Forest ecosystems are shaped by internal dynamics and external factors like forest management. A modified physical appearance of ecosystems may affect the short- and long-term provision of ecosystem goods and services (EG&S) from forests. Ecosystem functioning research provides powerful tools for both, the mechanistic description of ecosystem dynamics and the normative assessment of the value of ecosystems. Holistic insights into the functioning of ecosystems require interdisciplinary projects. However, disciplinary differences in terminology and implicit underlying assumptions about relationships between descriptive ecosystem characteristics and their relevance for society may lead to misunderstandings hindering conceptual integration of disciplines. This highlights the need for a common language in ecosystem functioning research. Here, we provide a classification scheme of different types of ecosystem characteristics applicable across disciplines. We introduce a terminology that unambiguously differentiates between normative and descriptive uses of the terms ‘ecosystem function’, ‘ecosystem functioning’ and ‘ecosystem goods and services’. Based on this terminology, we propose a framework facilitating the explicit communication of hypotheses, underlying assumptions, analyses and conclusions belonging to three different aspects of ecosystem functioning: (1) mechanistic ecosystem functioning, (2) supplying ecosystem functioning and (3) ecosystem integrity. ‘Mechanistic ecosystem functioning’ addresses relationships and interactions between purely descriptive ecosystem characteristics and their dependence on external factors. ‘Supplying ecosystem functioning’ translates descriptive ecosystem characteristics into benefits for society in providing EG&S. ‘Ecosystem integrity’ assesses the capabilities of ecosystems to provide EG&S in the long term. We further propose the use of a mathematical interpretation of the term ‘ecosystem function’ as ‘characteristic A is the function of characteristics B, C, …, X’. This definition clearly accounts for complex relationships between ecosystem characteristics and avoids ambiguity inherent in the discrimination of ‘ecosystem-functions’ and other (‘non-functional’) ecosystem characteristics. Our approach will increase understanding across disciplines and foster joint syntheses by providing a unified framework for interdisciplinary studies.