Examining tensions between the past and present uses of scientific concepts can help clarify their contributions as tools in experimental practices. This point can be illustrated by considering the concepts of mental imagery and hallucinations: despite debates over their respective referential reliabilities remaining unresolved within their interdependent histories, both are used as independently stable concepts in neuroimaging experiments. Building on an account of how these concepts function as tools structured for pursuit of diverging goals in experiments, this paper explores this tension by re-examining the continued reliance of each concept on inverse characterisations inherited from the nominally-discarded ‘mediator-view’ of sensory-like mental phenomena (SLMP). In doing so, I seek to demonstrate how examining unresolved tensions can help highlight that entrenched associations can remain both integral to, and obscured by, the uses of concepts as goal-directed tools within experimental practices.

检查科学概念的过去和现在使用之间的紧张关系有助于阐明它们作为实验实践工具的贡献。这一点可以通过考虑心理意象和幻觉的概念来说明：尽管关于它们各自的参考可靠性的争论在它们相互依赖的历史中仍未解决，但在神经成像实验中，两者都被用作独立稳定的概念。基于这些概念如何作为在实验中追求不同目标而构建的工具，本文通过重新检查每个概念对继承自名义上被丢弃的“中介观点”的逆向表征的持续依赖来探索这种紧张关系。感觉类心理现象(SLMP)。在这样做的过程中，我试图证明检查未解决的紧张关系如何有助于强调根深蒂固的关联既可以作为实验实践中以目标为导向的工具使用的概念的使用，也可以保持其整体性和模糊性。