Is there beer in here?

The enigmatic site of Castle Hill has long been a bit of a mystery. The site, near Crewkerne in South Somerset presented a challenge for the team - and not just the task of getting to the top.

The geophys team wasted no time in getting a survey under way and soon had results indicating a large structure built on the summit. But why would a castle be at this location? Local records referred to a castle 'at Crewkerne' - but modern Crewkerne is several miles from our site. Could the records be talking about the same castle? Only one way to find out: get digging.

Phil's first foray into the site involved a long trench over the geophysical survey results. Pretty soon 12th century pottery was appearing along with evidence of a large rectangular structure on top of the hill. Radar results suggested a number of features including a large ditch surrounding the summit, a large stone structure and a peculiar circular feature in its centre. If we were going to understand this site we needed more dating evidence and some idea as to whether the features shown on the geophys were all of the same date.

Meanwhile Alex and Mick began working on the landscape context of the site. The castle mentioned in the records seemed to have been referred to as 'Crow Castle', 'Crook Castle' and even 'Crewkerne Castle' - so which one was our site? Or were they different names for the same place? As day one drew to a close we certainly had what seemed to be a Norman structure on our hill but we were no closer to discovering if it was indeed a Norman Castle...

As day two began the several trenches opened on the hill began to give us some answers - just not the ones we were expecting. Mick was convinced we had a stone building with a timber-framed structure below it. John felt we could have a timber building with a stone foundation. Phil's trench included the hints of substantial stone walls and the strange circular feature on the geophysics - thought to be the base of a roof support - just kept going down. Not at all what we had expected.

Meanwhile Raksha had been quietly uncovering some star finds in from the outer ditch. A fancy clothing pin, and masses of 12th century pottery: just what we needed to back up our theory that we were dealing with a substantial Norman structure.

Just as we felt like we were getting a grip on the site, another theory emerged: could the site be an 'anarchy' castle dating to a period of turbulence in the middle of the 12th century and not one built in the 11th century? This could explain why the castle might have been timber framed, thrown up in a hurry, and why the records are so confused. All the 12th century pottery being uncovered was certainly backing up the theory, but could we prove it?

We needed more evidence from the trenches. Raksha soon delivered, recovering a spur from her trench over the ditch. Could this be evidence of knights at the site? More tentative evidence for the anarchy period perhaps? At least Phil was helping. His trench revealed substantial stone walls: pretty convincing support for the argument that we were dealing with a stone building erected over time rather than a timber one or even a timber building sitting on stone supports thrown up quickly during the anarchy. The significance of the site was growing but what did it all mean?

Day three was our last chance to get the bottom of the mystery of Castle Hill. The only way to show that the site could have been a castle was to prove the geophysics. These results had suggested a large rectangular building. If this was true then Mick would be satisfied that we were dealing with a permanent castle. More trenches! Thankfully Matt, Phil and the archaeological team were able to prove that the building was a large, stone rectangle.

But of course, this being Time Team, Phil decided to throw a spanner in the works. Evidence from our main trench suggested that the castle was never finished. This theory explained the lack of clarity in the records, the lack of stone rubble we had encountered and the small number of finds. So the team had gone from conquest castle to anarchy castle, from timber castle to stone castle... and from castle to, well... unfinished castle. If the structure had ever been finished it would have had three metre wide walls and stood over eighty metres high. Oh, and the mysterious circular feature we'd been looking at: an unfinished well.

The mystery of Castle Hill had been solved.