Reviewed by:

• Hume’s Science of Human Nature: Scientific Realism, Reason, and Substantial Explanation by David Landy
• Emily Kelahan

As the title suggests, David Landy’s Hume’s Science of Human Nature: Scientific Realism, Reason, and Substantial Explanation defends a staunchly realist interpretation of Hume on scientific explanation. Landy’s forward-looking view sees Hume’s methodology in the Treatise as anticipating developments much later in history. He gives Hume a Sellarsian update, making his philosophy of science more impactful and contemporary than previously thought. The motivation for his view is twofold. First, he wants to respond to the long line of Hume critics who think his “science of human nature” is hardly science at all. According to this interpretive tradition (the Deductive-Nomological view), Hume’s “science” is merely the collection of empirical generalizations with no explanation involved. Far from deploying this lazy, shallow method, Landy’s Hume has a sophisticated methodology at the heart of which is a “perceptible model.” Just as Bohr proposed that we understand the structure of an atom (not directly observable) as analogous to the solar system (directly observable), Hume is best understood as treating simple ideas as theoretical posits in an experiential model based on purely descriptive phenomenology that both resemble and differ from direct experience. Secondly, Landy wants to distance himself from other recent attempts to update our understanding of Hume on scientific explanation, most notably the New Humean view. While Landy and the New Humeans agree that [End Page 109] Hume intends to do more than collect generalizations, the New Humeans think Hume concludes that a satisfying explanation is impossible, as the explanans is unknowable. The Deductive-Nomological view sees Hume as failing to reach for substantial explanation. The New Humean view sees Hume as failing to grasp substantial explanation. Landy’s interpretation has Hume reaching for substantial explanation and it aligns Hume’s grasp with his reach.

Chapters 1 and 2 lay out Landy’s perceptible model-based interpretation. He begins with Hume’s theory of mental representation, introducing the basic distinctions between simple and complex perceptions and between impressions and ideas, but then quickly treks into controversial territory. Landy elucidates something about which Hume is not explicit but which he likely cannot deny. We never experience simple ideas as simple. Experience is always complex. Simple impressions always enter the mind as parts of complex impressions. Because experience is always complex, our revival set for “simple idea” consists exclusively of complex ideas. What does this mean for the status of simple ideas? Simple ideas are like atoms. We cannot see them but we commit to them as theoretical posits because they make sense of what we do see. Landy sees us as having a choice. Explain the novelty of human thought and give up strict nominalism or give up a satisfying explanation of the novelty of human thought and keep strict nominalism. Experience alone licenses only nominalism, but to properly explain experience, you cannot be a strict nominalist. Landy opts to give up strict nominalism by construing simple ideas as unobservable theoretical posits. By basing them on a perceptible model, he minimizes the risks posed to Hume’s naturalism. As for substantial explanation, Landy argues that Hume takes no issue with substantial explanation per se. His problem, Landy explains, is with substantial explanation of inappropriate putative objects of explanation. Hume rejects and strongly condemns non-empirical methods. If a substantial explanation is explaining a bona fide empirical phenomenon, there is no problem by Hume’s lights.

Chapter 3 offers a deeper analysis of Hume on substantial explanation and explores the role of reason. Because Hume’s science of human nature is so rich and covers such an array of experience, it can be difficult to track Hume’s own ontological commitments. The vulgar and especially the false philosopher often talk of the substance undergirding human experience, but Hume clearly denies the legitimacy of their claims. They are not advancing anything like an acceptable substantial explanation of descriptive...

审核人：

• 休谟的人性科学：科学实在论、理性和实质解释大卫·兰迪
• 艾米丽·克拉汉

正如标题所暗示的那样，大卫·兰迪的休谟的《人性科学：科学实在论、理性和实质性解释》捍卫了休谟在科学解释上的坚定实在论解释。兰迪的前瞻性观点在论文中看到了休谟的方法论作为预测历史上很晚的事态发展。他对休谟进行了塞拉斯式的更新，使他的科学哲学比以前想象的更具影响力和现代感。他观点的动机是双重的。首先，他想回应一长串休谟批评者，他们认为他的“人性科学”根本不是科学。根据这种解释传统（演绎法-规则论观点），休谟的“科学”只是经验概括的集合，不涉及任何解释。Landy's Hume 远非采用这种懒惰、肤浅的方法，而是拥有一种复杂的方法论，其核心是“可感知模型”。正如玻尔提出我们将原子的结构（不可直接观测）理解为类似于太阳系（可直接观测），休谟最好被理解为将简单的想法视为基于纯粹描述性现象学的经验模型中的理论假设，既类似于直接经验又不同于直接经验。其次，兰迪希望与最近在科学解释上更新我们对休谟的理解的其他尝试保持距离，尤其是新休谟观点。虽然兰迪和新休谟主义者同意[结束第 109 页]休谟打算做的不仅仅是收集概括，新休谟主义者认为休谟的结论是不可能有令人满意的解释，因为解释者是不可知的。演绎法则的观点认为休谟没有达到实质性的解释。新休谟主义认为休谟未能掌握实质性的解释。兰迪的解释使休谟达到了实质性的解释，它使休谟的理解与他的影响保持一致。

第 1 章和第 2 章列出了兰迪基于模型的可感知解释。他从休谟的心理表征理论开始，介绍了简单和复杂知觉之间以及印象和观念之间的基本区别，但随后迅速进入了有争议的领域。兰迪阐明了休谟没有明确但他可能无法否认的一些事情。我们从来没有体验过简单的想法那么简单。经验总是很复杂。简单的印象总是作为复杂印象的一部分进入脑海。因为经验总是复杂的，我们对“简单想法”的复兴完全由复杂的想法组成。这对简单想法的地位意味着什么？简单的想法就像原子。我们看不到它们，但我们将它们视为理论假设，因为它们使我们所看到的有意义。兰迪认为我们有选择。解释人类思想的新颖性并放弃严格的唯名论或放弃对人类思想的新颖性的令人满意的解释而保持严格的唯名论。单独的经验只允许唯名论，但要正确解释经验，你不能成为严格的唯名论者。兰迪选择通过将简单的想法解释为不可观察的理论假设来放弃严格的唯名论。通过将它们建立在一个可感知的模型上，他最大限度地减少了休谟的自然主义所面临的风险。至于实质解释，兰迪认为休谟不接受实质解释 你不能成为一个严格的唯名论者。兰迪选择通过将简单的想法解释为不可观察的理论假设来放弃严格的唯名论。通过将它们建立在一个可感知的模型上，他最大限度地减少了休谟的自然主义所面临的风险。至于实质解释，兰迪认为休谟对实质解释没有异议 你不能成为一个严格的唯名论者。兰迪选择通过将简单的想法解释为不可观察的理论假设来放弃严格的唯名论。通过将它们建立在一个可感知的模型上，他最大限度地减少了休谟的自然主义所面临的风险。至于实质解释，兰迪认为休谟对实质解释没有异议本身。兰迪解释说，他的问题是对不恰当的假定解释对象进行实质性解释。休谟拒绝并强烈谴责非经验方法。如果一个实质性的解释是在解释一个真正的经验现象，那么休谟的观点就没有问题。

第三章对休谟的实质解释进行了更深入的分析，并探讨了理性的作用。因为休谟的人性科学如此丰富，涵盖了如此多的经验，所以很难追踪休谟自己的本体论承诺。庸俗的，尤其是虚假的哲学家经常谈论支撑人类经验的实质，但休谟明确否认他们的主张的合法性。他们没有提出任何像描述性的可接受的实质性解释......