At Siobhan Davies Studios, a 49-year-old choreography is being rebuilt from memory — and the gaps are the point

The invitation, when it arrived, was deliberately thin. Five choreographers were given a single instruction: rebuild Sphinx, a 1977 work by Siobhan Davies, working only from materials that survived the intervening half-century — a handful of photographs, a few lines of notation, the muscle memory of dancers who performed the original and are now in their seventies. None of the five had ever seen the piece staged. The result, Living Image: Chapter 1, opens at Siobhan Davies Studios in London on 3 July 2026 and runs for five days, in a format that its organisers describe less as a revival than as a controlled act of reconstruction under pressure. The watching is the work.

What audiences are being offered is a rare kind of cultural object: an evening spent watching artists fail, recover, and fail again in public, with the failures left visible. Davies's original Sphinx is the kind of piece that, once gone, normally stays gone — its score, its staging, its contingent decisions absorbed into the bodies of the dancers who performed it and into the archive of those who didn't. The new project treats that loss as material rather than as a problem to be solved. The reconstruction is, in effect, a portrait of the gaps.

A working method, not a museum piece

The conceit is older than it sounds. Restaging canonical dance from fragmentary evidence is a well-trodden practice in the continental European tradition; the Bavarian State Ballet's periodic reconstructions of the Mary Wigman repertoire, and the Rambert Dance Company's long-running project to revive lost British modern works, both operate in a similar mode. What distinguishes the Davies project is the discipline of the constraint. The choreographers are not invited to honour the original, to interpret it, or to write a new piece in dialogue with it. They are asked to rebuild it, with the explicit understanding that what they will produce will, almost certainly, be wrong in places.

That last clause is the structural interest. The project implicitly rejects the museum logic by which an artist's later career becomes a commentary on, and a custodian of, the earlier work. Sphinx in 1977 was a piece by a young choreographer working at the edge of what British postmodern dance could hold. A 2026 reconstruction, even a faithful one, would be a piece by a 2026 choreographer about a 1977 piece. The Siobhan Davies Studios framing — five artists, none of them Davies, working without her direct authorship — is an attempt to honour the original by refusing to impersonate it. The unknowns do the work.

The counter-reading: archive as alibi

The more sceptical read is also available. A reconstruction project, however honest its brief, can function as a kind of cultural alibi: proof that the institution cares about its history, a programmatic gesture toward preservation that generates press and footfall without committing to the harder, slower work of supporting living artists making new work. The five-day run at a small south London studio is, in scale terms, modest — the inverse of the museum-as-monument. But the framing is not innocent. The choice to mark Davies's name in the venue, the choice to lead marketing with the 1977 date, the choice to describe the evening as a review rather than a premiere: each of these small decisions re-anchors the work to the past it cannot quite reach.

A defensible counter to that read is that the dancers who performed the original in 1977 are still alive, and still capable of consenting to be wrong in public alongside the new generation. The piece is, in that sense, a piece about transmission — about what can and cannot cross a forty-nine-year gap between bodies. The alibi reading depends on the assumption that reconstruction is always a form of enclosure. The Davies project proceeds from the opposite assumption: that the original, properly speaking, has no enclosure, only a series of partial resurrections.

Structural context: the dance archive as contested ground

The London dance world in 2026 is, by several measures, the most institutionally crowded it has been in a generation. The Royal Ballet's residency at the Linbury, Rambert's ongoing reconstruction programme at the Barbican, the Place's continued mid-scale programming, and a proliferating fringe of independent choreographers competing for the same small pool of mid-week audiences have all converged on a city in which the contemporary-dance ticket buyer is, increasingly, a curator of attention rather than a passive consumer. The reconstruction-as-event format is a logical response. It offers an institution a way to claim continuity with a canonical past while still doing the work — the labour of rehearsal, the risk of staging — that defines the present.

What is structurally interesting about the Davies project is the absence of the centralising gesture that usually accompanies such programming. There is no house choreographer credited, no single artistic director whose name is on the billing. The five artists are credited individually; the original work is credited to Davies. The institutional credit goes to the studio itself, which has, for the better part of a decade, operated as one of the more disciplined small-scale producing houses in the capital. That institutional restraint is the condition that makes the evening readable at all.

Stakes and what the evidence does not yet show

For Davies, now in her eighties, the project is, in the gentlest possible sense, a form of stewardship — a way of placing the original into other hands before the hands that knew it are no longer available. For the five choreographers, it is a high-stakes professional exercise in a field that rarely offers public rehearsal of its own failures. For the audience, it is the chance to watch a piece that, in 2026, almost certainly cannot be made the same way it was made in 1977, and to watch it being made anyway. The honest version of the evening is that nobody in the room — choreographers, dancers, audience — will know, at the moment of performance, how much of the original has been recovered and how much has been invented. The five-day run, ending on 7 July 2026, will not resolve that question. It will simply make the question visible.

What the available coverage does not specify — and what this publication cannot independently verify — is the precise scale of the surviving source material: how many photographs, how much notation, which dancers from the original cast have been available to the new team. The Guardian's review credits the five participating choreographers by name and dates the run, but does not itemise the archival holdings the project began from. The remaining uncertainty is, in a sense, the project's own argument: that reconstruction is a method, not a guarantee.

— Monexus framed this as a piece about working method, not nostalgia. The Guardian's review leads on the show's difficulty and reward; this publication treats the difficulty as the subject.