A 2019 study by Júlia Perczel introduced the 'historical fingerprint' of a museum collection — a statistically validated venue network derived from the pre-acquisition exhibition histories of the collection's artists — and used it to compare how the Tate, MoMA, and Centre Pompidou represent Central-Eastern European art. The paper framed a collection's representation as acting on the art world in four interconnected ways and empirically demonstrated only two of them (the congealing of the art world into a museum-specific venue structure, and museum positioning among peer museums). Two posited modes of agency were explicitly left for future research: whether a collection's representation enables some and inhibits other interpretations in situated reception, and whether museums' venue-structure positioning reinforces or modifies the overall structure of the art field over time. The open problem is to establish these two modes of agency empirically; the unquantified selection and aggregation biases of aggregated online exhibition databases (such as Artfacts.net) on which fingerprint construction rests remain an author-acknowledged limitation that conditions any continuation.

Background

Museums of modern and contemporary art build their collections through acquisition decisions that unfold over decades, and these collections in turn shape which artists and regions become visible in art history. Quantitative study of this process became possible through aggregated online exhibition databases, of which Artfacts.net — a Berlin-based company maintaining one of the most-used exhibition-history datasets, compiled partly by merging online databases and partly from user-submitted records — is a prominent example. Perczel (2019) proposed a data-driven method for comparing museum collections from such data. For each collection, the exhibition histories of its artists (acquired from 1989 onward) are encoded as a bipartite network — a graph with two node types, here venues and artists, where links only run between the types (a venue is linked to an artist it exhibited). This is projected onto venues: two venues are connected when they both exhibited many of the same collection artists. To keep only statistically meaningful links, each venue–venue link is tested against the null hypothesis of random artist co-occurrence using the hypergeometric distribution (the probability model for drawing a given number of successes in a sample without replacement), with a Bonferroni correction (dividing the significance level α=0.01\alpha=0.01 by the number of tests to control false positives) — an instance of the 'statistically validated network' methodology of Miccichè and Mantegna. The resulting filtered venue network is called the historical fingerprint of the collection: it cannot be directly controlled by the museum (unlike individual acquisitions), yet it uniquely characterizes the collection's representation at a point in time, and every new acquisition somewhat modifies it. Applying this to the Tate, Centre Pompidou, and MoMA with respect to the Central-Eastern European (CEE) region, the paper showed that the three museums rely on differently structured venue landscapes, reflecting distinctive takes on the region. In the theoretical framing, the representation enabled by a collection 'acts as an agent in four main, interconnected ways': (1) it enables some and inhibits other interpretations of the depicted phenomena, by acting as a context of interplay to the represented artistic positions; (2) it congeals the venues of the art world into a museum-specific venue structure; (3) it is a tool with which the museum positions itself among other museums; (4) the venue structure is a tool with which the museum participates in the reinforcement or modification of the structure of the (art) field — the latter in the sociological sense of a structured space of positions and position-takings, following Bourdieu. The paper's analyses substantiated the second and third aspects; aspects (1) and (4) were explicitly declared 'hopefully … the subject of future research'. The study also acknowledged two unresolved limitations that condition any continuation: the selection and aggregation biases of Artfacts-type databases (overrepresentation of artists with strong online presence and of self-reporting, entrepreneurial artists) are mitigated only by fixing the comparison to three collections, never quantified; and the deliberate restriction to acquisitions from 1989 onward leaves earlier periods and the longitudinal evolution of fingerprints under new acquisitions unexplored.

Problem Statement

In the historical-fingerprint framework of Perczel (2019), a museum collection's derived venue structure is posited to act as an agent in the art world in four interconnected ways, of which only two — the congealing of the art world into a museum-specific venue structure (aspect 2) and museum positioning among peer museums (aspect 3) — were empirically established. The open problem is to establish the two remaining, explicitly deferred modes of agency empirically: (1) reception — whether, and how, a collection's representation (as characterized by its historical fingerprint, i.e. its statistically validated venue structure over the pre-acquisition exhibition histories of its artists) enables some and inhibits other interpretations of the depicted phenomena in situated reception, where 'situated reception' means the concrete encounters in which the collection is interpreted — viewing situations, exhibition reviews, curatorial catalog framing, art-historical writing, audience response — rather than interpretation in the abstract; and (2) field dynamics — whether, and how, museums' use of venue-structure positioning reinforces or modifies the overall structure of the art field over time. Both questions share one scientific objective — demonstrating (or refuting) the receptive and field-transforming efficacy of collection representations. The selection and aggregation biases of aggregated online exhibition databases (Artfacts.net-type data) on which fingerprint construction and cross-collection comparability rest remain an author-acknowledged limitation of the source study.

References

  1. Júlia Perczel, "Is Structure Context or Content? A Data-Driven Method of Comparing Museum Collections", Život umjetnosti 105 (2019), pp. 76–109. DOI: 10.31664/zu.2019.105.04, https://doi.org/10.31664/zu.2019.105.04 (open-access full text: https://hrcak.srce.hr/233172)
  2. Salvatore Miccichè and Rosario Nunzio Mantegna, "A primer on statistically validated networks" (2019). arXiv: 1902.07074, https://arxiv.org/abs/1902.07074
  3. Júlia Perczel and Balazs Vedres, "Careers in the global art field: Geo-capital and globalizer venues in the consecration of Central-Eastern European artists", Poetics 108, article 101961 (2025). DOI: 10.1016/j.poetic.2024.101961, https://doi.org/10.1016/j.poetic.2024.101961
  4. Júlia Perczel, "Modes of Refraction: How Core Museums and Peripheral Art Collectors Co-Produce Cosmopolitan Elite Networks", Cultural Sociology (2026). DOI: 10.1177/17499755261473075, https://doi.org/10.1177/17499755261473075
  5. Edgar Buršić, Sanja Sekelj, Donato Ricci, and Davor Mišković, "Networking Through Persistence: Case Study on the Position and Development of the Clubture Network", Sociologija i prostor 62(1) (229) (2024), pp. 3–28. DOI: 10.5673/sip.62.1.1, https://hrcak.srce.hr/clanak/460459
  6. Mar Canet Solà, Antonina Korepanova, Ksenia Mukhina, and Maximilian Schich, "Quantifying Collection Lag in European Modern and Contemporary Art Museums" (2023). arXiv: 2305.14159, https://arxiv.org/abs/2305.14159; published in Proceedings of the 16th International Symposium on Visual Information Communication and Interaction (VINCI '23). DOI: 10.1145/3615522.3615562, https://doi.org/10.1145/3615522.3615562