Open science imaginaries: Meanings and values among natural scientists in Brazil,France and Peru

Abstract

‘Open science’ is a term that remains under constant discussion. Based on 29 semi-structured interviews with researchers from the natural sciences working in Brazil, France and Peru, this paper discusses the different meanings and values associated with open science. The different imaginaries concerning open science can be organized into three main categories: open science as research outputs, as embedded in the research process, and as related to the science–society relationship, the latter being more common in the two South American countries in this study. Openness, freedom and democracy are values commonly associated with this concept. In addition, Mertonian values such as universalism and organized skepticism are highlighted, although with nuanced variations due to open science practices.

1. Introduction

1.1 Open science meanings

The origin of the term open science can be traced back to the late 16th and early 17th centuries, marking a shift away from secrecy in research, though this remains a relatively modern origin (David, 2004). Open science institutions are historically rooted social and political constructs, external to scientific practice and therefore subject to redesign or policy manipulation in representative democracies (David, 2008). Grand et al. (2012) further frame open science as a ‘technology of trust’. Under this understanding, science is understood less as a body of definitive results than as a provisional and continually evolving endeavour. Open science has the capacity to emerge as a novel technology of trust, benefiting both the scientific community and the general public with its emphasis on transparency and unrestricted access to research outputs (Grand et al., 2012). Moreover, the concept entails a particular set of both values and behaviours. This relies on the idea that, given the right tools, scientists can systematically enhance scientific practice (Cohoon and Howison, 2021).

In contrast to proprietary approaches, open science foregrounds transparency, offering clear benefits such as faster expansion of knowledge through the use and dissemination of public domain goods (David, 2004). One of the key advantages of open science is its ability to accelerate the pace of research through increased accessibility and collaboration (Woelfle et al., 2011). Evidence further supports these benefits: as McKiernan et al. (2016) note, multiple studies highlight the different open science benefits, while an increasing number of foundations and academic institutions actively promote and reward open research practices. Consequently, given all these reasons, this movement holds the potential to foster a more coherent and globally shared system of knowledge (Méndez, 2021).

Practising open science has several advantages for early-career researchers, as highlighted by Allen and Mehler (2019). Adopting open science enhances the researcher's reputation, hence augmenting publication opportunities, even for null results. Moreover, it also enhances future prospects for collaboration, recognition and career progression, often leading to higher citation rates. At a broader level, open science practices improve research reliability by enabling wider dissemination, fostering replication and ensuring greater transparency in data presentation.

The term open science generally refers to the disclosure of all stages of scientific inquiry. Yet the concept remains highly complex and contested (David, 2004). For instance, the very notion of openness is under constant questioning (Albagli et al., 2015) and lacks consensus (Grubb and Easterbrook, 2011). Openness is not an intrinsically worthwhile objective of science, but rather a practice that must be encouraged and rewarded at each stage of research (Levin and Leonelli, 2017). Along these lines, Leonelli (2023) argues that, while sharing objects is central to sound science, the creation and exchange of research objects should not be mistaken for the ultimate objective of science.

Further debates highlight competing interpretations of what open science entails. Clinio (2019) identifies two contrasting perspectives: a utilitarian view prioritizing productivity, efficiency and competitiveness, and a normative approach emphasizing rights, social justice and cognitive justice. Echoing this critical stance, Mirowski (2018) argues against the commercialization of open science through its integration into platform capitalism. According to this view, the increasing platformization of science divides the research process into distinct segments to enhance efficiency and reduce costs, while simultaneously reinforcing market logics.

Divergent viewpoints exist around open science, acknowledging significant contextual variations. Lindemann and Häberlein (2023) argue that proposals for open science must be assessed within their specific institutional and cultural settings. Similarly, Arza et al. (2017) and Fressoli and Arza (2017) contend that openness is typically progressive and differentiated, rather than uniform or universally applicable.

Open science is characterized as a collaborative approach to disseminating knowledge and data among diverse stakeholders, including researchers, corporations and individuals, to address shared challenges and foster innovation (Lepage, 2010). At the same time, it is widely recognized as a highly contested concept. Open science encompasses a broad spectrum of theories about how knowledge should be produced and shared, and its meaning varies according to the perspectives of its advocates. As a result, the term provokes debates that, while grouped under the same banner, often pursue divergent motivations and goals (Fecher and Friesike, 2014).

Despite the lack of a single authoritative definition, several overlapping interpretations exist. For instance, open science is understood as a means to make scientific research more efficient and reliable (Burgelman et al., 2019), or more broadly as an umbrella term covering multiple assumptions about the future of knowledge production and dissemination (Fecher and Friesike, 2014). Reflecting this plurality, Thibault et al. (2023) emphasize that, although a final definition remains unsettled, most interpretations overlap significantly.

Efforts to formalize these principles are evident at the policy level. The most comprehensive articulation is offered by UNESCO (2021), which defines open science as an ‘inclusive construct combining various movements and practices aimed at making scientific knowledge openly available, accessible, and reusable for everyone, increasing scientific collaboration and information sharing for the benefit of science and society, and opening scientific knowledge creation, evaluation, and communication processes to societal actors outside the traditional scientific sphere’. Extending this vision, Gong (2022) argues that open science should be recognized as a basic human right. Similarly, Miedema (2022) contends that by promoting robust scientific practices, open science can contribute to solving pressing global challenges and advance science into a new phase.

In addition to definitional debates, other scholars have sought to systematize the concept through taxonomies, a way of organizing concepts. Pontika et al. (2015) identified six core components: open access, open data, open reproducible research, open science evaluation, open science policies and open science tools. Building on this, Da Silveira et al. (2021) proposed a taxonomy adapted to the Brazilian context, adding new facets such as open education, open licensing, citizen science, digital preservation and open innovation. More recently, Da Silveira et al. (2023) updated this taxonomy to integrate the UNESCO (2021) recommendation, highlighting elements such as open scientific knowledge, open infrastructures, societal engagement and dialogue with other knowledge systems. This latest version reflects a ‘Global South’ perspective, making the taxonomy more inclusive than previous versions.

1.2 Analytical framework

Imaginaries of science have a long history and a wide variety of focuses (McNeil et al., 2017). In particular, according to Jasanoff and Kim (2015), sociotechnical imaginaries are ‘collectively held, institutionally stabilized, and publicly performed visions of desirable futures, animated by shared understandings of forms of social life and social order attainable through, and supportive of, advances in science and technology’. In this case, this paper focuses on the different imaginaries about open science, specifically its meanings and values.

1.2.1 Platform capitalism

Platforms are reshaping the flow of the digital economy as sociotechnical intermediaries. They actively structure, standardize and programme exchanges using data, rules and code (Langley and Leyshon, 2017). Platforms lock in users and data by integrating sectoral services and infrastructure (Van Dijck, 2021).

The platformization of scholarly communication manifests in two main ways: first, through the erosion of negotiating power in shaping the standards, values and norms of knowledge production; and second, through the datafication and monetization of both scholarly outputs and users’ personal data (Ma, 2023). In other words, this process converts scholarly infrastructures into commercial ones that are better suited for competition, rating and monitoring (Ma, 2023), and even surveillance (Pooley, 2022).

This process is already integrated into different stages of the research process, ranging from data collection to outreach, embedded in a digital ecosystem of scientific platforms operated by both commercial and non-profit entities (Da Silva Neto and Chiarini, 2023; Fecher et al., 2024). In this scenario, data becomes a new type of formalized scholarly object (Plantin et al., 2018).

1.2.2 Mertonian norms and open science

The ethos of science, as articulated many years ago by Robert K Merton (1942), rests on four core norms (sometimes referred to as the ‘CUDOS’ norms). Universalism holds that scientific claims should be evaluated according to impersonal criteria, regardless of the researcher's status or identity. Communism (or communalism) asserts that scientific knowledge is a collective good and must be openly shared. Disinterestedness requires that scientists pursue knowledge for its own sake rather than for personal gain. Finally, organized skepticism emphasizes that all ideas must be subject to rigorous scrutiny and continual questioning. More recent research suggests that some norms are now more valid than others: academics continue to express strong support for at least one of Merton's standards, specifically, communism (Macfarlane and Cheng, 2008). This could be partially explained by the fact that the institutional and political context that gave rise to Mertonian norms has long since changed (Hosseini et al., 2022).

The relationship between Mertonian norms and open science has been interpreted in multiple ways. For instance, the sharing of results exemplifies the norm of communalism. Similarly, equal access to publications and data, along with practices such as open peer review, can be viewed as extensions of universalism. Open science also reinforces disinterestedness, insofar as transparency and surveillance mechanisms strengthen research integrity by reducing the scope for personal bias. Finally, organized skepticism is supported not only through academic scrutiny but also through broader societal engagement, facilitated by infrastructures such as repositories and open communication channels (Hosseini et al., 2022). Similarly, Cohoon and Howison (2021) argue that the open science movement is guided by a set of values, and it aims to match scientific practices with the norms established by Merton. In fact, open science advocates perceive the movement as a value-driven philosophy that aims to enhance scientific practices by using open technological infrastructure.

However, it is also argued that open science has established a new ethos of science, departing from the old Mertonian norms (Von Schomberg, 2024). For instance, scientists now promote their work in terms of its possible societal benefits rather than following the norm of ‘disinterestedness’ (Von Schomberg, 2024). According to von Schomberg, this new ethos is related to research conduct driven by information exchange and social cooperation through open science.

Others contend that it is the platformization of science that is reshaping its normative framework (Fecher et al., 2024). In this view, Mertonian norms are being reconfigured: communalism becomes conditional on platform-mediated ownership of findings; universalism is reframed through perceived inclusivity and standardization; in commercialization, platforms serve the interests of their providers; originality is encouraged through participation yet constrained by homogenized knowledge production; and organized skepticism may be reinforced through enhanced transparency (Fecher et al., 2024).

Given this complex background, the concept of open science remains a contested one. This study seeks to contribute by examining the diverse meanings and values that natural scientists attach to it. In particular, it addresses a notable gap: the limited representation of researchers from the Global South. Incorporating their perspectives is essential for fostering a more inclusive debate on the different interpretations and values of open science.

The research questions are therefore as follows:

What are the different meanings of open science according to natural science researchers?

What are the central values associated with these meanings?

2. Method

This paper adopts an exploratory qualitative research approach, which supports a better understanding of social reality by highlighting processes, meaning tendencies and structural characteristics (Flick et al., 2004).

Specifically, this study uses qualitative content analysis (QCA), a technique used to describe the meaning of qualitative data systematically (Schreier, 2014). Building on this foundation, meanings are constructed from both manifest and latent content while considering the context from which the data was drawn (Roller, 2023). Furthermore, QCA concentrates on specific facets of the data while maintaining adaptability in the coding framework (Schreier, 2012).

The main rationale for using semi-structured interviews is that they allow for a degree of consistency in the themes explored across interviews (Corbin and Strauss, 2014). Moreover, interviews help researchers understand how scientists see their own work and the field as a whole (Rodriguez Medina, 2013), while also encouraging interviewees to reflect on both their own lives and their disciplinary contexts (Rodriguez Medina, 2013).

2.1 Rationale for choosing subject areas and countries

This research focuses on researchers in the natural sciences, who have been early adopters of open science practices (Armeni et al., 2021; Bennett, 2022; Clinio and Albagli, 2017; Vale, 2015). Moreover, most social studies of science are deeply rooted in ‘situated’ contexts of knowledge production, despite the generalization claims often made in ‘classic’ science studies (Kreimer, 2022). This raises an important question about the validity of applying conceptual frameworks developed in high-income contexts to so-called ‘peripheral’ settings. Therefore, this study aims to produce results that can still hold under this open question, using conceptual frameworks developed in various contexts. Moreover, cross-country studies offer a unique opportunity to explore research practices comparatively. Identifying parallels among these practices underscores the context-dependent and socially constructed nature of knowledge production (Monteiro, 2023).

2.2 Participant selection

The participant selection process comprised several steps. First, using the QS World University Rankings 2022 as a guide, the five highest-ranked universities in Brazil, France and Peru were identified. The rationale for using the QS ranking as a guide is its ease of use. This specific ranking is a practical and useful tool to identify productive research institutions, carefully curated using several indicators. In this case, since these countries have a large number of universities, the use of this instrument helped in the identification and ranking of institutions, which would otherwise have been an enormous task because of the high number.

Following this initial identification process, natural sciences researchers, mainly working in the experimental areas of biology, chemistry and physics, were contacted through social media.

A small group of researchers (n = 6) was also contacted in person during local conferences or academic gatherings, and all agreed to be interviewed directly. In total, 85 invitations were sent through social media. Several individuals declined participation due to lack of time or limited experience with open science, while a few withdrew due to unexpected work commitments.

In total, 29 researchers participated in this study. Table 1 presents the full list of interviewees and their characteristics. The sample is composed of a diverse group of researchers working primarily in the experimental areas of the three aforementioned disciplines.

Table 1.

Description of the participants.

Final code	Interview name	Title	Age range	Country of work	Gender
P1.AP.F	Astrophysics 1	Associate professor	30–39	France	Male
B1.AP.F	Biology 1	Associate professor	50–59	France	Male
C1.D.F	Chemistry 1	PhD student	20–29	France	Female
P2.D.F	Physics 2	PhD student	20–29	France	Male
B2.PD.F	Biology 2	Postdoc	30–39	France	Female
C2.RS.P	Chemistry 2	Research scientist	40–49	Peru	Female
P3.AP.P	Physics 3	Associate professor	40–49	Peru	Male
C3.D.P	Chemistry 3	PhD student	20–29	Peru	Female
C4.PD.F	Chemistry 4	Postdoc	30–39	France	Female
B3.RS.P	Biology 3	Research scientist	30–39	Peru	Male
C5.RS.P	Chemistry 5	Research scientist	40–49	Peru	Male
P5.AP.B	Physics 5	Associate professor	50–59	Brazil	Male
C7.D.B	Chemistry 7	PhD student	20–29	Brazil	Male
P4.FP.P	Physics 4	Full professor	40–49	Peru	Male
B4.FP.P	Biology 4	Full professor	40–49	Peru	Male
P6.FP.P	Physics 6	Full professor	50–59	Peru	Male
C8.D.B	Chemistry 8	PhD student	20–29	Brazil	Female
BQ1.PD.B	Biochemistry 1	Postdoc	30–39	Brazil	Female
C9.PD.B	Chemistry 9	Postdoc	20–29	Brazil	Male
P9.PD.B	Physics 9	Postdoc	30–39	Brazil	Male
P10.PD.F	Physics 10	Postdoc	30–39	France	Male
C10.D.F	Chemistry 10	PhD student	20–29	France	Female
P12.FP.F	Physics 12	Full professor	50–59	France	Male
P11.FP.F	Physics 11	Full professor	40–49	France	Male
B5.AS.B	Biology 5	Associate professor	30–39	Brazil, Peru	Male
B6.D.P	Biology 6	PhD student	30–39	Peru, France	Female
B7.PD.B	Biology 7	PhD student	30–39	Brazil	Female
P13.PD.B	Physics 13	Postdoc	30–39	Brazil	Male
BQ2.FP.B	Biochemistry 2	Full professor	40–49	Brazil	Female

2.3 Participant consent and ethics review

Interviews were audio-recorded, and verbal consent was given at the beginning of each session. Written consent for participation and IRB/Ethical Committee review were not required for social science research in France, where this study was carried out (Vassy and Keller, 2008).

2.4 Interview process

Participants were informed of the research objectives prior to the start of each interview. Interviews took place between May 2022 and September 2023. All interviews were audio-recorded, and participants affirmed their consent before recording commenced.

Due to the geographical distribution of participants, most interviews were conducted online via Zoom or Google Meet, and only a few were held in person (a small number of participants who were in Lyon and Geneva at the time of interview). All 29 interviews were conducted one-to-one, with no one else present besides the participant and the interviewer. All interviews were conducted by the author.

The complete semi-structured interview guide is available in Appendix 1. No repeat interviews were carried out. The duration of each interview ranged from 40 minutes to 1 hour and 30 minutes, with an average length of approximately 1 hour.

2.5 Data analysis

Interview transcripts were analysed using NVivo, a widely used qualitative data analysis software (Dhakal, 2022; Siccama and Penna, 2008). The author was the sole coder of the data presented in this study.

The coding process was developed following the process outlined by Saldana (2009), using an inductive category development approach for the codes. Initial codes were developed with an open coding strategy (Benaquisto, 2008), emerging directly from the data. These codes were then refined through iterative coding (Gioia et al., 2013; Locke et al., 2022) in a systematic way, in order to reduce the initial number of codes. Afterwards, codes were organized into categories based on similarities, from which the initial themes were identified. The resulting themes and codes are shown in Table 2, while Appendix 2 presents the themes, codes and representative extracts.

Table 2.

Open science meanings and values: Themes and codes.

Open science values	Openness
	Freedom
	Transparency
	Reproducibility
	Universalization
Open science meanings	Research outputs	Open access
		Gold open access
		Open code
		Open data
		Open software
		Journals
		Infrastructure & platform
		Results
	Research process	Sharing culture
		Analysis
		Collaboration
		Discourse
		Production forms
		Instruments
		Interchange
		Interdisciplinary
		Methodology
		Novelty
	Science–society	Education
		Public good
		Citizen science
		Public communication of science
		Science communication
		Science translation
		Translation

Previous studies indicate that the mode of interview (online or face-to-face) does not significantly affect the significance of information obtained (Shapka et al., 2016). Consistently, in this study, there were also no significant differences in the number of codes or references between online and face-to-face interviews.

2.6 Data saturation

In this work, data saturation was reached when no new codes or themes emerged during the analysis process around the seventh interview, which coincides with patterns commonly observed in qualitative research (Francis et al., 2010). For this specific part of the research, a relatively narrow range of interviews was conducted within each epistemic community (12 in chemistry, 7 in biology, and 13 in physics), as saturation can be achieved with a small number of participants (Hennink and Kaiser, 2022).

3. Results

This section discusses the different meanings, interpretations and values associated with open science among natural-science researchers. It identifies three different kinds of open science meanings and the values associated with these meanings (see Table 3): (1) open science meanings as communication of research outputs (i.e., results); (2) open science meanings related to how the research process is carried out; (3) open science meanings more closely related to society; and values associated with open science.

Table 3.

Different imaginaries of open science.

Research outputs	Research process	Science–society
Open access	Sharing and interchange: Sharing culture	Levelling education
Open data	Sharing instruments	Science communication
Open code and open software	Institutional discourse
Infrastructure and platforms		Enabler of credibility

This section includes a diverse range of quotations and their respective participant codes to illustrate the themes discussed. The aim of providing representative interview quotations is to present evidence and illustrate the findings of each subsection (Eldh et al., 2020). Moreover, the bold text in each quotation of the interview extracts serves as an aid in highlighting the most essential and representative information available in the extract.

3.1 Open science meanings as research outputs

The interviewed researchers hold diverse conceptions of open science based on their day-to-day practices. One particularly expansive definition proposed by participants describes open science as the dissemination of the complete findings of a research project, centred on the distribution of scientific information and data.

Open science does not intervene in the research process; it intervenes at the final stage. It is a choice at the final stage. (P6.FP.P)

The issue of enhanced accessibility to data, papers and information in general was repeatedly mentioned by different researchers. Similarly, participants emphasized the freedom associated with being able to access different resources:

To have access to scientific literature and what is currently happening in science, which we are interested in, but then we can’t just be an expert since we are not affiliated with a user deal. So it's a huge freedom. (P10.PD.F)

3.1.1 Open science as open access

The dominant understanding of open science among researchers mostly centres on the unrestricted availability of knowledge in the form of academic papers. In this sense, open science is usually understood as a near-synonym for open access to articles. This conflation may stem from the fact that open access is the most visible aspect of open science policies and also the oldest of its elements: open-access journals began to appear in the 1990s, and preprint servers emerged even earlier. Consequently, this historical background has shaped an imaginary of open science as equivalent to open access.

I think it is, especially for papers, a better way to make the results available for more people. So, I think when we are working with open science, we are kind of choosing who can read. (BQ1.PD.B)

There is a confusion in my mind between open science and open access   …  I am more familiar with open access, and that is what we try to do when we publish a paper. For example, we always give open access to everyone, including a wide audience, non-scientific persons, so we pay extra fees to the papers to have open access. This is what it comes in my mind when you say open science. (B1.AP.F)

Similarly, in the field of chemistry across the three countries studied, it is evident that the concept of open science mainly focuses on making research papers easily accessible, rather than including a wider range of research outputs.

Open science to me is when after my studies [can] be able to read or [I can] access different papers. (C7.D.B)

3.1.1.1 Inequalities established in open access

Nevertheless, within the widespread association between open science and open access, more nuanced viewpoints also emerge. Some researchers highlight that open access extends beyond the simple availability of papers on journal websites. Instead, it involves addressing systemic problems such as high publication costs, profit-driven publishing models, and the necessity for fair mechanisms that ensure accessibility for all researchers, regardless of their financial limitations.

I just think there is a push for everything to become open access in the journal itself. And paying to have open access to the journal without solving the problems of too expensive companies that have high profit margins, and not responding to market pressures and without proper safeguards in place, means that people in Brazil, for example, are impeded from publishing because of paywalls, and it is not open access because open access is not just being able to download the paper on the journal website. Open access means that I can publish in the same place that my peers can publish without a paywall stopping me. (BQ2.FP.B)

Following this widely understood meaning, a clear position on gold open access becomes apparent.

3.1.1.2 Position on gold open access

Despite points of view arguing for fair open access/open science, some academics believe that paying article processing charges to get published is necessary to guarantee the widespread distribution of high-quality content, especially in internationally recognized (prestige) journals.

We pay. So, in my lab, because we are heavily funded, what we do is we pay to give, to grant open access in a paper in science. We pay to publish open access in Science or Nature … [some years ago] we paid Nature Genetics to grant open access, and it was like 2500 euros. It' s not so expensive … to us . But we accept doing this if the impact factor of the journal is huge. (B1.AP.F)

This position also involves an examination of equitable and rational charges for open-access publishing. While many acknowledge the legitimacy of publishers earning a profit and the need to cover operational costs, concerns persist regarding the fairness and transparency of pricing structures, especially when fees are perceived as excessively high. We have reached the point where the concept of open access, once widely celebrated, now also generates unease due to rising expenses and unanticipated consequences.

I don’t see a problem in that, as soon as the fee is fair and reasonable. The companies have to get some profit. They have to pay the cost, so someone has to pay. So, the only thing is that I don’t think the fee is fair enough. The fee is too expensive … I don’t know the costs of the publishers, but it doesn’t seem that 2000, [or] 3000 euros is fair. That's too high. (P13.PD.B)

Although open science is often equated with open access, and there is general consensus about this association, there is no unified understanding of what open access actually entails or ought to represent. The variety of perspectives reflects the contextual and contested nature of these debates, particularly concerning the commercial dimensions of open science and the availability of funds for gold open access. In well-funded laboratories, for example, paying to publish may be seen as a normal and justified practice, while in other contexts it raises questions about the acceptability and limits of such costs, even becoming an exclusionary barrier, another kind of paywall. These divergent views underscore that open access is only one aspect of open science. Other conceptions and values, distinct from open access, will be explored in the following sections.

3.1.2 Open science as open data

Another prevailing understanding of open science, specifically in relation to the availability of data, is embraced mainly by researchers deeply engaged in the field of physics. This trend is demonstrated in the viewpoints of physicists who emphasize the necessity of accessible data as a key concept of open science.

Open science means to me [that] we share our data and we share our methodology so people can reproduce our results. (P1.AP.F)

Open science is a possibility too; the fact [is] that we have to provide access to our production, scientific production and even our data. So even in my field in physics now, most reviews propose that our data can be in open access in order that colleagues can try again to do again the experiments using the given data and to confirm and provide new results. (P11.FP.F)

Despite this prevailing pattern, noticeable variations in the perception of open science can be observed. It is worth mentioning that physicists in France and Brazil commonly adopt open data practices, while their colleagues in Peru have only recently begun to do so, in a development described as relatively new.

This is much more recent, but there was already talk of a policy of making data available to all researchers … The data should be available so that researchers can do what they want, that is, have a greater capacity to exploit the data. (P6.FP.P)

The variation in the timing of adoption suggests that open research data has become widely integrated within physics communities. Still, the uneven pathways through which it is incorporated into research practices underscore the contextual nature of open science adoption across different parts of the world.

3.1.3 Open science as open code and open software

The connection between open code and open software is a significant topic in current discussions about open science. These two components are closely intertwined, as code requires software to run, and both play a crucial role in collaborative knowledge production.

Open codes and [their] development means that people can share with each other. (BQ1.PD.B)

However, open science in this sense is not limited to open data alone but also involves having open tools for data analysis, which eventually can create reproducibility.

Another thing is whether they publish the sources too, as tools to analyse the data. And another issue that I am very worried about is reproducibility. So, to what extent, if I take … The same data set, and I use the same tools, do they get the same result? … It's not clear. Because for instance, if you have computer codes, they will depend on the version of the compiler. It will depend on the modules and so on. And sometimes you have slight differences, and the results will not be the same. (P12.FP.F)

Nevertheless, there is a clear shift towards open-source technologies, demonstrated by the growing influence of collaborative repositories such as GitHub and GitLab, marking a significant transformation in research methodologies.

3.1.4 Open science as infrastructure and platforms

Within the understanding of open science as access to and communication of research outputs, an emerging perspective envisions the creation of scientific platforms that move beyond traditional formats. Conceived as dynamic spaces, these platforms would provide access not only to articles but also to a broader array of research outputs, including audio recordings, datasets and other scientific materials. This vision underscores the need for infrastructures capable of supporting and making such diverse outputs discoverable.

I would say we need scientific platforms, just some platforms which we can view as a kind of space where you can open[ly] access to audio, scientific inquiries, data, any request. I think it doesn’t matter what kind of form you are putting into … the code, or it could be a manuscript or it could be whatever, but it' s just open for the people to go and verify it . (P10.PD.F)

At the same time, this conceptualization resonates with broader debates on the platformization of science, a development that has been met with scepticism and even criticism by some researchers. Notably, only one interviewee articulated this understanding of open science as infrastructure, suggesting that it is not yet widespread among natural-science researchers. However, the fact that this view was expressed by an early-career researcher (a postdoc) makes it an imaginary worth further attention, as it may suggest a generational shift, from conceiving open science primarily in terms of open access to articles, towards imagining infrastructure itself as the central locus of openness. While promising in scope, this trajectory also raises concerns about dependency on centralized platforms and the further entrenchment of science within platform capitalism.

3.2 Open science meanings related to the research process

Across the different narratives about what open science is, several nuances emerge concerning how the research process is carried out, particularly in relation to researchers’ practices and lived experiences.

3.2.1 Open science as sharing and interchange: sharing culture

One prevailing understanding of open science is that it is the sharing of knowledge and the procedures of the production of that knowledge. From this perspective, researchers across disciplines and countries are rethinking how information is exchanged, questioning traditional notions of ownership over data, information and results, and moving towards more collaborative modes of working. In other words, this view emphasizes not only the dissemination of knowledge but also the openness of methodologies and research processes.

What emerges from this perspective is a broader cultural shift in scientific practice. Patterns of information sharing are evolving across fields and national contexts, signalling a reconfiguration of norms not only around ownership but also around collaboration. While uneven in scope, these changes suggest that open science is fostering a gradual transformation in how knowledge is produced, circulated and legitimized.

Because science, by definition, is meant to be shared. I don’t see the point of keeping information to oneself. (C1.D.F)

Nowadays people are more open to sharing information and all the data they have generated . The jealousy that often existed in the past, where people felt that the information they generated was theirs and they would not share it with anyone, I think has already been overcome. The trend now is for researchers to share as much information as possible with their peers, and it is a form of feedback. I think that, in that sense, things have changed. (B3.RS.P)

Moreover, this sharing behaviour is correlated with collaboration in science. Conversely, a reluctance to share not only obstructs the progression of science but also hinders collaborative initiatives.

They like to keep the things for themselves, not to share with others , and I think this is not good, and also one very important thing [is] willingness to collaborate. (B2.PD.F)

The sharing culture also has a long-term economic impact. A closed approach to information sharing can have significant economic consequences, as it emphasizes the potential inefficiencies in producing tools and software that could otherwise be readily accessible through shared knowledge.

So, you lose time, you lose money, and because you don’t have this information, and (I) talk about software and tools like that … if you don’t have these two available, and you don’t have the money to pay for it. You need to develop by yourself. So again, you spend time, [and] money, and develop things that could be available if someone shares them . (BQ1.PD.B)

Another dimension worth highlighting is the practice of disseminating research methods. This extends beyond the simple sharing of results and delves into the complexities of the research process itself.

Share our methodology. (P1.AP.F)

Throughout these comments, it is evident that there are evolving dynamics of information exchange among researchers from the three different disciplines interviewed in this study. This reflects a cultural shift with consequences for scientific cooperation and commercial viability. Moreover, there appear to be no significant disciplinary or geographical differences in this regard at present.

3.2.2 Open science as sharing instruments

Another aspect of open science involves the collective sharing of scientific instruments. This is particularly evident in disciplines such as physics and astrophysics, which rely heavily on specialized equipment for observation and experimentation.

I mean, the tools to observe … and so this is a bit more difficult … in the sense that investments done by different countries usually go to the nationals of that country. So, for instance, typically the telescopes in France, they are open to the French community only. Now, this has changed in the past, I would say, 40 years , and the reason is that you want to maximize the scientific return of the instruments … and so, then, we realized that by opening up the goals for observing time, to the international community, there were many good ideas coming from other countries that would actually use the instruments and maximize the scientific output. This has been the case in France for a number of years, but then, France and Germany in particular created what we call the European Southern Observatory, whose headquarters are in Munich, with all the telescopes in Chile. And so this was the union of all the astronomical facilities in Europe and budgets that went into building these telescopes in Chile, and so the example that was shown, each country didn’t have enough budget to produce a big accelerator and likewise in astronomy. If we wanted to build big instruments that were too expensive, and by actually uniting all the budgets from different countries, we were able to build the best telescopes. (P12.FP.F)

Overall, it is evident that there has been a notable shift towards a more accessible approach in the exchange and sharing of instruments as part of the open science framework. The transition from nationalistic approaches to a more communal one, albeit not yet universally adopted, marks a significant change.

3.2.3 Open science as translation (of language)

Another interpretation of open science concerns multilingualism, which is largely described as a challenge. It is argued that there is a need for improved translation methods within the scientific community. The issue of multilingualism in science is significant, as research can be conducted in several locations, but language barriers impede the effective sharing and understanding of findings. Therefore, better translation could make science more openly communicable.

I think it would be great to have a specialist in the field that speaks both of the languages that could translate to other languages that they know … Probably there are like papers that already exist, but I think it would be really nice because sometimes we have a lot of papers in Mandarin, and they are all already in ideograms. And if we put on automatic translation, it will have an accurate translation. So, I think open science for me is more like free communication. (C8.D.B)

3.2.4 Open science as a simple institutional discourse, a critical view

Finally, another interpretation of open science questions the truthfulness and usefulness of the concept. Open science may be viewed as a fashionable topic of discussion, with many people sharing opinions but without strong evidence of its effectiveness.

How can you say these pilgrims? I don ’t know how many years ago when they just gave the word and spoke with a book. Right? I think we have to show more things because science needs evidence. (P4.FP.P)

This perspective underscores a broader issue; namely, the institutional pressure to adopt open science practices, even when individuals themselves do not perceive much personal benefit in doing so.

3.3 Open science meanings closely related to society

Another set of open science meanings emphasizes its relationship with society and the public understanding of science. These imaginaries focus on reducing social disparities by making education and science communication more accessible, improving the dissemination and comprehension of scientific information, and facilitating the transfer of knowledge beyond academia (i.e., public understanding of science). In this sense, open science is framed less as a research practice and more as a public good.

Notably, such perspectives were most frequently articulated by researchers from Brazil and Peru. Their emphasis highlights the contextual nature of open science imaginaries, underscoring how social and regional conditions influence the value attributed to this concept.

The exhibition in the sense of displaying all this information relevant to the results of a particular research, not as an example, that is, freely accessible by all interested persons … obviously, students or [those] from the academic area, but logically also for any interested person from the society. (B5.AS.B)

Open science for me means that everyone, as long as they’re interested in science, can, by their own means, go and learn about science. That's what open science means. (P2.D.F)

However, this concept of open science is complex. There is particular emphasis on preventing knowledge appropriation, meaning that, regardless of the availability of scientific information through open science, having a solid scientific basis is essential for an actual comprehension. This meaning highlights the mutually beneficial relationship between open science and the importance of scientific literacy.

Open science needs a scientific background anyway to use the knowledge. (B4.FP.P)

3.3.1 Open science as levelling education

Another meaning related to society is that cross-disciplinary knowledge sharing in open science can be viewed as an effective strategy for reducing educational inequalities.

So open science not only helps to learn from other disciplines but also helps to level out the economic and educational differences that exist worldwide. (C2.RS.P)

This means that, in the eyes of researchers, open science has the potential to be a transformative force, leading to a paradigm shift in how scientific learning is conducted and eventually contributing even to global socio-economic equality.

3.3.2 Open science as science communication

Open science encompasses various aspects that extend beyond academia and researchers, and science communication is an essential component. Interviewees from Peru often mentioned an association of science communication with open science as one of its central meanings.

It's been like a bridge really for everybody at least has a little bit of knowledge ; everybody has the opportunity to access that kind of knowledge, right? And now it's like people have at least dedicated themselves. I feel like they’ve taken an interest in the matter … I mean everyone can have that access … and form a position on science. (C3.D.P)

From this perspective, open science is understood as disseminating knowledge beyond academia in order to empower the public to form well-informed views on scientific issues. It is framed as a bridge between scientists and society, with the potential to reduce disparities in science education, helping to close gaps in science literacy and enabling non-academic audiences — such as interested citizens or students — to access scientific content.

However, this interpretation also reveals a limitation: interviewees tended to equate openness with one-way communication, emphasizing the translation of complex concepts into accessible language rather than fostering dialogue. In this context, the term ‘open’ largely refers to the distribution of knowledge to the public, positioning open science as a vehicle for dissemination and education. While valuable in making research more understandable, such an approach risks reinforcing a top-down model of science communication (i.e., science popularization), rather than encouraging reciprocal exchanges between scientists and society.

There is a level of what it is to communicate to an undergraduate student, to the level of how to communicate [to] a postgraduate degree. And there is a level of how to communicate, for example, to the person in the bus, to the person in the shopping centre … There are different ways of communicating science, and not all of them are going to digest it in the way you would expect …, communicating, teaching, showing with evidence. (P4.FP.P)

The term ‘science’ has several definitions; the one that I see as the most relevant is that which has to do with basic research, in principle with mathematics and physics, which are the pillars of knowledge, from which we understand, and describe what exists in nature and what does not exist in nature and which we have created artificially … The term ‘open’ is, in my experience that I have had as a communicator and disseminator, to bring these basic principles that may suddenly appear in the university, wrapped in many equations, [and] in a lot of advanced mathematics, and convey them in a language and in a way that, to all people, the message, the idea, the concept can suddenly reach them in some way … I understand the term ‘open’ mainly in terms of dissemination to the public . (P3.AP.P)

The interviews revealed the complexity and diversity of open science, particularly in its close connection to science communication. In this view, open science is framed as a vehicle for dissemination, education and knowledge mobilization, functioning as a bridge between the scientific community and the public, and at times even compensating for deficiencies in basic science education.

Linked to the topic of science communication is a distinct meaning related to what interviewees called ‘science translation’, which is more closely associated with the public understanding of science. During the interviews, participants referred to it as a way of communicating science in a way that is understandable to people with varying levels of education, thereby making it more accessible to the broader community.

Like science, I think it has been like a bridge really for everybody at least have a little bit of knowledge; we all have the opportunity to access that kind of knowledge. (C3.D.P)

The issue extends beyond accessibility to include translating and sharing knowledge in a way that appeals to a broad range of audiences.

I think it means to me to give them (the public) the opportunity through the knowledge we have. Translate for them and disseminate to them in a more accessible way. (C3.D.P)

This specific open science meaning — ‘science translation’ — reveals how difficult it can be to make scientific information in the natural sciences understandable to a wide audience. The interviews clarified the complex nature of this meaning and highlighted the difficulty of articulating complicated scientific ideas in terms that people with different educational backgrounds can understand.

3.3.3 Open science as an enabler of credibility

This final meaning of open science relates to the fact that it has become a topic of significant discussion. In this sense, a meaning associated with prestige emerges, which translates into a question of institutional credibility:

I think it would be, when I start to publish, some type of article because, if I put [it] on a platform that has like open access, it will help our university to get more recognition and more research [and] to be more known. So, in this sense, I think there will be an impact . (C8.D.B)

It is evident that there are advantages associated with adopting open science practices, which can be translated into symbolic capital, benefiting both the researcher and their institution.

3.4 Values associated with open science: A departure from Mertonian values?

3.4.1 Science is open (scientific ethos adapting?)

According to some researchers, the fundamental nature of science lies in its openness. This characteristic has played a crucial role in the development of scientific understanding, as demonstrated by the historical shift from secrecy in alchemy and astrology to the establishment of the fundamental principles of modern science. It becomes clear that openness is not an added component but rather the essence of scientific research, deeply embedded in its historical origins.

Science is open, you know, science is a universal commodity for mankind. It should be as open as possible. (BQ2.FP.B)

Science must be open. In fact, this is the characteristic of science. If you think a bit about the history of science, you realize that in the old times, science was secret … alchemy was secret. Astrology was secret. Everyone had different recipients, right? And the sudden change in the seventeenth century was to say, no, this is no longer the rule. If I am to believe your result, I can … I should be able to reproduce it. And to reproduce it, you have to tell me how you did it. And that's also how the first scientific journals appeared … So that everyone knew about what observation you made, what experiment you carried out, so at the roots of science is openness. So open science, I think, is a misnomer because science is open. (P12.FP.F)

To summarize, ‘science is open’ is not only a current idea but a long-standing notion profoundly ingrained in the historical development of science. Therefore, it has become evident that openness is not a supplementary attribute but an intrinsic feature of the scientific ethos.

3.4.2 Open science as free and democratic

It is argued that access to current scientific literature can provide freedom to researchers, particularly those working independently and not affiliated with specific institutions:

To have access to current scientific literature and what is happening in science, which we are interested in, but then … we can just be an expert not being affiliated [with an institution] with a user [subscription] deal … So it's a huge freedom and, I think it's a huge thing. (P10.PD.F)

This specific value of open science highlights the significant influence that institutional affiliation can exert on an individual's capacity to participate in science. The contradiction of demanding competence while facing the constraints of institutional affiliation reveals the complex dynamics of the research environment and the role of open science in shaping them.

Similarly, another emerging theme is that open science is not intrinsically democratic, but it has the potential to become a transformative force that advocates for inclusion and accessibility.

It was nice how we are trying to find new ways to facilitate access for everyone; it is still not that democratic, but I think we are going to a good place. (C9.PD.B)

Ultimately, the trajectory of open science appears to be moving towards strengthening democratic principles.

3.4.3 Open science as universalism or universalization?

The universalization of research is seen as a crucial aspect within the broader framework of open science values, although this is a disputed issue. Dedication to universalization is presented as a guiding principle aimed at reducing disparities between research conducted by well-funded groups and others with fewer resources.

I believe in universalization when you don’t restrict countries. I believe that basically, as a principle, the universalization of information is something very important, especially for less favoured sectors, right? So there is research that is carried out by groups that are quite, let's say, with quite a lot of resources, and there are other groups that may not have those conditions that can benefit from this information that many times. (B5.AS.B)

Therefore, it is evident that the call for the universalization of information goes beyond simple rhetoric, entering the realm of inclusivity and aligning closely with democratic values.

3.4.4 Transparency and reproducibility as key values in open science

Transparency and reproducibility are seen as fundamental to the way researchers work and may be viewed as ingrained in the Mertonian value of organized skepticism. The recognition of transparency as a crucial value emerges across different disciplines, encompassing research methodology, the research process and the reliability of research outcomes.

It's a way to make research data more transparent so that people can check whether the research has been done correctly or in the best way, or if it can be improved. So, someone can … if everyone agrees that research has to be open, then there is less difficulty to access the details of the research, so it's about transparency. (P13.PD.B)

It is evident that transparency in research has become a widely recognized and highly valued attribute associated with open science practices.

In general, the process is transparent, that data is not erased, that sort of thing … I think it's a positive thing. If it is possible for discussions between peers to be published and that you have to show all your unedited data, your unedited photos, it can make you more confident that what you are seeing is recognised. Yes, it is a positive thing … one shows everything; one seeks to be transparent. (C5.RS.P)

Enhanced replicability of results by other colleagues is widely regarded as a cornerstone of rigorous research and is frequently identified by researchers as a core value of open science. Within this framing, reproducibility is conceptualized as a long-standing virtue of good scientific practice.

Our analyses are public, so people can reproduce them. (P1.AP.F)

I always preserve the idea that (science) should be verified , should be open to be verified by anyone who is interested in it. So, so it's a good idea … so, and I keep that in mind when I do research that my research should be easily verified … you have results and all your methods very properly organised and arranged. (P10.PD.F)

Nevertheless, researchers observe complex obstacles associated with achieving enhanced reproducibility. Challenges include issues such as the reliance of computer codes on variables, including compiler versions and modules, which may result in inconsistencies in outcomes. Therefore, obtaining this perceived value is not straightforward.

It's not clear, because for instance, if you have computer codes, they will depend on the version of the compiler. It will depend on the modules and so on. And sometimes you have slight differences, and the results will not be the same. And so, I think we should also work towards reproducibility in science. (P12.FP.F)

It is apparent that, although the scientific community — at least in the three data-intensive disciplines of biology, chemistry and physics — strongly encourages greater reproducibility as an open science value, there are challenges to overcome in order to achieve this objective.

4. Discussion

This section is articulated in two parts: first, the different meanings researchers ascribe to open science, and second, the values associated with it. Together, these represent the different imaginaries of open science, which are evidently linked to researchers’ own practices.

4.1 Open science meanings

The term open science encompasses a variety of definitions across disciplinary communities, but it is often grouped into three main categories: open science as research outputs, open science embedded in research processes and open science as engagement with society. This multiplicity of meanings reflects broader debates about openness, which is a contested and evolving concept (Albagli et al., 2015; Clinio, 2019; Fecher and Friesike, 2014) with no clear consensus on what it entails (Grubb and Easterbrook, 2011).

Previous literature, including comprehensive literature reviews (Vicente-Saez and Martinez-Fuentes, 2018), conceptual papers (Fecher and Friesike, 2014) and taxonomies (Da Silveira et al., 2021, 2023; Pontika et al., 2015), has examined what constitutes open science. However, very few studies have focused on experimental natural scientists, particularly those working in developing countries, or on the third meaning of open science, namely its societal dimension.

The most well-known interpretation understands open science as the sharing of outputs, particularly publications and datasets. Open science policies mainly focus on these two areas (Manco, 2022; Moradi and Abdi, 2023). This perspective aligns with Leonelli's (2023) conceptualization of open science as object sharing. Similarly, Miedema (2022) states that the most well-known initiative within open science is undoubtedly the open-access movement since it has been formalized since at least 2002, while David (2004) conceptualizes it as ‘public domain goods’. A related, though less common, interpretation is open code and open software, which was mentioned primarily by physicists, probably reflecting the epistemic culture in this discipline. This focus on outputs mirrors findings in other contexts: Lattu and Cai (2023), for example, argue that Finnish researchers typically equate open science with open-access publishing and, to a lesser extent, open research data, while rarely associating it with the research process itself. Such patterns suggest that many scientists still frame openness as a practice occurring mainly at the end of the research cycle, and thus outside of the process of generating results.

This clarity contrasts with previous research. Open science has often been portrayed as a highly ambiguous and contested notion, being in question or open to interpretation (Albagli et al., 2015; Fecher and Friesike, 2014), while Levin et al. (2016) observed that researchers often defined it by what it was not, rather than through concrete practices or values. More recent contributions, however, point to a consolidation of the concept. Vicente-Saez and Martinez-Fuentes (2018) describe open science in terms of four key principles — transparency, accessibility, sharing and collaboration — indicating a growing awareness among researchers of its core values.

Open science as embedded in the research process is the second category. This perspective highlights changes in how research is conducted, emphasizing the sharing of methodologies, instruments and infrastructures, including repositories and digital platforms. Previous research supports this focus, particularly regarding open instrumentation and science translation (Fiorini et al., 2023; Stirling, 2024). This second definition concurs with what has been previously stated by Levin et al. (2016), who note that researchers often stress the importance of repositories and databases, particularly in relation to the establishment of standards and metadata that shape possibilities for accessing, reusing and distributing research. However, the emphasis on infrastructures and platforms also raises critical concerns. As Plantin et al. (2018) argue, open science infrastructures could lead to a concentration in a few large, centralized entities, mirroring trends in scholarly communication infrastructures (Ma, 2023; Pooley, 2022). Or, even further, Mirowski (2018) interprets open science as embedded in platform capitalism, where ‘radically collaborative science’ becomes intertwined with market-driven logics. The fact that researchers increasingly frame open science through infrastructures and platforms suggests that the process of platformization has, to some extent, already been normalized.

The third category concerns the social dimension of open science and is closely intertwined with the relationship between science and society, such as the dissemination of scientific information through public and educational practices. This conception reflects ‘Mode 2’ science, in which research is embedded within societal contexts (Bucchi and Trench, 2016; Gibbons et al., 2010; Miedema, 2022). However, a tension arises: while researchers emphasize meanings and values tied to science communication, education and the social responsibility of scientists, open science policy frameworks continue to privilege research outputs over societal engagement (Chtena et al., 2025; Manco, 2024; Moradi and Abdi, 2023).

A further dimension identified by the interviewed scientists is open science as a means of science communication. Here, openness is often conceived in terms of disseminating ready-made knowledge to the public, reflecting a model of science popularization. On the contrary, the model of participatory science or citizen science, which may be closer to the ethos of open science (Bucchi and Trench, 2016; Trench and Bucchi, 2010), was largely absent from their accounts. This presents a paradox: although open science is frequently associated with public engagement, researchers tend to frame it as one-way communication rather than a dialogical exchange with society.

At the same time, a shift is visible in how open science is framed. Boon et al. (2022) observe a movement away from a narrow focus on open access to publications, data and software, towards a broader vision of open science that fosters partnerships between academia and society, which is closer to the meaning proposed by Grand et al. (2012). This meaning is paradoxical because, as illustrated by Miedema (2022), society remains largely absent from the current credibility cycle of science, even though, under a genuinely open science ecosystem, engagement and dialogue with the public could become integral to scientific legitimacy, forming an essential part of the credibility cycle of scientists.

Finally, it is worth noting that, even though an innovation-oriented framing of open science exists in the literature (Caulfield et al., 2012; Vicente-Saez et al., 2020), this meaning did not emerge during the interviews as a definition of open science in any of the cases.

Overall, the meanings proposed by the interviewed scientists overlap with the different open science taxonomies (Da Silveira et al., 2021, 2023; Pontika et al., 2015), but only to a certain extent. For instance, the following specific meanings are missing from these taxonomies: open science as sharing instruments, as a discourse levelling education, and as public science communication (including science translation). This could be further explained by the particularities of epistemic cultures and material practices in laboratories, especially among experimental natural scientists. Although these taxonomies were constructed using literature reviews and interviews with open science experts, they were developed in a normative manner.

4.2 Values associated with open science

This section outlines the different values associated with open science and science in general. According to the interviewed researchers, open science is associated with values such as transparency, reproducibility, universalization of research and democratization. Some of these can be interpreted as extensions of Merton's original norms, while others suggest a reconfiguration of the scientific ethos. For instance, several researchers contended that openness is not a novel idea but an intrinsic feature of science throughout history (‘science is open’). From this perspective, it is the commercialization of research that has diverted science from its original ethos. This view is echoed by Cohoon and Howison (2021) and David (2004, 2008), who emphasize that openness has long been embedded in the scientific process. In this sense, the concept of open science is often framed not as an innovation but as a return to the true essence of science.

Merton's CUDOS norms remain a reference point. Open science literature frequently encompasses these established norms. Despite facing criticism, advocates of open science prioritize the pursuit of Mertonian norms as a crucial aspect (Cohoon and Howison, 2021). At the same time, these norms serve more as aspirational frameworks for the scientific community rather than fully attainable standards of practice (Von Schomberg, 2024). Hosseini et al. (2022) note tensions between Mertonian values and open science, yet evidence suggests that open science practices often adapt and extend these norms rather than contradict them. For instance, although it is argued that current academic performative pressures may reshape the Mertonian norms (Macfarlane and Cheng, 2008), results suggest that practices such as transparency and reproducibility are aligned with organized skepticism and universalism. According to the results, certain researchers make references to the Mertonian norms as embedded in their daily practices related to open science values, and this concurs with Hosseini et al.'s (2022) argument that open science practices provide a way to maintain Mertonian values in contemporary research environments. Specifically, these norms manifest as universalization and organized skepticism through an approach involving transparency and reproducibility.

Researchers frequently associate open science with the dissemination of complete results, such as articles or datasets, reflecting adherence to Merton's communalism norm by prioritizing the sharing of novel and significant discoveries. However, a generational shift is evident: early-career researchers (including relatively young professors) tend to interpret open science more expansively, encompassing methods, code and the entire research process alongside final research outputs.

Hosseini et al. (2022) highlight this divergence as a potential source of friction between open science practices and Mertonian norms. However, rather than representing a contradiction, this tension might signal an ongoing reconfiguration of scientific norms. By extending communalism beyond the sharing of results to the openness of processes, younger researchers could be reshaping the ethos of science itself, which suggests that open science is not merely a revival of Mertonian ideals but a transformation of them in line with contemporary digital and collaborative practices, which aligns with Von Schomberg's (2024) proposition that the scientific ethos evolves under digital and collaborative conditions.

The value of open science as a democratic principle is strongly linked to the idea of open access, which ensures the circulation of knowledge regardless of the researcher's origin or institutional affiliation. In this sense, open access is intrinsically linked to Merton's communalism, which asserts that scientific knowledge belongs to the entire community rather than to individuals (Fernández Pinto, 2020, 2022), contrary to Hosseini et al.'s (2022) argument that open access is more closely related to the universalism value. This resonates with the democratic school of open science outlined by Fecher and Friesike (2014). However, in this sense, it departs from Miedema (2022), who contends that open science can only contribute to an ‘open society’ where democratic conditions already exist. Critics, meanwhile, caution that discourses of democracy have also facilitated the platformization of science, embedding open science within structures of platform capitalism (Mirowski, 2023).

Empirical results further highlight transparency and reproducibility as defining values of open science. Open research data promotes communalism and organized skepticism, advocating for collective ownership of scientific resources and critical examination of methodology and ethical standards (Uygun Tunç et al., 2023). However, the results suggest that transparency and reproducibility make open science closer to organized skepticism. These two elements reflect an emerging open research culture (Nosek et al., 2015) that updates Merton's original ethos for contemporary conditions. Universalism, in this context, is reinterpreted as the globalization of scientific information, reinforced by open access, open science and digital infrastructures; this is rather a reconfiguration of universalism as the principle that academic knowledge should transcend national, political or religious boundaries (Macfarlane and Cheng, 2008).

Overall, the findings suggest that open science values represent an update rather than a rejection of Mertonian norms. Communalism, universalism, disinterestedness and organized skepticism continue to serve as legitimizing references (Cohoon and Howison, 2021). Yet the emphasis on universalization and organized skepticism in particular points to a shift in how these values are enacted in practice.

Open science, understood mainly as access to papers and data, inherently relies on digital infrastructures; thus, open science practices are closely intertwined with technological infrastructures (Cohoon and Howison, 2021) and the contemporary trend of platform logics (Fecher et al., 2024; Ma, 2023; Plantin et al., 2018). This convergence suggests that the ethos of science, historically described as the normative structure of science (Merton, 1942), is dynamic and evolving in response to technological, institutional and generational change.

Finally, although open science is widely regarded positively, some researchers remain sceptical, viewing it as a rhetorical trend lacking empirical proof of effectiveness. Interpretations and implementations also vary across disciplines and national contexts. Nonetheless, a set of shared values, particularly transparency, reproducibility and access, emerges as central to the evolving ethos of science.

4.3 Recommendations

This study highlights that open science is understood in various ways, many of which remain underexplored. To capture this complexity, both policies and research assessment frameworks should be revised to account for the full spectrum of meanings and values associated with open science. After all, open science constitutes a broader ecosystem in which research outputs are inseparable from the infrastructures, instrumentation and societal contexts in which they are embedded. Recognizing these interconnections is crucial for designing policies and assessments that reflect the multidimensional character of open science and avoid reducing it to a single, output-focused metric.

4.4 Limitations

This study has several limitations. The decision to examine open science practices among natural science researchers in elite institutions provides a specific disciplinary perspective. Different findings might have emerged if participants from other fields, such as the social sciences, had been included. Moreover, the focus on experimental natural scientists may have introduced a bias, as their practices, infrastructures and epistemic cultures differ from those of other disciplines.

5. Conclusion

This study provides an overview of the diverse imaginaries of open science among natural-science communities in three countries, two of which are located in the Global South. The findings show that the meanings and values of open science are highly context- and discipline-dependent, and at times extend beyond the normative definitions promoted by policymakers and international frameworks. For instance, open science as a research output is a shared conceptualization among both scientists and policymakers in all three countries. In contrast, open science as embedded in research processes and sharing cultures emerges primarily from scientists’ perspectives, while open science rooted in science–society relations is particularly salient in Brazil and Peru. These divergences underscore not only the situated nature of open science imaginaries but also how Global South perspectives contribute to expanding its scope beyond standardized policy framings.

Furthermore, it is possible to categorize these meanings and values into practice-based values and a science-based ethos. Open science values appear to be relatively consistent across researchers from different countries, in contrast to the highly variable open science meanings, which differ significantly by nationality and discipline. This suggests the existence of a shared science ethos, which has evolved from the classic and normative Mertonian values. Further research could focus on why the emphasis on the social aspects of the meaning of open science (such as science education and science communication) are particularly prominent among respondents based in South America.