2023年更新：此文在2023年被会议GPPC接收。接受后发现原文多有疏漏之处，于是会议上基本是抛弃了原文重写了一份讲稿。原文附在最后，以下是会议的讲义。
Is Tesla Self-driving Car Moral Patient? A Reply to Goodpaster
Author: Santure Chen, Franklin & Marshall College. Email: hchen2@fandm.edu
In his argument for the moral consideration of plants, Kenneth E. Goodpaster (1987) asserts that plants possess interests, as evidenced by their tendency to maintain and heal themselves, and further argues that only beings with interests deserve moral consideration. I neither agree that plants have interest nor being with interest desire moral consideration. Therefore, in my freshman year, I use Tesla self-driving car as a counterexample to refute his argument. After this paper being accepted by this conference, I substantively revisited this question and developed several deeper arguments.
Initial Response: An obvious counter-example
Goodpaster’s defense for plant’s interest is ambiguous for me. He said:
In the face of their obvious tendencies to maintain and heal themselves, it is very difficult to reject the idea of interests on the part of trees (and plants generally) in remaining alive.
I understand his claim as
(1) Beings have an interest in X if and only if the being will take action to achieve X without external interference.
Hence, an obvious fact is
(2) Tesla self-driving cars will take action to achieve normal driving without external interference.
We may intermediately conclude
(3) Tesla self-driving cars have interest.
And given
(4) (Interest Principle) Beings who have (or can have) interests can deserve moral consideration.
We could conclude that
(5) Tesla self-driving cars deserve moral consideration.
Apparently, (5) is false. Thus either (1), (2), or (4) is false. (2) is an undeniable fact, so either (1) or (4) is false. Given falsifying either (1) or (4) will falsify the moral status of plants, plants must not deserve moral consideration.
Freshman-leveled response: a detailed look at “interest”
A weak point of the previous argument is Goodpaster did not directly claim (1). As I said before, his defense of plants’ interests is ambiguous for me in freshman year. May he define interest in a special way, that lets plants have interest but let self-driving car not have?
What does interest truly means, and where did the author extract the concept of interest?
Goodpaster said:
…(Plants) are not “mere things”; they are vital objects with inherited biological propensities determining their natural growth. Moreover, we do say that certain conditions are “good” or “bad” for plants, thereby suggesting that plants, unlike rocks, are capable of having a “good.”
We do say certain states of plants are good, like healthy, fast-growing, or fertile. But (1) part of these goods are judged according to human need, as if we say a car is properly functioning, and (2) I suspect other goods are analogized from other things that really have a good, probably under the influence of fairy tale stories.
Goodpaster also quoted this uncited paragraph in his defense for interest:
If we view pleasure as rooted in our sensory physiology, it is not difficult to see that our neurophysiological equipment must have evolved via variation and selective retention in such a way as to record a positive signal to adaptationally satisfactory conditions and a negative signal to adaptationally unsatisfactory conditions . . . The pleasure signal is only an evolutionarily derived indicator, not the goal itself.
In short, it argues that (a) interest is observed in animals, and (b) animals are a refined version of plants, so (c) plants have interests.
Goodpaster takes a biological reductionism here. So, why don’t we go a little bit further? We may argue that plants are merely physical systems fully consisting of atoms with sophisticated structures. Their tendency of self-healing, absorbing water, etc., are not mystery inherited biological propensities but mere results of applying physical and chemical laws to those atoms, which is as same as Tesla Self-driving cars. Plants clearly do not have Descartes’s “thinking substance” or any other special material that allows them to escape from physical laws. (However, it may have some special structure that allows them to become a non-physical being, which will be discussed later.)
Sophomore-leveled response: the metaphysics behind ethics
In this part, we will examine claim (4), the interest principle in a more broad philosophical background.
Goodpaster inherits his principle from Joel Feinberg (1974), though Feinberg denies that plants have interests. Feinberg argues that
I have come to this tentative conclusion for two reasons: ( I ) because a right holder must be capable of being represented and it is impossible to represent a being that has no interests, and ( 2 ) because a right holder must be capable of being a beneficiary in his own person, and a being without interests is a being that is incapable of being harmed or benefitted, having no good or “sake” of its own.
It should be noted that these two reasons only give the necessity of having interest but do not provide sufficiency.
All ethical theories do not get their principle from the void; they are all driven by some metaphysical theories. The range of moral patient are also governed by metaphysical theories. For example, you may see that philosophers who support animal rights are mostly utilitarian (e.g., Peter Singer) because utilitarianism’s metaphysics is most compatible with the animal rights’ claim.
| Normative ethics | The Metaphysics behind it | Implied Moral Patient |
|---|---|---|
| Deontology We ought to follow the prior rule set by reason, disregards consequences and inclination. | Kant’s Categorical imperative Act only if you wish everyone to do the same & people are ends but not means. | All things that are rational being. (Immanuel Kant) |
| Utilitarianism We ought maximize net happiness disregard whom and when it happens. (J.S. Mill) | Hume’s bundle theory Personal identity is an illusion, person is merely locus of experience. (Marya Schechtman) | All things that can have experience. (Peter Singer) |
| Moral Ethics We ought to cultivate proper emotional responses and act upon them. (Aristotle) | Tripartite Soul Harmony among parts of soul is sufficient and necessary for good judgement. (Plato) | All things that provokes our sympathy or will to care. (Martha Nussbaum) |
I will give a little bit more detailed illustration of utilitarianism here, because Goodpaster assumes utilitarianism as the default ethical theory. Utilitarianism believes that (1) happy experience is the thing that ultimately matters. (2) The value of happy experience is irrelevant to when, where, and whom it happens at. According to the bundle theory of personal identity, there is no “person” but a bundle of experiences. So experience is what ultimately exists (you may also consider Hume’s skepticism) and, therefore, ultimately valuable. Besides, given personal identity is an illusion, we shall not consider my experience more valuable than others’ experiences. So the value of happy experience is irrelevant to when, where, and whom it happens at. Therefore the basic principle of utilitarianism is justified.
Then, what metaphysics stands behind Goodpaster’s claim?
If we view pleasure as rooted in our sensory physiology, it is not difficult to see that our neurophysiological equipment must have evolved via variation and selective retention in such a way as to record a positive signal to adaptationally satisfactory conditions and a negative signal to adaptationally unsatisfactory conditions . . . The pleasure signal is only an evolutionarily derived indicator, not the goal itself.
In short, (a) (as a hidden premise) pleasure matters, and (b) pleasure is a means to interest, so (c) interest matters.
We do see many people argue that (a) pleasure matters. But people argue that pleasure matters because they believe in Hume’s metaphysics, while (b) is true only if we accept science realism. So one cannot coherently acknowledge both (a) and (b).
Junior-leveled response: intentionality beyond Goodpaster
In this section, I will revisit the concept of “interest” again. Goodpaster did not give too much illustration for plants’ interest, so this section will review discussion about intentionality in other philosophical literatures.
Here are several dominating schools of the mind-body problem:
| Theory | Nature of intentionality | Status of Plants |
|---|---|---|
| Dualism (Decartes) | A kind of mystery non-physical Material. | Open, but plants does not seem to have such material. |
| Functionalism (David M. Armstrong) | A device that has proper casual effect disregarding physical implementation. | Debatable, people may or may not attribute plants’ behavior to intentionality. |
| Embodied Cognition (Joh Haugeland) | A combination of a relatively simple brain and a complex world, connected through rich sensory input and output. | Mostly no. Plants do not have rich-enough sensory organs. |
This table is listed here merely for reference. I do not feel it’s necessary to explain this table in detail because (a) the theories listed above should sound familiar and (b) they are not very useful for today’s particular question. I feel the intentional system theory by Daniel C. Dennett (1983) is the best theory to explain my standpoint and Goodpaster’s mistake. Dennett argues that intentionality is what we bestow on things through our cognition but is still real.
There are multiple types of stances in our cognition; each type is a way to abstract. We say something is intentional if and only if psychological language – what do this person believe, what do this person desire, etc. – could best abstract the property of the thing for our purpose. All chemical laws are driven by physical laws, and all chemical reaction can be explained by physicists. However, we still say something is chemical if a chemical explanation of its behavior is shorter and easier to understand. Psychological language, design language, and physical language were not invented by philosophers who sat in the darkroom; they were invented in thousands of years of life to best describe different things, yet rhetoric like analogy or personification may blur the boundary occasionally.
But it does not mean they are subjective; all these descriptions are also real, because they represent those new patterns that emerge when things are looked at different scales. Conway’s life game is a very good example to illustrate this point, but I am not sure if I have time to present it. See Appendix II for a short introduction to it.
| Type of stance | Definition | Example |
|---|---|---|
| Physical stance | Things could be best explained by its physical property. | The chalk fails down because it is 1 oz weight. I am blooding because a knife cut my finger. |
| Design stance | Things could be best explained by its purpose. Physical details are abstracted away. | The valve could adjust volume of water, disregarding it’s made by plastic or copper. I need eat food to gain energy before hiking. |
| Intentional stance | Things could be best explained by its intention (beliefs, desires, and etc.). Physical and design details are abstracted away. | He is crying because he is sad. She might be the murderer because she had conflict with the victim. |
To my understanding, both self-driving car and plant are design stances. Without doubt, the car is designed by people for a clear-defined ask. But someone may worry that plants do not have a creator, if you don’t believe in religion. However, design stance is not about being designed; it is about its function or purpose. Clearly artificial things have a purpose, but natural things like plants, also have purpose as Goodpaster acknowledged – their organs are evolved to best sustain its life and reproduce.
Acknowledgement
I want to say happy birthday to Professor Nimershiem, a brilliant math professor who is my advisor. Her birthday is yesterday.
Appendix I: possible comments
- Tesla Self-driving cars’ “interest” is made by human, so it does not count.
First, in 2010, scientist has made the first fully artificial cell. It might not be far to create a totally artificial plant that is indistinguishable from natural plants. If artificial plants are indistinguishable from natural plants in anyway other than tracing its source, I do not see reason to let them have a different moral status. But more importantly, Goodpaster did not argue that artificial interest is not interest; he merely argues that interest matters. I am curious what reason could you give for defending that only natural interests deserve moral consideration.
Moreover, people are influenced by society in many ways, our interest are deeply shaped by our education and advertisement. If artificial interest does not count, then most of things you and me care today do not deserve moral consideration. On the other hand, almost the whole earth has been influenced by human to some degree. For example, after the industrial revolution, the pale moth turned to have the tendency of being black in many parts of Britain, allowing it to blend in against soot-covered trees and avoid predators. (Tozer, 2009)
- Intentional stance will be useless if we have a super-smart physicist.
I may answer this doubt in two ways. First, we do not have a super-smart physicist; to be limited is a nature of our cognition. You may recall Kant’s justification for casual law – it is universally necessary for being like us.
Second, as I show in appendix II, those patterns are real. There exists pattern beyond physical level. Saying intentional language is useless is a failure to capture those real things.
Appendix II: Conway’s game of life
The Conway’s game of life happens on an infinite two-dimensional orthogonal grid. Each square is a cell, which is either live or dead. Every cell has 8 neighbors, which are the cells that are horizontally, vertically, or diagonally adjacent. At each step in time, the following transitions occur:
- Any live cell with fewer than two live neighbors dies, .
- Any live cell with two or three live neighbors lives on to the next generation.
- Any live cell with more than three live neighbors dies, as if by overpopulation.
- Any dead cell with exactly three live neighbors becomes a live cell.
This is all rules. We could use cell-leveled language to describe everything happens in this game. Now open your browser and go to https://playgameoflife.com/. There are 5 live cells at the middle of the page. Then click start. What happens?
“It move!” you must say. But what is “it”?
Now go to https://playgameoflife.com/lexicon/Gosper_glider_gun, increase speed dial and zoom out using scale on the right bottom corner, and click start. What happens?
“It ejects babies!” you may say. But what is “it”?
New rules, that seems to be totally irrelevant to the fundamental rules, emerge in different scales. There even are turning machines built by thousands of baby producer.
Appendix III: Reference
David M. Armstrong, “The Causal Theory of Mind,” in The Nature of Mind and Other Essays (Ithaca, NY: Cornell University Press, 1980), 16–31.
Daniel C. Dennett, “True Believers: The Intentional Strategy and Why It Works,” in The Intentional Stance(Cambridge, MA: Bradford/MIT Press, 1987), 13–35
Joel Feinberg, “The Rights of Animals and Unborn Generations,” in Philosophy Et Environmental Crisis, by William Blackstone, Univ. of Georgia Press, 1974.
Kenneth E. Goodpaster, “On Being Morally Considerable.” in The Journal of Philosophy, vol. 75, no. 6, 1978, p. 308., doi:10.2307/2025709.
John Haugeland, “Mind Embodied and Embedded,” in Having Thought, 219–37.
Martha C. Nussbaum, “The Discernment of Perception: An Aristotelian Conception of Private and Public Rationality,” in Love’s Knowledge: Essays on Philosophy and Literature (Oxford: Oxford University Press, 1990), 54–84
Marya Schechtman, “The Same and the Same: Two Views of Psychological Continuity,” American Philosophical Quarterly 31, no. 3 (1994): 199–21
Appendix IV: Original Paper
（略）
2021年原文
突然意识到很久没发东西了——准确说，是很久没有写有趣的东西了。以下这篇文章是前两天写的一篇期末论文，私以为写的还算不错，于是就放上来充个数。
这篇文章试图反驳“植物是道德容受者”这样一个观点。道德容受者（Moral Patient），指的是其喜怒哀乐兴衰荣辱在道德上有价值的东西。例如因为路边的石头不是道德容受者，所以无论对石头做出多过分的事情，都不能说是不道德的。
最开始，康德认为只有会思考、拥有理性的人是道德容受者。接着，Peter Singer认为一切有感知能力的东西都是道德容受者。例如很多动物，他们虽然没有理性，但是能感觉到痛，有开心或者不开心。现在，Kenneth E. Goodpaster 宣称，一切有有“interest”（或译为“兴趣”“关切”）的东西都是道德容受者。而植物，表现出了对阳光和水的关切，因此是道德容受者——换句话说，残忍地对待植物是不道德的。这种观点被称作biocentrists，本文试图反驳这一观点。
本文从“自动驾驶汽车也有维持驾驶的关切，你难道要说自动驾驶汽车也是道德容受者吗”的反问入手，试图厘清“关切”和“趋势”的概念。植物有趋光趋水的特性，但是“关切”是形容一种思维态度的。无思维的物体可以有趋势，但不会有关切。
以下为原文：
Claiming that “only beings who have (or can have) interests can deserve moral consideration”, some philosophers, such as Kenneth E. Goodpaster, argues that not only rational being and animals but also plants, which has “the interests” of sunshine and water, are moral patient. (Goodpaster, 1972, p318) Tesla Model S is a car that can drive itself automatically. In this essay, I will argue that biocentrists confuse physical tendencies with subjective interests. Therefore, by their argument, we could prove Tesla Model S has “the interests” of driving appropriately and, hence, Tesla Model S is a moral patient.
Tesla Model S (or any other auto-drive car) is equipped with radars, sonars, and cameras. A computer will analyze these sensors’ data, then send proper instructions to the engine, the brake, and the steering system. The computer is programmed to drive normally – or, to be specific, safely, legally, and efficiently. (The Tesla Team, 2017) Similarly, according to biological study, plants are an assembly of cells, and cells are an assembly of molecules. Molecules are arranged as a particular structure in order to maintain the cell’s function, such as photosynthesis (the process of converting light energy to chemical energy) or absorption of water and nutrition. These cells are arranged as a particular structure in order to maintain the existence (life) and duplication (reproduction) of the plant.
In order to identify Tesla Model S has interests or not, we must clarify the definition of the word “interest.” The Merriam-Webster dictionary defines interest as “a feeling that accompanies or causes special attention to something or someone.” Obviously, only conscious beings can have feeling and attention. Since plants do not have consciousness, if it is claimed that plants have interests, philosophers must expand the definition of interest. “In the face of their obvious tendencies to maintain and heal themselves,” Goodpaster wrote, “it is very difficult to reject the idea of interests on the part of trees (and plants generally) in remaining alive.” (Goodpaster, 1972, p319) What does “tendencies to maintain and heal themselves” indicate? Straightforwardly, it means that trees will take action maintain and heal themselves without external interference. In a biological view, it means that cells of the tree will act (or react) in a certain way that can maintain the tree (or heal the tree). Therefore, the quotation implies the following claim:
(1) Beings have an interest in X if and only if the being will take action to achieve X without external interference.
Hence, if:
(2) Tesla Model X will take action to achieve normal driving without external interference.
And if:
(3) Beings who have (or can have) interests can deserve moral consideration.
We could conclude:
(4) Tesla Model X deserve moral consideration.
This argument is logically valid. Therefore, the conclusion must be true if all premises are true. However, the conclusion must be wrong[1]. Thus, at least one premise must be false. (2) is a fact; it cannot be wrong. (1) is implied by Goodpaster’s justification of plants’ interests, and (3) is Goodpaster’s claim. Therefore, either Goodpaster fails to prove plants have interests, or Goodpaster’s claim is false. Both leads to the falsification of the moral standing of plants.[2]
Someone may argue that plants are different from Tesla Model S because plants are organisms. Joel Feinberg wrote:…(Plants) are not “mere things”; they are vital objects with inherited biological propensities determining their natural growth. Moreover we do say that certain conditions are “good” or “bad” for plants, thereby suggesting that plants, unlike rocks, are capable of having a “good.” (Feinberg, 1974, p51)
Feinberg mentions two reasons here: plants are different from rocks, or Tesla Model S, because (1) plants have “inherited biological propensities” and (2) plants are “capable of having a good”. I can reply this objection by two ways.
First, A is different from B in some aspects does not necessarily mean A and B cannot belong to the same category. Women are different from men, but both women and men are moral patients. The objection above implies another principle that only beings who have “inherited biological propensities” and/or “capability of having a good” deserve moral consideration, which conflicts with the “interest principle” that only beings who have interest deserve moral consideration. I am interested in seeing how biocentrist will justify the new principle, though I am afraid that they can never finish the justification because of my second point.
Second, inherited biological propensity or capability of having a good is not something mysterious. Inherited biological propensities are just an application of basic physical and chemical rules. Obviously, plants do not have consciousness, which expels them from the umbrella of the mind-body problem. The biological study has clearly shown the mechanism of how organisms maintain their life. Simple molecules, such as DNA, are formed in an ingenious way, which allows them to have particular physical or chemical interactions with molecules nearby. Hence, energy and substance are induced to move, transfer, and assemble in a certain way, which results in the maintenance of a macroscopic object which is named as “tree”. Biological propensity is an engineering wonder, but nothing more. I acknowledge that heredity is an organism-exclusive feature now, but the self-replicating machine is also in developing. There is no fundamental obstacle for an artificial craft to behave like an organism. Besides, the capability of having a good is a result of popular culture. Feinberg’s justification of plants’ capability of having a good is that “we do say that certain conditions are ‘good’ or ‘bad’ for plants, unlike rocks.” However, we also say that certain conditions are good or bad for Tesla Model S – if it drives normally, it is good; if it crashes, it is bad. Plants have never expressed their desire of alive in any way other than the tendency. If following the tendency is good and vice versa, cars can also have good and bad. We cannot provide a rational reason for why we think being alive and healthy is good for plants. I guess it might be analogical – personalization of animals and plants is very frequent in fairy tales. It is to say that, since health is good for us (the human), we may use the same scale to describe plants. If this hypothesis is true, biocentrism is not different from human-centrism since it uses the similarity between plants and human to justify the moral standing of plants.
Some biocentrists may also try to refute my argument by claiming that Tesla Model S’s tendency is given by humans, but the plant’s tendency is natural. First, as I mentioned above, the source of the tendency is not counted in the original version of the interest principle. Additional work must be done to justify this modification. Second, genetically modified crops, which usually is still classified as plants, have some artificial interests. For example, the Golden Rice has the tendency to produce beta-carotene, a precursor of vitamin A, in its edible part. (Ye, 2000) Another example, if you deny that genetically modified crops are plants and deserve moral consideration, is the moth. After the industrial revolution, the pale moth turned to have the tendency of being black in many parts of Britain, allowing it to blend in against soot-covered trees and avoid predators. (Tozer, 2009) According to Darwin’s theory of evolution, organisms are changing to survive in this changing world. As an influential part of this world, intentional or not, directly or indirectly, human has changed many organisms’ tendencies. It is really hard to find an organism which is not influenced by human but lives on earth.
Although there might be more objections, above all, I hold that the Tesla Model S and a plant do not have a morally significant difference. They are both specific structures of molecules, which utilizes physical and chemical rules to create some macroscopic tendencies. Since we cannot capture any evidence other than tendencies to prove plants have interests, if it is claimed that plants have interests, it must imply tendency is equivalent to interests. Therefore, since Tesla Model S is not included in the moral circle, we should not consider plants as moral patients.
I believe biocentrism makes the mistake that it confuses physical tendencies and subjective interests. Any substance has its tendency according to physical laws. It implies that even breathing may interrupt the air’s tendency of staying at its original place, according to the law of inertia. We should realize that any action we take is an interruption of a natural tendency. Interests, however, is produced by consciousness. We literally cannot live if we have to respect those tendencies.
Reference:
Feinberg, Joel. “The Rights of Animals and Unborn Generations.” Philosophy Et Environmental Crisis, by William Blackstone, Univ. of Georgia Press, 1974.
Goodpaster, Kenneth E. “On Being Morally Considerable.” The Journal of Philosophy, vol. 75, no. 6, 1978, p. 308., doi:10.2307/2025709.
The Tesla Team. “Your Autopilot Has Arrived.” Tesla, 30 Nov. 2017, www.tesla.com/blog/your-autopilot-has-arrived.
Tozer, James. “Darwin’s ‘Evolution’ Moth Changes Back from Black to White Thanks to Soot-Free Skies.” Daily Mail Online, Daily Mail, 20 June 2009, www.dailymail.co.uk/sciencetech/article-1194281/Darwins-evolution-moth-changes-black-white-thanks-soot-free-skies.html.
Ye, X. “Engineering the Provitamin A (-Carotene) Biosynthetic Pathway into (Carotenoid-Free) Rice Endosperm.” Science, vol. 287, no. 5451, 2000, pp. 303–305., doi:10.1126/science.287.5451.303.
[1] It should be noticed that Tesla Model X is not a special case. Roly-poly toy, a traditional east-asian toy invented thousands of years ago, tends to right itself when pushed at an angle, and does this in seeming contradiction to how it should fall. Even water has the tendency of flowing down. The strategy that claiming Tesla Model S is artificial intelligence (therefore deserve moral considerability) cannot work.
[2] This argument can also be used to attack Peter Singer’s sentience principle, since Tesla Model S has cameras, it also has sentience. However, sentience principle can run to mind-body problem’s protection by claiming that qualia (the non-physical and organism-only part of consciousness) is necessary for being sentience.