Philosophy Within Its Proper Bounds: The Overall Argument

Does responsibility require the possibility to have done otherwise? Does knowledge require safety? Can causation be reduced to some form of counterfactual dependency? Could a material duplicate fail to be a psychological duplicate? To answer these and similar questions, one must gain knowledge about metaphysical possibilities and necessities. One must know, for instance, that there could not be a situation where an agent would be responsible for what she has done while being unable to have done otherwise. Philosophy Within Its Proper Bounds (OUP, 2017) argues that we typically cannot gain such knowledge. I conclude from our modal ignorance that the philosophical issues that turn on this knowledge (which I call “modally immodest philosophical issues”) cannot be solved and that they should be set aside despite the long, venerable history of philosophical theorizing about them. Philosophy should then be reoriented toward projects that do not require the incriminated modal knowledge, including conceptual analysis and conceptual explication (aka conceptual engineering).

This week, I will discuss the main claims and arguments defended at length in Philosophy Within Its Proper Bounds (for a list of reviews and symposia, see here). Today, however, I present the book’s overall argument to give the reader a sense of how all the pieces fit together.

Let’s go back to the modally immodest philosophical issues. For instance, how can we know whether someone could be responsible for what she has done even if she could not have done otherwise? To gain such knowledge, philosophers describe actual or, more commonly, merely possible situations where agents could not have done otherwise and assess whether in such situations the agents would be responsible for what they’ve done. Philosophers may come to the view that there are possible situations where an agent would be responsible for what she has done although she could not have done otherwise, and they would conclude that the proposed necessary condition for responsibility is false. That is, philosophers appeal to what has come to be known in recent years as “the method of cases.”

Experimental philosophers and psychologists have studied the judgments lay people and (to a smaller extent) philosophers make in response to the kind of cases used to gain the relevant modal knowledge. They have found that these judgments often differ among demographic groups (people in different cultures, men and women, etc.) or that trivial changes in the presentation of these cases (e.g., slight changes in their wording) impact judgment. These two kinds of effects (demographic and presentation effects) should be taken into consideration in any metaphilosophical discussion.

Demographic and presentation effects underlie the two arguments Philosophy Within Its Proper Bounds develops against the method of cases. According to Unreliability, these effects show that the cases modally immodest issues turn on elicit unreliable judgments and that we ought to suspend judgment. The second argument is a dilemma: When people seem to react differently in response to a case either they genuinely disagree or their disagreement is purely verbal. According to the first horn of the dilemma (“Dogmatism”), if people genuinely disagree about what to say in response to a case they ought to suspend judgment because the parties involved in the disagreement are on an equal epistemic footing; according to the second horn of the dilemma (“Parochialism”), if disagreements are merely verbal, then our first priority should be to learn about what others are talking about instead of using the method of cases to theorize about what we accidentally happen to be talking about.

These two arguments lead to the same conclusion: Philosophers ought to suspend judgment in response to cases. Philosophy Within Its Proper Bounds examines critically various alternatives to the method of cases that might allow us to gain knowledge about the relevant metaphysical necessities and possibilities. Ultimately, I defend a moderate form of modal skepticism: The metaphysical necessities and possibilities modally immodest philosophical issues typically turn on are beyond our epistemic reach. I conclude that these issues should be set aside.

But then what is left for philosophers? Analyze and engineer concepts, I propose. Philosophy Within Its Proper Bounds presents a new characterization of conceptual analysis, which takes concepts to be psychological entities to be studied, at least to some extent, empirically. While conceptual analysis 2.0 differs from traditional conceptual analysis (it does not provide analytic truths or a priori knowledge), it is philosophically significant. Conceptual explication aims at transforming concepts that are in some respect deficient, and Philosophy Within Its Proper Bounds calls for philosophers to become conceptual engineers.



  1. Nate

    Does modal skepticism entail skepticism about counterfactual reasoning more generally? I often think things like, “If I had packed a lunch today, I wouldn’t be snacking so much.” I take that to be knowable by modest reasoners, and that we frequently do know such things. If knowledge of those sorts of propositions aren’t beyond are ken, what’s distinctive about the philosophically relevant ones that puts them out of reach?

    • Edouard Machery

      Nate, I defend a restricted form of modal skepticism: one that bears on far-fetched propositions about metaphysical necessities. You ask, rightly, what is distinctive about those propositions. Here we can only speculate: We know that our judgments are unreliable about those, but the source of this unreliability is not fully clear. In any case, here is my diagnosis: Our epistemic tools break down when we try to assess these propositions, in part because they are about situations where the properties that we use to apply concepts in usual situations are split apart. In a Gettier case, the agent has a justified belief, but her belief is true by luck, for instance, while in actual situations or in possible situations where modal reasoning is reliable, justification and truth obtained by a reliable (vs. lucky) process go together.

Comments are closed.