Emergence of abstract rules in the primate brain.

Various aspects of human cognition are shaped and enriched by abstract rules, which help to describe, link and classify discrete events and experiences into meaningful concepts. However, where and how these entities emerge in the primate brain and the neuronal mechanisms underlying them remain the subject of extensive research and debate. Evidence from imaging studies in humans and single-neuron recordings in monkeys suggests a pivotal role for the prefrontal cortex in the representation of abstract rules; however, behavioural studies in lesioned monkeys and data from neuropsychological examinations of patients with prefrontal damage indicate substantial functional dissociations and task dependency in the contribution of prefrontal cortical regions to rule-guided behaviour. This Review describes our current understanding of the dynamic emergence of abstract rules in primate cognition, and of the distributed neural network that supports abstract rule formation, maintenance, revision and task-dependent implementation.

A One-Shot Shift from Explore to Exploit in Monkey Prefrontal Cortex.

Much animal learning is slow, with cumulative changes in behavior driven by reward prediction errors. When the abstract structure of a problem is known, however, both animals and formal learning models can rapidly attach new items to their roles within this structure, sometimes in a single trial. Frontal cortex is likely to play a key role in this process. To examine information seeking and use in a known problem structure, we trained monkeys in an explore/exploit task, requiring the animal first to test objects for their association with reward, then, once rewarded objects were found, to reselect them on further trials for further rewards. Many cells in the frontal cortex showed an explore/exploit preference aligned with one-shot learning in the monkeys' behavior: the population switched from an explore state to an exploit state after a single trial of learning but partially maintained the explore state if an error indicated that learning had failed. Binary switch from explore to exploit was not explained by continuous changes linked to expectancy or prediction error. Explore/exploit preferences were independent for two stages of the trial: object selection and receipt of feedback. Within an established task structure, frontal activity may control the separate processes of explore and exploit, switching in one trial between the two.SIGNIFICANCE STATEMENT Much animal learning is slow, with cumulative changes in behavior driven by reward prediction errors. When the abstract structure a problem is known, however, both animals and formal learning models can rapidly attach new items to their roles within this structure. To address transitions in neural activity during one-shot learning, we trained monkeys in an explore/exploit task using familiar objects and a highly familiar task structure. When learning was rapid, many frontal neurons showed a binary, one-shot switch between explore and exploit. Within an established task structure, frontal activity may control the separate operations of exploring alternative objects to establish their current role, then exploiting this knowledge for further reward.

Behavioral flexibility is associated with changes in structure and function distributed across a frontal cortical network in macaques.

One of the most influential accounts of central orbitofrontal cortex-that it mediates behavioral flexibility-has been challenged by the finding that discrimination reversal in macaques, the classic test of behavioral flexibility, is unaffected when lesions are made by excitotoxin injection rather than aspiration. This suggests that the critical brain circuit mediating behavioral flexibility in reversal tasks lies beyond the central orbitofrontal cortex. To determine its identity, a group of nine macaques were taught discrimination reversal learning tasks, and its impact on gray matter was measured. Magnetic resonance imaging scans were taken before and after learning and compared with scans from two control groups, each comprising 10 animals. One control group learned discrimination tasks that were similar but lacked any reversal component, and the other control group engaged in no learning. Gray matter changes were prominent in posterior orbitofrontal cortex/anterior insula but were also found in three other frontal cortical regions: lateral orbitofrontal cortex (orbital part of area 12 [12o]), cingulate cortex, and lateral prefrontal cortex. In a second analysis, neural activity in posterior orbitofrontal cortex/anterior insula was measured at rest, and its pattern of coupling with the other frontal cortical regions was assessed. Activity coupling increased significantly in the reversal learning group in comparison with controls. In a final set of experiments, we used similar structural imaging procedures and analyses to demonstrate that aspiration lesion of central orbitofrontal cortex, of the type known to affect discrimination learning, affected structure and activity in the same frontal cortical circuit. The results identify a distributed frontal cortical circuit associated with behavioral flexibility.

Prefrontal and Medial Temporal Lobe Cortical Contributions to Visual Short-Term Memory.

A number of recent studies have indicated that the medial temporal lobe (MTL) plays a critical role in working memory (WM) and perception, but these results have been highly controversial given the traditional association of MTL with long-term memory. We review the research and highlight important factors that need to be considered in determining the role of MTL in WM including set-size of used stimuli and feature complexity and/or feature conjunctions/bindings embedded in those stimuli. These factors relate to hierarchical and, accordingly, domain-specific theories of functional organization within the temporal lobe. In addition, one must consider process-specific theories too, because two key processes commonly understood to contribute recognition memory, namely, recollection and familiarity, also have robust support from neurophysiological and neuroimaging research as to their functional dissociations within MTL. PFC has long been heavily implicated in WM; however, relatively less is known about how the PFC contributes to recollection and familiarity, although dynamic prefrontal coding models in WM may help to explain their neural mechanisms. The MTL and PFC are heavily interconnected and do not operate independently in underlying WM. We propose that investigation of the interactions between these two regions in WM, particularly their coordinated neural activities, and the modeling of such interactions, will be crucial for the advancing understanding of the neural mechanisms of WM.

Integrated neural dynamics for behavioural decisions and attentional competition in the prefrontal cortex.

In the behaving monkey, complex neural dynamics in the prefrontal cortex contribute to context-dependent decisions and attentional competition. We used demixed principal component analysis to track prefrontal activity dynamics in a cued target detection task. In this task, the animal combined identity of a visual object with a prior instruction cue to determine a target/nontarget decision. From population activity, we extracted principal components for each task feature and examined their time course and sensitivity to stimulus and task variations. For displays containing a single choice object in left or right hemifield, object identity, cue identity and decision were all encoded in population activity, with different dynamics and lateralisation. Object information peaked at 100-200 ms from display onset and was largely confined to the contralateral hemisphere. Cue information was weaker and present even prior to display onset. Integrating information from cue and object, decision information arose more slowly and was bilateral. Individual neurons contributed independently to coding of the three task features. The analysis was then extended to displays with a target in one hemifield and a competing distractor in the other. In this case, the data suggest that each hemisphere initially encoded the identity of the contralateral object. The distractor representation was then rapidly suppressed, with the final target decision again encoded bilaterally. The results show how information is coded along task-related dimensions while competition is resolved and suggest how information flows within and across frontal lobes to implement a learned behavioural decision.

Managing competing goals - a key role for the frontopolar cortex.

Humans are set apart from other animals by many elements of advanced cognition and behaviour, including language, judgement and reasoning. What is special about the human brain that gives rise to these abilities? Could the foremost part of the prefrontal cortex (the frontopolar cortex), which has become considerably enlarged in humans during evolution compared with other animals, be important in this regard, especially as, in primates, it contains a unique cytoarchitectural field, area 10? The first studies of the function of the frontopolar cortex in monkeys have now provided critical new insights about its precise role in monitoring the significance of current and alternative goals. In human evolution, the frontopolar cortex may have acquired a further role in enabling the monitoring of the significance of multiple goals in parallel, as well as switching between them. Here, we argue that many other forms of uniquely human behaviour may benefit from this cognitive ability mediated by the frontopolar cortex.

Low-beta repetitive transcranial magnetic stimulation to human dorsolateral prefrontal cortex during object recognition memory sample presentation, at a task-related frequency observed in local field potentials in homologous macaque cortex, impairs subsequent recollection but not familiarity.

According to dual-process signal-detection (DPSD) theories, short- and long-term recognition memory draws upon both familiarity and recollection. It remains unclear how primate prefrontal cortex (PFC) contributes to these processes, but frequency-specific neuronal activities are considered to play a key role. In Experiment 1, nonhuman primate (NHP) local field potential (LFP) electrophysiological recordings in macaque left dorsolateral PFC (dlPFC) revealed performance-related differences in a low-beta frequency range during the sample presentation phase of a visual object recognition memory task. Experiment 2 employed a similar task in humans and targeted left dlPFC (and vertex as a control) with repetitive transcranial magnetic stimulation (rTMS) at 12.5 Hz during occasional sample presentations. This low-beta frequency rTMS to dlPFC decreased DPSD derived indices of recollection, but not familiarity, in subsequent memory tests of the targeted samples after short delays. The same number of rTMS pulses over the same total duration albeit at a random frequency had no effect on either recollection or familiarity. Neither stimulation protocols had any causal effect upon behaviour when targeted to the control site (vertex). In this study, our hypotheses for our human TMS study were derived from our observations in NHPs; this approach might inspire further translational research through investigation of homologous brain regions and tasks across species using similar neuroscientific methodologies to advance the neural mechanism of recognition memory in primates.

The Role of Primate Prefrontal Cortex in Bias and Shift Between Visual Dimensions.

Imaging and neural activity recording studies have shown activation in the primate prefrontal cortex when shifting attention between visual dimensions is necessary to achieve goals. A fundamental unanswered question is whether representations of these dimensions emerge from top-down attentional processes mediated by prefrontal regions or from bottom-up processes within visual cortical regions. We hypothesized a causative link between prefrontal cortical regions and dimension-based behavior. In large cohorts of humans and macaque monkeys, performing the same attention shifting task, we found that both species successfully shifted between visual dimensions, but both species also showed a significant behavioral advantage/bias to a particular dimension; however, these biases were in opposite directions in humans (bias to color) versus monkeys (bias to shape). Monkeys' bias remained after selective bilateral lesions within the anterior cingulate cortex (ACC), frontopolar cortex, dorsolateral prefrontal cortex (DLPFC), orbitofrontal cortex (OFC), or superior, lateral prefrontal cortex. However, lesions within certain regions (ACC, DLPFC, or OFC) impaired monkeys' ability to shift between these dimensions. We conclude that goal-directed processing of a particular dimension for the executive control of behavior depends on the integrity of prefrontal cortex; however, representation of competing dimensions and bias toward them does not depend on top-down prefrontal-mediated processes.

Focused Representation of Successive Task Episodes in Frontal and Parietal Cortex.

Complex cognition is dynamic, with each stage of a task requiring new cognitive processes appropriately linked to stimulus or other content. To investigate control over successive task stages, we recorded neural activity in lateral frontal and parietal cortex as monkeys carried out a complex object selection task, with each trial separated into phases of visual selection and learning from feedback. To study capacity limitation, complexity was manipulated by varying the number of object targets to be learned in each problem. Different task phases were associated with quasi-independent patterns of activity and information coding, with no suggestion of sustained activity linked to a current target. Object and location coding were largely parallel in frontal and inferior parietal cortex, though frontal cortex showed somewhat stronger object representation at feedback, and more sustained location coding at choice. At both feedback and choice, coding precision diminished as task complexity increased, matching a decline in performance. We suggest that, across successive task steps, there is radical but capacity-limited reorganization of frontoparietal activity, selecting different cognitive operations linked to their current targets.

Similar time course of fast familiarity and slow recollection processes for recognition memory in humans and macaques.

According to dual-process theory, recognition memory performance draws upon two processes, familiarity and recollection. The relative contribution to recognition memory are commonly distinguished in humans by analyzing receiver-operating-characteristics (ROC) curves; analogous methods are more complex and very rare in animals but fast familiarity and slow recollective-like processes (FF/SR) have been detected in nonhuman primates (NHPs) based on analyzing recognition error response time profiles. The relative utility of these methods to investigate familiarity and recollection/recollection-like processes across species is uncertain; indeed, even how comparable the FF/SR measures are across humans and NHPs remains unclear. Therefore, in this study a broadly similar recognition memory task was exploited in both humans and a NHP to investigate the time course of the two recognition processes. We first show that the FF/SR dissociation exists in this task in human participants and then we demonstrate a similar profile in the NHP which suggests that FF/SR processes are comparable across species. We then verified, using ROC-derived indices for each time-bin in the FF/SR profile, that the ROC and FF/SR measures are related. Hence, we argue that the FF/SR approach, procedurally easier in nonhuman animals, can be used as a decent proxy to investigate these two recognition processes in future animal studies, important given that scant data exists as to the neural basis underlying recollection yet many of the most informative techniques primarily exist in animal models.

Mnemonic Introspection in Macaques Is Dependent on Superior Dorsolateral Prefrontal Cortex But Not Orbitofrontal Cortex.

The human PFC has been associated more with meta-perceptual as opposed to meta-memory decisions from correlational neuroimaging investigations. Recently, metacognitive abilities have also been shown to be causally dependent upon anterior and dorsal PFC in nonhuman primate lesion studies. Two studies, using postdecision wagering paradigms and reversible inactivation, challenged this meta-perceptual versus meta-memory notion and showed that dorsal and anterior prefrontal areas are associated with metamemory for experienced objects and awareness of ignorance, respectively. Causal investigations are important but scarce; nothing is known, for example, about the causal contributions of prefrontal subregions to spatial metamemory. Here, we investigated the effects of dorsal versus ventral PFC lesions on two-alternative forced-choice spatial discrimination tasks in male macaque monkeys. Importantly, we were rigorous in approach and applied three independent but complementary indices used to quantify individual animals' metacognitive ability ("Type II sensitivity") by two variants of meta-d'/d' and phi coefficient (φ). Our results were consistent across indices: while neither lesions to superior dorsolateral PFC nor orbitofrontal cortex impaired spatial recognition performance, only monkeys with superior dorsolateral PFC lesions were impaired in meta-accuracy. Together with the observation that the same orbitofrontal cortex lesioned monkeys were impaired in updating rule value in a Wisconsin Card Sorting Test analog, we therefore document a functional double-dissociation between these two PFC regions. Our study presents important causal evidence that other dimensions, namely, domain-specific processing (e.g., spatial vs nonspatial metamemory), also need considerations in understanding the functional specialization in the neural underpinnings of introspection.SIGNIFICANCE STATEMENT This study demonstrates macaque monkeys' metacognitive capability of introspecting its own memory success is causally dependent on intact superior dorsolateral prefrontal cortices but not the orbitofrontal cortices. Combining neurosurgical techniques on monkeys and state-of-the-art measures of metacognition, we affirm a critical role of the PFC in supporting spatial meta-recognition memory and delineate functional specificity within PFC for distinct elements of metacognition.

Distinct Roles for the Anterior Cingulate and Dorsolateral Prefrontal Cortices During Conflict Between Abstract Rules.

Distinct patterns of activity within the anterior cingulate cortex (ACC) and dorsolateral prefrontal cortex (dlPFC) reported in neuroimaging studies during tasks involving conflict between competing responses have often been cited as evidence for their key contributions to conflict-monitoring and behavioral adaptation, respectively. However, supporting evidence from neuropsychological patients has been scarce and contradictory. We administered a well-studied analog of the Wisconsin Card Sorting Test, designed to elicit conflict between 2 abstract rules, to a cohort of 6 patients with damage to ACC or dlPFC. Patients who had sustained more significant damage to the ACC were not impaired either on a measure of "conflict cost" nor on measures of "conflict-induced behavioral adaptation." In contrast, damage to dlPFC did not affect the conflict cost measure but abolished the patients' ability to adapt their behavior following exposure to conflict, compared with controls. This pattern of results complements the findings from nonhuman primates with more circumscribed lesions to ACC or dlPFC on the same task and provides converging evidence that ACC is not necessary for performance when conflict is elicited between 2 abstract rules, whereas dlPFC plays a fundamental role in behavioral adaptation.

Behavioral consequences of selective damage to frontal pole and posterior cingulate cortices.

Frontal pole cortex (FPC) and posterior cingulate cortex (PCC) have close neuroanatomical connections, and imaging studies have shown coactivation or codeactivation of these brain regions during performance of certain tasks. However, they are among the least well-understood regions of the primate brain. One reason for this is that the consequences of selective bilateral lesions to either structure have not previously been studied in any primate species. We studied the effects of circumscribed bilateral lesions to FPC or PCC on monkeys' ability to perform an analog of Wisconsin Card Sorting Test (WCST) and related tasks. In contrast to lesions in other prefrontal regions, neither posttraining FPC nor PCC lesions impaired animals' abilities to follow the rule switches that frequently occurred within the WCST task. However, FPC lesions were not without effect, because they augmented the ability of animals to adjust cognitive control after experiencing high levels of conflict (whereas PCC lesions did not have any effect). In addition, FPC-lesioned monkeys were more successful than controls or PCC-lesioned animals at remembering the relevant rule across experimentally imposed distractions involving either an intervening secondary task or a surprising delivery of free reward. Although prefrontal cortex posterior to FPC is specialized for mediating efficient goal-directed behavior to maximally exploit reward opportunities from ongoing tasks, our data led us to suggest that FPC is, instead, specialized for disengaging executive control from the current task and redistributing it to novel sources of reward to explore new opportunities/goals.

Essential functions of primate frontopolar cortex in cognition.

Brodmann's area 10 is one of the largest cytoarchitecturally defined regions in the human cerebral cortex, occupying the most anterior part of the prefrontal cortex [frontopolar cortex (FPC)], and is believed to sit atop a prefrontal hierarchy. The crucial contributions that the FPC makes to cognition are unknown. Rodents do not possess such [corrected] a FPC, but primates do, and we report here the behavioral effects of circumscribed FPC lesions in nonhuman primates. FPC lesions selectively impaired rapid one-trial learning about unfamiliar objects and unfamiliar objects-in-scenes, and also impaired rapid learning about novel abstract rules. Object recognition memory, shifting between established abstract behavioral rules, and the simultaneous application of two distinct rules were unaffected by the FPC lesion. The distinctive pattern of impaired and spared performance across these seven behavioral tasks reveals that the FPC mediates exploration and rapid learning about the relative value of novel behavioral options, and shows that the crucial contributions made by the FPC to cognition differ markedly from the contributions of other primate prefrontal regions.

A Putative Multiple-Demand System in the Macaque Brain.

UNLABELLED: In humans, cognitively demanding tasks of many types recruit common frontoparietal brain areas. Pervasive activation of this "multiple-demand" (MD) network suggests a core function in supporting goal-oriented behavior. A similar network might therefore be predicted in nonhuman primates that readily perform similar tasks after training. However, an MD network in nonhuman primates has not been described. Single-cell recordings from macaque frontal and parietal cortex show some similar properties to human MD fMRI responses (e.g., adaptive coding of task-relevant information). Invasive recordings, however, come from limited prespecified locations, so they do not delineate a macaque homolog of the MD system and their positioning could benefit from knowledge of where MD foci lie. Challenges of scanning behaving animals mean that few macaque fMRI studies specifically contrast levels of cognitive demand, so we sought to identify a macaque counterpart to the human MD system using fMRI connectivity in 35 rhesus macaques. Putative macaque MD regions, mapped from frontoparietal MD regions defined in humans, were found to be functionally connected under anesthesia. To further refine these regions, an iterative process was used to maximize their connectivity cross-validated across animals. Finally, whole-brain connectivity analyses identified voxels that were robustly connected to MD regions, revealing seven clusters across frontoparietal and insular cortex comparable to human MD regions and one unexpected cluster in the lateral fissure. The proposed macaque MD regions can be used to guide future electrophysiological investigation of MD neural coding and in task-based fMRI to test predictions of similar functional properties to human MD cortex. SIGNIFICANCE STATEMENT: In humans, a frontoparietal "multiple-demand" (MD) brain network is recruited during a wide range of cognitively demanding tasks. Because this suggests a fundamental function, one might expect a similar network to exist in nonhuman primates, but this remains controversial. Here, we sought to identify a macaque counterpart to the human MD system using fMRI connectivity. Putative macaque MD regions were functionally connected under anesthesia and were further refined by iterative optimization. The result is a network including lateral frontal, dorsomedial frontal, and insular and inferior parietal regions closely similar to the human counterpart. The proposed macaque MD regions can be useful in guiding electrophysiological recordings or in task-based fMRI to test predictions of similar functional properties to human MD cortex.

Retrosplenial Cortical Contributions to Anterograde and Retrograde Memory in the Monkey.

Primate retrosplenial cortex (RSC) is important for memory but patient neuropathologies are diffuse so its key contributions to memory remain elusive. This study provides the first causal evidence that RSC in macaque monkeys is crucial for postoperative retention of preoperatively and postoperatively acquired memories. Preoperatively, monkeys learned 300 object-in-place scene discriminations across sessions. After RSC removal, one-trial postoperative retention tests revealed significant retrograde memory loss for these 300 discriminations relative to unoperated control monkeys. Less robust evidence was found for a deficit in anterograde memory (new postoperative learning) after RSC lesions as new learning to criterion measures failed to reveal any significant learning impairment. However, after achieving ≥90% learning criterion for the postoperatively presented novel 100 object-in-place scene discriminations, short-term retention (i.e., measured after 24 h delay) of this well-learnt set was impaired in the RSC monkeys relative to controls. A further experiment assessed rapid "within" session acquisition of novel object-in-place scene discriminations, again confirming that new learning per se was unimpaired by bilateral RSC removal. Primate RSC contributes critically to memory by supporting normal retention of information, even when this information does not involve an autobiographical component.

Adaptability to changes in temporal structure is fornix-dependent.

Recognition memory deficits, even after short delays, are sometimes observed following hippocampal damage. One hypothesis links the hippocampus with processes in updating contextual memory representation. Here, we used fornix transection, which partially disconnects the hippocampal system, and compares the performance of fornix-transected monkeys with normal monkeys on two versions of a delayed-matching-to-position task with short delays. Spatial recognition memory was affected by fornix transection only when the temporal structure of the task changed across trials, while differences in motor control, motivation, perception, or short-term memory were not critical. We attributed the deficit to a compromised ability in tracking changes in task temporal structure.

The essential role of primate orbitofrontal cortex in conflict-induced executive control adjustment.

Conflict in information processing evokes trial-by-trial behavioral modulations. Influential models suggest that adaptive tuning of executive control, mediated by mid-dorsal lateral prefrontal cortex (mdlPFC) and anterior cingulate cortex (ACC), underlies these modulations. However, mdlPFC and ACC are parts of distributed brain networks including orbitofrontal cortex (OFC), posterior cingulate cortex (PCC), and superior-dorsal lateral prefrontal cortex (sdlPFC). Contributions of these latter areas in adaptive tuning of executive control are unknown. We trained monkeys to perform a matching task in which they had to resolve the conflict between two behavior-guiding rules. Here, we report that bilateral lesions in OFC, but not in PCC or sdlPFC, impaired selection between these competing rules. In addition, the behavioral adaptation that is normally induced by experiencing conflict disappeared in OFC-lesioned, but remained normal in PCC-lesioned or sdlPFC-lesioned monkeys. Exploring underlying neuronal processes, we found that the activity of neurons in OFC represented the conflict between behavioral options independent from the other aspects of the task. Responses of OFC neurons to rewards also conveyed information of the conflict level that the monkey had experienced along the course to obtain the reward. Our findings indicate dissociable functions for five closely interconnected cortical areas suggesting that OFC and mdlPFC, but not PCC or sdlPFC or ACC, play indispensable roles in conflict-dependent executive control of on-going behavior. Both mdlPFC and OFC support detection of conflict and its integration with the task goal, but in contrast to mdlPFC, OFC does not retain the necessary information for conflict-induced modulation of future decisions.

Cognitive control functions of anterior cingulate cortex in macaque monkeys performing a Wisconsin Card Sorting Test analog.

Monkeys were trained to select one of three targets by matching in color or matching in shape to a sample. Because the matching rule frequently changed and there were no cues for the currently relevant rule, monkeys had to maintain the relevant rule in working memory to select the correct target. We found that monkeys' error commission was not limited to the period after the rule change and occasionally occurred even after several consecutive correct trials, indicating that the task was cognitively demanding. In trials immediately after such error trials, monkeys' speed of selecting targets was slower. Additionally, in trials following consecutive correct trials, the monkeys' target selections for erroneous responses were slower than those for correct responses. We further found evidence for the involvement of the cortex in the anterior cingulate sulcus (ACCs) in these error-related behavioral modulations. First, ACCs cell activity differed between after-error and after-correct trials. In another group of ACCs cells, the activity differed depending on whether the monkeys were making a correct or erroneous decision in target selection. Second, bilateral ACCs lesions significantly abolished the response slowing both in after-error trials and in error trials. The error likelihood in after-error trials could be inferred by the error feedback in the previous trial, whereas the likelihood of erroneous responses after consecutive correct trials could be monitored only internally. These results suggest that ACCs represent both context-dependent and internally detected error likelihoods and promote modes of response selections in situations that involve these two types of error likelihood.

Spatial and temporal distribution of visual information coding in lateral prefrontal cortex.

Prefrontal neurons code many kinds of behaviourally relevant visual information. In behaving monkeys, we used a cued target detection task to address coding of objects, behavioural categories and spatial locations, examining the temporal evolution of neural activity across dorsal and ventral regions of the lateral prefrontal cortex (encompassing parts of areas 9, 46, 45A and 8A), and across the two cerebral hemispheres. Within each hemisphere there was little evidence for regional specialisation, with neurons in dorsal and ventral regions showing closely similar patterns of selectivity for objects, categories and locations. For a stimulus in either visual field, however, there was a strong and temporally specific difference in response in the two cerebral hemispheres. In the first part of the visual response (50-250 ms from stimulus onset), processing in each hemisphere was largely restricted to contralateral stimuli, with strong responses to such stimuli, and selectivity for both object and category. Later (300-500 ms), responses to ipsilateral stimuli also appeared, many cells now responding more strongly to ipsilateral than to contralateral stimuli, and many showing selectivity for category. Activity on error trials showed that late activity in both hemispheres reflected the animal's final decision. As information is processed towards a behavioural decision, its encoding spreads to encompass large, bilateral regions of prefrontal cortex.