Somerset County WPA Cemetery Readings

A year on its own means only the year was written down. If you see 1922 with no month or day, the record gave no more than the year. Nothing here has been guessed at or filled in.

A question mark means the person copying the stone was unsure: a worn digit, or a name he could not quite make out. Those marks are kept exactly as he left them.

Ages are shown to one unit: 6 yrs, or 11 mths and 8 days for the very young. Where the stone recorded more precisely than that, the full figure is written in smaller type beneath it, and always in the saved file.

A date such as 19?? means a year was carved but could not be read. Those burials are left out of the year filters and the decade chart, and counted as having no readable year.

The last column, Source page, shows where the original worksheet sits: the page, the grave number on that page, and the file the sheet is in. The link opens the scanned volume at the Internet Archive. The scans there are packed inside a single 230 MB disc image (.iso); open it, then the named .pdf, and go to the page shown.

Red names are buttons that narrow this list. Blue links open another website in a new tab.

In the late 1930s, government workers walked the burial grounds of Somerset County and copied down whatever could still be read on the stones. For anyone who died before 1906, when Pennsylvania began keeping official death records, these pages are often the only surviving proof that a person lived.

They are copies of copies. A stone misread in 1938 is still misread here. Where this index and a gravestone disagree, believe the stone.

The spelling is the transcriber's, not the carver's. The readers wrote down what they made of each stone, and where they knew a family locally they tended to write the name the way that family was known. A stone cut MAST may well be indexed here as Maust.

This is not hypothetical. In the Maust burial ground on the old Jonas Maust farm in Elk Lick Township, one stone is plainly cut LYDIA MAST, yet the worksheet for that cemetery writes the family as Maust throughout, and does not list Lydia at all. Somerset County holds 92 Mausts and no Masts, which is partly true history (the branch that spelled it Mast moved west to Ohio and Indiana) and partly just how local men wrote down a family they knew as Mausts.

So an absence here is never proof. A name you cannot find under one spelling may be sitting under another. Try every spelling you can think of, and use the suggestions that appear under the Last name box.

And many graves were never lettered at all. This index records inscribed stones. The original worksheets also counted graves with no inscription on them, and those people are not here, because there was no name to write down.

That gap is far larger than it sounds. The worksheet for the Maust Burial Ground in Elk Lick counts fifteen graves: six lettered stones, four unlettered adult graves, four unlettered children’s graves, and one more burial known only by family memory. The Jonas Maust farm graveyard is the same: fifteen graves, three inscriptions. In both of those family plots the majority of the people buried there cannot appear in any name index, anywhere, and never could.

Most of these readings stop around 1940, because the bulk of the work was done in the late 1930s. Anyone who died after that is usually not here, and cemeteries still in use kept growing afterwards. Use this alongside Find a Grave, not instead of it.

But not every ground was read at the same time. The 2006 recompilation gathered in later surveys as well — worksheets from the 1950s for cemeteries the original project missed, a recompilation by Della Reagan Fischer in 1977, and at least one ground read as late as 1981. That is why a few burials here run into the 1990s. Just over 900 records post-date 1945, and 92% of those sit in only four cemeteries, chiefly Kingwood I.O.O.F., whose entry still carries “1981 reading” in its name, and St. Mary’s Russian Orthodox.

Not every grave is here. Unmarked and uninscribed graves were often left out. The compilers describe one graveyard with 2,471 known graves that produced only 1,850 records. … of the cemeteries listed have no burials recorded at all.

How old people were when they died

These are the ages recorded on the stones. The first band counts infants who did not reach their first birthday.

How accurate is this copy? … records were read from a stated total of 39,423; about 0.8% could not be read automatically. Two separately ordered copies of the same records were parsed and compared: they agreed exactly, field for field. Ages match the birth and death years given in 97 of every 100 records where both are present.