The Journal of Positive Psychology Dedicated to furthering research and promoting good practice

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rpos20

The Ordinary Concept of a Meaningful Life: The Role of Subjective and Objective Factors in Third- Person Attributions of Meaning

Michael Prinzing, Julian De Freitas & Barbara L. Fredrickson

To cite this article: Michael Prinzing, Julian De Freitas & Barbara L. Fredrickson (2021): The Ordinary Concept of a Meaningful Life: The Role of Subjective and Objective Factors in Third-Person Attributions of Meaning, The Journal of Positive Psychology, DOI: 10.1080/17439760.2021.1897866 To link to this article: https://doi.org/10.1080/17439760.2021.1897866

Published online: 07 Mar 2021.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rpos20 THE JOURNAL OF POSITIVE PSYCHOLOGY https://doi.org/10.1080/17439760.2021.1897866

ARTICLE The Ordinary Concept of a Meaningful Life: The Role of Subjective and Objective Factors in Third-Person Attributions of Meaning Michael Prinzing a, Julian De Freitasb and Barbara L. Fredricksonc aDepartment of Philosophy, University of North Carolina – Caldwell Hall, Chapel Hill, NC USA; bDepartment of Psychology, Harvard University – William James Hall, MA USA; cDepartment of Psychology and Neuroscience, University of North Carolina , Chapel Hill, NC USA

ABSTRACT ARTICLE HISTORY The desire for a meaningful life is ubiquitous, yet the ordinary concept of a meaningful life is poorly Received 5 June 2020 understood. Across six experiments (total N = 2,539), we investigated whether third-person Accepted 24 February 2021 attributions of meaning depend on the psychological states an agent experiences (feelings of KEYWORDS interest, engagement, and fulfillment), or on the objective conditions of their life (e.g., their effects Experimental philosophy; on others). Studies 1a–b found that laypeople think subjective and objective factors contribute folk theories; meaning in life; independently to the meaningfulness of a person’s life. Studies 2a–b found that positive mental moral psychology; positive states are thought to make a life more meaningful, even if derived from senseless activities (e.g., psychology hand-copying the dictionary). Studies 3a–b found that agents engaged in morally bad activities are not thought to have meaningful lives, even if they feel fulfilled. In short, both an agents’ subjective mental states and objective impact on the world affect how meaningful their lives appear.

The search for life’s meaning appears across cultures and However, at present, little is known about how study historical periods, and is commonly thought to reflect an participants use these terms, and, therefore, little is innate human need. Victor Frankl (1971) famously known about what people are reporting when they claimed that a human being’s ‘main concern is not to complete these questionnaires. gain pleasure or to avoid pain but rather to see A better understanding of the ordinary concept of a meaning in [their] life’ (p. 115). Empirical work on this a meaningful life will be important if this body of topic finds that most people do in fact report a greater research is to be put to practical use. At least one pri­ interest in living a meaningful life than a pleasant one mary reason for academics to study life’s meaning (Kim et al., 2014). Indeed, this desire is so strong that the should be to help people live more meaningfully. more a person feels that their life lacks meaning, the People make real life choices – everything from what greater their risk of suicide (Chen et al., 2020; Schnell to do with an afternoon to what career to pursue – on et al., 2018). On the other hand, those who do find their the basis of what they take to be meaningful. The trou­ lives meaningful tend to display superior mental and ble is that it is difficult to help people achieve a goal physical health across a wide range of metrics (Steger, without understanding what the goal is. 2017). For these reasons, the feeling that one’s life is One strategy for investigating the lay concept would be meaningful has been called a ‘flagship indicator of well- to simply ask people for their definitions of a meaningful being’ (Steger et al., 2013), and is the focus of an enor­ life, and to look for commonalities or themes (Wong, 1998). mous amount of research (for reviews, see: Baumeister & However, people are rarely in a position to explain how or Landau, 2018; King & Hicks, 2021; Wong, 2013). why they form the judgments that they do (Nisbett & Yet, something important has so far been missing Wilson, 1977; Wilson & Dunn, 2004). Just as competent from this research. The standard self-report measures English speakers can be unable to explain the rules of of meaning in life (e.g., the Purpose in Life Test, English grammar (Chomsky, 1965), a competent concept- Crumbaugh & Maholick, 1964; the Meaning in Life user may be unable to explain the criteria they use in Questionnaire, Steger et al., 2006; or the Perceived applying it. Hence, a theory of meaning in life developed Personal Meaning Scale, Wong, 1998) all require partici­ by asking people what makes life meaningful would be pants to define ‘meaning’ or ‘meaningful’ for them­ analogous to a theory of grammar developed by asking selves. This strategy of offloading definitional questions native speakers what makes a sentence grammatical. to participants can be quite useful for researchers. A better approach to understanding lay theories and

CONTACT Michael Prinzing [email protected] © 2021 Informa UK Limited, trading as Taylor & Francis Group 2 M. M. PRINZING ET AL. concepts involves, not asking for explicit definitions, but they think someone is (in their view) an accurate or inac­ observing participants’ use of their concepts. On this curate judge of the meaningfulness of their own life. approach, researchers present participants with short stor­ The six experiments reported here are organized ies or vignettes, and systematically vary different factors to into three pairs. Each pair includes an exploratory see what influences participants’ judgments. This metho­ study and a conceptual replication using a wider vari­ dology is ubiquitous in moral psychology (e.g., Clifford ety of vignettes. Studies 1a-b investigated the effects et al., 2015; Greene, 2001), and experimental philosophy of fulfillment and contribution on third-person attri­ (Knobe & Nichols, 2017; Phillips et al., 2017). butions of meaning and agent-perceived meaning. The present research investigates how lay people The subsequent studies investigated potential bound­ ordinarily think about what makes a life meaningful. To ary conditions – i.e., how much it matters whether this end, we considered philosophical theories to be an one’s activities are sensible (Studies 2a-b) or moral important source of insight. After all, philosophers have (Studies 3a-b). For each pair of experiments, we report spent much time reflecting on and attempting to clarify both sets of results before discussing. The exploratory this topic (for an overview, see Metz, 2013b). In the studies also included a within-subjects component academic philosophical literature, there are three main that is not reported here, though these data are avail­ kinds of theories: subjectivist, objectivist, and hybrid. able online, along with all other data, materials, and Broadly speaking, subjectivists claim that a life is mean­ the R scripts used for analyses: https://osf.io/cn5ed/. ingful when the person living it experiences certain The confirmatory studies were pre-registered on mental states – typically positive attitudes towards AsPredicted: https://aspredicted.org/35cb5.pdf. their life or activities (Frankfurt, 1998; Rowlands, 2015). For instance, Richard Taylor (1970) once argued that Study 1a: The Roles of Fulfillment and feeling fulfilled by one’s life is both necessary and suffi­ Contribution cient for it to be meaningful. Objectivists, on the other hand, claim that one’s life is meaningful when certain Our first research question was whether people consider subject-independent, evaluative facts obtain (Metz, a life meaningful when it is fulfilling but makes no con­ 2013a; Smuts, 2013). For instance, Frank Martela (2017) tribution to the world, or when it is unfulfilling but does argues that ‘lives are meaningful to the extent that they make a contribution. These are the cases in which the are able to contribute to something beyond them­ aforementioned philosophical theories have different selves . . . [i.e., insofar as they involve] promoting or implications. Each would agree that a life is meaningful realizing some intrinsic value that goes beyond the when it is both fulfilling and makes a contribution, and agent in question’ (pp. 233, 241). Objectivists typically that a life is not (or not especially) meaningful when it is argue that this value can be of various kinds: moral (e.g., unfulfilling and makes no contribution. The theories benefiting others), aesthetic (e.g., producing beautiful diverge in their predictions when only one of the two art), epistemic (e.g., making a scientific discovery), and conditions is met. We independently manipulated fulfill­ perhaps others. Hybrid views, meanwhile, claim that life ment and contribution and assessed attributions of is meaningful ‘when subjective attraction meets objec­ meaning and agent-perceived meaning. tive attractiveness’ (Wolf, 2010, p. 26). That is, one’s life is meaningful only when one possesses certain positive Method attitudes towards one’s activities, and those activities are objectively valuable (Evers et al., 2016; Wolf, 1997, Participants. Using Amazon’s Mechanical Turk, we 2010). Hence, philosophers typically claim that, to be recruited 402 adults from across the United States. meaningful, a life must meet a subjective condition, an Participants who responded to an item that read ‘This is objective condition, or both simultaneously. an attention check, please leave this question blank’ were We tested these claims against lay persons’ assessments excluded from analysis (n = 11). This left N = 391 partici­ using short vignettes that systematically varied subjective pants (MAge = 39.24, SDAge = 12.84; 43.2% women, 56.3% and objective factors in a series of 2 × 2 experimental men, <1% other or prefer not to say; 4.9% Asian, 7.4% Black designs. We assessed third-person attributions of meaning or African American, 4.3% Hispanic or Latinx, 76.7% White and agent-perceived meaning – that is, the extent to which or European American, 6.7% mixed, other, or prefer not participants believed that the agent in the vignette had to say). a meaningful life, and the extent to which participants Procedure and Measures. The vignettes in this study believed that the agent felt that their life was meaningful. described an agent named Teddy, who had This enabled us to explore not only the conditions under a comfortable life working in finance. Participants were which lay people think a life is meaningful, but also when randomized to one of four conditions in a 2 (Fulfillment: THE JOURNAL OF POSITIVE PSYCHOLOGY 3 fulfilled or unfulfilled) x 2 (Contribution: high or low) r = .57, p < .001. Thus, participants’ own assessments design. The fulfilled conditions included the following of meaning were related to – but clearly distinct description of the agent’s mental states: ‘Teddy loved from – their beliefs about the agent’s own assess­ working in finance. He was truly engaged with life, and ment of his life. his day-to-day activities left him with a sense of fulfill­ We used a 2 (unfulfilled, fulfilled) x 2 (low, high) ment.’ The unfulfilled conditions, by contrast, included factorial ANOVA to compare the effects of Fulfillment the following description: ‘Teddy really disliked working and Contribution on third-person attributions of mean­ in finance.He wasn’t truly engaged with life, and his day- ing (Figure 1, Panel A). This revealed main effects of 2 to-day activities left him feeling unfulfilled.’ In the high Fulfillment, F(1, 387) = 100.86, p < .001, ηp = .21 (95% contribution conditions, he was described as ‘a very CI: .15, .26), and Contribution, F(1, 387) = 135.60, p < .001, 2 generous philanthropist, giving away over half of his ηp = .26 (95% CI: .20, .32), as well as a significant inter­ annual earnings to charities.’ In the low contribution 2 action effect, F(1, 387) = 13.93, p < .001, ηp = .03 (95% CI: conditions, he was described as ‘a wine connoisseur, .01, .07). Tukey’s post-hoc tests found that all pairwise collecting and drinking very expensive vintage wines.’ comparisons were significant (ps < .001) except between After reading the vignette, participants used 7-point the fulfilled-low contribution and unfulfilled-high con­ Likert scales (1 = ‘Strongly disagree’, 2 = ‘Disagree’, tribution conditions (p = .379). The majority of partici­ 3 = ‘Slightly disagree’, 4 = ‘Neither agree nor disagree’, pants in the unfulfilled-high contribution condition 5 = ‘Slightly agree’, 6 = ‘Agree’, 7 = ‘Strongly agree’) to (87.9%) and fulfilled-low contribution condition (75.8%) respond to a series of statements: indicated ‘Slightly agree’ or higher to the statement that the agent lived a meaningful life. ● Teddy lived a meaningful life We then ran a similar ANOVA in which the depen­ ● Teddy felt that he was living a meaningful life dent variable was agent-perceived meaning (Figure 1, ● Teddy made a valuable contribution to the world Panel A). This revealed a very large main effect of ● Teddy felt fulfilled 2 Fulfillment, F(1, 387) = 367.99, p < .001, ηp = .49 (95% CI: .43, .54), a main effect of Contribution, F(1, 387) = These statements were mixed in with several additional 32.02, p < .001, and a small interaction effect, F(1, items, which concerned the attitudes Teddy held 387) = 8.38, p = .004, η2 = .02 (95% CI: .00, .05). towards his life, whether his life made for a good story, p Tukey’s post-hoc tests found that all pairwise compar­ and whether he lived a good life overall. These items are isons were significant (ps < .001), except between the beyond the focus of the present analyses but are fulfilled-low contribution and fulfilled-high contribu­ included in the online materials: https://osf.io/cn5ed/. tion conditions. Hence, when the agent in the vign­ ette felt fulfilled, participants thought that he Results considered his life meaningful regardless of whether he made a contribution. However, when the agent We first performed manipulation checks. Agreement was unfulfilled, they thought he considered his life with ‘Teddy felt fulfilled’ was significantly higher in the more meaningful only when he made a contribution. fulfilled conditions (M = 6.15, SD = .81) than unfulfilled Given the very large effect of Fulfillment on attri­ conditions (M = 3.12, SD = 1.54), t(388) = 24.23, p < .001, butions of agent-perceived meaning, we computed d = 2.45. Agreement with ‘Teddy made a valuable con­ the correlation between participants’ responses to tribution to the world’ was significantly higher in the ‘Teddy felt that he was living a meaningful life’ and high contribution conditions (M = 6.36, SD = .84) than ‘Teddy felt fulfilled’ across conditions. These two low contribution conditions (M = 4.05, SD = 1.55), t variables were highly correlated, r = .86, p < .001. (388) = 18.26, p < .001, d = 1.85. Note: ‘A-P Meaning’ indicates ‘agent-perceived mean­ Next, we tested whether third-person attributions ing.’ Error bars indicate 95% confidence intervals. of meaning differed from third-person attributions of agent-perceived meaning. Collapsing across condi­ tions, agreement with ‘Teddy lived a meaningful life’ Study 1b: Conceptual Replication (M = 5.21, SD = 1.51) differed significantly from agree­ ment with ‘Teddy felt that he was living a meaningful This study aimed to replicate the results of Study 1a life’ (M = 4.75, SD = 2.01), paired t(390) = 5.40 p with additional vignettes, and to address some of its < .001, d = .26. Moreover, the correlation between potential weaknesses. First, in Study 1a, the vignettes these third-person attributions of meaning and described a single agent, who was male and of rela­ agent-perceived meaning was not especially strong, tively high socioeconomic status. In this study, we 4 M. M. PRINZING ET AL.

Figure 1. Third-Person Attributions of Meaning and Agent-Perceived Meaning in Studies 1a–b. used three sets of vignettes with different agents. To address this issue, in Study 1b, each vignette Two of these three agents were female, each was mentioned only one activity that the agents were described as being either working-class or middle- engaged in. One agent, Ariana, was described as class, and each was given an ethnically ambiguous running a restaurant and eating by herself each name (Ariana, Marcus, and Naomi). Second, in Study night (low contribution) or as providing meals to 1a, the vignettes mentioned two parts of the agent’s local homeless people (high contribution). Another, life (his career in finance, and either his wine con­ Marcus, was described as playing on an amateur foot­ noisseurship or philanthropy). Participants might have ball team (low contribution) or as coaching a football believed the agent to have felt differently about his team for children with special needs (high contribu­ work versus non-work activities. For instance, he tion). The third, Naomi, was described as playing the might have been seen as feeling unfulfilled by cello in her garage (low contribution) or for thou­ a career in finance but fulfilled by charitable giving. sands of adoring fans (high contribution). THE JOURNAL OF POSITIVE PSYCHOLOGY 5

Method SD = 1.44), t(362.01) = 28.10, p < .001, d = 2.83. Participants. Using Amazon’s Mechanical Turk, we Agreement with ‘[Agent] made a valuable contribution recruited 411 adults from across the United States. to the world’ was higher in the high contribution condi­ Before they began, participants were alerted that there tions (M= 6.06, SD = 1.02) than the low contribution would be a comprehension check embedded in the sur­ conditions (M = 4.57, SD = 1.48), t(352.19) = 11.64, p < vey. This was a text-entry question that read: ‘Please .001, d = 1.17. summarize the life of the person you read about. You As in Study 1a, third-person attributions of mean­ only need to write a sentence or two. We just want to ing (i.e., agreement with ‘[Agent] lived a meaningful make sure you read carefully.’ Participants who wrote life’) were positively correlated with third-person attri­ nonsense, or obviously irrelevant or inaccurate informa­ butions of agent-perceived meaning (i.e., agreement tion (n= 5) were excluded from analysis. We also excluded with ‘[Agent] felt that [s]he was living a meaningful participants (n= 10) who spent less than 5 seconds read­ life’), r = .67, p < .001. However, they again differed ing the vignette (an impossibly fast time). This left N = 396 significantly, paired t(395) = 8.28, p < .001.

(MAge = 39.19, SDAge = 12.60; 57.8% women, 40.9% men, To determine the effects of Fulfillment and 1.3% other gender or prefer not to say; 8.3% Asian, 10.1% Contribution on third-person attributions of meaning Black or African American, 6.1% Hispanic or Latinx, 67.9% and agent-perceived meaning, we used the lmerTest White or European American 1.0% mixed race, 6.6% other package in R (Kuznetsova et al., 2017) to run mixed or prefer not to say). effects models (Figure 1, Panel B). Since Fulfillment Procedure and Measures. Participants in this study and Contribution were the factors of interest, they were randomized to one of three agents, and then to were treated as fixed factors. Since agent was simply one of four main conditions, parallel to those of Study a robustness check, it was treated as a random inter­ 1a. After reading the vignette, participants used 7-point cept. The first model, with third-person attributions of Likert scales (same as Study 1a) to respond to the follow­ meaning as the dependent variable, revealed signifi­ ing statements: cant main effects of Fulfillment, b = 1.95, t(389.99) = 11.45, p < .001, and Contribution, b = 1.09, t ● [Agent] lived a meaningful life (390.14) = 6.40, p < .001, with no significant interac­ ● [Agent] felt that [s]he was living a meaningful life tion, b = – .42, t(390.03) = – 1.75, p = .081. Across ● [Agent] made a valuable contribution to the world agents, the majority of participants in the unfulfilled- ● [Agent] felt fulfilled high contribution conditions (74.0%) and fulfilled-low ● [Agent’s] activities were sensible things to spend contribution conditions (90.0%) indicated ‘Slightly time on* agree’ or higher to the statement that the agent’s ● [Agent’s] activities were pointless* life was meaningful. ● [Agent’s] activities were valuable* In contrast to Study 1a, we found no difference ● [Agent] was a good person* between the high and low contribution versions of the unfulfilled conditions. This may reflect the fact that, as Finally, on a separate page of the survey, participants mentioned above, participants in Study 1a may have responded to the comprehension check (see above). seen the agent as feeling unfulfilled by a career in Note, the asterisked items above were not analyzed in finance but fulfilled by charitable giving. In this study, this study but were important for subsequent studies (2b we eliminated that potential confound by ensuring that and 3b). For consistency, we used the same set of items each vignette only mentioned one activity for each across the three confirmatory studies. This set also agent. included several items that were used for exploratory The second model, with third-person attributions of purposes. These are not discussed here and are available agent-perceived meaning as the dependent variable, with the rest of the online materials: https://osf.io/ revealed a significant main effect of Fulfillment only, cn5ed/. b = 3.50, t(392.00) = 21.86, p < .001. There was no significant effect of Contribution (p = .93), and no inter­ action (p = .23). As in Study 1a, responses to ‘[Agent] felt Results that [s]he was living a meaningful life’ and ‘[Agent] felt We first performed manipulation checks, collapsing fulfilled’ were highly correlated across conditions, r = .89, across agents. Agreement with ‘[Agent] felt fulfilled’ p < .001. was higher in the fulfilled conditions (M= 6.21, These results confirm our pre-registered hypotheses: SD = 1.07) than the unfulfilled conditions (M = 2.62, replicating Study 1a, we found main effects of both 6 M. M. PRINZING ET AL.

Fulfillmentand Contribution on third person attributions suggest that the two assessments are identical (rs = .57 of meaning in life. and .68). Second, while third-person attributions of meaning were affected by both fulfillment and contribu­ tion, third-person attributions of agent-perceived mean­ Discussion: Studies 1a–1b ing were affected only by fulfillment.1 Thus, when We sought to determine whether people consider a life people assess the meaningfulness of an agent’s life, the meaningful when it is fulfilling but makes no valuable data suggest that they do not simply try to determine contribution to the world, or when it is unfulfilling but that person’s own views on the question, and then defer does make a contribution (that is, when it meets to those views. Rather, it appears that they consider a subjective but not objective criterion, and vice versa). agents to be fallible judges of the meaningfulness of In two studies, we found effects of both fulfillment and their own lives. contribution on third-person attributions of meaning. Why might people think that agents are sometimes Hence, like hybrid theorists (Evers et al., 2016; Wolf, mistaken about the meaningfulness of their own lives? 1997, 2010), laypeople considered both factors to be One possibility is that agents are sometimes viewed as relevant. But, unlike hybrid theorists, the absence of an employing mistaken criteria of meaningfulness. For interaction effect suggests that they didn’t think that instance, some religious people might consider a life to meaning only arises when both factors are present. be meaningful only if it advances God’s plans. Instead, each was viewed as independently sufficient for Meanwhile some atheists might see this as mistaken, a meaningful life. In both studies, the majority of partici­ thinking that a life is meaningful only if it benefits pants in the mixed conditions (i.e., fulfilled-low contribu­ other people. (Of course, which criteria, if any, are the tion and unfulfilled-high contribution) indicated ‘Slightly ‘correct’ criteria of meaningfulness is a philosophical agree’ or higher to the statement that the agent had question, beyond the scope of the present work.) a meaningful life (percentages ranging from 74.0% to Another possibility is that, regardless of which criteria 90.0%). These results suggest that the majority of people of meaningfulness an agent employs, false beliefs might hold an implicit conception of meaningfulness that is lead agents to misapply their own criteria. For instance, inconsistent with the three theories that currently predo­ an agent might hold that their life is meaningful only if it minate the philosophical literature. Specifically, most par­ positively affects other people. That agent might spend ticipants thought that meaningfulness could be weeks painting a mural on the side of their home in the bestowed by either subjective states (i.e., feelings of fulfill­ belief that it will inspire joy in passersby. Those pas­ ment) or objective conditions (i.e., making a valuable sersby may, in fact, consider the mural an eyesore, and contribution). want it removed. In this case, according to the painter’s One possible explanation for these results is that our own criteria of meaningfulness, the project was not participants were epistemically deferential objectivists. meaningful because it did not positively affect others. That is, while they themselves thought that meaningful­ Yet the painter may mistakenly believe that it was mean­ ness depends solely on objective conditions, they also ingful because they believe that it positively affected deferred to the judgment of the person living the life. others. After all (the thinking may go), an agent typically holds Our results invite questions about boundary condi­ positive attitudes towards their life only when they tions on the effects of subjective and objective factors. believe that they are doing something valuable and For instance, while the agents in the low contribution worthwhile. Thus, knowing that someone holds such conditions didn’t make a positive difference in the lives positive attitudes, participants might have assumed of others, their activities were nevertheless sensible. that the agent had reason to believe that their activities Even if one doesn’t personally care about pursuits such were objectively valuable. Epistemically deferential indi­ as wine connoisseurship or cello playing, one can recog­ viduals might simply accede to that person’s judgment, nize them as worthwhile projects. In contrast, senseless even if they see no basis for it. activities – e.g., hand-copying War and Peace, or (like However, this explanation does not fit the data. First, Sisyphus from Greek mythology) endlessly rolling in both studies participants seemed to distinguish boulders up a mountain only for them to roll back between how meaningful the agents’ lives were, and down – have been considered paradigm examples of how meaningful those agents felt their lives were. Third- meaningless activities (Camus, 1969; Taylor, 1970; Wolf, person attributions of meaning differed significantly 2010). This suggests that feelings of fulfillment may be from third-person attributions of agent-perceived mean­ thought to add meaning to one’s life only when they ing. While these two assessments were positively corre­ result from engaging in an intelligible or sensible activ­ lated, those correlations were not high enough to ity. After all, an objective criterion of meaningfulness THE JOURNAL OF POSITIVE PSYCHOLOGY 7 need not claim that, for a life to be meaningful, it must sensible things to spend time on’ was significantly benefit others. Instead, it might be that one needs to do higher in the sensible conditions (M = 4.58, SD = 1.49) something that one has good reason to care about – than senseless conditions (M = 4.14, SD = 1.61), t(372) = something, for instance, with aesthetic or epistemic 2.78, p = .006, d = .29. Similarly, agreement with ‘Teddy’s value. In Studies 2a-b, we examined the effects of fulfill­ activities were valuable’ was also higher in the sensible ment and sensibility, in order to determine whether lay­ (M = 4.62, SD = 1.47) than senseless conditions (M = 4.01, people think that fulfillment derived from senseless SD = 1.64), t(373) = 3.80, p < .001, d = .39. activities can make a life meaningful. We used a 2 (unfulfilled, fulfilled) x 2 (senseless, sensible) ANOVA to compare the effects of Fulfillment and Sensibility on third-person attributions Study 2a: The Role of Sensibility of meaning across conditions (Figure 2, Panel A). This We hypothesized that fulfillment derived from senseless revealed a large main effect of Fulfillment, F(1, 372) = 2 activities would not be thought to make a life mean­ 100.22, p < .001, ηp = .21 (95% CI: .15, .27), no effect of ingful. We therefore independently manipulated fulfill­ Sensibility, F(1, 372) = .05, p = .821, and a marginally ment and sensibility and assessed participants’ significant interaction effect, F(1, 372) = 3.69, p = .055. attributions of meaning and agent-perceived meaning. Tukey’s post-hoc tests found significant (ps < .001) pairwise comparisons for all levels except across levels of sensibility for which fulfillment stayed constant. Method Though, the difference between the fulfilled- Participants. Using Amazon’s Mechanical Turk, we senseless (M = 5.73, SD = 1.25) and fulfilled-sensible recruited 400 adults from across the United States. (M = 5.21, SD = 1.33) conditions was marginally sig­ Participants who responded to an item that read ‘This nificant, p = .057. In the fulfilled-senseless conditions, is an attention check, please leave this question blank’ 92.7% of participants indicated ‘Slightly agree’ or were excluded from analysis (n = 24). This left N = 376 higher to the statement that the agent’s life was

(MAge = 37.24, SDAge = 11.69, 54.0% women, 45.2% men, meaningful. By contrast, only 37.6% of participants in <1% other gender or prefer not to say, 9.3% Asian or the unfulfilled-sensible conditions did the same. Asian American, 9.0% Black or African American, 5.3% A similar ANOVA, with third-person attributions of Hispanic or Latinx, 71.5% White or European American, agent-perceived meaning as the dependent variable, 3.5% mixed race, 1.4% other or prefer not to say). revealed an extremely large effect of Fulfillment, F(1, 2 Procedure and Measures. The experimental design 372) = 314.97, p < .001, ηp = .46, no effect of Sensibility, F and materials were nearly identical to Study 1a. Only, in (1, 372) = .17, p = .681, and no interaction, F(1, 372) = .001, this study, we crossed Fulfillment with an independently p = .979 (Figure 2, Panel A). Hence, as in Studies 1a-b, third- manipulated factor of Sensibility, which systematically person attributions of agent-perceived meaning were varied the sensibility of the agent’s projects. Wine con­ affected only by subjective factors. Also confirming the noisseurship was reused as the sensible project. For the results of the previous studies, responses to ‘Teddy felt senseless project, the agent, Teddy, was described as ‘an that he was living a meaningful life’ and ‘Teddy felt fulfilled’ avid collector of rubber bands. He didn’t use the bands were highly correlated across conditions, r = .85, p < .001. for anything. He just wanted to have them in his ever- Note: ‘A-P Meaning’ indicates ‘agent-perceived mean­ growing warehouse.’ The items and response scales ing.’ Error bars indicate 95% confidence intervals. were identical those used in Study 1a, only we included two new items for manipulation checks. Study 2b: Conceptual Replication ● Teddy’s activities were sensible things to spend This study aimed to replicate the results of Study 2a with time on a wider variety of vignettes. Since Study 2a found no ● Teddy’s activities were valuable effect of sensibility on third-person attributions of mean­ ing, we made two changes that might enable us to detect such an effect. First, we increased the sample size by Results approximately 50%, providing adequate statistical power We first performed manipulation checks. Agreement to detect even very small effects. Second, as in Study 1b, with ‘Teddy felt fulfilled’ was significantly higher in the we used three agents, and described them as engaging in fulfilled conditions (M = 5.95, SD = 1.00) than unfulfilled activities that we expected participants to find particularly conditions (M= 2.86, SD = 1.57), t(372) = 22.59, p < .001, senseless: digging for buried treasure, hand-copying the d = 2.36. Agreement with ‘Teddy’s activities were dictionary, and counting the bricks in local buildings. 8 M. M. PRINZING ET AL.

Figure 2. Third-Person Attributions of Meaning and Agent-Perceived Meaning in Studies 2a–b.

Method Hispanic or Latinx, 71.2% White or European Participants. Using Amazon’s Mechanical Turk, we American, 1.4% mixed or other race, 6.8% prefer not recruited 602 adults from across the United States. to say). After reading the vignettes and responding to the state­ Procedure and Measures. Participants were rando­ ments, participants were presented with the same com­ mized to one of three agents, and then to one of the prehension check as in Study 1b. We again excluded same four conditions as Study 2a. The descriptions of participants who failed this check (n = 3) or spent less the agents’ mental states in the fulfilled and unfulfilled than 5 seconds reading the vignette (n = 12). This left N = conditions were identical to those used in replication

587 participants (MAge = 41.61, SDAge = 14.23; 55.0% Study 1b. One agent, Ariana, was described as spend­ women, 43.9% men, 1.0% other gender; 9.0% Asian or ing time in her backyard, either building a treehouse Asian American, 7.7% Black or African American, 3.9% (sensible) or searching for buried treasure (senseless). THE JOURNAL OF POSITIVE PSYCHOLOGY 9

Another, Marcus, was described as correcting out­ means and the Satterthwaite method for estimating dated entries in the dictionary (sensible) or repeatedly degrees of freedom, found no significant differences hand-copying the dictionary (senseless). The third, between conditions differing in sensibility but not fulfill­ Naomi, was described as painting the buildings in ment. As in the previous studies, responses to ‘[Agent] her neighborhood (sensible) or counting the bricks in felt that [s]he was living a meaningful life’ and ‘[Agent] those buildings (senseless). To discourage participants felt fulfilled’ were highly correlated across conditions, r = from assuming that the agents in the senseless condi­ .91, p < .001. tions suffered from mental illness, each vignette Given that we found a main effect of Sensibility on described the agent as being an ordinary person in third person attributions of meaning in life, these results good health. We used the same measures as in contradict the results of Study 2a, and therefore our pre- Study 1b. registered hypothesis.

Results Discussion: Studies 2a–2b We first performed manipulation checks, collapsing Our aim in these two studies was to determine whether across agents. Agreement with ‘[Agent] felt fulfilled’ fulfillment is still thought to make a life meaningful was higher in the fulfilled conditions (M = 6.32, SD = when it is derived from senseless or unintelligible activ­ .91) than the unfulfilled conditions (M = 2.67, SD = ities. In Study 2a, we found a main effect of fulfillment 1.49), t(487.24) = 35.94, p < .001, d = 2.96. Agreement but no effect of sensibility on meaning ratings. with ‘[Agent’s] activities were sensible things to spend However, in Study 2b – with a larger sample and time on’ was higher in the sensible conditions (M = wider variety of vignettes – we found main effects of 4.90, SD = 1.38) than the senseless conditions (M = 3.09, both fulfillment and sensibility. This suggests that the SD = 1.59), t(578.62) = 14.75, p < .001, d = 1.21. sensibility of a person’s activities can affect third-person Similarly, agreement with ‘[Agent’s] activities were attributions of meaningfulness. In both studies, valuable’ was higher in the sensible conditions (M = a majority of participants in the fulfilled-senseless con­ 5.08, SD = 1.38) than the senseless conditions (M = 3.36, ditions indicated ‘Slightly agree’ or higher to the state­ SD = 1.64), t(572.32) = 13.70, p < .001, d = 1.13. ment that the agent’s life was meaningful (92.7% in Agreement with ‘[Agent’s] activities were pointless’ Study 2a, and 66.0% in Study 2b). Hence, it seems was lower in the sensible conditions (M = 2.78, SD = that most laypeople consider fulfillment to be sufficient 1.49) than the senseless conditions (M = 4.56, SD = to make a life meaningful – even if that fulfillment is 1.73), t(578.4) = – 13.36, p < .001, d = – 1.10. derived from senseless activities. In contrast, in the To determine the effects of Fulfillment and Sensibility unfulfilled-sensible conditions, only a minority of parti­ on third-person attributions of meaning and agent- cipants indicated ‘Slightly agree’ or higher to the state­ perceived meaning, we ran two mixed effects models, ment that the agent’s life was meaningful (37.6% in with Fulfillment and Sensibility as fixed factors, and Study 2a, and 36.3% in Study 2b). Hence, most lay­ Agent as a random intercept (Figure 2, Panel B). The people don’t consider sensibility sufficient to make an first model, with third-person attributions of meaning unfulfilling life meaningful. as the dependent variable, revealed significant main One explanation for these results is that participants effects of Fulfillment, b = 1.91, t(581.13) = 12.07, p < were thinking something like the following: ‘Hey, this .001, and Sensibility, b = .70, t(581.52) = 4.38, p < .001, person has found something that they love. It’s bizarre. and no significant interaction, b = .34, t(581.46) = 1.50, But they’re not hurting anyone.’ When one’s activities p = .134. Across agents, in the fulfilled-senseless condi­ don’t affect other people, their meaningfulness may be tions, 66.0% of participants indicated ‘Slightly agree’ or thought to depend primarily on one’s own mental higher to the statement that the agent’s life was mean­ states. After all, ‘find your passion’ is a common recom­ ingful. By contrast, only 36.3% of participants in the mendation for those seeking meaning in their lives. And unfulfilled-sensible conditions did the same. many of the things that people are passionate about can The second model, with third-person attributions of appear strange – even senseless – to others. agent-perceived meaning as the dependent variable, The qualification about not hurting others seems revealed a significant effect of Fulfillment, b = 3.28, t important, however. Meaningfulness is supposed to be (581.03) = 24.05, p < .001, no effect of Sensibility, b = – an aspect of the good life (King & Napa, 1998; Wolf, .13, t(581.25) = – .93, p = .351, and a marginally signifi­ 1997), and may therefore require a certain minimum of cant interaction, b = .38, t(581.22) = 1.96, p = .051. moral goodness. For instance, it seems strange to think However, post-hoc comparisons, using least squares that going to a racist rally or punching strangers could 10 M. M. PRINZING ET AL. make a person’s life more meaningful just because that across conditions (Figure 3, Panel A). This revealed person happens to find it fulfilling. In Studies 3a-b, we a medium-sized main effect of Fulfillment, F(1, 389) = 2 tested whether fulfilling but evil activities are thought to 23.51, p < .001, ηp = .06 (95% CI: .03, .10), a very large 2 add meaning to a life. main effect of Morality, F(1, 389) = 189.87, p < .001, ηp = .33 (95% CI: .27, .38), and no significant interaction, F Study 3a: The Role of Morality (1, 389) = .02, p = .902. In the fulfilled-bad conditions, only 36.4% of participants indicated ‘Slightly agree’ or We hypothesized that fulfillment derived from morally higher to the statement that the agent’s life was bad activities would not be thought to make a life more meaningful. By contrast, in the unfulfilled-good condi­ meaningful. Therefore, we independently manipulated tions, 82.0% of participants agreed that the agent’s life fulfillment and morality and assessed participants’ attri­ was meaningful. butions of meaning and agent-perceived meaning. A similar ANOVA, with third-person attributions of agent-perceived meaning as the dependent variable, Method revealed a very large main effect of Fulfillment, F(1, 2 388) = 158.60, p < .001, ηp = .29 (95% CI: .23, .35), no Participants. Using Amazon’s Mechanical Turk, we main effect of Morality, F(1, 388) = .78, p = .378, and recruited 401 adults from across the United States. 2 a small interaction, F(1, 388) = 4.18, p = .041, ηp = .01 Participants who responded to an item that read ‘This (95% CI: .00, .03). Tukey’s post-hoc tests found that all is an attention check, please leave this question blank’ pairwise comparisons between conditions differing in were excluded from analysis (n = 7). Two additional Fulfillment were significant (ps < .001), but no compar­ participants did not respond to ‘Teddy lived isons between conditions differing in Morality were sig­ a meaningful life’ and were also excluded from analysis. nificant while Fulfillment stayed constant. As in the This left N = 393 participants (MAge = 40.74, SDAge = previous studies, responses to ‘Teddy felt that he was 12.98, 59.7% women, 39.2% men, <1% other gender or living a meaningful life’ and ‘Teddy felt fulfilled’ were prefer not to say; 3.8% Asian or Asian American, 7.1% highly correlated across conditions, r = .78, p < .001. Black or African American, 3.8% Hispanic or Latinx, 76.7% Note: ‘A-P Meaning’ indicates ‘agent-perceived mean­ White or European American, 6.8% mixed race, 1.7% ing.’ Error bars indicate 95% confidence intervals. other or prefer not to say). Procedure and Measures. The experimental design and measures were nearly identical to Studies 1a and 2a. Study 3b: Conceptual Replication Only, in this study we crossed Fulfillment with an inde­ pendently manipulated factor of Morality, which system­ This study aimed to replicate the results of Study 3a with atically varied the morality of the agent’s activities. a wider variety of vignettes. Philanthropy was reused (from Study 1a) as the good activity. For the morally bad activity, the agent was Method described as ‘a religious extremist, giving away over half of his annual earnings to terrorist organizations.’ In Participants. Using Amazon’s Mechanical Turk, we this experiment, we also included a morality manipula­ recruited 409 adults from across the United States. tion check: ‘Teddy was a good person’. Participants were presented with the same comprehen­ sion check as in Studies 1b and 2b. We excluded partici­ pants (n = 4) who failed this check. We also excluded Results participants who spent less than 5 seconds reading the

We first performed manipulation checks. Agreement vignette (n = 8). This left N = 397 participants (MAge = with ‘Teddy felt fulfilled’ was significantly higher in the 40.5, SDAge = 13.94; 58.9% women, 41.1% men; 6.8% fulfilled conditions (M = 5.84, SD = 1.13) than unfulfilled Asian or Asian American, 8.3% Black or African conditions (M = 3.11, SD = 1.77), t(391) = 18.23, p < .001, American, 4.0% Hispanic or Latinx, 73.6% White or d = 1.84. Agreement with ‘Teddy was a good person’ was European American, 1.3% mixed race, 6.0% other or significantly higher in the morally good conditions (M = prefer not to say). 6.24, SD = 1.03) than bad conditions (M = 1.86, SD = Procedure and Measures. The procedure and mea­ 1.46), t(390) = 34.30, p < .001, d = 3.47. sures used in this study were identical to those used We used a 2 (unfulfilled, fulfilled) x 2 (bad, good) in Studies 1b and 2b. Participants were randomized ANOVA to compare the effects of Fulfillment and to one of three agents, and then to one of the four Morality on third-person attributions of meaning conditions of Study 3a. The descriptions of the THE JOURNAL OF POSITIVE PSYCHOLOGY 11

Figure 3. Third-Person Attributions of Meaning and Agent-Perceived Meaning in Studies 3a–b.

agent’s mental states in the fulfilled and unfulfilled Results conditions were identical to those used in Studies 1b We first performed manipulation checks, collapsing and 2b. To manipulate morality, we held the type of across agents. Agreement with ‘[Agent] felt fulfilled’ activity constant while varying its moral valence. One was higher in the fulfilled conditions (M = 6.10, SD = agent was described as working for a pharmaceutical .97) than unfulfilled conditions (M = 2.53, SD = 1.51), t company, and either reducing the cost of overpriced (342.35) = 27.98, p < .001, d = 2.79. Agreement with drugs (good) or as driving patients into debt by ‘[Agent’s] was a good person’ was higher in the morally increasing prices (bad). Another agent was described good conditions (M = 6.24, SD = .84) than bad conditions as working with at-risk teenagers, and either helping (M = 2.46, SD = 1.35), t(331.51) = 33.40, p < .001, d = 3.35. them to stay motivated at school and away from To determine the effects of Fulfillment and Morality drugs (good), or as encouraging them to use drugs on third-person attributions of meaning and agent- (bad). The third agent was described as either volun­ perceived meaning, we ran mixed effects models, with teering for a homeless outreach group (good), or as Fulfillment and Morality as fixed factors, and Agent as harassing the local homeless people (bad). 12 M. M. PRINZING ET AL. a random intercept (Figure 3, Panel B). The first model, can make an agent’s life meaningful. These studies also with third-person attributions of meaning as the depen­ found that, according to lay assessments, doing some­ dent variable, revealed significant effects of Fulfillment, thing that has value for others can also make an agent’s b = .70, t(393.00) = 3.91, p < .001, and Morality, b = 3.19, t life meaningful. These findings conflict with the predo­ (393.00) = 17.94, p < .001, and no significant interaction, minant philosophical theories of meaning in life. These b = .42, t(393.00) = 1.66, p = .097. Across agents, in the theories posit an exclusive role for either positive mental fulfilled-bad conditions, only 18.8% of participants indi­ states (subjectivist theories) or objective states of an cated ‘Slightly agree’ or higher to the statement that the agent’s life (objectivist theories), or they require that agent’s life was meaningful. By contrast, in the unful­ both criteria be met (hybrid theories). In contrast, we filled-good conditions, 82.9% of participants did the found that laypeople think an agent’s life is meaningful same. when either criterion is met. This indicates that the The second model, with third-person attributions of ordinary concept of a meaningful life does not fit neatly agent-perceived meaning as the dependent variable, with these three philosophical theories. Instead, they revealed a main effect of Fulfillment, b = 3.30, t seem to be captured by what we will call the indepen­ (392.00) = 17.83, p < .001, but no significant effect of dent-additive theory: subjective factors (positive mental Morality, b = .15, t(392.00) = .82, p = .413, and no inter­ states like fulfillment) and objective factors (like contri­ action, b = .36, t(392.00) = 1.37, p = .172. As in the bution, sensibility, and morality) each affect the mean­ previous studies, responses to ‘[Agent] felt that [s]he ingfulness of an agent’s life, and their effects are both was living a meaningful life’ and ‘[Agent] felt fulfilled’ independent and additive. were highly correlated across conditions, r = .87, We investigated the roles of sensibility and morality p < .001. as plausible boundary conditions for lay attributions of These results confirm our pre-registered hypotheses: meaningfulness. For sensibility, we saw somewhat replicating Study 3a, we found main effects of both mixed results. Study 2a found no evidence that a life Fulfillment and Morality on third person attributions of characterized by sensible activities (wine connoisseur­ meaning in life. ship) was seen as more meaningful than a life charac­ terized by senseless activities (rubber band collecting). However, Study 2b, with a larger sample and wider Discussion: Studies 3a–3b variety of vignettes, did find such an effect. Are agents who derive fulfillmentfrom immoral activities Nevertheless, in both studies, fulfilling lives were seen still viewed as leading meaningful lives? In both studies, as more meaningful than unfulfilling ones – regardless we observed main effects of Fulfillment and Morality, of whether that fulfillment was derived from sensible with no interaction. Thus, contrary to our original pre­ or senseless activities. Hence, on the ordinary concept, diction, the incremental effect of fulfillment on attribu­ sensibility contributes to meaningfulness, though not ted meaning was not eliminated by immorality. With as much as fulfillment does. Moreover, in alignment that said, the effect of Morality was far larger than that with the independent-additive theory, fulfillment 2 of Fulfillment( ηp = .33 versus .06 in Study 3a and b = 3.17 maintains its additive effect, independently of versus .70 in Study 3b). In both studies, a majority of sensibility. participants in the unfulfilled-good conditions indicated Regarding morality, Studies 3a-b found that morally ‘Slightly agree’ or higher to the statement that the good lives were viewed as much more meaningful than agent’s life was meaningful (82.0% in Study 3a, and morally bad ones. In fact, morally bad agents were not 82.9% in Study 3b). By contrast, only a minority of parti­ thought to live meaningful lives, even if those agents felt cipants in the fulfilled-bad conditions did the same very fulfilled. In contrast, morally good agents were seen (36.4% in Study 3a, and 18.8% in Study 3b). Hence, it as having meaningful lives even if they didn’t feel ful­ seems that most people think that moral goodness is filled. Nevertheless, though the effect of morality was sufficient to make an unfulfilling life meaningful, but larger than that of fulfillment, participants still thought that fulfillment is not sufficient to make an immoral life that a fulfilled, immoral agent was living more mean­ meaningful. ingfully than an unfulfilled, immoral agent. Supporting the independent-additive theory, the additive effect of fulfillment was independent of morality. General Discussion In short, we identified four factors (fulfillment, con­ What, according to the ordinary concept, makes a life tribution, sensibility, and morality) that seem to have meaningful? Studies 1a-b found that laypeople think independent, additive effects on third-person attribu­ positive mental states (interest, engagement, fulfillment) tions of meaningfulness. There may well be more such THE JOURNAL OF POSITIVE PSYCHOLOGY 13 factors. But the evidence from these six experiments perceived meaning in life is a ‘tripartite construct’ supports a model of third-person meaningfulness judg­ (George & Park, 2016; Martela & Steger, 2016). ments that – in contrast to subjectivist, objectivist, and However, more recent longitudinal research has found hybrid theories – emphasizes independent and additive that, of these three, only perceptions of mattering pre­ factors that contribute to the meaning in a person’s life. dict first-person perceptions of meaning (Costin & We have called such a model the ‘independent-additive Vignoles, 2020). This finding suggests that the lay con­ theory’. cept of a meaningful life might simply be the concept of These results, which concern third-person attribu­ a life that matters. If so, subjective and objective factors tions, align with past research that has identified numer­ could each generate meaningfulness because they are ous sources of first-person perceptions of meaning in different ways in which a life can matter. For instance, life. Positive mood inductions have been found to when something matters to a person, it is something increase first-personally perceived meaning (Hicks & they care about – something that elicits distinctive emo­ King, 2009; King et al., 2006). Other research has found tional responses like fulfillment. Hence, an agent’s feel­ that people report greater feelings of meaningfulness on ings of fulfillment are a sign that their life is filled with days when they report having achieved something activities that matter to them – even if, perhaps, they (Machell et al., 2015), and experiments have found that don’t matter to anyone else. Similarly, morally good prosocial activities increase first-personal perceptions of activities tend to matter to many people (most notably meaning (Klein, 2017; Tongeren et al., 2016). Hence, it the beneficiaries) and may do so even if they do not seems that both first-person as well as third-person matter to the agent in question. Hence, an agent’s moral attributions of meaningfulness are affected by subjec­ goodness is a sign their life is filled with activities that tive states and objective conditions (e.g., benefiting matter to others – even if, perhaps, those activities don’t others). matter to them. In all six experiments, the fulfillment manipulation Alternatively, the concept of a meaningful life may be produced a very large effect on third-person attributions one of an authentic life – i.e., one lived in accordance of agent-perceived meaning. We also found that with one’s true self. In past work, researchers have ana­ responses to the fulfillment and agent-perceived mean­ lyzed essays about times when people felt that their lives ing items were highly correlated (rs ranging from .78 to were particularly meaningful or meaningless (Debats .91). This suggests the possibility that ‘feeling fulfilled’ et al., 1995). This content analysis found that the essays and ‘feeling like life is meaningful’ may be more or less about meaningful times frequently mention feelings of synonymous. This is intriguing, given that none of the authenticity (i.e., of being true to oneself), while essays standard self-report measures of meaning in life (e.g., about meaningless times frequently mention feelings of the Purpose in Life Test, Crumbaugh & Maholick, 1964; alienation and inauthenticity. Moreover, the ease with the Meaning in Life Questionnaire, Steger et al., 2006; or which people can describe their perceived true self pre­ the Perceived Personal Meaning Scale, Wong, 1998) dicts first-person perceptions of meaning in life, over include any items that use the term ‘fulfilled’. Hence, and above their mood and self-esteem (Schlegel et al., our results suggest that it may be valuable to add 2011). If meaningfulness derives from expressing one’s items about feelings of fulfillment to self-report mea­ true self, this could explain why fulfillment influences sures of meaning in life. attributions of meaning. Feelings of fulfillment can be taken as evidence that one is living in accord with one’s true self. It could also explain why contribution and Why do Subjective and Objective Factors morality generate meaning. Previous studies find that Independently Affect Meaningfulness? people tend to assume that true selves – both their own Why do subjective and objective factors independently and others’ – are morally good (De Freitas et al., 2017; affect third-person attributions of meaningfulness? We Newman et al., 2014). Even when someone behaves suggest that two explanations hold promise: meaning­ badly, people tend to assume that, deep down, they fulness is mattering, and meaningfulness is authenticity. are good. This seems to be the default assumption Past research has suggested that first-person percep­ across cultures, even among misanthropes (De Freitas tions of meaning (i.e., the feeling that one’s own life is et al., 2018), and even when considering the true self of meaningful) are based on the feeling that one’s life is a potentially threatening outgroup member (De Freitas coherent or makes sense, that it has a purpose, and that & Cikara, 2018). If a person’s true self is good, then it matters. This has led to proposals that first-person people are likely to view prosocial activities (like 14 M. M. PRINZING ET AL. charitable donations) as highly authentic, assuming no et al., 2001) has found that third-person assessments of ulterior motive is evident (De Freitas et al., 2019). For the the overall goodness of a life are subject to the ‘peak- same reason, evil deeds may be viewed as conflicting end rule’ (Fredrickson & Kahneman, 1993). Future studies with this true (presumably good) self, reducing attribu­ may examine how the shape or progression of an tions of meaningfulness. agent’s life affects lay assessments of its meaningfulness. A particularly strong test of the claim that meaning­ fulness is authenticity would be if people view immoral agents as leading meaningful lives when they are Conclusion described as having evil true selves. Although people Across six experiments (total N = 2,539) we used vign­ do not tend to attribute evil true selves by default, ettes to investigate the ordinary concept of previous work findsthat if an agent’s true self is explicitly a meaningful life. We found that lay people attribute described as evil (versus good), then people think that greater meaning to lives that are experienced as ful­ the agent is happier when engaged in evil activities filling (Studies 1a–3b), as well as lives that are charac­ (Newman et al., 2015). Perhaps a similar effect occurs terized by activities that make a valuable contribution for ascriptions of meaningfulness. Future work might (Studies 1a–b), are sensible (Studies 2a–b), and morally explore this surprising possibility. good (Studies 3a–b). Overall, these results suggest that lay people hold an implicit theory of meaning in life Limitations and Scope of the Effect unlike the three theories that dominate the philosophi­ cal literature. In philosophy, subjectivist theories posit All of our participants were recruited on Amazon’s that positive mental states are necessary and sufficient Mechanical Turk. Workers on this platform have been for a meaningful life. Objectivist theories posit that found to be more demographically and geographically objective value (e.g., doing good for others) is neces­ diverse than typical college undergraduates (Difallah sary and sufficient for a meaningful life. Hybrid theories et al., 2018). However, it is possible that people who posit that positive subjective states and objective value spend time completing online surveys have somewhat are individually necessary and jointly sufficient for different intuitions about meaningfulness than the a meaningful life. We introduce here a new theory population at large. Individuals engaged in paradigma­ that appears to fit the intuitions of lay people: the tically meaningful activities (e.g., feeding the homeless independent-additive theory, according to which sub­ or conducting cancer research) are likely to be under­ jective and objective factors each exert independent represented on this platform. Future research could and additive effects on the meaningfulness of an compare the results obtained from samples differing in agent’s life. career or recreational pursuits. Another limitation is the use of a ‘Strongly Disagree’ – ‘Strongly Agree’ response scale, which assesses partici­ Note pants’ confidence that someone’s life was meaningful, but does not readily allow for comparisons of magni­ 1. The correlations between third-person attributions of agent-perceived meaning and fulfillment were perhaps tude – i.e., how meaningful a life is. Future studies may high enough to indicate identity (rs = 86, .89). We take benefit from using response scales that range from ‘Not up this possibility in the General Discussion. at all meaningful’ to ‘Extremely meaningful.’ Finally, an additional limitation of these studies is that they paint a somewhat static picture of meaning in life. Acknowledgments Some philosophers have argued that a life’s meaning The authors thank Joshua Knobe for his supportiveness, and depends in part on its ‘shape’ – i.e., its narrative struc­ helpful comments on prior drafts. ture, or temporal dynamics (Dorsey, 2015; Glasgow, 2013). Similarly, psychological research has found that the meaning people see in particular events in their lives Funding depends largely on how they conceive of the story of their life (Bauer et al., 2008; McLean & Pasupathi, 2011). It This research was supported by a grant from the Institute for may be that, for instance, that people attribute greater Humane Studies awarded to the lead author.; meaning to a life that starts unhappily but ends well – what personality psychologist Dan McAdams and collea­ ORCID gues (2001) call a ‘redemptive narrative’ – than to one that starts happily but ends badly. Past research (Diener Michael Prinzing http://orcid.org/0000-0002-7879-7131 THE JOURNAL OF POSITIVE PSYCHOLOGY 15

References Dorsey, D. (2015). The significance of a life’s shape. Ethics, 125 (2), 303–330. doi:10.1086/678373 Bauer, J., McAdams, D., & Pals, J. (2008). Narrative identity and Evers, D., Smeden, G. E., & Van. (2016). Meaning in life: In eudaimonic well-being. Journal of Happiness Studies, 9(1), defense of the hybrid view. The Southern Journal of 81–104. doi:10.1007/s10902-006-9021-6 Philosophy, 54(3), 355–371. doi:10.1111/sjp.12179 Baumeister, R. F., & Landau, M. J. (2018). Finding the meaning Frankfurt, H. (1998). Necessity, Volition, and Love . Cambridge of meaning: Emerging insights on four grand questions. University Press. https://www.cambridge.org/core/books/ Review of General Psychology, 22(1), 1–10. doi:10.1037/ necessity-volition-and-love/1E2F7EBB70738D2A8FC54DAD gpr0000145 13E81B6C Camus, A. (1969). The Myth of Sisyphus and Other Essays (J. Frankl, V. E. (1971). Man’s Search For Meaning. Ratna Sagar. O’Brien, Trans.). Knopf. Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in Chen, Z., Poon, K.-T., DeWall, C. N., & Jiang, T. (2020). Life lacks retrospective evaluations of affective episodes. Journal of meaning without acceptance: Ostracism triggers suicidal Personality and Social Psychology, 65(1), 45–55. doi:10.1037/ thoughts. Journal of Personality and Social Psychology, 119 0022-3514.65.1.45 (6), 1423–1443. doi:10.1037/pspi0000238 George, L. S., & Park, C. L. (2016). Meaning in life as comprehen­ Chomsky, N. (1965). Aspects of a Theory of Syntax. MIT Press. sion, purpose, and mattering: Toward integration and new Clifford, S., Iyengar, V., Cabeza, R., & Sinnott-Armstrong, W. research questions. Review of General Psychology, 20(3), (2015). Moral foundations vignettes: A standardized stimu­ 205–220. doi:10.1037/gpr0000077 lus database of scenarios based on moral foundations Glasgow, J. (2013). The shape of a life and the value of loss and theory. Behavior Research Methods, 47(4), 1178–1198. gain. Philosophical Studies, 162(3), 665–682. doi:10.1007/ doi:10.3758/s13428-014-0551-2 s11098-011-9788-0 Costin, V., & Vignoles, V. L. (2020). Meaning is about mattering: Greene, J. D. (2001). An fMRI Investigation of Emotional Evaluating coherence, purpose, and existential mattering as Engagement in Moral Judgment. Science, 293(5537), precursors of meaning in life judgments. Journal of 2105–2108. doi:10.1126/science.1062872 Personality and Social Psychology, 118(4), 864–884. Hicks, J. A., & King, L. A. (2009). Meaning in life as a subjective doi:10.1037/pspp0000225 judgment and a lived experience. Social and Personality Crumbaugh, J. C., & Maholick, L. T. (1964). An experimental Psychology Compass, 3(4), 638–653. doi:10.1111/j.1751- study in existentialism: The psychometric approach to 9004.2009.00193.x Frankl’s concept of noogenic neurosis. Journal of Clinical Kim, J., Kang, P., & Choi, I. (2014). Pleasure now, meaning later: Psychology, 20(2), 200–207. doi:10.1002/1097-4679(196404) Temporal dynamics between pleasure and meaning. Journal 20:2<200::AID-JCLP2270200203>3.0.CO;2-U of Experimental Social Psychology, 55, 262–270. doi:10.1016/j. De Freitas, J., & Cikara, M. (2018). Deep down my enemy is jesp.2014.07.018 good: Thinking about the true self reduces intergroup bias. King, L. A., & Hicks, J. A. The Science of Meaning in Life. (2021). Journal of Experimental Social Psychology, 74, 307–316. Annual Review of Psychology, 72(1), 561–584. annurev-psych doi:10.1016/j.jesp.2017.10.006 -072420-122921. 10.1146/annurev-psych-072420-122921 De Freitas, J., Cikara, M., Grossmann, I., & Schlegel, R. (2017). King, L. A., Hicks, J. A., Krull, J. L., & Del Gaiso, A. K. (2006). Origins of the belief in good true selves. Trends in Cognitive Positive affect and the experience of meaning in life. Journal Sciences, 21(9), 634–636. doi:10.1016/j.tics.2017.05.009 of Personality and Social Psychology, 90(1), 179–196. De Freitas, J., DeScioli, P., Thomas, K. A., & Pinker, S. (2019). doi:10.1037/0022-3514.90.1.179 Maimonides’ ladder: States of mutual knowledge and the King, L. A., & Napa, C. K. (1998). What makes a life good?. perception of charitability. Journal of Experimental Journal of Personality and Social Psychology, 75(1), 156–165. Psychology. General, 148(1), 158–173. doi:10.1037/ doi:10.1037/0022-3514.75.1.156 xge0000507 Klein, N. (2017). Prosocial behavior increases perceptions of De Freitas, J., Sarkissian, H., Newman, G. E., Grossmann, I., meaning in life. The Journal of Positive Psychology, 12(4), Brigard, F. D., Luco, A., & Knobe, J. (2018). Consistent belief 354–361. doi:10.1080/17439760.2016.1209541 in a good true self in misanthropes and three interdepen­ Knobe, J., & Nichols, S. (2017). Experimental philosophy. In dent cultures. Cognitive Science, 42(S1), 134–160. E. N. Zalta. (Ed.), The Stanford Encyclopedia of Philosophy. doi:10.1111/cogs.12505 Metaphysics Research Lab. https://plato.stanford.edu/ Debats, D. L., Drost, J., & Hansen, P. (1995). Experiences of archives/win2017/entries/experimental-philosophy/ meaning in life: A combined qualitative and Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). quantitative approach. British Journal of Psychology; lmerTest Package: Tests in Linear Mixed Effects Models. London, Etc., 86(3), 359–375. doi:10.1111/j.2044- Journal of Statistical Software, 82(13), 1–26. doi:10.18637/ 8295.1995.tb02758.x jss.v082.i13 Diener, E., Wirtz, D., & Oishi, S. (2001). End effects of rated life Machell, K. A., Kashdan, T. B., Short, J. L., & Nezlek, J. B. (2015). quality: The James Dean effect. Psychological Science, 12(2), Relationships between meaning in life, social and achieve­ 124–128. doi:10.1111/1467-9280.00321 ment events, and positive and negative affect in daily life. Difallah, D., Filatova, E., & Ipeirotis, P. (2018). Demographics and Journal of Personality, 83(3), 287–298. doi:10.1111/ Dynamics of Mechanical Turk Workers. Proceedings of the jopy.12103 Eleventh ACM International Conference on Web Search and Martela, F. (2017). Meaningfulness as contribution. The Data Mining - WSDM ’18, Marina Del Rey, CA, 135–143. Southern Journal of Philosophy, 55(2), 232–256. doi:10.1111/ 10.1145/3159652.3159661 sjp.12217 16 M. M. PRINZING ET AL.

Martela, F., & Steger, M. F. (2016). The three meanings of mean­ Schnell, T., Gerstner, R., & Krampe, H. (2018). Crisis of mean­ ing in life: Distinguishing coherence, purpose, and ing predicts suicidality in youth independently of significance. The Journal of Positive Psychology, 11(5), depression. Crisis: The Journal of Crisis Intervention and 531–545. doi:10.1080/17439760.2015.1137623 Suicide Prevention, 39(4), 294–303. doi:10.1027/0227- McAdams, D. P., Reynolds, J., Lewis, M., Patten, A. H., & 5910/a000503 Bowman, P. J. (2001). When bad things turn good and Smuts, A. (2013). The good cause account of meaning in life. good things turn bad: Sequences of redemption and con­ Southern Journal of Philosophy, 51(4), 536–562. doi:10.1111/ tamination in life narrative and their relation to psychosocial sjp.12044 adaptation in midlife adults and in students. Personality & Steger, M. F. (2017). Meaning in life and wellbeing. In M. Slade, Social Psychology Bulletin, 27(4), 474–485. doi:10.1177/ L. Oades, & A. Jarden (Eds.), Wellbeing, Recovery and Mental 0146167201274008 Health (pp. 75–85). Cambridge University Press. McLean, K. C., & Pasupathi, M. (2011). Old, New, Borrowed, Steger, M. F., Frazier, P., Oishi, S., & Kaler, M. (2006). The mean­ Blue? The Emergence and Retention of Personal Meaning ing in life questionnaire: Assessing the presence of and in Autobiographical Storytelling: Narrative Identity. Journal search for meaning in life. Journal of Counseling of Personality, 79(1), 135–164. doi:10.1111/j.1467- Psychology, 53(1), 80–93. doi:10.1037/0022-0167.53.1.80 6494.2010.00676.x Steger, M. F., Shin, J. Y., Shim, Y., & Fitch-Martin, A. (2013). Is Metz, T. (2013a). Meaning in Life: An Analytic Study. Oxford meaning in life a flagship indicator of well-being?. In University Press. A. S. Waterman (Ed.), The Best Within Us: Positive Psychology Metz, T. (2013b). The meaning of life. In E. N. Zalta. (Ed.), Perspectives on Eudaimonia (pp. 159–182). American Stanford Encyclopedia of Philosophy. Metaphysics Research Psychological Association. Lab. https://plato.stanford.edu/archives/sum2013/entries/ Taylor, R. (1970). Good and Evil. Prometheus Books. life-meaning/ Tongeren, D. R. V., Green, J. D., Davis, D. E., Hook, J. N., & Newman, G. E., Bloom, P., & Knobe, J. (2014). Value Judgments Hulsey, T. L. (2016). Prosociality enhances meaning in life. and the True Self. Personality & Social Psychology Bulletin, 40 The Journal of Positive Psychology, 11(3), 225–236. (2), 203–216. doi:10.1177/0146167213508791 doi:10.1080/17439760.2015.1048814 Newman, G. E., Freitas, J. D., & Knobe, J. (2015). Beliefs about Wilson, T. D., & Dunn, E. W. (2004). Self-Knowledge: Its Limits, the true self explain asymmetries based on moral judgment. Value, and Potential for Improvement. Annual Review of Cognitive Science, 39(1), 96–125. doi:10.1111/cogs.12134 Psychology, 55(1), 493–518. doi:10.1146/annurev. Nisbett, R., & Wilson, T. D. (1977). Telling more than we can psych.55.090902.141954 know: Verbal reports on mental processes.. Psychological Wolf, S. (1997). Happiness and meaning: Two aspects of the Review, 84(3), 231–259. doi:10.1037/0033-295X.84.3.231 good life. Social Philosophy & Policy, 14(1), 207. doi:10.1017/ Phillips, J., De Freitas, J., Mott, C., Gruber, J., & Knobe, J. (2017). S0265052500001734 True happiness: The role of morality in the folk concept of Wolf, S. (2010). Meaning in Life and Why It Matters. Princeton happiness. Journal of Experimental Psychology. General, 146 University Press. (2), 165–181. doi:10.1037/xge0000252 Wong, P. T. (1998). Implicit theories of meaningful life and Rowlands, M. (2015). The immortal, the intrinsic and the quasi the development of the personal meaning profile. In meaning of life. The Journal of Ethics, 19(3–4), 379–408. P. T. Wong & P. S. Fry (Eds.), The Human Quest for doi:10.1007/s10892-015-9212-7 Meaning: A Handbook of Psychological Research and Schlegel, R. J., Hicks, J. A., King, L. A., & Arndt, J. (2011). Feeling Clinical Applications (pp. 111–140). Lawrence Erlbaum like you know who you are: Perceived true self-knowledge Associates Publishers. and meaning in life. Personality & Social Psychology Bulletin, Wong, P. T. (2013). The Human Quest for Meaning: Theories, 37(6), 745–756. doi:10.1177/0146167211400424 Research, and Applications. Routledge.