Locational and Topographic Surnames: Names Connected to Places

Preston and Lee, Hamilton and Green — when “I live at that place” became “our family is from that place”; what the map can prove, what it can’t, and how to research a name that points somewhere

If Smith is a man of metal and Johnson a man’s son, then a name like Brook or Preston or Hamilton is a man of somewhere. It is the most intuitively obvious surname type of all — and, as this guide shows, the one where a careful editor has to warn readers most firmly: a surname from a place is a claim about geography, not a certificate of descent. Places get shared between unrelated families too.

The crucial split is between two related species, which surname scholars keep distinct because they fail in different ways:

• Locative (toponymic) surnames — from a named place: a town, village, manor, farm or estate. Robert de Preston froze into Preston; the Scottish Hamilton ultimately traces, per the Oxford Family Names project, to an English place-name; Fazackerley can only be from the one village of that name, in Lancashire.
• Topographic surnames — from the feature itself, whether or not a settlement carried that name: a dweller by the hill (Hill), wood (Wood, Woods), stream (Brooks), meadow (Lee, “clearing”), marsh, ford, lane, field or village green (Green).

And immediately, the polygenesis warning that runs through this whole series, from our guides to Welsh Surnames, Spelling Variations, Occupational Surnames and Patronymic Surnames: the Oxford dictionary team’s launch announcement for Family Names in Britain and Ireland (Hanks, Coates & McClure, OUP 2016) notes that the English surname Lee — a locative name from a Middle English word for “clearing in a wood” — has nothing to do with the same-spelled Chinese surname, of which Britain had over 9,000 bearers named Li in 2011, “not counting those who spell it Lee”; nearly half of Lee’s 84,000 British bearers, their research found, are of Chinese ancestry. Same letters; two worlds.

Key takeaways

• Britain today carries some 378,782 surnames with two or more bearers (Richard Webber’s analysis of 1997 electoral data, cited in the FaNUK project papers) — the core of long-established, fairly frequent names is about 43,877, and habitative names are a major slice of it.
• Distribution usually outranks resemblance. The team’s Rochester study found most modern bearers descend from a small Northumberland village, not the great city of Kent; Harmison traces to Hermiston (Roxburghshire), not Harmston (Lincolnshire).
• Medieval place-surnames wore particles — de, del, atte, atten — before the name froze: Robert del Hegges (1301) → Hedges; southern England froze a special -er suffix (Waterer, Forder, Felder).
• Spelling drift can bury the place: Greenwich → Greenidge, Leicester → Lester, Bromwich → Bromage, old Stopford = Stockport.
• Elite retinues carried place-names hundreds of miles — the fully documented Rootham/Wrotham story — so “surname geography” (like Surname Atlas maps) is a starting hypothesis, never a finish line.
• As ever: surname origin ≠ family history. The origin is one event; your line is a thread of documents.

1. Two species of place-name surname — and their neighbours

Locative: “of a place”

The person was identified by the settlement they came from or held. Three sub-classes matter for research:

1. Town and region names. Migrants into the towns were “John de Hereford”, “Emma de Lincoln”; big-city names like London or Lincoln survive as rare surnames. Scots took territory-names as clan-style identifiers — Murray (from Moray), Ross, Lennox — while a wholly different set of Scottish surnames are, as the FaNUK methodology paper stresses, “of midland English origin”: Ramsey, Lindsay, Coventry, Barclay, Hamilton — English places carried north.
2. Manor and estate names. Gentry and lairds named themselves from their holdings: the Welsh gentry pattern we saw in the Welsh guide (Mostyn, Pennant, Powys, Nanney — Powys itself, the old kingdom).
3. Minor place and farm names — the commonest and the most useful: hundreds of unremarkable hamlets, mills and farms that each begat one surname line. Fazackerley (one place, Lancashire); Greatorex — remarkably, from what is now Great Rocks Farm, Wormhill (Derbyshire); Harmison from Hermiston (Roxburghshire); the Scottish Peffer/Peffers, from streams so named in the Lothians, on a Cumbric/Pictish root meaning “radiant” (George Black’s explanation).

Topographic: “by the feature”

No specific place is meant; the label describes the dwelling. From the ONS National Health Service Central Register survey (1991–2000, as tabulated by the Guild of One-Name Studies), the big topographic names rank: Green 19th, Hall 20th (ambiguous — also “servant at the hall”), Wood 26th, Hill 30th, Lee 35th, Brooks 119th, Ford 117th, Wells 137th, Marsh 152nd, Woods 155th, Stone 158th, Lane 174th, Holmes 87th (Norse holmr, “river islet”), Booth 133rd (the cowshed), Riley 153rd (“rye-clearing”), Dale, Sykes (“gully/ditch”), Kirk (“dweller by the church”), Heath, Field, Knowles (“knolls”), Hurst (“wooded hill”), Hedges (from that 1301 del Hegges).

Near-neighbours to keep apart

• Ethnonyms — “where the people came from”: Fleming, Breton/Brett, Scott, Walsh (“the Briton”… or, in Ireland, “the foreigner”). These are origin-nicknames, adjacent to but distinct from place names.
• The southern topographic -er identified by Gustav Fransson (1935): Forder (“dweller by the ford”), Felder, and our case study Waterer — all southern, especially Sussex, Hampshire and Surrey, and all easily mistaken for occupation names.

2. When place-names froze — and how they got scrambled

Place-surnames obey the general English freeze: hereditary transmission broadly between the thirteenth and sixteenth centuries, earlier in towns and among landholding families, later in remoter districts. The byname era shows the particles doing live work — de (from), atte/atten (at the), del (of the) — and a few surnames preserve them as fossils: Noakes (from atten okes, “at the oaks”), Nye (from atte eye, “at the low river-land island”?) — hmm, hold to the safer pair: Noakes (atten oaks) and Byfield/Bytheway-type (skip the murky ones); the certain fossils are de-type names (Preston → de Preston) and Noakes.

Then the scramble. The FaNUK paper’s account of literacy and migration is the best one-paragraph warning you will find: without standard spelling, place-derived surnames drifted to pronunciation, and — crucially — “when a family had moved far enough away from the place of origin for the association to be forgotten,” phonetic spelling simply won: Bromwich → Bromage; Greenwich → Greenidge; Leicester → Lester; Stopford = an old form (not the dominant one) of Stockport. And with universal literacy, some families re-adopted the etymological spelling (Brown/Browne; Wooster/ Worcester), while others preferred difference. Finally, a demographic fact every place-name researcher needs: surname death. Sturges and Haggett’s statistical study (1987) projects that over half of thousand single-bearer surnames would die out within six centuries — and the Oxford team lists extinct habitative names to prove the type’s fragility. The survivors at one locality may be latecomers, not the founder’s line.

3. Why the map beats the dictionary

This pillar’s central method. The FaNUK project describes the principle with unusual clarity: place-names are comparatively stable, surnames are not; therefore “statistical association between surnames and localities needs to become a central component of surnames research.” The lineage of the technique is traceable:

• H.B. Guppy (1890), Homes of Family Names — the pioneer, mapping surnames from farmers’ entries in county trade directories. (Respect him, but note — as in our Welsh guide — his Wales data was crude: two big regional halves only.)
• Steven Archer’s Surname Atlas — computer mapping of every surname in the 1881 census; the FaNUK team credits it with confirming “the essential correctness of Guppy’s hypothesis of a relation between surnames and locations.”
• Richard Webber’s geodemographic work (1997 electoral data; the free gbnames public profiler) — the modern equivalent, and the source of Britain’s surname inventory above.
• FaNUK’s own rule of evidence: “A surname whose association with a particular locality is statistically significant very probably originated there or close by, and this possibility needs to be exhaustively investigated before other possibilities are considered.” Many surnames still cluster near their points of origin — but “the epicentre itself may move from century to century,” hence the standing advice to check continuity across every period, and (ideally) to let DNA adjudicate where two places compete.

4. The traps

1. Resemblance is not derivation. Reaney read the cluster Ravenshaw, Ramshaw, Ramshire, Renshaw… as “dweller by the raven wood”, with several candidate places. George Redmonds, Turi King and David Hey (Surnames, DNA and Family History, 2011) showed that distribution points decisively to a place named Ramshaw near Bishop Auckland, County Durham.
2. The same place-name exists everywhere. Newtons in dozens of counties; Lees in every woodland parish. For topographic names, polygenesis is total — treat them like patronymics.
3. Topographic and occupational blur. Hall = by the hall? or servant of the hall? (Both, in Reaney-type dictionaries — check medieval bearers for your line’s region.) Our occupational pillar’s Waterer case is the emblem: not an irrigation-man at all, but “dweller by the water”.
4. Names cross borders and change citizenship. Ireland’s Norman and English settlement planted English place-names in Gaelic soil: as FaNUK puts it, Walsh, Bermingham, Staunton and Stapleton “must be acknowledged as Irish names despite their English etymons”. A Bermuda… a Bermingham ancestor may be found in Connaught, not Warwickshire. (For Irish transitions, the Tudor Fiants — authorizations for letters patent — are being digitised precisely to watch surnames mid-metamorphosis.)
5. Garbled spellings hide the geography (Bromage, Greenidge, Lester, Stopford above) — your place-name research must ride on the variant-list method of our spelling-variations pillar.
6. “Of” can mean “in the service of”. Rootham’s Bedfordshire tenure came with a retinue, not a relocation of the family estate — see case study 1.

5. Four case studies

1. Rootham ← Wrotham (Kent): the retinue story. The surname’s 1881 association is Bedfordshire, historically also Somerset and Devon. One-name research (cited by FaNUK) resolved it: the name is from Wrotham, Kent, and the geography is explained by elite migration in the twelfth and thirteenth centuries — Geoffrey de Wrotham, a servant in the retinues of successive archbishops of Canterbury; his son William, Sheriff of Devon by 1198; a grandson, Archdeacon of Taunton, influential in developing King John’s navy; and the Bedfordshire link dating from 1206 at Harrold Priory. A complete, document-grade explanation of why the surname sits hundreds of miles from its origin.

2. Laidlaw ← Ludlow. The Scottish surname is a scotticised Ludlow (Shropshire) — and should shock no one, says FaNUK, when Scottish surnames like Ramsey, Lindsay, Coventry, Barclay and Hamilton are likewise “of midland English origin”: the Normanisation of Scotland’s administration under David I (early 12th c.), who took retainers north, opened a long migration of Norman barons and fortune-seekers bearing English and French surnames.

3. Greatorex ← Great Rocks Farm. A whole surname from a farm in Wormhill, Derbyshire — with the geography doing double duty: the obviously cognate Greatrix and Greatrex are preferred in adjacent counties, and Gratrix among emigrant bearers in Glamorgan, South Wales. (Reaney, the FaNUK team notes, cited only a 13th-century place spelling and an 18th-century Berkshire bearer — a demonstration of why continuity-matching matters.)

4. Rochester: Kent’s city versus a Northumberland village. Reaney asserted without qualification “from Rochester (Kent)”. Archer’s 1881 map shows the surname most strongly associated with Northumberland and County Durham, supported by early records of the name there from the sixteenth century — so, the Oxford team concludes, most or all modern bearers probably derive from a much smaller northern Rochester. The moral generalises: the bigger and more famous the place, the more likely you are to assume the wrong one. (A companion case: Dibden — Reaney said Kent; the evidence says Hampshire; and Winship, “from Wincheap Street, Canterbury”, has no modern bearers in Kent at all — 1881 points to County Durham, with bearers like Thomas Wynshop, 1543. Reaney himself called his own dictionary’s medieval evidence no guide to modern lines without checking.)

6. How to research a place surname — in the right order

1. Read the name in FaNUK (2016) before any other dictionary. Each entry gives current and 1881 frequencies, the name’s main region, and an evidence-based explanation — often correcting older ones (Waterer, Rochester, Greatorex are the pattern). Then go to the specialist literature: McKinley’s county surname series (Norfolk/Suffolk, Oxfordshire, Lancashire, Sussex…), Redmonds on Yorkshire, Black for Scotland, MacLysaght and Woulfe for Ireland, Morgan & Morgan and the Rowlands for Wales.
2. Map before you etymologise. Archer’s Surname Atlas (1881), the gbnames profiler (modern), and — for regional detail — the Rowlands’ hundred-level work (Welsh guide). The rule: exhaust the locality suggested by the map before theorising from the word.
3. List every candidate place with that name, county by county, using place-name society volumes (the English Place-Name Society’s county surveys) and national gazetteers; then rank candidates by the surname’s modern epicentre and neighbouring spellings.
4. Match medieval to modern, locally. The FaNUK-standard sources: the Hundred Rolls (1273), lay subsidy rolls, the poll taxes of 1377–81 (Fenwick), Feet of Fines, Chancery proceedings (15th-century ones are among the team’s own digitisation targets at The National Archives), inquisitions post mortem. Forms like de X, atte X, del X in the right county are the gold evidence.
5. Track spelling every step: Bromage/Greenidge-type garbles, -s additions, regional variants (Greatrex/Gratrix). Return to our spelling-variations pillar’s variant log discipline.
6. Test continuity parish by parish, then prove your own line with your family’s own records (per the research workflow of our earlier pillars: civil registration, census occupation and address columns, parish registers, wills, tithes, manorial and estate papers).
7. Where two places compete (Hermiston vs Harmston; Kent vs Northumberland), let DNA and one-name studies arbitrate — the FaNUK team’s explicit recommendation for settled origin questions.

7. FAQ

My surname is Preston/Hamilton/Ashton — am I descended from that town? One of the founding lines probably took the name at one of the many places of that spelling — and which one is a research question answered by distribution and medieval documents, not by the famous place’s glamour. Multiple unrelated founding families are the norm.

What is the difference between “locative” and “topographic” — and why does it matter? A locative name points to one named place (Fazackerley; Rootham); a topographic name to the feature (Hill, Brooks, Green) and therefore has unlimited independent origins. The research methods differ: locatives want place-identification; topographics want cluster-tracing.

Could my “Wood” ancestor have lived by the wood — or the wood be something else? Both, and more: Woods may be an original plural, a del Hegges-type genitive of a feature, or (as our patronymic guide shows in the Maud/Mowat case) a look-alike with an entirely different root. FaNUK’s Green — “a village green” — shows even the “obvious” topographic reading gets refined by modern scholarship.

Why is my place-surname spelled so little like the place? Post-medieval phonetic drift, especially at distance from the origin: Greenwich→Greenidge, Leicester→Lester, Bromwich→Bromage, Stopford=Stockport. Some families later re-adopted the etymological spelling; others didn’t. Build the variant list first.

How can I find which Preston my Preston came from? Map your precise spelling in 1881 (Surname Atlas) and today (gbnames); list every medieval and modern Preston; search the tax-and-property rolls for de Preston in the counties your map favours; then pursue the continuity of your line’s documents. If two candidates tie, DNA decides.

Sources and further reading

• Patrick Hanks, Richard Coates & Peter McClure (eds), Oxford Dictionary of Family Names in Britain and Ireland (OUP, 2016); UWE Bristol launch announcement (17 Nov 2016) — 45,600 surnames; Lee/”clearing in a wood” and the Chinese Li/Lee convergence; FaNUK methodology paper (Hanks, Coates & McClure) — 378,782 surnames statistic; Guppy/Archer/Webber method; Wrotham/Rootham; Laidlaw/Ludlow and the David I migration; Greatorex/Gratrix; Rochester, Dibden, Winship, Harmison; Peffer/Pever; Bromage, Greenidge, Lester, Stopford; extinct habitative names; Sturges & Haggett (1987) surname-death model.
• H.B. Guppy, Homes of Family Names in Great Britain (1890) — with the Wales caveat noted in our Welsh pillar.
• Steven Archer, The British 19th-Century Surname Atlas (Archer Software, 2003/2011); Richard Webber’s surname analyses (gbnames public profiler; the Welsh-diaspora study cited in our Welsh pillar).
• Richard A. McKinley, A History of British Surnames (1990) and county series; George Redmonds, Surnames and Genealogy: A New Approach (2002); George Redmonds, Turi King & David Hey, Surnames, DNA and Family History (2011) — the Ramshaw/Durham finding.
• Carolyn C. Fenwick (ed.), The Poll Taxes of 1377, 1379, 1381; P.H. Reaney & R.M. Wilson, A Dictionary of English Surnames (3rd edn, 1991); Gustav Fransson, Middle English Surnames of Occupation 1100–1350 (1935) — including his southern topographic -er names.
• George F. Black, The Surnames of Scotland (1946); Edward MacLysaght, The Surnames of Ireland; Patrick Woulfe, Irish Names and Surnames (1923) — plus the Tudor Fiants digitisation project noted by FaNUK.
• T.J. & Prys Morgan, Welsh Surnames (1985); John & Sheila Rowlands, The Surnames of Wales (1996/2007); English Place-Name Society county volumes.
• Guild of One-Name Studies — “Top 500 names in England and Wales” (ONS NHS Central Register survey 1991–2000) with the ONS ranks quoted here.
• Cross-links in this series: Welsh Surnames: Patronymics, Spelling Changes, and History; Surname Spelling Variations: How to Search Historical Records; Occupational Surnames; Patronymic and Matronymic Surnames Explained.

Individual place-name surnames — Preston, Hamilton, Ashton, Booth, Sykes, Riley, Greatorex, Rootham and the rest — carry their own profiles on Select Surname List, each with the candidate places, earliest bearers, 1881 epicentre and research notes.