Wednesday, May 31, 2017

summary of an article on the sense of agency

Here I share the summary of the recent review article on the sense of agency, authored by Patrick Haggard.
  • Haggard, P. (2017). Sense of agency in the human brain. Nature Reviews Neuroscience. doi: 10.1038/nm.2017.14

There are 8 sections in this article excluding introduction and conclusion.

1) Defining the sense of agency
In the first section, the author confirms the definition of the sense of agency. He writes, “the sense of agency is the feeling of making something happen.” And he also adds, “it is the experience of controlling one’s own motor acts and, through them, the course of external events.” According to the author, the sense of agency requires an action, and the subsequent outcome. Through overall article, he emphasizes that the association between a voluntary action and an outcome underlies the sense of agency.

2) Measuring the sense of agency
In the second section, the author explains that there are two different methods used to measure the sense of agency.
  • One is the explicit method. In an experimental setting, the participant is shown a certain image of hand movement which could be one’s own movement or another person’s movement. After watching it, the participant is asked to attribute it to the self or the other.
  • The other one is the implicit method, so-called “intentional binding.” The participant is asked to put the key voluntarily or involuntarily by TMS and listen to the sound after 250 mili-seconds. As a result, the participants report that the timespan between the putting action and the sound feels shorter, when they perform the action in a voluntary manner. So, the extent of binding can be considered as a quantitative marker for the sense of agency.
It is important to note that the author remarks that the sense of agency measured by an explicit manner and that measured by an implicit manner are different. (Though the former is social, the latter is intra-personal). And the relationship between them are to be explored further in the future investigations.

3) Cognitive processes that drive agency
In the third section, the author explains the cognitive processes that drive the sense of agency. In the first phase of agency, there are two important aspects; volition and action selection. As we know through intentional binding experiment, the voluntary action is needed to bring forth the agency. The second aspect is the selection of action. The author refers to the sensory attenuation in this regard. When the action is selected by oneself, the prediction of the outcome becomes redundant, and that makes the sensory feedback weaker (we can recall that the self-tickling is not ticklish).
4) The comparator model of agency
Then the author explains the comparator model. Traditionally, the sense of agency has been explained on the basis of this model.
According to the figure, when one plans to perform a certain action, the motor command makes a prediction based on the efference copy of it. And on the other hand, there is a sensory feedback route based on the perception of the outcome of the performed action. And we know the match or mismatch between the prediction and the sensory feedback. Basically, the sense of agency is considered to be generated through the match between them. However, this model includes one difficult issue, according to the author. If there is no prediction error, the perception of the outcome becomes weak, as it is known as the sensory attenuation. This is counter-intuitive, because the same situation means that one finds precisely the intended outcome in the situation. Thus, the author claims that the comparator model cannot be the only source for the sense of agency.

5) Prospective versus retrospective agency
This contradiction included in the comparator model leads the author to the next discussion. He emphasizes that not only the retrospective signals used in the comparator model, but also the prospective signals generated through volition and the selection of action are also important. There are empirical research that shows the importance of prospective signals. For example, in the experimental setting of the intentional binding, if the probability of the key sounds is controlled such as 75% or 50%, the binding becomes stronger when the probability is higher. This suggests that the participants experience the stronger sense of agency, when they anticipate the result in more intensive manner.

6) Brain mechanisms underlying agency
Traditionally, the parietal lobe, especially the angular gyrus has been said to play an important role in detecting non-agency condition, that is, the mismatched sensory feedback from the environment. Regarding the prospective signals, the frontal lobe and the pre-frontal cortex are the important areas. The author especially names pre-supplementary motor area, mentioning that the intentional bindings are reduced when the electric stimuli are added in this part. So, the pre-supplementary motor area seems to play an important role in generating the voluntary action or its planning. Within the frontal lobe, DLPFC is also considered to correspond with the selection of action. What the author emphasizes is the connectivity of the frontal lobe and the parietal lobe. This connectivity seems to correspond to the fact that the sense of agency has the prospective aspect and the retrospective aspect. There is no evidence so far that suggests a single unit in the bran that represents the comparator.

7) Pathological sense of agency
In this section, the author refers to the pathology of agency, especially focusing on the case of schizophrenia. What is well studied among the symptoms of schizophrenia in terms of agency is the "delusions of control". The patients subjectively feel that their actions or thoughts are inserted by an external force. And this is explained based on the comparator model, that the patients start performing in a voluntary manner but the prediction route is not functioning properly, so that the sensory feedback is not attributed to the self.

8) Society, agency and responsibility
In the last part, the author discusses the relationship between the sense of agency and the legal responsibility. Generally speaking, there must be an independent agent who is responsible for his/her action, in order that the society can punish or reward him/her. And that social agent must have a sense of agency to be fully responsible in these scenes. The author introduces one interesting experiment regarding this point; the participant is asked to give a painful shock to the partner participant, but by obeying the instruction of the experimenter. And comparing the condition with or without the coercion of the experimenter, the participant’s subjective time perception becomes different. With coercion, the participant’s time perception becomes longer, suggesting that the intentional binding reduces, and the sense of agency becomes weaker. Thus, this experiment suggests that the sense of agency actually becomes weaker, when one is forced to perform an action by someone else. This finding will stimulate a discussion in the field of ethics and law.