The use of fictional models is extensive and rewarding in modern science. This fact captured the attention of philosophers of science, who are focusing on questions such as the following: is it possible for a fictional model to be explanatory? And, if so, in virtue of what is such a fictional model explanatory? In this paper, I discuss these questions in relation to the realism vs. anti-realism debate in philosophy of science. I focus on work developed by Alisa Bokulich who has argued for the explanatory role played by fictional models, claiming that it can be accommodated by realism in ‘a broad sense’. To this end, Bokulich introduced what she called ‘an eikonic conception’ of explanation, as opposed to the so-called ‘ontic’ conception. This move, Bokulich claims, allows non-causal explanations to be acceptable to a modified version, as opposed to a ‘naïve’ brand of realism. Roughly, a fictional model can capture aspects of reality if it manages to act as a proxy for a theory targeting a physical system. As Bokulich explains with the example of James Clerk Maxwell’s use of mechanical models, this is due to structural correspondences between a model and a theory at a certain level of abstraction. However, Bokulich does not explain how the model grasps reality. In this paper, I fill this missing piece of the puzzle. While I agree with Bokulich on the explanatory function of fictional models, I shall invoke her reading of Maxwell’s work to draw a different conclusion. Specifically, I argue that structural correspondences at various levels form what I call ‘a ladder of abstractions’ which connects aspects of a target system to the corresponding depths of descriptions. Modifying in these terms the scope of ‘representing a system’, I suggest that Bokulich’s proposed modification of realism is obviated.

虚构模型的使用在现代科学中广泛且有益。这一事实引起了科学哲学家的注意，他们正在关注以下问题：虚构的模型是否有可能具有解释性？如果是这样，这样一个虚构的模型是凭借什么来解释的呢？在本文中，我将结合科学哲学中的实在论与反实在论辩论来讨论这些问题。我重点关注艾丽莎·博库利奇 (Alisa Bokulich) 的作品，她主张虚构模型所起的解释作用，声称它可以在“广义”的现实主义中得到适应。为此，博库利奇引入了她所谓的解释的“eikonic 概念”，而不是所谓的“ontic”概念。博库利奇声称，这一举措使得非因果解释可以被修改后的版本所接受，而不是“天真的”现实主义。粗略地说，如果虚构模型能够充当针对物理系统的理论的代理，那么它就可以捕获现实的各个方面。正如博库利奇以詹姆斯·克拉克·麦克斯韦使用机械模型为例所解释的那样，这是由于模型与理论在一定抽象层次上的结构对应关系。然而，博库利奇并没有解释该模型如何把握现实。在本文中，我填补了这个拼图中缺失的一块。虽然我同意博库利奇关于虚构模型的解释功能的观点，但我将援引她对麦克斯韦著作的阅读来得出不同的结论。具体来说，我认为各个层次的结构对应形成了我所说的“抽象阶梯”，它将目标系统的各个方面与相应的描述深度连接起来。通过用这些术语修改“代表一个系统”的范围，我建议避免博库利奇提出的对现实主义的修改。