| 摘要: | 臉孔吸引力（facial attractiveness）在人類演化歷程中有其價值，臉孔較吸引人的個體，在生存競爭中有其優勢。過往的實徵研究不僅展現了影響臉孔吸引力的普遍因素，還指出吸引力偏好會被生活經驗所影響。然而，儘管先前的研究在臉孔編碼（encoding）和辨識（recognition）歷程中可穩定觀察到其他種族效應（other-race effect，ORE，即相比於其他種族的臉孔，人們對自己種族的臉孔有更好的辨別和記憶能力），但並未發現自己種族臉孔（own-race faces）在吸引力判斷中具有優勢。有鑒於在臉孔吸引力的外顯判斷中缺乏ORE的證據，本研究的目的為探討在沒有意識到臉孔的情況下，對臉孔種族屬性的知覺是否會影響對其吸引力的判斷。
根據研究臉孔知覺的文獻，使用視覺遮蔽（visual masking）和持續閃現抑制（continuous flash suppression，CFS）派典的行為研究顯示，針對臉孔吸引力和種族這兩種屬性的知覺都是自動且可無意識地處理的。本研究採用這兩種派典，並同時記錄事件關聯電位（event-related potentials，ERPs），以檢驗在早期臉孔處理歷程中無意識（unconscious）和有意識（conscious）處理階段的臉孔種族和吸引力知覺。在視覺遮蔽派典中，臉孔刺激在前遮蔽刺激和後遮蔽刺激之間出現，以避免臉孔的視覺訊息在臉孔呈現之前或之後被處理。在CFS派典中，CFS刺激呈現於主視眼，使呈現於另一隻眼的臉孔刺激在試驗開始時無法被察覺，並透過臉孔被偵測到的反應時間來測量臉孔進入意識所需的時間。
在本論文中，我們首先進行了兩個行為實驗，分別在視覺遮蔽派典中呈現短時間（100毫秒或更短）的臉孔刺激，以檢驗臉孔種族和吸引力這兩個屬性在意識階段被處理到的相對速度（實驗一），以及探討這兩個臉孔屬性是否在CFS派典中可被無意識地處理、並展現出交互作用（實驗二）。基於這兩個行為實驗，我們進一步執行了兩個神經生理實驗（實驗三和四），以檢驗臉孔種族和吸引力這兩個屬性對行為反應的影響是否源於潛意識（subliminal）和意識（supraliminal）階段的早期臉孔知覺。
在實驗一的結果中，當臉孔刺激呈現33.34毫秒時，受試者能夠正確判斷臉孔的種族和吸引力程度，表示在有意識的判斷中，這兩個屬性均能夠被有效地處理。在實驗二中，我們發現在CFS流程下，對自己種族臉孔的偵測時間快於偵測其他種族的臉孔。此外，在有持續閃現刺激的抑制下，臉孔的吸引力與臉孔倒置效應（face inversion effect，FIE，即相對於倒立臉孔，正立臉孔的檢測時間更快）的強度呈正相關，而且該相關的強度在自己種族臉孔與其他種族臉孔之間並無顯著差異。這些實驗結果顯示，臉孔種族和吸引力均可被無意識地處理，而且這兩種屬性的處理並無交互作用、不相互影響。在實驗三和四的腦電波實驗中，我們從有意識情況下的N170成分大小觀察到臉孔種族和臉孔呈現方向之間存在交互作用，這表示臉孔配置（face configuration）的編碼受到臉孔種族的調節。然而，我們並沒有觀察到清楚的ERP證據可反映在無意識情況下對臉孔吸引力的處理。
總結來說，本論文的行為結果顯示，臉孔的吸引力和種族屬性都可以迅速且無意識地被處理。然而，我們的神經生理證據僅揭露了在早期臉孔知覺歷程中，臉孔種族在潛意識和有意識情況下的效果，但沒有清楚的ERP訊號可反映臉孔吸引力在無意識狀態下的處理歷程。為了解釋這些發現，我們提出了一個理論模型，說明在處理自己種族和其他種族的正立、倒立臉孔時，會影響N170成分強度的因素，以及視覺遮蔽和CFS如何阻斷視覺通路中的神經處理機制。
;Facial attractiveness reflects the evolutionary value of individuals. Empirical studies have not only demonstrated the general components contributing to facial attractiveness, but also indicated that attractiveness preferences are shaped by life experiences. However, previous studies did not find an own-race advantage in explicit attractiveness judgments of faces. In contrast, a robust other-race effect (ORE) in the encoding and recognition of faces has been consistently shown, indicating better discrimination and memorability for own-race faces compared to other-race faces. Given the absence of the ORE in explicit judgments of facial attractiveness, we aimed to investigate whether the perception of the race of faces affects perceived attractiveness in implicit judgments without conscious face awareness.
Both the perception of facial attractiveness and face race have been suggested to be processed automatically and unconsciously in behavioral studies using the visual masking and continuous flash suppression (CFS) paradigms. In the present study, we employed the two paradigms, in combination with simultaneous recordings of event-related potentials (ERPs), to investigate the perception of face race and attractiveness during early face processing in both subliminal/unconscious and supraliminal/conscious processing stages. In the visual masking paradigm, face stimuli are presented between forward and backward masks in order to prevent the processing of visual information from face stimuli before and after the face presentation. In the CFS paradigm, the CFS stimulus presented to the dominant eye renders the face stimulus presented to the other eye invisible at the beginning of a trial, and the duration the face needs to break into awareness is measured by the reaction time (RT) of face detection.
We first conducted two behavioral experiments to examine the relative processing speed of the race and attractiveness of faces for these two attributes to reach conscious stages during short face exposures of 100 ms or less in the visual masking paradigm (Experiment 1), and whether the two face attributes are unconsciously processed and interact with each other in the CFS paradigm (Experiment 2). Based on these two experiments, we further conducted two neurophysiological experiments (Experiments 3 & 4) to examine whether the effects of face race and attractiveness on behavioral responses arise from early face perception in both subliminal and supraliminal conditions.
In Experiment 1, participants showed sensitivity to the race and attractiveness of faces when the face stimulus was presentation for 33.34 ms, suggesting efficient processing of face race and attractiveness in explicit judgments. In Experiment 2, we found a significant difference on the face detection time between own-race and other-race faces under the CFS procedure. Also, the attractiveness of faces was positively correlated with the face inversion effect (FIE, the faster detection time for faces in the upright than in the inverted orientation) under the CFS, while this correlation was not significantly different between own-race and other-race faces. These face detection results suggest that face race and attractiveness are processed unconsciously, while the two types of processing do not interact with each other. In Experiments 3 and 4, we observed an interaction effect between face race and orientation on the N170 component under supraliminal conditions, suggesting that the encoding of face configuration is modulated by face race. However, we did not have strong ERP evidence indicating unconscious processing of facial attractiveness in subliminal conditions.
In summary, our behavioral findings suggest that both the attractiveness and race of faces are processed rapidly and unconsciously. However, our ERP findings in early face perception only showed the effects of face race in both subliminal and supraliminal conditions. We did not have strong neurophysiological evidence to suggest that facial attractiveness is processed in early face perception without conscious awareness. To account for these findings, we proposed a model which specifies the factors that affect the magnitude of the N170 component across unconscious and conscious levels of face processing when encountering own-race and other-race upright and inverted faces, and how visual masking and CFS may interrupt the neural mechanisms in visual pathways. |
