Friday, September 14, 2018

Spoilers vs. Keepers

There’s a certain kind of philosopher who plies the trade proving that things we believe in and care deeply about do not really exist: free will, mind, love, evil, knowledge, God, self, the external world. Stuff like that. Call them Spoilers.

The method of the Spoiler is straightforward: (1) Identify a susceptible entity and advance a definition of the term we use to refer to it. (2) Claim that this definition captures its Genuine Meaning. (3) Proceed to argue that there is, in fact, (or, better, logically could never be) anything in the world that satisfies this definition.

There’s another kind of philosopher whose mission it is to rescue these entities from the ignominy of non-existence. Call them Keepers. As I will use the term, a Keeper is not simply someone who attempts to refute a Spoiler’s argument; rather, she is someone who believes that the definition advanced misconceives the entity in question.

There are two different types of Keeper.

Keeper1, accepting the basic methodology of the Spoiler, engages him on his own terms, providing criticisms of his definition and defending an alternative Genuine Meaning. She then permits the Spoiler to respond in kind. Theoretically, this process continues until a mutually acceptable definition emerges. We call this the method of Reflective Equilibrium and it is how Analytic Philosophy gets done.

Keeper2 is someone who is disengaged from the search for Genuine Meaning, at least as a purely philosophical activity. She practices a different method known as Concept Explication. Explication typically occurs when the entity in question is the subject of ongoing scientific inquiry. Hence, when Keeper2 explicates a concept, she does so, not in an attempt to capture a quiddity, but rather in an attempt to clarify it for the purpose of further inquiry.

Consider an example: Knowledge. Most of us stipulate that knowledge is a fine thing and we humans have acquired quite a bit of it. A Spoiler might propose a definition of Genuine Knowledge that implies that to know P we must (in some sense) be certain that P is so. He may then argue that no such P could exist, since all beliefs must be based on evidence, which is inherently uncertain. Ergo, knowledge does not exist.

Whereas Keeper1 objects to the Spoiler’s definition on intuitive grounds, Keeper2 rejects it on methodological grounds. Keeper2 may allow that Spoiler’s definition accurately captures some traditional or intuitive meaning. But she knows that these often stem from immature theories of the entity in question. The failure of anything in the world to satisfy such a definition testifies to the shortcomings of the theory, not to the unreality of the entity.

In general, Keeper2 is interested only in definitions that serve some clear explanatory aim. So, if the subject is knowledge, her question will not be “What is Genuine Knowledge?” Her question will be “What notion of knowledge will best serve the attempt to explain how humans learn about the world?” (Clearly, scientific knowledge has progressed in the face of uncertainty, so the Spoiler’s definition will never do.)

Both types of Keepers should remain of interest to you, but I wish to better your acquaintance with Keeper2, whom I henceforth refer to simply as Keeper. Let’s look at a few other perennial targets of Spoiler and how Keeper might rejoin.

Free Will

Spoiler argues that everything we know about the physical world entails that free will is an illusion. For him, genuine free will involves a moment of choice between multiple options, all of which are equally available to the agent. Spoiler argues that choice, like all events leading up to it, is a purely physical process. Our feeling that it occurs in a “causal gap” is an illusion that itself must have a physical explanation.

Keeper suggests that Spoiler misconceives the nature of free will. He may be credited with accurately characterizing a traditional theory as well as its flaws, but all he has shown is that it is a poor theory. It is the theory that must be discarded, not free will itself. Keeper counters that free will just is the observable human capacity to consider various possible futures and to make decisions aimed at bringing one of them about. The important explanatory questions are how this ability evolved and how it is implemented in the human mind.

The Self

For Spoiler, the self is necessarily a Genuine Me that persists unaltered through all physical and mental change. Among other things it is supposed to explain how all of our various sensory modalities can be unified into a single coherent perspective. Spoiler argues that we have no introspective evidence for such an entity, nor any clue how such a thing could perform the functions assigned to it.

Again, Keeper’s response is to suggest that Spoiler is mistaking an inadequate theory for the thing itself. Again, she identifies the self with something that patently exists, namely a being with a self-conception, one that, as the renowned Keeper John Locke observed, “considers itself as itself, the same thing in different times and places.” For Keeper, a successful theory of the self so conceived will be one that explains, inter alia, how humans become aware of their own existence.


God, of course, is a perennial target of the Spoiler. For him, the genuine meaning of God is a traditional one: an eternal supreme person-like being who created the universe and the physical and moral laws by which it is governed. Spoilers argue for the non existence of such a being on a variety of grounds, but generally its radical implausibility, the paucity of evidence in its favor and varieties of fatal incoherence.

Keeper has no commitment to preserving an ancient theory of a creator God. Rather, she identifies God with what indisputably exists, and that is the operating principles of the universe itself. (This, of course, is the God of Spinoza and Einstein.) On the Keeper’s view a theory of God is simply a Theory of Everything.

G. Randolph Mayes
Department of Philosophy
Sacramento State


  1. Randy, Nice account. It occurs to me you may be too modest/cautious in saying that explication typically occurs where the concept in question is the subject of scientific enquiry. Your account can also be applied to normative concepts. E.g. Rights. The explicator ditches "natural rights" as nonsense on stilts and defines rights legalistically or behaviouristically. Or art. The explicator scraps the idea that art has an essence and redefines art as whatever people call art, or whatever is likely to be found in galleries and museums.

  2. Em, thanks, and yes you are completely right about that. That's a point I had to sacrifice to the 1000 word limit, but explication occurs in any context in which the term is being assigned a meaning for some technical purpose. It is in the explanatory context, however, that the explicator is in the best position to assert that she is continuing a recognizable effort to clarify, rather than to simply appropriate for an entirely new purpose.

  3. Interesting metaphilosophical post. Yet, at what point does the explicator(2) change the meaning of terms, such as with ‘God’? It appears that the spoiler can say that the explicator's notion of 'God' has a different meaning than the traditional one. Hence, there is an incommensurability of 'God' concepts, and there are different truth conditions to the two inquiries. The two are talking past one another in that they're talking about different things. Let us explore what follows if this incommensurability is the case. The spoiler can then say that relative to the traditional meaning of 'God,' there is no such God. This can be said without being rejected by the explicator. Contrary to your post, if there’s a difference in meaning, the spoiler need not butt heads with the explicator since they are talking past one another, like comparing apples and oranges. The explicator can’t reject the spoiler with the explicator’s own concept because their respective concepts are individuated. The explicator can’t say to the spoiler that you have “misconceive[d] the nature of free will.” The explicator can’t say you have debunked a traditional immature theory but “not free will itself,” where this is understood in the sense that the spoiler and explicator are debating against each other. This is because there is no genuine disagreement between the two. Notice that a restricted spoiler could even agree with the explicator's project while being an eliminativist on the traditional one. 'God,' 'free will,' etc. don't refer on the traditional meaning of the terms. Yet, on the explicator's disparate concepts and meaning, they do refer.

  4. Thanks John,

    I think I agree with most of what you are saying.

    Keeper agrees that there is nothing that corresponds to the definition proposed by Spoiler. But Spoiler is not content to say simply that there is nothing that corresponds to this definition. He says that this definition captures the Genuine Meaning of the term, and therefore that entity of interest does not exist. Keeper thinks this is a confusion. Whether this is a Genuine Disagreement (or just, as you say, philosophers talking past each other) depends on what they both mean by Genuine Meaning, which I haven't addressed.

    One way to construe it as a Genuine Disagreement, though, is to think about the way that science works.

    When scientists learn that their theory of X is wrong, they propose a new theory. They don’t typically conclude that X does not exist. (Though they often conclude that there are many different types of X, or that some things they thought were X’s are not.) So, Keeper’s move is to construe philosophical debates in these terms, and typically it is to characterize the term of interest, not as attaching to any particular theory, but to the phenomenon we are trying to explain.

    Of course, Keeper understands that there are times when scientists explicitly discard theoretical entities, such as phlogiston, impetus and the luminiferous ether. But this is because they unequivocally belonged to the theoretical vocabulary, a distinct observational vocabulary is relatively well-established and and it has become possible to define experimental procedures that most agree are sufficient to detect the entity in question.

    In any given case it is open to the Spoiler to respond that Keeper is just wrong about how these terms are used. For example, he might insist that Free Will refers essentially to a specific theory of how human decision-making occurs, not the process of human decision-making itself.

    My view is that one of the defining characteristics of philosophical inquiry is a certain degree of fluidity in this area. If this is so, then Keeper’s rebuttal ends up being that if this is an entity to which we attach a great deal of importance, we might as well attach it to something that we agree to exist.

  5. Randy,

    Could your view make room for a ‘Keeper3’?

    Your Keeper1 seems to be a logician and semanticist whose goal is to chisolm away at definitions until we arrive at one that fits intuitions common to Seekers and Keepers.

    Your Keeper2 is a system builder who holds that clarifying concepts is in part a constructive enterprise. For instance, explicators in the philosophy of mathematics don’t need to believe that any definition of, say, ‘ordered pair’ will actually ‘capture a quiddity’. It seeks an explanatory theory of them that fits into a broader useful understanding.

    But what about us in the quiddity capturing business? Our self-understanding seems to fit neither. We think that some things we believe in and care deeply apart form genuine kinds: features of reality independent of our explanatory and classificatory practices. Yet we have a cognitive access to those things that points beyond our conceptual scheme. Such access derives from different sources, but generally from our capacity for demonstrative thought. We encounter them because, well, they’re part of reality and so are we. If our concepts are inadequate to grasp them, or our explanatory practices to fit them into our current broadest useful understanding, so much the worse for our concepts and explanatory practices.

    Indeed on this view that’s how philosophy progresses: by developing novel concepts and practices in order to understand those things. And we cannot know in advance what novel concepts and practices we will develop, since to predict such an advance would be just to make the advance.

    Here’s another example besides the ones you cite: causality. The Humean Spoiler asserts that if there are causal relations, we have no cognitive access to them; all we can hope for are correlations. But we just do have such access. We observe causes quite frequently. As John Searle pointed out, when a car backfiring makes you jump, you experience the causal tie. You don’t need two of them to form the connection.

  6. Tom, thanks for the close reading!

    I suppose I would have been content to think of your Keeper3 as a Keeper1, since I don’t detect any deep methodological differences. To me your Keeper3 is just someone like yourself who has thought far more deeply than the average chisolmer about what sort of commitments would tend to justify this method in this day and age. Still, I am very happy to acknowledge such a creature.

    I’m glad you brought up causality. I tried to incorporate that into the piece, but it just got too long. I agree with your assessment of Hume on this, but I don’t find Searle’s argument compelling. To me that is just one of his usual appeals to intuition that he tends to inflate with a lot of patented bluster and condescension than compelling argument (unlike, e.g., your own far more compelling defenses of formal causality in these pages.)

    Hume’s rejection of causality fits with his rejection of the self, as well as his external world skepticism in the sense that they all depend on construing observation in Cartesian terms (i.e. phenomenalist and internalist). He would never have anticipated the success of 20th century empiricism, which fundamentally depends on positing unseen entities.

    I think you and I agree that a coherent notion of observation presupposes the existence of an external world. But we may depart somewhat in how we construe evidence for cognitive access to things in this world.

    As you know, I think more like Keeper2, and what I would say on her behalf is this:

    Causes are theoretical entities that explain certain features of our experience. What we experience in the world is a combination of regularity and randomness, but one obvious way of explaining the regularities is that they are caused. This theory has been wildly successful. The cognitive process most responsible for our separation from the rest of the animal world is our ability to do causal reasoning. So, anyone who does not believe in causes has to explain why our ability to think in terms of them has been so incredibly fecund.

    Luckily for Keeper2 there is today a very robust scientific research program into the nature of causal reasoning (both neurological, psychological and Bayesian), so if a Keeper2 wants to have an impact, that’s where she will shine her analytical torch.

    The obvious question for Keeper3 is why he doesn’t just avail himself of this kind of thinking, rather than claim to have a different kind of cognitive access that it seems unnecessary to posit. My guess would be that it is because Keeper3 is more into incubating new research programs rather than contributing to existing ones.

  7. Is it possible to be a Spoiler about some things, a Keeper1 about others, and a Keeper2 about yet others?

    More weirdly perhaps, is it possible to be a Spoiler or a Keeper1 or a Keeper2 about some types of thing on different occasions?

    My somewhat awkward example is St. Paul, who seems (on one reading) to have taken a different tack in the city of Ephesus than he did in Athens concerning whether (and how) pagan idols might refer to a deity.

    In Athens, Paul's speech to the Epicurean and Stoic philosophers seems to incorporate the idea that one of the idols in their city "to an unknown god" was actually pointing to a known god (who, Paul explains, really does not want idols): "as I walked around and looked carefully at your objects of worship, I even found an altar with this inscription: to an unknown god. So you are ignorant of the very thing you worship—and this is what I am going to proclaim to you." (Acts 17:23) This might paint him as a Keeper2 in the following way:

    Spoiler: "to an unknown god?! Pshaw. Ain't any such thing."
    Keeper2: "wait a minute, the definition's more flexible. Course, I'm a Jewish monotheist. But let's have a little epistemic charity here; "unknown" says more about the manufacturer of the idol ("it's unknown to us") than the essential properties of the deity ("it's unknown to anyone"). Once you get the definition tweaked a little bit, you can see the thing in question really exists..."

    In Ephesus, on the other hand, Paul's reputation led the local idol-makers to say this: "this fellow Paul has convinced and led astray large numbers of people here in Ephesus and in practically the whole province of Asia. He says that gods made by human hands are no gods at all." (Acts 19:26) This sounds like Paul might be acting like (or at least is being perceived as) a Spoiler concerning the deity supposedly corresponding to an idol.

    In re-reading over my example, I'm not sure St. Paul was the best case of being a Spoiler on one occasion and a Keeper2 on another.

    Still, perhaps the first two questions of my response make sense?

  8. Thanks Russell.

    I'm not competent to evaluate your St. Paul example, but I think the answer to both of your questions is yes. The truth is that, while I think there are clearly those who just really enjoy smashing idols, I think most of us tend to keep here and spoil there depending on a lot of often pretty inscrutable factors.

    Your observation might steer us in the direction of recognizing that there is also such a thing as a Spoiler2, someone who will argue for the non existence of an entity in a way that is compatible with Keeper2's explanation-based methodology. The general form of Spoiler2’s argument might be that, even though we are still operating in a recognizably philosophical context, the entity in question has very clearly been introduced for the purposes of explaining phenomena for which we already have a well-developed observational vocabulary.

    For example, a Spoiler2 might argue that there is no such thing as a witch, not based on the Genuine Meaning of ‘witch,’ but on the basis of the fact that typically when we call someone a witch we are attributing the power to cast spells with certain causal powers, such as turning people into toads. She then she might argue that since there are no confirmed observations of people spontaneously turning into toads, there is no reason to believe in witches. In this context, a Keeper2 might conceivably suggest that we should use the word ‘witch’ to refer to what uncontroversially does exist, and that is people believed to have this power. But I think her argument is less compelling than those I mention in the post, because the explanandum is now completely different; i.e., why people turn into toads vs. why people think some people can turn other people into toads. There is nothing wrong with repurposing the term in this way, of course.

    In other contexts Spoiler2 might allow that the explanandum occurs, but that the theoretical entity just can’t do the job it has been summoned for. Philosophical attempts to develop a theory of sexual reproduction provides a lot of great examples of this. The theory of homunculi, e.g., never really got past the philosophical level, but it was discarded before microscopy could show that there are no little men inside human sperm. It just didn’t have the power to explain why children resembled there mother just as much (or little) as their father.