Terraces · Catalonia
Terraces above the vines: what a Catalan marge asks of a Scottish waller
The first time I stood in front of a fallen marge I thought I knew what to do, because I had rebuilt fallen walls before. By the end of the day I had learned that a wall holding back a hill is a different animal from a wall standing in a field, and that I had been looking at it the wrong way round.
Laid 13 July 2025Checked again 7 December 20258 min on the hill
Lift 1 of 4
A gap in the terraces after a wet autumn
The terrace in question sat about halfway up a slope planted with old vines, a couple of hours from Barcelona by train and car, in a district where the hillsides are stepped from the valley floor almost to the ridge. After a wet autumn a stretch about six metres long had bulged and then let go, spilling its face stones and a good deal of the terrace soil into the vines below. The owner, an elderly man who still worked the vines himself, had asked one of the volunteer groups to help, and eight of us turned up on a cold Saturday in November.
The first two hours were spent the way my first morning in Scotland had been spent: clearing, sorting and piling. The difference was the soil. A Scottish dyke that falls leaves you stone and grass. A terrace that falls leaves you stone buried in a slope of wet earth that has to be dug out, sieved roughly by hand and set aside, because it will be needed again to fill the terrace behind the rebuilt wall. By the time we reached the original foundation stones we had moved something like four tonnes of soil, and I had revised my ideas about why terraces are rebuilt so rarely.
Lift 2 of 4
One face, one bank and a different lean
The biggest adjustment was the shape. A Scottish field dyke is symmetrical: both faces lean in towards the centre line, the base is roughly twice the width of the top, and the whole thing stands on its own. A terrace wall has only one visible face, and that face leans back into the hillside. The lean, the batter, does the work that a second face would do in a field wall, because a wall that leans into the load it is holding needs far less stone to hold it. The foundation course is set tilted slightly back into the slope for the same reason.
My instinct, trained on dykes, was to keep the face closer to vertical than the older volunteers wanted. The section leader corrected me three times before lunch, patiently each time, and then made me stand at the end of the gap and sight along the surviving wall on either side. It leaned back further than I would ever have built by eye. It had also stood for longer than anyone could remember, apart from the six metres at my feet, which had failed where a blocked drain had let the terrace behind it turn to mud.
Lift 3 of 4
Building the face and the drain together
The rebuilding followed a rhythm I recognised from home, with one extra step. Each course of face stones was laid with the length running back into the bank and the joints broken over the course below, one over two and two over one, exactly as I had been taught. Behind the face went the reble, packed by hand. Behind the reble went the soil, returned from the heaps in shallow layers and trodden down. The extra step was a small gap left low in the face at intervals, a deliberate opening through which water can leave, which on this terrace had silted up and caused the collapse in the first place.
The stone was a local limestone, pale, sharp-edged and much lighter than the whinstone I learned on. It splits in flat beds and faces up well, but it breaks unexpectedly if you hit it in the wrong place, and my Scottish hammer habits cost us two good face stones before I learned to leave the hammer alone and search the heap for a better fit instead. The volunteers who grew up with this stone rarely dressed anything. They turned stones over, tried them, turned them again and found the one that sat.
Lift 4 of 4
What a terrace wall taught me about looking
By four o'clock the six metres were back to the height of the surviving wall, and the older volunteers were laying the top course, flat stones set to make a firm edge for the terrace above. We did not finish the soil behind it that day. The owner came down from the vines, looked along the wall without saying anything for a long time and then walked the terrace above it, testing the ground with his heel. He said something to the section leader that I did not catch, and she laughed, and I chose to take it well.
What I took home was a change in where I look. On a field dyke I look along the faces for bulges and at the cope for gaps. On a terrace I now look first at the drainage openings at the foot, then at the batter against the wall on either side, and only then at the face. Most of the failures I have seen since, on half a dozen terraces across two districts, started with water and not with the stonework, and the stonework was simply where the failure became visible.
Measurements in this note were taken with a folding rule and a level on the day and copied up the same evening. It is a hobbyist's record, not a specification, and any wall that holds up a path or a field deserves a proper builder.