Introduction 引言

Individuals tend to value their own possessions more favorably than those they do not own, a bias that has been termed as mere ownership effect [1], [2]. This effect occurs even when the actual possessions are imaged, not physically present [3]. The ownership effect has been linked to the self-enhancement motivation in which individuals overvalue an object owned by or associated with self in order to improve their self-image [1].
个人倾向于比非拥有的物品更高度评价他们自己的财产，这种偏见被称为纯粹的拥有效应[1]、[2]。即使实际的财产被想象而不是实际存在，这种效应也会发生[3]。拥有效应与自我提升动机有关，其中个体过度评价由自己拥有或与自己关联的物体，以改善自我形象[1]。

Recent studies suggested that the perception of ownership modulates other-regarding behavior in economic decision-making [4], [5], [6]. Oxoby and Spraggon [6] asked participants to play a dictator game (DG; [7]) in which the allocator decided how to distribute asset and the recipient had no right but to accept the allocation. The asset (a certain amount of monetary reward) was initially earned either by the allocator or by the recipient through an unrelated task. Result showed that offers to the recipient were lower in the “allocator-earned” condition and higher in the “recipient-earned” condition, highlighting the importance of property right in determining individuals’ other-regarding behavior [8]. Using a related task, Leliveld et al. [4] investigated how the perception of ownership affects the allocator’s other-regarding behavior in an ultimatum game (UG). This game, originally developed by Güth et al. [9], is similar to the DG but has one major difference: the recipient can either accept or reject the allocator’s offer. If accepted, the pie is divided as proposed; if rejected, both the allocator and the recipient end empty handed. Leliveld et al. [4] put the chips (related to monetary reward later on) either at the allocator’s side of the table or at the recipient’s side of the table. Results showed that allocations to the recipient were higher in the latter case than in the former case; moreover, this modulation of other-regarding behavior by the perceptions of ownership reflected a true concern for other’s welfare rather than fear of rejection.
近期的研究表明，所有权感知在经济决策中的感知影响了他人的行为 [4]，[5]，[6]。Oxoby 和 Spraggon [6] 让参与者进行了一种名为“施舍者游戏”（DG；[7]）的实验，在这个实验中，分配者决定如何分配资产，而接收者只能接受分配。资产（一定数量的货币奖励）最初是由分配者还是接收者通过与实验无关的任务获得的。结果显示，在“分配者获得”条件下，对接收者的提议较低，在“接收者获得”条件下较高，这突显了财产权利在决定个人对他人的行为的重要性 [8]。Leliveld 等人 [4] 使用相关任务，研究了所有权感知如何影响分配者在“最后通牒游戏”（UG）中的对他人的行为。这个游戏最初由 Güth 等人 [9] 开发，与 DG 类似，但有一个主要区别：接收者可以接受或拒绝分配者的提议。如果接受，蛋糕按照提议分配；如果拒绝，分配者和接收者都将空手而归。Leliveld 等人。 [4] 将芯片（与后续的货币奖励相关）放置在分配者一侧的桌子或接收者一侧的桌子上。结果显示，在后者情况下向接收者分配的金额高于前者；此外，这种由所有权感知调节的利他行为反映了对他人福利的真正关心，而不是害怕被拒绝。

The ownership effect is closely related to the concept of entitlement. Entitlement is a kind of feeling that may result from ownership: because I feel I own it, I have a right to end up with it. In the above studies, the allocators distributed more assets to themselves when they had the initial ownership of the assets and felt entitled to have more. It should be noted, however, that the previous behavioral studies on the effect of ownership or entitlement have exclusively focused on the allocator’s decision-making behavior. It is not clear how the recipient’s fairness consideration in economic bargain would be affected by the initial ownership or the feeling of entitlement, and more close to the purpose of the present study, whether and how the brain responses to different levels of fairness in asset allocation are modulated by the initial ownership.

This study was therefore conducted to investigate how initial ownership of a bargaining property modulates recipient’s fairness consideration; this was measured through behavioral reactions (i.e., accepting vs. rejecting offers) and electrophysiological recordings. We randomly assigned the property (a certain amount of monetary reward) to either the allocator or the recipient before the presentation of the division scheme and measured the recipient’s event-related potentials (ERPs) evoked by the division scheme. We manipulated the level of fairness in asset allocation by letting the recipient receive disadvantageous unequal offers (1, 2, or 3 out of 10 Chinese yuan), equal offers (5 out of 10 yuan) or advantageous unequal offers (7 or 8 out of 10 yuan). Behaviorally, we were interested in the acceptance rate for different offers. This rate should decrease as the level of fairness in the division scheme decreases. Importantly, this rate could be lower when the assets were initially owned by the recipient than by the allocator, especially when the offers were disadvantageously unequal. The feeling of entitlement might lead the recipient to demand a larger portion of the pie.

Electrophysiologically, we focused on MFN and P300, two ERP components that are sensitive to the evaluation of fairness in asset distribution. The medial frontal negativity (MFN) or the feedback-related negativity (FRN) was originally observed in studies on performance monitoring and the evaluation of decision outcome [10], [11], [12], [13]. The FRN is a negative deflection peaking between 200 ms and 350 ms at frontocentral recording sites, and is more pronounced for negative feedback associated with unfavorable outcomes, such as incorrect responses or monetary loss, than for positive feedback. Later studies showed that these differential responses to decision outcome can be modulated by social factors, such as interpersonal relationship between the evaluator and the decision maker [14], [15], [16], [17], [18], [19], [20] and the extent of personal responsibility for the outcome [21], [22]. These studies suggest the FRN may reflect the processes of assessing the motivational/affective impact of the outcome events (i.e., the processes of putting subjective values onto the outcomes; [10], [23]. Importantly, recent studies extended the role of FRN in outcome evaluation and performance monitoring to the social domain and to the online decision process and found that violations of social norms, such as unfair or unequal offers in asset division, also elicit more negative-going FRN (or MFN) than fair offers [24], [25], [26], [27], [28]. In such studies, participants were offered either fair (e.g., receiving 50%) or unfair (e.g., receiving 10%) divisions of assets (monetary reward) and ERPs were time-locked to the presentation of such division schemes. Although participants were not directly provided with feedback contingent upon their actions or choices, a division scheme may nevertheless be compared with implicit, long-established social norms (e.g., equal division) concerning asset distribution and the scheme’s motivational/affective significance is hence derived. Any violation of the norms would elicit the MFN (FRN) responses. A study showed that individuals highly appreciating moral norms such as fairness and honesty exhibited larger MFN amplitudes when processing unfair offers than individuals with less regard for such norms [24].

For the present design, we predicted a more negative-going MFN for the disadvantageous unequal offers, as compared to the equal offers. The situation for the advantageous unequal offers could be more complex. If the MFN reflects the violation of expectancy at some abstract level, then the advantageous unequal offers should elicit a more negative-going MFN, similar to the disadvantageous unequal offers. If the MFN reflects simply the valence of offers or the relevance of offers to self-interests, then we should predict a less negative-going MFN for the advantageous unequal offers, as compared to the equal offers. Oliveira et al. [29] found that both unexpected positive reward and unexpected negative reward elicited more negative-going FRN responses than expected feedback in an anticipation-timing task and that the two FRN effects were of equal magnitudes, consistent with the first possibility outlined above.

More importantly, we predicted that the initial ownership may modulate the MFN effects for the disadvantageous and advantageous unequal offers. Previous behavioral studies have already shown that people’s fairness consideration is influenced by various social factors, such as the valence (gain vs. loss) of a bargaining property [30], [31], the power of those involved in the transaction [32] and the context in which a division scheme is presented [33], [34]. We also conducted an ERP study asking participants to act as recipients in DG and received either equal or disadvantageous unequal offers from friends or strangers [27]. The MFN effect for disadvantageous unequal offers was present only in the friend-allocation condition, not in the stranger-allocation condition. In the present design, the (dis-)advantageous unequal offers could elicit stronger MFN effects when the bargaining property was initially owned by the recipients rather than by the allocators, as the feeling of entitlement could enhance the recipients’ expectancy towards an equal or larger portion of the pie.

Another ERP component, the P300, which is the most positive peak in the period of 200–600 ms post-onset of feedback and which typically increases in magnitude from frontal to parietal electrodes, has also been found to be related to various aspects of outcome evaluation. Earlier studies employing the oddball paradigm suggested that the P300 is related to higher-order cognitive operations, such as memory updating, selective attention and resource allocation [35]. The P300 has also been found to be related to various aspects of outcome evaluation. Some studies found that the P300 is sensitive to the magnitude of reward, with a more positive response to a larger than to a smaller reward [13], [36]. Other studies suggested that the P300 is also sensitive to reward valence, with a more positive amplitude for positive feedback than for negative outcome [18], [37], [38], [39], [40]. It is possible that during outcome evaluation more attentional resources have been devoted to the outcome magnitude or valence that has stronger motivational significance to the participants. In asset distribution, Wu and colleagues found that the P300 is more positive to fair offers than to disadvantageous unfair offers, suggesting that differential distribution of attentional resources to the two types of offers which had different affective/motivational significance [27]. We hence predicted to observe the same pattern on the P300 for the two types of offers. For advantageous unequal offers, one would normally predict an even more positive P300 as the P300 has been found to increase with the magnitude of reward [13], [36].

It was not clear, however, whether and how the P300 would be modulated by initial ownership. Turk et al. [41] observed a more positive P300 to an object when it was assigned to the participant rather than to another person. The authors interpreted this effect as reflecting increased attention to the self-relevant objects. In the present design, however it was not clear how this initial ownership would affect the P300 responses to the later division schemes.

Methods

Design and procedures

The experiment had a 3×2 within-participant factorial design, with the first factor referring to the offer type (disadvantageous unequal vs. equal vs. advantageous unequal offer) and the second factor referring to the initial ownership (self vs. other). Disadvantageous unequal offers could be 1, 2, or 3 yuan (out of 10 yuan), equal offers could be 5 yuan (out of 10 yuan), and advantageous unequal offers could be 7 or 8 yuan (out of 10 yuan). The bargaining property (10 yuan) was assigned ostensibly by the computer to either the recipient or the allocator in random order before the division scheme was presented to the recipient.

When the EEG participant came to the laboratory, he/she and the four confederates were told that they would sit in separate rooms to finish a task together through the computer network. The EEG participant was ostensibly selected through lottery to undergo the EEG test. This participant was then told that he/she would play as a recipient in UG and the others would be allocators. He/she was also informed about the rules of UG and the manipulation of ownership. That is, at the beginning of each round the computer would randomly assign 10 yuan to either the allocator or himself/herself, and the allocator would then offer a scheme on how to divide the amount. The EEG participant was asked to press a button with the index finger of his/her left or right hand, without elaborative thinking, to indicate whether he/she would accept or reject the offer. He/she was reminded that the allocators made their division schemes individually and independently, and his/her response would not be sent back to the allocator immediately and therefore could not affect the allocators’ offers in following rounds.

Each trial began with the presentation of a fixation sign (a white dot subtended 0.3? of visual angle) for 500 ms against a black background (see Fig. 1). The sentence “The computer is randomly pairing” in Chinese (white and Song font, size 32) was presented for either 800, 850, 900, 950, or 1000 ms, indicating to the EEG participant that one of the other four persons was randomly selected to play as an allocator in the current round of game. Then the EEG participant’s own head portrait and a silhouette (each subtended 1.5 × 1.6°, separated for 2.3° between the centers of the two figures) were presented at the left side of the screen for either 800, 900, 1000, 1100 or 1200 ms, along with Chinese words “please wait” (white and Song font, size 32) at the left. This was to suggest to the participant that the computer was assigning the initial ownership of the 10 yuan. The positions of these two figures were counterbalanced over trials. After this frame, the assignment of initial ownership, with a photo of a 10 yuan bill (2.6° × 1.3°) aligned with either of the figures, was presented for 1500 ms. It was explained to the participant that the computer had endowed the money initially to the person involved. After the presentation of a blank screen for a jittered time between 500 and 800 ms, the allocator’s division scheme, in two lines of words (e.g., “he 8, you 2”, white and Song font, size 32) was revealed at the center of screen for 1200 ms. The screen turned blank again for 500 ms, followed by the presentation of two options, “accept” and “reject”, on the left and right side of the screen, with the positions of the two options counterbalanced over participants. The EEG participant was asked to make the “accept” or “reject” decision as quickly as possible and the next trial began after 1000 ms after the button press.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Sequence of events in a single trial.

https://doi.org/10.1371/journal.pone.0039627.g001

The participant was seated comfortably about 1.5 m in front of a computer screen in a dimly lit room. The experiment was administered on a computer with a Del 22-in. CRT display, using Presentation software (Neurobehavioral System Inc.) to control the presentation and timing of the stimuli. The experiment consisted of 4 blocks of 75 trials each. Under each of the two types of initial ownership, the disadvantageous unequal condition consisted of 20 trials of 1/9 offer, 20 trials of 2/8 offer and 10 trials of 3/7 offer; the equal condition consisted of 40 trials of 5/5 offer; and the advantageous unequal condition consisted of 20 trials of 7/3 offer and 20 trials of 8/2 offer. In addition, 10 trials of 4/6 offer and another 10 trials of 6/4 offer were used as fillers. The number before the slash indicated the offered amount to the recipient and the number after the slash indicated the amount left to the allocator. Without the participant’s knowledge, all the offers were predetermined by a computer program. The 300 trials were pseudo-randomized with the restriction that no more than 3 consecutive trials were of the same offer type and no more than 3 consecutive trials were of the same initial ownership.

A practice block of 20 trials was administered before the formal test. To check the manipulation of initial ownership, we presented the participants, after the experiment, with the fourth frame of Fig. 1. and asked them to indicate on a 7-point Likert Scale to what extent they felt that the property should be in their own possession (1 =  absolutely not in their own possession, 7 =  absolutely in their own possession) and to what extent they felt that the property should be in the allocator’s possession (1 =  absolutely not in the allocator’s possession, 7 =  absolutely in the allocator’s possession) in each initial ownership condition. We also presented the participants with the fourth frame of Fig. 1 and asked the participants to indicate the minimal amount (out of 10 yuan) they wanted from the pie and the fairest division they perceived when the property was assigned to themselves or to the allocator. The participants were debriefed, paid and thanked at the end of the experiment.

Results

Among the thirty EEG participants, three participants stated that they completely disbelieved the setup of the experiment in the interview after the EEG test, four participants displayed excessive artifacts in EEG recording, one participant misunderstood the game rule, and one participant accepted all the offer types. These participants were excluded from data analysis, leaving twenty-one participants (8 females) for the following analysis.

Behavioral Results 行为结果

The acceptance rates for different division schemes are presented in Fig. 2. A 2 (initial ownership: self vs. other) × 3 (offer type: disadvantageous unequal vs. equal vs. advantageous unequal offer) repeated measures ANOVA revealed a significant main effect of offer type, F(2, 40) = 108.31, p<0.001, indicating that the acceptance rate for disadvantageous unequal offers (0.24±0.05) was lower than for either equal (0.91±0.03) or advantageous unequal offers (0.94±0.03), as confirmed by the post-hoc tests (p<0.001). The differences between the equal and the advantageous unequal offer conditions were not significant (p>0.1). The main effect of initial ownership was also significant, F(1, 20) = 12.24, p<0.01, suggesting that the acceptance rate was higher when the 10 yuan bill was initially aligned with the other’s silhouette (0.76±0.02) than when bill was presented with the participant’s own portrait (0.63±0.03). Importantly, the interaction between initial ownership and offer type was significant, F(2, 40) = 7.25, p<0.01. Simple-effect tests showed that the acceptance rate to disadvantageous unequal offers was significantly higher in the “other” condition (0.36±0.07) than in the “self” condition (0.11±0.04), t(20) = 3.81, p<0.01. A similar pattern was observed for equal offers (0.99±0.003 vs. 0.83±0.06), t(20) = 2.59, p<0.05. However, this effect was absent for advantageous unequal offers, t(20) = –0.70, p>0.1.
不同分组方案的接受率在图 2 中呈现。2×3 重复测量 ANOVA（初始所有权：自我 vs.他人）×（报价类型：不利不平等 vs.平等 vs.有利不平等报价）显示，报价类型的主要效应显著，F(2, 40) = 108.31，p<0.001，表明不利不平等报价（0.24±0.05）的接受率低于平等报价（0.91±0.03）或有利不平等报价（0.94±0.03），后验检验也证实了这一点（p<0.001）。平等报价与有利不平等报价条件之间的差异不显著（p>0.1）。初始所有权的主要效应也显著，F(1, 20) = 12.24，p<0.01，表明当 10 元钞票与他人的轮廓对齐时（0.76±0.02），接受率高于展示参与者自己的肖像时（0.63±0.03）。重要的是，初始所有权与报价类型之间的交互作用也显著，F(2, 40) = 7.25，p<0.01。简单效应检验显示，在“他人”条件下，不利不平等报价的接受率显著较高（0.36±0.）。07) 比“自我”条件（0.11±0.04）中的要高，(20) = 3.81，p<0.01。对于等额提议，也观察到了类似的趋势（0.99±0.003 vs. 0.83±0.06），(20) = 2.59，p<0.05。然而，这种效应在有利的不平等提议中不存在，(20) = –0.70，p>0.1。

Download:  下载：

• PPT
  PowerPoint slide PowerPoint 幻灯片
• PNG
  larger image 更大图片
• TIFF
  original image 原始图像

Figure 2. The acceptance rate in the ultimatum game as a function of the offer type.
图 2. 极限提议游戏中的接受率作为提议类型的功能。

Error bars represent standard errors of the means.
误差棒代表平均值的标准误差。

https://doi.org/10.1371/journal.pone.0039627.g002

ERP Responses to the Presentation of Division Schemes
ERP 对分部方案呈现的响应

For the mean amplitudes in the 280–380 ms (MFN) time window (Fig. 3A and 3B), a 2 (initial ownership: self vs. other) × 3 (offer type: disadvantageous unequal vs. equal vs. advantageous unequal offer) repeated-measures ANOVA showed a significant main effect of offer type, F(2, 40) = 8.66, p<0.01, indicating that ERP responses were more negative-going for disadvantageous (−2.06 µV) and advantageous unequal offers (−2.72 µV) than for equal offers (−1.19 µV), p = 0.06 and p<0.01, respectively. The ERP responses to the two types of unequal offers did not differ, p>0.1. However, we found no significant main effect of initial ownership, F(1, 20) = 0.18, p>0.1, nor the interaction between initial ownership and offer type, F(2, 40) = 1.13, p>0.1.
对于 280-380 ms（MFN）时间窗口的平均振幅（图 3A 和 3B），进行了 2（初始所有权：自我 vs.他人）× 3（提议类型：不利不平等 vs.平等 vs.有利不平等提议）的重复测量 ANOVA，结果显示提议类型的主要效应显著，F(2, 40) = 8.66，p<0.01，表明 ERP 响应对于不利（-2.06 µV）和有利不平等提议（-2.72 µV）比平等提议（-1.19 µV）更负向，p = 0.06 和 p<0.01，分别。两种不平等提议的 ERP 响应没有差异，p>0.1。然而，我们没有发现初始所有权的主要效应显著，F(1, 20) = 0.18，p>0.1，也没有初始所有权与提议类型之间的交互作用，F(2, 40) = 1.13，p>0.1。

Download:  下载：

• PPT
  PowerPoint slide PowerPoint 幻灯片
• PNG
  larger image 更大图片
• TIFF
  original image 原始图像

Figure 3. ERP responses and topographic maps.
图 3. ERP 响应和地形图。

(A) ERP responses time-locked to the onset of different offers at the midline FCz, Cz and Pz. The shaded 280–380 ms time window was for the calculation of the mean amplitudes of the MFN. The shaded 400–600 ms time window was for the calculation of the mean amplitudes of the P300. (B) Topographic maps for the MFN effects in the 280–380 ms time window. (C) Topographic maps for the P300 effects in the 400–600 ms time window.
(A) ERP 响应，锁定在中线 FCz、Cz 和 Pz 不同提议的开始时间。计算 MFN 平均幅度的阴影 280-380 毫秒时间窗口。计算 P300 平均幅度的阴影 400-600 毫秒时间窗口。 (B) 280-380 毫秒时间窗口内 MFN 效应的地形图。 (C) 400-600 毫秒时间窗口内 P300 效应的地形图。

https://doi.org/10.1371/journal.pone.0039627.g003

Similarly, for the mean amplitudes in the 400–600 ms (P300) time window (Fig. 3A and 3C), the 2 × 3 ANOVA showed also a main effect of offer type, F(2, 40) = 32.98, p<0.001, indicating that the mean amplitudes were more positive for equal offers (4.36 µV) than for disadvantageous unequal offers (0.94 µV) or advantageous unequal offers (0.12 µV). The differences between conditions were all significant after Bonferroni correction, p<0.001 or p<0.01. The main effect of initial ownership was also significant, F(1, 20) = 8.28, p<0.01, suggesting that the ERP responses were more positive for the “other” (2.12 µV) than for the “self” condition (1.49 µV). The interaction between offer type and initial ownership did not reach significance, F(2,40) = 0.54, p>0.1, indicating that the initial ownership effect on the P300 was essentially the same across the three types of offers.
同样地，对于 400-600 毫秒（P300）时间窗口的平均振幅（图 3A 和 3C），2×3ANOVA 也显示了主要的提供类型效应，F(2, 40) = 32.98，p<0.001，表明平均振幅对于等价提供（4.36 µV）比对于不利的不等提供（0.94 µV）或有利的不等提供（0.12 µV）更为正向。在 Bonferroni 校正后，各条件之间的差异均显著，p<0.001 或 p<0.01。初始所有权的主要效应也显著，F(1, 20) = 8.28，p<0.01，表明 ERP 响应对于“他人”（2.12 µV）比对于“自我”条件（1.49 µV）更为正向。提供类型与初始所有权之间的交互作用未达到显著性，F(2,40) = 0.54，p>0.1，表明初始所有权对 P300 的影响在三种类型的提供中基本上是相同的。

Discussion 讨论

This study demonstrated that initial ownership influenced recipients’ brain responses to unfair asset allocation schemes and their behavioral decisions to accept offers. Participants were more reluctant to accept disadvantageous unequal and equal offers when the bargaining property was initially in their own possession than when it was in the other’s possession, and this distinction disappeared for advantageous unequal offers. Electrophysiologically, both disadvantageous and advantageous unequal offers elicited more negative going ERP responses compared to equal offers in an earlier, MFN time window (280–380 ms), with no obvious differences between the two types of unequal offers. These earlier effects were not affected by the initial ownership. In a later time window (400–600 ms), however, the P300 was more positive for equal offers than for disadvantageous unequal or advantageous unequal offers and were more positive when the bargaining property was initially owned by the allocator than by the recipient. In the following paragraphs, we explore the implications of our behavioral and electrophysiological findings, focusing on the effects of fairness in asset allocation and the effects of initial ownership.
本研究显示，初始所有权影响了接收者对不公平资产分配方案的大脑反应以及他们接受提议的行为决策。参与者在协商财产最初属于他们自己时，对于不利的不平等和平等提议的接受意愿比财产最初属于对方时更为犹豫。然而，对于有利的不平等提议，这种区别消失。从电生理学角度来看，不利和有利的不平等提议在较早的时间窗口（280-380 ms）内，与平等提议相比，引发了更多的负向 ERP 响应，两种类型的不平等提议之间没有明显差异。这些早期效应不受初始所有权的影响。然而，在较晚的时间窗口（400-600 ms）内，对于平等提议，P300 的正向反应比不利的不平等或有利的不平等提议更明显，且当协商财产最初属于分配者而非接收者时，这种正向反应更为显著。 在以下段落中，我们将探讨我们行为和电生理学发现的含义，重点是资产分配的公平性影响以及初始所有权的影响。

Previous studies have shown that noncausal forms of association between an individual and an object (e.g., the numbers corresponding to an individual’s birthday, prior touch or use of the object) can significantly increase the individual’s preference or valuation of the object [43], [44], [45], [46]. In this study, the initial random assignment of the 10 yuan bill strongly affected the participants’ perception of ownership and the feeling of entitlement in subsequent asset distribution, as demonstrated by their post-experiment self-report. Although the participants had been explicitly told that the initial assignment was randomly conducted by computer and it did not imply that they would eventually have the money, this perception of ownership and feeling of entitlement had nevertheless affected the participants’ subsequent acceptance or rejection of disadvantageously unfair and even fair (equal) offers.
先前的研究表明，个体与物体之间非因果形式的关联（例如，与个体生日对应的数字，或个体先前触摸或使用该物体的次数）可以显著增加个体对物体的偏好或价值评估 [43]，[44]，[45]，[46]。在本研究中，10 元钞票的初始随机分配对参与者对所有权的感知以及后续资产分配中的满足感产生了强烈影响，这一点通过他们的实验后自我报告得到了证明。尽管参与者明确被告知初始分配是通过计算机随机进行的，并不意味着他们最终会得到这笔钱，但这种对所有权的感知和满足感仍然影响了参与者对不利不公平甚至公平（等同）提议的后续接受或拒绝。

One of the prominent motivations for individuals rejecting disadvantageous unequal offers in asset distribution is to preserve self-image/self-esteem and/or to punish unfair behavior [47], [48]. The assignment of initial ownership to a participant may increase his feeling of entitlement, and disadvantageous unequal divisions and even equal divisions would be perceived as challenges to his/her self-image or self-esteem. These challenges would then meet strong reactions, resulting in lower acceptance rates. Thus, the perceived fairness or equity in asset distribution can be highly context-dependent [33], [34].
拒绝资产分配中不利不平等提议的主要动机之一是为了保护自我形象/自尊，或惩罚不公平行为 [47]，[48]。参与者初始所有权的分配可能会增加其对所有权的归属感，不利的不平等分配甚至平等分配都会被视为对其自我形象或自尊的挑战。这些挑战会引发强烈的反应，导致接受率降低。因此，资产分配的公平性或公正性可能高度依赖于上下文 [33]，[34]。

On the other hand, the high acceptance rate and the absence of initial ownership effect for advantageous unequal offers can also be taken as evidence for the context-dependent nature of fairness consideration. Although individuals care for fairness in asset distribution, particularly when they are in a disadvantageous position [30], [31], they are nevertheless self-interested. This care for self-interests may be strategic and is shown when their self-interests are not likely to be negated. In this situation, effects of other social factors (including the initial ownership) are dwarfed or overshadowed.
另一方面，对有利的不平等提议的高接受率以及不存在初始所有权效应，也可以作为公平考虑具有上下文依赖性的证据。尽管个人关心资产分配的公平性，尤其是在不利位置时 [30]、[31]，但他们仍然是自私的。这种对自身利益的关心可能是策略性的，并在他们的自身利益不太可能被否定的情况下表现出来。在这种情况下，其他社会因素（包括初始所有权）的影响被缩小或掩盖。

The finding of an MFN effect, with more negative going responses to disadvantageous unequal offers than to equal offers, replicated previous studies [24], [25], [26], [27], [28]. This effect may reflect the detection of social expectancy violation as egalitarian distribution of assets is an expected social norm in our life [49], [50], [51]. The human brain might have developed specific mechanisms to detect ongoing deviation from social norms [52] and these mechanisms might be based on similar neural substrates as those engaged in detecting errors during non-social reinforcement leaning [53]. For instance, a recent fMRI study on social conformity in facial attractiveness judgment showed that conflict with group opinions, regardless of whether the opinions were given by human peers or by computers, triggered activation of brain regions implicated in reinforcement learning, i.e., rostral cingulate zone and the ventral striatum, and these neural signals can predict whether the participants would subsequently change their initial judgment [54].
发现 MFN 效应，即对不利的不平等提议的反应比对平等提议的反应更负向，重复了之前的研究所发现的结果 [24]，[25]，[26]，[27]，[28]。这种效应可能反映了对社会期望违反的检测，因为在我们的生活中，资产的平等分配是一个预期的社会规范 [49]，[50]，[51]。人类的大脑可能发展了特定的机制来检测社会规范的持续偏离 [52]，这些机制可能基于在非社会强化学习中检测错误时所涉及的相似神经基础 [53]。例如，最近的一项关于面部吸引力判断的社会顺应性的 fMRI 研究显示，无论是由人类同伴还是计算机给出的意见与群体意见的冲突，都会触发与强化学习相关的脑区的激活，即前扣带皮层和腹侧基底核，这些神经信号可以预测参与者是否会随后改变他们的初始判断 [54]。

Importantly, we found that the advantageous unequal offers also elicited more negative-going MFN responses than equal offers, and this effect appeared to be of equal magnitude as for disadvantageous unequal offers. This finding is novel and important because it allows us to differentiate theoretical proposals concerning the nature of MFN or FRN (assuming they are essential the same, as we argued in the Introduction). One proposal is that the FRN reflects the impact of midbrain dopamine signals on the anterior cingulate cortex (ACC) [11], [55]. The phasic decreases in dopamine inputs elicited by negative prediction errors (i.e., “the result is worse than expected”) give rise to the increased ACC (anterior cingulate cortex) activity that is reflected as larger MFN amplitude, whereas the phasic increases in dopamine signals elicited by positive prediction errors (i.e., “the result is better than expected”) give rise to decreased ACC activity that is reflected as smaller MFN amplitudes. By this account, we should expect to observe less negative (or more positive) going MFN responses to advantageous unequal offers than to equal offers. Another more viable conception is that the FRN (the associated ACC) serves as a general performance monitoring system which detects violation of (social and non-social) expectancy, irrespective of whether the violated expectancy is positive or negative [29], [39], [56]. By this account, although the advantageous unequal offers could benefit the recipients (i.e., better than expected), they nevertheless violated the equity rule in asset distribution [57], [58], just as the disadvantageous unequal offers. Thus any division schemes in violation of the equity rule would be detected by the monitoring system, resulting in more negative-going MFN responses (see also [29], [39], [56]). Indeed, recent studies suggest that the short-latency phasic responses in the dopamine system are related to a general process of switching attention to unexpected, behaviorally relevant stimuli [59]. The feedback, whether positive or negative, elicits a phasic increase in the activity of mesencephalic dopamine neurons which, in turn, induces increased excitability in the ACC, thereby giving rise to the FRN/MFN effect.
重要的是，我们发现有利的不平等提议也引发了比平等提议更多的负向 MFN 响应，而且这种效应似乎与不利的不平等提议的效应相等。这一发现是新颖且重要的，因为它使我们能够区分关于 MFN 或 FRN 本质的理论提议（假设它们本质上是相同的，正如我们在引言中所争论的）。一种提议是，FRN 反映了中脑多巴胺信号对前扣带回（ACC）的影响[11]，[55]。由负面预测误差（即，“结果比预期的更糟”）引发的多巴胺输入的短暂减少导致了 ACC 活动的增加，这反映为更大的 MFN 幅度，而由正面预测误差（即，“结果比预期的好”）引发的多巴胺信号的短暂增加导致了 ACC 活动的减少，这反映为较小的 MFN 幅度。据此，我们应预期观察到有利的不平等提议的 MFN 响应比平等提议的响应更少（或更积极）。 另一种更可行的观念是，FRN（关联的 ACC）作为通用性能监控系统，能够检测到无论是社会的还是非社会的期望被违反，无论被违反的期望是积极的还是消极的 [29]，[39]，[56]。根据这一观点，尽管有利的不平等提议可能对接受者有利（即比预期的更好），但它们仍然违反了资产分配的公平规则 [57]，[58]，就像不利的不平等提议一样。因此，任何违反公平规则的分配方案都会被监控系统检测到，导致更强烈的 MFN 响应（参见 [29]，[39]，[56]）。事实上，最近的研究表明，多巴胺系统的短时相响应与对意外、行为相关刺激的注意力转移过程有关 [59]。无论是正面的还是负面的反馈，都会引发中脑多巴胺神经元活动的短时相增加，进而导致 ACC 的兴奋性增加，从而产生 FRN/MFN 效应。

A perhaps surprising finding in this study was that the initial ownership of the distributed asset had no obvious effect on the MFN responses to division schemes. This absence of an initial ownership effect appears to be at odds with Wu et al. [27] in which the social distance between the recipient and the allocator (being a friend or a stranger) modulated the MFN responses to fair and disadvantageous unfair offers. We believe that the discrepancy between the two studies may be related to paradigms adopted for the experiments. Wu et al. [27] used a DG task in which the recipient had no choice but to accept any offers given by the allocator. However, in the UG task used here, the recipient could choose either to accept or to reject offers. Thus the outcome to the recipient was deterministic in DG and was negotiable in UG. This difference in the certainty of outcome may affect the level of affective/motivation significance assessment for the offers, as deeper or more comprehensive assessment in DG allows social/affective factors to play a bigger role. In UG, however, the system may adopt a “wait-and-see” strategy and conduct deeper assessment of offers only at a later stage involving more top-down processes [18], [19], [60], [61], rendering a P300 effect for the initial ownership (see later discussion).
在这项研究中，一个可能令人惊讶的发现是，分布式资产的初始所有权对 MFN 响应对分割方案的影响并不明显。这种初始所有权效应的缺失似乎与吴等人[27]的研究结果相矛盾，在该研究中，接受者与分配者之间的社会距离（是朋友还是陌生人）调节了对公平和不利的不公平提议的 MFN 响应。我们认为，这两项研究之间的差异可能与所采用的实验范式有关。吴等人[27]使用了 DG 任务，在该任务中，接受者别无选择，只能接受分配者提供的任何提议。然而，在这里使用的 UG 任务中，接受者可以选择接受或拒绝提议。因此，在 DG 中，结果是确定性的，在 UG 中是可协商的。结果的确定性可能会影响对提议的情感/动机重要性评估水平，因为在 DG 中进行更深入或更全面的评估，社会/情感因素可以发挥更大的作用。 然而，在 UG 中，系统可能会采取“观望”策略，在后续涉及更多自上而下过程的较晚阶段进行更深入的评估提议，[18]，[19]，[60]，[61]，从而为初始所有权产生 P300 效应（请参阅后续讨论）。

Note that the assignment of bargaining property to the participant may enhance their demand for a larger portion of the pie, evidenced by the reduced acceptance rate for the equal division in the “self” condition than in the “other” condition. One might view this enhanced demand for the pie in self-ownership as a kind of social norm. Consequently if the MFN reflects the violation of social norm, we should expect to observe more negative-going MFN responses to the disadvantageous offers in the “self” condition than in the “other” condition. Although we did obtained numerically larger MFN responses for the former (−2.21 µV) than for the latter (−1.92 µV), the difference between the two conditions did not reach statistical significance. It is possible that when different social norms are involved in evaluating schemes of asset division, the equity rule, which is ubiquitous in the society, might dominate over other rules, including the rule for a larger portion of pie in self-ownership, in determining the MFN responses. Further studies are needed to investigate how different social norms or rules might interact to modulate the brain activity in outcome evaluation or interpersonal interaction.
请注意，将讨价还价属性分配给参与者可能会增强他们对更大份额的需求，这从“自我”条件中对等分的接受率低于“他人”条件的情况中可以得到证据。人们可能会将自我所有权中的这种对蛋糕的增强需求视为一种社会规范。因此，如果 MFN 反映了社会规范的违反，我们应预期在“自我”条件中对不利提议的 MFN 响应比在“他人”条件中更呈现为负向变化。尽管我们确实获得了前者（-2.21 µV）比后者（-1.92 µV）数值上更大的 MFN 响应，但两个条件之间的差异并未达到统计学显著性。当涉及资产分配方案的评估时，如果涉及不同的社会规范，普遍存在于社会中的公平原则可能在决定 MFN 响应时会超过其他规则，包括自我所有权中更大份额的规则。 需要进一步的研究来探讨不同的社会规范或规则如何相互作用，以调节结果评估或人际互动中的大脑活动。

In contrast to the MFN, we found that the P300 was modulated by both the offer type and the initial ownership, although these modulations were independent from each other. Previous studies on outcome evaluation have indicated that the P300 is related to processes of attentional allocation [62], [63] and/or high-level motivational/affective evaluation [13], [64]. According to the equity theory [57], [58], individuals who are facing inequity would feel distressed and are less satisfied with asset distribution than individuals who are facing equity. The stronger P300 responses to equal offers than to unequal offers may suggest that participants (recipients of asset distribution) in this study attached more motivational/affective significance to the equal divisions than to unequal divisions, consistent with the social fairness norms cultured in individuals.
与最惠国待遇（MFN）不同，我们发现 P300 不仅受到提议类型和初始所有权的影响，而且这些影响是相互独立的。关于结果评估的先前研究已经表明，P300 与注意力分配过程[62]、[63]和/或高级动机/情感评估[13]、[64]有关。根据公平理论[57]、[58]，面对不公平的人会感到困扰，并且对资产分配的满意度低于面对公平的人。研究中对等提议比不等提议产生更强的 P300 响应可能表明，参与者（资产分配的接受者）更重视对等分配，而不是不等分配，这与个体中培养的社会公平规范一致。

In addition, we found that disadvantageous unequal offers elicited more positive P300 than advantageous unequal offers. Although both types of offers violate the equity rule of social norms, it is possible that different amount of attentional resources are used to process the two types of offers. For disadvantageous unequal offers, participants might be in a difficult position to assess the pros and cons of accepting or rejecting offers; for advantageous unequal offers, participants might not have such dilemma and they can assess the implications of offers, as demonstrated by their 94% acceptance rate.
此外，我们发现不利的不平等提议比有利的不平等提议激发了更多的积极 P300。尽管两种类型的提议都违反了社会规范的公平原则，但可能使用了不同数量的注意力资源来处理这两种类型的提议。对于不利的不平等提议，参与者可能难以评估接受或拒绝提议的利弊；对于有利的不平等提议，参与者可能没有这样的困境，他们可以评估提议的含义，正如他们 94%的接受率所显示的那样。

On the other hand, we observed a small, but significant initial ownership effect on the P300, with the offers from the bargaining property initially owned by allocator eliciting more positive P300 responses than the offers from the property initially owned by the recipient himself/herself. A number of studies on outcome evaluation have shown that the P300 is sensitive to reward valence in gambling tasks, with positive outcomes eliciting more positive P300 than negative outcomes [18], [37], [38], [39], [40]. In the present study, any amount proposed by the allocator in the “other” condition might be considered, implicitly, as a kind of extra “gain”, even though the recipient may eventually decide to reject the offer and lose it. Conversely, any amount proposed by the allocator in the “self” condition might be considered as a kind of “loss” as the bargaining property was initially assigned to the recipient and he/she might implicitly declare the ownership of the whole lot (see also [65]).
另一方面，我们观察到一个小但显著的初始所有权效应在 P300 上，来自协商财产最初由分配者拥有的提议引发的 P300 响应比最初由接收者自己拥有的财产的提议更积极。许多关于结果评估的研究表明，P300 对赌博任务中的奖励价值敏感，正向结果引发的 P300 比负向结果更积极 [18]，[37]，[38]，[39]，[40]。在本研究中，分配者在“其他”条件下提出的任何金额，可以被隐含地视为一种额外的“收益”，尽管接收者最终可能决定拒绝提议并失去它。相反，分配者在“自我”条件下提出的任何金额可以被视为一种“损失”，因为协商财产最初被分配给接收者，他/她可能隐含地宣布了整个财产的所有权（也参见[65]）。

An important finding here was that the modulations of the P300 by offer type and initial ownership appeared to be independent from each other, consistent with the absence of an interaction between fairness of offers and social distance between the allocator and recipient in DG [27]. We would like to suggest that there are two top-down processes associated with the P300. One process cares for fairness of different offers, with different levels of attentional resources being devoted to the elaborative processing of the social/affective significance of offers. Another process cares more for self-interests and is sensitive to gain/loss. Either of two processes can modulate the P300 magnitude, although it needs further investigation to elucidate under what circumstances the two processes work independently when they are manipulated concurrently.
重要发现之一是，提供类型和初始所有权对 P300 的调节似乎彼此独立，这与公平提供与分配者和接收者之间的社会距离在 DG [27] 中不存在交互作用一致。我们建议 P300 与两个自上而下的过程相关联。一个过程关注不同提供的公平性，不同级别的注意力资源被投入到对提供的社会/情感意义的详细处理中。另一个过程更关注个人利益，并对收益/损失敏感。两个过程中的任何一个都可以调节 P300 的大小，尽管需要进一步的研究来阐明在同时操纵这两个过程时，它们在何种情况下独立工作。

In summary, by assigning a bargaining property to either the allocator or the recipient and presenting the recipient with offers of different fairness levels, we found that the participant, acting as the recipient, were more likely to reject disadvantageous unequal and equal offers when they initially owned the property than when they did not. The two types of unequal offers evoked more negative-going MFN than the equal offers in an early time window (280–380 ms) and these differential effects were not modulated by the initial ownership. In a late time window (400–600 ms), however, the P300 responses to division schemes were affected not only by offer types but also by whom the property was initially assigned to. These findings suggest that while the MFN may function as a general mechanism that evaluates whether the offer is consistent or inconsistent with the equity rule, the P300 is sensitive to later, top-down controlled processes, into which factors related to the allocation of attentional resources, including initial ownership and personal interests, come to play.
总结而言，通过将讨价还价属性赋予分配者或接收者，并向接收者提供不同公平水平的提议，我们发现参与者（作为接收者）在最初拥有属性时，更有可能拒绝不利的不平等和平等提议，与他们不拥有属性时相比。两种类型的不平等提议在早期时间窗口（280-380 ms）中引发的 MFN 反应比平等提议更多，这些差异效应不受初始所有权的影响。然而，在晚期时间窗口（400-600 ms），对分配方案的 P300 响应不仅受到提议类型的影响，还受到最初将属性分配给谁的影响。这些发现表明，虽然 MFN 可能作为评估提议是否与公平规则一致或不一致的一般机制发挥作用，但 P300 对后续的、自上而下控制的过程敏感，其中包括注意力资源分配、初始所有权和个人利益等因素扮演了角色。

Acknowledgments 致谢

We thank Prof. Albrecht Inhoff, Mr. Stephen Politzer-Ahles, Mr. Stijn Massar and three anonymous reviewers for their comments on early versions of the manuscript.
我们感谢阿尔布雷希特·因霍夫教授、斯蒂芬·波利策-阿赫尔斯先生、斯蒂恩·马萨尔先生以及三位匿名审稿人对手稿早期版本的评论。