My Objectivist phase was one of the more shameful periods of my intellectual development. I'm offering this by way of penance for that unfortunate year and a half. I feel compelled to atone somehow after hearing that an old friend of mine that I'd recommended Atlas Shrugged to long ago actually followed up on my suggestion - and has become a full blown Objectivist since I last saw him. This essay will examine Ayn Rand's metaethics and some fundamental aspects of her normative ethics. The clearest articulation of her views on these subjects I have at hand is Leonard Peikoff's Objectivism: The Philosophy of Ayn Rand, and it is on the account given there that I will base my criticism. I will then conclude by summarizing three other criticisms of Ayn Rand's metaethics which are to be found in the (limited) literature. I ask forgiveness in advance for my sometimes wordy and inelegant writing, I haven't edited any of this and it was sort of hastily thrown together in a day and a half. Suggestions for ways of simplifying or clarifying my points are welcome.
Peikoff/Rand begins OPAR's consideration of the good with the question, "...does man need to judge and select values at all? Is morality necessary or not, and if it is, why?" (207). For Rand the term, "value," is defined empirically on the basis of, "the fact of goal-directed action," as, "that which one acts to obtain or keep" (208). Peikoff goes on to claim that this understanding entails that values are always relative to agents and to specific goals.
Already we are confronted with controversial and unquestioned premises. To begin with, why we should look to goal directed action to understand the nature of value is never explained. She seems to just pluck an area of inquiry out of the air, pick a term that has some intuitive relation to it and claim that she has somehow arrived at an empirical understanding of value. Rand says that she derives her definition of value from observation of goal directed action, but doesn't expand on what she means by such action. Action is understood broadly in her work as a feature of causation, defined earlier in OPAR as the fact that, "an entity must act in accordance with its nature" (14). What exactly constitutes an entity is extremely unclear, and Rand's ontology is left extraordinarily vague both by her and her posthumous devotees. At the very least, most of the things in our usual day to day perception are entities on her view, so people, plants, pebbles, hydrogen gas and light bulbs are all entities and therefore capable of action. So what is meant by goal-directed? "Action toward an object," (209) is the closest we get, though this isn't terribly illuminating. Without a rigorous account of what it means for action to be goal directed, Rand threatens to let in a lot of things that are fatal to her analysis. Is water flowing downhill goal-directed action? It's certainly action under her view of causation, and it seems oriented towards attaining some equilibrium state. Is the melting of ice at room temperature goal directed action? Is the process of biological evolution? The accretion of galaxies? While Peikoff claims that, "one does not observe desks or pebbles pursuing goals," the simple orientation of action towards some object or state of affairs doesn't provide the isolation of the qualitative features from which Rand attempts to generalize a claim about the nature of value.
Peikoff tries to rein this in by limiting the scope of what behaviors can be considered goal-directed, though still not providing necessary or sufficient conditions for it. "Goal directed behavior is possible only because an entity's action, its pursuit of a certain end, can make a difference to the outcome. 'Alternative' does not imply choice; it means that the entity is confronted by two possible results: either it acts successfully, gaining the object it seeks, or it does not (and thus fails to gain the object)" (208). First, this claim by Peikoff is completely unsubstantiated. Second, it doesn't mesh with our intuitions on goal directed actions. Consider the case of Christians who pray for a cancer patient's health. Their actions appear goal directed, even though they don't affect whether the patient will be healed of their terminal illness or not. Alternately, consider a female applicant for a job who sends in her resume. That action seems goal directed even if one stipulates that she has ignorantly applied to a firm run by a misogynist who refuses to hire women. Third, this verges on incoherence if we continue to take seriously Rand's account of causation. What an entity does just is a result of its nature. That means that whether or not an entity achieves a given goal just is a result of its nature. This would imply that all supposed alternatives are merely apparent. Since my achievement of some goal is solely a result of my nature, it cannot be the case that I could have the same nature and fail to achieve the goal, ceteris paribus. The supposed alternative can only come to pass if things are not as they are, but as Rand insists on noting on seemingly every page, OMG CONTRADICTIONS CANNOT EXIST. Rand's view of causation is incompatible with the existence of any genuinely goal directed action on Peikoff's account, and thus incompatible with her analysis of value.
Having defined values, Rand/Peikoff's next step is to consider who possesses them. She says that the very concept, "presupposes an entity capable of generating action towards an object," (209). This may be true given her definition, but it's far from some sort of discovery about the nature of value in any interesting sense. It seems at best an artifact of her choice to limit the question of value to the actions of entities in persuit of goals. That value should then be agent and goal relative given her arbitrary, unmotivated starting point should come as no surprise. The sorts of entities she thinks have values are living organisms, as they are, "capable of self-generated, goal directed action - because they are the conditional entities, which face the alternative of life or death" (209). The self-generated bit is snuck in, and only mentioned this once. It doesn't appear to be a part of her definition of goal directed action and isn't connected by anything else in this section to her understanding of value. It also makes little sense given the account of causality already discussed, as it isn't clear in what sense an action taken by an entity as a result of its nature could be anything but self-generated, that is, a result of its nature. Also, recall that as noted before lots of things other than living organisms - ice, stars, heat - seem capable of acting (in Rand's sense of action) in ways that are directed at certain outcomes. She can't escape her problems by arbitrarily limiting goal directed action to life, as that would simply make her definition of value question begging.
Peikoff gives one reason to think that only living organisms can undertake goal directed action: they are the only entities which face the alternatives of life and death, which Peikoff understands as synonymous with existence and non-existence in this context. "The alternative of existance or non-existence is the precondition of all values" (209). Notably, it is left unspoken how living organisms face the possibility of non-existence in a sense which planets and pebbles do not. Peikoff proceeds to justify the claim that existance is the only end it itself by offering Rand's example of an immortal robot. We are told that such a robot couldn't possibly have any needs, desires or goals. The idea that removing the alternatives of life and death removes the possibility of need satisfaction or need frustration is motivated here almost solely by an appeal to the unstated intuition that all of our physical and psychological needs reduce to our need to survive. But consider a slight alteration of the scenario, where the aforementioned robot is ignorant of his own immortality. He thinks he is destructible and builds up some set of values on the Randian account. His creators then tell him that he is immortal. It hardly seems right to think that our robot will suddenly freeze like some old fashioned sci-fi robot laid low by some circular request or a demand to compute pi, but this is precisely Rand's prediction. To bring the example closer to home, imagine your answer to the question, "what would you do if you were immortal?" If your answer is anything other than, "lie inert," you can't possibly find Rand's attempt to draw out your intuitions convincing. Having certain pleasurable or painful mental states and acting to experience the one and not experience the other lacks the sort of necessary connection to survival that she postulates.
We can consider some aspects of this empirically. Rand says through John Galt in Atlas Shrugged that, "An animal is equipped for sustaining its life; its senses provide it with an automatic knowledge of what is good for it or evil. It has no power to extend its knowledge or to evade it. In conditions where its knowledge proves inadequate, it dies. But so long as it lives, it acts on its knowledge, with automatic safety and no power of choice, it is unable to ignore its own good, unable to decide to choose the evil and act as its own destroyer." (as quoted in Rand, 1960) If living organisms are really the only things that can undertake goal directed actions, if existence is really the fundamental value, and animals are naturally directed at that value, then this should be true. Fatally for Rand, it's not. Examples abound in nature of animals that act in ways contrary to their own survival, putting higher value on other things - notably copulation. The male praying mantis, for instance, is well known for running the risk of getting decapitated by its mate during sex - which doesn't exactly discourage them from getting it on anyways. Dolphins have been known to commit suicide in captivity by battering themselves to death on the walls of their tank. Common octopuses often refuse to eat after laying their eggs, to leave their body as a food source for the young. No, the choice of life over death as Rand lays it out is an arbitrary starting point for value. I can clearly hold other values without grounding them out in survival. Martyrs and suicides act on the basis of some code of values that doesn't ground out in survival as the sole end in itself. The very fact that the question, "Would it be better to die or live a life of constant suffering," makes sense to us means that value cannot reduce to the imperative, "Live!".
It's important to observe very clearly that Rand is not arguing that life is necessary to attaining all other hypothetical values, so instrumentally we have to preserve it prior to achieving anything else. Peikoff observes, "The Objectivist viewpoint... is not that life is a precondition of other values - not that one must remain alive in order to act. This idea is a truism, not a philosophy" (213).
To summarize, Objectivism attempts to ground out value in generalization from goal directed action. Its inadequate account of goal directed action, however, forces us to use the word "value" in extraordinarily non-standard, counterintuitive ways and contradicts the Objectivist account of causation. Rand tries to limit goal direction to organisms, but gives no compelling reasons and in facts contradicts our intuitions that actions (in the Randian sense) can be directed at outcomes (Peikoff's very definition of goal-directed) without being originated by living things. Lastly, her attempt to use the alternative between life and death to pick out life as an end in itself gets little support save from her defused intuition pump and fails by the empirical measures she commits herself to.
Stupidly, of course, I only thought to check the critical literature on Rand once I was basically finished with this. It was mildly disappointing, therefore, to find that several of my objections have already been considered in the rather limited critical literature on Rand. What follows is a survey of three other critiques of the Objectivist metaethics.
King (1984) raises several important objections and anticipates at least one of mine. First, he grants that in any system of values there must be at least one end in itself, an argument made by Rand. He observes, however, that this does not entail that there must be a single ultimate value, a single end in itself which undergirds the entire moral system. It's entirely reasonable to suppose that life, truth and beauty could all be ends in themselves. Next, King proposes an alteration of the immortal robot scenario much like that I put forward above and uses it to demonstrate that it is the capacity for desire, not the alternative between life and death, that makes goal directed action possible. In our experience, certainly, the activity of desiring presupposes life, but we have no reason to think that this is neccessarily the case. To most of us, life is simply very high on our ordering of preferences and desires. And true, in our experience valuing most things requires valuing life instrumentally to acquire it, but Peikoff explicitly denies that this is what Rand's view consists in.
Huemer leverages an attack on the Objectivist rejection of the analytic/synthetic division and of a priori truths to show that the Objectivist project of inferring the meaning of a term (in this case "good" or value") from observations is misguided. He compares it to attempting to define the term "blue" by observing the color of the sky. Clearly, looking up tells you the color of the sky, it doesn't reveal any information about "blueness". One can arbitrarily define "blue" as "the color of the sky", but one can define blue in any way one wishes if that's acceptable. This impossible prospect is the only direction left to the objectivist who has rejected sources of knowledge justification other than sense perception.
Robbins' Answer to Ayn Rand is an overlooked response by a devout Christian to Ayn Rand's entire corpus. Google Scholar has no record of it except as a citation in two bibliographies on Objectivism. I'm inclined to think this is more on account of Objectivism's lack of scholarly legitimacy that that of Robbins. He undertakes a painstaking reconstruction of Rand's entire account of ethics and a simultaneous rebuttal of it, but two things are of particular interest in this present context. Robbins argues that Rand's account of value begs the question, insofar as she assumes her conclusion about the nature of value when she simply asserts that the concept of value is by its very nature agent and goal relative. Most interestingly, Robbins observes that John Galt actually threatens to commit suicide at one point in Atlas Shrugged. Galt says that if Dagny is captured and tortured by the socialists, there would be no values left for him, and he doesn't care to exist without values. Of course, this runs entirely contrary to the metaethical claim by Rand that the alternative of existence and non existance is the basis of all value. Presumably Galt would have at least one value left - his own continued survival.
Huemer, Michael. "Why I am not an Objectivist". http://home.sprynet.com/~owl1/rand.h
King, J. Charles. "Life and the Theory of Value: The Randian Argument Reconsidered". Published in The Philosophical Thought of Ayn Rand. Eds Den Uyl and Rassmussen. University of Illionois Press, Chicago: 1984.
Peikoff, Leonard. Objectivism: The Philosophy of Ayn Rand. Penguin Books, New York: 1991.
Rand, Ayn. "Faith and Force: The Destroyers of the Modern World". 1960. http://freedomkeys.com/faithandforce.ht
Robbins, John W. Answer to Ayn Rand. Mount Vernon Publishing Co, Washington DC: 1974.