Willem Halffman

One imaginary sees the scientific literature as a gigantic library, with texts to be read. Another sees it as a database, with facts to be mined. Our analysis clarifies contrasting expectations informing current publishing innovations – and their epistemic and political risks.

Halffman, Willem, and Serge P. J. M. Horbach. ‘The Library and the Database: Contrasting Expectations in Two Imaginaries for the Research Literature’. Minerva, ahead of print, 3 December 2025. https://doi.org/10.1007/s11024-025-09618-7.

When journalists report news based on scientific publications, they may have to follow up on the fate of their sources. Sometimes, scientific publications get retracted and they may invalidate earlier reporting. We used retracted covid papers to see whether journalists correct news that was based on papers that were later retracted. The results are rather disappointing. The good news is that at least quality media outlets distinguish themselves with better correction practices. We conclude our paper with suggestions for how to improve journalistic practices.

Valinciute, Auste, and Willem Halffman. ‘Do Journalists Update Retracted Science News?’ Journalism Practice, Routledge, 2025, 1–23. https://doi.org/10.1080/17512786.2025.2540460.

The slow rise of non-mechanistic explanatory language in the neurosciences: an analysis of the neuroscience literature using large-volume text analysis.
Or: how an empirical study of scientific language can inform a debate about types of explanations among philosophers of neuroscience.

Kostić, Daniel, and Willem Halffman. ‘Mapping Explanatory Language in Neuroscience’. Synthese 202, no. 4 (29 September 2023): 112. https://lnkd.in/eBhVSPFe.

Now available as preprint:
Wang, Jing, Willem Halffman, and Serge P. J. M. Horbach. ‘The Journal Attention Cycle: Indicators as Assets in the Chinese Scientific Publishing Economy.’, 13 September 2023. https://doi.org/10.31222/osf.io/6uvpz.

How journals use attention indicators such as the Journal Impact Factor as an asset, but also how the possibility to do so depends on the evaluative regime under which they function, using Chinese journals as examples.

What it means and what it does to have lists of ‘good’ and ‘bad’ journals and why China has so many of them: Wang, Jing, Willem Halffman, and Yuehong Helen Zhang. ‘Sorting out Journals: The Proliferation of Journal Lists in China’. Journal of the Association for Information Science and Technology n/a, no. n/a (13 July 2023). https://doi.org/10.1002/asi.24816. 

Observational grids specify what can be seen and how observations can be characterised. We use the notion of grids to analyse observational practices of dragonfly enthusiasts. We show how grids are readjusted to align with the concerns of the dragonfly community through the lens of a succession of dragonfly field guides.

Turnhout, Sander, and Willem Halffman. ‘Readjusting Observational Grids in Dragonfly Field Guides’. Social Studies of Science, 8 July 2023, 03063127231183011. https://doi.org/10.1177/03063127231183011.

Software tools that suggest citable sources as your are writing may sound like a really convenient idea, but we see a lot of disadvantages. Sloppy citing is one of them: the temptation grows to refer to sources without properly reading them. This aggravates already sloppy citation practices in research. Details in our new publication:

Horbach, Serge P. J. M., Wytske M. Hepkema, Willem Halffman, and Freek J. W. Oude Maatman. “Automated Citation Recommendation Tools Encourage Questionable Citations.” Research Evaluation, 2022, rvac016. https://doi.org/10.1093/reseval/rvac016.

Journal guidelines help precent the publication of misidentified biomedical resources, but they are not the quick fix you might expect. Journals’ regulatory action is important and yet only part of the solution, as there are limits to what journals can do. Our paper analyses the effects of journal guidelines for cell line identification, antibody validation, and lab animal reporting is out in pre-pub version in the International Journal of Cancer.

Hepkema, W.M., Horbach, S.P., Hoek, J.M. and Halffman, W. (2021), Misidentified biomedical resources: Journal guidelines are not a quick fix. Int. J. Cancer. Accepted Author Manuscript. https://doi.org/10.1002/ijc.33882

I’m really excited about the NanoBubbles project, bringing together STS, data scientists, library scientists. computer scientists, historians, and nonscientists. We investigate controversies, errors and over-stretched promises in nanobiology from this exciting interdisciplinary mix, with a team that is really committed to cooperation. For now, not a lot is visible to the outside world of the cooperative explorations that are going on, apart from our beautiful webpage and of course the adverts out for hiring people throughout the project, such as my two PhD candidates and two Postdocs. One of the big challenges is to cooperate in a multi-method approach, ranging from quantitative bibliometrics, qualitative interviews and ethnographies, all the way into experimentation.

If the variability of lab animals contributes to the failure to replicate experiments, should this lead to further standardisation? The microbiome can significantly affect experimental results and this presents an interesting challenge. A perspective from philosophy of science offers clarification and investigates the alternative avenue of embracing variation. Our new publication is a close cooperation with biomedical researchers.

Witjes, Vera, Annemarie Boleij, & Halffman, W. (2020). The Microbiome of Laboratory Mice: Reducing versus Embracing Variation, Animals, 10(12), 2415. doi: 10.3390/ani10122415.