~/,   /skills,   /person,   /projects,   /blog

Free Agency for Frankfurt and Watson

Published at Mar 25, 2025 | Back to blog page




Compare and contrast the accounts of free agency presented by Frankfurt and Watson.

Both Frankfurt and Watson develop compatibilist accounts of freedom that preserve a notion of agency within determinism by examining the relations of the agents’ mental states, or their experience of the will, though they differ significantly in their approach.

Firstly, desire or to want are “elusive” (324) terms; they can range from intensely needing an act in a certain way to having no emotional response in relation to the action. The want F\&W explore is the abstract quality of desire, not its intensity or direction.

Frankfurt posits that the human will is unique because we can have desires about desires, which he calls second-order desires. An n-1 desire could be wanting to smoke a cigarette, while an n-2 desire would be wanting to want to smoke. Frankfurt places no limit on these hierarchical orders, allowing for theoretically infinite dimensions of desires about desires. For Frankfurt, freedom involves having the will one wants to have, meaning all of the agent’s orders of desire are in harmony with the action itself.

Watson, while agreeing that free will involves different modes of desire, rejects Frankfurt’s potentially infinite hierarchical structure. Instead, Watson proposes two systems: motivational (representing immediate appetitive desires) and evaluative (representing judgments about what is valuable or good). In ancient terms, this would be the rational vs. the appetitive soul, or in Nietzschean terms, Dionysian vs. Apollonian acts. These systems don’t command each other hierarchically but interact in conversation. However, Watson sees the evaluative system as superior, stating freedom is acting in accordance with our evaluative system rather than being controlled by immediate desires.

How do they both work out the idea that freedom involves the ability to get what one most wants?

Both philosophers conceive freedom as getting what one most wants, but differ crucially in how they identify what one “most wants.” For Frankfurt, what we most want is determined by our second-order volitions—our desires about which first-order desires should be effective. Freedom means having the will one wants to have and acting on it.

For Watson, what we most want is determined by what we value—our ethical or aesthetic judgments of good. Freedom means acting from our evaluative system rather than being controlled by our motivational system. This mirrors Frankfurt’s system but articulates it with a degree more confinement, not letting higher-order desires be anything but only what the agent values.

Explain the implications of these approaches for Frankfurt’s cases of the unwilling and willing addicts.

Frankfurt’s unwilling addict has a first-order desire for drugs but a second-order volition not to act on this desire. When they succumb to addiction, their desires are in conflict, and Frankfurt concludes they lack freedom of will because they cannot bring their effective will in line with their second-order volitions.

Under Watson’s approach, the unwilling addict similarly lacks freedom because their motivational system (craving the drug) overwhelms their evaluative system (judging that taking the drug is not good). They are in agreement here.

The willing addict has both a first-order desire for drugs and a second-order volition that affirms this desire. Frankfurt argues this agent has freedom of will (their will is what they want it to be) but lacks freedom of action (they cannot do otherwise).

From Watson’s perspective, the willing addict appears free only if their evaluative judgment genuinely affirms drug use as valuable. However, Watson could question whether addiction has corrupted the evaluative system itself, suggesting that what appears to be evaluative endorsement might actually be rationalization driven by addiction.

Do you agree that the willing addict acts freely, even though they lack freedom of will?

I disagree with Frankfurt’s assessment that the willing addict acts freely as his conclusion appears to be incongruent with his theory. The willing addict’s n1 and n2 desires align, but an nth degree could show the contradiction of its subordinate desires with a desire for health or existence.

To give a personal example, I consume nicotine and like that I like nicotine; the aesthetic and setting the action demands lubricates social connection. However, there are times when I wish I didn’t, yet see my n-2 and n-1 desires as fixed. My n-3 desire, wanting to never have had the want to desire nicotine, is still persistent as I feel the habit hurting my body and restricting the total set of actions I can perform.

It is as Spinoza says, “men believe themselves free, because they are conscious of their volitions and desires, but are ignorant of the causes by which they are led to wish and desire” (Ethics, Part II, Proposition 35). Sinful action (avarice, vanity, gluttony, lust, sloth, etc.), or desires that subsume the will cannot be considered to be chosen freely. We are led by flaws in the human condition (laws of nature) towards these desires, flaws that if acknowledged are incongruent with Frankfurt’s free will.

A truly free agent needs more than just alignment between known or operational levels of desire; they need desires formed under conditions allowing for critical reflection free from distorting influences.

If the unwilling addict is not free, what about ordinary weak-willed agents? Is it plausible to suppose they, too, lack freedom of will?

The comparison between unwilling addicts and ordinary weak-willed agents reveals an important distinction about freedom of will. Both exhibit a similar structure of conflict—acting against their better judgment or higher-order volitions. However, there are crucial differences in degree and duration.

The unwilling addict experiences a chronic, compulsive state where physiological factors consistently override their higher-order volitions. In contrast, weak-willed agents experience episodic lapses that don’t completely define their agency. When an agent falls victim to the kiss of a drunk cigarette, he experiences incontinence (akrasia) but not a complete lack of freedom.

It is implausible to claim that every instance of weakness of will constitutes a complete absence of freedom. Thus, Watson’s degrees of free will better extend an understanding of an agent’s evaluative capacities, and thus the extent to which they are free agents.

Weak-willed agents retain substantial freedom of will despite momentary lapses. Their evaluative system remains intact and functional most of the time, even if occasionally overridden. The unwilling addict, however, suffers from a more systematic disruption of their ability to act from their evaluative judgments. This distinction is crucial for moral assessment as it maintains that weak-willed agents are still responsible agents, agents who possess freedom of will, even if that freedom is imperfectly exercised.

Is this idea better conceptualized in terms of higher-order volitions, or in terms of the agent’s valuation system?

Within the binary bounds of Watson and Frankfurt’s description of free agency, Watson’s valuation system offers a more compelling conceptualization. Frankfurt’s hierarchical model faces an infinite regress problem—authenticity of subordinate orders of desire are dependent on their n+1 order of desire. If we are to say the unwilling addict doesn’t want to take the drug, who is to say the second order is legitimate—this process is repeated for the third, on the second to n∞-1 legitimizing n∞. This makes it impossible to truly determine if a will is free, as infinity is impossible to calculate and grasp. Watson binds the orders to rationale (evaluative) and appetitive without a hierarchy of legitimizing, making harmony between the two the condition for free will.

This system seems more functional; however, it still has its own flaws. Free will under compatibilism is only true in the agent’s eyes, and for the sake of moral blame. However, everything is necessitated and the laws behind human nature are still largely unknown. As action is a necessitated reaction to antecedent events, does it not help to understand why that action is necessitated? Both philosophers operate under the presupposition that the social self or the perceived self is the correct self. The tension of wanting to act in accordance with convention (morality and what society deems valuable), and wanting to act in accordance with myself is not resolved. Both Frankfurt and Watson place value on morality rather than desire or the will of the body; the correct self is the self in the context of their embodied social.

To use another personal example, Bitches Brew by Miles Davis is an album I struggle to feel desire for. My social identity as a Jazz musician compels me to “ought” to desire it, in order to truly verify that quantification of identity. I still listen to it, I have the album on vinyl, I play it on Spotify not out of enjoyment but pure discordant worship of the social. This simple example can be expanded to other moral actions that disagree with the agent’s conditions.

In this sense, Diogenes and Crates may be the only truly free men in history. Athens did not bind them to action like the psychiatrist acting against his first-order desires for excellence in his social identity (arete). They acted independently of good and evil to venerate the higher nature. Their actions are still under the command of the spectacle; however, it seems to be in agreement with the only good truly part of our constitution, that being nature. In this sense, I feel the infinite examination of will Frankfurt demands is a better conception of free will. Diogenes wants to eat, Diogenes wants to want to satisfy his natural desires, Diogenes wants to only want to want to want to satisfy his natural desires. Thus, Diogenes freely chooses his asceticism, eating without a bowl and eating plainly, outside of the pageantry/spectacle of dining.

Watson and Frankfurt are in a sense dishonest to the determinism they slate to believe. Spinoza is a brand of the cynics, with the metaphysical backing the ancients had no time for. Freedom through acknowledging the limitations of the will is the only way to grade and understand agency. However, the end of this venture will be complete harmony of will and determinism, or Dionysian veneration of the sublime command of nature.

References

Nietzsche, F. (2003). The Birth of Tragedy (S. Whiteside, Trans.). Penguin Classics. (Original work published 1872)

Spinoza, B. (1996). Ethics (E. Curley, Trans.). Penguin Classics. (Original work published 1677)

Debord, G. (1990). The Society of the Spectacle (D. Nicholson-Smith, Trans.). Zone Books.

Hard, R. (2012). Diogenes the Cynic: Sayings and Anecdotes, With Other Popular Moralists. Oxford World’s Classics.

Watson, G. (2003). Free Will. Oxford University Press.

Plato. (2007). Republic (D. Lee, Trans.). Penguin Classics. (Original work c. 375 BCE)

Plato. (2005). Protagoras and Meno (A. Beresford, Trans.). Penguin Classics. (Original work c. 380 BCE)

Aristotle. (2004). Nicomachean Ethics (J.A.K. Thomson, Trans., revised by H. Tredennick). Penguin Classics. (Original work c. 350 BCE)

Deleuze, G. (1990). Expressionism in Philosophy: Spinoza (M. Joughin, Trans.). Zone Books.

Email me at me@danielokita.com

as well as the source code for this website here