provide a hint regarding family origins. I expect that this sort of analysis will prove controversial.


My own experience has told me that whenever a genealogy states a fact, but does not include a reference to its source material, I am frustrated. Then it becomes impossible to say how that fact may fit with other information, particularly if there be any inconsistency. To avoid frustrating others I have cited my information as best as possible.

Many times I have not seen the primary sources for material I cite. Sometimes those sources no longer exist. Other times I have been forced (or out of laziness, have chosen) to rely upon the secondary source. In all cases I have documented what source I have seen.

There are a few intentional exceptions to this kind of obsession. U. S. federal census primary source information is now easily obtainable from ancestry.com and familysearh.org, among others. Municipal directory information, particularly from Historic Pittsburgh, is available from the Historic Pittsburgh web site. Some family oral history can not be confirmed in the historical record, but I do try to indicate its source within the family when I cite it.

For general historical statements, I have avoided footnotes.