Archive for February 8th, 2010


Short review: J. D. Wallis, „Single Neuron Activity Underlying Behavior-Guiding Rules”

Short review: J. D. Wallis, „Single Neuron Activity Underlying Behavior-Guiding Rules


  1. Using abstract information to guide behavior is an important and potentially unique function of Prefrontal Cortex (PFC). This capacity might underlie two of the hallmarks of PFC, flexibility and ability to deal with novelty.
  2. The mechnisms of learning abstract information remain unclear. Reason: this is usually analyzed (the brain imaging data are recorded) when the task has already been learned.

Experimantal setting:

Three monkeys were to apply the rules „act X if the two pictures are the same, otherwise act Y” and „act X when the two pictures are different, otherwise act Y.” With the first picture there was given a cue (a drop of juice or no juice) as to which of the two rules is to be applied (juice – rule 1; no juice – rule 2). The monkeys performed above the chance (70% correct responses) applying the rules the first time they encountered a new picture. In subsequent behavioral tests the monkeys performed the task easily when new pictures were used on every trial (90% success). This means, according to the author, that the monkeys had to be solving the task by using the abstract rule.

The monkey’s brain avtivity of four regions was recorded: PFC, inferior temporal cortex (ITC), Premotor Cortex (PMC) and the striatum (STR). The imaging showed that PFC was the only area from which the data was recorded that encoded all of the task-relevant information, namely both the picture and the rules. PMC and STR encoded the rule only, while ITC encoded the sole picture. It is further assumed that the activity of PMC (which encodes the rule more strongly and earlier than PFC) is such only in the case of familiar rules; for learning new rules PFC is more critical.

Additional argument in favor of the role of PFC in learning and applying abstract rules includes:

(a) PFC damage leads to ‘disexecutive syndrome’, i.e. inability to use high-level, abstract rules to control behavior;

(b) some species are able to solve ‘matching-to-sample rule’; these include chimpanzees, rhesus monkeys, dolphins and sea lions, as well as some birds (corvids ) in which the analogon of PFC, mesopallium and nidopallium, is bigger than in other birds. Even some insects (bees) are ‘capable’ of using the ‘sameness’ rule.


It must be asked whether the behavior of the monkeys in the experiment can justly be called rule-governed. The monkeys spend some significant time learning – in a stimulus-response way – the constraints of the task. They do exhibit some ‘second level stimulus response’, i.e. thay adapt their behavior in the pictures-action task according to an additional stimulus (juice/no juice). The question is whether this amounts to applying a rule?



Short review: Michael Petrides, Selection between Competing Responses Based on Conditional Rules, [in:] S.A. Bunge, J.D. Wallis (eds.), Neuroscience of Rule-Governed Behavior, Oxford University Press 2008, pp. 3-22.

Short review:

Michael Petrides, Selection between Competing Responses Based on Conditional Rules, [in:] S.A. Bunge, J.D. Wallis (eds.), Neuroscience of Rule-Governed Behavior,Oxford University Press 2008, pp. 3-22.

1. Summary

(a) Theses

Thesis 1: Lateral frontal cortex plays a major role in both the acquisition and the execution of conditional rules that enable appropriate selection of a response from several competing alternatives.

Thesis 2: In this process, a functional interaction between the frontal cortex and the hippocampal memory system is essential.

(b) Experimental setting

Experiment 1: Patients with lateral frontal cortex excisions were given the following task. Five hand gestures were associated with five color signals. Each color signal was a stimuli to present one of the gestures. The associations were taught by a trial&error method. The patients learned the gestures easily and easily reproduced them from memory. However, they were seriously impaired when it came to connect the signal (the color) with the response (the gesture).

Experiment 2: In a similar experiment, the learning technique was changed from trial&error to demonstration (an experimenter demonstrated which of the stimuli is connected to which of the responses). The results were similar: patients with lateral frontal cortex excisions were severely impaired. It seems that the impairment is not dependent of the method of learning the „rules”.

Experiment 3: Similar results were obtained in experiments involving macaque monkeys.

Experiment 4: The activity of the lateral frontal cortex in conditional rules acquisition and execution was confirmed on brain scans (Brass et. al. 2005).

2. Critical comments

(a) The main problem with this research is whether it is justified to speak of conditional rules here. Certainly, this is an instance of a „stimulus – response” learning and executing. However, one can object that we have to do with rules here. What we call rules is probably connected to a more complex brain functions.

(b) It would be interesting to compare these results with the research on brain activity during speech production (the learning and utilization of linguistic rules), as well as on moral reasoning.