The complexity and sophistication of human learning are increasingly appreciated. Enduring theoretical models illustrate that learners utilise ‘prediction errors’ to refine their predictions of future states (e.g., Rescorla–Wagner [RW] and temporal difference models; O’Doherty et al., 2003; Rescorla and Wagner, 1972; Schultz et al., 1997; Sutton and Barto, 2018). An explosion of studies, however, illustrates that this simple mechanism lies at the heart of more complex and sophisticated systems that enable humans (and other species) to learn from, keep track of the utility of, and integrate information from, multiple learning sources (Behrens et al., 2009; Biele et al., 2009; Li et al., 2011), meaning that one can learn from many sources of information simultaneously (Daw et al., 2006). Such complexity enables individuals to, for example, rank colleagues according to the utility of their advice and learn primarily from the top-ranked individual (Kendal et al., 2018; Laland, 2004; Morgan et al., 2012; Rendell et al., 2011) whilst also tracking the evolving utility of advice from others (Behrens et al., 2008; Biele et al., 2011). Recent studies have further revealed that learning need not rely solely on directly experienced associations since one can also learn via inference (Bromberg-Martin et al., 2010; Dolan and Dayan, 2013; Jones et al., 2012; Langdon et al., 2018; Moran et al., 2021; Sadacca et al., 2016; Sharpe and Schoenbaum, 2018). This growing appreciation of the complexity and sophistication of human learning may help to explain contradictory findings in various fields. Here, we focus on the field of social learning.

The existence in the human brain of neural and/or neurochemical pathways that are specialised for learning from social information and individual experience respectively is the topic of much debate (Heyes, 2012; Heyes and Pearce, 2015). Indeed, the claim that humans have social-specific learning mechanisms that are adaptive specialisations moulded by natural selection to cope with the pressures of group living lies at the heart of some theories of cultural evolution (Kendal et al., 2018; Morgan et al., 2012; Templeton et al., 1999). Since cultural evolution is argued to be specific to humans (Richerson and Boyd, 2005), establishing whether humans do indeed possess social-specific learning mechanisms has attracted many scholars with its promise of elucidating the key ingredient that ‘makes us human’.

Cognitive neuroscience offers tools that are ideally suited to investigating whether the mechanisms underpinning social learning (learning from others) do indeed differ from those that govern learning from one’s individual experience (individual learning). Cognitive neuroscientific studies, however, present mixed evidence for social-specific learning mechanisms. Some studies find dissociable neural correlates for social and individual learning (Apps et al., 2016; Behrens et al., 2008; Hill et al., 2016; Zhang and Gläscher, 2020). For example, a study by Behrens et al., 2008 reported that whilst individual learning was associated with activity in dopamine-rich regions such as the striatum that are classically associated with reinforcement learning, social learning was associated with activity in a dissociable network that instead included the anterior cingulate cortex gyrus (ACCg) and temporoparietal junction. Further supporting this dissociation, studies have revealed correlations between personality traits, such as social dominance (Cook et al., 2014) and dimensions of psychopathy (Brazil et al., 2013) and social, but not individual, learning, as well as atypical social, but not individual, prediction error-related signals in the ACCg in autistic individuals (Balsters et al., 2017). Together, these studies support the existence of social-specific learning mechanisms. In contrast, other studies have reported that the same computations, based on the calculation of prediction error, are involved in both social and individual learning (Diaconescu et al., 2014), and that social learning is associated with activity in dopamine-rich brain regions typically linked to individual learning (Biele et al., 2009; Braams et al., 2014; Campbell-Meiklejohn et al., 2010; Delgado et al., 2005; Diaconescu et al., 2017; Klucharev et al., 2009). Diaconescu et al., 2017, for example, observed that social learning-related prediction errors covaried with naturally occurring genetic variation that affected the function of the dopamine system. Further supporting this overlap between social and individual learning, behavioural studies have observed that social and individual learning are subject to the same contextual influences. For example, Tarantola et al., 2017 observed that prior preferences bias social learning, just as they do individual learning. Such findings promote the view that ‘domain-general’ learning mechanisms underpin social learning: we learn from other people in the same way that we learn from any other stimulus in our environment (Heyes, 2012; Heyes and Pearce, 2015). That is, there are no social-specific learning mechanisms.

One potential resolution to this conflict in the literature hinges on (1) an appreciation of the complexity and sophistication of human learning systems and (2) a difference in study design between tasks that have, and have not, found evidence of social-specific mechanisms. In studies that have linked social learning with the dopamine-rich circuitry typically associated with individual learning (and which are therefore consistent with the domain-general view), participants have been encouraged to learn primarily from social information. Indeed, in many cases the social source has been the sole information source (Campbell-Meiklejohn et al., 2010; Diaconescu et al., 2017; Klucharev et al., 2009). For example, in the paradigm employed by Diaconescu and colleagues (2014, 2017), participants were required to choose between a blue and green stimulus and were provided with social advice which was sometimes valid and sometimes misleading; on each trial, participants received information about the time-varying probability of reward associated with the blue and green stimuli, thus participants did not have to rely on their own individual experience of blue/green reward associations and could fully dedicate themselves to social learning. That is, participants did not learn from multiple sources (i.e., social information and individual experience); participants only engaged in social learning. In contrast, in studies where social learning has been associated with neural correlates outside of the dopamine-rich regions classically linked to individual learning (and which are therefore consistent with the domain-specific view), social information has typically comprised a secondary, additional source (Behrens et al., 2008; Cook et al., 2014). Typically, the non-social (individual) information is presented first to participants, represented in a highly salient form, and is directly related to the feedback information. The social information, in contrast, is presented second, is typically less salient in form, and is not directly related to the feedback information. For example, in the Behrens et al. study (2008) (and in our own work employing this paradigm; Cook et al., 2014; Cook et al., 2019), participants were required to choose between two, highly salient, blue and green boxes to accumulate points. The boxes were the first stimuli that participants saw on each trial. Outcome information came in the form of a blue or green indicator, thus primarily informing participants about whether they had made the correct choice on the current trial (i.e., if the outcome indicator was blue, then the blue box was correct). In addition, each trial also featured a thin red frame, which represented social information, surrounding one of the two boxes. The red frame was the second stimulus that participants saw on each trial and indirectly informed participants about the veracity of the frame: if the outcome was blue and the frame surrounded the blue box, then the frame was correct. In such paradigms, participants must learn from multiple sources of information with one source taking primary status over the other. Consequently, in studies that have successfully dissociated social and individual learning the two forms of learning differ both in terms of social nature (social or non-social) and rank (primary versus secondary status). Thus, it is unclear which of these two factors accounts for the dissociation.

This study tests whether social and individual learning share common neurochemical mechanisms when they are matched in terms of (primary versus secondary) status. Given its acclaimed role in learning (Glimcher and Bayer, 2005; Schultz, 2007), we focus specifically on the role of the neuromodulator dopamine. Drawing upon recent studies illustrating the complexity and sophistication of human learning (Daw et al., 2005; Gläscher et al., 2011; Moran et al., 2021), we hypothesise that pharmacological modulation of the human dopamine system will dissociate learning from two sources of information along a primary versus secondary, but not along a social versus individual axis. In other words, we hypothesise that social learning relies upon the dopamine-rich mechanisms that also underpin individual learning when social information is the primary source, but not when it comprises a secondary, additional element. Such a finding would offer a potential resolution to the aforementioned debate concerning the existence of social-specific learning mechanisms.

Preliminary support for our hypothesis comes from three lines of work. First, studies have convincingly argued for flexibility within learning systems. For example, in a study by Daw et al., 2006, participants tracked the utility of four uncorrelated bandits, with particular brain regions – such as the ventromedial prefrontal cortex – consistently representing the value of the top-ranked bandit, even though the identity of this bandit changed over time. Second, studies are increasingly illustrating the flexibility of social brain networks (Ereira et al., 2020; Garvert et al., 2015). The medial prefrontal cortex (mPFC), for example, is not – as was once thought – specialised for representing the self; if the concept of ‘other’ is primarily relevant for the task at hand, then the mPFC will prioritise representation of other over self (Cook, 2014; Nicolle et al., 2012). Finally, in a recent study (Cook et al., 2019), we provided preliminary evidence of a catecholaminergic (i.e., dopaminergic and noradrenergic) dissociation between learning from primary and secondary, but not social and individual, sources of information. In this work (Cook et al., 2019), we employed a between-groups design, wherein both groups completed a version of the social learning task adapted from Behrens et al., 2008 described above. For one group, the secondary source was social in nature (social group). For the non-social group, the secondary source comprised a system of rigged roulette wheels and was thus non-social in nature. We observed that, in comparison to placebo (PLA), the catecholaminergic transporter blocker methylphenidate only affected learning from the primary source, which, in this paradigm, always comprised participant’s own individual experience. Methylphenidate did not affect learning from the secondary source, irrespective of its social or non-social nature. That is, we found positive evidence supporting a dissociation between primary and secondary learning but no evidence to support a distinction between learning from social and non-social sources. Nevertheless, since we did not observe an effect of methylphenidate on learning from the (social or non-social) secondary source of information, this study was unable to provide positive evidence of shared mechanisms for learning from social and non-social sources. If it is truly the case that domain-general (neurochemical) mechanisms underpin social learning, it should follow that pharmacological manipulations that affect individual learning when individual information is the primary source also affect social learning when social information is the primary source.

The current (pre-registered) experiment tested this hypothesis by orthogonalising social versus individual and primary versus secondary learning. We perturbed learning using the dopamine D2 receptor antagonist haloperidol (HAL), in a double-blind, counter-balanced, PLA-controlled design. To test whether pharmacological manipulation of dopamine dissociates learning along a primary-secondary and/or a social-individual axis, we developed a novel between-groups manipulation wherein one group of participants learned primarily from social information and could supplement this learning with their own individual experience, and a second group learned primarily from individual experience and could supplement this learning with socially learned information. To foreshadow our results, we demonstrate that HAL specifically affects learning from the primary (not secondary) source of information. Bayesian statistics confirmed that the effects of haloperidol were comparable between the groups, thus, HAL affected individual learning when individual information was the primary source and, to the same extent, social learning when social information was the primary source. Our data support an expanding field showing that, rather than being fixedly specialised for particular inputs, neurochemical pathways in the human brain can process both social and non-social cues and arbitrate between the two depending upon which cue is primarily relevant for the task at hand (Cook, 2014; Garvert et al., 2015; Nicolle et al., 2012).

Participants (n = 43; aged 19–38, mean [standard error] ${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 25.950 [0.970]; 24 males, 19 females; see Materials and methods) completed an adapted version of the behavioural task originally developed by Behrens et al., 2008. Participants were randomly allocated to one of two groups. Participants in the individual-primary group (n = 21) completed the classic version of this task (Figure 1A; Behrens et al., 2008) in which they were required to make a choice between a blue and green box in order to win points. A red frame (the social information), which represented the most popular choice made by a group of four participants who had completed the task previously, surrounded either the blue or green box on each trial, and participants could use this to help guide their choice. The actual probability of reward associated with the blue and green boxes and the probability that the red frame surrounded the correct box varied according to uncorrelated pseudo-randomised schedules (Appendix 2—figure 1). For the individual-primary group, the individual information (blue and green stimuli) was primary, and the social information (red stimulus) was secondary on the basis that the blue/green stimuli appeared first on the screen, were highly salient (large boxes versus a thin frame) and were directly related to the feedback information. That is, after making their selection, participants saw a small blue or green box which primarily informed them whether a blue or green choice had been rewarded on the current trial. From this information, the participant could, secondarily, infer whether the social information (red frame) was correct or incorrect.

Figure 1

Download asset Open asset

Our social-primary group (n = 22; groups matched on age, gender, body mass index [BMI], and verbal working memory [VWM] span; Table 1) completed an adapted version of this task (Figure 1B) wherein the social information (red stimulus) was primary and the individual information (blue/green stimuli) was secondary. Participants first saw two placeholders; one empty and one containing a red box which indicated the social information. Subsequently, a thin green and a thin blue frame appeared around each placeholder. Participants were told that the red box represented the group’s choice. They were then required to choose whether to go with the social group (red box) or not. After making their choice, a tick or cross appeared which primarily informed participants whether going with the social information was the correct option. From this they could, secondarily, infer whether the blue or green frame was correct. Consequently, for the social-primary group the social information was primary on the basis that it appeared first on the screen, highly salient (a large red box versus thin green/blue frames), and directly related to the feedback information.

Participants in both the individual-primary and social-primary groups performed 120 trials of the task on each of two separate study days. To perturb learning, on one day participants took 2.5 mg of HAL, previously shown to affect learning (Pessiglione et al., 2006) via multiple routes including perturbation of phasic dopamine signalling (Schultz, 2007; Schultz et al., 1997) facilitated by action at mesolimbic D2 receptors (Camps et al., 1989; Grace, 2002; Lidow et al., 1991). On the other day, they took a PLA under double-blind conditions, with the order of the days counterbalanced. 43 participants took part in at least one study day, 33 participants completed both study days. Two participants performed at below-chance-level accuracy and were excluded from further analysis. We present an analysis of data from the 31 participants who completed both study days with above-chance accuracy (Table 1) in this article, which we complement with a full analysis of all 41 datasets in Appendix 4i.

We used the following strategy to analyse our data. First, we sought to validate our manipulation by testing (under PLA) whether participants in both the individual-primary and social-primary groups learned in a more optimal fashion from the primary, versus secondary, source of information. Next, we tested our primary hypothesis that both social and individual learning would be modulated by HAL when they are the primary source of learning, but not when they comprise the secondary source. To do so, we estimated learning rates for primary and secondary sources of information, for each group (social-primary, individual-primary), under HAL and PLA, by fitting an adapted RW learning model to choice data. To ascertain that our model accurately described choices, we used simulations and parameter recovery. We used random-effects Bayesian model selection (BMS) to compare our model with alternative models. These analyses provided confidence that our model accurately described participants’ behaviour. After testing our primary hypothesis, we explored the relationship between parameters from our computational model and performance. To accomplish this, we first used an optimal learner model, with the same architecture and priors as our adapted RW model, to assess the extent to which HAL made participants’ learning rates more (or less) optimal. Finally, we regressed estimated model parameters against accuracy to gain insight into the extent to which variation in these parameters (and the effect of the drug thereupon) contributed to correct responses on the task.

Social information is the primary source of learning for participants in the social-primary group

Our novel manipulation orthogonalised primary versus secondary and social versus individual learning. To validate our manipulation, we tested whether participants in both the individual-primary and social-primary group learned in a more optimal fashion from the primary versus secondary source of information in our PLA condition. For this validation analysis, we used a Bayesian learner model to create two optimal models: (1) an optimal primary learner and (2) an optimal secondary learner (Materials and methods). Subsequently, we regressed both models against participants’ choice data, resulting in two β_optimal values capturing the extent to which a participant made choices according to the optimal primary, and optimal secondary learner models, respectively. β_optimal values were submitted to a repeated-measures analysis of variance (RM-ANOVA) with factors information source (primary, secondary) and group (social-primary, individual-primary), revealing main effects of information source (F(1,29) = 6.594, p=0.016) and group (F(1,29) = 10.423, p=0.003). β_optimal values (averaged across individual-primary and social-primary groups) were significantly higher for the primary information (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.872 (0.101)) compared with secondary information source (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.438 (0.101); t(29) = 2.568, p_holm = 0.016). β_optimal values (averaged across primary and secondary conditions) were significantly higher for the social-primary group (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.833 (0.078)) compared with the individual-primary group (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.477 (0.078); t(29) = 3.228, p_holm = 0.003) (Figure 2). Crucially, we did not observe a significant interaction between information and group (F(1,29) = 0.067, p=0.797), meaning that participants’ choices were more influenced by the primary information source, regardless of whether it was social or individual in nature. Furthermore, β_optimal values for primary information alone did not significantly differ between groups (t(29) = –1.982, p_holm = 0.257). Note that β_optimal weights for both information sources were significantly greater than zero (primary: t(30) = 7.534, p<0.001; secondary: t(30) = 4.789, p<0.001), thus our optimal models of information use explained a significant amount of variance in the use of both primary and secondary learning sources. These data show that, irrespective of social (or individual) nature, participants learned in a more optimal fashion from the primary (relative to secondary) learning source, which was first in the temporal order of events, highly salient and directly related to the reward feedback.

Figure 2

Download asset Open asset

Haloperidol reduces the rate of learning from primary sources

We hypothesised that both social and individual learning would be modulated by administration of the dopamine D2 receptor antagonist HAL when they were the primary source of learning, but not when they comprised the secondary source. To test this hypothesis, we fitted an adapted RW learning model (Rescorla and Wagner, 1972) to participants’ choice data, enabling us to estimate various parameters that index learning from primary and secondary sources of information, for HAL and PLA conditions, for participants in the social-primary and individual-primary groups. Our adapted RW model provided estimates, for each participant, of $\alpha ,\mathrm{\beta },\mathrm{}\mathrm{a}\mathrm{n}\mathrm{d}\mathrm{}\mathrm{\zeta }.$ The learning rate (α) controls the weighting of prediction errors on each trial. A high $\alpha$ favours recent over (outdated) historical outcomes, while a low $\alpha$ suggests a more equal weighting of recent and more distant trials. Since our pseudo-random schedules included stable phases (where the reward probability associated with a particular option was constant for >30 trials), and volatile phases (where reward probabilities changed every 10–20 trials), α was estimated separately for volatile and stable phases (for both primary and secondary learning) to accord with previous research (Behrens et al., 2007; Cook et al., 2019; Manning et al., 2017). β captures the extent to which learned probabilities determine choice, with a larger β meaning that choices are more deterministic with regard to the learned probabilities. ζ represents the relative weighting of primary and secondary sources of information, with higher values indicating a bias towards the over-weighting of secondary relative to primary (see Materials and methods and Appendix 3 for further details of the model, model fitting, and model comparison).

We hypothesised an interaction between drug and (primary versus secondary) information source such that HAL would affect learning from the primary information source only, regardless of its social/individual nature. To test this hypothesis, we employed a linear mixed effects model with fixed factors information source (primary, secondary), drug (HAL, PLA), environmental volatility (volatile, stable), and group (social-primary, individual-primary) and dependent variable α (square-root transformed to meet assumptions of normality). We controlled for inter-individual differences by including random intercepts for subject. Including pseudo-randomisation schedule as a factor in all analyses did not change the pattern of results. The mixed model revealed a drug by information interaction (F(1, 203) = 6.852, p=0.009, beta estimate (${\sigma }_{\stackrel{-}{x}}$) = 0.026 (0.010), t = 2.62, confidence interval [CI] = [0.010–0.050]) (Figure 3). There were no significant main effects of drug (F(1, 203) = 0.074, p=0.786), group (F(1, 29) = 3.148, p=0.087), or volatility (F(1, 203) = 1.470, p=0.227) on $\alpha$ values, nor any other significant interactions involving drug (all p-values>0.05, see Appendix 4v–vi for analysis, including schedule, session, and working memory). Planned contrasts showed that, whilst under PLA, α_primary (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.451 (0.025), collapsed across volatility and group) was significantly greater than α_secondary (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.370 (0.025); z(30) = 2.861, p=0.004); this was not the case under HAL (α_primary ${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.393 (0.025), α_secondary ${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.417 (0.025); z(30) = –0.843, p=0.400). Furthermore, α_primary was decreased under HAL relative to PLA (z(30) = –2.050, p=0.040). Although α_secondary was, in contrast, numerically increased under HAL (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.417 (0.025)) relative to PLA (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.370 (0.025)), this difference was not significant (z(30) = 1.654, p=0.098). This drug × information interaction therefore illustrated that whilst HAL significantly reduced α_primary it had no significant effect on α_secondary. Furthermore, under PLA there was a significant difference between α_primary and α_secondary, which was nullified by HAL administration. Consequently, under PLA participants’ rate of learning was typically higher for learning from the primary relative to the secondary source; however, under the D2 receptor antagonist HAL the rate of learning from the primary source was reduced and thus there was no significant difference in the rate of learning from primary and secondary sources.

Figure 3

Download asset Open asset

Linear mixed models, with fixed factors group and drug, and random intercepts for subject, were also used to explore drug effects on ζ values (representing the relative weighting of primary/secondary information) and $\mathrm{\beta }$ values. For ζ, there were no significant main effects of drug (F(1, 29) = 1.941, p=0.174, beta estimate ( ${\sigma }_{\stackrel{-}{x}}$)= −0.07 (0.050), t = −1.390, CI = [–0.170 to 0.003]) or group (F(1, 51) = 0.184, p=0.669, beta estimate(${\sigma }_{\stackrel{-}{x}}$ )=0.020 (0.040), t = 0.430, CI = [–0.070 to 0.100]), nor drug by group interaction (F(1, 29) = 0.039, p=0.845, beta estimate(${\sigma }_{\stackrel{-}{x}}$ )=−0.001 (0.050), t = −0.200, CI = [-0.110 to 0.090]). Similarly, our analysis of $\mathrm{\beta }$ values revealed no main/interaction effect(s) of drug, group, or drug by group (all p>0.05).

Haloperidol brings ${\displaystyle \alpha }$_primary estimates within the optimal range

To assess whether the effects of HAL on α_primary are harmful or beneficial with respect to performance, we first explored drug effects on accuracy (see Appendix 4ii for a detailed analysis including randomisation schedule). There was no significant difference in accuracy between HAL (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.600 (0.013)) and PLA (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.611 (0.010); F(1,29) = 0.904, p=0.349, η_p² = 0.030) conditions.

The lack of a significant main effect of drug on accuracy was somewhat surprising given the significant (interaction) effect on learning rates, that is, a decrease in α_primary under HAL relative to PLA. To investigate whether HAL resulted in learning rates that were less, or alternatively, more, optimal, we compared our estimated $\alpha$ values with optimal $\alpha$ estimates. Since trial-wise outcomes were identical to those utilised by Cook et al., 2019, optimal values are also identical and are described here for completeness. An optimal learner model, with the same architecture and priors as the model employed in the current task, was fit to 100 synthetic datasets, resulting in average optimal learning rates: α_{optimal_primary_stable} = 0.16, α_{optimal_primary_volatile} = 0.21, α_{optimal_secondary_stable} = 0.17, α_{optimal_secondary_volatile} = 0.19. Scores representing the difference between (untransformed) $\alpha$ estimates and optimal $\alpha$ scores were calculated (${\alpha }_{diff}$ = $\alpha -\alpha$_optimal). A linear mixed model analysis on ${\alpha }_{diff}$ values with factors group, drug, volatility and information source, and random intercepts for subject was conducted. A significant interaction between drug and information source was observed (F(1, 203) = 4.895, p=0.028, beta estimate (${\sigma }_{\stackrel{-}{x}}$) = 0.019 (0.010), t = 2.212, CI = [0.000–0.040]) (Figure 4). Planned contrasts showed that, for primary information, ${{\alpha }_{diff\_primary}}_{}$ was higher under PLA (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.052 (0.023)) compared with HAL (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.009 (0.028)); z(30) = 1.806, p=0.071). In contrast, ${{\alpha }_{diff\_secondary}}_{}$ was lower under PLA (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = −0.011 (0.023)) compared with HAL (${\displaystyle \overline{x}({\sigma }_{\overline{x}})}$ = 0.021 (0.021)); z(30) = 1.323, p=0.186. Learning rates for learning from the primary source were higher than optimal under PLA, with ${{\alpha }_{diff\_primary}}_{}$ significantly differing from 0 (one-sample t-test; t(30) = 2.259, p=0.031. HAL reduced learning rates that corresponded to learning from the primary source, thus bringing them within the optimal range, with ${{\alpha }_{diff\_primary}}_{}$ not significantly differing from 0 under HAL (one-sample t-test; t(30) = 0.319, p=0.752). Consequently, under HAL relative to PLA, learning rates for learning from primary sources were more optimal.

Figure 4

Download asset Open asset

To explore whether α values were in some way related to accuracy scores, we used two separate backward regression models, for PLA and HAL conditions separately, with α_primary and α_secondary as predictors and accuracy as the dependent variable (see Appendix 4iii for details of a regression model with all model parameters). PLA accuracy was predicted by α_secondary though this model only approached significance (R = 0.121, F(1,29) = 3.981, p=0.055). Under HAL, however, accuracy was predicted by a model with α_secondary and α_primary (R = 0.450, F(2,28) = 3.560, p=0.042), with α_primary a significant positive predictor of accuracy (β = 0.404, p=0.028). Removing α_secondary as a predictor did not significantly improve the fit of this model (R² change = 0.014, F change (1,29) = 0.495, p=1.000). When combined with our optimality analysis, these results suggest that under PLA α_primary was outside of the optimal range of α values and thus accuracy was primarily driven by α_secondary. However, HAL reduced α_primary, bringing it within the optimal range. Thus, under HAL accuracy was driven by both α_primary and α_secondary.

In sum, relative to PLA, the dopamine D2 receptor antagonist HAL significantly decreased learning rates relating to learning from primary, but not secondary sources of information, likely via mediation of phasic dopaminergic signalling (see Appendix 4iv). Interestingly, learning rates for learning from the primary source were higher than optimal under PLA and HAL brought them within the optimal range. Consequently, both primary and secondary learning contributed to accuracy under HAL but not under PLA. Importantly, the effects of HAL did not vary as a function of group allocation, which dictated whether the primary source was of social or individual nature. A Bayesian analysis confirmed that we had moderate evidence to support the conclusion that there was no interaction between drug, learning source and group. These data, thus, illustrate a dissociation along the primary-secondary but not social-individual axis.

This study tested the hypothesis that social and individual learning share common neurochemical mechanisms when they are matched in terms of (primary versus secondary) status. Specifically, we predicted that HAL would perturb learning from the primary but not the secondary source, irrespective of social or individual nature. Supporting our hypothesis, we observed an interaction between drug and information source (social versus individual) such that under HAL (compared to PLA) participants exhibited reduced learning rates with respect to learning from the primary, but not the secondary, source of information. Crucially, we did not observe an interaction between drug, information source, and group (social-primary versus individual-primary). Bayesian statistics revealed that, given the observed data, a model that excludes this interaction is 7.5 times more likely than models which include the interaction.

An important question concerns whether the lack of a dopaminergic dissociation between social and individual learning could be explained by participants not fully appreciating the social nature of the red shape (the social information source). In opposition to this, we argue that since our participants could not commence the task until reaching 100% accuracy in a pre-task quiz, which questioned participants about the social nature of the red shape, we can be confident that all participants knew that the red shape indicated information from previous participants. Participants also completed a post-task questionnaire (Appendix 5), which required them to reflect upon the extent to which their decisions were influenced by the social (red shape) and individual (blue/green shapes) information. If participants had not fully believed that the red shape represented social information, one might expect that they would indicate that they were not influenced by this source. In contrast, participants in both the individual-primary and social-primary groups believed that they were influenced by the red shape (as well as the blue/green stimuli). Furthermore, in our previous work, using the same social manipulation, we demonstrated that the personality trait social dominance significantly predicts social, but not individual, learning (Cook et al., 2014). Thus, illustrating that participants treat the social information differently from the non-social information in this type of paradigm. Finally, based on previous studies, we argue that even with a more overtly social manipulation it is highly likely that social learning would still be perturbed by dopaminergic modulation when social information is the primary source. Indeed, in a study by Diaconescu et al., 2017 social information was represented by a video of a person indicating one of the two options. Even with this overtly social stimulus, Diaconescu et al. still observed that social learning covaried with genetic polymorphisms that affect the functioning of the dopamine system.

The first part of our analysis illustrated that our manipulation produced the expected effect: when social information was first in the temporal order of events, highly salient and directly related to reward feedback participants learned in a more optimal fashion from this source of information. Such a result may be a surprise to some since one might think that, relative to learning from one’s own experience, learning from others will always take a ‘backseat’. Here, we clearly demonstrate that, when cast as the primary task, participants can make good use of social information. This paradigm may comprise a step towards developing a system to support accelerated social learning. Future studies could, for instance, investigate whether similar manipulations can be used to improve learning about (as opposed to from) other individuals. Since temporal order, saliency, and reward feedback were manipulated simultaneously, we cannot determine which manipulation is the most influential. Future work may therefore also seek to manipulate these factors independently to establish the most effective method for promoting social learning.

Our results comprise an important contribution to the debate concerning the existence of social-specific learning mechanisms. We find that, like individual learning, social learning is modulated by a dopaminergic manipulation when it is the primary source of information. This result marries well with previous studies that have linked social learning with dopamine-rich mechanisms when the social source has been the primary (or in many cases the sole) information source (Campbell-Meiklejohn et al., 2010; Diaconescu et al., 2017; Klucharev et al., 2009). Our results are also consistent with studies that have associated social learning with different neural correlates, outside of the dopamine-rich regions classically linked to individual learning, when it is a secondary source of information (Behrens et al., 2008; Hill et al., 2016; Zhang and Gläscher, 2020). Our data suggest that social and individual learning share common dopaminergic mechanisms when they are the primary learning source and that previous dissociations between these two learning types may be more appropriately thought of as dissociations between learning from a primary and secondary source. Extant studies (e.g., Cook et al., 2019) were not able to illustrate the importance of the primary versus secondary distinction because they did not fully orthogonalise primary versus secondary and social versus individual learning.

Though our results suggest shared neurochemical mechanisms for social and individual learning when they are matched in status, it is, nevertheless, essential to highlight that it does not follow that there are no dimensions along which social learning may be dissociated from individual learning. It is possible that although social and individual learning are affected by dopaminergic modulation – when they are the primary source – there are differences in the location of neural activity that could be revealed by neuroimaging. For instance, although social and individual learning are both associated with activity within the striatum (Burke et al., 2010; Cooper et al., 2012), social-specific activation patterns have been observed in other brain regions, including the temporoparietal junction (Behrens et al., 2008; Lindström et al., 2018) and the gyrus of the anterior cingulate cortex (Behrens et al., 2008; Hill et al., 2016; Zhang and Gläscher, 2020). Consequently, it is possible that HAL has comparable effects on social and individual learning but that these effects (seen at an ‘algorithmic level of analysis’, Lockwood et al., 2020) are associated with activity in different brain regions (i.e., dissociations at an ‘implementation level of analysis’, Lockwood et al., 2020). For example, HAL may comparably affect the BOLD signal associated with social and individual prediction errors, but the effect may be localised to dissociable neural pathways. Such a location-based dissociation requires further empirical investigation as well as further consideration of the possible functional significance of such location-based differences, if they are indeed present when primary versus secondary status is accounted for. Nevertheless, whilst such location-based differences are possible, we argue that they are not probable since, given different distributions of dopamine neurons, receptors, and reuptake mechanisms throughout the brain (Grace, 2002; Korn et al., 2021; Matsumoto et al., 2003; Sulzer et al., 2016), differences in location are relatively likely to result in differences in the magnitude of the effect of HAL (Wächtler et al., 2020; Yael et al., 2013). Additionally, since we did not observe significant effects of HAL on learning from social or individual sources when they were secondary in status, it remains a logical possibility that social and individual learning can be neurochemically dissociated when they are the secondary source of information – though it is admittedly difficult to conceive of a parsimonious explanation for the existence of two neurochemical mechanisms for social and individual learning from secondary sources. Finally, it is possible that social and individual learning share common dopaminergic mechanisms when they are the primary source, but differentially recruit other neurochemical systems. For instance, some have argued that social learning may heavily rely upon serotonergic mechanisms (Crişan et al., 2009; Frey and McCabe, 2020; Roberts et al., 2020). The abovementioned avenues should be further explored; however, in the interim, it must be concluded that since existing studies have not controlled for primary versus secondary status, we do not currently have convincing evidence that social and individual learning can be dissociated in the human brain.

Notably, our results reveal a clear dissociation between learning from primary and secondary sources. For learning from primary sources HAL made learning rates more optimal, HAL did not have this effect on learning rates for secondary learning. Interestingly, a combined optimality analysis and regression model suggested that, under PLA, learning rates for learning from the primary source were ‘too high’ and fell outside of the optimal range (for this specific task). Consequently, under PLA, variance in accuracy was primarily explained by learning rates for learning from the secondary source. However, HAL reduced learning rates for learning from the primary source, bringing them within the optimal range. Thus, under HAL, accuracy was driven by learning rates for learning from both the primary and secondary sources. An open question concerns whether HAL truly optimises or simply reduces learning rate. Since the current paradigm was not designed to test this hypothesis, a reduction in learning rate herein also corresponds to an optimisation of learning rate. To dissociate the two, one would need a paradigm that generates sufficient numbers of participants with learning rates (in the PLA condition) that are suboptimally low such that one can observe whether, in these critical test cases, HAL increases (i.e., optimises) learning rate.

An intriguing question concerns the synaptic mechanisms by which HAL affects learning rates. Non-human animal studies have shown that phasic signalling of dopaminergic neurons in the mesolimbic pathway encodes reward prediction error signals (Schultz, 2007; Schultz et al., 1997). Since HAL has high affinity for D2 receptors (Grace, 2002), which are densely distributed in the mesolimbic pathway (Camps et al., 1989; Lidow et al., 1991), dopamine antagonists including HAL can affect phasic dopamine signals (Frank and O’Reilly, 2006) – either via binding at postsynaptic D2 receptors (which blocks the effects of phasic dopamine bursts) or via presynaptic autoreceptors (which has downstream effects on the release and reuptake of dopamine and thus modulates bursting itself) (Benoit-Marand et al., 2001; Ford, 2014; Schmitz et al., 2003). That is, HAL may affect learning rate via blockade of the postsynaptic D2 receptors, which may mute the effects of phasic dopamine signalling (either directly or via reduction in the background tonic rate of activity which, in turn, reduces the amplitude of phasic responses; Belujon and Grace, 2015; Grace, 2016), thus reducing the weight of prediction error signals on value updating (i.e., reducing the learning rate). Indeed, a number of studies have shown that HAL can attenuate prediction error-related signals (Diederen et al., 2017; Haarsma et al., 2018; Menon et al., 2007; Pessiglione et al., 2006). For example, in the context of individual learning, Pessiglione et al., 2006 demonstrated that HAL attenuated prediction error signals in the striatum, indexed via changes in blood oxygen levels (BOLD). In addition to effects on postsynaptic D2 receptors, HAL may modulate prediction errors via effects on presynaptic autoreceptors. Autoreceptor binding is suggested to increase phasic bursting (Dugast et al., 1997; Frank and O’Reilly, 2006; Garris et al., 2003; Pehek, 1999), thus enhancing the phasic signal that is indicative of positive prediction errors. A combination of pre- and postsynaptic effects could feasibly result in more optimal learning rates wherein dopamine signalling is muted via postsynaptic blockade, thus muting (tonic background) ‘noise’ (and signal) but where the phasic ‘signal’ is enhanced via presynaptic effects, potentially resulting in an overall increased signal-to-noise ratio which may translate into more optimal learning rates.

Perhaps the most novel contribution of our work is that we here illustrate that, whilst dopaminergic modulation affects learning from the primary source, it does not significantly affect learning from the secondary source. Previous studies have illustrated that humans can learn – ostensibly simultaneously – from multiple sources of information and tend to organise this information in a hierarchical fashion such that the source which is currently of highest value has the greatest influence on a learner’s behaviour (Daw et al., 2006). Here, we extend this work by showing that the primary source, at the top of the hierarchy, is more heavily influenced by modulation of the dopamine system, thus suggesting a graded involvement of the dopamine system according to a source’s status in the ‘learning hierarchy’. Extant studies (Daw et al., 2006) suggest that such learning hierarchies are flexible and can be rapidly remodelled according to a source’s current value. The success of our orthogonalisation of social versus individual and primary versus secondary learning depended on a within-subjects design, wherein the status (primary or secondary) of the learning source varied only between participants. Although our study was therefore not optimised for studying the rapid remodelling of learning hierarchies, our results pave the way for future studies to investigate whether the impact of dopaminergic modulation of learning from a particular source quickly changes according to the source’s current status in the learning hierarchy.

In sum, in previous paradigms that dissociate social and individual learning, the social information comprised a secondary or additional information source, differing from individual information both in terms of its social nature (social/individual) and status (secondary/primary). We here provide evidence that dissociable effects of dopaminergic manipulation on different learning types are better explained by primary versus secondary status, than by social versus individual nature. Specifically, we showed that, relative to PLA, HAL reduced learning rates relating to learning from the primary, but not secondary, source of information irrespective of social versus individual nature. Results illustrate that social and individual learning share a common dependence on dopaminergic mechanisms when they are the primary learning source.

All raw data and analysis scripts can be accessed at the Open Science Framework data repository.

1. 1. Rybicki A
   2. Cook J

   (2021) Open Science Framework

   ID 398w4. Dopaminergic challenge dissociates learning from primary versus secondary sources of information.

   https://osf.io/398w4/

1. 1. Apps MAJ
   2. Rushworth MFS
   3. Chang SWC

   (2016) The anterior cingulate gyrus and social cognition: tracking the motivation of others

   Neuron 90:692–707.

   https://doi.org/10.1016/j.neuron.2016.04.018

   • PubMed
   • Google Scholar
2. 1. Bagby RM
   2. Taylor GJ
   3. Parker JD

   (1994) The Twenty-item Toronto Alexithymia Scale--II. Convergent, discriminant, and concurrent validity

   Journal of Psychosomatic Research 38:33–40.

   https://doi.org/10.1016/0022-3999(94)90006-x

   • PubMed
   • Google Scholar
3. 1. Balsters JH
   2. Apps MAJ
   3. Bolis D
   4. Lehner R
   5. Gallagher L
   6. Wenderoth N

   (2017) Disrupted prediction errors index social deficits in autism spectrum disorder

   Brain 140:235–246.

   https://doi.org/10.1093/brain/aww287

   • PubMed
   • Google Scholar
4. 1. Baron-Cohen S
   2. Wheelwright S
   3. Skinner R
   4. Martin J
   5. Clubley E

   (2001) The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians

   Journal of Autism and Developmental Disorders 31:5–17.

   https://doi.org/10.1023/a:1005653411471

   • PubMed
   • Google Scholar
5. Book

   1. Beck AT
   2. Steer RA
   3. Brown G

   (1996)

   Beck Depression Inventory-II

   APA PsycTests.

   • Google Scholar
6. 1. Behrens TEJ
   2. Woolrich MW
   3. Walton ME
   4. Rushworth MFS

   (2007) Learning the value of information in an uncertain world

   Nature Neuroscience 10:1214–1221.

   https://doi.org/10.1038/nn1954

   • PubMed
   • Google Scholar
7. 1. Behrens TEJ
   2. Hunt LT
   3. Woolrich MW
   4. Rushworth MFS

   (2008) Associative learning of social value

   Nature 456:245–249.

   https://doi.org/10.1038/nature07538

   • PubMed
   • Google Scholar
8. 1. Behrens TEJ
   2. Hunt LT
   3. Rushworth MFS

   (2009) The computation of social behavior

   Science (New York, N.Y.) 324:1160–1164.

   https://doi.org/10.1126/science.1169694

   • PubMed
   • Google Scholar
9. 1. Belujon P
   2. Grace AA

   (2015) Regulation of dopamine system responsivity and its adaptive and pathological response to stress

   Proceedings. Biological Sciences 282:1805.

   https://doi.org/10.1098/rspb.2014.2516

   • PubMed
   • Google Scholar
10. 1. Benoit-Marand M
    2. Borrelli E
    3. Gonon F

    (2001) Inhibition of dopamine release via presynaptic D2 receptors: Time course and functional characteristics in vivo

    Journal of Neuroscience 21:9134–9141.

    https://doi.org/10.1523/jneurosci.21-23-09134.2001

    • Google Scholar
11. 1. Bestmann S
    2. Ruge D
    3. Rothwell J
    4. Galea JM

    (2014) The role of dopamine in motor flexibility

    Journal of Cognitive Neuroscience 27:365–376.

    https://doi.org/10.1162/jocn_a_00706

    • Google Scholar
12. 1. Biele G
    2. Rieskamp J
    3. Gonzalez R

    (2009) Computational models for the combination of advice and individual learning

    Cognitive Science 33:206–242.

    https://doi.org/10.1111/j.1551-6709.2009.01010.x

    • Google Scholar
13. 1. Biele G
    2. Rieskamp J
    3. Krugel LK
    4. Heekeren HR

    (2011) The neural basis of following advice

    PLOS Biology 9:1001089.

    https://doi.org/10.1371/journal.pbio.1001089

    • Google Scholar
14. 1. Braams BR
    2. Güroğlu B
    3. de Water E
    4. Meuwese R
    5. Koolschijn PC
    6. Peper JS
    7. Crone EA

    (2014) Reward-related neural responses are dependent on the beneficiary

    Social Cognitive and Affective Neuroscience 9:1030–1037.

    https://doi.org/10.1093/scan/nst077

    • PubMed
    • Google Scholar
15. 1. Brazil IA
    2. Hunt LT
    3. Bulten BH
    4. Kessels RPC
    5. de Bruijn ERA
    6. Mars RB

    (2013) Psychopathy-related traits and the use of reward and social information: A computational approach

    Frontiers in Psychology 4:1–11.

    https://doi.org/10.3389/fpsyg.2013.00952

    • PubMed
    • Google Scholar
16. 1. Bromberg-Martin ES
    2. Matsumoto M
    3. Hong S
    4. Hikosaka O

    (2010) A pallidus-habenula-dopamine pathway signals inferred stimulus values

    Journal of Neurophysiology 104:1068–1076.

    https://doi.org/10.1152/jn.00158.2010

    • PubMed
    • Google Scholar
17. 1. Burke CJ
    2. Tobler PN
    3. Baddeley M
    4. Schultz W

    (2010) Neural mechanisms of observational learning

    PNAS 107:14431–14436.

    https://doi.org/10.1073/pnas.1003111107

    • PubMed
    • Google Scholar
18. 1. Campbell-Meiklejohn DK
    2. Bach DR
    3. Roepstorff A
    4. Dolan RJ
    5. Frith CD

    (2010) How the opinion of others affects our valuation of objects

    Current Biology 20:1165–1170.

    https://doi.org/10.1016/j.cub.2010.04.055

    • PubMed
    • Google Scholar
19. 1. Camps M
    2. Cortés R
    3. Gueye B
    4. Probst A
    5. Palacios JM

    (1989) Dopamine receptors in human brain: Autoradiographic distribution of D2 sites

    Neuroscience 28:275–290.

    https://doi.org/10.1016/0306-4522(89)90179-6

    • PubMed
    • Google Scholar
20. 1. Carver CS
    2. White TL

    (1994) Behavioral Inhibition, Behavioral Activation, and Affective Responses to Impending Reward and Punishment: The BIS/BAS Scales

    Journal of Personality and Social Psychology 2:319–333.

    https://doi.org/10.1037/0022-3514.67.2.319

    • Google Scholar
21. 1. Cook J. L

    (2014) Task-relevance dependent gradients in medial prefrontal and temporoparietal cortices suggest solutions to paradoxes concerning self/other control

    Neuroscience and Biobehavioral Reviews 42:298–302.

    https://doi.org/10.1016/j.neubiorev.2014.02.007

    • PubMed
    • Google Scholar
22. 1. Cook JL
    2. den Ouden HEM
    3. Heyes CM
    4. Cools R

    (2014) The social dominance paradox

    Current Biology 24:2812–2816.

    https://doi.org/10.1016/j.cub.2014.10.014

    • PubMed
    • Google Scholar
23. 1. Cook JL
    2. Swart JC
    3. Froböse MI
    4. Diaconescu AO
    5. Geurts DE
    6. den Ouden HE
    7. Cools R

    (2019) Catecholaminergic modulation of meta-learning

    eLife 8:e51439.

    https://doi.org/10.7554/eLife.51439

    • PubMed
    • Google Scholar
24. 1. Cools R
    2. Gibbs SE
    3. Miyakawa A
    4. Jagust W
    5. D’Esposito M

    (2008) Working memory capacity predicts dopamine synthesis capacity in the human striatum

    The Journal of Neuroscience 28:1208–1212.

    https://doi.org/10.1523/JNEUROSCI.4475-07.2008

    • PubMed
    • Google Scholar
25. 1. Cooper JC
    2. Dunne S
    3. Furey T
    4. O’Doherty JP

    (2012) Human dorsal striatum encodes prediction errors during observational learning of instrumental actions

    Journal of Cognitive Neuroscience 24:106–118.

    https://doi.org/10.1162/jocn_a_00114

    • PubMed
    • Google Scholar
26. 1. Crişan LG
    2. Panǎ S
    3. Vulturar R
    4. Heilman RM
    5. Szekely R
    6. Drugǎ B
    7. Dragoş N
    8. Miu AC

    (2009) Genetic contributions of the serotonin transporter to social learning of fear and economic decision making

    Social Cognitive and Affective Neuroscience 4:399–408.

    https://doi.org/10.1093/scan/nsp019

    • Google Scholar
27. 1. Davis MH

    (1983) Measuring individual differences in empathy: Evidence for a multidimensional approach

    Journal of Personality and Social Psychology 44:113–126.

    https://doi.org/10.1037/0022-3514.44.1.113

    • Google Scholar
28. 1. Daw ND
    2. Niv Y
    3. Dayan P

    (2005) Uncertainty-based competition between prefrontal and dorsolateral striatal systems for behavioral control

    Nature Neuroscience 8:1704–1711.

    https://doi.org/10.1038/nn1560

    • Google Scholar
29. 1. Daw ND
    2. O’Doherty JP
    3. Dayan P
    4. Seymour B
    5. Dolan RJ

    (2006) Cortical substrates for exploratory decisions in humans

    Nature 441:876–879.

    https://doi.org/10.1038/nature04766

    • Google Scholar
30. 1. Delgado MR
    2. Frank RH
    3. Phelps EA

    (2005) Perceptions of moral character modulate the neural systems of reward during the trust game

    Nature Neuroscience 8:1611–1618.

    https://doi.org/10.1038/nn1575

    • Google Scholar
31. 1. Diaconescu AO
    2. Mathys C
    3. Weber LAE
    4. Daunizeau J
    5. Kasper L
    6. Lomakina EI
    7. Fehr E
    8. Stephan KE

    (2014) Inferring on the intentions of others by hierarchical Bayesian learning

    PLOS Computational Biology 10:e1003810.

    https://doi.org/10.1371/journal.pcbi.1003810

    • Google Scholar
32. 1. Diaconescu AO
    2. Mathys C
    3. Weber LAE
    4. Kasper L
    5. Mauer J
    6. Stephan KE

    (2017) Hierarchical prediction errors in midbrain and septum during social learning

    Social Cognitive and Affective Neuroscience 12:618–634.

    https://doi.org/10.1093/scan/nsw171

    • PubMed
    • Google Scholar
33. 1. Diederen KMJ
    2. Ziauddeen H
    3. Vestergaard MD
    4. Spencer T
    5. Schultz W
    6. Fletcher PC

    (2017) Dopamine modulates adaptive prediction error coding in the human midbrain and striatum

    The Journal of Neuroscience 37:1708–1720.

    https://doi.org/10.1523/JNEUROSCI.1979-16.2016

    • PubMed
    • Google Scholar
34. 1. Dienes Z

    (2014) Using Bayes to get the most out of non-significant results

    Frontiers in Psychology 5:1–17.

    https://doi.org/10.3389/fpsyg.2014.00781

    • PubMed
    • Google Scholar
35. 1. Dolan RJ
    2. Dayan P

    (2013) Goals and habits in the brain

    Neuron 80:312–325.

    https://doi.org/10.1016/j.neuron.2013.09.007

    • PubMed
    • Google Scholar
36. 1. Dugast C
    2. Brun P
    3. Sotty F
    4. Renaud B
    5. Suaud-Chagny MF

    (1997) On the involvement of a tonic dopamine D2-autoinhibition in the regulation of pulse-to-pulse-evoked dopamine release in the rat striatum in vivo

    Naunyn-Schmiedeberg’s Archives of Pharmacology 355:716–719.

    https://doi.org/10.1007/pl00005004

    • PubMed
    • Google Scholar
37. 1. Ereira S
    2. Hauser TU
    3. Moran R
    4. Story GW
    5. Dolan RJ
    6. Kurth-Nelson Z

    (2020) Social training reconfigures prediction errors to shape Self-Other boundaries

    Nature Communications 11:1–14.

    https://doi.org/10.1038/s41467-020-16856-8

    • PubMed
    • Google Scholar
38. 1. Ford CP

    (2014) The role of D2-autoreceptors in regulating dopamine neuron activity and transmission

    Neuroscience 282:13–22.

    https://doi.org/10.1016/j.neuroscience.2014.01.025

    • PubMed
    • Google Scholar
39. 1. Frank MJ
    2. O’Reilly RC

    (2006) A mechanistic account of striatal dopamine function in human cognition: psychopharmacological studies with cabergoline and haloperidol

    Behavioral Neuroscience 120:497–517.

    https://doi.org/10.1037/0735-7044.120.3.497

    • PubMed
    • Google Scholar
40. 1. Frey AL
    2. McCabe C

    (2020) Effects of serotonin and dopamine depletion on neural prediction computations during social learning

    Neuropsychopharmacology 45:1431–1437.

    https://doi.org/10.1038/s41386-020-0678-z

    • PubMed
    • Google Scholar
41. 1. Garris PA
    2. Budygin EA
    3. Phillips PEM
    4. Venton BJ
    5. Robinson DL
    6. Bergstrom BP
    7. Rebec GV
    8. Wightman RM

    (2003) A role for presynaptic mechanisms in the actions of nomifensine and haloperidol

    Neuroscience 118:819–829.

    https://doi.org/10.1016/s0306-4522(03)00005-8

    • PubMed
    • Google Scholar
42. 1. Garvert MM
    2. Moutoussis M
    3. Kurth-Nelson Z
    4. Behrens TEJ
    5. Dolan RJ

    (2015) Learning-Induced plasticity in medial prefrontal cortex predicts preference malleability

    Neuron 85:418–428.

    https://doi.org/10.1016/j.neuron.2014.12.033

    • PubMed
    • Google Scholar
43. 1. Gläscher J
    2. Daw N
    3. Dayan P
    4. Doherty JPO

    (2011) States versus Rewards: Dissociable neural prediction error signals underlying model-based and model-free reinforcement learning

    Neuron 66:585–595.

    https://doi.org/10.1016/j.neuron.2010.04.016.States

    • Google Scholar
44. 1. Glimcher PW
    2. Bayer HM

    (2005) Midbrain dopamine neurons encode a quantitative reward prediction error signal

    Neuron 103:2304–2312.

    https://doi.org/10.1038/mp.2011.182.doi

    • Google Scholar
45. Book

    1. Grace AA

    (2002) Dopamine

    In: Davis KL, editors. Neuropsychopharmacology: The Fifth Generation of Progress. John Wiley & Sons, Inc. pp. 120–132.

    https://doi.org/10.1002/hup.431

    • Google Scholar
46. 1. Grace AA

    (2016) Dysregulation of the dopamine system in the pathophysiology of schizophrenia and depression

    Nature Reviews. Neuroscience 17:524–532.

    https://doi.org/10.1038/nrn.2016.57

    • PubMed
    • Google Scholar
47. 1. Haarsma J
    2. Fletcher PC
    3. Ziauddeen H
    4. Spencer TJ
    5. Diederen KMJ
    6. Murray GK

    (2018) Precision weighting of cortical unsigned prediction errors is mediated by dopamine and benefits

    Neuroscience 10:1–24.

    https://doi.org/10.1101/288936

    • Google Scholar
48. 1. Heyes CM

    (2012) What’s social about social learning

    Journal of Comparative Psychology (Washington, D.C 126:193–202.

    https://doi.org/10.1037/a0025180

    • PubMed
    • Google Scholar
49. 1. Heyes CM
    2. Pearce JM

    (2015) Not-so-social learning strategies

    Proceedings. Biological Sciences 282:1802.

    https://doi.org/10.1098/rspb.2014.1709

    • PubMed
    • Google Scholar
50. 1. Hill MR
    2. Boorman ED
    3. Fried I

    (2016) Observational learning computations in neurons of the human anterior cingulate cortex

    Nature Communications 7:12722.

    https://doi.org/10.1038/ncomms12722

    • PubMed
    • Google Scholar
51. 1. Jones JL
    2. Esber GR
    3. McDannald MA
    4. Gruber AJ
    5. Hernandez A
    6. Mirenzi A
    7. Schoenbaum G

    (2012) Orbitofrontal cortex supports behavior and learning using inferred but not cached values

    Science (New York, N.Y.) 338:953–956.

    https://doi.org/10.1126/science.1227489

    • PubMed
    • Google Scholar
52. 1. Kendal RL
    2. Boogert NJ
    3. Rendell L
    4. Laland KN
    5. Webster M
    6. Jones PL

    (2018) Social learning strategies: Bridge-building between fields

    Trends in Cognitive Sciences 22:651–665.

    https://doi.org/10.1016/j.tics.2018.04.003

    • PubMed
    • Google Scholar
53. 1. Klucharev V
    2. Hytönen K
    3. Rijpkema M
    4. Smidts A
    5. Fernández G

    (2009) Reinforcement learning signal predicts social conformity

    Neuron 61:140–151.

    https://doi.org/10.1016/j.neuron.2008.11.027

    • PubMed
    • Google Scholar
54. 1. Korn C
    2. Akam T
    3. Jensen KHR
    4. Vagnoni C
    5. Huber A
    6. Tunbridge EM
    7. Walton ME

    (2021) Distinct roles for dopamine clearance mechanisms in regulating behavioral flexibility

    Molecular Psychiatry 26:7188–7199.

    https://doi.org/10.1038/s41380-021-01194-y

    • PubMed
    • Google Scholar
55. 1. Laland KN

    (2004) Social learning strategies

    Animal Learning & Behavior 32:4–14.

    https://doi.org/10.3758/BF03196002

    • Google Scholar
56. 1. Langdon AJ
    2. Sharpe MJ
    3. Schoenbaum G
    4. Niv Y

    (2018) Model-based predictions for dopamine

    Current Opinion in Neurobiology 49:1–7.

    https://doi.org/10.1016/j.conb.2017.10.006

    • PubMed
    • Google Scholar
57. Book

    1. Lee MD
    2. Wagenmakers EJ

    (2013) Bayesian Cognitive Modeling: A Practical Course

    Cambridge University Press.

    https://doi.org/10.1017/CBO9781139087759

    • Google Scholar
58. 1. Li J
    2. Delgado MR
    3. Phelps EA

    (2011) How instructed knowledge modulates the neural systems of reward learning

    PNAS 108:55–60.

    https://doi.org/10.1073/pnas.1014938108

    • PubMed
    • Google Scholar
59. 1. Lidow MS
    2. Goldman-Rakic PS
    3. Gallager DW
    4. Rakic P

    (1991) Distribution of dopaminergic receptors in the primate cerebral cortex: Quantitative autoradiographic analysis using [3H]raclopride

    Neuroscience 40:657–671.

    https://doi.org/10.1016/0306-4522(91)90003-7

    • PubMed
    • Google Scholar
60. 1. Lindström B
    2. Haaker J
    3. Olsson A

    (2018) A common neural network differentially mediates direct and social fear learning

    NeuroImage 167:121–129.

    https://doi.org/10.1016/j.neuroimage.2017.11.039

    • PubMed
    • Google Scholar
61. 1. Lockwood PL
    2. Apps MAJ
    3. Chang SWC

    (2020) Is There a ‘Social’ Brain? Implementations and Algorithms

    Trends in Cognitive Sciences 24:802–813.

    https://doi.org/10.1016/j.tics.2020.06.011

    • PubMed
    • Google Scholar
62. Book

    1. Lovibond PF
    2. Lovibond SH

    (1995)

    Manual for the Depression Anxiety Stress Scales

    DASS.

    • Google Scholar
63. 1. Manning C
    2. Kilner J
    3. Neil L
    4. Karaminis T
    5. Pellicano E

    (2017) Children on the autism spectrum update their behaviour in response to a volatile environment

    Developmental Science 20:12435.

    https://doi.org/10.1111/desc.12435

    • Google Scholar
64. 1. Matsumoto M
    2. Weickert CS
    3. Akil M
    4. Lipska BK
    5. Hyde TM
    6. Herman MM
    7. Kleinman JE
    8. Weinberger DR

    (2003) Catechol O-methyltransferase mRNA expression in human and rat brain: Evidence for a role in cortical neuronal function

    Neuroscience 116:127–137.

    https://doi.org/10.1016/S0306-4522(02)00556-0

    • Google Scholar
65. 1. Menon M
    2. Jensen J
    3. Vitcu I
    4. Graff-Guerrero A
    5. Crawley A
    6. Smith MA
    7. Kapur S

    (2007) Temporal Difference Modeling of the Blood-Oxygen Level Dependent Response During Aversive Conditioning in Humans: Effects of Dopaminergic Modulation

    Biological Psychiatry 62:765–772.

    https://doi.org/10.1016/j.biopsych.2006.10.020

    • Google Scholar
66. 1. Moran R
    2. Dayan P
    3. Dolan RJ

    (2021) Human subjects exploit a cognitive map for credit assignment

    PNAS 118:1–12.

    https://doi.org/10.1073/pnas.2016884118

    • Google Scholar
67. 1. Morgan TJH
    2. Rendell LE
    3. Ehn M
    4. Hoppitt W
    5. Laland KN

    (2012) The evolutionary basis of human social learning

    Proceedings of the Royal Society B: Biological Sciences 279:653–662.

    https://doi.org/10.1098/rspb.2011.1172

    • Google Scholar
68. 1. Nicolle A
    2. Klein-Flügge MC
    3. Hunt LT
    4. Vlaev I
    5. Dolan RJ
    6. Behrens TEJ

    (2012) An agent independent axis for executed and modeled choice in medial prefrontal cortex

    Neuron 75:1114–1121.

    https://doi.org/10.1016/j.neuron.2012.07.023

    • PubMed
    • Google Scholar
69. 1. Niv Y
    2. Daw ND
    3. Joel D
    4. Dayan P

    (2007) Tonic dopamine: opportunity costs and the control of response vigor

    Psychopharmacology 191:507–520.

    https://doi.org/10.1007/s00213-006-0502-4

    • PubMed
    • Google Scholar
70. 1. O’Doherty JP
    2. Dayan P
    3. Friston K
    4. Critchley H
    5. Dolan RJ

    (2003) Temporal difference models and reward-related learning in the human brain

    Neuron 38:329–337.

    https://doi.org/10.1016/s0896-6273(03)00169-7

    • PubMed
    • Google Scholar
71. 1. Pehek EA

    (1999)

    Comparison of effects of haloperidol administration on amphetamine- stimulated dopamine release in the rat medial prefrontal cortex and dorsal striatum

    The Journal of Pharmacology and Experimental Therapeutics 289:14–23.

    • PubMed
    • Google Scholar
72. 1. Pessiglione M
    2. Seymour B
    3. Flandin G
    4. Dolan RJ
    5. Frith CD

    (2006) Dopamine-dependent prediction errors underpin reward-seeking behaviour in humans

    Nature 442:1042–1045.

    https://doi.org/10.1038/nature05051

    • PubMed
    • Google Scholar
73. 1. Porges SW

    (1993)

    Body Perception Questionnaire (BPQ) Manual

    Stress: The International Journal on the Biology of Stress 15:1–7.

    • Google Scholar
74. 1. Rendell L
    2. Fogarty L
    3. Hoppitt WJE
    4. Morgan TJH
    5. Webster MM
    6. Laland KN

    (2011) Cognitive culture: Theoretical and empirical insights into social learning strategies

    Trends in Cognitive Sciences 15:68–76.

    https://doi.org/10.1016/j.tics.2010.12.002

    • PubMed
    • Google Scholar
75. Book

    1. Rescorla RA
    2. Wagner AR

    (1972)

    A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement

    In: Rescorla RA, editors. Clasical Conditioning II: Current Research and Theory. Appleton Century Crofts. pp. 64–99.

    • Google Scholar
76. Book

    1. Richerson PJ
    2. Boyd R

    (2005) Not By Genes Alone: How Culture Transformed Human Evolution

    The University of Chicago Press.

    https://doi.org/10.7208/chicago/9780226712130.001.0001

    • Google Scholar
77. 1. Roberts C
    2. Sahakian BJ
    3. Robbins TW

    (2020) Psychological mechanisms and functions of 5-HT and SSRIs in potential therapeutic change: Lessons from the serotonergic modulation of action selection, learning, affect, and social cognition

    Neuroscience and Biobehavioral Reviews 119:138–167.

    https://doi.org/10.1016/j.neubiorev.2020.09.001

    • PubMed
    • Google Scholar
78. 1. Rouder JN
    2. Morey RD
    3. Speckman PL
    4. Province JM

    (2012) Default Bayes factors for ANOVA designs

    Journal of Mathematical Psychology 56:356–374.

    https://doi.org/10.1016/j.jmp.2012.08.001

    • Google Scholar
79. 1. Sadacca BF
    2. Jones JL
    3. Schoenbaum G

    (2016) Midbrain dopamine neurons compute inferred and cached value prediction errors in a common framework

    eLife 5:e13665.

    https://doi.org/10.7554/eLife.13665

    • PubMed
    • Google Scholar
80. 1. Schmitz Y
    2. Benoit-Marand M
    3. Gonon F
    4. Sulzer D

    (2003) Presynaptic regulation of dopaminergic neurotransmission

    Journal of Neurochemistry 87:273–289.

    https://doi.org/10.1046/j.1471-4159.2003.02050.x

    • PubMed
    • Google Scholar
81. 1. Schultz W
    2. Dayan P
    3. Montague PR

    (1997) A neural substrate of prediction and reward

    Science (New York, N.Y.) 275:1593–1599.

    https://doi.org/10.1126/science.275.5306.1593

    • PubMed
    • Google Scholar
82. 1. Schultz W

    (2007) Behavioral dopamine signals

    Trends in Neurosciences 30:203–210.

    https://doi.org/10.1016/j.tins.2007.03.007

    • PubMed
    • Google Scholar
83. 1. Sharpe MJ
    2. Schoenbaum G

    (2018) Evaluation of the hypothesis that phasic dopamine constitutes a cached-value signal

    Neurobiology of Learning and Memory 153:131–136.

    https://doi.org/10.1016/j.nlm.2017.12.002

    • PubMed
    • Google Scholar
84. 1. Stephan KE
    2. Penny WD
    3. Daunizeau J
    4. Moran RJ
    5. Friston KJ

    (2009) Bayesian model selection for group studies

    NeuroImage 46:1004–1017.

    https://doi.org/10.1016/j.neuroimage.2009.03.025

    • PubMed
    • Google Scholar
85. 1. Sternberg S

    (1969)

    Memory-scanning: mental processes revealed by reaction-time experiments

    American Scientist 57:421–457.

    • PubMed
    • Google Scholar
86. 1. Sulzer D
    2. Cragg SJ
    3. Rice ME

    (2016) Striatal dopamine neurotransmission: Regulation of release and uptake

    Basal Ganglia 6:123–148.

    https://doi.org/10.1016/j.baga.2016.02.001

    • PubMed
    • Google Scholar
87. Book

    1. Sutton R
    2. Barto AG

    (2018)

    Reinforcement Learning: An Introduction

    MIT press.

    • Google Scholar
88. 1. Tarantola T
    2. Kumaran D
    3. Dayan P
    4. De Martino B

    (2017) Prior preferences beneficially influence social and non-social learning

    Nature Communications 8:817.

    https://doi.org/10.1038/s41467-017-00826-8

    • PubMed
    • Google Scholar
89. 1. Templeton JJ
    2. Kamil AC
    3. Balda RP

    (1999) Sociality and social learning in two species of corvids: The pinyon jay (Gymnorhinus cyanocephalus) and the Clark’s nutcracker (Nucifraga columbiana)

    Journal of Comparative Psychology (Washington, D.C 113:450–455.

    https://doi.org/10.1037/0735-7036.113.4.450

    • PubMed
    • Google Scholar
90. 1. van den Bergh D
    2. van Doorn J
    3. Marsman M
    4. Draws T
    5. van Kesteren EJ
    6. Derks K
    7. Dablander F
    8. Gronau QF
    9. Kucharský Š
    10. Gupta A
    11. Sarafoglou A
    12. Voelkel JG
    13. Stefan A
    14. Ly A
    15. Hinne M
    16. Matzke D
    17. Wagenmakers EJ

    (2020) A tutorial on conducting and interpreting a bayesian ANOVA in JASP

    L’Année Psychologique 120:73–96.

    https://doi.org/10.3917/anpsy1.201.0073

    • Google Scholar
91. 1. Wächtler CO
    2. Chakroun K
    3. Clos M
    4. Bayer J
    5. Hennies N
    6. Beaulieu JM
    7. Sommer T

    (2020) Region-specific effects of acute haloperidol in the human midbrain, striatum and cortex

    European Neuropsychopharmacology 35:126–135.

    https://doi.org/10.1016/j.euroneuro.2020.04.008

    • PubMed
    • Google Scholar
92. 1. Watson D
    2. Clark L
    3. Tellegen A

    (1988) Development and validation of brief measures of positive and negative affect: the PANAS scales

    Journal of Personality and Social Psychology 54:1063–1070.

    https://doi.org/10.1037//0022-3514.54.6.1063

    • Google Scholar
93. 1. Yael D
    2. Zeef DH
    3. Sand D
    4. Moran A
    5. Katz DB
    6. Cohen D
    7. Temel Y
    8. Bar-Gad I

    (2013) Haloperidol-induced changes in neuronal activity in the striatum of the freely moving rat

    Frontiers in Systems Neuroscience 7:1–11.

    https://doi.org/10.3389/fnsys.2013.00110

    • PubMed
    • Google Scholar
94. 1. Zhang L
    2. Gläscher J

    (2020) A brain network supporting social influences in human decision-making

    Science Advances 6:eabb4159.

    https://doi.org/10.1126/sciadv.abb4159

    • PubMed
    • Google Scholar

Author details

1. Alicia J Rybicki

   Centre for Human Brain Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

   Contribution

   Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review and editing

   For correspondence

   axr783@bham.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6668-1214

2. Sophie L Sowden

   Centre for Human Brain Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

   Contribution

   Investigation, Project administration

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9913-0515

3. Bianca Schuster

   Centre for Human Brain Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

   Contribution

   Investigation

   Competing interests

   No competing interests declared

4. Jennifer L Cook

   Centre for Human Brain Health, School of Psychology, University of Birmingham, Birmingham, United Kingdom

   Contribution

   Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Software, Supervision, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4916-8667

• 1,382

  views

• 144

  downloads

• 13

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.