This study addresses the question how to characterise subject-specific requirements for research infrastructure with a focus on the influences of Open Access (OA), in the general sense covering open access to literature, open data and open science.

The chapters in this study present subject-specific views on OA infrastructure for research by analysing research workflows as well as researcher behaviours. They specifically take into account two aspects, namely (i) working with literature and (ii) working with data. Throughout the preceding chapters and throughout this chapter, the topic of OA infrastructure is centred on digital resources. Even though there are many transitions between physical and digital resources mentioned – for example, between the human researcher

and the computer or a digital resource, and the physical, experimental as well computational facilities – these transitions will not be addressed explicitly in most of the cases for the sake of lingual simplicity.

The following sections will discuss commonalities of and differences between the different presented views on OA infrastructure and formulate recommendations for supporting the development of infrastructure (e.g. through funding initiatives) under specific consideration of the question how principles of “openness” or OA can be applied. In line with the qualitative approach of

this whole study, the synthesis will be provided as an interpretative account.

When comparing the chapters, the most obvious observation can be summarised in one word: diversity.

Attempting to provide a synthesis, the following sections will consequently analyse the commonalities and differences. This is done first on a high conceptual level and then on a detailed, systematic case-by-case basis. Thus, the resulting qualitative, rather than quantitative, account shall inform strategic decisions for future developments with respect to conceptual rather than procedural aspects. The specific measures, programmes or plans are assumed to be the result of these strategic decisions.