Robot Philosopher, following Sam Harris, suggests that since we do not choose our preferences, we do not have free will. This is an odd line of attack for a determinist because the concept of a preference is mind-dependent. We never refer to mindless phenomena as having “preferences,” unless metaphorically, and determinists cannot allow that minds are causally efficacious because then a non-material phenomenon encompassing ideas and concepts would be able to bypass mere physical causation. Physical determinism involves the postulate of causal chains stretching back into the distant past. Given cause A, effect B must occur. B then becomes the cause of C, ad infinitum. When fuel is ignited in the cylinder of a car engine, the piston is forced down. It is meaningless to refer to “preferences” in that context. If we are mindless automatons following the rules of physics, as determinism conjectures, then it makes no sense to be discussing preferences in the first place. It is not a “preference” that is causing us to act in a particular manner, but physics. Such inconsistencies are commonly found on this topic.
One doubts whether Robot Philosopher or Sam Harris are actually interested in the phenomenon of preferences per se, so it seems a little disingenuous to write a response taking the topic seriously. Nonetheless, it might be rewarding to do so.
Part of the concept of a preference is that we do not always do what we prefer. Since we often do not do what we might prefer, then preferences have a perfectly voluntary aspect. Even if, counterfactually, preferences are not chosen, we certainly do not always follow them. Given this element of choice, determinism would not follow. I might prefer to read philosophy rather than drive my wife to the airport, or to just relax in the evening rather than supervising my son’s music practice and later reading to him (although that latter item was more fun). But, I do what I think I ought to do out of love for wife and son, rather than what I might prefer.
Now we come to the reason Sam Harris might have picked on preferences as having deterministic implications. Harris can respond to this description of not following one’s preferences by saying that whatever one in fact does reveals one’s real preference. If you do not remain lodged on the couch reading philosophy instead of driving one’s wife, or watching TV instead of supervising a music practice, then your real preference must be to do what is in the best interests of your wife and son in these instances. The trouble with this line of reasoning is that it involves a self-sealing fallacy, AKA the No True Scotsman fallacy. Harris would be making it true by definition that one always follows one’s preferences and is no longer making an assertion of empirical fact. With the No True Scotsman fallacy, someone hears of a terrible crime committed in London. The man says, “No Scotsman would do a thing like that.” The crime later turns out to have been committed by a Scotsman and the man responds, “No true Scotsman would do a thing like that.” He is making it tautologically true that Scotsmen do not do terrible things because if they do do terrible things they, by definition, are not a Scotsman. If “we always follow our preferences” is to be a statement about empirical reality, then it must permit of hypothetical/counterfactual counterexamples. Under what circumstances could one be considered not to be following one’s preferences? If there are none, then it is simply being asserted as a matter of the definition of words, not as a statement about the way the world works.
We routinely do what we do not prefer and do something else instead. The character Bartleby who in refusing to do what his boss asks him to replies, “I would prefer not to.” This is the normal use of the term “preference” and it is obvious that Bartleby is not outright refusing to do what his boss asks, at least verbally. It will be disconcerting to the boss because, if I remember the story correctly, Bartleby does not intend to do what he is told to do, but phrases his refusal in the subjunctive rather than giving a definitive answer. If his boss threatens to fire him, Bartleby could say, “I didn’t say I wouldn’t do it. I just said I would prefer not to!” So, the very notion of a preference entails that one can fail to follow them and thus that they have no deterministic implications. If it is insisted that whatever we actually do reveals our real preference, then that is simply to engage in a fallacy; to play with words and to cease making statements of fact.
We can concede to Harris and Robot Philosopher that lots of aspects of preferences are happenstance and not chosen by the individual, but neither are they determinative. Some preferences are not definitely not chosen. The preference not to starve, for instance, is built into us by evolution. This is a fixed element in our behavior. However, what exactly we do in order to satisfy that preference, such as a choice of job to pay for the food, and which kind of food to eat, is what biologists call the orienting element. The latter requires flexibility and the ability to improvise which is what this mysterious thing called consciousness enables us to do. The fox is driven to the henhouse in order to eat (fixed) but he must figure out how to get in to satisfy this desire (orienting). Once in, he must chase the hens. The hens will run in unpredictable directions and the fox must alter his chasing on the fly. He cannot follow a rule, because no one knows how any particular hen will react at any given moment. So, the fox cannot be a mindless automaton following an algorithm since algorithms cannot be created for one-off events. At the most, a heuristic might be involved; a rule of thumb. Once the fox has caught the hen, in retrospect, an algorithm could be written to achieve the same result, but the algorithm would be useless because it would have no application for the next time. This could be compared with a recipe for wooing a woman. If successful and one becomes happily married, it will not work for someone to do exactly what the husband did to win the heart of his beloved. If he tries, the wife will simply find it creepy. This was a plot element in The Eternal Sunshine of the Spotless Mind. Kate Winslet’s character, even though she has had the memory of a past relationship burned from the memory traces of her brain, reacts with disgust when a creepy Elijah Wood recapitulates major events and “dates” that won her over to Jim Carey’s character. Some vestigial memory remains to be reactivated, as with Plato’s conception of the River of Lethe, where people forget but can be reminded of Truth, Beauty, and Goodness by experiences on Earth, and she ends up running away from Wood.
So, in this instance of preferring not to starve, the preference is not chosen, but satisfying the preference requires free will. So, rather than counting in favor of determinism, these kinds of preferences actually require free will to be realized. Since preferences like this are life-preserving and healthy it is good that they exist.
In keeping with the usual use of the term “preference,” this preference not to starve can be overridden. I know someone who received notice that he had a terminal illness – a brain tumor. He was unaware of this until told by the hospital, and still felt fit and healthy. He went home and simply refused to eat or drink and was dead within three or four days. I am still intimidated when I think of the indomitable self-control he demonstrated in doing this.
Robot Philosopher goes on, in response to this argument, to assert that the fox remains a rule-following device (a machine). He has already stated that we are like a mechanical doll, but with a larger set of preprogrammed responses. These responses, he says, take the form of conditionals, “if…then…” statements. This cannot be true because many many events have never happened before and cannot be anticipated. The Big Bang would have had to program a more or less infinite number of anticipatory conditionals into all organisms in a manner approaching omniscience. For instance, Barbara McClintock performed experiments on maize plants involving bombarding them with radiation in a way that broke the ends off the chromosomes. The plant cells had to experiment using educated guesses as to how these chromosomes might be repaired. This involved transposition, reordering their own DNA code around until the two disjoint ends could be joined and cell reproduction could occur. Sometimes the cells succeeded, sometimes they did not. Similarly, scientists intentionally made holes in the funnels of mud wasps. Their eggs are laid underground and the nest entered by the funnel. The sides of the funnel as so smooth other insects cannot climb in or out of them. While the wasps were away, the scientists made holes in the funnels that would and could never occur in nature. The holes were entirely novel. The mud wasps, upon returning, inspected the holes and then made repairs to them, improvising. In one case building a funnel on top of the hole in the funnel. All these thoroughly unnatural events can have had no evolutionarily derived solution waiting to be activated because they had never happened before in the history of the world.
In all these cases, the fox, the maize cells, and the mud wasps, we are seeing goal-directed, teleological behavior. Determinism rules this out. Determinism is supposed to work by cause and effect pushing from behind. Goals exist in the future, not physically, and pull the organism forward. Goal-driven behavior is not compatible with the causal processes permitted by the determinist. Most importantly, to repeat, if something cannot be anticipated, there can be no rule waiting to be implemented.
The job of a mathematician is to discover new mathematical truths. We know, definitively, that mathematics cannot be reduced to the manipulation of symbols and the meaning of symbols must be considered at certain points in the process, particularly for finding novel mathematical truths. Thus, even mathematics is not merely a matter of following algorithms and “if…then…” statements. It is common for brilliant mathematicians to intuit a new mathematical truth long before they are able to construct a proof proving it. Thus, they are not following a mechanical procedure. As an extreme case, the proofs of Solomon Lefschetz were almost always wrong though his insights correct.
Some preferences are aspirational. They are modified as we go along in life. As a teenager, I aspired to be “educated” and knowledgeable. Some of this was shame based. I did not want to be the ignoramus who had to have culturally sophisticated jokes explained to me. This was presumably a preference I picked up from my social milieu – namely the relatively intellectual elite of 1970s and 1980s Christchurch, New Zealand. I would not say I was forced to adopt this preference in any way. Part of it involved a quest for accomplishment. How exactly I was to go about achieving this goal was relative to my interests, abilities, and tastes. I tried reading Das Kapital as a sixteen year old and had to put it down. I did not understand a word. I could, however, read Gustav Flaubert and Thomas Mann, so I did that instead. Crime and Punishment was attempted later and was so depressing I gave up again, only to be finally finished in firm middle age.
When I came to the US, I lived with my mentor and childhood hero. He listened assiduously to the highest of high art music and loved Proust. Wagner’s Ring Cycle featured prominently. These were not my preferences, but since I admired him, I hoped to make them my own. Alas, it was not to be, though I did come to genuinely enjoy certain symphonies of Mahler. Since I grew up playing the violin in orchestras and thus have plenty of exposure to classical music, I must be remarkably impervious to the charms of high culture music. A philistine in this regard. And Proust, with his willfully self-destructive self-deceiving central protagonists, such as Monsieur Swann, was not to my taste either. Chasing after a prostitute and hoping she will come to love him and be faithful to him was just too much for my twenty-four year old self.
So, preferences are what they are to a large extent. However, one’s goals are up to us and are chosen. Thus, preferences will be modified by other things that we do have control over. Given the goal of becoming cultured and admirable one can try to coax one’s preferences in the desired direction. I prefer Dostoevsky to reality TV. In fact, just about anything is preferable to reality TV. Reading Dostoevsky is a big effort, but, as it resonates with my sensibility, for whatever reason, I find it tremendously rewarding. I aspire to prefer one thing to the other.
One’s goals are chosen for a number of reasons. One reason might involve persuasive argument, that would be wholly consistent with free will; logic and reason proving to be efficacious. Some of it will be goals that one finds beautiful; that appeal to one’s imagination. Goals that appeal to other sets of beliefs and yet other goals. One might adopt short term goals in order to achieve a long-term goal. One then has preferences consistent with these visions of the good life, or not. Maybe there is a permanent conflict between what one wants to want, and what one actually wants along the lines of Augustine’s prayer to God to be a good man, but not yet. Let me have some more hedonistic fun before committing myself to goodness. Later in life, Augustine felt guilty for having had a girlfriend. When I first read Augustine’s Confessions I thought it rather silly. Only when I was older did I also feel guilty for having had a practice wife with whom, among other things, to have sex but who was to be dumped in favor of someone else subsequently; rather shameful behavior. Having had one’s fun, it is all very well to regret it later. Let me be good, but not yet, indeed.
The four quadrant model of Ken Wilber could be helpful here. The intersubjective culture provides many of the options to choose from; both limiting choice and expanding it. The interobjective aspects involve possible occupations that might affect one’s choices. Also, any equipment that goals require must be available for purchase. A stereo, a piano, books, tennis racquets, LPs or CDs, and so on. Without the ability to satisfy it, a goal might be abandoned. Objective elements would include DNA, hormones, XY or XX chromosomes, and any biological items. And finally, the subjective and personal, the key elements relevant to free will. Questions of taste, ideals, values, states of mind and feelings will all come into it.
A story I have related elsewhere involves someone who somewhat reluctantly did his piano practice. One day a concert pianist performed in his small town and he became inspired. He decided that he too wanted to be a concert pianist. From that moment on he had a different attitude to his piano practice and his preference to do the minimum changed.
So, the question as to whether we choose our preferences is complicated. Some are probably built in and some are relative to goals and purposes which are chosen. Sometimes we aspire to have new preferences, try, and fail. Goals and purposes are relative to our values, and what we find inspiring; what appeals to our imaginations. Whom we prefer will involve all sorts of things, from shared interests and intelligence, to what we find admirable. Not all that is within our control and there will be elements of luck concerning the people we happen to meet and who influences us. A vision of preference selection involving pure rationalism is no more appealing than determinism and does not seem particular “free” either. There should be an element of artistry in the whole thing. An autistic man developed a twenty six element spreadsheet of his ideal woman. He finally met a woman who satisfied all twenty six elements, and she wanted nothing to do with him. I do not blame her!
“I went through a similar analysis on free will after reading Sam Harris’ book on the subject. His argument is basically that we did not choose the situation in which we are born (an arguable point) and all the situations which influenced us along the way. Therefore, it is the shaping of these situations (outside of our control) which governs our choices. There is something to that argument, however, even if our choices are 99% shaped by factors outside of our control and 1% free will, then free will exists.”
The situation could be compared to the women’s studies professor who says that gender is a social construction with no input from biology. “Society” tells biological females what it means to be a woman and with what toys nice little girls play, and so on. And yet, mysteriously, the professor has her own opinion on the matter. Her views have not conformed to social determinism. This is the “subjective” quadrant coming into play. Likewise, there is some truth to the claim that gender, class, and race influence our ideas. However, members of the same gender, or class, or race, can have radically opposing and incompatible views so those things must be just one thing among many. If it were not, the belief that gender, class and race determine beliefs will itself be merely the product of gender, class and race and thus not true.
Robot Philosopher’s response to this article ignored most of it, following a circular logic. He thinks he knows that preferences are always wholly outside our control and that they are always causally determinative, and he knows this because determinism is true. However, the preference argument was used by him and Harris to try to prove that determinism is true.
 Organisms alter their DNA in a top-down goal-driven manner when necessary, putting the lie to the selfish gene concept promoted by Richard Dawkins in his stupid book of that title. Biology as conspiracy theory. One still hears supposedly knowledgeable people praise the book. Genes are often altered and even eliminated in the course of the life of an organism. Certainly, the way the gene gets expressed is frequently changed, resulting in epigenetics. How much a rat mother licks her pups, for instance, has been proved to make a difference in gene expression.
 See Gödel’s Theorem and The Halting Problem.