Fragment No. 70

Working from the Back

Severely damaged during the Second World War, the historic Dragon Carpet survives in pieces; its contemporary replica was cut into 100 slices and now circulates globally in distributed form. I received Fragment #70.

At the start of the project I struggled to identify its location on the carpet, to hold it in my mind’s eye as I switched between the views of the original and its mirrored reproduction.

I have limited visual recall; I can only hold an image in my mind as a 2D picture of something — and not a high res photo either. It’s more like a fuzzy, faded photo print from the 1970s. To really see something, I have to know it through handling it, conserving it, reconstructing it, protecting it. Just like if you put together an artwork reproduction as a jigsaw puzzle over several hours or days, you know it more intimately than from seeing the original in person for a few minutes.

Now as I write up my experience, I have been able to absorb each knot of the design and can immediately pick out “MY” fragment on the overall carpet. I recognise its shape in black and white, back or front, mirrored, oriented east, west, north or south.

For me, the naming of each section as a “fragment” in this project implies incompleteness and dispersal. The fragmentation of the historic carpet found an echo in the fragmentation of my own working hours. I began this project in the margins of domestic time, following the birth of my second child: in the car during my eldest’s football practice, late at night when the house was quiet, in brief intervals between work and childcare. These conditions have always influenced needlework: produced in intervals, within domestic constraints. Realising that my own working pattern lay within this continuity made it feel less like a constraint and instead I tried to embrace it as a chosen methodology.

My fragmented time was complemented by the fragmentation of my work spaces, which I then linked by journaling this project on Instagram whenever I had the opportunity. For several years, Instagram functioned for me as a kind of distributed studio. No matter where in the world I was working, I would come back to my tiled Instagram grid where I could assemble the “bigger picture” of my research, where I could document processes, analyse projects, and vitally, engage with other textile conservators and makers across continents. Over time I built a space which made my work visible as a whole. I have previously described that network as an informal community of practice: iterative, reciprocal, sustained through shared documentation of making. I previously also wrote about social media as a tool for cultural collaboration, and how one of the things that appealed to me was the use of hashtags to create an open source, gridded visual archive, where anyone with an interest in the subject could add their content. It gave me great pleasure to click through and see everything displayed neatly in modular fashion – from the mundane to the specific – whilst allowing for an unlimited multiplicity of viewpoints.

Over my years of scrolling, the spatial logic of Instagram’s grid has given way to an algorithmic feed. Now it’s a vertical stream where the conversations and context which gave the images their meaning dissolve, and the pictures themselves are increasingly filtered and, ironically, “optimised”. What use is a photo of a stained or crumpled textile where the filter has smoothed and blended away the very thing I’m trying to document?

And the content has shifted away from that sense of exchange and reciprocity, towards something more extractive. Post visibility seems dictated by its potential for consumption, an algorithm that churns out an infinite supply of whatever originally caught your eye, but its significance diminishing with each iteration. In the process, the modular logic of those hashtag archives, which I used to navigate like a visual filing system, has been disrupted by new rules and formats.

My contributions to that community of practice have also been eroded by motherhood, which simultaneously limited opportunities for textile conservation whilst rewriting the algorithms that shape my Instagram feed. As I began working with Fragment 70, thinking about the fragmentation of my professional time and space, I found myself returning to my dispersed Instagram studio once again, even though my usage of this tool has become superannuated. This allowed me to journal my work, to document different stages, and of course to connect once more with a global community of textile lovers.

Almost as soon as I sat down to start, I turned the fragment over. The rough back of a knotted carpet exposes structure and logic, in contrast with the front where its lush pile melds into a seamless patterned surface.

Recently I have been working with a vast set of historic needle tapestries which were made by hundreds of participants across France. In looking to understand more about the individual makers, I found the reverse more revealing than the front. The back records decision-making: carried threads, knots, tensions, the disciplined management of material. Historically, the reverses of textiles have often remained unseen, despite containing the clearest record of labour. My automatic gesture of turning the carpet over is aligned with my longer-standing interest in the social and gendered histories of needlework; that is to say, in what is structurally essential yet visually marginal.

On the reverse of Fragment #70, the dense pile of the front divides into discrete units of colour, forming a reproducible code. My analogy is not accidental; the structural logic of textiles, most notably the punched-card system of the Jacquard loom, is a precursor to binary code and computing. Looking at the back of the carpet, I began to read it as data for a pattern. When I am stitching a large or particularly complex piece of embroidery, I first grid my fabric to match the grid of the pattern; here, in order to convert the carpet into a pattern, I stitched into it, using red thread to outline a grid.

Working square by square, I counted knots within five-by-five units, assigned them a colour and then by colouring in squares on an empty grid in Photoshop, translated them into a digital pattern. It was slow, repetitive work, which is not to say it was tedious, but rather meditative. At times it felt like reverse-engineered counted work; at others, like drafting a map. The emerging design resembled a topographical chart, terrain translated into abstract code.

The resemblance to mapping reminded me of questions I have been exploring in my research into decolonial practices. The act of converting a living textile surface into a charted pattern uncomfortably echoes the broader impulse and colonial urge to translate complex cultural practices into systems of classification and representation. When I first envisaged my “open source, visual archives” through Instagram hashtags, it was partly to counter the overly precise attempts at fixing living cultural knowledge into rigid categories and artificial boundaries.

My charting was however less an attempt at perfect facsimile than a form of inquiry; more an act of translation that fits into a larger scheme of reproduction, where there is no “original”; pattern exists through reiteration. The dragon design itself has passed through multiple reiterations, originating in China, reinterpreted in the Caucasus; the once-whole historic carpet burned into fragments; the fragmented carpet reproduced in a replica that was then gridded and cut again into further fragments.

Now in my hands, the fragment was converted to a physical grid, then manually translated into a digital pattern, and then from that digital grid once again converted into a stitched woolen surface. When I translated the mapped pattern into wool, I restricted myself to a dozen colours. None of my yarns matched perfectly. Many were remnants from other projects: British yarns sent with the fragment from its previous host in the UK; wools purchased for the restoration of an Iranian Kilim; antique cashmere darning wool found at a brocante in rural Creuse. Each yarn carried its own provenance, its own chain of hands, including those involved in its manufacture.

My reproduction process felt like something algorithmic, governed by translation protocols: if this, then that — so how did this semi-mechanical reproduction engage with authenticity and aura? My small stitched square is not a copy of Fragment #70; it is a product of constraints: time constraints, domestic constraints, material constraints, translation constraints. So in what ways does it differ from the frustrating, and sadly characteristic, algorithmic replication of increasingly inferior content?

My needle tapestry is distorted into a slight diamond shape because I stitched it in hand rather than on a frame. That decision was practical — portability matters when working around young children! — but also conceptual. The work remains visibly human, visibly uneven, and resists perfect alignment. Its distortion is not a technical error; it records the circumstances of its making, much as historic needlework bears the imprint of the bodies and spaces that produced it.

It is hard not to sound like I am making excuses, especially in an environment like Instagram, increasingly saturated by frictionless digital reproduction and synthetic pattern generation, but these idiosyncrasies and unique combinations all mattered to me. 

When I started writing more reflexively I was struck by how the grid recurs throughout this project. It structures the back of the carpet; it structures my digital translation; and once upon a time, it structured Instagram: the tiled portfolio through which one’s identity could be apprehended spatially and relationally.

Using CulturalxCollabs to return temporarily to my Instagram grid was a kind of reclamation of a creative space that has lost its potency. I realised I had barely posted anything to my grid in years, shifting away from documenting my work, towards impermanence and consumption. Ironically, algorithmic optimisations have dissolved the adjacency and discipline of the archives I used to contribute to, where this square next to that square, this decision next to that one, once revealed patterns that felt significant to me. Now those contributions feel scattered and dissociated.

In this project however I can remind myself how fragmentation need not signify loss alone, that it can also generate multiplicity. Fragment #70, both prior to and after the creation of its redefined borders, circulated through several pairs of hands before reaching mine, and will continue beyond it. My stitched response is another fragment, a small, situated iteration within a larger distributed artwork.

The Dragon Carpet now exists not as a singular self-contained object but as a network of responses. In that sense, it mirrors both textile history and digital culture: a pattern sustained because it continues to be repeated, translated and altered. But where algorithmic reproduction typically involves acceleration and erasure of the original, this process insists on slow thoughtfulness and acknowledgement. Even if I only know some of the hands involved, even if it is not significant to anyone else, I have been able to document the provenance of my materials. My little square contains both literal and figurative DNA from across the planet.

Fragment #70 began as the representation of a piece of a damaged carpet. In turning it over, mapping it, translating it, and stitching it into a new scrap of tapestry, I found my own medium to visualise the conditions under which making now circulates. Even if the square I produced does not lay perfectly flat, it holds within it a tangle of relations: between front and back, analogue and digital, historic and contemporary, domestic and institutional, visibility and opacity.