How historians date an old pub

Two questions, not one

Dating a pub takes two kinds of evidence: dendrochronology, the only method that returns a year — it dated the upper timbers of the George Inn at Norton St Philip to 1430–32 — and the licensing record, which in England and Wales runs from the Ale Houses Act of 1551–52 and is the only thing that proves drink was sold. Almost every argument about a pub’s age is really two arguments wearing the same coat. How old is the building? And how long has it been a pub? They have different answers far more often than not, and mixing them up is how a fifteenth-century farmhouse ends up advertising six hundred years of hospitality.

The professional advice on this is blunt. The buildings-history guide to researching inns puts it in one line: “Believe nothing that cannot be substantiated from primary sources.” What follows is the toolkit historians and buildings archaeologists actually use, roughly in descending order of how much weight it carries.

Tree rings: the closest thing to a date

Dendrochronology is the only method that will hand you a year. Every ring in an oak beam marks one growing season, and because trees in the same region respond to the same weather, their ring-width patterns match. By overlapping sequences from timber of successive generations — a process called cross-dating — researchers have built master chronologies running back thousands of years. The British Isles is one of only three regions in the world with a continuous sequence reaching into prehistory, the others being the Northern Alps and the American south-west. That is why English and Welsh pub timbers can be dated so precisely.

The catch is what a tree-ring date actually tells you. To get a felling year you need the outermost ring, under the bark, and most structural timber was trimmed square. Without it you get a terminus post quem — the earliest possible date — plus an estimate for the missing sapwood rings, and then a further allowance for however long the wood sat seasoning before the carpenter used it. Reused timber muddies things further: a beam felled in 1400 can perfectly well be sitting in a wall built in 1600, and old ships’ timbers were reused constantly in port towns.

Even with those caveats, this is the evidence that settles cases. The jettied upper floors of the George Inn at Norton St Philip are dated 1430–32, which is why Historic England regards it as England’s earliest surviving purpose-built inn. The timber frame at the Bell Inn in Nottingham is mid-fifteenth century, a century and a half older than its first documentary mention as a pub, which is exactly the kind of result that keeps a disputed claim alive. At the Scotch Piper in Lydiate, whose sign says 1320, tree-ring work points to timbers of the mid-sixteenth century — old, listed Grade II*, and 230 years younger than advertised.

Documents: where the pub, rather than the building, appears

Paper is the only thing that proves drink was sold. In England and Wales the reliable series begins with the Ale Houses Act of 1551–52, which required alehouse keepers to be licensed by justices of the peace at quarter sessions and to give sureties for good behaviour. Those quarter-sessions records, held in county record offices, run from 1552 until licensing was reorganised in 1888. From 1753 the victuallers’ recognizances were entered into annual registers, and these sometimes name the sign over the door, which is the moment a building becomes an identifiable pub rather than an anonymous house.

Around that spine sit the sources that fill in the gaps:

• Deeds, wills and manorial court rolls for anything before 1552. A bequest of a tenement is often the earliest trace of a house, though it rarely says what happened inside.
• Parish tithe maps of about 1840 and the large-scale Ordnance Survey sheets, which name pubs and fix their footprints.
• Trade directories, which list coaching inns twice — under the street and under the coaching tables.
• Census returns, including a searchable database of English pubs in the 1881 census.
• Newspapers, which are frequently the best single source. A licensing hearing, a fire, an inquest or a sale notice puts a named landlord in a named house on a known day.

The Irish and Scottish pictures are thinner for structural reasons rather than through carelessness. Much of Ireland’s administrative record burned with the Public Record Office at the Four Courts in 1922, which is why the National Inventory of Architectural Heritage — a survey of buildings rather than of documents — does so much of the work on Irish claims.

Carpentry, plan and materials

A buildings archaeologist can date a structure to within a few decades without any paperwork at all, because construction methods changed and are visible if you know where to look.

Roof structure is the single most useful feature. A cruck truss, where paired curved timbers rise from near ground level to carry the ridge, belongs to a medieval and early-modern tradition and is a strong signal of age: there is one inside the Old Bridge Inn at Ripponden, behind an eighteenth-century stone face. A king-post roof, as at the Brocket Arms at Ayot St Lawrence, points the same way.

Framing and jetties. Close studding — vertical timbers set narrowly, using far more oak than necessary — was a display of wealth, and its upper-floor overhang, or jetty, is characteristic of the fifteenth and sixteenth centuries. The carved figure posts and bressumer at the Old Queen’s Head in Sheffield put it at about 1475.

Internal joinery. A plank-and-muntin screen, as survives at the Church House Inn at Harberton, is a Tudor and early Stuart feature. Bar counters, by contrast, are almost always Victorian or later: for most of the history of the English pub, drink was fetched from the cellar and carried to you. The Adam and Eve in Norwich served ale from the barrel until a bar was installed in 1971.

Brick, stone and render carry their own chronology, from brick sizes and bonds to the arrival of sawn rather than pit-sawn timber. A datestone is helpful and not conclusive — Historic England describes the 1636 rebuild at Ye Olde Man & Scythe in Bolton as merely “reputed”, and a stone can be moved.

Marks in the fabric

Apotropaic marks, popularly called witch marks, are carved or scratched at the entry points of a building — doorways, windows and above all fireplaces — to keep harm out. The commonest are the daisy wheel, a compass-drawn rosette, and the crossed V or double-V invoking Mary, Virgin of Virgins. Historic England dates their common use from the medieval period to the early nineteenth century, so they narrow a range rather than fix a year, but they tell you a room was lived in and worried about. There are well-preserved examples by the hearths at the Fleece Inn at Bretforton, where chalked circles were renewed on the flagstones for generations.

Radiocarbon, and the test that cannot be produced

Radiocarbon dating works on organic material and returns a range rather than a year, which makes it blunter than tree rings for a standing building. It also produces the single most instructive cautionary tale in this field. The claim that the Porch House at Stow-on-the-Wold was radiocarbon-dated to about AD 1000 has been repeated for fifty years. When the buildings archaeologist James Wright went looking for the report, he found none published, and the county archaeologist had no record of any samples ever being taken from the building. English Heritage attributes the standing structure to the fifteenth century. We set the whole case out in is the Porch House England’s oldest inn.

The lesson is not that scientific dating is unreliable. It is that a citation to a test is not the same thing as a test.

The continuity trap

Even with a firm building date and a firm document, one gap remains: was this house ever a pub in between? The examples are everywhere once you start looking. The Parrot in Canterbury occupies a genuine fourteenth-century Wealden hall house and has been a pub only since 1987. The Tobie Norris in Stamford has stonework of about 1280, was a bell foundry from 1617, spent the 1950s onwards as an RAF Association club, and became a pub around 2005. The Fleece Inn at Bretforton is an early-fifteenth-century longhouse first licensed in 1848. The Cridford Inn in Devon was a farmhouse until August 1982.

None of those are frauds, and all four are lovely buildings. They are simply old houses in which you can now buy a drink, which is a different and less romantic claim than the one usually printed on the sign.

A short checklist

If you want to test a pub’s date yourself, in roughly this order:

1. Read its listing entry. National heritage bodies describe what they can see and often note the pub’s own claim before contradicting it.
2. Ask what kind of date is being claimed — a building, a licence, a first mention, or a tradition.
3. Look for a named person in a named year. “Licensed to David How in 1716” is evidence; “since 1189” is not.
4. Check for a break. A closure, a change of use, a demolition or a move across the road defeats any claim of continuity.
5. Trace the earliest printed appearance of the date itself. If it first shows up on an Edwardian postcard, you have learned something about 1905 rather than about 1189.

That is the method behind every grade on this site, and it is set out in full on our methodology page. If the vocabulary is unfamiliar, start with inn, tavern or alehouse: what is the difference, and for the most famous English case of a date that will not survive contact with the archive, see the oldest pubs in England.

Found something wrong?Suggest a correction