A hillfort is the easiest archaeological site to notice and one of the hardest to research properly. The ramparts show up beautifully on a winter walk; the story behind them rarely does. The Wessex Hillforts Database was built to close that gap, and it rewards anyone willing to spend an afternoon learning how it works. Whether you are writing a parish history, preparing a walk leaflet, or trying to understand why there are three fortified hills within sight of each other, the records will give you far more than a name and a grid reference.

What the records actually hold

The project behind the database grew out of an English Heritage survey of hillfort interiors in central southern England, published in 2006, with the underlying records and plans archived through the Archaeology Data Service. That origin explains both its strength and its limits. Earlier work on Wessex hillforts concentrated heavily on ramparts and entrances; this survey went inside the defences, recording hut platforms, pits, tracks, later field boundaries and the damage done by ploughing, quarrying and conifer planting.

So expect detailed earthwork descriptions, geophysical survey results, interpretative plans and grid references for dozens of sites across Hampshire, Wiltshire, Dorset, Somerset, Berkshire and the Isle of Wight. Do not expect a complete national inventory, a dating summary you can quote, or photographs of every bank and ditch. The database is one excellent source among several, and treating it that way will keep your writing honest.

Begin with a question, not a search box

Typing a hillfort name into a search field is satisfying but rarely productive on its own. Decide first what you are trying to find out. Three questions work well for local projects: which hillforts sit within a particular parish, ridge or river valley; which share a characteristic such as multiple ramparts or an inturned entrance; and which show signs of activity in a specific period, whether Late Bronze Age, Iron Age or Roman.

Each question shapes your search differently. The first needs place-name and administrative searching, the second needs monument-type filters, and the third needs you to read the finds and dating sections of individual records rather than trusting a summary line.

Search tactics that save time

• Try every name variant. Hillforts appear as Badbury, Bury, Castle, Camp, Dyke and dozens of local names. The place-name element -bury is a strong hint, but not a guarantee, and some sites are recorded only under a parish name.
• Search by parish and district first, then widen to a grid reference if the results are thin. Administrative boundaries have changed, so an old county name may return nothing where a modern district name succeeds.
• Use the monument-type terms deliberately. "Hillfort", "univallate hillfort", "multivallate hillfort", "promontory fort" and "enclosure" are not interchangeable in these records.
• Record what you find as you go. Keep a simple table of site name, grid reference, source, date you checked it and the exact wording of any description you might quote. You will not remember which of six similar records said "probable".
• Note the archive references, not just the web page. If you later need the full survey report or a plan, having the reference number makes the request straightforward.

Putting the records on a map

Grid references are the key that turns a list into a landscape. Convert them to six-figure references if they are given in a longer form, then plot them in a free GIS such as QGIS, or simply in a spreadsheet alongside an Ordnance Survey base map. Seeing the sites together often raises the question that drives a good project: why here, and not on the equally suitable hill two kilometres east?

Two extras are worth the effort. Freely available Environment Agency LiDAR reveals earthworks under woodland and pasture, including features that were invisible to the original surveyors. Georeferenced historical Ordnance Survey maps, hosted by the National Library of Scotland, show how a hillfort looked before twentieth-century forestry and ploughing. Overlay both, and you can often trace lost ramparts and old trackways without leaving your desk.

If you want to test ideas about visibility, a viewshed analysis will show what can be seen from each entrance. Treat the results as a prompt for fieldwork rather than proof of anything, since vegetation and slight changes in ground level make a large difference.

Reading a record like an archaeologist

Terminology in hillfort records is precise, and misusing it in print is the quickest way to lose a knowledgeable reader's confidence. A short glossary helps:

• Univallate, bivallate, multivallate: one, two, or more than two lines of rampart and ditch. "Multivallate" does not automatically mean a later or grander site.
• Contour, promontory, hill-slope: where the defences sit in relation to the hill. A contour fort follows the contour line right around the summit.
• Slight and developed univallate: conventional labels describing scale, not age.
• Annexe, outworks, cross-ridge dyke: attached or detached earthworks that may be contemporary with the main fort, or centuries later.
• Scheduled area: the legally protected boundary, which is not the same as the archaeological site's full extent.

Read records as snapshots. A description of "well-preserved rampart" reflects the day of survey, and a century of rabbit warrens, ploughing or tree roots will have altered it. Where the record says "probable", keep that word in your own writing.

Cross-checking with other sources

Before you commit anything to a final draft, check the county Historic Environment Record, which usually holds more records and more recent amendments than any national dataset. The National Heritage List for England confirms scheduling and gives a concise official description; the Atlas of Hillforts of Britain and Ireland is useful for wider comparison; and the Portable Antiquities Scheme can tell you whether finds have been reported from the surrounding fields. Local archaeological society journals and old excavation reports fill in the human detail, though their dating evidence deserves scrutiny.

Where sources disagree, say so. "The database records two ramparts; the county record suggests three" is a perfectly respectable sentence and often the most interesting one in a local history chapter.

An afternoon that produces something useful

1. Pick one hillfort you can walk to and gather every record for it, including name variants.
2. Note grid reference, monument type, scheduled status and the date of the survey.
3. Plot it against the two nearest hillforts and compare their settings.
4. Download LiDAR for the area and look for features the written record misses.
5. Write a single page of description, clearly separating what is recorded from what you have inferred.

If that page leads to a visit, remember that most hillforts stand on private land, and many are scheduled monuments. Get the landowner's permission before entering, keep to public rights of way where you have none, and leave gates, fences and vegetation alone. Works on a scheduled monument need consent, and metal detecting on one is not permitted without a licence from Historic England. A short email to the county archaeology service before you start will usually save you time later and may put you in touch with someone who knows the site well.

Photo: RDNE Stock project / Pexels