We examined whether self-biases in perceptual matching reflect the positive valence of self-related stimuli. Participants associated geometric shapes with either personal labels (e.g., you, friend, stranger) or faces with different emotional expressions (e.g., happy, neutral, sad). They then judged whether shape-label or shape-face pairs were as originally shown or re-paired. Match times were faster to self-associated stimuli and to stimuli associated with the most positive valence. In addition, both the self-bias and the positive emotion bias were reliable across individuals in different test sessions. In contrast there was no sign of a correlation between the self-bias and the emotion-bias effects. We argue that self-bias and the bias to stimuli linked to positive emotion are separate and may reflect different underlying processes.

No abstract

Although theoretical discourse and experimental studies on the self- and reward-biases have a long tradition, currently we have only a limited understanding of how the biases are represented in the brain and, more importantly, how they relate to each other. We used multi-voxel pattern analysis to test for common representations of self and reward in perceptual matching in healthy human subjects. Voxels across an anterior–posterior axis in ventromedial prefrontal cortex (vmPFC) distinguished (i) self–others and (ii) high–low reward, but cross-generalization between these dimensions decreased from anterior to posterior vmPFC. The vmPFC is characterized by a shift from a common currency for value to independent, distributed representations of self and reward across an anterior–posterior axis. This shift reflected changes in functional connectivity between the posterior part of the vmPFC and the frontal pole when processing self-associated stimuli, and the middle frontal gyrus when processing stimuli associated with high reward. The changes in functional connectivity were correlated with behavioral biases, respectively, to the self and reward. The distinct representations of self and reward in the posterior vmPFC are associated with self- and reward-biases in behavior.

Linking arbitrary shapes (e.g., circles, squares, and triangles) to personal labels (e.g., self, friend, or stranger) or reward values (e.g., £18, £6, or £2) results in immediate processing benefits for those stimuli that happen to be associated with the self or high rewards in perceptual matching tasks. Here we further explored how social and reward associations interact with multisensory stimuli by pairing labels and objects with tones (low, medium, and high tones). We also investigated whether self and reward biases persist for multisensory stimuli with the label removed after an association had been made. Both high reward stimuli and those associated with the self, resulted in faster responses and improved discriminability (i.e., higher d’), which persisted for multisensory stimuli even when the labels were removed. However, these self- and reward-biases partly depended on the specific alignment between the physical tones (low, medium, and high) and the conceptual (social or reward) order. Performance for reward associations improved when the endpoints of low or high rewards were paired with low or high tones; meanwhile, for personal associations, there was a benefit when the self was paired with either low or high tones, but there was no effect when the stranger was associated with either endpoint. These results indicate that, unlike reward, social personal associations are not represented along a continuum with two marked endpoints (i.e., self and stranger) but rather with a single reference point (the self vs. other).

In their seminal paper 'Is our self nothing but reward', Northoff and Hayes (Biol Psychiatry 69(11):1019-1025, Northoff, Hayes, Biological Psychiatry 69(11):1019-1025, 2011) proposed three models of the relationship between self and reward and opened a continuing debate about how these different fields can be linked. To date, none of the proposed models received strong empirical support. The present study tested common and distinct effects of personal relevance and reward values by de-componenting different stages of perceptual decision making using a drift-diffusion approach. We employed a recently developed associative matching paradigm where participants (N = 40) formed mental associations between five geometric shapes and five labels referring personal relevance in the personal task, or five shape-label pairings with different reward values in the reward task and then performed a matching task by indicating whether a displayed shape-label pairing was correct or incorrect. We found that common effects of personal relevance and monetary reward were manifested in the facilitation of behavioural performance for high personal relevance and high reward value as socially important signals. The differential effects between personal and monetary relevance reflected non-decisional time in a perceptual decision process, and task-specific prioritization of stimuli. Our findings support the parallel processing model (Northoff & Hayes, Biol Psychiatry 69(11):1019-1025, Northoff, Hayes, Biological Psychiatry 69(11):1019-1025, 2011) and suggest that self-specific processing occurs in parallel with high reward processing. Limitations and further directions are discussed.