Date of Degree


Document Type


Degree Name





Eric Fertuck

Committee Members

Diana Diamond

Robert Melara

Lissa Weinstein

Kevin Meehan

Subject Categories

Clinical Psychology | Mental Disorders | Other Psychiatry and Psychology | Psychiatric and Mental Health | Psychiatry | Psychoanalysis and Psychotherapy | Psychological Phenomena and Processes


self and other differentiation, personality disorders, narcissistic personality, borderline personality, psychopathology, perception, mental representations, psychophysics, personality pathology


Self-Other Differentiation (SOD) refers to a developmental process of acquiring a consolidated, integrated, and individuated sense of self. SOD develops at a) perceptual (e.g., facial perception) and b) representational (e.g., traits, mental states, and beliefs) levels. Impairments in representational SOD (R-SOD) are associated with many forms of psychopathology, particularly borderline personality disorder (BPD) and narcissistic personality disorder (NPD). Few studies to date have examined the perceptual aspects of SOD (P-SOD), which are hypothesized to develop from infancy onwards in tandem and in interaction with R-SOD. Given that the human face is one of the key characteristics that humans use to identify themselves and others, we developed a novel method using facial stimuli of self and other to assess SOD by way of a facial-morphing task (FM-SOD).

Our study aim was to validate the FM-SOD task by assessing criterion, convergent and divergent validity, and to examine whether subjects with SOD impairments (e.g., participants with pathological narcissism and BPD features) differ in terms of their SOD on their perceptual and cognitive responses (i.e., sensitivity, discriminability, and response range). Undergraduates (N=87, 38% female) appraised a series of facial images, which comprised features of the self and other in varying degrees (from 0% to 100%, with 0% being “no morphing of self”) on the FM-SOD task. They made self/other appraisals on randomly ordered presentations of these morphs on a Likert scale ranging from 1 (100% other) to 5 (100% self). The task measures sensitivity (i.e. a propensity to see the self in the other), discriminability (i.e. the ability to make finer distinctions between self relative to other), and response range.

Our results did not provide support for criterion validity of FM-SOD by utilizing the Differentiation and Relatedness Scale (DR-S). However, we showed that individuals with lower developmental quality with respect to self-other differentiation—based on their descriptions of their mothers on the Object Relations Inventory (ORI)—provided a constricted response range in the FM-SOD task. This suggests impairments in R-SOD manifest as a lack of cognitive flexibility and black and white response patterns in perceptual tasks. In terms of construct validity, we found that participants characterized by lower levels of personality organization and higher levels of identity diffusion and who employ lower level defenses (all three as assessed by the IPO) showed poorer sensitivity (i.e. a greater propensity to “see” themselves on facial morphs. These participants’ response range while rating facial morphs was also constricted. Similarly, there was a higher likelihood that participants whose self-esteem was inordinately affected by others’ opinion and who reported poorly differentiated sense of self [assessed by Self Other Differentiation scale (SOD)] see their features more than others in the facial morphs.

Our participants did not show any perceptual deficits in discriminating between morphs of self and other. However, our findings lend some support to convergent validity of FM-SOD particularly in terms of sensitivity, suggesting a link between R-SOD and P-SOD on self and other facial recognition. Our results demonstrated divergent validity of the FM-SOD task showing that discriminability, sensitivity and response range on the FM-SOD task measure constructs other than self-esteem or mood states. However, sensitivity on the FM-SOD task was found to be correlated with depression scores on the Center for Epidemiologic Studies Depression Scale (CESD). This might be due to the fact that in a nonclinical population, it is hard to differentiate between depression and specific pathology that would impact sensitivity parameters of the FM-SOD task. Lastly, in terms of psychopathology, BPD features were found to be significantly correlated with constricted response range and impaired sensitivity. Specifically, we found that participants with higher BPD features on the Zanarini – BPD scale (ZAN-BPD) showed a more constricted response range in differentiating themselves from others and further showed a higher likelihood of seeing themselves compared to others in facial morphs.Our results showed that perceptual sensitivity, but not perceptual discriminability or constricted response range, was associated with greater pathological narcissistic features. Lastly, individuals with high scores on narcissism, particularly vulnerable narcissism, assessed by the PNI-52 were more likely to rate the facial morphs similar to themselves. This study allowed us to validate a measure which assesses perceptual (as compared to representational) aspects of SOD in an adult population. With this study, we developed a methodology to investigate R-SOD and P-SOD impairments in a nonclinical population, as well as their relationship.