Guild of One-Name Studies
One-name studies, Genealogy
Study: Gray   
Variants: Graie, Grey
Category: 2 - A study where research using core genealogical datasets and transcriptions is well under way, but currently in some countries only.
Website: www.gray-ons.org
Guild hosted website: gray.one-name.net
Contact: Mr Christopher Gray
Profile link: https://one-name.org/name_profile/gray/
This study into the name GRAY grew out of a personal interest in Family History and Genealogy, kindled in the late 1990s. Researching my own Gray ancestors turned up far more data than I needed for my own family history — data that seemed too useful to set aside. That surplus became the foundation of this one-name study, registered with the Guild of One-Name Studies in 2008.
The study aims to cover the name world-wide, though the data gathered so far is concentrated on England and Wales, with particular depth in Yorkshire. Progress continues both through my own research and through the many people who get in touch about their own Gray and Grey connections — their questions often lead to new research and new finds.
More recently, the study has been able to move forward more quickly with the help of AI, which has taken on much of the organising, drafting, and day-to-day project management that a study of this size demands — freeing up time for the research itself.
The main variant is GREY, along with its associated forms DE GREY and LE GREY.
Early records also show CROY, GROY, and GRAIE (also written GRAYE) — spelling was not a strong point when surnames evolved in and after the 13th century, and people often wrote down what they heard rather than a fixed form. Some families move between variants, sometimes reverting after only a generation.
One GREY line has been traced back to the village of LE GRAIS in North West France, and there are families with the name GRAIS(?). Whether this is part of the same study is not yet resolved, and is noted here rather than left unexplained.
There are many theories about how the name evolved, and it seems clear there is more than one geographic origin rather than a single source. Two sources I've drawn on give the following:
The Dictionary of American Family Names (Oxford University Press, ISBN 0-19-508137-4):
Some researchers have focused specifically on de Gray and related names, and I have started to look into this myself, though this is not a current priority — a good deal of this ground has already been covered in print and validated as far as the surviving evidence allows. The earliest person I have traced so far is Anchetil de Greye, a vassal of William the Conqueror who accompanied him in the conquest of England — his lineage is covered in the History section.
I have not researched the origins of some other variants, such as Croy and Groy.
The frequency figures I have gathered so far — covering both GRAY and GREY — date back around twenty years, drawn from third-party name-frequency sites and census indexes available at the time. Fuller detail, including the sources and the figures themselves, is on the study website.
A great deal more data is now available online than when this work was first done, and modern tools make it far easier to access, extract and present. I intend to revisit this with fresh research of my own, rather than relying solely on what has already been published.
The distribution work done so far — like the frequency work (see above) — dates back around twenty years, and draws on the same kind of third-party sources: Public Profiler's worldwide "per million" figures, the Surname Profiler at Spatial Literacy in Teaching, the British 19th Century Surname Atlas, and Hamrick Software's US state-level data.
Even with that caveat, some patterns are worth noting. In Great Britain, GRAY has consistently been many times more common than GREY — in 2002, for instance, 67,549 people were recorded as GRAY in England and Wales against 6,518 GREY. The concentration of GRAYs has historically been in the north of England and Scotland — particularly Lanarkshire, County Durham, Yorkshire and Lancashire — which may point to a migration pattern from Scotland, though this hasn't been investigated in depth. In the USA, GRAY numbers rose sharply between 1840 and 1880 before falling back by 1920 — a change I haven't yet looked into.
As with frequency, a great deal more data and far better tools for accessing and presenting it are available today than when this work was done. I intend to revisit distribution with fresh research of my own, rather than relying solely on what has already been published. Fuller detail, including the full country-by-country figures and the underlying sources, is on the study website.
Data collection began with the 1881 UK census index compiled by the LDS — at the time, the only substantial source available. Since then, coverage has grown considerably: UK census data now runs from 1841 through to the 1921 census and the 1939 Register, and further material has been gathered for Australia, Canada, New Zealand and the USA, alongside UK birth, marriage and death indexes. Other countries have not yet been looked at in any depth.
Much of this early work involved harvesting large datasets in bulk. That approach has now largely given way to following specific families or smaller geographic areas as they come up — the big datasets are comparatively easy to access today, so the more valuable use of time is tracing individual lines and localities rather than gathering data wholesale. Much of this tree-building work is done slowly, and mainly in response to questions raised by others about their own family connections.
There is a known gap in GRO (General Register Office) certificate records, arising because the project to provide online GRO indices was never completed. This gap is well understood and is mitigated to some extent by using alternative sources.
The study website lists the major data sources gathered to date, though that list itself is now dated and doesn't reflect everything above — it will be brought up to date in due course.
The investigation into the DNA is at an early stage. I have tested with three companies: