Alchemy & algorithms : perspectives on the philosophy and history of open science

This paper gives the reader a chance to experience, or revisit, PHOS16: a conference on the History and Philosophy of Open Science. In the winter of 2016, we invited a varied international group to engage with these topics at the University of Helsinki, Finland. Our aim was a critical assessment of the defining features, underlying narratives, and overall objectives of the contemporary open science movement. The event brought together contemporary open science scholars, publishers, and advocates to discuss the philosophical foundations and historical roots of openness in academic research. The eight sessions combined historical views with more contemporary perspectives on topics such as transparency, reproducibility, collaboration, publishing, peer review, research ethics, as well as societal impact and engagement. We gathered together expert panelists and 15 invited speakers who have published extensively on these topics, which allowed us to engage in a thorough and multifaceted discussion. Together with our involved audience we charted the role and foundations of openness of research in our time, considered the accumulation and dissemination of scientific knowledge, and debated the various technical, legal, and ethical challenges of the past and present. In this article, we provide an overview of the topics covered at the conference as well as individual video interviews with each speaker. In addition to this, all the talks were recorded and they are offered here as an openly licensed community resource in both video and audio form. ‡ § | § § © Lahti L et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are


Introduction
The development of information technologies continues to transform science and society.This transformation is pushing the academic system towards greater openness on various fronts of the research process.The open science movement calls for increased openness in data collection, analysis, and collaboration; in education and in publishing and evaluating academic work (Fecher and Friesike 2013, Levin and Leonelli 2017, Nosek 2017).But how does the contemporary open science movement relate to earlier ideals and pragmatic implementations of science?
Since the 17th century, the academic system has gradually evolved from a 'language of alchemy' towards a more openly communicated and collaborative discipline; the definitions of the scientific process and scientific knowledge have become refined, and it has become more clear how they are distinguished from other, non-scientific forms of experimentation and reasoning.While technical and institutional aspects have largely dominated the contemporary open science debate, broader historical and philosophical perspectives have received less attention.This is paradoxical, as open communication and reproducible experimentation have been cornerstones of academic practice since the early modern period or even antiquity, and many of the current opportunities and challenges are conceptually similar to those of the past.More thorough historical and philosophical analysis can help motivate and guide the contemporary open science movement in addition to aiding it in recognising the varieties, significance, and limitations of openness in the scientific process.
Early motivations for open academic communication in the 16th and 17th centuries ranged from ideas of Christian charity to pragmatic needs of patrons to identify and attract the best scientists (David 2008, Vermeir andMargocsy 2012).The latter was greatly facilitated by clear and transparent reporting that could be subjected to peer review, and this operational procedure has become the organized peer review that is central to the contemporary scientific process.
Openness and transparency are essential for the quality and the efficiency of research.They can, for instance, accelerate reuse of data, methods, reported findings, and other research outputs; by helping to avoid overlapping activities and facilitating collaboration; and by promoting transparency and reproducibility.While the histories of science in general and of open science in particular overlap and parallel each other in significant ways, it is important also to recognise the fact that openness is an exceptional feature of the scientific venture.Digitisation and the evolution of the internet have in an unprecedented way enabled, and even created in some sense, what we now call the open science movement.There are also many aspects of open science today that do not have any conceptual precedents in history.But this is by no means a reason for us not to reflect on the history of open science and study, for example, Enlightenment ideas of science.What it does mean is that understanding the nature of open science includes accepting that the story of open science is not a straightforward continuum from antiquity and origins of the scientific system to the contemporary, often technically oriented debates on the role of openness in academic research.

Key outcomes and discussions
Here we see the reason that the concept of open science, viewed in light of its philosophical grounding and historical development, remains paradoxical: on the one hand openness is a central component of the definition of science; on the other hand various economic, ethical, technical, and social challenges tend to place practical restrictions.While technical innovations and the ideas of open science are increasingly utilised in the natural and social sciences as well as in the humanities (Fecher and Friesike 2013, Lahti et al. 2015, Nosek 2017), the overall role of open science in the history and present practice of science remains somewhat ambiguous.At PHOS16 (Fig. 1) we wanted to conceptualise these inbuilt tensions, economic, ethical, technical, and social, by examining open science under the following three complementary themes: as a discourse, a movement, and a practice based on the eight sessions including 15 talks and a panel, as detailed in the conference programme (Fig. 2).

Open science as a discourse
One purpose of PHOS16 was to offer complementary perspectives on openness as a core scientific value.Varying contemporary definitions of open science and their roles and limitations in academic value creation were covered by Benedikt Fecher's talk on the open science schools of thought ("Publishing and the Limits of Openness"; see also Fecher and Friesike 2013).This talk was complemented by Caroline Bassett's paper which considered  the broader historical and philosophical context of open science, and its associated uncertainties and risks, in terms of the concept of expertise ("Enough of Experts?Publics, public knowledge, and expertise").
The challenges for openness were considered on both structural and microsocial levels.Examples of structural challenges include technical and institutional aspects such as funding models, science policy and regulation, and communication technologies as well as social aspects such as gender bias and family upbringing, which can remarkably limit an individual's chances of entering and succeeding in academia.An example of the latter is the fact that access to the scientific community has historically been largely restricted to men, as was showcased by Camilla Mørk Røstvik in her talk on women and the Royal Society: "Too Ambitious'?The History of Women and Publishing at the Royal Society in the 20th and 21st century".Sociological mechanisms, technological development, and market forces are closely related issues that have been central drivers for, as well as threats to, openness in research practices since the Enlightenment.Manuela Fernández Pinto addressed these issues in her talk entitled "Open Science closed for business?"Sociological perspectives to open science include thinking about epistemic cultures and "how we know what we know" (Cetina 1999).These issues, as well as social structures in general, were discussed by Werner Reichmann in his talk "Open Science, Epistemic Culture, and Social Structures".
One hotly debated set of issues at the conference included the analysis of the concept of socially relevant knowledge, the idea of co-production, where non-academic communities can contribute to the knowledge generating process, and the tensions between expert knowledge and democratisation of research, where participation in the research and knowledge generation process could be open for everyone (Koskinen 2016).Open research practices often seem to refer to transparent reporting, equal opportunity, open communication, and the various technical aspects of research rather than to democratisation, or open participation in academic research.These issues were treated by Inkeri Koskinen in her talk on transdisciplinary research, titled "Commercialisation Threatening Openness in Transdisciplinary Research".

Open science as a movement
Open science is also an active movement which is driven by an increasing number of scientists adopting, innovating, and promoting open practices to improve the quality of research, collaboration, and overall visibility of their research.There are also conscious endeavours towards explicit policy goals seeking to influence prevailing institutions: the various national campaigns to increase transparency in the current academic publishing model provide one topical example.These campaigns aim at redefining the relationship between the academic community on the one hand and commercial publishers on the other (a Finnish example of such a campaign is www.tiedonhinta.fi).Various forms of, and efforts at, open participation, or democratisation constitute a crucial part of and a driving force behind the contemporary movement, including increased levels of participation, activist research, extra-academic collaboration, citizen science, or transdisciplinarity.For instance, data collection in ecology, biomedicine, and many other fields often relies on lay volunteers, and sharing research data with others is now making it possible for individuals to compare their genetic makeup or microbial fingerprints.An important related aspect is the question of authority and different kinds of expert knowledge as contrasted with citizen science.Caroline Bassett linked this question with what's called post-truth or alt-fact politics.On the other hand, citizen science enterprises can help build trust in and give insights into scientific practice.In the future, we are likely to encounter increasingly interesting, and at times difficult questions about citizens who do science, and scientists who are representing the views of citizens as opposed to an institutional point of view.
While there have been tensions between openness and profit-seeking, secretive, and often private interests (David 2008, Vermeir andMargocsy 2012), for instance, those of commercial academic publishers, the transition into more and more open science is ongoing on all levels within the academic community and a number of publishers and commercial companies have identified rapidly emerging business opportunities.Public interest has a remarkable role in defining the rules and regulations within the scientific community, including the standards for openness.While societal impact does not equal financial impact, open science can also be profitable.Openness can increase the overall efficiency of research and dissemination thereof, thus bringing added value to the allocation of resources.Its value can be challenging to quantify, however, and adoption of open practices has concrete costs that need to be accounted for within academic research and funding communities.Open science is simultaneously driven by various parties and its development is in part due to the tensions just enumerated rather than despite them.
Since the early modern period, scientific endeavour has been motivated by a desire for knowledge driven by ideas, not authorship, with the result that texts have often been anonymous.Opening medieval codices, one is struck by how often they are compilations of sources on one topic rather than assembled groups of works by one author.This creates significant challenges in assigning authorship and new technologies are providing new opportunities to study medieval scholarship, for instance by cross-linking and collaborative analysis of texts, over great distances and at an historically unprecedented scale.These new collaborative methods and ways of accessing sources are significantly changing the ways in which historical sources are contextualised and understood as well as highlighting their influence in the long term.The most recent developments in this field were covered by Jeffrey Witt in his talk referencing the work done at the Scholastic Commentaries and Texts Archive ("Texts as Networks: The Promise and Challenge of Publishing Humanities Texts as Open Data Networks").
The concluding panel discussed, among other things, the fact that in addition to openness in academic processes having intrinsic value in itself, openness can also have an instrumental role in advancing the generation, accumulation, and dissemination of knowledge which are central to the practical questions of shaping research policies and funding models.Moreover, issues of authority and the gatekeeping role of the academic community in validating knowledge was another aspect that was actively debated during the panel session.Whereas open participation and open dissemination of preliminary research outcomes were seen to pose risks to the credibility and authority of research, it was also brought up that these will also allow expanded forms of review, criticism, and engagement.

Open science as practice
In many ways, open science is a very pragmatic concept that promotes access to scientific outputs to facilitate the process of knowledge generation.All this is facilitated by technological development, as was illustrated by Scott Chamberlain's talk on contemporary Software and Best Practices to Facilitate Open Science.In the age of digital information, data processing and analysis algorithms are enabling new levels of scientific discovery, and have become increasingly central elements of the research process.At the same time, this is setting new challenges for transparency, reproducibility, and openness.Communitydriven initiatives such as rOpenSci, have emerged to support the creation and maintenance of high-quality software based on best practices, community engagement, and open source.Ultimately, however, much of the progress will depend on the academic incentives that can be set up to promote best practices and openness in scientific information processing.
Improved access to scientific research data, methods, publications as well as other outputs of scientific research will not only serve to increase transparency.It will also provide opportunities for further use, and innovative combinations, of ideas, information, and data.This, in turn, opens new avenues for creative thinking and scientific activity.
The recently proposed European Open Access Platform and the so-called diamond open access model have gained attention and provide one alternative that could replace the current commercial publishing system in the long term (see e.g.Fecher et al. 2017).
Modern technologies have led to revolutions in science and dissemination of knowledge, driving a culture shift in how we share, collaborate, trust, and evaluate scientific work (Laine et al. 2015, Levin andLeonelli 2017).Making peer review, including review comments, and potentially even the identities of reviewers, open and attributable can improve the quality of reviews, help to avoid conflicts of interest, and facilitate collaboration (Aleksic et al. 2014).However, further research is needed to evaluate the relative merits or potential disadvantages of opening the various elements of the review process.These issues were discussed in Michael Markie's talk on freeing peer review, "OPEN SESAME -Let's Free Peer Review and the Sharing of Research".His talk emphasized the role and responsibilities that academic publishers have in facilitating open peer review processes, and also addressed the technical challenges and criticism of open peer review.Whereas Markie's talk focused on open peer review from the publishing perspective, Samuli Ollila's talk on "Open Collaboration Method Developed in NMRlipids Project" considered aspects of open review in the context of academic software and experimental outputs.In contemporary science, critical assessment of the research process can form an important, yet often neglected, component of peer review.
New technologies come with remarkable practical and social challenges such as the storage of massive data collections, the need for increasingly sophisticated algorithms, and the overall dependence on technology, as discussed for instance by Scott Chamberlain and Samuli Ollila.Culture and norms do not always follow technical opportunities.While the digital revolution of recent decades has changed the way we communicate irreversibly, many popular features of the present open science movement, such as open data or electronic notebooks, are of concern to a very specific geographically, economically, and technically privileged set.We can therefore with good reason ask the following: on what grounds should we talk about the history of open science that extends beyond the digital era?These themes were discussed from a bird's eye perspective by Koen Vermeir in his talk "Open Science: The Big Picture".
Looking back at the historical and philosophical foundations of openness in the scientific process generally will be valuable for understanding, and also guiding, the ongoing transformation of the academic system into a more open way of doing science.One way to do this is to utilise established ethical frameworks as tools for the evaluation of and decision-making about open science (Mustajoki and Mustajoki 2017).The ethical aspects of open science research were addressed by Arto Mustajoki in his talk on fresh angles on the ethics of scientific research: "Open Data in the Framework of a New Approach to Research Ethics".
We believe these kinds of well-founded and systematic approaches to open science can and should be interwoven also into discussions of open science and its role in research ethics and integrity.An example of how this might be done was provided by Katrien Maes' talk on how European research institutions can be a force in making openness in science reality ("Europe Needs Ongoing Efforts to Promote Research Integrity -What LERU Universities Contribute to the Debate.").
The reasons for using cryptic alchemical notations in the 17th century have been transformed into new kinds of challenges associated with the complexity of modern information technologies.Despite this transformation, there are strong parallels between the different eras regarding issues of reproducibility, priority, scientific communication, funding, commercialization, and other aspects of doing science.These new forms of challenges contribute both to the modern reproducibility crisis and to the push towards more open science (Pinto 2015).Jennifer Rampling's recent analysis of medieval and early modern science and medicine (Rampling 2013) explored the issues of how early experimenters dealt with problems of reproducibility.Using the historical example of alchemy, she demonstrated some ways in which this was done also in her engaging talk on "Reproducibility and the Language of Alchemy".
The complex alchemical recipes that featured in Rampling's talk were then sharply contrasted by Scott Chamberlain's aforementioned talk on the latest modern practices of shared and reproducible software as well as Mikael Laakso's talk on the latest research on exploring the ecosystems surrounding traditional scientific publishing and feasible open access ("Charting the Evolving Landscape between Paywalls and Sustainable Open Access: Practices, Problems, and Solutions").The emergence and growth of modern open access publishing are making scientific literature available to the public at an unprecedented scale, although actual access to this literature is still limited by economic and social factors.It remains to be seen how the open science revolution will be able to shape societies in the future.

Conclusions
The open science movement has led to a transformation in the funding, regulation and communication of science with potentially far-reaching implications for the way the scientific system is organised.Open practices can help increase the quality of research and help create added value for society in the form of new economic incentives generated by enhancing the flow of information and access to digital methods, and by enabling new forms of collaboration.But we are still debating what precise role openness has, or should have, in the scientific process.This understanding can be furthered by philosophical and historical analysis.Viewed from a broader historical and philosophical perspective, what the open science movement highlights is the importance of adapting community norms and practices in a changing world and of ensuring that core values of science can receive continued support.
Open science is a vast concept that touches on nearly all aspects of research practice and dissemination.It covers social, cultural, ethical, and structural issues, and reflects broader trends in science and society.The PHOS16 conference highlighted a variety of historical parallels and pointed out new avenues for understanding and promoting openness in academic research.This was achieved by linking contemporary topics with closely related philosophical and historical ideas and processes with a particular focus on the early modern period.Whereas open science is an overarching concept that is associated with nearly every aspect of a research process, particular elements that were repeatedly debated included the challenges for transparency and reproducibility brought up by modern measurement and information processing technologies, the increasing role of digital communication and knowledge dissemination, research ethics, and openness of research process, evaluation, collaboration, and societal impact.
At PHOS16 we sketched the overall foundations of openness as a core scientific value, discussed the scope and limits of openness (Fecher and Friesike 2013), highlighted specific aspects of open science and its implications that deserve further conceptual analysis, and pointed at new avenues for promoting openness in contemporary research.We hope that this brief perspective, as well as the associated audiovisual material, will help the reader to better conceive the current open science movement as part of a continuum with historical roots.This conception, we believe, will serve to clarify the overall implications the open science movement has for science and society.

Audiovisual material
We are looking forward to this perspective article and the associated audiovisual material accelerating further debate on open science.Recordings of the live online streams of the

Figure 1 .
Figure 1.Around 100 open science scholars and advocates joined the PHOS16 conference in Helsinki to discuss the philosophical and historical roots of the contemporary open science movement.