Taxonomies of Error
an observation of digital museum glitches
Fig.1
Curatorial statement
This project is not an attempt to change it, but rather an attempt to observe the strangeness of these digital spaces via illustration-specific tools: a mixture of observational drawing, 3D modelling, and animation. Based on both physical and digital visits to the Pitt Rivers Museum and the Museum of Natural History in Oxford, the project focuses on the ideas of mistake, observation, and docufiction. The glitches in the virtual tours, the patchiness and layered nature of annotations, and scientific illustration cliches, as well as detailed observational drawing, modelling and animation become a phenomenological and psychogeographical account of digital-physical experience of the museum. The affective encounter with the museum is articulated back through an image-based account of being in a museum both physically and digitally – in a language that's aware of the museum curation tropes.
I initially started this project from an interest in museum virtual tours and the ways 3D images participate in them.
Despite the sporadic appearance of breakthrough projects in this area, the promise of 3D in the context of museums is often unfulfilled – at least when it comes to presenting the museum collections to an online audience. Lots of ‘3D museums’ and virtual tours occupy a space between a clunky game prototype and an interior scan in real estate marketing. They have to balance the pressure to generate physical footfall and digital likes and shares with sharing knowledge and defining what a digital museum can be.
Despite the sporadic appearance of breakthrough projects in this area, the promise of 3D in the context of museums is often unfulfilled – at least when it comes to presenting the museum collections to an online audience. Lots of ‘3D museums’ and virtual tours occupy a space between a clunky game prototype and an interior scan in real estate marketing. They have to balance the pressure to generate physical footfall and digital likes and shares with sharing knowledge and defining what a digital museum can be.
Fig.2
Non-scientific illustration
The process
The 'taxonomy' and the names of the species were devised with ChatGPT 5.0 and were based on my descriptions of the glitches. The statistical hallucination of AI here illustrates how I, having no sufficient knowledge in biology, perceive scientific illustration in the context of a museum: a decorative semblance, indication of structure, order, and authority.
The focus on glitch is also a way of looking at what a digital museum is through its 'negative space': ‘Noise becomes a way of understanding how normal communication works' (Parikka 2012: 110)
The focus on glitch is also a way of looking at what a digital museum is through its 'negative space': ‘Noise becomes a way of understanding how normal communication works' (Parikka 2012: 110)
I started from observing the glitches in the 3D-scanned part of the Pitt Rivers Museum virtual tour, as if they were biological species, trying to document them with the precision of the botanical illustrations displayed next door, in the Museum of Natural History.
During the physical visit, I also took 3D scans of some of the museum objects myself. The resulting scans, with their glitches (holes, stretches, spikes) indeed looked like intricate geological or biological specimens, which I placed into the Blender-modelled cases that replicate the ones at Pitt Rivers Museum.
These objects are still trapped behind the glass of the cases and the surface of the screen, but they don't behave like their static physical prototypes: drift like meteorites in space, and breathe like organs displaced from their bodies.
These objects are still trapped behind the glass of the cases and the surface of the screen, but they don't behave like their static physical prototypes: drift like meteorites in space, and breathe like organs displaced from their bodies.
Similarly, in a digital museum tour, what I see is not the objects, but the 'digital space': the objects are all flattened into surfaces surrounding the digital void. Surprisingly, the digital version gives even less access to the objects that the physical one: most of the signs and details are rendered illegible due to the resolution.
I see the digital 'pillars' holding up the knowledge I'm supposed to get, but not the knowledge itself.
Photogrammetry or LIDAR-scanning conveys information about the texture and tactile qualities in a very limited way, flattening all touch-oriented information into perfectly flat blurry surfaces. ‘The ocular-centric framing of data visualisation that underpins 3D digital object creation diminishes other modes of engagement and what our understanding of a whole object is’ (Geismar 2018: 101)
I see the digital 'pillars' holding up the knowledge I'm supposed to get, but not the knowledge itself.
Photogrammetry or LIDAR-scanning conveys information about the texture and tactile qualities in a very limited way, flattening all touch-oriented information into perfectly flat blurry surfaces. ‘The ocular-centric framing of data visualisation that underpins 3D digital object creation diminishes other modes of engagement and what our understanding of a whole object is’ (Geismar 2018: 101)
When I go to a museum, I often look at the 'museumness' itself more than at the exhibits. At Pitt Rivers, it is the overwhelmingly packed space, heavy wood, darkness of the space, dusty smells, echoing footsteps, stiff air and resulting headache; pieces of paper and screens instructing me, lecturing, explaining, entertaining with 'fun facts', inviting to leave a review or participate in a shared space of a guest book.
Fig. 3
But what 3D objects can do is convey this information visually through the behaviors of 3D that physical objects are incapable of. Distortions, overlapping, glitches act as metaphoric representations of touch – or rather its impossibility through screens or behind glass displays.
Museumness
Digital sensing and knowledge-making
These representations of touch, or some other sensing, do not always directly allude to known physical sensations. When 'breath' is referenced, for example, the 3D model directly engages with physical rhythms, becoming a form of what Laura Marks called 'haptic visuality' (Marks, 2000).
However, in many cases movement or interaction with the objects (zooming in, rotating, overlapping etc) are disconnected with directly embodied sensing. The 3D scans of the real museum artefacts 'invading' the 3D models of the display cases allude to a different kind of experience: that of early games glitches, where the player can get 'stuck' in textures. Such glitches are not neutral technical errors. They are experiences: disturbing, funny, uncanny, annoying. Even if such collisions of 3D shapes do not refer to an embodied sense of weight, they allude to lived sensations. That moment of getting trapped wasn’t just a bug — it was a rupture in the contract of spatial logic.
You were no longer “playing the game” — you were exiled from it, but still inside.
'Being stuck' here, in the context of neither digital-nor physical museums, is being stuck between various regimes of production of 'museumness', where museum logic (predictable, buildable) meets machine vision logic (statistically valid, haunted, prone to failure). Knowledge-making is 'stuck' in the structural parts of the museum: both physical and digital.
These representations of touch, or some other sensing, do not always directly allude to known physical sensations. When 'breath' is referenced, for example, the 3D model directly engages with physical rhythms, becoming a form of what Laura Marks called 'haptic visuality' (Marks, 2000).
However, in many cases movement or interaction with the objects (zooming in, rotating, overlapping etc) are disconnected with directly embodied sensing. The 3D scans of the real museum artefacts 'invading' the 3D models of the display cases allude to a different kind of experience: that of early games glitches, where the player can get 'stuck' in textures. Such glitches are not neutral technical errors. They are experiences: disturbing, funny, uncanny, annoying. Even if such collisions of 3D shapes do not refer to an embodied sense of weight, they allude to lived sensations. That moment of getting trapped wasn’t just a bug — it was a rupture in the contract of spatial logic.
You were no longer “playing the game” — you were exiled from it, but still inside.
'Being stuck' here, in the context of neither digital-nor physical museums, is being stuck between various regimes of production of 'museumness', where museum logic (predictable, buildable) meets machine vision logic (statistically valid, haunted, prone to failure). Knowledge-making is 'stuck' in the structural parts of the museum: both physical and digital.
Disease of data
Fig. 4
The glitch becomes the monstrous sublime, which the museum cannot contain. Embodiment itself becomes the dreadful, yet alluring entity haunting the museums.
I may be able touch the fur of a stuffed bear in the Museum of Natural History, and I may be able to scale up the 3D model online. But the movement of bodies and manipulation of data is still a threat to the museum.
Embodiment
Fig. 5
My own experience of ‘learning’ at Pitt Rivers is anarchic and bluntly ignorant of the things that are certainly important for the historians, anthropologists, curators, conservationists, and everyone working on the collections. However, it might be sensitive towards some of the other things: the museum's aversion towards embodiment, the role of sense-making strategies learnt from other mediums like games, or the role of 3D and scientific imaging cliches in constructing credibility.
This project is ultimately a visualised account of what an affective encounter with a museum looks like – an extended version of a guest book entry; a museum experience articulated back, reflected back through image-making. It is also a reflection on embodiment within the museum: digital and physical.
Some of the guest book entries amuse with how frivolous, random, emotional, or irrelevant they might be – even though they are heavily interrupted with ‘orderly’ and ‘well-mannered’ expressions of gratitude and appreciation - i.e. normative practices of spectatorship. The first type of entries are often dismissed, but for me they trace the reality of learning and experiencing the museum.
Some of the guest book entries amuse with how frivolous, random, emotional, or irrelevant they might be – even though they are heavily interrupted with ‘orderly’ and ‘well-mannered’ expressions of gratitude and appreciation - i.e. normative practices of spectatorship. The first type of entries are often dismissed, but for me they trace the reality of learning and experiencing the museum.
Bibliography
Anderson, G. (2013) Isomorphology. [online] Available at:
https://repository.falmouth.ac.uk/400/1/isomorphology-catalogue.pdf
Accessed 17 November 2025.
Anderson, G. (2017). Drawing as a Way of Knowing in Art and Science. Bristol: Intellect.
Carré, M. "Post-Truth: Archives, GPT-2 and Fake News". AI in Museums: Reflections, Perspectives and Applications, edited by Sonja Thiel and Johannes C. Bernhardt, Bielefeld: transcript Verlag, 2023, pp. 187-198. https://doi.org/10.1515/9783839467107-017
Geismar, H. (2018) Museum Object Lessons for the Digital Age. London: UCL Press
Graham, H. (2025) Deconstituting Museums: Participation’s affective work. London: UCL Press
Marks, L.U. (2002). Touch: Sensuous Theory and Multisensory Media. Minneapolis: University of Minnesota Press.
Marstine, J. (2006). New museum theory and practice : an introduction. Malden: Blackwell.
Parikka, J. (2012) What Is Media Archaeology? Cambridge: Polity Press
Thiel, S., Bernhardt J.S. (eds.) (2024) AI in Museums. Verlag, Bielefeld
Varutti, M. (2022). The affective turn in museums and the rise of affective curatorship. Museum Management and Curatorship. 38. 1-15. 10.1080/09647775.2022.2132993.
Anderson, G. (2013) Isomorphology. [online] Available at:
https://repository.falmouth.ac.uk/400/1/isomorphology-catalogue.pdf
Accessed 17 November 2025.
Anderson, G. (2017). Drawing as a Way of Knowing in Art and Science. Bristol: Intellect.
Carré, M. "Post-Truth: Archives, GPT-2 and Fake News". AI in Museums: Reflections, Perspectives and Applications, edited by Sonja Thiel and Johannes C. Bernhardt, Bielefeld: transcript Verlag, 2023, pp. 187-198. https://doi.org/10.1515/9783839467107-017
Geismar, H. (2018) Museum Object Lessons for the Digital Age. London: UCL Press
Graham, H. (2025) Deconstituting Museums: Participation’s affective work. London: UCL Press
Marks, L.U. (2002). Touch: Sensuous Theory and Multisensory Media. Minneapolis: University of Minnesota Press.
Marstine, J. (2006). New museum theory and practice : an introduction. Malden: Blackwell.
Parikka, J. (2012) What Is Media Archaeology? Cambridge: Polity Press
Thiel, S., Bernhardt J.S. (eds.) (2024) AI in Museums. Verlag, Bielefeld
Varutti, M. (2022). The affective turn in museums and the rise of affective curatorship. Museum Management and Curatorship. 38. 1-15. 10.1080/09647775.2022.2132993.