Character Trouble (2022) needed to be written. Doris’s earlier volume, Lack of Character (2002), framed a wide range of debates about the relationship between normative and descriptive theories of moral character. But the ground has shifted in the roughly two decades since its publication. Normative theorizing in philosophy began to engage more deeply with empirical psychology, ironically, just as psychology itself was beset with a replication crisis.Lack of Character set its sights on the entire research program of Aristotelian virtue ethics, pairing philosophical ambition with a pugilistic ethos. Character Trouble is more measured than its antecedent. It engages carefully, charitably, and forthrightly with the work of critics from the intervening decades. Although Doris has exchanged the moniker of “situationist” for “character skeptic,” he hews closely to the original thesis. The core commitment of character skepticism is that “dispositional variables have weaker influence, and situational variables stronger influence, than one should expect if one understands dispositional differences in terms of robust traits issuing in cross-situationally consistent behavior. If you like: the influence of dispositional variables falls short of, and the influence of situational variable exceeds, many people’s priors” (217).The book is divided into eleven chapters. A useful appendix (coauthored with Edouard Machery) offers practical advice for carrying out interdisciplinary research in moral psychology. The first ten chapters consist of collected essays, starting with “Persons, Situations, and Virtue Ethics” (1998). The essays advance theses that are startling (a person can be held morally responsible for tragic mistakes that do not result from any moral defect) (137), radical (criminal law should do without character assessment) (52), and deeply unsettling (perpetrators of atrocity typically occupy excusing conditions and are therefore not morally responsible for their conduct) (79). But the main event is the eleventh chapter, “The Future of Character,” itself divided into nine subchapters. It contains Doris’s assessment of how best to think about character’s role in moral psychology and normative ethics. Doris defends the significance of classic studies by Stanley Milgram and Philip Zimbardo and advances a new argument about effect sizes that he identifies as the “fulcrum” (189) of the character debate.Arguably the central exhibit in the character skeptic’s case, Milgram’s iconic obedience experiments have come under renewed scrutiny since the expiration of the seal on Yale’s Milgram archives. Drawing from newly accessible exit interviews, critics have cast doubt on whether Milgram’s subjects believed the experiment’s cover story. Matthew M. Hollander and Jason Turowetz (2017: 661) find that 72 percent of obedient subjects in their sample explained their compliance with beliefs coded as “learner not being harmed,” while just 11 percent of defiant subjects did so, suggesting that incredulity was implicated in obedient behavior. Here Doris points out the likely possibility that professed incredulity consists of self-serving rationalization, and that attributing sincere incredulity renders mysterious the extreme stress that subjects displayed. Doris maintains that the canonical interpretation that Milgram’s subjects demonstrated “unexpectedly high levels of hesitant and conflicted obedience” is not in need of revision (202). Undoubtedly, many were surprised by the results. But should they have been? In my view, Milgram arranged the perfect storm: formidable social pressure to comply combined with ready resources to rationalize compliance (natural doubts about the authenticity of the setup). Milgram induced serious moral failure, but the task was not easily accomplished. It required exquisite calibration of both pressure and doubt sufficient to sway but not to exculpate.Zimbardo’s infamous 1971 Stanford Prison Experiment (SPE), in which ordinary young men slipped into the role of bullying, sadistic guards or sullen, passive prisoners, is another key chapter in the character skeptic narrative. It has long been observed that the SPE was not in fact an “experiment” since it lacked controls. More recently, critics have pressed the more troubling concern that the participants in the demonstration were merely playacting in response to financial incentives (Blum 2018; Le Texier 2019). Here Doris points out that neither “I was paid to” nor “I was just pretending” morally excuses: “Cruelty in the service of pretense remains cruelty” (208). Doris maintains that what happened in Palo Alto in the summer of 1971 was indeed a moral calamity: “What the SPE suggests is that ordinary people can be brought to do shitty things (or worse), under the sway of depressingly modest situational influences. Once again, that’s the disconcerting disproportion motivating character skepticism” (206).Doris’s novel argument for character skepticism derives from what he dubs the Lotta-Little Principle: “Typically, many factors are implicated in complex psychological outcomes, and relatively seldom are individual factors implicated especially strongly—i.e., with large effect sizes…. Moreover, a lot of little influences may be unexpected, and interact with one another in myriad ways” (172). Doris argues that the Lotta-Little Principle explains why effect sizes for multivariate psychological phenomena will be low and so behavioral consistency associated with any given trait will be limited. In light of this, “debates about the interpretation of particular experiments recede in importance” (233). Doris is surely right about human behavior’s intricate roots, but I am skeptical that Lotta-Little makes too much trouble for attributing traits associated with imperfect duties like helping strangers. If my “batting average” for helping is sufficiently high, do I need to be worried that on particular occasions whether or not I help strangers is influenced by a variety of factors? Granted, I do need to worry if I fail to help when helping is morally required. But as anyone who lives in a city will know, there is often ambiguity surrounding the moral requirement to help strangers. There is less ambiguity about the requirement to help one’s friends (and I would expect, also more consistency in behavior). But this gets much less attention in social psychology.Doris envisions a “future for character highly revised” (241) that calls for a rule-based ethics over one that is character-based. Indeed, if our intuitive understanding of moral character undergoes significant upheaval it would be very surprising if there weren’t deep implications for moral life. Rachel Cohon and I have argued that the denial of cross-situational consistency threatens our ability to make or accept promises in good faith (Cohon and D’Cruz 2016), and I’ve argued that it undermines the warrant for interpersonal trust (D’Cruz 2015). Good-faith promises and warranted trust both presuppose a capacity for cross-situational consistency without which the normative expectation that promises be kept and trust honored, along with our practices of holding each other accountable, simply do not make sense. We (prudentially) should not extend trust to those with weak dispositions to honor it, and we (morally) ought not to invite the trust of others. But these practices are at the very center of moral life. If they are under threat then the situationist thesis is radically morally revisionary in a way that has not been fully appreciated.In the new volume, Doris addresses the concern head-on: “My inclination here is to a considerable extent concessive. I’ll allow that both promising and promised may both feel a bit different to the true-believer situationist” (250). However, Doris maintains that “institutions, relationships, and norms all contribute to promises being kept, so whether we acknowledge it or not, we’re not depending on autonomous individuals in our practices of trust, but socially sustained individuals” (250).It is surely right that institutions, relationships, and norms support a wide range of behavioral regularities. It is also right that trust is very rarely “unqualified.” Plausibly, trusting wisely includes supporting trustworthiness rather than setting people up for failure. But promises and trustworthiness are valuable to us precisely in those circumstances in which supportive situational pressure goes absent. If I lend a friend my credit card to pick up beer, I expect them to refrain from a spending spree even if they know I never check my statements. Certainly there are cases where we take up a defensive stance toward others because we are pessimistic. As Doris says, “Many of us engage in some self-protective social wariness, and that’s a good thing” (250). But as socially embedded beings our capacity to economize on trust is limited. Even in the instances when trust is abused or disappointed, our practices of forgiveness presuppose a capacity for durable reform.If character skepticism is too pessimistic about the human capacity for cross-situational consistency, then I think that it is radically (and dismally) morally revisionary. But I do not think that such pessimism is warranted, especially when it comes to our interactions with intimates. Interestingly, there is a strain in Doris’s work on agency that appears more optimistic about consistency than his work on character. In Talking to Ourselves (2015) Doris proposes that people’s identities and actions are shaped by the self-depictions or “biographies” that they express to themselves and others. Could this confer the stability necessary to make promises in good faith? In the new book, Doris’s own analysis offers grounds for optimism: “That one understands one has made a promise, and further understands oneself to be a person who honors her promises, may help ensure that the promise is kept. One’s biography can secure behavior expressing one’s values, even in the face of unfavorable circumstance or instable inclination” (156). Here knotty questions lurk about whether “having promised” belongs to the “person” or to the “situation.” But clearly there is an important difference between “having promised” and smelling freshly baked bread.Much of the evidence for character skepticism comes from the social psychology of one-off interactions with strangers. In contrast, the resources we have for consistency are best exemplified in extended interpersonal relationships (Westra 2022). Relatedly, in a version of the Milgram experiment in which the two participants knew each other for at least two years, the rate of obedience is very low (Athanassoulis 2023).Of course, moral failure will occur even in our closest relationships. The way a person takes responsibility, seeks forgiveness, and effects meaningful change speaks volumes regarding their character. My own sense is that vitalized theories of character will require attentiveness to the longitudinal, the fine-grained, and the narratological that we find in biography, history, and literature as much as to advancements in the methods of psychological science.Character Trouble is essential reading for anyone who wants to understand the last decades of debate over moral character and who seeks to shape its future. Doris’s stimulating book opens up space for a conception of character that is nuanced, capacious, and morally ambiguous. The great value of Doris’s work consists in its commitment to follow the evidence even when it is messy and does not reassure. We are well served to appreciate more deeply the power of situations and the closeness of moral calamity. Character Trouble directs us to reckon honestly and unflinchingly with moral fragility.Earlier drafts benefited from comments from Bradley Armour-Garb, Nomy Arpaly, Rachel Cohon, John Doris, Iskra Fileva, and Ariel Zylberman.

中文翻译：

---

性格问题：关于道德能动性和人格的散漫散文


需要写《人物麻烦》（2022）。多丽丝的早期著作《缺乏品格》（Lack of Character，2002）提出了关于道德品格的规范性理论和描述性理论之间关系的广泛争论。但自该书出版以来的大约二十年里，情况发生了变化。具有讽刺意味的是，哲学中的规范理论开始更深入地与经验心理学联系起来，正如心理学本身也陷入了复制危机一样。缺乏品格将目光投向了亚里士多德美德伦理学的整个研究计划，将哲学野心与拳击精神结合起来。性格问题比它的前身更加有分寸。它谨慎、仁慈、坦率地参与了几十年来批评家的工作。尽管多丽丝将“情境主义者”的绰号改为“品格怀疑论者”，但他仍然坚持最初的论点。性格怀疑论的核心承诺是，“如果人们从跨情境一致行为中产生的稳健特征来理解性格差异，那么性格变量的影响力会更弱，而情境变量的影响力会更强。”如果你愿意的话：性格变量的影响低于许多人的先验，而情境变量的影响超出了许多人的先验”（217）。本书分为十一章。有用的附录（与爱德华·马切里合着）为开展道德心理学的跨学科研究提供了实用的建议。前十章由论文集组成，从《人、情境与美德伦理》（1998）开始。 这些文章提出的论点令人震惊（一个人可以对并非由任何道德缺陷造成的悲剧性错误承担道德责任）（137），激进（刑法应该不进行品格评估）（52），并且令人深感不安（肇事者）暴行通常占据借口条件，因此不对其行为承担道德责任）（79）。但主要事件是第十一章“性格的未来”，它本身分为九个子章节。它包含多丽丝对如何最好地思考人物在道德心理学和规范伦理学中的作用的评估。多丽丝捍卫了斯坦利·米尔格拉姆和菲利普·津巴多经典研究的重要性，并提出了关于效应大小的新论点，他将其视为性格辩论的“支点”（189）。可以说，性格怀疑论者案例中的核心表现是米尔格拉姆标志性的服从自从耶鲁大学米尔格拉姆档案封印到期以来，实验受到了新的审查。根据新近获得的退出访谈，批评者对米尔格拉姆的受试者是否相信该实验的封面故事表示怀疑。 Matthew M. Hollander 和 Jason Turowetz (2017: 661) 发现，样本中 72% 的服从受试者解释了他们遵守编码为“学习者不会受到伤害”的信念，而只有 11% 的违抗受试者这样做，这表明怀疑是涉及服从行为。在这里，多丽丝指出了一种可能的可能性，即公开表示怀疑是出于自私的合理化，而将真诚的怀疑归咎于受试者所表现出的极端压力变得神秘。 多丽丝坚持认为，米尔格拉姆的受试者表现出“出乎意料的高水平的犹豫和冲突的服从”的规范解释不需要修改（202）。毫无疑问，许多人对这个结果感到惊讶。但他们应该是这样吗？在我看来，米尔格拉姆安排了一场完美的风暴：要求遵守的强大社会压力，加上现成的资源，使遵守合理化（自然怀疑这一安排的真实性）。米尔格拉姆造成了严重的道德败坏，但任务并不容易完成。它需要对压力和怀疑进行精确的校准，足以动摇但不能开脱。津巴多臭名昭著的 1971 年斯坦福监狱实验 (SPE) 是另一个关键，在该实验中，普通年轻人扮演了欺凌、虐待狂的看守或闷闷不乐、消极被动的囚犯的角色。人物怀疑论叙述的一章。长期以来，人们一直认为 SPE 实际上并不是一个“实验”，因为它缺乏控制。最近，批评者提出了更令人不安的担忧，即示威活动的参与者只是为了回应经济激励而进行的表演（Blum 2018；Le Texier 2019）。多丽丝在这里指出，“我是被付钱的”和“我只是假装”都不是道德借口：“为假装服务的残酷仍然是残酷”（208）。多丽丝坚持认为 1971 年夏天在帕洛阿尔托发生的事情确实是一场道德灾难：“SPE 表明，在令人沮丧的温和情境影响下，普通人可能会做出糟糕的事情（或更糟）。这又是令人不安的不成比例引发的性格怀疑”（206）。多丽丝关于性格怀疑论的新颖论点源自他所谓的洛塔-利特尔原则：“通常，许多因素都与复杂的心理结果有关，而个人因素相对很少有特别强烈的影响——即具有很大的影响……。此外，许多微小的影响可能是意想不到的，并以多种方式相互作用”（172）。多丽丝认为，洛塔-利特尔原理解释了为什么多元心理现象的效应大小会很低，因此与任何给定特征相关的行为一致性都会受到限制。有鉴于此，“关于特定实验解释的争论不再重要”（233）。多丽丝对于人类行为错综复杂的根源的看法肯定是正确的，但我怀疑洛塔-利特尔在归因与帮助陌生人等不完美职责相关的特征时会带来太多麻烦。如果我帮助陌生人的“成功率”足够高，我是否需要担心在特定情况下我是否帮助陌生人会受到多种因素的影响？诚然，当道德上需要帮助时，我确实需要担心如果我没有提供帮助。但居住在城市的任何人都知道，帮助陌生人的道德要求常常含糊不清。帮助朋友的要求不再那么含糊（我希望行为上也更加一致）。但这在社会心理学中受到的关注要少得多。多丽丝设想了一个“对性格进行高度修改的未来”（241），它要求基于规则的道德观而不是基于性格的道德观。事实上，如果我们对道德品质的直觉理解经历了重大变革，那么如果没有对道德生活产生深刻的影响，那将是非常令人惊讶的。 Rachel Cohon 和我认为，否认跨情境一致性会威胁到我们真诚地做出或接受承诺的能力（Cohon 和 D'Cruz 2016），并且我认为这会破坏人际信任的保证（D'克鲁兹 2015）。善意的承诺和有保证的信任都以跨情境一致性的能力为前提，如果没有这种能力，信守承诺和尊重信任的规范期望以及我们相互问责的做法根本就没有意义。我们（谨慎地）不应该向那些性格较弱的人提供信任以兑现它，我们（道德上）不应该邀请他人的信任。但这些实践是道德生活的核心。如果他们受到威胁，那么情境主义的论点就会以一种尚未得到充分理解的方式从根本上进行道德修正。 在新书中，多丽丝正面解决了这个问题：“我在这里的倾向在很大程度上是让步的。我承认，对于真正的信徒情境主义者来说，承诺和承诺可能都感觉有点不同”（250）。然而，多丽丝坚持认为，“制度、关系和规范都有助于兑现承诺，因此无论我们承认与否，我们在信任实践中并不依赖于自主的个人，而是依赖于社会维持的个人”（250）。制度、关系和规范支持广泛的行为规律，这无疑是正确的。同样正确的是，信任很少是“无条件的”。似乎，明智的信任包括支持信任，而不是让人们陷入失败。但正是在缺乏支持性情境压力的情况下，承诺和信任对我们来说才有价值。 如果我借给朋友信用卡去买啤酒，我希望他们不要大肆消费，即使他们知道我从不检查我的账单。当然，在某些情况下，我们会因为悲观而对他人采取防御立场。正如多丽丝所说，“我们中的许多人都有一些自我保护的社交警惕性，这是一件好事”（250）。但作为融入社会的人，我们节省信任的能力是有限的。即使在信任被滥用或失望的情况下，我们的宽恕实践也以持久改革的能力为前提。如果性格怀疑论对人类跨情境一致性的能力过于悲观，那么我认为这在道德上是根本性的（而且令人沮丧的）修订版。但我认为这种悲观主义是没有道理的，特别是当涉及到我们与亲密的人的互动时。有趣的是，多丽丝在代理方面的工作似乎比他在性格方面的工作对一致性更加乐观。在《与我们自己对话》（2015）中，多丽丝提出，人们的身份和行为是由他们向自己和他人表达的自我描述或“传记”塑造的。这能否赋予真诚做出承诺所需的稳定性？在新书中，多丽丝自己的分析提供了乐观的理由：“一个人明白自己已经做出了承诺，并进一步了解自己是一个信守承诺的人，可能有助于确保信守诺言。一个人的传记可以确保表达一个人价值观的行为，即使面对不利的环境或不稳定的倾向”（156）。这里潜伏着一个棘手的问题：“已许诺”是属于“人”还是属于“情况”。但显然，“做出承诺”和闻到新鲜出炉的面包之间存在着重要的区别。性格怀疑论的大部分证据都来自与陌生人一次性互动的社会心理。相比之下，我们拥有的保持一致性的资源在扩展的人际关系中得到了最好的体现（Westra 2022）。与此相关的是，在米尔格拉姆实验的一个版本中，两个参与者彼此认识至少两年，服从率非常低（Athanassoulis 2023）。当然，即使在我们最亲密的关系中，道德失败也会发生。一个人承担责任、寻求宽恕和实现有意义的改变的方式充分说明了他们的性格。我自己的感觉是，充满活力的性格理论需要关注我们在传记、历史和文学中发现的纵向、细粒度和叙事性，以及心理科学方法的进步。性格问题是必不可少的适合任何想要了解过去几十年有关道德品质的争论并寻求塑造道德未来的人。多丽丝这本激动人心的书为细致入微、广阔且道德上模糊的人物概念开辟了空间。多丽丝工作的巨大价值在于它致力于遵循证据，即使证据是混乱的并且不能让人放心。我们应该更深入地认识形势的力量和道德灾难的迫在眉睫。 《性格问题》引导我们诚实而坚定地正视道德脆弱性。早期的草稿受益于 Bradley Armour-Garb、Nomy Arpaly、Rachel Cohon、John Doris、Iskra Fileva 和 Ariel Zylberman 的评论。