Instead of being (even partially) defined, thought experiments and models can be characterized. That is, we can describe their most central feature. My proposal is that thought experiments and models alike can be considered as representations the content of which is simultaneously hypothetical and empirical. I call this the “representational mingling” which is the merging of representations of empirical conditions alongside representations of hypothetical state of affairs. Goodman’s theory of representation is applied and greatly helps to describe the representational strategies that bring to the fore the mingled representations. The same piece of information can be transmitted through different representational tools and an exploration thereof reveals a great many of tools that are used across sciences.