Transformative experiences are experiences that radically change theexperiencer in both an epistemic and personal way (see Paul 2014).When a person transformsepistemically, they gain knowledgeof “what the experience is like” that they could not havehad without the experience. For instance, falling in love for thefirst time—be it with a romantic partner, child, friend, orpet—is the only way to truly know what it’s like to fallin love and allows one to better understand human bonds andrelationships. A person might also transformpersonally, orsuch that key agential features—like their core preferences,life goals, and way they maneuver through the world—change. Forinstance, falling in love has a tendency to radically reshape thepeople and activities one prefers, change one’s major futureplans, and alter the way one moves through the world. While thesepersonal transformations might be caused epistemic ones, they alsojust could arise alongside the epistemic transformation as a result ofthe new experience. Experiences like falling in love that transformpeople both epistemically and personally aretransformativeexperiences.

Though the concept of transformative experience sketched above comesfrom Paul (2014), elements of her account can be traced back toearlier work. For instance, antecedents include Lewis’s (1988[1990]) discussion of what experience teaches with respect to thephenomenal character of the experience, as well asUllmann-Margalit’s (2006) work that highlights how making bigdecisions isn’t subject to rationality precisely because thesedecisions tend to involve personal transformation. However,Paul’s work has brought a resurgence of attention to the topic.Transformative experiences give us a way to understand the richness ofour experiences, and also raise important questions in diverse areasof philosophy. Thus, this topic permeates both our lives andphilosophical conversations. This entry discusses the differentcontexts in which transformative experiences may arise and the wideset of philosophical questions that surround transformativeexperience.

• 1. Varieties of Transformative Experience
  • 1.1 Parenthood
  • 1.2 Sense Modality and Sensory Experiences
  • 1.3 Love and Relationships
  • 1.4 Social Identity
  • 1.5 Tragedy
  • 1.6 Ideology
  • 1.7 Art and Fiction
• 2. Epistemology: Can We Know What It’s Like?
• 3. Metaphysics: Can I Become Someone New?
• 4. Decision Theory: Is It Rational to Choose Transformative Experiences?
  • 4.1 “What It’s Like” Uncertainty
  • 4.2 Future Preference Uncertainty
  • 4.3 Self Uncertainty
  • 4.4 Reasonable, Not Rational
• 5. Existentialism: Choosing Authentically
• 6. Applied Ethics: In Light of Transformation, How Should We Treat Others?
  • 6.1 Choosing for Others
  • 6.2 Long Term Decisions
  • 6.3 Transformative Environments
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. Varieties of Transformative Experience

Transformative experiences permeate our everyday lives—at leastif we’re to believe the growing literature that has sprung uparound this topic. To fully appreciate its significance, this sectionoffers a brief, non-exhaustive sampling of various purported cases oftransformative experience. It’s up to the reader to determinewhether these examples qualify as transformative in the sense of thestrict definition offered above. But perhaps more important thanwhether each of these qualifies, is the fact that considering whetherthey do raises important issues that help us better understand themyriad dimensions of our own lives.

1.1 Parenthood

Parenthood is Paul’s paradigm case of transformation (2014,2015a). Prior to having a child, one cannot anticipate what it will belike to become a parent. Further, having a child changes the parent ina personally transformative way. Core preferences and life goals areoften reshaped around a new priority: the child. The way the newparent sees the world and perceives terror and joy shifts. Zadie Smith(2013 [2018]) eloquently puts the previously unknown complexity ofhaving a child thusly:

Occasionally the child, too, is a pleasure, though mostly she is ajoy, which means in fact she gives us not much pleasure at all butrather that strange admixture of terror, pain, and delight that I havecome to recognize as joy and now must find some way to live withdaily. This is a new problem. (Smith 2013 [2018, 331])

And of course, the transformation associated with parenthood is notrestricted to one’s first biological child. It extends toadopted children, and even can extend beyond humans to feline, canine,and other beings who depend on us and with whom we can bond.

1.2 Sense Modality and Sensory Experiences

Children who grow up in the United States rather than Australiagenerally don’t experience the distinctive taste of vegemite.People born with color blindness don’t see certain colors.Tasting vegemite and seeing color for the first time are radically newsensory experiences with distinctive phenomenal characters, andprobably only involve epistemic transformation. Some people experienceeven more radical changes when they gain or lose a sense modality.Someone who has always been sighted or hearing cannot appreciate whatit is like to be blind or deaf. Likewise, people who have always beenblind or deaf cannot appreciate what it is like to be sighted orhearing. In addition to not knowing what gaining or losing a sensemodality might be like, one cannot anticipate how that gain or lossmight personally change them and the way they navigate the world. Thefact that most of society is structured to accommodate sighted andhearing people increases the difference between a change in sensemodality. And of course, appreciating the transformative nature ofgaining a sense modality carries practical implications since manypeople are in a position to decide for themselves or others whether togain a sense modality via surgical interventions.

1.3 Love and Relationships

Love has the power to transform us little by little. In fact, thisincremental change is one of the fascinating features of love when putin the context of transformative experience (see Paul 2014). Smilingat a new acquaintance is not transformative, nor is grabbing coffeewith them. Once you’ve had coffee, eating dinner together is nottransformative either. On and on each step goes. Yet, by the end ofthings, when you look back at a relationship that’s decadeslong, you realize that you could not have imagined what it would belike to bond with another person in such an intimate way. The way youapproached life, the preferences and life projects you developed, andthe way love’s rose-tinted glasses filtered the way you see theworld have radically changed you. On the flip side, ending this sortof relationship is also wildly transformative, especially ifyou’ve formed a sense of identity that revolves around yourbeloved. Furthermore, because it’s not clear that reason everrequires that you love someone in a romantic way, it’s difficultto explain how choosing love or rejecting it could be done rationally.Thus, opening oneself up to love or boldly ending it invitestransformation and raises questions about whether doing so could everbe defended rationally.

1.4 Social Identity

All of us inhabit a place in society that’s determined in partby race, gender, sexual orientation, ability, religion, country oforigin, education, socioeconomic status, political affiliation, andmuch more. Due to contingent but very real injustices, our place insociety shapes our experiences. Sometimes, it’s possible tochange our place in society (McKinnon 2015 discusses gendertransitions and Barnes 2015 offers the fictional case of Pip fromGreat Expectations), and when we do so, that change istransformative. But the way in which these social changes aretransformative underlines some troubling features these social divideshave created. First, just as, for instance, Pip, cannot explain to hispast self what social elevation is like, it is generally extremelydifficult for people in one social group to communicate what it islike to be a member of that group to someone who is outside of it.This partially explains the phenomenon of situated knowledge and alsogives rise to the possibility of testimonial epistemic injustice (see,e.g., Fricker 2007). Hermeneutical issues also arise. People’sself-conceptions aren’t developed in a vacuum, and shared normsand concepts contribute to the way in which one’sself-conception can develop, thus influencing the ways in which onecan transform. For instance, “brave inspiration” is areadily available self-conception for disabled people while“thriving person in an unconventional body” isnot—and this is an unjust state of affairs (Barnes 2015). Socialidentities and structures contribute an ethically rich dimension tothe conversation on transformative experience.

1.5 Tragedy

Unfortunately, life also contains tragedies. The trauma of war leavesveterans and civilians physically and emotionally scarred in ways thatcan change their outlook on life forever. We aren’t morallyperfect, and often fail in extremely regrettable ways. To top it alloff, people die, sometimes after extended periods of pain andsuffering. Experiences of negative valence are often epistemicallysurprising. There’s no way to predict what the death of a closeloved one will feel like, even if the death is anticipated.Life’s tragedies also have the capacity to change who we are.Sometimes, this change can happen for the better. Cashman and Cushman(2020) suggest that although we don’t want to intentionallymorally fail, moral failure can often teach incredibly valuablelessons and change how we act for the better. When it comes to death,dying creates a chasm of understanding between those who are dying andthose who are not. Chung (2017 [Other Internet Resources]), a philosopher who wrote about his impending death, tells of the chasmthat death opens up between those who are dying and those who are not.Truly acknowledging that one is dying changes your priorities and theway you experience the world in a way that can’t be explained tosomeone who isn’t dying. The same holds for someone forced towitness their beloved slowly fade away. Thompson (2020) offers aninsightful meta-point, arguing that dying is an “ultimate”transformation because it forces us to adopt a perspective of our lifeas a whole that allows us to reconsider which of our prior experiencestruly were transformative.

1.6 Ideology

Our beliefs shape who we are, and changes to our core beliefs have thepotential to transform us. For instance, De Cruz (2018) and Chan(2016, 2019) argue that religious conversions are transformative. Themission statements of universities appear to acknowledge and strivefor transformation among students (Paul & Quiggin 2020).Schwenkler (2020) and Stump (2020) both investigate the nuances ofopening oneself up to ideological conversion in light of the potentialfor doxastic and philosophical transformation, respectively.Crucially, when ideological difference are too great, people lose theability to adopt the perspective of others. This applies bothintrapersonally, such as when we look back at our past selves andmarvel at how we believed in Santa Claus, and interpersonally, whichleads to the sort of breakdown we see in hyper-polarized politicalcontexts. Speaking of politics, holding office or participating inpolitics in some deep way might also be transformative, both becausethey involve encountering people with different perspectives and alsobecause they require representing them in a meaningful way (see Allen2017). It’s tempting to think that changes in belief are merelyepistemically transformative since they clearly change our epistemicstates. But in the New Testament (Acts 9), Saul’s conversion onthe road to Damascus into the Apostle Paul isn’t just a changein beliefs, it’s a deeper change that completely alters the wayhe lives his life. The scales falling off his eyes does not merelysignify seeing the truth; it also represents a change to who heis.

1.7 Art and Fiction

Paul (2014) opensTransformative Experience with thoughtexperiment about transforming into glamorous, slightly murderous,creatures of the night: vampires. Obviously, none of us actually facethis choice. However, rhetorically, thinking about turning into afictional creature helps motivate the idea that doing so would be bothepistemically and personally transformative. After all, who can sayfor sure what becoming a vampire would really be like or how it wouldchange a person? Furthermore, this case reveals something furtherabout the connection between fiction, art, and transformativeexperience. It could be that art is transformative—or at leastquasi-transformative—because it helps us understand things thatwe have not (or cannot) experience (Aumann 2022). Furthermore, art andfiction might be transformative from the perspective of the creator aswell as the audience. For instance, the expression of oneself throughart may be a transformative experience (Riggle 2020). Finally, artclearly involves our imaginations, and imagination may provide a wayto gain knowledge about what a transformative is like prior toexperiencing it. (More on this in§2.)

2. Epistemology: Can We Know What It’s Like?

Epistemically transformative experiences are experiences where onecannot know what they are like before having them. In the paradigmcase of parenthood, the experience of becoming a parent for the firsttime has “an epistemically unique phenomenal character”(Paul 2015a, 157). One cannot know what these are experiences are likeprior to having them.

Prior work, such as Jackson (1986), provides support for thisepistemic claim. Jackson’s “What Mary Didn’tKnow” invites us to engage in a thought experiment about Mary, aperson who has lived in a black and white room her entire life. Priorto encountering color, can Mary know what it is like to see the colorred? Does it matter whether Mary is a color scientist who has studiedthe physics of color and read extensive testimonies about what thecolor red is like? Jackson’s objective—to show thatphysicalism is false because it cannot give us knowledge ofqualia—applies to transformative experience as well. In general,experiential, or “what it’s like”, knowledge cannotbe obtained prior to experience. Thus, what transformative experiencesare like cannot be known without having the relevant experience.

Many philosophers reject the claim that one cannot know whattransformative experiences are like prior to having them. Theserejections tend to center upon one of two strategies. First, one mightturn to paradigm cases such parenthood and offer reasons to believethat one can in fact know quite a lot about what it’s like to bea parent prior to becoming one. For instance, Harman (2015) arguesthat caring for a significantly younger sibling in a quasi-paternalway or witnessing a close friend or family member become a parentprovides at leastsome evidence regarding what becoming aparent is like. Krishnamurthy (2015, 180) offers a similar suggestion,arguing that “we can on the basis of experiences of similartypes know what it is like to have a child”. For instance,babysitting or raising a cat shares common elements with parenting.Knowing what these elements are like allows one to know somethingabout what being a parent is like.

Sharadin (2015) offers a variation of this type of strategy thatfocuses on non-phenomenal elements of parenthood. For instance,becoming a parent typically involves things like financial and sleepdeprivation, and we canknow that becoming a parent typicallyinvolves these things. Plausibly, there is a supervenience relationbetween these non-phenomenal elements and an agent’s phenomenalexperience of them. Further, many people, such as philosophy graduatestudents, are intimately familiar with the phenomenal experience offinancial and sleep deprivation. By building upon the non-phenomenalelements that parenthood shares with other experiences a person hashad, that person can develop an idea of what parenthood is like priorto actually experiencing parenthood.

Some empirical research work focusing on how people model uncertaintysuggests that people do in fact attempt to use this first strategy ofbuilding on what one knows to figure out what one does not know.Zimmerman and Ullman (2020, 73) model epistemic uncertainty andpurportedly demonstrate that “people can make informed decisionsabout trying unfamiliar things”. For instance, suppose thatsomeone encounters an unfamiliar yellow grape. Though theywouldn’t know what the yellow grape is like, they might appealto their knowledge of what green and red grapes are like to form ajudgment about what the yellow grape might be like. At the very least,it seems like they are justified in believing that the yellow grapewill not wildly diverge from their other grape experiences. Based onthis information, they can rationally choose to try the yellow grape.Even if they are completely unfamiliar with grapes, they might usetheir knowledge of other fruits to form an informed decision. Ingeneral, novel experiences fall into a taxonomy of the types ofexperiences one might have, and agents can use the next closestexperiences to project what the novel one might be like.

Even if this strategy is successful, it’s unclear that it showsthat we can know what transformative experiences are like prior tohaving them. Recall that transformative experiences are supposed tohave “unique phenomenal character” (Paul 2015a, 157). If oneknows enough about what the experience is like, then it’s not infact transformative. For instance, Paul (2015b) responds toHarman’s case by suggesting that through having similar enoughexperiences, Harman might in fact have known what becoming a parent islike. Thus, becoming a parent for the first time would not have beentransformative because the relevant transformative experience alreadyoccurred earlier. Whether a type of experience is transformative maybe agent-relative. Furthermore, while this strategy is promising, somecases may still elude it. For instance, when it comes to gaining orlosing a sense modality, it really does seem that the pre-transformedperson cannot have any relevantly similar experiences.

The second strategy does not deny that we often don’t know whattransformative experiences are like. Rather, it locates ourexperiential ignorance in a lack of imagination rather than an inprinciple difference in types of knowledge (i.e., propositional andexperiential knowledge). Kind (2020), for instance, argues that thedifference between projecting what transformative andnon-transformative experiences amounts to a difference of degreerather than kind. Plausibly, it’s possible to imagine whatit’s like to babysit for the first time. Further, imagining whatit’s like to be a first-time babysitter might be easier thanimagining what it’s like to become first-time parent. Forsomeone with a sufficiently good imagination—and our imaginativeabilities do also come in degrees—one can gradually scale upfrom the babysitting experience to the experience of becoming aparent. It may even be that we can augment our imaginations throughaids such as art and thus approximate what foreign experiences arelike (see Aumann 2022). The effectiveness of artistic aids as well asthe understanding gained from them also plausibly admit of degrees.Thus, the imagination strategy can explain why it’s easier tofigure out what some experiences are like than others, and why thiscan vary from individual to individual. Since paradigm transformativeexperiences tend to be among the most difficult to imagine, positingdistinct “what it’s like” knowledge is tempting.However, the difficulty in imagining what paradigm transformativeexperiences are like merely generates theappearance of adistinct kind of knowledge. The upshot is that in principle, we canknow what a transformative experience is like prior to experiencingit.

Arpaly (2020) offers a wrinkle on the imagination strategy, blamingour “crappy” imaginations for our inability to projectwhat transformative experiences are like. Sometimes, what anexperience is like is opaque because the “devilishdetails” of the experience simply haven’t occurred to us.For example, most people have not considered that if a giraffe were todrink coffee, the coffee would cold by the time it reaches thegiraffe’s stomach. Other times, we overestimate our ability toimagine things, and our overconfidence prevents us from acceptingtestimony or other evidence that would be illuminating. These casescan be tragic, such as when a depressed person’s familydoesn’t believe their reports of depression. The failure tocorrectly imagine what things would be like thus causes it toappear as if our lack of experience is the problem, sinceexperience would quickly correct this misimpression. Crucially, thereis nothing special about the “what it’s like”quality of these experiences. Though we are ignorant, we are not inprinciple ignorant and could theoretically figure out what thetransformative experience would be like.

Like the first strategy, this second one is most vulnerable when itcomes to extreme transformations in experience such as gaining orlosing a sense modality. One might also wonder whether the secondstrategy significantly undermines the core epistemic claim: that onecannot know what transformative experiences are like prior to havingthem. Even if there is no in principle difference propositional and“what it’s like” knowledge, it does seem likeimagination-based responses do acknowledge that some experiences eludeour imaginative abilities.

Finally, Moss (2018) offers an interestingly different defense of theclaim that we cannot know what a transformative experience is likeprior to having it. Moss argues that we often have“probabilistic knowledge”. For instance, suppose that 50%of the people who experience childbirth report that the experience ismiraculous. Based on that, it seems reasonable to conclude that thereis a 50% chance that experiencing childbirth is miraculous. While anindividual can’t know that the experience will be miraculous forthem, it does seem that they can know that there’s a 50% chancethat the experience will be miraculous for them. But crucially, beforeone knows that probabilistic claim, one must know that theyaren’t exceptional. For instance, suppose it turns out that ofthe people who decline epidurals, only 10% of them report that thechildbirth experience is miraculous. If one is going to decline anepidural, then they do not know that there’s a 50% chance theexperience will be miraculous since they are part of agroup—what Moss (2018, 173) calls “an alternative referenceclass”—for whom the probability does not apply.Furthermore, those who can’t rule out that they are in thealternative reference class (in this case, perhaps those who areunsure of whether they will have an epidural) also lack probabilisticknowledge. Plausibly in the case of parenthood and othertransformative experiences, probabilistic claims can be formed basedon testimony gathered from those who have had the experience. However,it could be that among those giving the testimony, there issignificant variance that’s based on whether individuals belongto an alternative reference class. Given the nature of transformativeexperience, it may be unobvious whether a particular individual has atrait that places them into an alternative reference class. Thus, evenif there’s ample data regarding how likely it is to have aparticular parenting experience, that won’t result inprobabilistic knowledge of what it’s like to be a parent. Thisdefense also works even if it turns out that there isn’t adistinctive “what it’s like” knowledge and ourimaginations can sometimes help us understand what foreign experiencesare like.

3. Metaphysics: Can I Become Someone New?

Personally transformative experiences are ones that change who anagent is. Certainly, there are epistemic elements associated withpersonal transformation: agents may not knowhow anexperience will personally transform them and they won’t knowwhat it’s like to be transformed. There might also bede se knowledge, or self-locating knowledge of the kind thatoccurs when one knows who or where one is within a world (see Lewis1979). Significantly for the context of transformative experience,de se knowledge can apply to a temporal part of a person, soit’s possible that someone who undergoes a transformativeexperience will “locate” themselves as being the personwho results from the transformation, while their earlier self wouldhave lacked thisde se knowledge since they may not havetaken the future, radically transformed person to be themselves. Inany case, many of the considerations in the previous section apply tothese epistemic elements. However, personal transformation alsointroduces interesting metaphysical issues, and this section focusesupon those.

Paul describes the situation as being one that “radicallychanges what it is like to be you, perhaps by replacing your corepreferences with very different ones” (Paul 2015a, 156). Life goals,core preferences, or the way one experiences the world may constitutepart of the change that occurs. Ullmann-Margalit (2006, 159) adds thatbecause our beliefs and desires shape the core of who we are, changesto them turn us into a “different person”. There’salso a terrifying sense of finality since these personaltransformations are “irrevocable” (Ullman-Margalit 2006, 160) or at least potentiallyirrevocable. This picture painted by Paul and Ullmann-Margalit issometimes characterized as a “Replacement Model” (seeLambert & Schwenkler 2020b, § 5) because the agent contemplatingundergoing transformation ends up “replaced” by thetransformed person. There’s supposed to be a literal sense inwhich the pre-transformation agent ceases to exist.

Spelling out the sense in which the agent no longer exists is tricky.Focusing on core features is one natural way to explain the differencebetween the pre-transformed agent and future transformed person.Perhaps when enough core features are lost, the pre-transformed agentceases to exist. (In essentialist language, perhaps some of these corefeatures are essential, or such that the agent cannot survive the lossof them.) There’s an open question regarding which of ourfeatures count as core in this way, and some empirical studies suggestthat we at least do regard some our features as being essential inthis way. Molouki, Chen, Urminsky, and Bartels (2020) have doneempirical work to see whether people exhibit semi-uniform views aboutwhich features are central to their selves. Their findings supportearlier work by Strohminger and Nichols (2014) suggesting thatfeatures concerning morality (e.g., character traits) and personalityare taken to be most central. Changes that are perceived as mostdrastic involve morality and personality, as well as features that aretaken to be causally central. For instance, if someone takes“being a philosopher” to be causally central to how theynavigate the world, then they tend to regard losing that feature as adrastic change. Valence and desires also matter: people take negativeand undesired change to be more drastic. The explanation forasymmetric valence and desire perceptions rests upon the fact thatpeople project their future selves to undergo positive development, sonegative and undesired changes lie outside their self-conception.These finding cohere with many of the sorts of personaltransformations that constitute personal transformations.

A second sort of claim that’s frequently made is that thepre-transformed agent cannot identify first-personally with the futuretransformed person, and vice versa. For instance, Velleman (1996)draws a distinction between imagining what Napoleon saw when lookingout at Austerlitz and imagining oneself as Napoleon looking out atAusterlitz. The latter removes all first-personal traces of theimaginer (and is likely impossible to do). Paul (2017) makes a similarclaim when distinguishing between affective and cognitive empathy. Theformer involves projecting our current selves into potential futurescenarios, like if I imagine what it will be for me (a childlessperson) to have a child. It’s akin to the way we empathize withothers. The latter is much richer, and it involves experiencing thepotential future from the perspective of the future, rather thancurrent self. To cognitively empathize with my future self who has achild, I would have to put myself into the first-personal experienceof that person who now has the preferences and perspective of aparent. According to Paul, I lack the ability to cognitively empathizewith my future parent self because that self is too unrelated to mycurrent self. In general, we are not able to empathize with ourdistant future selves, even if those selves haven’t undergone adramatic transformation. If Velleman and Paul are right, then thesevering of the first-personal perspective and empathy that occursafter transformation is one way in which the pre-transformed agentceases to exist.

Finally, it’s helpful to take a step back and consider the waysin which the future-transformed person is a different person. Supposethat losing central features and a disconnect in first-personalperspective are sufficient for replacing the pre-transformed agentwith a new person. How do we explain the connection between the futureand past person? On the replacement theory, transformative changeresults in a new person. Can we coherently say that thepre-transformed person hasbecome the new person? Or doestransformation annihilate and replace them?

The trick to explaining how one can become someone new without beingannihilated involves articulating two different senses of“sameness”: one in which the past and future persons arethe same and one in which they are not. For instance, Glazier (2020)notes that the type of explanation demanded mirrors a debate aboutwhethercontingentism is a coherent view. On contingentism,it’s possible that I (or anyone else) could be someone else.Philosophers who champion the necessity of identity tend to rejectthis claim as straightforwardly absurd. To explain how their accountis coherent, the contingentist must identify two senses in whichidentity claims are necessary. Glazier’s suggestion rests on theclaim that that no one’s perspective is impossible. Roughly,there’s aproprial, or perspectival, sense in which aclaim like “I am Fred” is necessary. It’s necessaryfrom “I’s” perspective that they are Fred becausefrom their perspective the world is centered upon both the“I” and “Fred”. However, it’s possiblethat the world could have centered on someone else, say Georgia. Inthat case, there’s a non-proprial sense in which “I”should have been Georgia. This latter, non-proprial sense in which“I should have been Georgia” generates the contingency ofidentity.

Chan (2019) offers a different account on which the past and futurepersons might be the samemetaphysically speaking, butdifferent in terms of theirpractical identities. Thesepractical identities are characterized by the traits that are changedin personal transformation: core preferences, life goals, ways ofnavigating the world, continuity of first-personal identity, and thelike. Importantly, these traits form the person’s self, or atleast the self that’s relevant for considerations ofself-interest. This explains why ordinary decision-making procedureshave difficulty when it comes to transformative experiences. (More onthis in the next section.) Furthermore, distinguishing betweenmetaphysical and practical senses of sameness explains the (practical)sense in which the person becomes someone new and the (metaphysical)sense in which they endure and are not annihilated.

4. Decision Theory: Is It Rational to Choose Transformative Experiences?

Transformative experiences present several challenges for traditionalmodels of decision theory. First, the epistemically transformativecomponent of the transformative experience involves a lack of“what it’s like” knowledge. Purportedly, this inturn involves a type of uncertainty that is different from the type ofuncertainty decision theory is designed to handle. Second, thepersonally transformative component means that the post-transformationperson may be extremely different from the pre-transformed person.Their preferences may radically change in unpredictable ways, anddecision procedures must find a way to accommodate these changingpreferences. Third, the personal transformation may be so extreme thatthe pre-transformed person considering whether to undergo thetransformative experience can’t make the decision in thestraightforwardly self-interested way that’s presupposed instandard decision-making models. Any of these challenges make usingdecision-making procedures impossible in the case of transformativeexperience. And if that’s the case, then it looks like onecan’t rationally choose to undergo a transformativeexperience—the choice is either arational or irrational. Giventhat so many of our momentous choices in life involve potentiallytransformative outcomes, it would be disturbing if we discovered thatthese choices could not be made rationally.

4.1 “What It’s Like” Uncertainty

Standard decision models take probabilities and values as inputs andproduce expected values as outputs. For instance, suppose ESPN’soutcome predictor projects the Kansas City Chiefs as having a 50%chance to win the Super Bowl, yet the moneyline, which is based onbetting activity, is +105 (meaning that if you bet $100 on them andthey win, you receive $105 in winning plus your initial $100 wager).By weighting the possible outcomes with their correspondingprobabilities, we can calculate the expected value of betting $100 onthe Chiefs:

If one uses a decision procedure where one maximizes expected value,it’s rational to bet on the Chiefs since betting $100 has anexpected value of $2.50 while not betting has an expected value of 0since there are no gains or losses. There are other decisionprocedures one might use as well. For instance, one might prefer amaximin procedure, in which one chooses the option with thebest worst outcome. Since the worst outcome of betting on the Chiefsis losing money and the worst outcome of not betting is neithergaining nor losing money, not betting is the best worst outcome. Thus,a maximin procedure would recommend not betting. Crucially for ourpurposes, it doesn’t matterwhich decision procedureone utilizes. Rather, what matters is that values and probabilities(or at least possibilities) are required for any decision procedure tooperate.

On the first challenge for decision theory, transformative experiencesare epistemically transformative, and agents do not know what theexperience will be like. What an experience is like determines thevalue associated with the experience. At the most basic level, part ofwhat an experience is like involves the pain or pleasure of theexperience. Pain detracts from the value of the experience whilepleasure enhances it. More complex components of what the experienceis like also affect the value of the experience. While some otherfactors—such as moral or social ones—may affect the valueof the experience, what the experience itself is like clearly plays alarge role in determining value. Without knowing what the experienceis like, agents can’t reliably project its value. If the valueof the experience is not known, then the value of any outcomeinvolving transformative experiences also remains unknown. Thus, anagent choosing between options that may involve transformativeexperience cannot use standard decision procedures since thoseprocedures require values as inputs.

This purported challenge is highly contested. Recall from§2 that many philosopher reject the claim that experiences likeparenthood are epistemically transformative (e.g., Harman,Krishnamurthy, etc.). Denying that the experiences are epistemicallytransformative means that agents can knowsomething aboutwhat the experiences are like prior to having them. For instance, thetestimony of parents or the non-phenomenal elements of the experiencemight provide some sense of what becoming a parent is like. Ifthat’s right, then perhaps at least rough values can be used ina decision procedure.

Alternatively, one might grant that we can’t know anything aboutwhat transformative experiences are like. Nevertheless, one mightstill be able to use standard decision procedures since we may knowabout the values associated with the experience even if we don’tknow what the experience is like. For instance, Dougherty, Horowitz,and Sliwa (2015) invite us to consider a mystery box. Everyone wholooks into the mystery box has a positive experience, though theycannot describe what the experience is like. On the basis of thisinformation—that the value associated with the box is alwayspositive—one can rationally choose to look in the box withoutknowing what that will be like.

Similarly, knowing the valence of outcomes can often provide rationalgrounds on which to make a decision even if what the outcome is likeis unknown. Pascal’s Wager offers one such scenario: theinfinite positive value of eternal happiness swamps any finite value.Thus, one is rationally justified—or perhaps evenrequired!—to try to believe in God. McKinnon (2015) offers theopposite sort of case by inviting us to consider an individualconsidering transitioning. Prior to transitioning, many trans peoplesuffer from severe depression and are at risk of suicide. Even if theydo not know exactly what it will be like to transition, thealternative is sufficiently horrible, and thus the decision totransition can be rational.

Paul (2015b) suggests that all these responses miss the mark. First,people who choose transformative experience purely on the basis ofvalues without knowing what the experience will be like are notchoosing on the basis of what the experience is like. Of course, itcan often be appropriate to make choices without considering what theoutcomes will be like—cases involving moral stakes are one suchscenario. However, in many paradigm cases such as parenthood, agentsoften do envision what their future lives will be like and base theirdecisions, at least partially, on that vision. Second, Paul emphasizesthe importance of the values beingsubjective, or assigned bythe agent. Because decisions involving transformative experience arepersonally significant, agents need to make the choiceauthentically. On Paul’s account, authenticity requiresthat the agents themselves ascribe value to the experience. (More onauthenticity in§5.) Thus, testimony from others about the values they’ve ascribedto an experience won’t help an agent choose authentically forthemselves.

4.2 Future Preference Uncertainty

Suppose that the approximate values of associated with outcomes can beknown. The second challenge for standard decision theory proceduresarises from the fact that transformative experiences have thepotential to be personally transformative. Since personaltransformation can involve extreme preference changes, the wayoutcomes are valued might not remain stable from before and after thetransformation. For instance, a prospective parent might not valuesleep, but once they become a parent confronted with a screaminginfant that needs to be fed every few hours throughout the night, theway they value sleep may drastically shift. The values of outcomesused as inputs typically correspond to the preferences of thepre-transformed agent making the decision. If those preferences willshift unpredictably post-transformation, then knowing the valuesassociated with the outcomes still may not be enough to allow an agentto make a rational decision.

In some sense, the challenge of changing preferences blends twofamiliar problems. First, preferences change over time. For instance,people with the means to save for retirement must balance theircurrent preferences with what they believe will be their future ones.An extreme saver’s later self may look back and wish theirearlier self had spent more on vacations. Transformative experiencesadd a further wrinkle by making the future preferences unknown.Second, changing preferences present a diachronic problem in whichpast, present, and future preferences must be balanced against eachother. Presumably, this balancing involves aggregating all of thesepreferences, as well as deciding how to weigh them. For instance, manypeople are near-biased, and tend to weigh the preferences of theircurrent and near-future selves far heavier than the preferences oftheir distant-future selves. The preferences of past selves rarelyreceive any attention at all. (Parfit [1984] has a series of thoughtexperiments generating the intuition that it’s better for painto be in the past, even if past pain is more severe than future pain.Sullivan [2018] and Boonin [2019] offer uncommonly held dissentingviews: the former argues for temporal neutrality, or not engaging inany temporal discounting of the past or distant future, while thelatter argues that the past preferences—even those belonging topeople who are no longer alive—matter.)

Pettigrew (2019) offers a sophisticated account that has twocomponents aimed at tackling both of these problems. The first, aimedat the uncertainty surroundingwhat the future preferenceswill be, bears similarity to responses to the challenge of assigningvalues given the “what it’s like” uncertainty.Basically, just as we can gather information about how people valuetransformative experiences, we can gather information about howpeople’s preferences change after undergoing transformativeexperiences. In an oversimplified example, if it turned out that 95%of new parents would prefer to be childless again, then given that onechooses to have a child, there would be a 95% chance that their futureself would have the childless preference while there’d be a 5%chance that their future self would have a child-full preference.Then, just as decision procedures weigh the values of potentialoutcomes by their likelihood when we make decisions, they couldinclude weighted preferences.

The second component, aimed at tackling the diachronic aspect ofchanging preferences, introduces a weighted general value functionthat aggregates the “local” value functions thatcorrespond to the past, present, and future selves that make up theentire life of the agent. Further, the way in which the local valuefunctions are aggregated can incorporate differences in the ways inwhich local selves handle probability assignments, decisionprocedures, and the weighing of, e.g., past versus future preferences.Aggregating the local value functions thus begins to resemble a socialchoice problem. Agents faced with choosing for changing selves oughtto maximize the general value function, which aggregates the localvalue functions accounting for each temporally located local self.

Paul (2022) points out that Pettigrew’s account is vulnerable intwo ways. First, the type of solution proposed by Pettigrew likensfuture selves to third parties. As noted, the weighted general valuefunction closely resembles functions that aggregate individualpreferences to group ones in social contexts. But it seemsinappropriate to treat our future selves as third persons instead offully integrated selves—and this is precisely the thrust of theproblem that transformative experience presents! Second, it assumesthat meaningful intrapersonal comparisons can be made. There’ssome empirical work, especially regarding how people with disabilitiesjudge their own welfare, that suggests these comparisons are difficultto make. Pettigrew (2019, § 8.6) attempts to address this issue with his“matching intervals solution”, which attempts to scalethe different utility functions to each other. Briggs (2015) offers aprudence-inspired alternative that rests on privileging aperson’s actual preferences (as opposed to their counterfactualpreferences) without privileging their present preferences (as opposedto privileging present over future ones). Though success of any ofthese intrapersonal comparisons is an open question, there iswidespread agreement that they are necessary to explain how futurepreferences can matter.

4.3 Self Uncertainty

Finally, the personally transformative aspect of transformativeexperiences creates the possibility of future selves who are sodifferent from their pre-transformed selves that neither the earliernor later selves regards the other as being “the same”person. As Ullmann-Margalit (2006, 158) puts it, “big”decisions are “core-affecting” and “transform one’sfuture self in a significant way” such that one emerges from thesituation a different person. As we saw in§3, spelling out the exact sense of in which one can become a differentperson is tricky. However, examples illustrate how one might regard atemporally related self as a different person, and shed light on theexact nature of the problem for standard decision-makingprocedures.

Parfit (1984, 326) tells the story of a young Russian who holds hissocialist ideals close to his heart. The young Russian also happensdescend from the aristocracy, and stands to inherit a fortune when hisrelatives die. He realizes that when this happens, he may lose hissocialist ideals. His instructions to his wife—“if I losethese ideals, I want you to think that I cease toexist”—reveals that he regards that potential future selfas a different person. This type of example reinforces Paul’s(2017) claim that we cannot first-personally project into futures thatare too distant or different. Agents who fully appreciate thepersonally transformative element of transformative experiencesrealize that these experiences may potentially transform them intosomeone different. Thus, from the perspective of the agent decidingwhether to undergo a transformative experience, they must treatoutcomes involving transformation as ones in which they may no longerexist.

The possibility of non-existence (or something regarded as equivalentto non-existence) means that even if an agent knows all the fact aboutvalue and future preferences, they cannot use standard decision-makingprocedures in a straightforward way—at least not ifthey’re self-interested. After all, even if a future transformedself would be incredibly happy and live a good, meaningful life, thatdoesn’t help the agent contemplating transformation if thatfuture transformed self is not them in the relevant sense. Chan (2019)makes exactly this point in the context of Pascal’s Wager. Theinfinite happiness of the believer doesn’t benefit thenon-believer precisely because becoming transformed by faith wouldchange them into a different person (which in turn undermines theWager’s appeal to self-interest). This point is closely relatedto the challenge raised by Paul for Pettigrew’s view in theprevious subsection (§4.2). The future self is alien, and though there might be some decisionprocedure by which to make a rational decision to bring about thatfuture self, that procedure is not the one related to the sense ofself-interested that is under discussion.

A response to this challenge requires choosing in a way that bridgesthe gap between the pre-transformed agent and the future transformedself. Given the nature of transformative experience, it’sdifficult to see how this could occur precisely since thepre-transformed agent lacks the ability to connect in the right way tothe future transformed self. That connection goes beyond knowing therelevant values, probabilities, and value functions. It gets at theagent’s ability to choose authentically, to which§5 turns.

4.4 Reasonable, Not Rational

Finally, it’s worth considering the possibility that rationalchoice is simply not possible when it comes to transformativeexperience. Ullmann-Margalit and Morgenbesser (1977) discuss“picking,” which is characterized by indifference betweenalternatives, such as when one pick's between cans of the same flavorof Cambell's soup. By stipulation, picking can't appeal to values,preferences, or the sense of rationality discussed above—thosethings undertermine which alternative is optimal, which gives rise tothe indifference in the first place. Importantly for Ullmann-Margalitand Morgenbesser, picking extends beyond trivial “Cambell's soup”choices to bigger ones, including transformative experiences. Even ourvalues and preferences themselves (assuming we have control overthem) may be subject to picking, since it's would be circularto appeal to them in order to explain why we hold them. Thus,many of our major life decisions, including potentially transformativeones, are subject to picking, rather than rational choice.

Continuing this line of thought, Ullmann-Margalit (2006) suggests thatrather than focusing on acting rational in the sense of“optimizing” (as one does with standarddecision-procedures), one ought to focus on “actingreasonably”. For instance, transformation involves bothdiscontinuity in one’s life as well as a point of no return. Areasonable way of navigating choices that involve these twofeatures involves attempting to minimize those features. For instance,a person contemplating a transformative decision like marriage mighttry to ease into it by first moving in with their potential spouse.Doing so minimizes the discontinuity and gives them an out, whichmakes the decision revocable. If Ullmann-Margalit is correct insuggesting that one might be able to build up to a transformativeexperience little by little, this raises an extremely interestingquestion for transformations that unfold gradually over time. Is itpossible that each incremental step can be made rationally while agiant leap to transformation cannot be made rationally?

5. Existentialism: Choosing Authentically

Investigating decisions involving transformative experiences magnifiesan existential problem as well. Recall from§4 that many of the strategies for making rational transformativedecisions fail to also preserve authenticity. When transformation ison the table, who we are is at stake. And the choice of who we are ismore fundamental than the other types of decisions we make. Who we areprovides the framework from which we make other choices, so a choicethat could change who we are is a choice between frameworks thatitself lacks a framework from which to deliberate. As Ullmann-Margalitputs it:

At bottom, we make our most fundamental choices of the canons ofmorality, logic and rationality in total freedom and without appeal toreasons. They embody acts that this literature variously describes asnihilist, absurd, or leaps (of faith). (2006: 172)

On the flip side, what “this literature” does permit isauthentic choice. For instance, deciding to transform withoutappealing to reason (perhaps because one cannot appeal to reason)coheres well to the Sartrean idea that persons do not antecedentlyhave a particulartelos. Rather, we are free to be ourauthentic selves precisely because are unconstrained and can choosewho we want to be.

On this existentialist gloss, authenticity stands in tension withrationality when it comes to transformative experiences. On the onehand, we want to be able to rationally defend our big decisions tohave a child, marry a particular person, or commit to a vocation. Buton the other, these are the types of decisions that we also want tomake authentically, and that doesn’t seem possible ifwe’re merely following what a decision procedure recommends. Infact, empirical research suggests that people do care deeply aboutauthenticity. Furthermore, people approach decisions perceived toinvolve authenticity differently from more ordinary ones. Oktar andLombrozo (2022) gave subjects vignettes in which deciding based on“deliberation” (which was akin to standard decisionmaking) conflicted with deciding based on intuition. For instance, onescenario involved picking between two people with whom to pursue aromantic relationship. One person is a better “on paper”match, and it’s stipulated that you believe the method thatdetermines them as such is accurate. The other person is one with whomyou vibe with better in that sort of intangible, gut-feeling sort ofway. Subjects leaned strongly toward pursuing a relationship with thelatter person (the “intuition” choice) instead of theformer (the “deliberation” choice). They also perceivedthis case as one that involves authenticity. In fact, when a choicewas perceived to involve authenticity, subjects favoredintuition-based decisions over deliberation-based ones. The preferenceof for intuition over deliberation raises the question of whetherthere are non-deliberative (in the sense of standard decision-making)procedures that can help us choose in transformative decisions in away that preserves authenticity.

In fact, there are accounts that purport to preserve authenticitywhile also preserving rationality. Paul (2014) favorschoosing on the basis of revelation. According to Paul, the decisionshould be reframed as one that asks whether the agent wants todiscover what life will be like and who the agent will become,post-transformation. Essentially, the agent chooses transformation forits own sake. In doing so, the agent chooses rationally by choosing inaccordance with the preference for discovery. The agent also choosesauthentically by choosing based on the desire to know what theexperience will be like forthem and whothey willbecome once transformed. The problem is that choosing based onrevelation does not always seem appropriate given the import of manytransformative experience. For instance, many transformativeexperiences involve other people, such as becoming a parent ormarrying someone. Choosing these experiences on the basis ofrevelation fails to capture concern for the resulting child or thefuture spouse. Paul might respond that this objection to revelation isa moral one, and that a similar objection could be levied if thechoice were made on the basis of self-interest. Thus, the objectionmight not count specifically against revelation. However,there’s a second worry, which is that revelation doesn’thelp in cases where the decision lies between two transformativeoutcomes as opposed to a decision between the status quo and onetransformative outcome.

Callard (2018) picks up on this last worry and points out thatrevelation for its own sake lacks a mechanism for valuing differentrevelations differently. Instead, Callard favors an account based onaspiration andproleptic reasons. For Callard,proleptic reasons are provisional ones aimed at a good to which oneaspires but may only have an “inchoate, anticipatory, andindirect grasp”. For instance, someone who aspires to be a wineconnoisseur initially lacks any conception of what they are trying toappreciate or experience when they take a sip of wine. Importantly,proleptic reasons allow the aspirant to act rationally when they startthe process even though they lack knowledge of what the end goal islike. Interestingly, these types of cases typically lack the sharpdiscontinuity that are associated with paradigm transformativeexperiences like parenthood. In fact, Callard suggests that parenthoodis not as discontinuous as assumed because potential parents oftendedicate a lot of mental energy to the decision and learning all theycan about parenthood. What appears to be a climactic“let’s go for it” is “embedded in a longertransformative journey” (Callard 2018, 63). Putting all of this together,transformative decisions can be made rationally via proleptic reasonsand are authentic since they result from intentional deliberationabout who the agent hopes to become.

Chang (2015) offers an interestingly different way of approachingtransformative decision making. Chang starts by distinguishing betweenevent-based andchoice-based transformations. Theformer are the type that fit how transformative experiences aretypically described. They involve experiences that are downstream fromthe choice—like becoming a parent—that transform theagent. The latter is where Chang’s interest lies. Choice-basedtransformation is one in which “the making of the choice itselftransforms you” (Chang 2015, 240). Here, Chang has in mind cases in which oneinvokes voluntarist reasons when making a decision. Voluntaristreasons are reasons generated by the will that—on Chang’saccount—are appropriate when all other existing reasonsunderdetermine what an agent ought to do. (Chang favors an account inwhich the will strengthens an existing reasons, but there are othervoluntarists like Korsgaard (1996) who hold that the will can createreasonsex nihilo.) Crucially, in willing a voluntaristreason, one simultaneously reveals themself to be the type of personfor whom that choice is rational. This type of account also appears topreserve authenticity (since what could be more authentic thanone’s own will) and rationality (since adding the voluntaristreason to the mix tips the scales in favor of the choice that’sbeen made). Of course, voluntarist accounts are somewhatcontroversial, so Chang’s account of how to make transformativechoices would inherit all the worries that voluntarism does.

Although these accounts—Paul’s, Callard’s, orChang’s—all face worries, it is worth noting thatif any of them work, then there is a way to preserve both authenticityand rationalityeven while respecting the original challenge oftransformative experience to standard decision making procedures.None require knowing what the experience will be like or who will comeout the other end. Nevertheless, one can still rationally andauthentically plunge into the unknown.

6. Applied Ethics: In Light of Transformation, How Should We Treat Others?

Transformative experiences also demand that we revisit how we treatothers in light of the fact that they potentially may undergotransformative experiences. As seen in§1, transformation may arise in many contexts, far too many to attempt totackle here. Instead, this section will focus on three types ofcontexts in which the issue of how we should treat others given thepossibility of transformation might arise: when we choose for others,when we make long-term decisions involving others, and when we createenvironments that (dis)favor transformation.

6.1 Choosing for Others

Plausibly, it’s wrong to interfere with the autonomy of others,especially when it comes to important personal decisions. As Akhlaghi(2022, 7) points out, interference is even less defensible whenothers are choosing to undergo transformative experiences becausethere is a

moral duty not to interfere with the autonomous self-making of others,through their choosing to undergo transformative experience todiscover who they will become.

However, people often find themselves in situations in which they needto make decisions on behalf of others that may lead to transformativeexperiences. In fact, those who take the leap and become parents thenfind themselves in a position to make transformative decisions onbehalf of their children. For instance, some parents of deaf childrenmust decide whether to get cochlear implants for their child. Ideally,children receive these implants when they are extremely young, usuallybefore the age of three. The early age at which the decision must bemade forces parents to choose on behalf of their children. In additionto being a choice that affects how their children will perceive andnavigate the world, the choice carries incredibly significant socialconsequences. Deaf communities offer distinctive goods that hearingcommunities cannot, and vice versa. Furthermore, as disabilityadvocates point out, the difference between being deaf and non-deaf is“mere-difference” rather than “bad-difference”(see, e.g., Barnes 2014). This offers a further reason to believe thatthe decision to give a child a cochlear implant is a transformativeone that must be made on the basis of the child’s futureexperiences as deaf or hearing as opposed to on the basis ofvalue.

Put this way, parental choices, especially significant ones likedeciding whether to get a child a cochlear implant, also resemble thenon-identity problem. Parfit (1984) raises the non-identity problem asa puzzle for decisions that affectwho will exist in thefuture. For instance, suppose a potential parent is choosing betweenimplanting one of two embryos, and chooses one that has a geneticdefect that causes monthly migraines that the other embryo lacks.Given that there are no other relevant differences between the twoembryos, it appears that the parent has done something wrong bybringing into existence a person who will have a lower level ofwelfare than the other person who could have existed instead. However,it’s not obviouswho the potential parent has harmed.Assuming the migraine-inflicted person prefers existing to neverhaving existed, they have not been harmed since the alternative isnon-existence. If that’s right, then perhaps the parents havedone nothing wrong since no one has been harmed.

Similarly, the personally transformative aspect of transformativeexperiences affects who exists after undergoing the experience.Receiving a cochlear implant and growing up in a hearing communitywill transform the child into a radically different person from whothe child would be if they did not receive an implant and grew up in adeaf community. The person who develops won’t have grounds tocomplain since wishing their parents had chosen otherwise would belike wishing they did not exist. But even if one does not believe thatthe cochlear implant is an existence-affecting decision point, thefact that the two potential futures are so foreign to each other meansthat a person situated in one of those futures cannot project what itwould be like to have had the other life. If the earlier points aboutprojecting value are correct, the person who has grown up in the deafcommunity cannot assign a subjective value to the alternate life inwhich they had the cochlear implant, and vice versa. Without thosevalue assignments, it doesn’t appear that either person canclaim that they would have been better off in the alternate scenario.Thus, whether the person receives the cochlear implant or not,articulating how that transformation harmed them will be challenging.Perhaps the parent can choose either option without harming orwronging anyone. The situation is also similar from the perspective ofthe parents and their preferences, for they will have grown to lovetheir child and the particular person into whom they havedeveloped.

Philosophers diverge on how parents should choose in these cases.(See, e.g., Harman (2015) for an interesting discussion on when“I’ll be glad I did it” reasoning is appropriate.)However, there’s a more interesting lesson to be drawn fromthese types of cases. Recall from§4.1 and§4.2 that Paul objects to certain types of decision-making strategiesbecause they treat future selves as if they were third persons ratherthan integrated parts of the same self. These cases show that even inthe third person case, standard decision-making procedures are notparticularly helpful. Transformative experiences still create majorchallenges for decision-making.

At the other end of life, decisions must sometimes be made for peoplepost-transformation. People suffering from dementia orAlzheimer’s transform in myriad ways, and their preferencesoften diverge from the preferences of their earlier selves. Sometimes,these preferences regard end of life care. Since the person withdementia or Alzheimer’s is not considered legally competent tochange their advanced directives, others must make important medicaldecisions on their behalf. These people must decide whether to respectthe preferences of the past self, or the current one. In some ways,this a questions about the weight that ought to be assigned to pastpreferences, and we might resolve in the way Pettigrew (2019) suggestsby constructing a general value function that aggregates the past andcurrent preferences. However, if the personal transformation is takenseriously, the current self is a different person. Binding them to thepreferences of the past self is akin to binding them to thepreferences of another person.

6.2 Long Term Decisions

In addition to transforming in the early and late stages of life,people transform throughout their lives, either via momentous“big” changes or via an accumulation of“small” changes. Long-term decisions must factor thepotential for transformative change into account. For instance, Lackey(2020) and Chan (2020) independently address this issue in the contextof punishment. Lackey’s argument demands that those with thepower to issue long-term punishments, like judges, must consider thepossibility of transformative change. Plausibly, punishment should besensitive to the relevant evidence available at that time. We knowthat people change, often in transformative ways, and imprisonmentseems like exactly the type of experience that would betransformative. Issuing strict, long-term prison sentences—suchas life sentences without the possibility of parole—screens offthe possibility of taking relevant future evidence into account. Thus,recognizing the possibility of transformation offers a powerfulargument for sentencing reform.

Extended partnerships also provide another context in whichtransformative changes need to be accounted for when making long termdecisions. For instance, marriage vows typically involve some type oflife-long pledge. And, as anyone familiar with married people knows,people change in ways that disrupt the marriage. In a way, identifyingthese changes as transformative makes sense of an already commonphenomenon. But it also raises a further question: how can onefelicitously make a life-long promise to someone knowing that boththey and their partner may transform in a way involving a radicalshift of preferences and life goals? The centrality of transformationto this decision is doubled because the marriage itself is oftentransformative. It’s true that transformations connected withrelationships tend to involve incremental changes because therelationship develops incrementally. But these incremental changes canlead people away from each other and won’t be noticeable untilthe larger schism develops due to their incremental nature.

These long-term decisions involving others suggest that the normativequestions transformative experiences raise go beyond self-interesteddecision making. Furthermore, some of the proposed ways of makingtransformative decisions have interesting implications for these typesof cases. For instance, Ullmann-Margalit suggests that there simply isnot a rational way to make transformative decisions. Applying that tothese cases gives the radical result that decision to marry orsentence someone to prison cannot be made rationally!Ullmann-Margalit’s further suggestion that perhaps thesedecisions can be made reasonably also has interesting implications.She suggests that minimizing the discontinuity of a big decision andbacking off the point of no return might help, and includes moving inbefore marriage as one such way to ease reasonably into a big decisionlike marriage. The application for the prison case is even moreinteresting, for it suggests that life in prison should not be such aradical departure from life outside of it, and that there must bemeaningful short term possibilities for parole.

6.3 Transformative Environments

Finally, special attention must be paid to our social environmentssince they shape the way in which members of society may transform.Education provides the most obvious instance of this. Fights overparental rights to control their children’s educations are socontentious precisely because early experiences shape who childrengrow up to become. Even higher education is taken to have atransformative goal (see Paul & Quiggin 2020). College studentsare typically in their early twenties—the last age at whichpeople tend to undergo radical change before developing fairly stablepreferences and goals that tend to endure through most of adulthood.In addition, the transformative environment of education shapes morethan the individual students who experience it. Students go on tobecome members of society who in turn create an environment that willshape those who come after them. Morton (2021) points out that thistends to have a self-reinforcing effect because elite institutionslike the Ivy League schools disproportionately influence socialpolicy. Even if students enter these institutions without the goal ofmaintaining the status quo of the elite, by the time they graduate,they can be transformed into members of the elite who are now investedin safeguarding the interests of their newfound group. This, in turn,shapes social policies that reinforce which transformative experiencesare most readily available to members of society.

This idea that social policies can shape individual choices andpreferences creates both an opportunity and a responsibility toreconsider which policies ought to be adopted in light of thepotential for transformative experience. For instance, Pettigrew(2023) examines the concept ofnudging introduced by Thalerand Sunstein (2008). In a nutshell, nudging is a sort of interventionthat pushes people toward making a particular choice while leavingopen the possibility that they choose otherwise. Opt-out retirementprograms offer a paradigm case of nudging since they increase thelikelihood that people will put money away for retirement while stillallowing people the ability to opt-out. Thaler and Sunstein labelnudging policies as “libertarian paternalism” becausewhile these interventions look paternalistic, they purportedly helppeople make the decision that their future idealized selves will deemto be best “as judged by themselves” (Thaler and Sunstein 2008, 4). Of course, if thedecisions are transformative, or incremental parts of multipledecisions that become transformative, simply looking at the judgmentof the future self does not seem appropriate at all! (And Pettigrewmakes precisely this argument before proposing a modification to the“as judged by themselves” test that’s based on hisweighted general value function.)

There are at least two bigger-picture issues that come out of thisdiscussion of nudging people, potentially toward transformation.First, the norms regarding nudging people toward potentialtransformation requires further examination. Most policies, nudging orotherwise, exhibit some degree of interventionism. In addition towhich transformations might be preferable for society as a whole,there’s also the question of whether we should adopt policiesthat maximize individual autonomy. Second, we might wonder why nudgingworks in the first place, and if it really does work fortransformative experiences. Recall that part of the nature oftransformative experiences is that one cannot know what they will belike or how they will change prior to having the experience. Nudging,on the other hand, only works if these things are somewhatpredictable. Of course, unpredictability with respect to an individualis compatible with predictability with respect to overall socialtrends. But, if nudging really works, it does lend credence to theidea certain elements of transformative experience are predictable andperhaps strengthens the case of those from§4 who believe that one can make rational choices with respect totransformative experiences.

Finally, we’ve structured our society in many unfortunate ways.This goes beyond the discrimination that people suffer on the basis ofrace, sex, ability, socioeconomic status, and the like. More subtleconceptual structures shape and limit the way we form ourself-identities. These injustices are obviously bad in themselves, andthey also interact with transformative experience in a way thatfurthers the injustice. Barnes (2015) argues that one’s“self-conception and self-identity aren’t developed incultural isolation” and that “social norms and structuresmake certain ways of interpreting or thinking about ourselves readilyavailable” (185). For example, society makes it easy for womento “re-shape their self-conception to cohere with the image of adutiful, submissive wife” (186). Similarly, “braveinspiration” is a readily available self-conception for disabledpeople while “thriving person in an unconventional body”is not (185–186). These states of affairs are bad for women anddisabled people. Since personal transformation involves changes toone’s self-conception, society plays an outsized role in thetypes of transformations that are available to us.

Bibliography

• Akhlaghi, Farbod, 2022, “Transformative Experienceand the Right to Revelatory Autonomy”,Analysis, firstonline: 31 December 2022. doi:10.1093/analys/anac084
• Allen, Peter, 2017, “Transformative Experiences in PoliticalLife: Transformative Experiences in Political Life”,Journalof Political Philosophy, 25(4): e40–e59.doi:10.1111/jopp.12131
• Arpaly, Nomy, 2020, “What is it Like to Have a CrappyImagination?”, in Lambert and Schwenkler 2020a: 122–132(ch. 6). doi:10.1093/oso/9780198823735.003.0007
• Aumann, Antony, 2022, “Art and Transformation”,Journal of the American Philosophical Association, 8(4):567–585. doi:10.1017/apa.2021.40
• Barnes, Elizabeth, 2014, “Valuing Disability, CausingDisability”,Ethics, 125(1): 88–113.doi:10.1086/677021
• –––, 2015, “Social Identities andTransformative Experience”,Res Philosophica, 92(2):171–187. doi:10.11612/resphil.2015.92.2.3
• Boonin, David, 2019,Dead Wrong: The Ethics of PosthumousHarm, Oxford/New York, NY: Oxford University Press.doi:10.1093/oso/9780198842101.001.0001
• Briggs, R.A., 2015, “Transformative Experience andInterpersonal Utility Comparisons”,Res Philosophica,92(2): 189–216. doi:10.11612/resphil.2015.92.2.7
• Callard, Agnes, 2018,Aspiration: the agency of becoming,Oxford: Oxford University Press.
• –––, 2020, “TransformativeActivities”, in Lambert and Schwenkler 2020a: 147–161 (ch.8). doi:10.1093/oso/9780198823735.003.0009
• Cashman, Matthew and Fiery Cushman, 2020, “Learningfrom Moral Failure”, in Lambert and Schwenkler 2020a:182-195.
• Chan, Rebecca, 2016, “Religious Experience, VoluntaristReasons, and the Transformative Experience Puzzle”,ResPhilosophica, 93(1): 269–287.doi:10.11612/resphil.2016.93.1.16
• –––, 2019, “Transformed By Faith”:,Faith and Philosophy, 36(1): 4–32.doi:10.5840/faithphil2019115116
• –––, 2020, “Punishing the WrongPerson”,Public Affairs Quarterly, 34(2):124–141. doi:10.2307/26921123
• Chang, Ruth, 2015, “Transformative Choices”,ResPhilosophica, 92(2): 237–282.doi:10.11612/resphil.2015.92.2.14
• Chituc, Vladimir, Laurie Paul, and Molly Crockett, 2021,“Evaluating Transformative Decisions”,Proceedings ofthe Annual Meeting of Cognitive Science Society, 43: 973-978. [Chituc, Paul, and Crockett 2021 available online]
• De Cruz, Helen, 2018, “Religious Conversion, TransformativeExperience, and Disagreement”:,Philosophia Christi,20(1): 265–275. doi:10.5840/pc201820125
• Dougherty, Tom, Sophie Horowitz, and Paulina Sliwa, 2015,“Expecting the Unexpected”,Res Philosophica,92(2): 301–321. doi:10.11612/resphil.2015.92.2.5
• Duncan, Samuel, 2020, “Commitment and TransformativeChoice”,European Journal of Philosophy, 28(4):942–953. doi:10.1111/ejop.12528
• Fricker, Miranda, 2007,Epistemic Injustice, Oxford:Oxford University Press.
• Glazier, Martin, 2020, “Being Someone Else”, inLambert and Schwenkler 2020a: 37–52 (ch. 2).doi:10.1093/oso/9780198823735.003.0003
• Harman, Elizabeth, 2015, “Transformative Experiences andReliance on Moral Testimony”,Res Philosophica, 92(2):323–339. doi:10.11612/resphil.2015.92.2.8
• Jackson, Frank, 1986, “What Mary Didn’t Know”,The Journal of Philosophy, 83(5): 291–295.doi:10.2307/2026143
• Kind, Amy, 2020, “What Imagination Teaches”, inLambert and Schwenkler 2020a: 133–146 (ch. 7).doi:10.1093/oso/9780198823735.003.0008
• Krishnamurthy, Meena, 2015, “We Can Make Rational Decisionsto Have a Child: On the Grounds for Rejecting L.A. Paul’sArguments”, inPermissible Progeny?, Sarah Hannan,Samantha Brennan, and Richard Vernon (eds.), New York: OxfordUniversity Press, 170–183.doi:10.1093/acprof:oso/9780199378111.003.0008
• Korsgaard, Christine M., 1996,The Sources ofNormativity, Cambridge/New York: Cambridge University Press.doi:10.1017/CBO9780511554476
• Lackey, Jennifer, 2020, “Punishment andTransformation”, in Lambert and Schwenkler 2020a: 230–250(ch. 13). doi:10.1093/oso/9780198823735.003.0014
• Lambert, Enoch and John Schwenkler (eds.), 2020a,BecomingSomeone New: Essays on Transformative Experience, Choice, andChange, Oxford: Oxford University Press.doi:10.1093/oso/9780198823735.001.0001
• –––, 2020b, “Editors’Introduction”, in Lambert and Schwenkler 2020a: 1–15.doi:10.1093/oso/9780198823735.003.0001
• Lewis, David, 1979, “AttitudesDe Dicto andDeSe”,The Philosophical Review, 88(4):513–543. doi:10.2307/2184843
• –––, 1988 [1990], “What ExperienceTeaches”,Proceedings of the Russellian Society(University of Sydney), 13: 29–57. Reprinted inMind andCognition: A Reader, William G. Lycan (ed.), Oxford/New York:Basil Blackwell, 1990, 499–519. Also reprinted in David Lewis,Papers in Metaphysics and Epistemology (Cambridge Studies inPhilosophy), Cambridge: Cambridge University Press, 1999,262–290. doi:10.1017/CBO9780511625343.018
• McKinnon, Rachel, 2015, “Trans*formative Experiences”,Res Philosophica, 92(2): 419–440.doi:10.11612/resphil.2015.92.2.12
• Molouki, Sarah, Stephanie Y. Chen, Oleg Urminsky, and Daniel M.Bartels, 2020, “How Personal Theories of the Self Shape Beliefsabout Personal Continuity and Transformative Experience”, inLambert and Schwenkler 2020a: 52–72 (ch. 3).doi:10.1093/oso/9780198823735.003.0004
• Morton, Jennifer M., 2021, “The Miseducation of theElite”,Journal of Political Philosophy, 29(1):3–24. doi:10.1111/jopp.12208
• Moss, Sarah, 2018,Probabilistic Knowledge, Oxford:Oxford University Press. doi:10.1093/oso/9780198792154.001.0001
• Oktar, Kerem and Tania Lombrozo, 2022, “Deciding to BeAuthentic: Intuition Is Favored over Deliberation When AuthenticityMatters”,Cognition, 223: article 105021.doi:10.1016/j.cognition.2022.105021
• Parfit, Derek, 1984,Reasons and Persons, Oxford:Clarendon Press. doi:10.1093/019824908X.001.0001
• Paul, L. A., 2014,Transformative Experience, Oxford:Oxford University Press.doi:10.1093/acprof:oso/9780198717959.001.0001
• –––, 2015a, “What You Can’t ExpectWhen You’re Expecting”,Res Philosophica, 92(2):149–170. doi:10.11612/resphil.2015.92.2.1
• –––, 2015b, “Transformative Choice:Discussion and Replies”,Res Philosophica, 92(2):473–545. doi:10.11612/resphil.2015.92.2.15
• –––, 2017, “The Subjectively EnduringSelf”, inThe Routledge Handbook of the Philosophy ofTemporal Experience, Ian Phillips (ed.), New York: Routledge, pp.262–272.
• –––, 2020, “Who Will I Become?”, inLambert and Schwenkler 2020a: 16–36 (ch. 1).doi:10.1093/oso/9780198823735.003.0002
• –––, 2021, “The Paradox of Empathy”,Episteme, 18(3): 347–366. doi:10.1017/epi.2021.31
• –––, 2022, “Review ofChoosing forChanging Selves, by Richard Pettigrew”,ThePhilosophical Review, 131(2): 230–235.doi:10.1215/00318108-9554756
• Paul, L. A. and John Quiggin, 2020, “TransformativeEducation”,Educational Theory, 70(5): 561–579.doi:10.1111/edth.12444
• Pettigrew, Richard, 2015, “Transformative Experience andDecision Theory”,Philosophy and PhenomenologicalResearch, 91(3): 766–774. doi:10.1111/phpr.12240
• –––, 2019,Choosing for ChangingSelves, Oxford/New York: Oxford University Press.doi:10.1093/oso/9780198814962.001.0001
• –––, 2023, “Nudging for ChangingSelves”,Synthese, 201(1): article 22.doi:10.1007/s11229-022-04020-2
• Riggle, Nick, 2020, “Transformative Expression”,in Lambert and Schwenkler 2020a: 162-181.
• Schwenkler, John, 2020, “Risking Belief”, in Lambertand Schwenkler 2020a: 196–211 (ch. 11).doi:10.1093/oso/9780198823735.003.0012
• Sharadin, Nathaniel, 2015, “How You Can Reasonably FormExpectations When You’re Expecting”,ResPhilosophica, 92(2): 441–452.doi:10.11612/resphil.2015.92.2.2
• Smith, Zadie, 2013 [2018], “Joy”,The New YorkReview of Books, 10 January 2013. Reprinted inFeel Free:Essays, New York: Penguin Press, pp. 329–335.
• Strohminger, Nina and Shaun Nichols, 2014, “The EssentialMoral Self”,Cognition, 131(1): 159–171.doi:10.1016/j.cognition.2013.12.005
• Stump, Jacob, 2020, “On Socrates’ Project ofPhilosophical Conversion”,Philosophers’ Imprint,20: article 32. [Stump 2020 available online]
• Sullivan, Meghan, 2018,Time Biases: A Theory of RationalPlanning and Personal Persistence, Oxford: Oxford UniversityPress. doi:10.1093/oso/9780198812845.001.0001
• Thaler, Richard H. and Cass R. Sunstein, 2008,Nudge:Improving Decisions about Health, Wealth, and Happiness, NewHaven, CT: Yale University Press.
• Thompson, Evan, 2020, “Death: The Ultimate TransformativeExperience”, in Lambert and Schwenkler 2020a: 269–288 (ch.15). doi:10.1093/oso/9780198823735.003.0016
• Ullmann-Margalit, Edna, 2006, “Big Decisions: Opting,Converting, Drifting”,Royal Institute of PhilosophySupplement, 58: 157–172. doi:10.1017/S1358246100009358
• Ullmann-Margalit, Edna and Sydney Morgenbesser, 1977,“Picking and Choosing”,Social Research, 44(4):757–785.
• Velleman, J. David, 1996, “Self to Self”,ThePhilosophical Review, 105(1): 39–76.doi:10.2307/2185763
• Yudkin, Daniel A., Annayah M. B. Prosser, S. Megan Heller, KateriMcRae, Aleksandr Chakroff, and M. J. Crockett, 2022, “ProsocialCorrelates of Transformative Experiences at Secular Multi-Day MassGatherings”,Nature Communications, 13: article 2600.doi:10.1038/s41467-022-29600-1
• Zimmerman, Samuel and Tomer Ullman, 2020, “Models ofTransformative Decision-Making”, in Lambert and Schwenkler2020a: 74–99 (ch. 4).doi:10.1093/oso/9780198823735.003.0005

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

• Chung, Ken, 2017, “Is Dying a Transformative Experience?”, Ken Chung’s blog, 23 August 2017,
• Paul, L.A., “Teaching Guide for Transformative Experience”
• “L.A. Paul on Transformative Experience”,Philosophy Bites Podcast, 8 October, 2022
• “Transformative Experiences”,Philosophy Talk Podcast, 16 November 2014

Related Entries

authenticity |decision theory |existentialism |imagination |parenthood and procreation |personal identity |preferences |rational choice, normative: expected utility