In the problem of free will, the higher order theory exists as an attempt to defend compatibilism. Even if all of our desires are derived causally from the laws of nature, some philosophers argue that we still have free will if we have the ability to form desires and the agency to act on these desires. Wanting to do a particular thing is a first order desire. Wanting to want or not want this thing is a second order desire (Frankfurt, 1971). As the chain of desires grows, all desires other than the first are collectively known as higher order desires (Frankfurt, 1971). The higher order theory states that we have freedom of will if we act on our first order desires, which are supported by our higher order desires. However, I believe a contemporary discussion of free will encompasses not only the freedom to will a certain action, but also the freedom to act on said will (Doyle, n.d.). Since the higher order theory does not always guarantee freedom of action, it is clear that it cannot be a solution to the free will problem. Furthermore, merely a congruency between first order and second order desires is not enough to make soft determinism plausible. There must be an honest cause and effect relationship between belief, desires, and action which reflects our true identity.
The reasons for the higher order theory can be seen in cases of mental illnesses or cognitive disabilities. Such cases explain why acting on first order desires alone is not enough to constitute free will. The first order desires of someone with obsessive compulsive disorder (OCD) come in the form of obsessions and compulsions. While the person acts upon these compulsions, they do not actually have control over their actions (Glannon, 2012b; Meynen, 2012). This is why it is important to consider higher order desires. The second order desire of a person with OCD would most often be that they do not want these compulsions, thereby contradicting their first order desire (Glannon, 2012b). Similarly, a kleptomaniac’s first order desire would be to steal. However, their second order desire would be that they do not want to want to steal. Again, this contradicts their first order desire. Therefore, these cases illustrate why it is important for one’s first order desires to be supported by their higher order desires in order for free will to be possible.
Not only is it necessary for one’s first order desires to be supported by higher orders, there must also be a genuine causal relationship between first order desires and action. In order to establish genuinity of such a relationship, we must distinguish between a desire and an effective desire. An effective desire is one that actually causes an effect in action (Frankfurt, 1971). If a person merely acts in accordance with their desire, and yet this desire in itself was not the main motivator for such an action, then it is not an effective desire (Frankfurt, 1971).
Consider, as an example, that I desire to donate to charity for two reasons. The first reason being that I want to seem cool, and the second reason being that I want to help the less fortunate. My second order desire is that I want to want to help the less fortunate. If it turns out that the desire which actually motivates me to donate is that I want to help the less fortunate, then that desire is the effective one. Since it is also supported by my second order desire, I have free will.
Alternatively, if the real motivation for my donation is that I want to seem cool, then this becomes the effective desire. Since my effective desire contradicts my higher order desire, I do not have free will. Although the desire to help the less fortunate still exists in this second example, I do not have free will since it is not my effective desire.
The effectiveness of higher order desires must also be considered. Frankfurt (1971) puts forth the example of a physician who practices psychotherapy with drug addicts. While he does not want to actually do drugs, he seeks to accurately comprehend the feeling of drug addiction. Therefore, his second order desire is that he wants to want to do drugs. In this case, the physician clearly acts on his first order desire, that he does not want to do drugs. Even though this contradicts his second order desire, instinct would tell us that he does have free will. This is because although the physician wants to desire drugs, he does not want this desire to be effective (Frankfurt, 1971). This example adds a new level of complication to the higher order theory. This is a peculiar case in which the conditions for free will, as outlined by the higher order theory, are not met, and yet the subject still has free will. Perhaps it is not enough to confirm that first order desires are supported by higher order desires. We must consider whether or not desires of higher orders include effectiveness of lower order desires.
It is important to note that even if there is a causal relationship between effective desires and action, which is motivated by desires of higher orders, this may still prohibit freedom of action (Doyle, n.d.). Consider the possibility that a kleptomaniac’s second order desire is that they want to want to steal. In other words, they are okay with being a kleptomaniac and their first order desire is in fact supported by their higher order desires. While the higher order theory would suggest that this person does in fact have freedom of will, I believe that freedom of action is also necessary in order to constitute free will. In this regard, it is clear that the kleptomaniac does not have free will as their actions are still being controlled by their mental illness (Glannon, 2012b; Meynen, 2012). They have simply surrendered to their mental illness. In this example, the fault of the higher order theory is evident; it does not account for cases in which the causal relationship between desires and action hinders freedom of action (Doyle, n.d.).
Another fault of the higher order theory is that it does not account for the possibility that one’s higher order desires may be controlled by an exterior being or entity (Glannon, 2012a; Kane, 2000b). David Kyle Johnson (2016) argues that even if the origin of one’s desires, “is completely out of your control, it doesn’t matter” (p. 15). He goes on to say that free will still exists even if our desires are programmed by our genes or environment. Upon initial reading of this argument, I would have to agree with Johnson. However, the issue becomes more complex when considering a different external source of our desires. Kane (2000a) introduced the idea of a “covert, non-constraining control” (CNC) that could manipulate one’s higher order desires (p. 401). This includes, but is not limited to, drugs, hypnosis, and propaganda. An interesting example of a CNC is the Frankfurt-style case of the “nefarious neurosurgeon” (Frankfurt, as cited in Fischer, 2000). In this thought experiment, we must imagine that a rogue neuroscientist is controlling our higher order desires, which are in alliance with our first order desires. While Johnson (2016) argues that manipulation or programming by external factors would not have any weight in the problem of free will, I believe that one would instinctively say that they did not have free will in the neuroscientist case. This discrepancy forms because Johnson (2016) has only addressed external factors that abide by laws of nature, such as genes and environment. However, he has not considered the possibility of a much more unnatural source of higher order desires, such as a rogue neuroscientist, hypnosis, or propaganda (Frankfurt, as cited in Fischer, 2000). When we consider such possibilities, it seems as though we would not have free will in these cases, even if we acted on desires that do not contradict desires of higher orders (Haji & Cuypers, 2001). Therefore, this thought experiment is substantial as an objection to the higher order theory.
The reason why the higher order theory fails in both the case of the kleptomaniac and the case of the neuroscientist is that both examples prove that the existence of free will is not only dependent on the congruency between first order desires and higher order desires. Instead, it is dependent on how these desires are formed, how these desires motivate actions, and whether such desires reflect one’s true identity. In the case of someone with OCD or kleptomania, their identity, and hence their desires, are manipulated by their mental disorders (Glannon, 2012b; Meynen, 2012). In the case of the neuroscientist, their true identity is mutable by an external factor which does not abide by the laws of nature (Fischer, 2000; Frankfurt, 1971). Both of these lead to a changeability in desires. Therefore, in order to determine if a person is acting in accordance with their true identity, we must consider what one would do if nothing had gone wrong. One must imagine a “perfect world” in which the causal relationships between beliefs, desires, and actions are properly working, without the interference by mental disorders or “covert, non-constraining controls” (Kane, 2000a, p. 401). For example, if the kleptomaniac did not have such a disorder, what would their desires be? If they are acting from the motivation of such desires, then we can conclude that they have free will, as these desires would be derived accurately from their identity.
The neuroscientist scenario, on the other hand, poses much more of a complication. If one was in fact controlled by a neuroscientist, or some other “covert, non-constraining control”, there would be no way to actually confirm this (Kane, 2000a, p. 401). Therefore, it would be impossible to even conceive what a life without the external factor would look like, and what one’s desires would be. Evidently, the neuroscientist thought experiment weakens the higher order theory substantially. (Fischer, 2000; Frankfurt, 1971).
It is evident that the higher order theory, in its current form, fails to provide a solution to the free will problem. This theory should be revised so that it demands a genuine and effective causal relationship between the natural origin of our desires, our desires, and our actions, in a way that reflects our identity. If this can be accomplished there may be hope for free will.