Abstract

A number of thorny issues, such as the nature of time, free will, the clash of the manifest image and the scientific image, the possibility of a naturalistic foundation of morality, and perhaps even the possibility of accounting for consciousness in naturalistic terms, seem to be plagued by the conceptual confusion nourished by a single fallacy: the old fisherman’s mistake. This is the mistake that consists in disregarding the fact that knowledge is not just learning new facts about old concepts. Knowledge is often to realise the inadequacy and misleading character of some of our old concepts, and the intuitions that ground them.

Once upon a time, an old fisherman used to enjoy the sunset. The sky would explode in fiery colours, the sun would descend majestically and dive into the ocean, the sky would turn dark blue, and one by one the stars would light up. But a man from the city came along and told the old fisherman: “You know: the sun does not really dive into the ocean. It stands still out there and is always shining. What you see is only a perspectival show due to our movement.” The old fisherman was stunned. He trusted the man from the city and began to worry. The sunset is an illusion. Hence it is not real. He had been watching a non-existent event for years. He had been deluded all his life. If the sunset is an illusion—he thought—we cannot rely upon it. We must learn to think without sunsets. He tried, and it was a disaster: he did not know when to go to sleep anymore, he did not expect a sunset in the evening, and if he saw one, he repeated to himself: “It is an illusion, it is not real, there is no sunset, the sun never dives into the ocean: the sun is always shining, I should take this seriously, I should not go to sleep.” So, he couldn’t sleep anymore; he lost his mind.

The old man was obviously making a mistake. It seems a simple one, but it is a subtle one. I find precisely this same mistake repeated over and over in a number of domains. I think that this mistake confuses the discussion on several thorny and much debated issues. In this essay, I describe what precisely the mistake is and suggest that recognizing it could make those issues a bit less troublesome.

1 THE FALLACY

The question bothering the old fisherman is whether the sunset is real or illusory. On the one hand, the reality of the sunset is denied by the knowledge of the city man, which the old fisherman trusts. On the other hand, denying the reality of the sunset appears ridiculous and leads the old man to dramatic and absurd deductions. Where is the catch?

The catch is in the meaning of the concept “sunset.” The old man grew up with a clear-cut notion of what a sunset is:

(a) The sunset is the diving of the sun into the water of the ocean.

This, for the old fisherman, was what a sunset is. Its very nature. Its definition. The very meaning of the concept “sunset.”

A careful conceptual analysis of the old fisherman’s notion of sunset would have identified it as the dive of the sun into the ocean. When he is told that the sun does not dive into the ocean, the unavoidable conclusion is that there is no sunset.

But the rest of us, who know our Copernicus, are still very happy to talk about sunsets. We rely upon sunsets, we enjoy sunsets, and it would not even cross our mind to state that there is no sunset in our universe. Why?

Because—here is the key point—we have re-conceptualized the notion of “sunset,” adjusting it to our increased knowledge. Say:

(b) A sunset is the collection of real phenomena happening around us when the rotation of Earth moves us out of the lit part of the globe.

Let us assign different names to these two definitions: (a) defines “sunset-(a),” while (b) defines “sunset-(b).” The knowledge of the man of the city denies the reality of sunset-(a) but does not deny the reality of sunset-(b). So, a reasonable reaction by the old fisherman should have been something like: “Oh gosh, I always thought that the sunset I used to watch was a sunset-(a), but this is not true. The sunsets I saw were all sunsets-(b).” But sunsets-(b) provide the very same vision of the sun majestically descending towards the ocean, the sky exploding in fiery colours, the sky turning dark blue, one by one the stars lighting up, and so on. Furthermore, sunsets-(b) equally stir emotions, they equally tell us the time to go to bed, they are as reliable in their recurrence as sunsets have always been. Hence, the notion of sunset can play exactly the same role as before in the fisherman’s life, even if according to his strict definition (a) there are no sunsets at all.

To summarize: there is a concept (“sunset”) that plays an important role in the old fisherman’s life. He naturally understands this notion as sunset-(a). But sunsets-(a), he finds out, do not exist in nature. Sunsets-(b) do exist, however, and they play precisely the role that sunsets have always played all along. No trouble follows in replacing the definition (a) with the definition (b) and using the notion of sunset just as before. Believing that there is trouble is the mistake that the old fisherman makes.

Let’s draw the general lesson from the parable. There are concepts that play an important role in our dealing with the world. We can have a good conceptual understanding of them. We know what they mean. But it happens that increased knowledge about the structure of the world shows us that our understanding of the phenomena they designate contains something incorrect. There is, however, a modification of the same concept that does play precisely the role the original concept played. No trouble follows in replacing the old understanding with the new one.

Believing there is trouble is the old fisherman’s mistake.

In the next section, I illustrate how precisely this mistake plays out in a number of current debates. Before getting to that, however, one more observation is due. The account above might not yet capture the full story.

Some old fishermen are stubborn guys. Our old fisherman may well end up understanding all of the above but still feel a sense of loss. After all, he was emotionally attached to the idea of the sun actually diving into the water. He could almost visualize the splashes and the boiling around the sun immersing itself into the waves. So, he may mourn the loss of sunsets-(a). This might be psychologically comprehensible, and we would be culturally insensitive to deny him the right to his loss. We may feel sorry for him and perhaps even empathize a bit with him.

This is comprehensible. But if the old man fails to see that sunsets-(b) are real and if he denies that they are fully entitled to play the vast number of functions that “sunset” used to play in his life, then he is clearly just mistaken.

2 EXAMPLES

In what follows, I list several issues that are debated today, sometimes fiercely. I do not pretend to offer conclusive solutions to all of them, perhaps not even to say anything particularly original about them. My point is to notice what seems to me to be a common fallacy in the way all these issues are sometimes addressed.

Needless to stay, not all the authors involved in these debates fall into this fallacy. The fallacy is well recognized by some players in each of the debates. My aim here is to point out that it is the very same fallacy that creates confusion in each of these issues. This may make it easier to avoid it.

3 FREE WILL

We have the distinct experience of being able to make decisions that are free: they depend on us. Moral responsibility, legal theory, and our own understanding of ourselves rely upon the idea that the source of some decisions is the individual, who therefore bears responsibility for the decisions taken.

A naive way we can think about the nature of free will is the following:

(a) We have the power to steer the future along different paths, in a way that is not due to random chance, and without different future paths implying any difference at all in the past state of the world (including ourselves).

Call this “free-will-(a).” Now, the man from town, steeped in modern physics, tells us that free-will-(a) is incompatible with the laws of nature. In the classical limit, different futures require different pasts. In quantum physics, the past fully determines the probability distributions of future events.

Like the old fisherman, we panic. If there is no free will, our idea of ourselves crumbles. We agonize. We oscillate between trusting science or remaining faithful to our deep intuition about our freedom. We fear the scientific advances will jeopardize the very foundation of our legal system and our morality.

It is the same panic as the old fisherman’s. Consider a possible definition of free-will-(b). Say:

(b) At a given state of the external world, with given external stimuli and given internal memories, our future actions are still not determined. They are affected by complex mental processes, carried on by the physical system that we are.

Nothing in modern science denies the existence of free-will-(b), of course (see, e.g., Dennett 1984). Free-will-(b) is perfectly sufficient to underpin our sense of internal freedom, our legal system, and our morality; it justifies our sense that it is we who decide, and it accounts for all phenomena that we traditionally attribute to free will.

For instance, to the extent legal theory requires a notion of free choice, free-will-(b) is largely sufficient. We incarcerate persons on the basis of the assumption that their mental processes have led to crime in contexts and under circumstances in which other people would not have done the same. This notion of choice is coherent with free-will-(b). Whether we hold a dissuasive, a punitive, a redeeming, or a vengeful motivation for incarceration, in all these cases, the reality of free-will-(b) suffices not only to define legal responsibility but also to provide the rational motivation for incarceration in the concrete cases in which free choice is invoked. (I touch on moral responsibility later.)

Furthermore, it is also fully comprehensible why our (free, in the sense just clarified) decisions have led us to establish laws accordingly. Hence, there is no contradiction between any of this and modern science. To think that there is a contradiction is the old fisherman’s mistake.

Of course, there is more. Remember that some old fishermen, stubborn guys, could still feel a sense of loss at the idea that the sun is not actually diving into the ocean with a splash. In the same vein, some of us, stubborn guys, may feel a sense of loss about free-will-(a): namely, the loss of the idea that something in us is external to the natural realm and can influence it from the outside. This fact may be hard to swallow for some. During the Renaissance, some found it hard to swallow the loss of an immovable Earth. For this, I see no remedy but mourning.

But the point I am making is a different one: there is no contradiction between the reality of the phenomenon we commonly call “free will” and the full use we make of this concept, including in our psychology, our morality and law, and the discoveries of science. The reason for thinking that there is a contradiction is the old fisherman’s fallacy: failing to distinguish a rigid (mistaken) understanding of a concept from the actual (fluid) role that it plays within our overall conceptual structure.

To learn is to adjust our concepts, not to be enslaved by the way we intended them at some moment in time.

4 MANIFEST AND SCIENTIFIC IMAGES OF THE WORLD

The example of the old fisherman is a special case of the supposed tension between the “manifest” image and the “scientific” image of the world (Sellars 1962). A piece of matter appears to be continuous, whereas science tells us that it is mostly empty and just scattered with atoms. The ground appears static, whereas Earth is spinning.

The clash between the two images is easily resolved in the way illustrated above: realizing that that the manifest image is neither “illusory” nor in contradiction with the underlying complexity revealed by scientific research.

The error is to mistake an appearance for a definition. We have analogous situations in everyday life: seen from a distance, a forest along the side of a mountain appears as a uniform velvet green. This is a “manifest” image of a forest. But if I walk into the forest, I see trees, trunks, leaves, insects—a rich complexity. Does this make the “uniform velvet green” illusory? Does this create a conflict? It does not. Things look one way seen from a distance and another way seen from close up. The scientific image of the world does not clash with the manifest image; it is a more detailed perspective on the same objects. It does not falsify the manifest image; it grounds it, as the perspectival understanding of a sunset explains the image of the sun diving into the ocean.

Once again, it is the analysis of the meaning of our words that can be misleading. If we define a table as

(a) a continuous piece of wood,

we are posing continuity to be part of the conceptual definition of matter. Then we get endlessly confused in learning atomic theory. But if we are flexible in redefining what we mean by continuous matter at the light of the advances of science between the nineteenth century and the twentieth, there is no tension between the continuous surface of a table and its atomic structure. A table is simply

(b) an atomic structure that is continuous at scales much larger than the atomic scale.

The mistake of believing there is a clash is to take the manifest image as the definition of reality, instead of taking it, as we should, as a provisional, approximate, effective, useful level of conceptualization of reality, to be kept flexible in view of the development of knowledge.

5 THE NATURE OF TIME

I focus here on one specific aspect (among the many) of the debate on the nature of time: orientation, or “the arrow of time.” The macroscopic world we experience is time oriented. We witness phenomena that we never see happening in time-reversed order. Hence, our naive intuition about time is that it is necessarily oriented. Here is a possible definition of what time is:

(a) Time is an oriented flow.

Science has discovered, however, that all elementary evolution equations are time reversal invariant: if a phenomenon happens, the phenomenon obtained by time reversing it is allowed by these equations to happen as well (Price 1996).

If, like the old fisherman, we think that time is by nature oriented (as in definition [a] above), we are confused by finding out that irreversibility is a contingent fact regarding a macroscopic description of events, rooted in the entropy happening to be low in the past. If we hold on to definition (a), we deduce that there is no time. We panic. We see a dramatic clash between microphysics and experience. Should we believe in microphysics and declare that our entire experience of the unidirectional flow of time is illusory? Should we hold on to the cogency of our experience and alter the microphysics? Should we declare current physics incomplete, because it does not include time-(a) in its foundations?

None of the above is necessary, or wise. We are making the old fisherman’s mistake. We fail to recognize that we are indicating two different notions with the same name. There is a more refined notion of time useful to describe elementary dynamics, where time is not fundamentally oriented, while orientation appears only in the macroscopic account of phenomena. Say:

(b) Time is an independent parameter in the equations of motion, in terms of which the macroscopic account of physical phenomena in our universe turns out to be irreversible.

Time-(b) accounts for our experience, hence justifies our intuition, underpins all irreversible phenomena of the world, and is compatible with current microphysics. There is no tension.

Once again, some of us, stubborn guys, will keep mourning for the lost notion of a fundamental oriented time. O.K. to mourn. But to argue that a fundamental orientation of time (independent from the macroscopic approximation) is required to make sense of our experience of the word is like arguing that the sun truly splashing into the ocean’s waves is required for the sunset we see.

6 THE POSSIBILITY OF A NATURALISTIC FOUNDATION OF VALUES

I now touch a wider theme: the possibility of a naturalistic foundation of values, such as moral ones. Values can be held and studied by themselves and as such. But in the context of a naturalistic worldview, we can also ask for an understanding of how they relate to—say—biology, natural evolution, cultural evolution, and the like (Darwin 1982 [1871]). In this context, one considers the natural origin of values: they are not literally undersigned by God, they are not absolute Kantian imperatives, they are not valid by their abstract nature.

Here I am not concerned with the extent to which specific accounts of the natural origin of values are convincing. Rather, I am interested in a surprisingly widespread reaction to any such account: the idea that it empties the significance or the cogency of the values themselves, either because it relativizes them or because it undercuts their supposed foundation. The point I want to make here is only that this reaction is another example of the old fisherman’s mistake.

To put it pictorially, let us consider a wiser man than the fisherman of the initial story, one who does not panic so easily. This wise man loves his children and helps the poor. When asked why he does so, he answers: because this is what God wants from me. But one day the man from the city ends up convincing him that the reason he loves his children and helps the poor is because of his biology.

The old man listens carefully, thinks awhile, and then says: “It may be so, as you say, but I still love my children and I still help the poor.”

What has happened to him? He has simply allowed for the possibility of re-conceptualizing a real phenomenon, his morality, by changing his definition from:

(a) Love and charity are what God asks me to do

to:

(b) Love and charity are what I do because of my nature and culture.

The denial of the reality of (a) does not imply that the phenomenology of love and charity is false or ill founded, mistaken, or illusory. It remains real, because it can be re-conceptualized, due to a change of general cultural context. Whether (a) or (b) is more compelling depends on the cultural context, of course.

If in the same situation the wise man had panicked: “Oh gosh, if there is no God telling me what to do, there is no morality,” he would have committed the old fisherman’s mistake. This is the reaction illustrated by Dostoyevsky in a renowned passage in The Brothers Karamazov, which well represents the entire (misplaced) rhetoric against supposed twentieth-century “nihilism.” The mistake is to be stubbornly anchored to an interpretation of a rich phenomenology—morality—in terms of a strict definition of this phenomenology that relies on incomplete knowledge.

The remarks above allow us to consider the question whether moral responsibility (as opposed to legal responsibility) depends on free-will-(a). Is it put in jeopardy by the determinism of classical physics or the random probabilism of quantum physics? As mentioned above, we can consider values, including moral values, as such or study the coherence of the ensemble of moral phenomena with a naturalistic worldview. In the first case, there is nothing wrong in assuming free-will-(a): we are in the context of an incomplete account of [to get: of] reality, and within this account choice can be coherently taken as literally un-caused. If instead we inquire about the compatibility between free will and modern science, we are doing something else: we are placing (Price 2011) the moral discourse within a naturalistic worldview. Once again, there is no contradiction, because any naturalistic account of morality refers to free-will-(b), not to free-will-(a). The confusion comes only if we pretend the second remains valid outside the context where we can use it.

7 CONSCIOUSNESS

Finally, I find that the same fallacy informs the debate about consciousness. Chalmers’s well-known distinction between the “easy” and the “hard” problem of consciousness is rooted in this fallacy (Chalmers 1996).

There are substantial differences between the situation of the old fisherman and the other cases. In the case of the sunset, for instance, we have the “easy” problem fully solved: Copernican theory. In the case of consciousness, we are far from an equally solid understanding of the phenomenon.

But the recurring claim that there necessarily is a “hard” problem, over and above the unsolved “easy” problem, is to base the very formulation of the issue on a conceptual analysis of what it means for us to be conscious. Our intuition is that—whatever the phenomenon—to be conscious must be something over and above whatever process is happening in the brain.

Chalmers’s key argument, for instance, is the zombie argument: the same brain processes that happen together with a subjective experience can also be conceived to happen in the absence of this subjective experience. But why could one conceive this to happen? One can conceive this to happen only if one holds on to a conceptual understanding of ourselves as entities independent from the physical processes in the brain—“consciousness-(a).” It is the presumed understanding that implies we can conceive the separation between brain and mind.

A claim is made that we directly know the intrinsic nature of consciousness by being conscious, so we couldn’t be mistaken about its nature. Centuries of advances in knowledge teach us that there are no intuitions we can’t be mistaken about. This, I believe, is precisely what naturalism is all about.

It may well be true that consciousness-(a) is incompatible with the current naturalistic account of the world, but this is no argument against the possibility that the “easy” problem could lead to an understanding of brain processes where a “consciousness-(b)” concept can account for the full experience of subjectivity.

In simpler words, the intuition that authors like Chalmers (1996) and Nagel (2012) defend is just what it is: an intuition promoted to a definition, or a certainty. Like the intuition of the old fisherman, also promoted to a definition, that the very nature of sunset is the sun splashing into the ocean. Intuitions turn out to be wrong. Especially when they find themselves in ever-increasing contrast with everything we are learning about the natural world.

8 CONCLUDING REMARKS

The old fisherman’s mistake is to fail to understand that strict definitions based on a careful conceptual analysis of the way we think and the way we speak become misleading once knowledge increases.

In a sense, all this is trivial. After all, if we just define things carefully, everything goes in order. Sunset-(a) has a definition different from sunset-(b); free-will-(a) has a definition different from free-will-(b); and so on. The mistake is the blindness to this distinction. Philosophers are supposed to know better.

The subtle point is that a complex phenomenology (sunset, free will, flowing of time, consciousness, … ) is commonly named, then rigidly characterized by a concept. But the concept so generated does not refer solely to the phenomenology: it refers also to a certain understanding we have of this phenomenology, and this particular understanding can turn out to be wrong in the light of increased knowledge. The confusion comes because we confuse the part of the definition that refers to the phenomenology with the part that depends on our wrong beliefs.

There is no pure phenomenology independent from our general conceptual structure. The other way around, it seems to me that we have learned, at least since Quine, if not since the second Wittgenstein, that there is no concept that may not be challenged by empirical knowledge. The historical growth of knowledge has repeatedly shown us that concepts that seemed “clear and distinct,” and appropriate, can in fact turn out to be inadequate. The pre-relativistic concept of “simultaneity” or the pre-Copernican concept of “motion” turned out just to be inadequate.

The moral is that to understand is to recognize that we have to modify concepts, not only to clarify and combine them. Stein expresses this point beautifully : “This mistake is the assumption that a clarification of ‘ideas’, or concepts, should always or can always precede the advance of knowledge” (2004, 164), And: “[T]he two enterprises, that of knowledge and that of understanding, are inextricably intertwined” (135). Where “‘understanding’ here refer[s] to the grasp of ideas, or concepts” (135). The formulation of the useful concepts may come after knowledge, not before it, because “old” concepts may be internally plagued by wrong knowledge.

Once this is clear, we have always two options. The first option is to keep a name attached to the inadequate comprehension and say, for instance, that there is no sunset, there is no free will, there is no time, there is no table, there is no consciousness: these are “illusions.” This is correct, if we take these names to indicate their (a)-version.

But there is also the second alternative: there are sunsets, there is free will, there is time, there are tables, there is consciousness; they are just what their (b)-version implies.

What we decide to indicate by a world is conventional, and whether we take the first or the second option is a matter of convenience, not substance. (If I think that the shadows in my room are ghosts and then realize they are cast by branches outside the window, I may prefer to drop the ghost language altogether and call it illusory, rather than reinterpreting “ghosts” as “shadows of the branches.”) But to believe that there is a problem is the old fisherman’s mistake.

This implies that (taking the second option) concepts that we use can play the full spectrum of their role and still denote what they used to denote (sunsets, forests, pieces of matter, oriented time, values, moral imperatives, subjective experience, … ), even if we change our understanding of what they actually refer to: namely, if we redefine them taking into account increased knowledge. We got confused because we took them rigidly.

“But this is not what I mean by sunset!” answers the old fisherman to anybody trying to talk him out of this confusion by illustrating what sunsets really are. “But this is not what I mean by free will, by time, by morality, by consciousness,” answers the philosopher who resists what science is teaching us about the nature of time or the functioning of ourselves. And this answer is exactly right. The problem is that to remain anchored to “this is what I mean by” is to refuse to take away the screen that blinds us to the comprehension of the actual phenomena that we witness.

It is not the nature of the real sunset to have the sun splashing into the water, it is not in the nature of our actual freedom to be independent of the natural order, it is not in the nature of natural time to be necessarily oriented, it not in the nature of real continuous matter to be continuous at any scale, it is not in the nature of the actual moral behaviour of people or in the nature of actual subjective experience to be unaccountable in terms of known elementary laws.

Using this language, we can say that it is not that sunsets, free will, tables, consciousness, the oriented flow of time, and so on are “illusory.” They are perfectly real. What is an illusion is that a sunset is literally the sun diving into the waves, free will is literally an external intervention above physical evolution, and so on. Which is how we (mistakenly) might “intend” these phenomena.

Conceptual analysis in the sense of precise clarification of what we now intend by the concepts we use can become dangerously misleading as soon as we learn something new.

The reason is that knowledge is not just learning new facts about old concepts. It is often to realise the inadequacy and misleading character of some of our old concepts, and the intuitions that ground them.

I find it surprising that a single specific fallacy nourishes the confusion around so many issues.

ACKNOWLEDGMENTS

I am deeply indebted to Jenann Ismael for enlightening exchanges and comments, from which several of the ideas presented in this paper originated. This work was made possible through the support of the FQXi Grant FQXi-RFP-1818 and by the QISS grant ID# 60609 of the John Templeton Foundation.

REFERENCES

Citing Literature