Scientific imaging techniques have played an increasingly significant role in nanoscale research. But how should the resulting images be interpreted? What kind of information do they offer about their targets (the objects and relations they are about)? And how should that information be understood and disseminated (in particular, how reliable should the information be)? Given the plurality of images found in nanoscale research, it is unlikely that a single unified account can be articulated that accommodates all the images in question. However, this doesn’t mean that a general framework that helps us address central issues in the interpretation of nanoscale imaging cannot be provided.

In this paper, I offer such a framework. It provides a new conceptualization of nanoscale images and their content, by highlighting the following aspects: (a) In order to understand a nanoscale image, it’s crucial to determine which kind of image we are dealing with: How was the image obtained? And which sort of information is it intended to convey? (b) We should then examine the sources of bias that the images may contain: What kind of artifacts may be included in the image? (c) Finally, nanoscale images are often used as exemplars in the domain from which they emerged: How can such images be used as inferential devices that allow researchers to generalize the information provided in a particular image to other samples (whether in the same domain or in related ones)?

After motivating and articulating such a framework, I provide some case studies illustrating how the proposal works, by considering a variety of nanoscale images from microscopy (including probe and electron microscopy).